diff --git "a/train_job_output.txt" "b/train_job_output.txt"
--- "a/train_job_output.txt"
+++ "b/train_job_output.txt"
@@ -1,4 +1,4 @@
-slurm submission log: 2024-05-08 15:15:18.502971
+slurm submission log: 2024-05-09 15:03:33.199064
 created following sbatch script: 
 
 ###############################
@@ -7,24 +7,23 @@ created following sbatch script:
 
 #SBATCH --account=nlp
 #SBATCH --cpus-per-task=16
-#SBATCH --dependency=afterok:7590690
 #SBATCH --gres=gpu:2
-#SBATCH --job-name=tthrush-job-2892522
+#SBATCH --job-name=tthrush-job-3941143
 #SBATCH --mem=400G
 #SBATCH --nodelist=sphinx2
 #SBATCH --open-mode=append
-#SBATCH --output=/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms/pythia-70m_sciq/train_job_output.txt
+#SBATCH --output=/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms_2/pythia-70m_sciq/train_job_output.txt
 #SBATCH --partition=sphinx
 #SBATCH --time=14-0
 
 # activate your desired anaconda environment
-. /nlp/scr/tthrush/miniconda3/etc/profile.d/conda.sh ; conda activate pretraining-coreset-selection
+. /nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/etc/profile.d/conda.sh ; conda activate pretraining-coreset-selection
 
 # cd to working directory
 cd .
 
 # launch commands
-srun --unbuffered run_as_child_processes 'torchrun --master_port 29505 --nproc_per_node=2 train_llm.py --dataset_id /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/training_data/sciq --output_dir /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms/pythia-70m_sciq --output_hub_id pythia-70m_sciq --model_id /pythia-70m --num_train_epochs 1 --learning_rate 1e-3 --warmup_ratio=0.1 --gradient_accumulation_steps 2'
+srun --unbuffered run_as_child_processes 'torchrun --master_port 29502 --nproc_per_node=2 train_llm.py --dataset_id /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/train_data_2/sciq --output_dir /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms_2/pythia-70m_sciq --output_hub_id pythia-70m_sciq --model_id EleutherAI/pythia-70m --num_train_epochs 14 --learning_rate 1e-3 --warmup_ratio=0.1 --gradient_accumulation_steps 2'
 
 ###############################
 
@@ -34,611 +33,83 @@ submission to slurm complete!
 ###############################
 slurm submission output
 
-Submitted batch job 7590693
+Submitted batch job 7592321
 
 
 
 ###############################
 
-###############################
-start time: 2024-05-08 16:40:36.203264
-machine: sphinx2
-conda env: pretraining-coreset-selection
-###############################
-running following processes
-
-	torchrun --master_port 29505 --nproc_per_node=2 train_llm.py --dataset_id /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/training_data/sciq --output_dir /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms/pythia-70m_sciq --output_hub_id pythia-70m_sciq --model_id /pythia-70m --num_train_epochs 1 --learning_rate 1e-3 --warmup_ratio=0.1 --gradient_accumulation_steps 2
+/var/lib/slurm/slurmd/job7592321/slurm_script: line 15: /nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/etc/profile.d/conda.sh: No such file or directory
 
+CommandNotFoundError: Your shell has not been properly configured to use 'conda activate'.
+To initialize your shell, run
 
-###############################
-command outputs: 
-
+    $ conda init <SHELL_NAME>
 
-[2024-05-08 16:40:38,281] torch.distributed.run: [WARNING] 
-[2024-05-08 16:40:38,281] torch.distributed.run: [WARNING] *****************************************
-[2024-05-08 16:40:38,281] torch.distributed.run: [WARNING] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed. 
-[2024-05-08 16:40:38,281] torch.distributed.run: [WARNING] *****************************************
-05/08/2024 16:40:44 - INFO - __main__ - Script parameters ScriptArguments(dataset_id='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/training_data/sciq', output_dir='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms/pythia-70m_sciq', output_hub_id='pythia-70m_sciq', hf_hub_token=True, model_id='/pythia-70m', per_device_train_batch_size=256, num_train_epochs=1, learning_rate=0.001, gradient_accumulation_steps=2, from_scratch=True, warmup_ratio=0.1, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, weight_decay=0.01, lr_scheduler_type='cosine', local_rank=0, resume_from_checkpoint=False, deepspeed=None, peft=False)
-05/08/2024 16:40:44 - INFO - __main__ - Script parameters ScriptArguments(dataset_id='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/training_data/sciq', output_dir='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms/pythia-70m_sciq', output_hub_id='pythia-70m_sciq', hf_hub_token=True, model_id='/pythia-70m', per_device_train_batch_size=256, num_train_epochs=1, learning_rate=0.001, gradient_accumulation_steps=2, from_scratch=True, warmup_ratio=0.1, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, weight_decay=0.01, lr_scheduler_type='cosine', local_rank=0, resume_from_checkpoint=False, deepspeed=None, peft=False)
-Traceback (most recent call last):
-  File "/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/transformers/utils/hub.py", line 398, in cached_file
-    resolved_file = hf_hub_download(
-  File "/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/huggingface_hub/utils/_validators.py", line 110, in _inner_fn
-    validate_repo_id(arg_value)
-  File "/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/huggingface_hub/utils/_validators.py", line 164, in validate_repo_id
-    raise HFValidationError(
-huggingface_hub.utils._validators.HFValidationError: Repo id must use alphanumeric chars or '-', '_', '.', '--' and '..' are forbidden, '-' and '.' cannot start or end the name, max length is 96: '/pythia-70m'.
+Currently supported shells are:
+  - bash
+  - fish
+  - tcsh
+  - xonsh
+  - zsh
+  - powershell
 
-The above exception was the direct cause of the following exception:
+See 'conda init --help' for more information and options.
 
-Traceback (most recent call last):
-  File "/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/train_llm.py", line 202, in <module>
-    train_model()
-  File "/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/train_llm.py", line 166, in train_model
-    config = AutoConfig.from_pretrained(script_args.model_id)
-  File "/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/transformers/models/auto/configuration_auto.py", line 1138, in from_pretrained
-    config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs)
-  File "/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/transformers/configuration_utils.py", line 631, in get_config_dict
-    config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs)
-  File "/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/transformers/configuration_utils.py", line 686, in _get_config_dict
-    resolved_config_file = cached_file(
-  File "/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/transformers/utils/hub.py", line 462, in cached_file
-    raise EnvironmentError(
-OSError: Incorrect path_or_model_id: '/pythia-70m'. Please provide either the path to a local folder or the repo_id of a model on the Hub.
-Traceback (most recent call last):
-  File "/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/transformers/utils/hub.py", line 398, in cached_file
-    resolved_file = hf_hub_download(
-  File "/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/huggingface_hub/utils/_validators.py", line 110, in _inner_fn
-    validate_repo_id(arg_value)
-  File "/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/huggingface_hub/utils/_validators.py", line 164, in validate_repo_id
-    raise HFValidationError(
-huggingface_hub.utils._validators.HFValidationError: Repo id must use alphanumeric chars or '-', '_', '.', '--' and '..' are forbidden, '-' and '.' cannot start or end the name, max length is 96: '/pythia-70m'.
+IMPORTANT: You may need to close and restart your shell after running 'conda init'.
 
-The above exception was the direct cause of the following exception:
-
-Traceback (most recent call last):
-  File "/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/train_llm.py", line 202, in <module>
-    train_model()
-  File "/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/train_llm.py", line 166, in train_model
-    config = AutoConfig.from_pretrained(script_args.model_id)
-  File "/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/transformers/models/auto/configuration_auto.py", line 1138, in from_pretrained
-    config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs)
-  File "/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/transformers/configuration_utils.py", line 631, in get_config_dict
-    config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs)
-  File "/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/transformers/configuration_utils.py", line 686, in _get_config_dict
-    resolved_config_file = cached_file(
-  File "/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/transformers/utils/hub.py", line 462, in cached_file
-    raise EnvironmentError(
-OSError: Incorrect path_or_model_id: '/pythia-70m'. Please provide either the path to a local folder or the repo_id of a model on the Hub.
-[2024-05-08 16:40:48,301] torch.distributed.elastic.multiprocessing.api: [ERROR] failed (exitcode: 1) local_rank: 0 (pid: 2311216) of binary: /nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/bin/python
-Traceback (most recent call last):
-  File "/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/bin/torchrun", line 8, in <module>
-    sys.exit(main())
-  File "/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 347, in wrapper
-    return f(*args, **kwargs)
-  File "/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/torch/distributed/run.py", line 812, in main
-    run(args)
-  File "/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/torch/distributed/run.py", line 803, in run
-    elastic_launch(
-  File "/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/torch/distributed/launcher/api.py", line 135, in __call__
-    return launch_agent(self._config, self._entrypoint, list(args))
-  File "/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/torch/distributed/launcher/api.py", line 268, in launch_agent
-    raise ChildFailedError(
-torch.distributed.elastic.multiprocessing.errors.ChildFailedError: 
-============================================================
-train_llm.py FAILED
-------------------------------------------------------------
-Failures:
-[1]:
-  time      : 2024-05-08_16:40:48
-  host      : sphinx2.stanford.edu
-  rank      : 1 (local_rank: 1)
-  exitcode  : 1 (pid: 2311217)
-  error_file: <N/A>
-  traceback : To enable traceback see: https://pytorch.org/docs/stable/elastic/errors.html
-------------------------------------------------------------
-Root Cause (first observed failure):
-[0]:
-  time      : 2024-05-08_16:40:48
-  host      : sphinx2.stanford.edu
-  rank      : 0 (local_rank: 0)
-  exitcode  : 1 (pid: 2311216)
-  error_file: <N/A>
-  traceback : To enable traceback see: https://pytorch.org/docs/stable/elastic/errors.html
-============================================================
-###############################
-end time: 2024-05-08 16:40:56.231524
-elapsed time: 0:00:20.028260
-slurm submission log: 2024-05-09 07:34:40.262041
-created following sbatch script: 
-
-###############################
-
-#!/bin/bash
-
-#SBATCH --account=nlp
-#SBATCH --cpus-per-task=16
-#SBATCH --gres=gpu:2
-#SBATCH --job-name=tthrush-job-2100676
-#SBATCH --mem=400G
-#SBATCH --nodelist=sphinx2
-#SBATCH --open-mode=append
-#SBATCH --output=/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms/pythia-70m_sciq/train_job_output.txt
-#SBATCH --partition=sphinx
-#SBATCH --time=14-0
-
-# activate your desired anaconda environment
-. /nlp/scr/tthrush/miniconda3/etc/profile.d/conda.sh ; conda activate pretraining-coreset-selection
-
-# cd to working directory
-cd .
-
-# launch commands
-srun --unbuffered run_as_child_processes 'torchrun --master_port 29505 --nproc_per_node=2 train_llm.py --dataset_id /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/training_data/sciq --output_dir /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms/pythia-70m_sciq --output_hub_id pythia-70m_sciq --model_id EleutherAI/pythia-70m --num_train_epochs 1 --learning_rate 1e-3 --warmup_ratio=0.1 --gradient_accumulation_steps 2'
-
-###############################
-
-submission to slurm complete!
-
-
-###############################
-slurm submission output
-
-Submitted batch job 7591654
-
-
-
-###############################
 
 ###############################
-start time: 2024-05-09 12:20:20.802372
+start time: 2024-05-09 16:40:46.819377
 machine: sphinx2
 conda env: pretraining-coreset-selection
 ###############################
 running following processes
 
-	torchrun --master_port 29505 --nproc_per_node=2 train_llm.py --dataset_id /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/training_data/sciq --output_dir /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms/pythia-70m_sciq --output_hub_id pythia-70m_sciq --model_id EleutherAI/pythia-70m --num_train_epochs 1 --learning_rate 1e-3 --warmup_ratio=0.1 --gradient_accumulation_steps 2
+	torchrun --master_port 29502 --nproc_per_node=2 train_llm.py --dataset_id /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/train_data_2/sciq --output_dir /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms_2/pythia-70m_sciq --output_hub_id pythia-70m_sciq --model_id EleutherAI/pythia-70m --num_train_epochs 14 --learning_rate 1e-3 --warmup_ratio=0.1 --gradient_accumulation_steps 2
 
 
 ###############################
 command outputs: 
 
 
-[2024-05-09 12:20:22,106] torch.distributed.run: [WARNING] 
-[2024-05-09 12:20:22,106] torch.distributed.run: [WARNING] *****************************************
-[2024-05-09 12:20:22,106] torch.distributed.run: [WARNING] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed. 
-[2024-05-09 12:20:22,106] torch.distributed.run: [WARNING] *****************************************
-05/09/2024 12:20:26 - INFO - __main__ - Script parameters ScriptArguments(dataset_id='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/training_data/sciq', output_dir='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms/pythia-70m_sciq', output_hub_id='pythia-70m_sciq', hf_hub_token=True, model_id='EleutherAI/pythia-70m', per_device_train_batch_size=256, num_train_epochs=1, learning_rate=0.001, gradient_accumulation_steps=2, from_scratch=True, warmup_ratio=0.1, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, weight_decay=0.01, lr_scheduler_type='cosine', local_rank=0, resume_from_checkpoint=False, deepspeed=None, peft=False)
-05/09/2024 12:20:26 - INFO - __main__ - Script parameters ScriptArguments(dataset_id='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/training_data/sciq', output_dir='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms/pythia-70m_sciq', output_hub_id='pythia-70m_sciq', hf_hub_token=True, model_id='EleutherAI/pythia-70m', per_device_train_batch_size=256, num_train_epochs=1, learning_rate=0.001, gradient_accumulation_steps=2, from_scratch=True, warmup_ratio=0.1, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, weight_decay=0.01, lr_scheduler_type='cosine', local_rank=0, resume_from_checkpoint=False, deepspeed=None, peft=False)
+[2024-05-09 16:40:54,330] torch.distributed.run: [WARNING] 
+[2024-05-09 16:40:54,330] torch.distributed.run: [WARNING] *****************************************
+[2024-05-09 16:40:54,330] torch.distributed.run: [WARNING] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed. 
+[2024-05-09 16:40:54,330] torch.distributed.run: [WARNING] *****************************************
+05/09/2024 16:40:58 - INFO - __main__ - Script parameters ScriptArguments(dataset_id='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/train_data_2/sciq', output_dir='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms_2/pythia-70m_sciq', output_hub_id='pythia-70m_sciq', hf_hub_token=True, model_id='EleutherAI/pythia-70m', per_device_train_batch_size=256, num_train_epochs=14, learning_rate=0.001, gradient_accumulation_steps=2, from_scratch=True, warmup_ratio=0.1, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, weight_decay=0.01, lr_scheduler_type='cosine', local_rank=0, resume_from_checkpoint=False, deepspeed=None, peft=False)
+05/09/2024 16:40:58 - INFO - __main__ - Script parameters ScriptArguments(dataset_id='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/train_data_2/sciq', output_dir='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/llms_2/pythia-70m_sciq', output_hub_id='pythia-70m_sciq', hf_hub_token=True, model_id='EleutherAI/pythia-70m', per_device_train_batch_size=256, num_train_epochs=14, learning_rate=0.001, gradient_accumulation_steps=2, from_scratch=True, warmup_ratio=0.1, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, weight_decay=0.01, lr_scheduler_type='cosine', local_rank=0, resume_from_checkpoint=False, deepspeed=None, peft=False)
 
  0%|          | 0/10682 [00:00<?, ?it/s][rank1]:[W reducer.cpp:1360] Warning: find_unused_parameters=True was specified in DDP constructor, but did not find any unused parameters in the forward pass. This flag results in an extra traversal of the autograd graph every iteration,  which can adversely affect performance. If your model indeed never has any unused parameters in the forward pass, consider turning this flag off. Note that this warning may be a false positive if your model has flow control causing later iterations to have unused parameters. (function operator())
 [rank0]:[W reducer.cpp:1360] Warning: find_unused_parameters=True was specified in DDP constructor, but did not find any unused parameters in the forward pass. This flag results in an extra traversal of the autograd graph every iteration,  which can adversely affect performance. If your model indeed never has any unused parameters in the forward pass, consider turning this flag off. Note that this warning may be a false positive if your model has flow control causing later iterations to have unused parameters. (function operator())
-
  0%|          | 1/10682 [00:02<6:30:11,  2.19s/it]
  0%|          | 2/10682 [00:02<3:55:55,  1.33s/it]
  0%|          | 3/10682 [00:03<2:57:32,  1.00it/s]
  0%|          | 4/10682 [00:04<2:22:40,  1.25it/s]
  0%|          | 5/10682 [00:04<2:03:22,  1.44it/s]
  0%|          | 6/10682 [00:05<1:51:43,  1.59it/s]
  0%|          | 7/10682 [00:05<1:44:17,  1.71it/s]
  0%|          | 8/10682 [00:06<1:39:25,  1.79it/s]
  0%|          | 9/10682 [00:06<1:36:18,  1.85it/s]
  0%|          | 10/10682 [00:07<1:34:04,  1.89it/s]
  0%|          | 11/10682 [00:07<1:32:30,  1.92it/s]
  0%|          | 12/10682 [00:08<1:31:19,  1.95it/s]
  0%|          | 13/10682 [00:08<1:30:37,  1.96it/s]
  0%|          | 14/10682 [00:09<1:30:03,  1.97it/s]
  0%|          | 15/10682 [00:09<1:29:40,  1.98it/s]
  0%|          | 16/10682 [00:10<1:29:25,  1.99it/s]
  0%|          | 17/10682 [00:10<1:29:13,  1.99it/s]
  0%|          | 18/10682 [00:11<1:29:02,  2.00it/s]
  0%|          | 19/10682 [00:11<1:28:59,  2.00it/s]
  0%|          | 20/10682 [00:12<1:28:50,  2.00it/s]
  0%|          | 21/10682 [00:12<1:28:43,  2.00it/s]
  0%|          | 22/10682 [00:13<1:28:42,  2.00it/s]
  0%|          | 23/10682 [00:13<1:28:48,  2.00it/s]
  0%|          | 24/10682 [00:14<1:28:44,  2.00it/s]
  0%|          | 25/10682 [00:14<1:28:39,  2.00it/s]{'loss': 10.6567, 'grad_norm': 1.401993989944458, 'learning_rate': 2.3386342376052384e-05, 'epoch': 0.0}
-
                                                    

  0%|          | 25/10682 [00:14<1:28:39,  2.00it/s]
  0%|          | 26/10682 [00:15<1:28:57,  2.00it/s]
  0%|          | 27/10682 [00:15<1:28:52,  2.00it/s]
  0%|          | 28/10682 [00:16<1:29:01,  1.99it/s]
  0%|          | 29/10682 [00:16<1:28:55,  2.00it/s]
  0%|          | 30/10682 [00:17<1:28:51,  2.00it/s]
  0%|          | 31/10682 [00:17<1:28:40,  2.00it/s]
  0%|          | 32/10682 [00:18<1:28:37,  2.00it/s]
  0%|          | 33/10682 [00:18<1:28:36,  2.00it/s]
  0%|          | 34/10682 [00:19<1:28:34,  2.00it/s]
  0%|          | 35/10682 [00:19<1:28:30,  2.00it/s]
  0%|          | 36/10682 [00:20<1:28:24,  2.01it/s]
  0%|          | 37/10682 [00:20<1:28:19,  2.01it/s]
  0%|          | 38/10682 [00:21<1:28:19,  2.01it/s]
  0%|          | 39/10682 [00:21<1:28:21,  2.01it/s]
  0%|          | 40/10682 [00:22<1:28:18,  2.01it/s]
  0%|          | 41/10682 [00:22<1:28:23,  2.01it/s]
  0%|          | 42/10682 [00:23<1:28:22,  2.01it/s]
  0%|          | 43/10682 [00:23<1:28:23,  2.01it/s]
  0%|          | 44/10682 [00:23<1:28:18,  2.01it/s]
  0%|          | 45/10682 [00:24<1:28:13,  2.01it/s]
  0%|          | 46/10682 [00:24<1:28:14,  2.01it/s]
  0%|          | 47/10682 [00:25<1:28:24,  2.00it/s]
  0%|          | 48/10682 [00:25<1:28:13,  2.01it/s]
  0%|          | 49/10682 [00:26<1:28:17,  2.01it/s]
  0%|          | 50/10682 [00:26<1:28:15,  2.01it/s]
                                                    {'loss': 9.9425, 'grad_norm': 1.3134275674819946, 'learning_rate': 4.677268475210477e-05, 'epoch': 0.0}
-
  0%|          | 50/10682 [00:26<1:28:15,  2.01it/s]
  0%|          | 51/10682 [00:27<1:28:29,  2.00it/s]
  0%|          | 52/10682 [00:27<1:28:21,  2.01it/s]
  0%|          | 53/10682 [00:28<1:28:20,  2.01it/s]
  1%|          | 54/10682 [00:28<1:28:09,  2.01it/s]
  1%|          | 55/10682 [00:29<1:28:09,  2.01it/s]
  1%|          | 56/10682 [00:29<1:28:05,  2.01it/s]
  1%|          | 57/10682 [00:30<1:28:04,  2.01it/s]
  1%|          | 58/10682 [00:30<1:28:06,  2.01it/s]
  1%|          | 59/10682 [00:31<1:27:58,  2.01it/s]
  1%|          | 60/10682 [00:31<1:28:02,  2.01it/s]
  1%|          | 61/10682 [00:32<1:27:58,  2.01it/s]
  1%|          | 62/10682 [00:32<1:27:54,  2.01it/s]
  1%|          | 63/10682 [00:33<1:27:55,  2.01it/s]
  1%|          | 64/10682 [00:33<1:27:49,  2.02it/s]
  1%|          | 65/10682 [00:34<1:27:49,  2.01it/s]
  1%|          | 66/10682 [00:34<1:27:52,  2.01it/s]
  1%|          | 67/10682 [00:35<1:27:54,  2.01it/s]
  1%|          | 68/10682 [00:35<1:27:54,  2.01it/s]
  1%|          | 69/10682 [00:36<1:27:57,  2.01it/s]
  1%|          | 70/10682 [00:36<1:27:53,  2.01it/s]
  1%|          | 71/10682 [00:37<1:27:50,  2.01it/s]
  1%|          | 72/10682 [00:37<1:27:46,  2.01it/s]
  1%|          | 73/10682 [00:38<1:27:54,  2.01it/s]
  1%|          | 74/10682 [00:38<1:27:54,  2.01it/s]
  1%|          | 75/10682 [00:39<1:27:48,  2.01it/s]
                                                    {'loss': 9.2271, 'grad_norm': 1.1519968509674072, 'learning_rate': 7.015902712815715e-05, 'epoch': 0.01}
-
  1%|          | 75/10682 [00:39<1:27:48,  2.01it/s]
  1%|          | 76/10682 [00:39<1:27:57,  2.01it/s]
  1%|          | 77/10682 [00:40<1:27:48,  2.01it/s]
  1%|          | 78/10682 [00:40<1:27:47,  2.01it/s]
  1%|          | 79/10682 [00:41<1:27:47,  2.01it/s]
  1%|          | 80/10682 [00:41<1:27:42,  2.01it/s]
  1%|          | 81/10682 [00:42<1:27:53,  2.01it/s]
  1%|          | 82/10682 [00:42<1:27:52,  2.01it/s]
  1%|          | 83/10682 [00:43<1:27:48,  2.01it/s]
  1%|          | 84/10682 [00:43<1:27:48,  2.01it/s]
  1%|          | 85/10682 [00:44<1:27:46,  2.01it/s]
  1%|          | 86/10682 [00:44<1:27:43,  2.01it/s]
  1%|          | 87/10682 [00:45<1:27:47,  2.01it/s]
  1%|          | 88/10682 [00:45<1:27:43,  2.01it/s]
  1%|          | 89/10682 [00:46<1:27:46,  2.01it/s]
  1%|          | 90/10682 [00:46<1:27:48,  2.01it/s]
  1%|          | 91/10682 [00:47<1:27:45,  2.01it/s]
  1%|          | 92/10682 [00:47<1:27:47,  2.01it/s]
  1%|          | 93/10682 [00:48<1:27:40,  2.01it/s]
  1%|          | 94/10682 [00:48<1:27:38,  2.01it/s]
  1%|          | 95/10682 [00:49<1:27:40,  2.01it/s]
  1%|          | 96/10682 [00:49<1:27:37,  2.01it/s]
  1%|          | 97/10682 [00:50<1:27:36,  2.01it/s]
  1%|          | 98/10682 [00:50<1:27:33,  2.01it/s]
  1%|          | 99/10682 [00:51<1:27:45,  2.01it/s]
  1%|          | 100/10682 [00:51<1:27:53,  2.01it/s]{'loss': 8.4443, 'grad_norm': 0.8334071040153503, 'learning_rate': 9.354536950420954e-05, 'epoch': 0.01}
                                                     
-
  1%|          | 100/10682 [00:51<1:27:53,  2.01it/s]
  1%|          | 101/10682 [00:52<1:27:53,  2.01it/s]
  1%|          | 102/10682 [00:52<1:27:47,  2.01it/s]
  1%|          | 103/10682 [00:53<1:27:47,  2.01it/s]
  1%|          | 104/10682 [00:53<1:27:47,  2.01it/s]
  1%|          | 105/10682 [00:54<1:27:45,  2.01it/s]
  1%|          | 106/10682 [00:54<1:27:43,  2.01it/s]
  1%|          | 107/10682 [00:55<1:27:44,  2.01it/s]
  1%|          | 108/10682 [00:55<1:27:46,  2.01it/s]
  1%|          | 109/10682 [00:56<1:27:41,  2.01it/s]
  1%|          | 110/10682 [00:56<1:27:44,  2.01it/s]
  1%|          | 111/10682 [00:57<1:27:42,  2.01it/s]
  1%|          | 112/10682 [00:57<1:27:37,  2.01it/s]
  1%|          | 113/10682 [00:58<1:27:40,  2.01it/s]
  1%|          | 114/10682 [00:58<1:27:42,  2.01it/s]
  1%|          | 115/10682 [00:59<1:27:43,  2.01it/s]
  1%|          | 116/10682 [00:59<1:27:33,  2.01it/s]
  1%|          | 117/10682 [01:00<1:27:39,  2.01it/s]
  1%|          | 118/10682 [01:00<1:27:33,  2.01it/s]
  1%|          | 119/10682 [01:01<1:27:34,  2.01it/s]
  1%|          | 120/10682 [01:01<1:27:37,  2.01it/s]
  1%|          | 121/10682 [01:02<1:27:33,  2.01it/s]
  1%|          | 122/10682 [01:02<1:27:29,  2.01it/s]
  1%|          | 123/10682 [01:03<1:27:35,  2.01it/s]
  1%|          | 124/10682 [01:03<1:27:21,  2.01it/s]
  1%|          | 125/10682 [01:04<1:27:23,  2.01it/s]
                                                     
{'loss': 7.7875, 'grad_norm': 0.4627320170402527, 'learning_rate': 0.00011693171188026193, 'epoch': 0.01}
-
  1%|          | 125/10682 [01:04<1:27:23,  2.01it/s]
  1%|          | 126/10682 [01:04<1:27:35,  2.01it/s]
  1%|          | 127/10682 [01:05<1:27:30,  2.01it/s]
  1%|          | 128/10682 [01:05<1:27:30,  2.01it/s]
  1%|          | 129/10682 [01:06<1:27:39,  2.01it/s]
  1%|          | 130/10682 [01:06<1:27:35,  2.01it/s]
  1%|          | 131/10682 [01:07<1:27:34,  2.01it/s]
  1%|          | 132/10682 [01:07<1:27:42,  2.00it/s]
  1%|          | 133/10682 [01:08<1:27:32,  2.01it/s]
  1%|▏         | 134/10682 [01:08<1:27:32,  2.01it/s]
  1%|▏         | 135/10682 [01:09<1:27:31,  2.01it/s]
  1%|▏         | 136/10682 [01:09<1:27:21,  2.01it/s]
  1%|▏         | 137/10682 [01:10<1:27:24,  2.01it/s]
  1%|▏         | 138/10682 [01:10<1:27:27,  2.01it/s]
  1%|▏         | 139/10682 [01:11<1:27:22,  2.01it/s]
  1%|▏         | 140/10682 [01:11<1:27:27,  2.01it/s]
  1%|▏         | 141/10682 [01:12<1:27:25,  2.01it/s]
  1%|▏         | 142/10682 [01:12<1:27:23,  2.01it/s]
  1%|▏         | 143/10682 [01:13<1:27:25,  2.01it/s]
  1%|▏         | 144/10682 [01:13<1:27:21,  2.01it/s]
  1%|▏         | 145/10682 [01:14<1:27:17,  2.01it/s]
  1%|▏         | 146/10682 [01:14<1:27:21,  2.01it/s]
  1%|▏         | 147/10682 [01:15<1:27:17,  2.01it/s]
  1%|▏         | 148/10682 [01:15<1:27:17,  2.01it/s]
  1%|▏         | 149/10682 [01:16<1:27:23,  2.01it/s]
  1%|▏         | 150/10682 [01:16<1:27:17,  2.01it/s]{'loss': 7.3151, 'grad_norm': 0.45290571451187134, 'learning_rate': 0.0001403180542563143, 'epoch': 0.01}
                                                     
-
  1%|▏         | 150/10682 [01:16<1:27:17,  2.01it/s]
  1%|▏         | 151/10682 [01:17<1:27:20,  2.01it/s]
  1%|▏         | 152/10682 [01:17<1:27:23,  2.01it/s]
  1%|▏         | 153/10682 [01:18<1:27:15,  2.01it/s]
  1%|▏         | 154/10682 [01:18<1:27:17,  2.01it/s]
  1%|▏         | 155/10682 [01:19<1:27:21,  2.01it/s]
  1%|▏         | 156/10682 [01:19<1:27:13,  2.01it/s]
  1%|▏         | 157/10682 [01:20<1:27:14,  2.01it/s]
  1%|▏         | 158/10682 [01:20<1:27:12,  2.01it/s]
  1%|▏         | 159/10682 [01:21<1:27:08,  2.01it/s]
  1%|▏         | 160/10682 [01:21<1:27:12,  2.01it/s]
  2%|▏         | 161/10682 [01:22<1:27:12,  2.01it/s]
  2%|▏         | 162/10682 [01:22<1:27:13,  2.01it/s]
  2%|▏         | 163/10682 [01:23<1:27:18,  2.01it/s]
  2%|▏         | 164/10682 [01:23<1:27:13,  2.01it/s]
  2%|▏         | 165/10682 [01:24<1:27:09,  2.01it/s]
  2%|▏         | 166/10682 [01:24<1:27:15,  2.01it/s]
  2%|▏         | 167/10682 [01:25<1:27:15,  2.01it/s]
  2%|▏         | 168/10682 [01:25<1:27:24,  2.00it/s]
  2%|▏         | 169/10682 [01:26<1:27:18,  2.01it/s]
  2%|▏         | 170/10682 [01:26<1:27:16,  2.01it/s]
  2%|▏         | 171/10682 [01:27<1:27:04,  2.01it/s]
  2%|▏         | 172/10682 [01:27<1:27:11,  2.01it/s]
  2%|▏         | 173/10682 [01:28<1:26:59,  2.01it/s]
  2%|▏         | 174/10682 [01:28<1:26:59,  2.01it/s]
  2%|▏         | 175/10682 [01:29<1:26:50,  2.02it/s]{'loss': 6.9184, 'grad_norm': 0.5876092910766602, 'learning_rate': 0.00016370439663236668, 'epoch': 0.02}
-
                                                     

  2%|▏         | 175/10682 [01:29<1:26:50,  2.02it/s]
  2%|▏         | 176/10682 [01:29<1:26:54,  2.01it/s]
  2%|▏         | 177/10682 [01:30<1:26:56,  2.01it/s]
  2%|▏         | 178/10682 [01:30<1:26:53,  2.01it/s]
  2%|▏         | 179/10682 [01:31<1:26:56,  2.01it/s]
  2%|▏         | 180/10682 [01:31<1:26:52,  2.01it/s]
  2%|▏         | 181/10682 [01:32<1:26:49,  2.02it/s]
  2%|▏         | 182/10682 [01:32<1:26:48,  2.02it/s]
  2%|▏         | 183/10682 [01:33<1:26:53,  2.01it/s]
  2%|▏         | 184/10682 [01:33<1:27:00,  2.01it/s]
  2%|▏         | 185/10682 [01:34<1:26:57,  2.01it/s]
  2%|▏         | 186/10682 [01:34<1:26:57,  2.01it/s]
  2%|▏         | 187/10682 [01:35<1:26:56,  2.01it/s]
  2%|▏         | 188/10682 [01:35<1:26:53,  2.01it/s]
  2%|▏         | 189/10682 [01:36<1:26:54,  2.01it/s]
  2%|▏         | 190/10682 [01:36<1:26:52,  2.01it/s]
  2%|▏         | 191/10682 [01:37<1:26:51,  2.01it/s]
  2%|▏         | 192/10682 [01:37<1:26:53,  2.01it/s]
  2%|▏         | 193/10682 [01:38<1:26:54,  2.01it/s]
  2%|▏         | 194/10682 [01:38<1:26:58,  2.01it/s]
  2%|▏         | 195/10682 [01:39<1:26:52,  2.01it/s]
  2%|▏         | 196/10682 [01:39<1:26:58,  2.01it/s]
  2%|▏         | 197/10682 [01:40<1:26:53,  2.01it/s]
  2%|▏         | 198/10682 [01:40<1:26:52,  2.01it/s]
  2%|▏         | 199/10682 [01:41<1:26:58,  2.01it/s]
  2%|▏         | 200/10682 [01:41<1:26:44,  2.01it/s]
                                                     {'loss': 6.5885, 'grad_norm': 0.4638342559337616, 'learning_rate': 0.00018709073900841907, 'epoch': 0.02}
-
  2%|▏         | 200/10682 [01:41<1:26:44,  2.01it/s]
  2%|▏         | 201/10682 [01:42<1:26:50,  2.01it/s]
  2%|▏         | 202/10682 [01:42<1:26:46,  2.01it/s]
  2%|▏         | 203/10682 [01:43<1:26:47,  2.01it/s]
  2%|▏         | 204/10682 [01:43<1:26:45,  2.01it/s]
  2%|▏         | 205/10682 [01:44<1:26:39,  2.02it/s]
  2%|▏         | 206/10682 [01:44<1:26:38,  2.02it/s]
  2%|▏         | 207/10682 [01:45<1:26:33,  2.02it/s]
  2%|▏         | 208/10682 [01:45<1:26:35,  2.02it/s]
  2%|▏         | 209/10682 [01:46<1:26:34,  2.02it/s]
  2%|▏         | 210/10682 [01:46<1:26:43,  2.01it/s]
  2%|▏         | 211/10682 [01:47<1:26:51,  2.01it/s]
  2%|▏         | 212/10682 [01:47<1:26:41,  2.01it/s]
  2%|▏         | 213/10682 [01:48<1:26:41,  2.01it/s]
  2%|▏         | 214/10682 [01:48<1:26:39,  2.01it/s]
  2%|▏         | 215/10682 [01:49<1:26:37,  2.01it/s]
  2%|▏         | 216/10682 [01:49<1:26:47,  2.01it/s]
  2%|▏         | 217/10682 [01:50<1:26:42,  2.01it/s]
  2%|▏         | 218/10682 [01:50<1:26:39,  2.01it/s]
  2%|▏         | 219/10682 [01:51<1:26:38,  2.01it/s]
  2%|▏         | 220/10682 [01:51<1:26:29,  2.02it/s]
  2%|▏         | 221/10682 [01:52<1:26:30,  2.02it/s]
  2%|▏         | 222/10682 [01:52<1:26:27,  2.02it/s]
  2%|▏         | 223/10682 [01:53<1:26:28,  2.02it/s]
  2%|▏         | 224/10682 [01:53<1:26:21,  2.02it/s]
  2%|▏         | 225/10682 [01:53<1:26:32,  2.01it/s]
                                                     {'loss': 6.3284, 'grad_norm': 0.5081571936607361, 'learning_rate': 0.00021047708138447147, 'epoch': 0.02}
-

  2%|▏         | 225/10682 [01:53<1:26:32,  2.01it/s]
  2%|▏         | 226/10682 [01:54<1:26:53,  2.01it/s]
  2%|▏         | 227/10682 [01:54<1:26:46,  2.01it/s]
  2%|▏         | 228/10682 [01:55<1:26:47,  2.01it/s]
  2%|▏         | 229/10682 [01:55<1:26:48,  2.01it/s]
  2%|▏         | 230/10682 [01:56<1:26:41,  2.01it/s]
  2%|▏         | 231/10682 [01:56<1:26:48,  2.01it/s]
  2%|▏         | 232/10682 [01:57<1:26:35,  2.01it/s]
  2%|▏         | 233/10682 [01:57<1:26:38,  2.01it/s]
  2%|▏         | 234/10682 [01:58<1:26:37,  2.01it/s]
  2%|▏         | 235/10682 [01:58<1:26:28,  2.01it/s]
  2%|▏         | 236/10682 [01:59<1:26:28,  2.01it/s]
  2%|▏         | 237/10682 [01:59<1:26:21,  2.02it/s]
  2%|▏         | 238/10682 [02:00<1:26:26,  2.01it/s]
  2%|▏         | 239/10682 [02:00<1:26:12,  2.02it/s]
  2%|▏         | 240/10682 [02:01<1:26:26,  2.01it/s]
  2%|▏         | 241/10682 [02:01<1:26:24,  2.01it/s]
  2%|▏         | 242/10682 [02:02<1:26:23,  2.01it/s]
  2%|▏         | 243/10682 [02:02<1:26:28,  2.01it/s]
  2%|▏         | 244/10682 [02:03<1:26:26,  2.01it/s]
  2%|▏         | 245/10682 [02:03<1:26:20,  2.01it/s]
  2%|▏         | 246/10682 [02:04<1:26:19,  2.01it/s]
  2%|▏         | 247/10682 [02:04<1:26:14,  2.02it/s]
  2%|▏         | 248/10682 [02:05<1:26:13,  2.02it/s]
  2%|▏         | 249/10682 [02:05<1:26:10,  2.02it/s]
  2%|▏         | 250/10682 [02:06<1:26:13,  2.02it/s]
                                                     {'loss': 6.1249, 'grad_norm': 0.8689858317375183, 'learning_rate': 0.00023386342376052386, 'epoch': 0.02}
-
  2%|▏         | 250/10682 [02:06<1:26:13,  2.02it/s]
  2%|▏         | 251/10682 [02:06<1:26:15,  2.02it/s]
  2%|▏         | 252/10682 [02:07<1:26:20,  2.01it/s]
  2%|▏         | 253/10682 [02:07<1:26:15,  2.02it/s]
  2%|▏         | 254/10682 [02:08<1:26:22,  2.01it/s]
  2%|▏         | 255/10682 [02:08<1:26:20,  2.01it/s]
  2%|▏         | 256/10682 [02:09<1:26:14,  2.01it/s]
  2%|▏         | 257/10682 [02:09<1:26:17,  2.01it/s]
  2%|▏         | 258/10682 [02:10<1:26:16,  2.01it/s]
  2%|▏         | 259/10682 [02:10<1:26:11,  2.02it/s]
  2%|▏         | 260/10682 [02:11<1:26:09,  2.02it/s]
  2%|▏         | 261/10682 [02:11<1:26:04,  2.02it/s]
  2%|▏         | 262/10682 [02:12<1:26:02,  2.02it/s]
  2%|▏         | 263/10682 [02:12<1:25:59,  2.02it/s]
  2%|▏         | 264/10682 [02:13<1:26:04,  2.02it/s]
  2%|▏         | 265/10682 [02:13<1:25:58,  2.02it/s]
  2%|▏         | 266/10682 [02:14<1:26:01,  2.02it/s]
  2%|▏         | 267/10682 [02:14<1:25:58,  2.02it/s]
  3%|▎         | 268/10682 [02:15<1:26:05,  2.02it/s]
  3%|▎         | 269/10682 [02:15<1:26:07,  2.02it/s]
  3%|▎         | 270/10682 [02:16<1:26:04,  2.02it/s]
  3%|▎         | 271/10682 [02:16<1:26:01,  2.02it/s]
  3%|▎         | 272/10682 [02:17<1:26:01,  2.02it/s]
  3%|▎         | 273/10682 [02:17<1:26:01,  2.02it/s]
  3%|▎         | 274/10682 [02:18<1:25:58,  2.02it/s]
  3%|▎         | 275/10682 [02:18<1:25:57,  2.02it/s]
                                                     {'loss': 5.9562, 'grad_norm': 0.7329740524291992, 'learning_rate': 0.00025724976613657625, 'epoch': 0.03}
-
  3%|▎         | 275/10682 [02:18<1:25:57,  2.02it/s]
  3%|▎         | 276/10682 [02:19<1:26:00,  2.02it/s]
  3%|▎         | 277/10682 [02:19<1:26:01,  2.02it/s]
  3%|▎         | 278/10682 [02:20<1:26:01,  2.02it/s]
  3%|▎         | 279/10682 [02:20<1:25:58,  2.02it/s]
  3%|▎         | 280/10682 [02:21<1:25:54,  2.02it/s]
  3%|▎         | 281/10682 [02:21<1:25:59,  2.02it/s]
  3%|▎         | 282/10682 [02:22<1:26:00,  2.02it/s]
  3%|▎         | 283/10682 [02:22<1:25:59,  2.02it/s]
  3%|▎         | 284/10682 [02:23<1:26:02,  2.01it/s]
  3%|▎         | 285/10682 [02:23<1:26:00,  2.01it/s]
  3%|▎         | 286/10682 [02:24<1:25:59,  2.01it/s]
  3%|▎         | 287/10682 [02:24<1:25:53,  2.02it/s]
  3%|▎         | 288/10682 [02:25<1:25:53,  2.02it/s]
  3%|▎         | 289/10682 [02:25<1:25:54,  2.02it/s]
  3%|▎         | 290/10682 [02:26<1:25:50,  2.02it/s]
  3%|▎         | 291/10682 [02:26<1:25:56,  2.02it/s]
  3%|▎         | 292/10682 [02:27<1:25:50,  2.02it/s]
  3%|▎         | 293/10682 [02:27<1:25:51,  2.02it/s]
  3%|▎         | 294/10682 [02:28<1:25:47,  2.02it/s]
  3%|▎         | 295/10682 [02:28<1:25:52,  2.02it/s]
  3%|▎         | 296/10682 [02:29<1:25:49,  2.02it/s]
  3%|▎         | 297/10682 [02:29<1:25:53,  2.02it/s]
  3%|▎         | 298/10682 [02:30<1:25:54,  2.01it/s]
  3%|▎         | 299/10682 [02:30<1:25:51,  2.02it/s]
  3%|▎         | 300/10682 [02:31<1:25:48,  2.02it/s]
                                                     {'loss': 5.8154, 'grad_norm': 0.6380777955055237, 'learning_rate': 0.0002806361085126286, 'epoch': 0.03}
-
  3%|▎         | 300/10682 [02:31<1:25:48,  2.02it/s]
  3%|▎         | 301/10682 [02:31<1:25:55,  2.01it/s]
  3%|▎         | 302/10682 [02:32<1:25:54,  2.01it/s]
  3%|▎         | 303/10682 [02:32<1:25:53,  2.01it/s]
  3%|▎         | 304/10682 [02:33<1:25:46,  2.02it/s]
  3%|▎         | 305/10682 [02:33<1:25:41,  2.02it/s]
  3%|▎         | 306/10682 [02:34<1:25:40,  2.02it/s]
  3%|▎         | 307/10682 [02:34<1:25:37,  2.02it/s]
  3%|▎         | 308/10682 [02:35<1:25:42,  2.02it/s]
  3%|▎         | 309/10682 [02:35<1:25:39,  2.02it/s]
  3%|▎         | 310/10682 [02:36<1:25:37,  2.02it/s]
  3%|▎         | 311/10682 [02:36<1:25:43,  2.02it/s]
  3%|▎         | 312/10682 [02:37<1:25:43,  2.02it/s]
  3%|▎         | 313/10682 [02:37<1:25:44,  2.02it/s]
  3%|▎         | 314/10682 [02:38<1:25:45,  2.02it/s]
  3%|▎         | 315/10682 [02:38<1:25:49,  2.01it/s]
  3%|▎         | 316/10682 [02:39<1:25:42,  2.02it/s]
  3%|▎         | 317/10682 [02:39<1:25:39,  2.02it/s]
  3%|▎         | 318/10682 [02:40<1:25:31,  2.02it/s]
  3%|▎         | 319/10682 [02:40<1:25:30,  2.02it/s]
  3%|▎         | 320/10682 [02:41<1:25:26,  2.02it/s]
  3%|▎         | 321/10682 [02:41<1:25:34,  2.02it/s]
  3%|▎         | 322/10682 [02:42<1:25:29,  2.02it/s]
  3%|▎         | 323/10682 [02:42<1:25:31,  2.02it/s]
  3%|▎         | 324/10682 [02:43<1:25:34,  2.02it/s]
  3%|▎         | 325/10682 [02:43<1:25:35,  2.02it/s]{'loss': 5.6903, 'grad_norm': 0.7655269503593445, 'learning_rate': 0.00030402245088868103, 'epoch': 0.03}
                                                     
-
  3%|▎         | 325/10682 [02:43<1:25:35,  2.02it/s]
  3%|▎         | 326/10682 [02:44<1:25:40,  2.01it/s]
  3%|▎         | 327/10682 [02:44<1:39:35,  1.73it/s]
  3%|▎         | 328/10682 [02:45<1:49:05,  1.58it/s]
  3%|▎         | 329/10682 [02:46<1:42:00,  1.69it/s]
  3%|▎         | 330/10682 [02:46<1:37:01,  1.78it/s]
  3%|▎         | 331/10682 [02:47<1:33:37,  1.84it/s]
  3%|▎         | 332/10682 [02:47<1:31:06,  1.89it/s]
  3%|▎         | 333/10682 [02:48<1:29:30,  1.93it/s]
  3%|▎         | 334/10682 [02:48<1:28:12,  1.96it/s]
  3%|▎         | 335/10682 [02:49<1:27:27,  1.97it/s]
  3%|▎         | 336/10682 [02:49<1:26:49,  1.99it/s]
  3%|▎         | 337/10682 [02:50<1:26:26,  1.99it/s]
  3%|▎         | 338/10682 [02:50<1:26:11,  2.00it/s]
  3%|▎         | 339/10682 [02:51<1:25:55,  2.01it/s]
  3%|▎         | 340/10682 [02:51<1:25:45,  2.01it/s]
  3%|▎         | 341/10682 [02:52<1:25:38,  2.01it/s]
  3%|▎         | 342/10682 [02:52<1:25:34,  2.01it/s]
  3%|▎         | 343/10682 [02:53<1:25:26,  2.02it/s]
  3%|▎         | 344/10682 [02:53<1:25:26,  2.02it/s]
  3%|▎         | 345/10682 [02:54<1:25:17,  2.02it/s]
  3%|▎         | 346/10682 [02:54<1:25:20,  2.02it/s]
  3%|▎         | 347/10682 [02:55<1:25:25,  2.02it/s]
  3%|▎         | 348/10682 [02:55<1:25:20,  2.02it/s]
  3%|▎         | 349/10682 [02:56<1:25:22,  2.02it/s]
  3%|▎         | 350/10682 [02:56<1:25:21,  2.02it/s]
                                                     
{'loss': 5.5783, 'grad_norm': 0.6206523776054382, 'learning_rate': 0.00032740879326473337, 'epoch': 0.03}
-
  3%|▎         | 350/10682 [02:56<1:25:21,  2.02it/s]
  3%|▎         | 351/10682 [02:57<1:25:31,  2.01it/s]
  3%|▎         | 352/10682 [02:57<1:25:22,  2.02it/s]
  3%|▎         | 353/10682 [02:58<1:25:24,  2.02it/s]
  3%|▎         | 354/10682 [02:58<1:25:14,  2.02it/s]
  3%|▎         | 355/10682 [02:59<1:25:16,  2.02it/s]
  3%|▎         | 356/10682 [02:59<1:25:16,  2.02it/s]
  3%|▎         | 357/10682 [03:00<1:25:20,  2.02it/s]
  3%|▎         | 358/10682 [03:00<1:25:22,  2.02it/s]
  3%|▎         | 359/10682 [03:01<1:25:22,  2.02it/s]
  3%|▎         | 360/10682 [03:01<1:25:25,  2.01it/s]
  3%|▎         | 361/10682 [03:01<1:25:25,  2.01it/s]
  3%|▎         | 362/10682 [03:02<1:25:30,  2.01it/s]
  3%|▎         | 363/10682 [03:02<1:25:24,  2.01it/s]
  3%|▎         | 364/10682 [03:03<1:25:17,  2.02it/s]
  3%|▎         | 365/10682 [03:03<1:25:16,  2.02it/s]
  3%|▎         | 366/10682 [03:04<1:25:13,  2.02it/s]
  3%|▎         | 367/10682 [03:04<1:25:07,  2.02it/s]
  3%|▎         | 368/10682 [03:05<1:25:10,  2.02it/s]
  3%|▎         | 369/10682 [03:05<1:25:08,  2.02it/s]
  3%|▎         | 370/10682 [03:06<1:25:10,  2.02it/s]
  3%|▎         | 371/10682 [03:06<1:25:03,  2.02it/s]
  3%|▎         | 372/10682 [03:07<1:25:09,  2.02it/s]
  3%|▎         | 373/10682 [03:07<1:25:02,  2.02it/s]
  4%|▎         | 374/10682 [03:08<1:25:16,  2.01it/s]
  4%|▎         | 375/10682 [03:08<1:25:08,  2.02it/s]
                                                     
{'loss': 5.4977, 'grad_norm': 0.5949773192405701, 'learning_rate': 0.0003507951356407858, 'epoch': 0.04}
-
  4%|▎         | 375/10682 [03:08<1:25:08,  2.02it/s]
  4%|▎         | 376/10682 [03:09<1:25:16,  2.01it/s]
  4%|▎         | 377/10682 [03:09<1:25:07,  2.02it/s]
  4%|▎         | 378/10682 [03:10<1:25:07,  2.02it/s]
  4%|▎         | 379/10682 [03:10<1:25:00,  2.02it/s]
  4%|▎         | 380/10682 [03:11<1:25:01,  2.02it/s]
  4%|▎         | 381/10682 [03:11<1:24:59,  2.02it/s]
  4%|▎         | 382/10682 [03:12<1:24:58,  2.02it/s]
  4%|▎         | 383/10682 [03:12<1:25:00,  2.02it/s]
  4%|▎         | 384/10682 [03:13<1:24:57,  2.02it/s]
  4%|▎         | 385/10682 [03:13<1:25:01,  2.02it/s]
  4%|▎         | 386/10682 [03:14<1:25:10,  2.01it/s]
  4%|▎         | 387/10682 [03:14<1:25:05,  2.02it/s]
  4%|▎         | 388/10682 [03:15<1:25:03,  2.02it/s]
  4%|▎         | 389/10682 [03:15<1:25:01,  2.02it/s]
  4%|▎         | 390/10682 [03:16<1:25:03,  2.02it/s]
  4%|▎         | 391/10682 [03:16<1:24:56,  2.02it/s]
  4%|▎         | 392/10682 [03:17<1:24:55,  2.02it/s]
  4%|▎         | 393/10682 [03:17<1:24:57,  2.02it/s]
  4%|▎         | 394/10682 [03:18<1:24:57,  2.02it/s]
  4%|▎         | 395/10682 [03:18<1:24:55,  2.02it/s]
  4%|▎         | 396/10682 [03:19<1:25:04,  2.02it/s]
  4%|▎         | 397/10682 [03:19<1:24:54,  2.02it/s]
  4%|▎         | 398/10682 [03:20<1:24:57,  2.02it/s]
  4%|▎         | 399/10682 [03:20<1:25:06,  2.01it/s]
  4%|▎         | 400/10682 [03:21<1:25:05,  2.01it/s]
                                                     {'loss': 5.3969, 'grad_norm': 0.654660165309906, 'learning_rate': 0.00037418147801683815, 'epoch': 0.04}
-
  4%|▎         | 400/10682 [03:21<1:25:05,  2.01it/s]
  4%|▍         | 401/10682 [03:21<1:25:17,  2.01it/s]
  4%|▍         | 402/10682 [03:22<1:25:10,  2.01it/s]
  4%|▍         | 403/10682 [03:22<1:25:10,  2.01it/s]
  4%|▍         | 404/10682 [03:23<1:25:10,  2.01it/s]
  4%|▍         | 405/10682 [03:23<1:25:06,  2.01it/s]
  4%|▍         | 406/10682 [03:24<1:25:03,  2.01it/s]
  4%|▍         | 407/10682 [03:24<1:24:59,  2.01it/s]
  4%|▍         | 408/10682 [03:25<1:25:00,  2.01it/s]
  4%|▍         | 409/10682 [03:25<1:24:57,  2.02it/s]
  4%|▍         | 410/10682 [03:26<1:24:58,  2.01it/s]
  4%|▍         | 411/10682 [03:26<1:25:05,  2.01it/s]
  4%|▍         | 412/10682 [03:27<1:25:02,  2.01it/s]
  4%|▍         | 413/10682 [03:27<1:24:59,  2.01it/s]
  4%|▍         | 414/10682 [03:28<1:25:07,  2.01it/s]
  4%|▍         | 415/10682 [03:28<1:25:00,  2.01it/s]
  4%|▍         | 416/10682 [03:29<1:24:59,  2.01it/s]
  4%|▍         | 417/10682 [03:29<1:24:50,  2.02it/s]
  4%|▍         | 418/10682 [03:30<1:24:51,  2.02it/s]
  4%|▍         | 419/10682 [03:30<1:24:46,  2.02it/s]
  4%|▍         | 420/10682 [03:31<1:24:47,  2.02it/s]
  4%|▍         | 421/10682 [03:31<1:24:44,  2.02it/s]
  4%|▍         | 422/10682 [03:32<1:24:43,  2.02it/s]
  4%|▍         | 423/10682 [03:32<1:24:41,  2.02it/s]
  4%|▍         | 424/10682 [03:33<1:24:39,  2.02it/s]
  4%|▍         | 425/10682 [03:33<1:24:39,  2.02it/s]{'loss': 5.328, 'grad_norm': 0.5211993455886841, 'learning_rate': 0.0003975678203928906, 'epoch': 0.04}
                                                     
-
  4%|▍         | 425/10682 [03:33<1:24:39,  2.02it/s]
  4%|▍         | 426/10682 [03:34<1:24:51,  2.01it/s]
  4%|▍         | 427/10682 [03:34<1:24:49,  2.02it/s]
  4%|▍         | 428/10682 [03:35<1:24:41,  2.02it/s]
  4%|▍         | 429/10682 [03:35<1:24:42,  2.02it/s]
  4%|▍         | 430/10682 [03:36<1:24:39,  2.02it/s]
  4%|▍         | 431/10682 [03:36<1:24:40,  2.02it/s]
  4%|▍         | 432/10682 [03:37<1:24:37,  2.02it/s]
  4%|▍         | 433/10682 [03:37<1:24:35,  2.02it/s]
  4%|▍         | 434/10682 [03:38<1:24:32,  2.02it/s]
  4%|▍         | 435/10682 [03:38<1:24:39,  2.02it/s]
  4%|▍         | 436/10682 [03:39<1:24:39,  2.02it/s]
  4%|▍         | 437/10682 [03:39<1:24:38,  2.02it/s]
  4%|▍         | 438/10682 [03:40<1:24:42,  2.02it/s]
  4%|▍         | 439/10682 [03:40<1:24:33,  2.02it/s]
  4%|▍         | 440/10682 [03:41<1:24:33,  2.02it/s]
  4%|▍         | 441/10682 [03:41<1:24:35,  2.02it/s]
  4%|▍         | 442/10682 [03:42<1:24:41,  2.02it/s]
  4%|▍         | 443/10682 [03:42<1:24:37,  2.02it/s]
  4%|▍         | 444/10682 [03:43<1:24:38,  2.02it/s]
  4%|▍         | 445/10682 [03:43<1:24:39,  2.02it/s]
  4%|▍         | 446/10682 [03:44<1:24:34,  2.02it/s]
  4%|▍         | 447/10682 [03:44<1:24:35,  2.02it/s]
  4%|▍         | 448/10682 [03:45<1:24:36,  2.02it/s]
  4%|▍         | 449/10682 [03:45<1:24:42,  2.01it/s]
  4%|▍         | 450/10682 [03:46<1:24:38,  2.01it/s]
                                                     {'loss': 5.2521, 'grad_norm': 0.6177807450294495, 'learning_rate': 0.00042095416276894293, 'epoch': 0.04}
-
  4%|▍         | 450/10682 [03:46<1:24:38,  2.01it/s]
  4%|▍         | 451/10682 [03:46<1:24:42,  2.01it/s]
  4%|▍         | 452/10682 [03:47<1:24:36,  2.02it/s]
  4%|▍         | 453/10682 [03:47<1:24:33,  2.02it/s]
  4%|▍         | 454/10682 [03:48<1:24:36,  2.01it/s]
  4%|▍         | 455/10682 [03:48<1:24:35,  2.01it/s]
  4%|▍         | 456/10682 [03:49<1:24:35,  2.01it/s]
  4%|▍         | 457/10682 [03:49<1:24:33,  2.02it/s]
  4%|▍         | 458/10682 [03:50<1:24:34,  2.01it/s]
  4%|▍         | 459/10682 [03:50<1:24:28,  2.02it/s]
  4%|▍         | 460/10682 [03:51<1:24:31,  2.02it/s]
  4%|▍         | 461/10682 [03:51<1:24:32,  2.02it/s]
  4%|▍         | 462/10682 [03:52<1:24:33,  2.01it/s]
  4%|▍         | 463/10682 [03:52<1:24:32,  2.01it/s]
  4%|▍         | 464/10682 [03:53<1:24:31,  2.01it/s]
  4%|▍         | 465/10682 [03:53<1:24:26,  2.02it/s]
  4%|▍         | 466/10682 [03:54<1:24:27,  2.02it/s]
  4%|▍         | 467/10682 [03:54<1:24:32,  2.01it/s]
  4%|▍         | 468/10682 [03:55<1:24:26,  2.02it/s]
  4%|▍         | 469/10682 [03:55<1:24:27,  2.02it/s]
  4%|▍         | 470/10682 [03:56<1:24:19,  2.02it/s]
  4%|▍         | 471/10682 [03:56<1:24:25,  2.02it/s]
  4%|▍         | 472/10682 [03:57<1:24:20,  2.02it/s]
  4%|▍         | 473/10682 [03:57<1:24:26,  2.01it/s]
  4%|▍         | 474/10682 [03:58<1:24:20,  2.02it/s]
  4%|▍         | 475/10682 [03:58<1:24:20,  2.02it/s]
                                                     
{'loss': 5.1791, 'grad_norm': 0.5205311179161072, 'learning_rate': 0.0004443405051449954, 'epoch': 0.04}
-
  4%|▍         | 475/10682 [03:58<1:24:20,  2.02it/s]
  4%|▍         | 476/10682 [03:59<1:24:31,  2.01it/s]
  4%|▍         | 477/10682 [03:59<1:24:21,  2.02it/s]
  4%|▍         | 478/10682 [04:00<1:24:18,  2.02it/s]
  4%|▍         | 479/10682 [04:00<1:24:10,  2.02it/s]
  4%|▍         | 480/10682 [04:01<1:24:17,  2.02it/s]
  5%|▍         | 481/10682 [04:01<1:24:13,  2.02it/s]
  5%|▍         | 482/10682 [04:02<1:24:17,  2.02it/s]
  5%|▍         | 483/10682 [04:02<1:24:20,  2.02it/s]
  5%|▍         | 484/10682 [04:02<1:24:20,  2.02it/s]
  5%|▍         | 485/10682 [04:03<1:24:14,  2.02it/s]
  5%|▍         | 486/10682 [04:03<1:24:09,  2.02it/s]
  5%|▍         | 487/10682 [04:04<1:24:09,  2.02it/s]
  5%|▍         | 488/10682 [04:04<1:24:06,  2.02it/s]
  5%|▍         | 489/10682 [04:05<1:24:07,  2.02it/s]
  5%|▍         | 490/10682 [04:05<1:24:09,  2.02it/s]
  5%|▍         | 491/10682 [04:06<1:24:10,  2.02it/s]
  5%|▍         | 492/10682 [04:06<1:24:05,  2.02it/s]
  5%|▍         | 493/10682 [04:07<1:24:08,  2.02it/s]
  5%|▍         | 494/10682 [04:07<1:24:03,  2.02it/s]
  5%|▍         | 495/10682 [04:08<1:24:08,  2.02it/s]
  5%|▍         | 496/10682 [04:08<1:23:58,  2.02it/s]
  5%|▍         | 497/10682 [04:09<1:24:05,  2.02it/s]
  5%|▍         | 498/10682 [04:09<1:23:58,  2.02it/s]
  5%|▍         | 499/10682 [04:10<1:24:06,  2.02it/s]
  5%|▍         | 500/10682 [04:10<1:24:07,  2.02it/s]
                                                     
{'loss': 5.1148, 'grad_norm': 0.6353130340576172, 'learning_rate': 0.0004677268475210477, 'epoch': 0.05}
-
  5%|▍         | 500/10682 [04:10<1:24:07,  2.02it/s]
  5%|▍         | 501/10682 [04:11<1:24:11,  2.02it/s]
  5%|▍         | 502/10682 [04:11<1:24:12,  2.01it/s]
  5%|▍         | 503/10682 [04:12<1:24:06,  2.02it/s]
  5%|▍         | 504/10682 [04:12<1:24:04,  2.02it/s]
  5%|▍         | 505/10682 [04:13<1:24:00,  2.02it/s]
  5%|▍         | 506/10682 [04:13<1:23:59,  2.02it/s]
  5%|▍         | 507/10682 [04:14<1:23:58,  2.02it/s]
  5%|▍         | 508/10682 [04:14<1:24:05,  2.02it/s]
  5%|▍         | 509/10682 [04:15<1:24:04,  2.02it/s]
  5%|▍         | 510/10682 [04:15<1:24:03,  2.02it/s]
  5%|▍         | 511/10682 [04:16<1:24:12,  2.01it/s]
  5%|▍         | 512/10682 [04:16<1:24:07,  2.01it/s]
  5%|▍         | 513/10682 [04:17<1:24:05,  2.02it/s]
  5%|▍         | 514/10682 [04:17<1:24:00,  2.02it/s]
  5%|▍         | 515/10682 [04:18<1:24:01,  2.02it/s]
  5%|▍         | 516/10682 [04:18<1:23:56,  2.02it/s]
  5%|▍         | 517/10682 [04:19<1:23:58,  2.02it/s]
  5%|▍         | 518/10682 [04:19<1:23:56,  2.02it/s]
  5%|▍         | 519/10682 [04:20<1:23:58,  2.02it/s]
  5%|▍         | 520/10682 [04:20<1:24:00,  2.02it/s]
  5%|▍         | 521/10682 [04:21<1:23:58,  2.02it/s]
  5%|▍         | 522/10682 [04:21<1:23:52,  2.02it/s]
  5%|▍         | 523/10682 [04:22<1:23:53,  2.02it/s]
  5%|▍         | 524/10682 [04:22<1:23:47,  2.02it/s]
  5%|▍         | 525/10682 [04:23<1:23:52,  2.02it/s]{'loss': 5.0582, 'grad_norm': 0.47787991166114807, 'learning_rate': 0.0004911131898971, 'epoch': 0.05}
                                                     
-
  5%|▍         | 525/10682 [04:23<1:23:52,  2.02it/s]
  5%|▍         | 526/10682 [04:23<1:24:10,  2.01it/s]
  5%|▍         | 527/10682 [04:24<1:24:06,  2.01it/s]
  5%|▍         | 528/10682 [04:24<1:24:03,  2.01it/s]
  5%|▍         | 529/10682 [04:25<1:23:55,  2.02it/s]
  5%|▍         | 530/10682 [04:25<1:23:51,  2.02it/s]
  5%|▍         | 531/10682 [04:26<1:23:45,  2.02it/s]
  5%|▍         | 532/10682 [04:26<1:23:45,  2.02it/s]
  5%|▍         | 533/10682 [04:27<1:23:40,  2.02it/s]
  5%|▍         | 534/10682 [04:27<1:23:41,  2.02it/s]
  5%|▌         | 535/10682 [04:28<1:23:36,  2.02it/s]
  5%|▌         | 536/10682 [04:28<1:23:39,  2.02it/s]
  5%|▌         | 537/10682 [04:29<1:23:41,  2.02it/s]
  5%|▌         | 538/10682 [04:29<1:23:45,  2.02it/s]
  5%|▌         | 539/10682 [04:30<1:23:50,  2.02it/s]
  5%|▌         | 540/10682 [04:30<1:23:47,  2.02it/s]
  5%|▌         | 541/10682 [04:31<1:23:43,  2.02it/s]
  5%|▌         | 542/10682 [04:31<1:23:40,  2.02it/s]
  5%|▌         | 543/10682 [04:32<1:23:40,  2.02it/s]
  5%|▌         | 544/10682 [04:32<1:23:39,  2.02it/s]
  5%|▌         | 545/10682 [04:33<1:23:39,  2.02it/s]
  5%|▌         | 546/10682 [04:33<1:23:32,  2.02it/s]
  5%|▌         | 547/10682 [04:34<1:23:35,  2.02it/s]
  5%|▌         | 548/10682 [04:34<1:23:36,  2.02it/s]
  5%|▌         | 549/10682 [04:35<1:23:36,  2.02it/s]
  5%|▌         | 550/10682 [04:35<1:23:33,  2.02it/s]
                                                     
{'loss': 5.005, 'grad_norm': 0.4891902506351471, 'learning_rate': 0.0005144995322731525, 'epoch': 0.05}
-
  5%|▌         | 550/10682 [04:35<1:23:33,  2.02it/s]
  5%|▌         | 551/10682 [04:36<1:23:40,  2.02it/s]
  5%|▌         | 552/10682 [04:36<1:23:36,  2.02it/s]
  5%|▌         | 553/10682 [04:37<1:23:35,  2.02it/s]
  5%|▌         | 554/10682 [04:37<1:23:33,  2.02it/s]
  5%|▌         | 555/10682 [04:38<1:23:32,  2.02it/s]
  5%|▌         | 556/10682 [04:38<1:23:37,  2.02it/s]
  5%|▌         | 557/10682 [04:39<1:23:39,  2.02it/s]
  5%|▌         | 558/10682 [04:39<1:23:34,  2.02it/s]
  5%|▌         | 559/10682 [04:40<1:23:29,  2.02it/s]
  5%|▌         | 560/10682 [04:40<1:23:22,  2.02it/s]
  5%|▌         | 561/10682 [04:41<1:23:27,  2.02it/s]
  5%|▌         | 562/10682 [04:41<1:23:19,  2.02it/s]
  5%|▌         | 563/10682 [04:42<1:23:24,  2.02it/s]
  5%|▌         | 564/10682 [04:42<1:23:19,  2.02it/s]
  5%|▌         | 565/10682 [04:43<1:23:24,  2.02it/s]
  5%|▌         | 566/10682 [04:43<1:23:24,  2.02it/s]
  5%|▌         | 567/10682 [04:44<1:23:27,  2.02it/s]
  5%|▌         | 568/10682 [04:44<1:23:21,  2.02it/s]
  5%|▌         | 569/10682 [04:45<1:23:23,  2.02it/s]
  5%|▌         | 570/10682 [04:45<1:23:27,  2.02it/s]
  5%|▌         | 571/10682 [04:46<1:23:26,  2.02it/s]
  5%|▌         | 572/10682 [04:46<1:23:25,  2.02it/s]
  5%|▌         | 573/10682 [04:47<1:23:24,  2.02it/s]
  5%|▌         | 574/10682 [04:47<1:23:20,  2.02it/s]
  5%|▌         | 575/10682 [04:48<1:23:35,  2.02it/s]
                                                     {'loss': 4.9542, 'grad_norm': 0.6182771921157837, 'learning_rate': 0.0005378858746492049, 'epoch': 0.05}
-

  5%|▌         | 575/10682 [04:48<1:23:35,  2.02it/s]
  5%|▌         | 576/10682 [04:48<1:23:47,  2.01it/s]
  5%|▌         | 577/10682 [04:49<1:23:35,  2.01it/s]
  5%|▌         | 578/10682 [04:49<1:23:37,  2.01it/s]
  5%|▌         | 579/10682 [04:50<1:23:26,  2.02it/s]
  5%|▌         | 580/10682 [04:50<1:23:24,  2.02it/s]
  5%|▌         | 581/10682 [04:51<1:23:17,  2.02it/s]
  5%|▌         | 582/10682 [04:51<1:23:21,  2.02it/s]
  5%|▌         | 583/10682 [04:52<1:23:28,  2.02it/s]
  5%|▌         | 584/10682 [04:52<1:23:24,  2.02it/s]
  5%|▌         | 585/10682 [04:53<1:23:23,  2.02it/s]
  5%|▌         | 586/10682 [04:53<1:23:26,  2.02it/s]
  5%|▌         | 587/10682 [04:54<1:23:23,  2.02it/s]
  6%|▌         | 588/10682 [04:54<1:23:17,  2.02it/s]
  6%|▌         | 589/10682 [04:55<1:23:25,  2.02it/s]
  6%|▌         | 590/10682 [04:55<1:23:17,  2.02it/s]
  6%|▌         | 591/10682 [04:56<1:23:19,  2.02it/s]
  6%|▌         | 592/10682 [04:56<1:23:20,  2.02it/s]
  6%|▌         | 593/10682 [04:56<1:23:14,  2.02it/s]
  6%|▌         | 594/10682 [04:57<1:23:14,  2.02it/s]
  6%|▌         | 595/10682 [04:57<1:23:23,  2.02it/s]
  6%|▌         | 596/10682 [04:58<1:23:17,  2.02it/s]
  6%|▌         | 597/10682 [04:58<1:23:17,  2.02it/s]
  6%|▌         | 598/10682 [04:59<1:23:13,  2.02it/s]
  6%|▌         | 599/10682 [04:59<1:23:12,  2.02it/s]
  6%|▌         | 600/10682 [05:00<1:23:10,  2.02it/s]{'loss': 4.9009, 'grad_norm': 0.635250449180603, 'learning_rate': 0.0005612722170252572, 'epoch': 0.06}
-
                                                     

  6%|▌         | 600/10682 [05:00<1:23:10,  2.02it/s]
  6%|▌         | 601/10682 [05:00<1:23:19,  2.02it/s]
  6%|▌         | 602/10682 [05:01<1:23:15,  2.02it/s]
  6%|▌         | 603/10682 [05:01<1:23:14,  2.02it/s]
  6%|▌         | 604/10682 [05:02<1:23:09,  2.02it/s]
  6%|▌         | 605/10682 [05:02<1:23:12,  2.02it/s]
  6%|▌         | 606/10682 [05:03<1:23:14,  2.02it/s]
  6%|▌         | 607/10682 [05:03<1:23:15,  2.02it/s]
  6%|▌         | 608/10682 [05:04<1:23:11,  2.02it/s]
  6%|▌         | 609/10682 [05:04<1:23:11,  2.02it/s]
  6%|▌         | 610/10682 [05:05<1:23:05,  2.02it/s]
  6%|▌         | 611/10682 [05:05<1:23:00,  2.02it/s]
  6%|▌         | 612/10682 [05:06<1:23:07,  2.02it/s]
  6%|▌         | 613/10682 [05:06<1:23:08,  2.02it/s]
  6%|▌         | 614/10682 [05:07<1:23:05,  2.02it/s]
  6%|▌         | 615/10682 [05:07<1:23:03,  2.02it/s]
  6%|▌         | 616/10682 [05:08<1:23:00,  2.02it/s]
  6%|▌         | 617/10682 [05:08<1:23:02,  2.02it/s]
  6%|▌         | 618/10682 [05:09<1:23:00,  2.02it/s]
  6%|▌         | 619/10682 [05:09<1:23:06,  2.02it/s]
  6%|▌         | 620/10682 [05:10<1:23:01,  2.02it/s]
  6%|▌         | 621/10682 [05:10<1:23:09,  2.02it/s]
  6%|▌         | 622/10682 [05:11<1:22:57,  2.02it/s]
  6%|▌         | 623/10682 [05:11<1:23:06,  2.02it/s]
  6%|▌         | 624/10682 [05:12<1:23:08,  2.02it/s]
  6%|▌         | 625/10682 [05:12<1:23:05,  2.02it/s]
                                                     {'loss': 4.8538, 'grad_norm': 0.6464534997940063, 'learning_rate': 0.0005846585594013096, 'epoch': 0.06}
-
  6%|▌         | 625/10682 [05:12<1:23:05,  2.02it/s]
  6%|▌         | 626/10682 [05:13<1:23:10,  2.01it/s]
  6%|▌         | 627/10682 [05:13<1:23:05,  2.02it/s]
  6%|▌         | 628/10682 [05:14<1:23:02,  2.02it/s]
  6%|▌         | 629/10682 [05:14<1:23:02,  2.02it/s]
  6%|▌         | 630/10682 [05:15<1:23:11,  2.01it/s]
  6%|▌         | 631/10682 [05:15<1:23:08,  2.02it/s]
  6%|▌         | 632/10682 [05:16<1:23:02,  2.02it/s]
  6%|▌         | 633/10682 [05:16<1:22:58,  2.02it/s]
  6%|▌         | 634/10682 [05:17<1:23:09,  2.01it/s]
  6%|▌         | 635/10682 [05:17<1:23:02,  2.02it/s]
  6%|▌         | 636/10682 [05:18<1:22:58,  2.02it/s]
  6%|▌         | 637/10682 [05:18<1:22:55,  2.02it/s]
  6%|▌         | 638/10682 [05:19<1:22:57,  2.02it/s]
  6%|▌         | 639/10682 [05:19<1:22:51,  2.02it/s]
  6%|▌         | 640/10682 [05:20<1:22:53,  2.02it/s]
  6%|▌         | 641/10682 [05:20<1:23:00,  2.02it/s]
  6%|▌         | 642/10682 [05:21<1:22:57,  2.02it/s]
  6%|▌         | 643/10682 [05:21<1:22:51,  2.02it/s]
  6%|▌         | 644/10682 [05:22<1:22:50,  2.02it/s]
  6%|▌         | 645/10682 [05:22<1:22:45,  2.02it/s]
  6%|▌         | 646/10682 [05:23<1:22:48,  2.02it/s]
  6%|▌         | 647/10682 [05:23<1:22:43,  2.02it/s]
  6%|▌         | 648/10682 [05:24<1:22:46,  2.02it/s]
  6%|▌         | 649/10682 [05:24<1:22:44,  2.02it/s]
  6%|▌         | 650/10682 [05:25<1:22:45,  2.02it/s]{'loss': 4.8238, 'grad_norm': 0.5110419392585754, 'learning_rate': 0.0006080449017773621, 'epoch': 0.06}
                                                     
-
  6%|▌         | 650/10682 [05:25<1:22:45,  2.02it/s]
  6%|▌         | 651/10682 [05:25<1:22:53,  2.02it/s]
  6%|▌         | 652/10682 [05:26<1:22:59,  2.01it/s]
  6%|▌         | 653/10682 [05:26<1:22:59,  2.01it/s]
  6%|▌         | 654/10682 [05:27<1:22:57,  2.01it/s]
  6%|▌         | 655/10682 [05:27<1:22:53,  2.02it/s]
  6%|▌         | 656/10682 [05:28<1:22:45,  2.02it/s]
  6%|▌         | 657/10682 [05:28<1:22:39,  2.02it/s]
  6%|▌         | 658/10682 [05:29<1:22:36,  2.02it/s]
  6%|▌         | 659/10682 [05:29<1:22:48,  2.02it/s]
  6%|▌         | 660/10682 [05:30<1:22:45,  2.02it/s]
  6%|▌         | 661/10682 [05:30<1:22:44,  2.02it/s]
  6%|▌         | 662/10682 [05:31<1:22:37,  2.02it/s]
  6%|▌         | 663/10682 [05:31<1:22:37,  2.02it/s]
  6%|▌         | 664/10682 [05:32<1:22:39,  2.02it/s]
  6%|▌         | 665/10682 [05:32<1:22:40,  2.02it/s]
  6%|▌         | 666/10682 [05:33<1:22:42,  2.02it/s]
  6%|▌         | 667/10682 [05:33<1:22:40,  2.02it/s]
  6%|▋         | 668/10682 [05:34<1:22:40,  2.02it/s]
  6%|▋         | 669/10682 [05:34<1:22:43,  2.02it/s]
  6%|▋         | 670/10682 [05:35<1:22:43,  2.02it/s]
  6%|▋         | 671/10682 [05:35<1:22:39,  2.02it/s]
  6%|▋         | 672/10682 [05:36<1:22:44,  2.02it/s]
  6%|▋         | 673/10682 [05:36<1:22:42,  2.02it/s]
  6%|▋         | 674/10682 [05:37<1:22:42,  2.02it/s]
  6%|▋         | 675/10682 [05:37<1:22:34,  2.02it/s]
                                                     {'loss': 4.7865, 'grad_norm': 0.45274272561073303, 'learning_rate': 0.0006314312441534145, 'epoch': 0.06}
-
  6%|▋         | 675/10682 [05:37<1:22:34,  2.02it/s]
  6%|▋         | 676/10682 [05:38<1:22:42,  2.02it/s]
  6%|▋         | 677/10682 [05:38<1:22:34,  2.02it/s]
  6%|▋         | 678/10682 [05:39<1:22:38,  2.02it/s]
  6%|▋         | 679/10682 [05:39<1:22:32,  2.02it/s]
  6%|▋         | 680/10682 [05:40<1:22:37,  2.02it/s]
  6%|▋         | 681/10682 [05:40<1:22:36,  2.02it/s]
  6%|▋         | 682/10682 [05:41<1:22:32,  2.02it/s]
  6%|▋         | 683/10682 [05:41<1:22:31,  2.02it/s]
  6%|▋         | 684/10682 [05:42<1:22:32,  2.02it/s]
  6%|▋         | 685/10682 [05:42<1:22:33,  2.02it/s]
  6%|▋         | 686/10682 [05:43<1:22:28,  2.02it/s]
  6%|▋         | 687/10682 [05:43<1:22:27,  2.02it/s]
  6%|▋         | 688/10682 [05:44<1:22:26,  2.02it/s]
  6%|▋         | 689/10682 [05:44<1:22:30,  2.02it/s]
  6%|▋         | 690/10682 [05:45<1:22:24,  2.02it/s]
  6%|▋         | 691/10682 [05:45<1:22:26,  2.02it/s]
  6%|▋         | 692/10682 [05:46<1:22:22,  2.02it/s]
  6%|▋         | 693/10682 [05:46<1:22:25,  2.02it/s]
  6%|▋         | 694/10682 [05:47<1:22:21,  2.02it/s]
  7%|▋         | 695/10682 [05:47<1:22:23,  2.02it/s]
  7%|▋         | 696/10682 [05:48<1:22:21,  2.02it/s]
  7%|▋         | 697/10682 [05:48<1:22:28,  2.02it/s]
  7%|▋         | 698/10682 [05:49<1:22:24,  2.02it/s]
  7%|▋         | 699/10682 [05:49<1:22:26,  2.02it/s]
  7%|▋         | 700/10682 [05:50<1:22:22,  2.02it/s]
                                                     
{'loss': 4.747, 'grad_norm': 0.4373401701450348, 'learning_rate': 0.0006548175865294667, 'epoch': 0.07}
-
  7%|▋         | 700/10682 [05:50<1:22:22,  2.02it/s]
  7%|▋         | 701/10682 [05:50<1:22:28,  2.02it/s]
  7%|▋         | 702/10682 [05:50<1:22:25,  2.02it/s]
  7%|▋         | 703/10682 [05:51<1:22:18,  2.02it/s]
  7%|▋         | 704/10682 [05:51<1:22:19,  2.02it/s]
  7%|▋         | 705/10682 [05:52<1:22:15,  2.02it/s]
  7%|▋         | 706/10682 [05:52<1:22:22,  2.02it/s]
  7%|▋         | 707/10682 [05:53<1:22:15,  2.02it/s]
  7%|▋         | 708/10682 [05:53<1:22:19,  2.02it/s]
  7%|▋         | 709/10682 [05:54<1:22:16,  2.02it/s]
  7%|▋         | 710/10682 [05:54<1:22:27,  2.02it/s]
  7%|▋         | 711/10682 [05:55<1:22:19,  2.02it/s]
  7%|▋         | 712/10682 [05:55<1:22:20,  2.02it/s]
  7%|▋         | 713/10682 [05:56<1:22:15,  2.02it/s]
  7%|▋         | 714/10682 [05:56<1:22:16,  2.02it/s]
  7%|▋         | 715/10682 [05:57<1:22:10,  2.02it/s]
  7%|▋         | 716/10682 [05:57<1:22:19,  2.02it/s]
  7%|▋         | 717/10682 [05:58<1:22:25,  2.01it/s]
  7%|▋         | 718/10682 [05:58<1:22:26,  2.01it/s]
  7%|▋         | 719/10682 [05:59<1:22:36,  2.01it/s]
  7%|▋         | 720/10682 [05:59<1:22:31,  2.01it/s]
  7%|▋         | 721/10682 [06:00<1:22:29,  2.01it/s]
  7%|▋         | 722/10682 [06:00<1:22:20,  2.02it/s]
  7%|▋         | 723/10682 [06:01<1:22:21,  2.02it/s]
  7%|▋         | 724/10682 [06:01<1:22:15,  2.02it/s]
  7%|▋         | 725/10682 [06:02<1:22:17,  2.02it/s]
                                                     {'loss': 4.7143, 'grad_norm': 0.3921816945075989, 'learning_rate': 0.0006782039289055192, 'epoch': 0.07}
-
  7%|▋         | 725/10682 [06:02<1:22:17,  2.02it/s]
  7%|▋         | 726/10682 [06:02<1:22:35,  2.01it/s]
  7%|▋         | 727/10682 [06:03<1:22:26,  2.01it/s]
  7%|▋         | 728/10682 [06:03<1:22:21,  2.01it/s]
  7%|▋         | 729/10682 [06:04<1:22:12,  2.02it/s]
  7%|▋         | 730/10682 [06:04<1:22:12,  2.02it/s]
  7%|▋         | 731/10682 [06:05<1:22:15,  2.02it/s]
  7%|▋         | 732/10682 [06:05<1:22:11,  2.02it/s]
  7%|▋         | 733/10682 [06:06<1:22:09,  2.02it/s]
  7%|▋         | 734/10682 [06:06<1:22:09,  2.02it/s]
  7%|▋         | 735/10682 [06:07<1:22:07,  2.02it/s]
  7%|▋         | 736/10682 [06:07<1:22:10,  2.02it/s]
  7%|▋         | 737/10682 [06:08<1:22:12,  2.02it/s]
  7%|▋         | 738/10682 [06:08<1:22:11,  2.02it/s]
  7%|▋         | 739/10682 [06:09<1:22:06,  2.02it/s]
  7%|▋         | 740/10682 [06:09<1:22:07,  2.02it/s]
  7%|▋         | 741/10682 [06:10<1:22:03,  2.02it/s]
  7%|▋         | 742/10682 [06:10<1:22:05,  2.02it/s]
  7%|▋         | 743/10682 [06:11<1:21:59,  2.02it/s]
  7%|▋         | 744/10682 [06:11<1:22:08,  2.02it/s]
  7%|▋         | 745/10682 [06:12<1:22:00,  2.02it/s]
  7%|▋         | 746/10682 [06:12<1:22:11,  2.01it/s]
  7%|▋         | 747/10682 [06:13<1:22:13,  2.01it/s]
  7%|▋         | 748/10682 [06:13<1:22:12,  2.01it/s]
  7%|▋         | 749/10682 [06:14<1:22:11,  2.01it/s]
  7%|▋         | 750/10682 [06:14<1:22:07,  2.02it/s]
                                                     {'loss': 4.6826, 'grad_norm': 0.45835670828819275, 'learning_rate': 0.0007015902712815716, 'epoch': 0.07}
-
  7%|▋         | 750/10682 [06:14<1:22:07,  2.02it/s]
  7%|▋         | 751/10682 [06:15<1:22:16,  2.01it/s]
  7%|▋         | 752/10682 [06:15<1:22:12,  2.01it/s]
  7%|▋         | 753/10682 [06:16<1:22:08,  2.01it/s]
  7%|▋         | 754/10682 [06:16<1:22:12,  2.01it/s]
  7%|▋         | 755/10682 [06:17<1:22:03,  2.02it/s]
  7%|▋         | 756/10682 [06:17<1:22:04,  2.02it/s]
  7%|▋         | 757/10682 [06:18<1:21:59,  2.02it/s]
  7%|▋         | 758/10682 [06:18<1:22:01,  2.02it/s]
  7%|▋         | 759/10682 [06:19<1:21:53,  2.02it/s]
  7%|▋         | 760/10682 [06:19<1:21:58,  2.02it/s]
  7%|▋         | 761/10682 [06:20<1:21:50,  2.02it/s]
  7%|▋         | 762/10682 [06:20<1:21:53,  2.02it/s]
  7%|▋         | 763/10682 [06:21<1:22:00,  2.02it/s]
  7%|▋         | 764/10682 [06:21<1:21:59,  2.02it/s]
  7%|▋         | 765/10682 [06:22<1:22:02,  2.01it/s]
  7%|▋         | 766/10682 [06:22<1:21:59,  2.02it/s]
  7%|▋         | 767/10682 [06:23<1:22:00,  2.02it/s]
  7%|▋         | 768/10682 [06:23<1:21:55,  2.02it/s]
  7%|▋         | 769/10682 [06:24<1:21:55,  2.02it/s]
  7%|▋         | 770/10682 [06:24<1:21:50,  2.02it/s]
  7%|▋         | 771/10682 [06:25<1:21:50,  2.02it/s]
  7%|▋         | 772/10682 [06:25<1:21:53,  2.02it/s]
  7%|▋         | 773/10682 [06:26<1:21:54,  2.02it/s]
  7%|▋         | 774/10682 [06:26<1:21:55,  2.02it/s]
  7%|▋         | 775/10682 [06:27<1:22:01,  2.01it/s]{'loss': 4.6574, 'grad_norm': 0.40098676085472107, 'learning_rate': 0.0007249766136576241, 'epoch': 0.07}
                                                     
-
  7%|▋         | 775/10682 [06:27<1:22:01,  2.01it/s]
  7%|▋         | 776/10682 [06:27<1:22:01,  2.01it/s]
  7%|▋         | 777/10682 [06:28<1:21:56,  2.01it/s]
  7%|▋         | 778/10682 [06:28<1:21:53,  2.02it/s]
  7%|▋         | 779/10682 [06:29<1:21:53,  2.02it/s]
  7%|▋         | 780/10682 [06:29<1:21:48,  2.02it/s]
  7%|▋         | 781/10682 [06:30<1:21:50,  2.02it/s]
  7%|▋         | 782/10682 [06:30<1:21:48,  2.02it/s]
  7%|▋         | 783/10682 [06:31<1:21:50,  2.02it/s]
  7%|▋         | 784/10682 [06:31<1:21:59,  2.01it/s]
  7%|▋         | 785/10682 [06:32<1:21:56,  2.01it/s]
  7%|▋         | 786/10682 [06:32<1:21:47,  2.02it/s]
  7%|▋         | 787/10682 [06:33<1:21:52,  2.01it/s]
  7%|▋         | 788/10682 [06:33<1:21:45,  2.02it/s]
  7%|▋         | 789/10682 [06:34<1:21:45,  2.02it/s]
  7%|▋         | 790/10682 [06:34<1:21:38,  2.02it/s]
  7%|▋         | 791/10682 [06:35<1:21:50,  2.01it/s]
  7%|▋         | 792/10682 [06:35<1:21:39,  2.02it/s]
  7%|▋         | 793/10682 [06:36<1:21:45,  2.02it/s]
  7%|▋         | 794/10682 [06:36<1:21:37,  2.02it/s]
  7%|▋         | 795/10682 [06:37<1:21:37,  2.02it/s]
  7%|▋         | 796/10682 [06:37<1:21:31,  2.02it/s]
  7%|▋         | 797/10682 [06:38<1:21:39,  2.02it/s]
  7%|▋         | 798/10682 [06:38<1:21:34,  2.02it/s]
  7%|▋         | 799/10682 [06:39<1:21:43,  2.02it/s]
  7%|▋         | 800/10682 [06:39<1:21:37,  2.02it/s]{'loss': 4.622, 'grad_norm': 0.40164539217948914, 'learning_rate': 0.0007483629560336763, 'epoch': 0.07}
                                                     
-
  7%|▋         | 800/10682 [06:39<1:21:37,  2.02it/s]
  7%|▋         | 801/10682 [06:40<1:21:47,  2.01it/s]
  8%|▊         | 802/10682 [06:40<1:21:49,  2.01it/s]
  8%|▊         | 803/10682 [06:41<1:21:38,  2.02it/s]
  8%|▊         | 804/10682 [06:41<1:21:41,  2.02it/s]
  8%|▊         | 805/10682 [06:42<1:21:38,  2.02it/s]
  8%|▊         | 806/10682 [06:42<1:21:39,  2.02it/s]
  8%|▊         | 807/10682 [06:43<1:21:29,  2.02it/s]
  8%|▊         | 808/10682 [06:43<1:21:34,  2.02it/s]
  8%|▊         | 809/10682 [06:44<1:21:32,  2.02it/s]
  8%|▊         | 810/10682 [06:44<1:21:30,  2.02it/s]
  8%|▊         | 811/10682 [06:45<1:21:28,  2.02it/s]
  8%|▊         | 812/10682 [06:45<1:21:30,  2.02it/s]
  8%|▊         | 813/10682 [06:46<1:21:32,  2.02it/s]
  8%|▊         | 814/10682 [06:46<1:21:27,  2.02it/s]
  8%|▊         | 815/10682 [06:47<1:21:25,  2.02it/s]
  8%|▊         | 816/10682 [06:47<1:21:24,  2.02it/s]
  8%|▊         | 817/10682 [06:48<1:21:33,  2.02it/s]
  8%|▊         | 818/10682 [06:48<1:21:38,  2.01it/s]
  8%|▊         | 819/10682 [06:49<1:21:43,  2.01it/s]
  8%|▊         | 820/10682 [06:49<1:21:42,  2.01it/s]
  8%|▊         | 821/10682 [06:50<1:21:32,  2.02it/s]
  8%|▊         | 822/10682 [06:50<1:21:26,  2.02it/s]
  8%|▊         | 823/10682 [06:50<1:21:26,  2.02it/s]
  8%|▊         | 824/10682 [06:51<1:21:26,  2.02it/s]
  8%|▊         | 825/10682 [06:51<1:21:23,  2.02it/s]
                                                     
{'loss': 4.5943, 'grad_norm': 0.43268200755119324, 'learning_rate': 0.0007717492984097287, 'epoch': 0.08}
-
  8%|▊         | 825/10682 [06:51<1:21:23,  2.02it/s]
  8%|▊         | 826/10682 [06:52<1:21:34,  2.01it/s]
  8%|▊         | 827/10682 [06:52<1:21:26,  2.02it/s]
  8%|▊         | 828/10682 [06:53<1:21:25,  2.02it/s]
  8%|▊         | 829/10682 [06:53<1:21:16,  2.02it/s]
  8%|▊         | 830/10682 [06:54<1:21:24,  2.02it/s]
  8%|▊         | 831/10682 [06:54<1:21:31,  2.01it/s]
  8%|▊         | 832/10682 [06:55<1:21:28,  2.01it/s]
  8%|▊         | 833/10682 [06:55<1:21:35,  2.01it/s]
  8%|▊         | 834/10682 [06:56<1:21:33,  2.01it/s]
  8%|▊         | 835/10682 [06:56<1:21:36,  2.01it/s]
  8%|▊         | 836/10682 [06:57<1:21:35,  2.01it/s]
  8%|▊         | 837/10682 [06:57<1:21:27,  2.01it/s]
  8%|▊         | 838/10682 [06:58<1:21:27,  2.01it/s]
  8%|▊         | 839/10682 [06:58<1:21:25,  2.01it/s]
  8%|▊         | 840/10682 [06:59<1:21:26,  2.01it/s]
  8%|▊         | 841/10682 [06:59<1:21:20,  2.02it/s]
  8%|▊         | 842/10682 [07:00<1:21:20,  2.02it/s]
  8%|▊         | 843/10682 [07:00<1:21:19,  2.02it/s]
  8%|▊         | 844/10682 [07:01<1:21:17,  2.02it/s]
  8%|▊         | 845/10682 [07:01<1:21:17,  2.02it/s]
  8%|▊         | 846/10682 [07:02<1:21:12,  2.02it/s]
  8%|▊         | 847/10682 [07:02<1:21:15,  2.02it/s]
  8%|▊         | 848/10682 [07:03<1:21:10,  2.02it/s]
  8%|▊         | 849/10682 [07:03<1:21:10,  2.02it/s]
  8%|▊         | 850/10682 [07:04<1:21:08,  2.02it/s]
                                                     
{'loss': 4.5736, 'grad_norm': 0.4063767194747925, 'learning_rate': 0.0007951356407857812, 'epoch': 0.08}
-
  8%|▊         | 850/10682 [07:04<1:21:08,  2.02it/s]
  8%|▊         | 851/10682 [07:04<1:21:15,  2.02it/s]
  8%|▊         | 852/10682 [07:05<1:21:17,  2.02it/s]
  8%|▊         | 853/10682 [07:05<1:21:12,  2.02it/s]
  8%|▊         | 854/10682 [07:06<1:21:07,  2.02it/s]
  8%|▊         | 855/10682 [07:06<1:21:08,  2.02it/s]
  8%|▊         | 856/10682 [07:07<1:21:07,  2.02it/s]
  8%|▊         | 857/10682 [07:07<1:21:07,  2.02it/s]
  8%|▊         | 858/10682 [07:08<1:21:05,  2.02it/s]
  8%|▊         | 859/10682 [07:08<1:21:05,  2.02it/s]
  8%|▊         | 860/10682 [07:09<1:21:06,  2.02it/s]
  8%|▊         | 861/10682 [07:09<1:21:10,  2.02it/s]
  8%|▊         | 862/10682 [07:10<1:21:09,  2.02it/s]
  8%|▊         | 863/10682 [07:10<1:21:04,  2.02it/s]
  8%|▊         | 864/10682 [07:11<1:21:10,  2.02it/s]
  8%|▊         | 865/10682 [07:11<1:21:05,  2.02it/s]
  8%|▊         | 866/10682 [07:12<1:21:10,  2.02it/s]
  8%|▊         | 867/10682 [07:12<1:21:06,  2.02it/s]
  8%|▊         | 868/10682 [07:13<1:21:05,  2.02it/s]
  8%|▊         | 869/10682 [07:13<1:20:58,  2.02it/s]
  8%|▊         | 870/10682 [07:14<1:20:54,  2.02it/s]
  8%|▊         | 871/10682 [07:14<1:20:54,  2.02it/s]
  8%|▊         | 872/10682 [07:15<1:20:59,  2.02it/s]
  8%|▊         | 873/10682 [07:15<1:21:01,  2.02it/s]
  8%|▊         | 874/10682 [07:16<1:21:01,  2.02it/s]
  8%|▊         | 875/10682 [07:16<1:20:55,  2.02it/s]
                                                     
{'loss': 4.5502, 'grad_norm': 0.3670870363712311, 'learning_rate': 0.0008185219831618334, 'epoch': 0.08}
-
  8%|▊         | 875/10682 [07:16<1:20:55,  2.02it/s]
  8%|▊         | 876/10682 [07:17<1:21:00,  2.02it/s]
  8%|▊         | 877/10682 [07:17<1:20:56,  2.02it/s]
  8%|▊         | 878/10682 [07:18<1:20:59,  2.02it/s]
  8%|▊         | 879/10682 [07:18<1:20:50,  2.02it/s]
  8%|▊         | 880/10682 [07:19<1:20:51,  2.02it/s]
  8%|▊         | 881/10682 [07:19<1:20:49,  2.02it/s]
  8%|▊         | 882/10682 [07:20<1:20:45,  2.02it/s]
  8%|▊         | 883/10682 [07:20<1:20:49,  2.02it/s]
  8%|▊         | 884/10682 [07:21<1:20:48,  2.02it/s]
  8%|▊         | 885/10682 [07:21<1:20:50,  2.02it/s]
  8%|▊         | 886/10682 [07:22<1:20:54,  2.02it/s]
  8%|▊         | 887/10682 [07:22<1:20:49,  2.02it/s]
  8%|▊         | 888/10682 [07:23<1:20:56,  2.02it/s]
  8%|▊         | 889/10682 [07:23<1:20:47,  2.02it/s]
  8%|▊         | 890/10682 [07:24<1:20:49,  2.02it/s]
  8%|▊         | 891/10682 [07:24<1:20:45,  2.02it/s]
  8%|▊         | 892/10682 [07:25<1:20:45,  2.02it/s]
  8%|▊         | 893/10682 [07:25<1:20:48,  2.02it/s]
  8%|▊         | 894/10682 [07:26<1:20:50,  2.02it/s]
  8%|▊         | 895/10682 [07:26<1:20:47,  2.02it/s]
  8%|▊         | 896/10682 [07:27<1:20:51,  2.02it/s]
  8%|▊         | 897/10682 [07:27<1:20:47,  2.02it/s]
  8%|▊         | 898/10682 [07:28<1:20:46,  2.02it/s]
  8%|▊         | 899/10682 [07:28<1:20:45,  2.02it/s]
  8%|▊         | 900/10682 [07:29<1:20:45,  2.02it/s]
                                                     {'loss': 4.5241, 'grad_norm': 0.4166002869606018, 'learning_rate': 0.0008419083255378859, 'epoch': 0.08}
-
  8%|▊         | 900/10682 [07:29<1:20:45,  2.02it/s]
  8%|▊         | 901/10682 [07:29<1:20:54,  2.01it/s]
  8%|▊         | 902/10682 [07:30<1:20:50,  2.02it/s]
  8%|▊         | 903/10682 [07:30<1:20:49,  2.02it/s]
  8%|▊         | 904/10682 [07:31<1:20:50,  2.02it/s]
  8%|▊         | 905/10682 [07:31<1:20:53,  2.01it/s]
  8%|▊         | 906/10682 [07:32<1:20:49,  2.02it/s]
  8%|▊         | 907/10682 [07:32<1:20:47,  2.02it/s]
  9%|▊         | 908/10682 [07:33<1:20:43,  2.02it/s]
  9%|▊         | 909/10682 [07:33<1:20:43,  2.02it/s]
  9%|▊         | 910/10682 [07:34<1:20:39,  2.02it/s]
  9%|▊         | 911/10682 [07:34<1:20:41,  2.02it/s]
  9%|▊         | 912/10682 [07:35<1:20:44,  2.02it/s]
  9%|▊         | 913/10682 [07:35<1:20:47,  2.02it/s]
  9%|▊         | 914/10682 [07:36<1:20:46,  2.02it/s]
  9%|▊         | 915/10682 [07:36<1:20:44,  2.02it/s]
  9%|▊         | 916/10682 [07:37<1:20:37,  2.02it/s]
  9%|▊         | 917/10682 [07:37<1:20:40,  2.02it/s]
  9%|▊         | 918/10682 [07:38<1:20:39,  2.02it/s]
  9%|▊         | 919/10682 [07:38<1:20:38,  2.02it/s]
  9%|▊         | 920/10682 [07:39<1:20:29,  2.02it/s]
  9%|▊         | 921/10682 [07:39<1:20:32,  2.02it/s]
  9%|▊         | 922/10682 [07:40<1:20:27,  2.02it/s]
  9%|▊         | 923/10682 [07:40<1:20:30,  2.02it/s]
  9%|▊         | 924/10682 [07:41<1:20:38,  2.02it/s]
  9%|▊         | 925/10682 [07:41<1:20:31,  2.02it/s]{'loss': 4.4973, 'grad_norm': 0.3467422127723694, 'learning_rate': 0.0008652946679139383, 'epoch': 0.09}
-
                                                     

  9%|▊         | 925/10682 [07:41<1:20:31,  2.02it/s]
  9%|▊         | 926/10682 [07:42<1:20:35,  2.02it/s]
  9%|▊         | 927/10682 [07:42<1:20:34,  2.02it/s]
  9%|▊         | 928/10682 [07:43<1:20:33,  2.02it/s]
  9%|▊         | 929/10682 [07:43<1:20:30,  2.02it/s]
  9%|▊         | 930/10682 [07:44<1:20:31,  2.02it/s]
  9%|▊         | 931/10682 [07:44<1:20:24,  2.02it/s]
  9%|▊         | 932/10682 [07:45<1:20:29,  2.02it/s]
  9%|▊         | 933/10682 [07:45<1:20:22,  2.02it/s]
  9%|▊         | 934/10682 [07:46<1:20:27,  2.02it/s]
  9%|▉         | 935/10682 [07:46<1:20:23,  2.02it/s]
  9%|▉         | 936/10682 [07:46<1:20:30,  2.02it/s]
  9%|▉         | 937/10682 [07:47<1:20:24,  2.02it/s]
  9%|▉         | 938/10682 [07:47<1:20:25,  2.02it/s]
  9%|▉         | 939/10682 [07:48<1:20:20,  2.02it/s]
  9%|▉         | 940/10682 [07:48<1:20:25,  2.02it/s]
  9%|▉         | 941/10682 [07:49<1:20:20,  2.02it/s]
  9%|▉         | 942/10682 [07:49<1:20:24,  2.02it/s]
  9%|▉         | 943/10682 [07:50<1:20:21,  2.02it/s]
  9%|▉         | 944/10682 [07:50<1:20:21,  2.02it/s]
  9%|▉         | 945/10682 [07:51<1:20:16,  2.02it/s]
  9%|▉         | 946/10682 [07:51<1:20:18,  2.02it/s]
  9%|▉         | 947/10682 [07:52<1:20:15,  2.02it/s]
  9%|▉         | 948/10682 [07:52<1:20:20,  2.02it/s]
  9%|▉         | 949/10682 [07:53<1:20:18,  2.02it/s]
  9%|▉         | 950/10682 [07:53<1:20:17,  2.02it/s]{'loss': 4.4828, 'grad_norm': 0.34193843603134155, 'learning_rate': 0.0008886810102899908, 'epoch': 0.09}
                                                     
-
  9%|▉         | 950/10682 [07:53<1:20:17,  2.02it/s]
  9%|▉         | 951/10682 [07:54<1:20:27,  2.02it/s]
  9%|▉         | 952/10682 [07:54<1:20:24,  2.02it/s]
  9%|▉         | 953/10682 [07:55<1:20:22,  2.02it/s]
  9%|▉         | 954/10682 [07:55<1:20:19,  2.02it/s]
  9%|▉         | 955/10682 [07:56<1:20:17,  2.02it/s]
  9%|▉         | 956/10682 [07:56<1:20:16,  2.02it/s]
  9%|▉         | 957/10682 [07:57<1:20:15,  2.02it/s]
  9%|▉         | 958/10682 [07:57<1:20:18,  2.02it/s]
  9%|▉         | 959/10682 [07:58<1:20:16,  2.02it/s]
  9%|▉         | 960/10682 [07:58<1:20:15,  2.02it/s]
  9%|▉         | 961/10682 [07:59<1:20:22,  2.02it/s]
  9%|▉         | 962/10682 [07:59<1:20:18,  2.02it/s]
  9%|▉         | 963/10682 [08:00<1:20:14,  2.02it/s]
  9%|▉         | 964/10682 [08:00<1:20:10,  2.02it/s]
  9%|▉         | 965/10682 [08:01<1:20:11,  2.02it/s]
  9%|▉         | 966/10682 [08:01<1:20:11,  2.02it/s]
  9%|▉         | 967/10682 [08:02<1:20:09,  2.02it/s]
  9%|▉         | 968/10682 [08:02<1:20:16,  2.02it/s]
  9%|▉         | 969/10682 [08:03<1:20:10,  2.02it/s]
  9%|▉         | 970/10682 [08:03<1:20:11,  2.02it/s]
  9%|▉         | 971/10682 [08:04<1:20:05,  2.02it/s]
  9%|▉         | 972/10682 [08:04<1:20:05,  2.02it/s]
  9%|▉         | 973/10682 [08:05<1:20:08,  2.02it/s]
  9%|▉         | 974/10682 [08:05<1:20:07,  2.02it/s]
  9%|▉         | 975/10682 [08:06<1:20:00,  2.02it/s]
                                                     {'loss': 4.4593, 'grad_norm': 0.34431901574134827, 'learning_rate': 0.000912067352666043, 'epoch': 0.09}
-
  9%|▉         | 975/10682 [08:06<1:20:00,  2.02it/s]
  9%|▉         | 976/10682 [08:06<1:20:12,  2.02it/s]
  9%|▉         | 977/10682 [08:07<1:20:05,  2.02it/s]
  9%|▉         | 978/10682 [08:07<1:20:11,  2.02it/s]
  9%|▉         | 979/10682 [08:08<1:20:05,  2.02it/s]
  9%|▉         | 980/10682 [08:08<1:20:05,  2.02it/s]
  9%|▉         | 981/10682 [08:09<1:20:01,  2.02it/s]
  9%|▉         | 982/10682 [08:09<1:20:07,  2.02it/s]
  9%|▉         | 983/10682 [08:10<1:20:06,  2.02it/s]
  9%|▉         | 984/10682 [08:10<1:20:02,  2.02it/s]
  9%|▉         | 985/10682 [08:11<1:20:11,  2.02it/s]
  9%|▉         | 986/10682 [08:11<1:20:02,  2.02it/s]
  9%|▉         | 987/10682 [08:12<1:20:01,  2.02it/s]
  9%|▉         | 988/10682 [08:12<1:19:57,  2.02it/s]
  9%|▉         | 989/10682 [08:13<1:19:57,  2.02it/s]
  9%|▉         | 990/10682 [08:13<1:19:52,  2.02it/s]
  9%|▉         | 991/10682 [08:14<1:20:01,  2.02it/s]
  9%|▉         | 992/10682 [08:14<1:19:56,  2.02it/s]
  9%|▉         | 993/10682 [08:15<1:20:03,  2.02it/s]
  9%|▉         | 994/10682 [08:15<1:20:00,  2.02it/s]
  9%|▉         | 995/10682 [08:16<1:19:59,  2.02it/s]
  9%|▉         | 996/10682 [08:16<1:19:52,  2.02it/s]
  9%|▉         | 997/10682 [08:17<1:19:53,  2.02it/s]
  9%|▉         | 998/10682 [08:17<1:19:49,  2.02it/s]
  9%|▉         | 999/10682 [08:18<1:19:51,  2.02it/s]
  9%|▉         | 1000/10682 [08:18<1:19:50,  2.02it/s]{'loss': 4.4359, 'grad_norm': 0.3537678122520447, 'learning_rate': 0.0009354536950420954, 'epoch': 0.09}
                                                      
-
  9%|▉         | 1000/10682 [08:18<1:19:50,  2.02it/s]
  9%|▉         | 1001/10682 [08:19<1:19:59,  2.02it/s]
  9%|▉         | 1002/10682 [08:19<1:19:56,  2.02it/s]
  9%|▉         | 1003/10682 [08:20<1:20:00,  2.02it/s]
  9%|▉         | 1004/10682 [08:20<1:19:52,  2.02it/s]
  9%|▉         | 1005/10682 [08:21<1:19:52,  2.02it/s]
  9%|▉         | 1006/10682 [08:21<1:19:49,  2.02it/s]
  9%|▉         | 1007/10682 [08:22<1:19:53,  2.02it/s]
  9%|▉         | 1008/10682 [08:22<1:19:57,  2.02it/s]
  9%|▉         | 1009/10682 [08:23<1:19:57,  2.02it/s]
  9%|▉         | 1010/10682 [08:23<1:19:47,  2.02it/s]
  9%|▉         | 1011/10682 [08:24<1:19:50,  2.02it/s]
  9%|▉         | 1012/10682 [08:24<1:19:49,  2.02it/s]
  9%|▉         | 1013/10682 [08:25<1:19:52,  2.02it/s]
  9%|▉         | 1014/10682 [08:25<1:19:52,  2.02it/s]
 10%|▉         | 1015/10682 [08:26<1:19:53,  2.02it/s]
 10%|▉         | 1016/10682 [08:26<1:19:49,  2.02it/s]
 10%|▉         | 1017/10682 [08:27<1:19:48,  2.02it/s]
 10%|▉         | 1018/10682 [08:27<1:19:50,  2.02it/s]
 10%|▉         | 1019/10682 [08:28<1:19:49,  2.02it/s]
 10%|▉         | 1020/10682 [08:28<1:19:47,  2.02it/s]
 10%|▉         | 1021/10682 [08:29<1:19:45,  2.02it/s]
 10%|▉         | 1022/10682 [08:29<1:19:38,  2.02it/s]
 10%|▉         | 1023/10682 [08:30<1:19:44,  2.02it/s]
 10%|▉         | 1024/10682 [08:30<1:19:40,  2.02it/s]
 10%|▉         | 1025/10682 [08:31<1:19:42,  2.02it/s]
                                                      {'loss': 4.41, 'grad_norm': 0.3315429091453552, 'learning_rate': 0.0009588400374181479, 'epoch': 0.1}
-
 10%|▉         | 1025/10682 [08:31<1:19:42,  2.02it/s]
 10%|▉         | 1026/10682 [08:31<1:19:48,  2.02it/s]
 10%|▉         | 1027/10682 [08:32<1:19:42,  2.02it/s]
 10%|▉         | 1028/10682 [08:32<1:19:47,  2.02it/s]
 10%|▉         | 1029/10682 [08:33<1:19:41,  2.02it/s]
 10%|▉         | 1030/10682 [08:33<1:19:47,  2.02it/s]
 10%|▉         | 1031/10682 [08:34<1:19:42,  2.02it/s]
 10%|▉         | 1032/10682 [08:34<1:19:43,  2.02it/s]
 10%|▉         | 1033/10682 [08:35<1:19:38,  2.02it/s]
 10%|▉         | 1034/10682 [08:35<1:19:39,  2.02it/s]
 10%|▉         | 1035/10682 [08:36<1:19:33,  2.02it/s]
 10%|▉         | 1036/10682 [08:36<1:19:35,  2.02it/s]
 10%|▉         | 1037/10682 [08:37<1:19:34,  2.02it/s]
 10%|▉         | 1038/10682 [08:37<1:19:34,  2.02it/s]
 10%|▉         | 1039/10682 [08:38<1:19:30,  2.02it/s]
 10%|▉         | 1040/10682 [08:38<1:19:33,  2.02it/s]
 10%|▉         | 1041/10682 [08:39<1:19:33,  2.02it/s]
 10%|▉         | 1042/10682 [08:39<1:32:39,  1.73it/s]
 10%|▉         | 1043/10682 [08:40<1:28:40,  1.81it/s]
 10%|▉         | 1044/10682 [08:40<1:25:56,  1.87it/s]
 10%|▉         | 1045/10682 [08:41<1:24:02,  1.91it/s]
 10%|▉         | 1046/10682 [08:41<1:22:40,  1.94it/s]
 10%|▉         | 1047/10682 [08:42<1:21:49,  1.96it/s]
 10%|▉         | 1048/10682 [08:43<1:33:59,  1.71it/s]
 10%|▉         | 1049/10682 [08:43<1:29:39,  1.79it/s]
 10%|▉         | 1050/10682 [08:44<1:26:41,  1.85it/s]
                                                      
{'loss': 4.3985, 'grad_norm': 0.3065096437931061, 'learning_rate': 0.0009822263797942, 'epoch': 0.1}
-
 10%|▉         | 1050/10682 [08:44<1:26:41,  1.85it/s]
 10%|▉         | 1051/10682 [08:44<1:24:38,  1.90it/s]
 10%|▉         | 1052/10682 [08:44<1:23:05,  1.93it/s]
 10%|▉         | 1053/10682 [08:45<1:21:54,  1.96it/s]
 10%|▉         | 1054/10682 [08:45<1:21:08,  1.98it/s]
 10%|▉         | 1055/10682 [08:46<1:20:37,  1.99it/s]
 10%|▉         | 1056/10682 [08:46<1:20:13,  2.00it/s]
 10%|▉         | 1057/10682 [08:47<1:20:02,  2.00it/s]
 10%|▉         | 1058/10682 [08:47<1:19:42,  2.01it/s]
 10%|▉         | 1059/10682 [08:48<1:19:40,  2.01it/s]
 10%|▉         | 1060/10682 [08:48<1:19:30,  2.02it/s]
 10%|▉         | 1061/10682 [08:49<1:19:39,  2.01it/s]
 10%|▉         | 1062/10682 [08:49<1:19:27,  2.02it/s]
 10%|▉         | 1063/10682 [08:50<1:19:27,  2.02it/s]
 10%|▉         | 1064/10682 [08:50<1:19:21,  2.02it/s]
 10%|▉         | 1065/10682 [08:51<1:19:19,  2.02it/s]
 10%|▉         | 1066/10682 [08:51<1:19:23,  2.02it/s]
 10%|▉         | 1067/10682 [08:52<1:19:22,  2.02it/s]
 10%|▉         | 1068/10682 [08:52<1:19:22,  2.02it/s]
 10%|█         | 1069/10682 [08:53<1:19:30,  2.02it/s]
 10%|█         | 1070/10682 [08:53<1:19:24,  2.02it/s]
 10%|█         | 1071/10682 [08:54<1:19:23,  2.02it/s]
 10%|█         | 1072/10682 [08:54<1:19:19,  2.02it/s]
 10%|█         | 1073/10682 [08:55<1:19:20,  2.02it/s]
 10%|█         | 1074/10682 [08:55<1:19:18,  2.02it/s]
 10%|█         | 1075/10682 [08:56<1:19:19,  2.02it/s]{'loss': 4.3728, 'grad_norm': 0.31585821509361267, 'learning_rate': 0.0009999990387768293, 'epoch': 0.1}
                                                      
-
 10%|█         | 1075/10682 [08:56<1:19:19,  2.02it/s]
 10%|█         | 1076/10682 [08:56<1:19:21,  2.02it/s]
 10%|█         | 1077/10682 [08:57<1:19:25,  2.02it/s]
 10%|█         | 1078/10682 [08:57<1:19:24,  2.02it/s]
 10%|█         | 1079/10682 [08:58<1:19:22,  2.02it/s]
 10%|█         | 1080/10682 [08:58<1:19:21,  2.02it/s]
 10%|█         | 1081/10682 [08:59<1:19:16,  2.02it/s]
 10%|█         | 1082/10682 [08:59<1:19:15,  2.02it/s]
 10%|█         | 1083/10682 [09:00<1:19:09,  2.02it/s]
 10%|█         | 1084/10682 [09:00<1:19:13,  2.02it/s]
 10%|█         | 1085/10682 [09:01<1:19:17,  2.02it/s]
 10%|█         | 1086/10682 [09:01<1:19:12,  2.02it/s]
 10%|█         | 1087/10682 [09:02<1:19:14,  2.02it/s]
 10%|█         | 1088/10682 [09:02<1:19:13,  2.02it/s]
 10%|█         | 1089/10682 [09:03<1:19:06,  2.02it/s]
 10%|█         | 1090/10682 [09:03<1:19:08,  2.02it/s]
 10%|█         | 1091/10682 [09:04<1:19:04,  2.02it/s]
 10%|█         | 1092/10682 [09:04<1:19:08,  2.02it/s]
 10%|█         | 1093/10682 [09:05<1:18:58,  2.02it/s]
 10%|█         | 1094/10682 [09:05<1:19:07,  2.02it/s]
 10%|█         | 1095/10682 [09:06<1:19:00,  2.02it/s]
 10%|█         | 1096/10682 [09:06<1:19:07,  2.02it/s]
 10%|█         | 1097/10682 [09:07<1:19:04,  2.02it/s]
 10%|█         | 1098/10682 [09:07<1:19:07,  2.02it/s]
 10%|█         | 1099/10682 [09:08<1:19:04,  2.02it/s]
 10%|█         | 1100/10682 [09:08<1:19:02,  2.02it/s]
                                                      
{'loss': 4.3563, 'grad_norm': 0.30548256635665894, 'learning_rate': 0.0009999743408927195, 'epoch': 0.1}
-
 10%|█         | 1100/10682 [09:08<1:19:02,  2.02it/s]
 10%|█         | 1101/10682 [09:09<1:19:10,  2.02it/s]
 10%|█         | 1102/10682 [09:09<1:19:11,  2.02it/s]
 10%|█         | 1103/10682 [09:10<1:19:02,  2.02it/s]
 10%|█         | 1104/10682 [09:10<1:19:05,  2.02it/s]
 10%|█         | 1105/10682 [09:11<1:19:01,  2.02it/s]
 10%|█         | 1106/10682 [09:11<1:19:03,  2.02it/s]
 10%|█         | 1107/10682 [09:12<1:18:58,  2.02it/s]
 10%|█         | 1108/10682 [09:12<1:18:58,  2.02it/s]
 10%|█         | 1109/10682 [09:13<1:18:56,  2.02it/s]
 10%|█         | 1110/10682 [09:13<1:18:58,  2.02it/s]
 10%|█         | 1111/10682 [09:14<1:18:52,  2.02it/s]
 10%|█         | 1112/10682 [09:14<1:18:59,  2.02it/s]
 10%|█         | 1113/10682 [09:15<1:18:56,  2.02it/s]
 10%|█         | 1114/10682 [09:15<1:19:01,  2.02it/s]
 10%|█         | 1115/10682 [09:16<1:19:03,  2.02it/s]
 10%|█         | 1116/10682 [09:16<1:19:00,  2.02it/s]
 10%|█         | 1117/10682 [09:17<1:19:00,  2.02it/s]
 10%|█         | 1118/10682 [09:17<1:18:57,  2.02it/s]
 10%|█         | 1119/10682 [09:18<1:18:58,  2.02it/s]
 10%|█         | 1120/10682 [09:18<1:18:53,  2.02it/s]
 10%|█         | 1121/10682 [09:19<1:19:00,  2.02it/s]
 11%|█         | 1122/10682 [09:19<1:18:52,  2.02it/s]
 11%|█         | 1123/10682 [09:20<1:18:57,  2.02it/s]
 11%|█         | 1124/10682 [09:20<1:19:01,  2.02it/s]
 11%|█         | 1125/10682 [09:21<1:19:00,  2.02it/s]
                                                      
{'loss': 4.3264, 'grad_norm': 0.32466068863868713, 'learning_rate': 0.0009999162690918172, 'epoch': 0.11}
-
 11%|█         | 1125/10682 [09:21<1:19:00,  2.02it/s]
 11%|█         | 1126/10682 [09:21<1:19:11,  2.01it/s]
 11%|█         | 1127/10682 [09:22<1:19:05,  2.01it/s]
 11%|█         | 1128/10682 [09:22<1:19:03,  2.01it/s]
 11%|█         | 1129/10682 [09:23<1:18:57,  2.02it/s]
 11%|█         | 1130/10682 [09:23<1:18:50,  2.02it/s]
 11%|█         | 1131/10682 [09:24<1:18:47,  2.02it/s]
 11%|█         | 1132/10682 [09:24<1:18:46,  2.02it/s]
 11%|█         | 1133/10682 [09:25<1:18:47,  2.02it/s]
 11%|█         | 1134/10682 [09:25<1:18:40,  2.02it/s]
 11%|█         | 1135/10682 [09:26<1:18:45,  2.02it/s]
 11%|█         | 1136/10682 [09:26<1:18:42,  2.02it/s]
 11%|█         | 1137/10682 [09:27<1:18:44,  2.02it/s]
 11%|█         | 1138/10682 [09:27<1:18:39,  2.02it/s]
 11%|█         | 1139/10682 [09:28<1:18:44,  2.02it/s]
 11%|█         | 1140/10682 [09:28<1:18:39,  2.02it/s]
 11%|█         | 1141/10682 [09:29<1:18:47,  2.02it/s]
 11%|█         | 1142/10682 [09:29<1:18:44,  2.02it/s]
 11%|█         | 1143/10682 [09:30<1:18:43,  2.02it/s]
 11%|█         | 1144/10682 [09:30<1:18:40,  2.02it/s]
 11%|█         | 1145/10682 [09:31<1:18:43,  2.02it/s]
 11%|█         | 1146/10682 [09:31<1:18:41,  2.02it/s]
 11%|█         | 1147/10682 [09:32<1:18:41,  2.02it/s]
 11%|█         | 1148/10682 [09:32<1:18:41,  2.02it/s]
 11%|█         | 1149/10682 [09:33<1:18:46,  2.02it/s]
 11%|█         | 1150/10682 [09:33<1:18:46,  2.02it/s]{'loss': 4.3127, 'grad_norm': 0.2933719754219055, 'learning_rate': 0.0009998248272504882, 'epoch': 0.11}
                                                      
-
 11%|█         | 1150/10682 [09:33<1:18:46,  2.02it/s]
 11%|█         | 1151/10682 [09:34<1:18:51,  2.01it/s]
 11%|█         | 1152/10682 [09:34<1:18:56,  2.01it/s]
 11%|█         | 1153/10682 [09:35<1:18:46,  2.02it/s]
 11%|█         | 1154/10682 [09:35<1:18:38,  2.02it/s]
 11%|█         | 1155/10682 [09:36<1:18:37,  2.02it/s]
 11%|█         | 1156/10682 [09:36<1:18:44,  2.02it/s]
 11%|█         | 1157/10682 [09:37<1:18:37,  2.02it/s]
 11%|█         | 1158/10682 [09:37<1:18:38,  2.02it/s]
 11%|█         | 1159/10682 [09:37<1:18:43,  2.02it/s]
 11%|█         | 1160/10682 [09:38<1:18:45,  2.01it/s]
 11%|█         | 1161/10682 [09:38<1:18:40,  2.02it/s]
 11%|█         | 1162/10682 [09:39<1:18:39,  2.02it/s]
 11%|█         | 1163/10682 [09:39<1:18:45,  2.01it/s]
 11%|█         | 1164/10682 [09:40<1:18:41,  2.02it/s]
 11%|█         | 1165/10682 [09:40<1:18:40,  2.02it/s]
 11%|█         | 1166/10682 [09:41<1:18:38,  2.02it/s]
 11%|█         | 1167/10682 [09:41<1:18:40,  2.02it/s]
 11%|█         | 1168/10682 [09:42<1:18:37,  2.02it/s]
 11%|█         | 1169/10682 [09:42<1:18:38,  2.02it/s]
 11%|█         | 1170/10682 [09:43<1:18:36,  2.02it/s]
 11%|█         | 1171/10682 [09:43<1:18:31,  2.02it/s]
 11%|█         | 1172/10682 [09:44<1:18:28,  2.02it/s]
 11%|█         | 1173/10682 [09:44<1:18:33,  2.02it/s]
 11%|█         | 1174/10682 [09:45<1:18:30,  2.02it/s]
 11%|█         | 1175/10682 [09:45<1:18:28,  2.02it/s]{'loss': 4.2985, 'grad_norm': 0.2867088317871094, 'learning_rate': 0.0009997000214725905, 'epoch': 0.11}
                                                      
-
 11%|█         | 1175/10682 [09:45<1:18:28,  2.02it/s]
 11%|█         | 1176/10682 [09:46<1:18:46,  2.01it/s]
 11%|█         | 1177/10682 [09:46<1:18:37,  2.01it/s]
 11%|█         | 1178/10682 [09:47<1:18:36,  2.02it/s]
 11%|█         | 1179/10682 [09:47<1:18:29,  2.02it/s]
 11%|█         | 1180/10682 [09:48<1:18:29,  2.02it/s]
 11%|█         | 1181/10682 [09:48<1:18:27,  2.02it/s]
 11%|█         | 1182/10682 [09:49<1:18:25,  2.02it/s]
 11%|█         | 1183/10682 [09:49<1:18:25,  2.02it/s]
 11%|█         | 1184/10682 [09:50<1:18:25,  2.02it/s]
 11%|█         | 1185/10682 [09:50<1:18:17,  2.02it/s]
 11%|█         | 1186/10682 [09:51<1:18:27,  2.02it/s]
 11%|█         | 1187/10682 [09:51<1:18:20,  2.02it/s]
 11%|█         | 1188/10682 [09:52<1:18:22,  2.02it/s]
 11%|█         | 1189/10682 [09:52<1:18:17,  2.02it/s]
 11%|█         | 1190/10682 [09:53<1:18:15,  2.02it/s]
 11%|█         | 1191/10682 [09:53<1:18:17,  2.02it/s]
 11%|█         | 1192/10682 [09:54<1:18:31,  2.01it/s]
 11%|█         | 1193/10682 [09:54<1:18:27,  2.02it/s]
 11%|█         | 1194/10682 [09:55<1:18:25,  2.02it/s]
 11%|█         | 1195/10682 [09:55<1:18:24,  2.02it/s]
 11%|█         | 1196/10682 [09:56<1:18:20,  2.02it/s]
 11%|█         | 1197/10682 [09:56<1:18:17,  2.02it/s]
 11%|█         | 1198/10682 [09:57<1:18:11,  2.02it/s]
 11%|█         | 1199/10682 [09:57<1:18:12,  2.02it/s]
 11%|█         | 1200/10682 [09:58<1:18:07,  2.02it/s]{'loss': 4.2671, 'grad_norm': 0.2738240361213684, 'learning_rate': 0.000999541860089067, 'epoch': 0.11}
-
                                                      

 11%|█         | 1200/10682 [09:58<1:18:07,  2.02it/s]
 11%|█         | 1201/10682 [09:58<1:18:19,  2.02it/s]
 11%|█▏        | 1202/10682 [09:59<1:18:14,  2.02it/s]
 11%|█▏        | 1203/10682 [09:59<1:18:18,  2.02it/s]
 11%|█▏        | 1204/10682 [10:00<1:18:19,  2.02it/s]
 11%|█▏        | 1205/10682 [10:00<1:18:18,  2.02it/s]
 11%|█▏        | 1206/10682 [10:01<1:18:15,  2.02it/s]
 11%|█▏        | 1207/10682 [10:01<1:18:17,  2.02it/s]
 11%|█▏        | 1208/10682 [10:02<1:18:17,  2.02it/s]
 11%|█▏        | 1209/10682 [10:02<1:18:12,  2.02it/s]
 11%|█▏        | 1210/10682 [10:03<1:18:09,  2.02it/s]
 11%|█▏        | 1211/10682 [10:03<1:18:05,  2.02it/s]
 11%|█▏        | 1212/10682 [10:04<1:18:04,  2.02it/s]
 11%|█▏        | 1213/10682 [10:04<1:18:03,  2.02it/s]
 11%|█▏        | 1214/10682 [10:05<1:18:02,  2.02it/s]
 11%|█▏        | 1215/10682 [10:05<1:18:03,  2.02it/s]
 11%|█▏        | 1216/10682 [10:06<1:17:58,  2.02it/s]
 11%|█▏        | 1217/10682 [10:06<1:18:04,  2.02it/s]
 11%|█▏        | 1218/10682 [10:07<1:18:02,  2.02it/s]
 11%|█▏        | 1219/10682 [10:07<1:18:04,  2.02it/s]
 11%|█▏        | 1220/10682 [10:08<1:18:05,  2.02it/s]
 11%|█▏        | 1221/10682 [10:08<1:18:03,  2.02it/s]
 11%|█▏        | 1222/10682 [10:09<1:18:03,  2.02it/s]
 11%|█▏        | 1223/10682 [10:09<1:18:04,  2.02it/s]
 11%|█▏        | 1224/10682 [10:10<1:18:07,  2.02it/s]
 11%|█▏        | 1225/10682 [10:10<1:18:06,  2.02it/s]
                                                      
{'loss': 4.2427, 'grad_norm': 0.288103312253952, 'learning_rate': 0.0009993503536573894, 'epoch': 0.11}
-
 11%|█▏        | 1225/10682 [10:10<1:18:06,  2.02it/s]
 11%|█▏        | 1226/10682 [10:11<1:18:13,  2.01it/s]
 11%|█▏        | 1227/10682 [10:11<1:18:07,  2.02it/s]
 11%|█▏        | 1228/10682 [10:12<1:18:10,  2.02it/s]
 12%|█▏        | 1229/10682 [10:12<1:18:05,  2.02it/s]
 12%|█▏        | 1230/10682 [10:13<1:18:06,  2.02it/s]
 12%|█▏        | 1231/10682 [10:13<1:18:03,  2.02it/s]
 12%|█▏        | 1232/10682 [10:14<1:18:07,  2.02it/s]
 12%|█▏        | 1233/10682 [10:14<1:18:06,  2.02it/s]
 12%|█▏        | 1234/10682 [10:15<1:18:07,  2.02it/s]
 12%|█▏        | 1235/10682 [10:15<1:18:13,  2.01it/s]
 12%|█▏        | 1236/10682 [10:16<1:18:07,  2.02it/s]
 12%|█▏        | 1237/10682 [10:16<1:18:06,  2.02it/s]
 12%|█▏        | 1238/10682 [10:17<1:18:05,  2.02it/s]
 12%|█▏        | 1239/10682 [10:17<1:18:04,  2.02it/s]
 12%|█▏        | 1240/10682 [10:18<1:18:01,  2.02it/s]
 12%|█▏        | 1241/10682 [10:18<1:17:57,  2.02it/s]
 12%|█▏        | 1242/10682 [10:19<1:18:01,  2.02it/s]
 12%|█▏        | 1243/10682 [10:19<1:17:54,  2.02it/s]
 12%|█▏        | 1244/10682 [10:20<1:18:04,  2.01it/s]
 12%|█▏        | 1245/10682 [10:20<1:17:59,  2.02it/s]
 12%|█▏        | 1246/10682 [10:21<1:17:56,  2.02it/s]
 12%|█▏        | 1247/10682 [10:21<1:17:52,  2.02it/s]
 12%|█▏        | 1248/10682 [10:22<1:17:53,  2.02it/s]
 12%|█▏        | 1249/10682 [10:22<1:17:48,  2.02it/s]
 12%|█▏        | 1250/10682 [10:23<1:17:50,  2.02it/s]{'loss': 4.2362, 'grad_norm': 0.2934815287590027, 'learning_rate': 0.0009991255149608528, 'epoch': 0.12}
                                                      
-
 12%|█▏        | 1250/10682 [10:23<1:17:50,  2.02it/s]
 12%|█▏        | 1251/10682 [10:23<1:17:56,  2.02it/s]
 12%|█▏        | 1252/10682 [10:24<1:17:53,  2.02it/s]
 12%|█▏        | 1253/10682 [10:24<1:17:54,  2.02it/s]
 12%|█▏        | 1254/10682 [10:25<1:17:54,  2.02it/s]
 12%|█▏        | 1255/10682 [10:25<1:17:57,  2.02it/s]
 12%|█▏        | 1256/10682 [10:26<1:17:51,  2.02it/s]
 12%|█▏        | 1257/10682 [10:26<1:17:57,  2.01it/s]
 12%|█▏        | 1258/10682 [10:27<1:17:55,  2.02it/s]
 12%|█▏        | 1259/10682 [10:27<1:17:50,  2.02it/s]
 12%|█▏        | 1260/10682 [10:28<1:17:49,  2.02it/s]
 12%|█▏        | 1261/10682 [10:28<1:17:47,  2.02it/s]
 12%|█▏        | 1262/10682 [10:29<1:17:49,  2.02it/s]
 12%|█▏        | 1263/10682 [10:29<1:17:48,  2.02it/s]
 12%|█▏        | 1264/10682 [10:30<1:17:50,  2.02it/s]
 12%|█▏        | 1265/10682 [10:30<1:17:50,  2.02it/s]
 12%|█▏        | 1266/10682 [10:31<1:17:50,  2.02it/s]
 12%|█▏        | 1267/10682 [10:31<1:17:48,  2.02it/s]
 12%|█▏        | 1268/10682 [10:32<1:17:48,  2.02it/s]
 12%|█▏        | 1269/10682 [10:32<1:17:40,  2.02it/s]
 12%|█▏        | 1270/10682 [10:33<1:17:44,  2.02it/s]
 12%|█▏        | 1271/10682 [10:33<1:17:42,  2.02it/s]
 12%|█▏        | 1272/10682 [10:33<1:17:43,  2.02it/s]
 12%|█▏        | 1273/10682 [10:34<1:17:44,  2.02it/s]
 12%|█▏        | 1274/10682 [10:34<1:17:46,  2.02it/s]
 12%|█▏        | 1275/10682 [10:35<1:17:42,  2.02it/s]
                                                      {'loss': 4.2214, 'grad_norm': 0.27766427397727966, 'learning_rate': 0.0009988673590077238, 'epoch': 0.12}
-

 12%|█▏        | 1275/10682 [10:35<1:17:42,  2.02it/s]
 12%|█▏        | 1276/10682 [10:35<1:17:49,  2.01it/s]
 12%|█▏        | 1277/10682 [10:36<1:17:46,  2.02it/s]
 12%|█▏        | 1278/10682 [10:36<1:17:47,  2.01it/s]
 12%|█▏        | 1279/10682 [10:37<1:17:44,  2.02it/s]
 12%|█▏        | 1280/10682 [10:37<1:17:37,  2.02it/s]
 12%|█▏        | 1281/10682 [10:38<1:17:35,  2.02it/s]
 12%|█▏        | 1282/10682 [10:38<1:17:34,  2.02it/s]
 12%|█▏        | 1283/10682 [10:39<1:17:35,  2.02it/s]
 12%|█▏        | 1284/10682 [10:39<1:17:33,  2.02it/s]
 12%|█▏        | 1285/10682 [10:40<1:17:42,  2.02it/s]
 12%|█▏        | 1286/10682 [10:40<1:17:37,  2.02it/s]
 12%|█▏        | 1287/10682 [10:41<1:17:33,  2.02it/s]
 12%|█▏        | 1288/10682 [10:41<1:17:29,  2.02it/s]
 12%|█▏        | 1289/10682 [10:42<1:17:29,  2.02it/s]
 12%|█▏        | 1290/10682 [10:42<1:17:25,  2.02it/s]
 12%|█▏        | 1291/10682 [10:43<1:17:25,  2.02it/s]
 12%|█▏        | 1292/10682 [10:43<1:17:26,  2.02it/s]
 12%|█▏        | 1293/10682 [10:44<1:17:29,  2.02it/s]
 12%|█▏        | 1294/10682 [10:44<1:17:23,  2.02it/s]
 12%|█▏        | 1295/10682 [10:45<1:17:26,  2.02it/s]
 12%|█▏        | 1296/10682 [10:45<1:17:23,  2.02it/s]
 12%|█▏        | 1297/10682 [10:46<1:17:24,  2.02it/s]
 12%|█▏        | 1298/10682 [10:46<1:17:22,  2.02it/s]
 12%|█▏        | 1299/10682 [10:47<1:17:24,  2.02it/s]
 12%|█▏        | 1300/10682 [10:47<1:17:22,  2.02it/s]{'loss': 4.2098, 'grad_norm': 0.2935454249382019, 'learning_rate': 0.0009985759030302366, 'epoch': 0.12}
                                                      
-
 12%|█▏        | 1300/10682 [10:47<1:17:22,  2.02it/s]
 12%|█▏        | 1301/10682 [10:48<1:17:33,  2.02it/s]
 12%|█▏        | 1302/10682 [10:48<1:17:24,  2.02it/s]
 12%|█▏        | 1303/10682 [10:49<1:17:30,  2.02it/s]
 12%|█▏        | 1304/10682 [10:49<1:17:29,  2.02it/s]
 12%|█▏        | 1305/10682 [10:50<1:17:32,  2.02it/s]
 12%|█▏        | 1306/10682 [10:50<1:17:32,  2.02it/s]
 12%|█▏        | 1307/10682 [10:51<1:17:24,  2.02it/s]
 12%|█▏        | 1308/10682 [10:51<1:17:31,  2.02it/s]
 12%|█▏        | 1309/10682 [10:52<1:17:22,  2.02it/s]
 12%|█▏        | 1310/10682 [10:52<1:17:23,  2.02it/s]
 12%|█▏        | 1311/10682 [10:53<1:17:22,  2.02it/s]
 12%|█▏        | 1312/10682 [10:53<1:17:28,  2.02it/s]
 12%|█▏        | 1313/10682 [10:54<1:17:35,  2.01it/s]
 12%|█▏        | 1314/10682 [10:54<1:17:32,  2.01it/s]
 12%|█▏        | 1315/10682 [10:55<1:17:28,  2.02it/s]
 12%|█▏        | 1316/10682 [10:55<1:17:23,  2.02it/s]
 12%|█▏        | 1317/10682 [10:56<1:17:19,  2.02it/s]
 12%|█▏        | 1318/10682 [10:56<1:17:18,  2.02it/s]
 12%|█▏        | 1319/10682 [10:57<1:17:15,  2.02it/s]
 12%|█▏        | 1320/10682 [10:57<1:17:15,  2.02it/s]
 12%|█▏        | 1321/10682 [10:58<1:17:15,  2.02it/s]
 12%|█▏        | 1322/10682 [10:58<1:17:15,  2.02it/s]
 12%|█▏        | 1323/10682 [10:59<1:17:18,  2.02it/s]
 12%|█▏        | 1324/10682 [10:59<1:17:18,  2.02it/s]
 12%|█▏        | 1325/10682 [11:00<1:17:12,  2.02it/s]{'loss': 4.1857, 'grad_norm': 0.28335994482040405, 'learning_rate': 0.0009982511664834452, 'epoch': 0.12}
                                                      
-
 12%|█▏        | 1325/10682 [11:00<1:17:12,  2.02it/s]
 12%|█▏        | 1326/10682 [11:00<1:17:24,  2.01it/s]
 12%|█▏        | 1327/10682 [11:01<1:17:19,  2.02it/s]
 12%|█▏        | 1328/10682 [11:01<1:17:19,  2.02it/s]
 12%|█▏        | 1329/10682 [11:02<1:17:10,  2.02it/s]
 12%|█▏        | 1330/10682 [11:02<1:17:14,  2.02it/s]
 12%|█▏        | 1331/10682 [11:03<1:17:07,  2.02it/s]
 12%|█▏        | 1332/10682 [11:03<1:17:08,  2.02it/s]
 12%|█▏        | 1333/10682 [11:04<1:17:12,  2.02it/s]
 12%|█▏        | 1334/10682 [11:04<1:17:10,  2.02it/s]
 12%|█▏        | 1335/10682 [11:05<1:17:14,  2.02it/s]
 13%|█▎        | 1336/10682 [11:05<1:17:14,  2.02it/s]
 13%|█▎        | 1337/10682 [11:06<1:17:14,  2.02it/s]
 13%|█▎        | 1338/10682 [11:06<1:17:13,  2.02it/s]
 13%|█▎        | 1339/10682 [11:07<1:17:12,  2.02it/s]
 13%|█▎        | 1340/10682 [11:07<1:17:19,  2.01it/s]
 13%|█▎        | 1341/10682 [11:08<1:17:24,  2.01it/s]
 13%|█▎        | 1342/10682 [11:08<1:17:18,  2.01it/s]
 13%|█▎        | 1343/10682 [11:09<1:17:17,  2.01it/s]
 13%|█▎        | 1344/10682 [11:09<1:17:13,  2.02it/s]
 13%|█▎        | 1345/10682 [11:10<1:17:12,  2.02it/s]
 13%|█▎        | 1346/10682 [11:10<1:17:08,  2.02it/s]
 13%|█▎        | 1347/10682 [11:11<1:17:11,  2.02it/s]
 13%|█▎        | 1348/10682 [11:11<1:17:09,  2.02it/s]
 13%|█▎        | 1349/10682 [11:12<1:17:35,  2.00it/s]
 13%|█▎        | 1350/10682 [11:12<1:17:32,  2.01it/s]{'loss': 4.1724, 'grad_norm': 0.2608835995197296, 'learning_rate': 0.0009978931710439229, 'epoch': 0.13}
                                                      
-
 13%|█▎        | 1350/10682 [11:12<1:17:32,  2.01it/s]
 13%|█▎        | 1351/10682 [11:13<1:17:41,  2.00it/s]
 13%|█▎        | 1352/10682 [11:13<1:17:29,  2.01it/s]
 13%|█▎        | 1353/10682 [11:14<1:17:21,  2.01it/s]
 13%|█▎        | 1354/10682 [11:14<1:17:16,  2.01it/s]
 13%|█▎        | 1355/10682 [11:15<1:17:06,  2.02it/s]
 13%|█▎        | 1356/10682 [11:15<1:17:10,  2.01it/s]
 13%|█▎        | 1357/10682 [11:16<1:17:11,  2.01it/s]
 13%|█▎        | 1358/10682 [11:16<1:17:12,  2.01it/s]
 13%|█▎        | 1359/10682 [11:17<1:17:04,  2.02it/s]
 13%|█▎        | 1360/10682 [11:17<1:17:01,  2.02it/s]
 13%|█▎        | 1361/10682 [11:18<1:17:00,  2.02it/s]
 13%|█▎        | 1362/10682 [11:18<1:16:54,  2.02it/s]
 13%|█▎        | 1363/10682 [11:19<1:17:03,  2.02it/s]
 13%|█▎        | 1364/10682 [11:19<1:16:55,  2.02it/s]
 13%|█▎        | 1365/10682 [11:20<1:17:00,  2.02it/s]
 13%|█▎        | 1366/10682 [11:20<1:16:57,  2.02it/s]
 13%|█▎        | 1367/10682 [11:21<1:17:01,  2.02it/s]
 13%|█▎        | 1368/10682 [11:21<1:16:58,  2.02it/s]
 13%|█▎        | 1369/10682 [11:22<1:16:56,  2.02it/s]
 13%|█▎        | 1370/10682 [11:22<1:16:53,  2.02it/s]
 13%|█▎        | 1371/10682 [11:23<1:16:53,  2.02it/s]
 13%|█▎        | 1372/10682 [11:23<1:16:58,  2.02it/s]
 13%|█▎        | 1373/10682 [11:24<1:16:52,  2.02it/s]
 13%|█▎        | 1374/10682 [11:24<1:16:54,  2.02it/s]
 13%|█▎        | 1375/10682 [11:25<1:16:47,  2.02it/s]{'loss': 4.1589, 'grad_norm': 0.2611101269721985, 'learning_rate': 0.000997501940608316, 'epoch': 0.13}
                                                      
-
 13%|█▎        | 1375/10682 [11:25<1:16:47,  2.02it/s]
 13%|█▎        | 1376/10682 [11:25<1:16:56,  2.02it/s]
 13%|█▎        | 1377/10682 [11:26<1:16:53,  2.02it/s]
 13%|█▎        | 1378/10682 [11:26<1:16:49,  2.02it/s]
 13%|█▎        | 1379/10682 [11:27<1:16:45,  2.02it/s]
 13%|█▎        | 1380/10682 [11:27<1:16:49,  2.02it/s]
 13%|█▎        | 1381/10682 [11:28<1:16:45,  2.02it/s]
 13%|█▎        | 1382/10682 [11:28<1:16:46,  2.02it/s]
 13%|█▎        | 1383/10682 [11:29<1:16:43,  2.02it/s]
 13%|█▎        | 1384/10682 [11:29<1:16:42,  2.02it/s]
 13%|█▎        | 1385/10682 [11:30<1:16:42,  2.02it/s]
 13%|█▎        | 1386/10682 [11:30<1:16:41,  2.02it/s]
 13%|█▎        | 1387/10682 [11:31<1:16:43,  2.02it/s]
 13%|█▎        | 1388/10682 [11:31<1:16:45,  2.02it/s]
 13%|█▎        | 1389/10682 [11:32<1:16:43,  2.02it/s]
 13%|█▎        | 1390/10682 [11:32<1:16:41,  2.02it/s]
 13%|█▎        | 1391/10682 [11:32<1:16:47,  2.02it/s]
 13%|█▎        | 1392/10682 [11:33<1:16:45,  2.02it/s]
 13%|█▎        | 1393/10682 [11:33<1:16:41,  2.02it/s]
 13%|█▎        | 1394/10682 [11:34<1:16:39,  2.02it/s]
 13%|█▎        | 1395/10682 [11:34<1:16:40,  2.02it/s]
 13%|█▎        | 1396/10682 [11:35<1:16:37,  2.02it/s]
 13%|█▎        | 1397/10682 [11:35<1:16:33,  2.02it/s]
 13%|█▎        | 1398/10682 [11:36<1:16:35,  2.02it/s]
 13%|█▎        | 1399/10682 [11:36<1:16:34,  2.02it/s]
 13%|█▎        | 1400/10682 [11:37<1:16:33,  2.02it/s]{'loss': 4.1435, 'grad_norm': 0.2577407658100128, 'learning_rate': 0.0009970775012917489, 'epoch': 0.13}
                                                      
-
 13%|█▎        | 1400/10682 [11:37<1:16:33,  2.02it/s]
 13%|█▎        | 1401/10682 [11:37<1:16:43,  2.02it/s]
 13%|█▎        | 1402/10682 [11:38<1:16:39,  2.02it/s]
 13%|█▎        | 1403/10682 [11:38<1:16:39,  2.02it/s]
 13%|█▎        | 1404/10682 [11:39<1:16:38,  2.02it/s]
 13%|█▎        | 1405/10682 [11:39<1:16:39,  2.02it/s]
 13%|█▎        | 1406/10682 [11:40<1:16:38,  2.02it/s]
 13%|█▎        | 1407/10682 [11:40<1:16:36,  2.02it/s]
 13%|█▎        | 1408/10682 [11:41<1:16:29,  2.02it/s]
 13%|█▎        | 1409/10682 [11:41<1:16:27,  2.02it/s]
 13%|█▎        | 1410/10682 [11:42<1:16:30,  2.02it/s]
 13%|█▎        | 1411/10682 [11:42<1:16:32,  2.02it/s]
 13%|█▎        | 1412/10682 [11:43<1:16:31,  2.02it/s]
 13%|█▎        | 1413/10682 [11:43<1:16:28,  2.02it/s]
 13%|█▎        | 1414/10682 [11:44<1:16:23,  2.02it/s]
 13%|█▎        | 1415/10682 [11:44<1:16:25,  2.02it/s]
 13%|█▎        | 1416/10682 [11:45<1:16:21,  2.02it/s]
 13%|█▎        | 1417/10682 [11:45<1:16:26,  2.02it/s]
 13%|█▎        | 1418/10682 [11:46<1:16:23,  2.02it/s]
 13%|█▎        | 1419/10682 [11:46<1:16:31,  2.02it/s]
 13%|█▎        | 1420/10682 [11:47<1:16:24,  2.02it/s]
 13%|█▎        | 1421/10682 [11:47<1:16:26,  2.02it/s]
 13%|█▎        | 1422/10682 [11:48<1:16:45,  2.01it/s]
 13%|█▎        | 1423/10682 [11:48<1:16:35,  2.01it/s]
 13%|█▎        | 1424/10682 [11:49<1:16:32,  2.02it/s]
 13%|█▎        | 1425/10682 [11:49<1:16:27,  2.02it/s]{'loss': 4.1318, 'grad_norm': 0.2662830948829651, 'learning_rate': 0.00099661988142608, 'epoch': 0.13}
                                                      
-
 13%|█▎        | 1425/10682 [11:49<1:16:27,  2.02it/s]
 13%|█▎        | 1426/10682 [11:50<1:16:32,  2.02it/s]
 13%|█▎        | 1427/10682 [11:50<1:16:26,  2.02it/s]
 13%|█▎        | 1428/10682 [11:51<1:16:28,  2.02it/s]
 13%|█▎        | 1429/10682 [11:51<1:16:25,  2.02it/s]
 13%|█▎        | 1430/10682 [11:52<1:16:31,  2.01it/s]
 13%|█▎        | 1431/10682 [11:52<1:16:34,  2.01it/s]
 13%|█▎        | 1432/10682 [11:53<1:16:28,  2.02it/s]
 13%|█▎        | 1433/10682 [11:53<1:16:30,  2.01it/s]
 13%|█▎        | 1434/10682 [11:54<1:16:28,  2.02it/s]
 13%|█▎        | 1435/10682 [11:54<1:16:29,  2.01it/s]
 13%|█▎        | 1436/10682 [11:55<1:16:24,  2.02it/s]
 13%|█▎        | 1437/10682 [11:55<1:16:21,  2.02it/s]
 13%|█▎        | 1438/10682 [11:56<1:16:17,  2.02it/s]
 13%|█▎        | 1439/10682 [11:56<1:16:16,  2.02it/s]
 13%|█▎        | 1440/10682 [11:57<1:16:17,  2.02it/s]
 13%|█▎        | 1441/10682 [11:57<1:16:16,  2.02it/s]
 13%|█▎        | 1442/10682 [11:58<1:16:11,  2.02it/s]
 14%|█▎        | 1443/10682 [11:58<1:16:19,  2.02it/s]
 14%|█▎        | 1444/10682 [11:59<1:16:15,  2.02it/s]
 14%|█▎        | 1445/10682 [11:59<1:16:13,  2.02it/s]
 14%|█▎        | 1446/10682 [12:00<1:16:11,  2.02it/s]
 14%|█▎        | 1447/10682 [12:00<1:16:10,  2.02it/s]
 14%|█▎        | 1448/10682 [12:01<1:16:07,  2.02it/s]
 14%|█▎        | 1449/10682 [12:01<1:16:09,  2.02it/s]
 14%|█▎        | 1450/10682 [12:02<1:16:06,  2.02it/s]
                                                      {'loss': 4.1207, 'grad_norm': 0.2722988426685333, 'learning_rate': 0.0009961291115580116, 'epoch': 0.14}
-
 14%|█▎        | 1450/10682 [12:02<1:16:06,  2.02it/s]
 14%|█▎        | 1451/10682 [12:02<1:16:12,  2.02it/s]
 14%|█▎        | 1452/10682 [12:03<1:16:13,  2.02it/s]
 14%|█▎        | 1453/10682 [12:03<1:16:15,  2.02it/s]
 14%|█▎        | 1454/10682 [12:04<1:16:16,  2.02it/s]
 14%|█▎        | 1455/10682 [12:04<1:16:15,  2.02it/s]
 14%|█▎        | 1456/10682 [12:05<1:16:11,  2.02it/s]
 14%|█▎        | 1457/10682 [12:05<1:16:07,  2.02it/s]
 14%|█▎        | 1458/10682 [12:06<1:16:05,  2.02it/s]
 14%|█▎        | 1459/10682 [12:06<1:16:08,  2.02it/s]
 14%|█▎        | 1460/10682 [12:07<1:16:03,  2.02it/s]
 14%|█▎        | 1461/10682 [12:07<1:16:23,  2.01it/s]
 14%|█▎        | 1462/10682 [12:08<1:16:14,  2.02it/s]
 14%|█▎        | 1463/10682 [12:08<1:16:09,  2.02it/s]
 14%|█▎        | 1464/10682 [12:09<1:16:06,  2.02it/s]
 14%|█▎        | 1465/10682 [12:09<1:16:12,  2.02it/s]
 14%|█▎        | 1466/10682 [12:10<1:16:04,  2.02it/s]
 14%|█▎        | 1467/10682 [12:10<1:16:03,  2.02it/s]
 14%|█▎        | 1468/10682 [12:11<1:16:03,  2.02it/s]
 14%|█▍        | 1469/10682 [12:11<1:16:00,  2.02it/s]
 14%|█▍        | 1470/10682 [12:12<1:16:01,  2.02it/s]
 14%|█▍        | 1471/10682 [12:12<1:16:07,  2.02it/s]
 14%|█▍        | 1472/10682 [12:13<1:16:07,  2.02it/s]
 14%|█▍        | 1473/10682 [12:13<1:16:10,  2.01it/s]
 14%|█▍        | 1474/10682 [12:14<1:16:06,  2.02it/s]
 14%|█▍        | 1475/10682 [12:14<1:16:05,  2.02it/s]{'loss': 4.1037, 'grad_norm': 0.25797051191329956, 'learning_rate': 0.0009956052244470509, 'epoch': 0.14}
                                                      
-
 14%|█▍        | 1475/10682 [12:14<1:16:05,  2.02it/s]
 14%|█▍        | 1476/10682 [12:15<1:16:06,  2.02it/s]
 14%|█▍        | 1477/10682 [12:15<1:16:08,  2.02it/s]
 14%|█▍        | 1478/10682 [12:16<1:16:09,  2.01it/s]
 14%|█▍        | 1479/10682 [12:16<1:16:06,  2.02it/s]
 14%|█▍        | 1480/10682 [12:17<1:16:07,  2.01it/s]
 14%|█▍        | 1481/10682 [12:17<1:16:09,  2.01it/s]
 14%|█▍        | 1482/10682 [12:18<1:16:14,  2.01it/s]
 14%|█▍        | 1483/10682 [12:18<1:16:04,  2.02it/s]
 14%|█▍        | 1484/10682 [12:19<1:16:06,  2.01it/s]
 14%|█▍        | 1485/10682 [12:19<1:16:07,  2.01it/s]
 14%|█▍        | 1486/10682 [12:20<1:16:01,  2.02it/s]
 14%|█▍        | 1487/10682 [12:20<1:16:07,  2.01it/s]
 14%|█▍        | 1488/10682 [12:21<1:15:58,  2.02it/s]
 14%|█▍        | 1489/10682 [12:21<1:15:59,  2.02it/s]
 14%|█▍        | 1490/10682 [12:22<1:15:56,  2.02it/s]
 14%|█▍        | 1491/10682 [12:22<1:16:00,  2.02it/s]
 14%|█▍        | 1492/10682 [12:23<1:15:53,  2.02it/s]
 14%|█▍        | 1493/10682 [12:23<1:15:49,  2.02it/s]
 14%|█▍        | 1494/10682 [12:24<1:15:49,  2.02it/s]
 14%|█▍        | 1495/10682 [12:24<1:15:49,  2.02it/s]
 14%|█▍        | 1496/10682 [12:25<1:15:45,  2.02it/s]
 14%|█▍        | 1497/10682 [12:25<1:15:46,  2.02it/s]
 14%|█▍        | 1498/10682 [12:26<1:15:42,  2.02it/s]
 14%|█▍        | 1499/10682 [12:26<1:15:47,  2.02it/s]
 14%|█▍        | 1500/10682 [12:27<1:15:41,  2.02it/s]
                                                      {'loss': 4.0958, 'grad_norm': 0.26330456137657166, 'learning_rate': 0.0009950482550633215, 'epoch': 0.14}
-

 14%|█▍        | 1500/10682 [12:27<1:15:41,  2.02it/s]
 14%|█▍        | 1501/10682 [12:27<1:15:55,  2.02it/s]
 14%|█▍        | 1502/10682 [12:28<1:15:53,  2.02it/s]
 14%|█▍        | 1503/10682 [12:28<1:15:51,  2.02it/s]
 14%|█▍        | 1504/10682 [12:28<1:15:49,  2.02it/s]
 14%|█▍        | 1505/10682 [12:29<1:15:48,  2.02it/s]
 14%|█▍        | 1506/10682 [12:29<1:15:52,  2.02it/s]
 14%|█▍        | 1507/10682 [12:30<1:15:44,  2.02it/s]
 14%|█▍        | 1508/10682 [12:30<1:15:48,  2.02it/s]
 14%|█▍        | 1509/10682 [12:31<1:15:56,  2.01it/s]
 14%|█▍        | 1510/10682 [12:31<1:15:53,  2.01it/s]
 14%|█▍        | 1511/10682 [12:32<1:15:56,  2.01it/s]
 14%|█▍        | 1512/10682 [12:32<1:15:44,  2.02it/s]
 14%|█▍        | 1513/10682 [12:33<1:15:48,  2.02it/s]
 14%|█▍        | 1514/10682 [12:33<1:15:44,  2.02it/s]
 14%|█▍        | 1515/10682 [12:34<1:15:43,  2.02it/s]
 14%|█▍        | 1516/10682 [12:34<1:15:38,  2.02it/s]
 14%|█▍        | 1517/10682 [12:35<1:15:41,  2.02it/s]
 14%|█▍        | 1518/10682 [12:35<1:15:39,  2.02it/s]
 14%|█▍        | 1519/10682 [12:36<1:15:41,  2.02it/s]
 14%|█▍        | 1520/10682 [12:36<1:15:37,  2.02it/s]
 14%|█▍        | 1521/10682 [12:37<1:15:42,  2.02it/s]
 14%|█▍        | 1522/10682 [12:37<1:15:46,  2.01it/s]
 14%|█▍        | 1523/10682 [12:38<1:15:39,  2.02it/s]
 14%|█▍        | 1524/10682 [12:38<1:15:42,  2.02it/s]
 14%|█▍        | 1525/10682 [12:39<1:15:40,  2.02it/s]{'loss': 4.0833, 'grad_norm': 0.26850929856300354, 'learning_rate': 0.0009944582405852314, 'epoch': 0.14}
                                                      
-
 14%|█▍        | 1525/10682 [12:39<1:15:40,  2.02it/s]
 14%|█▍        | 1526/10682 [12:39<1:15:42,  2.02it/s]
 14%|█▍        | 1527/10682 [12:40<1:15:37,  2.02it/s]
 14%|█▍        | 1528/10682 [12:40<1:15:35,  2.02it/s]
 14%|█▍        | 1529/10682 [12:41<1:15:37,  2.02it/s]
 14%|█▍        | 1530/10682 [12:41<1:15:38,  2.02it/s]
 14%|█▍        | 1531/10682 [12:42<1:15:37,  2.02it/s]
 14%|█▍        | 1532/10682 [12:42<1:15:43,  2.01it/s]
 14%|█▍        | 1533/10682 [12:43<1:15:41,  2.01it/s]
 14%|█▍        | 1534/10682 [12:43<1:15:33,  2.02it/s]
 14%|█▍        | 1535/10682 [12:44<1:15:34,  2.02it/s]
 14%|█▍        | 1536/10682 [12:44<1:15:25,  2.02it/s]
 14%|█▍        | 1537/10682 [12:45<1:15:26,  2.02it/s]
 14%|█▍        | 1538/10682 [12:45<1:15:22,  2.02it/s]
 14%|█▍        | 1539/10682 [12:46<1:15:31,  2.02it/s]
 14%|█▍        | 1540/10682 [12:46<1:15:28,  2.02it/s]
 14%|█▍        | 1541/10682 [12:47<1:15:41,  2.01it/s]
 14%|█▍        | 1542/10682 [12:47<1:15:28,  2.02it/s]
 14%|█▍        | 1543/10682 [12:48<1:15:32,  2.02it/s]
 14%|█▍        | 1544/10682 [12:48<1:15:25,  2.02it/s]
 14%|█▍        | 1545/10682 [12:49<1:15:23,  2.02it/s]
 14%|█▍        | 1546/10682 [12:49<1:15:16,  2.02it/s]
 14%|█▍        | 1547/10682 [12:50<1:15:15,  2.02it/s]
 14%|█▍        | 1548/10682 [12:50<1:15:19,  2.02it/s]
 15%|█▍        | 1549/10682 [12:51<1:15:25,  2.02it/s]
 15%|█▍        | 1550/10682 [12:51<1:15:24,  2.02it/s]
                                                      {'loss': 4.0684, 'grad_norm': 0.23654504120349884, 'learning_rate': 0.0009938352203969901, 'epoch': 0.15}
-
 15%|█▍        | 1550/10682 [12:51<1:15:24,  2.02it/s]
 15%|█▍        | 1551/10682 [12:52<1:15:28,  2.02it/s]
 15%|█▍        | 1552/10682 [12:52<1:15:25,  2.02it/s]
 15%|█▍        | 1553/10682 [12:53<1:15:30,  2.02it/s]
 15%|█▍        | 1554/10682 [12:53<1:15:22,  2.02it/s]
 15%|█▍        | 1555/10682 [12:54<1:15:22,  2.02it/s]
 15%|█▍        | 1556/10682 [12:54<1:15:15,  2.02it/s]
 15%|█▍        | 1557/10682 [12:55<1:15:17,  2.02it/s]
 15%|█▍        | 1558/10682 [12:55<1:15:15,  2.02it/s]
 15%|█▍        | 1559/10682 [12:56<1:15:18,  2.02it/s]
 15%|█▍        | 1560/10682 [12:56<1:15:16,  2.02it/s]
 15%|█▍        | 1561/10682 [12:57<1:15:18,  2.02it/s]
 15%|█▍        | 1562/10682 [12:57<1:15:12,  2.02it/s]
 15%|█▍        | 1563/10682 [12:58<1:15:15,  2.02it/s]
 15%|█▍        | 1564/10682 [12:58<1:15:11,  2.02it/s]
 15%|█▍        | 1565/10682 [12:59<1:15:12,  2.02it/s]
 15%|█▍        | 1566/10682 [12:59<1:15:10,  2.02it/s]
 15%|█▍        | 1567/10682 [13:00<1:15:06,  2.02it/s]
 15%|█▍        | 1568/10682 [13:00<1:15:10,  2.02it/s]
 15%|█▍        | 1569/10682 [13:01<1:15:11,  2.02it/s]
 15%|█▍        | 1570/10682 [13:01<1:15:09,  2.02it/s]
 15%|█▍        | 1571/10682 [13:02<1:15:09,  2.02it/s]
 15%|█▍        | 1572/10682 [13:02<1:15:05,  2.02it/s]
 15%|█▍        | 1573/10682 [13:03<1:15:15,  2.02it/s]
 15%|█▍        | 1574/10682 [13:03<1:15:20,  2.01it/s]
 15%|█▍        | 1575/10682 [13:04<1:15:17,  2.02it/s]{'loss': 4.0663, 'grad_norm': 0.2444663792848587, 'learning_rate': 0.0009931792360859792, 'epoch': 0.15}
                                                      
-
 15%|█▍        | 1575/10682 [13:04<1:15:17,  2.02it/s]
 15%|█▍        | 1576/10682 [13:04<1:15:20,  2.01it/s]
 15%|█▍        | 1577/10682 [13:05<1:15:13,  2.02it/s]
 15%|█▍        | 1578/10682 [13:05<1:15:14,  2.02it/s]
 15%|█▍        | 1579/10682 [13:06<1:15:07,  2.02it/s]
 15%|█▍        | 1580/10682 [13:06<1:15:08,  2.02it/s]
 15%|█▍        | 1581/10682 [13:07<1:15:03,  2.02it/s]
 15%|█▍        | 1582/10682 [13:07<1:15:08,  2.02it/s]
 15%|█▍        | 1583/10682 [13:08<1:14:59,  2.02it/s]
 15%|█▍        | 1584/10682 [13:08<1:15:00,  2.02it/s]
 15%|█▍        | 1585/10682 [13:09<1:14:56,  2.02it/s]
 15%|█▍        | 1586/10682 [13:09<1:14:56,  2.02it/s]
 15%|█▍        | 1587/10682 [13:10<1:14:55,  2.02it/s]
 15%|█▍        | 1588/10682 [13:10<1:14:58,  2.02it/s]
 15%|█▍        | 1589/10682 [13:11<1:14:56,  2.02it/s]
 15%|█▍        | 1590/10682 [13:11<1:15:07,  2.02it/s]
 15%|█▍        | 1591/10682 [13:12<1:15:09,  2.02it/s]
 15%|█▍        | 1592/10682 [13:12<1:15:05,  2.02it/s]
 15%|█▍        | 1593/10682 [13:13<1:15:00,  2.02it/s]
 15%|█▍        | 1594/10682 [13:13<1:14:59,  2.02it/s]
 15%|█▍        | 1595/10682 [13:14<1:14:56,  2.02it/s]
 15%|█▍        | 1596/10682 [13:14<1:14:56,  2.02it/s]
 15%|█▍        | 1597/10682 [13:15<1:14:54,  2.02it/s]
 15%|█▍        | 1598/10682 [13:15<1:14:56,  2.02it/s]
 15%|█▍        | 1599/10682 [13:16<1:14:57,  2.02it/s]
 15%|█▍        | 1600/10682 [13:16<1:15:01,  2.02it/s]
                                                      
{'loss': 4.0559, 'grad_norm': 0.26061466336250305, 'learning_rate': 0.0009924903314399776, 'epoch': 0.15}
-
 15%|█▍        | 1600/10682 [13:16<1:15:01,  2.02it/s]
 15%|█▍        | 1601/10682 [13:17<1:15:06,  2.02it/s]
 15%|█▍        | 1602/10682 [13:17<1:15:30,  2.00it/s]
 15%|█▌        | 1603/10682 [13:18<1:15:15,  2.01it/s]
 15%|█▌        | 1604/10682 [13:18<1:15:10,  2.01it/s]
 15%|█▌        | 1605/10682 [13:19<1:14:58,  2.02it/s]
 15%|█▌        | 1606/10682 [13:19<1:15:00,  2.02it/s]
 15%|█▌        | 1607/10682 [13:20<1:14:54,  2.02it/s]
 15%|█▌        | 1608/10682 [13:20<1:14:53,  2.02it/s]
 15%|█▌        | 1609/10682 [13:21<1:14:52,  2.02it/s]
 15%|█▌        | 1610/10682 [13:21<1:14:59,  2.02it/s]
 15%|█▌        | 1611/10682 [13:22<1:14:55,  2.02it/s]
 15%|█▌        | 1612/10682 [13:22<1:14:51,  2.02it/s]
 15%|█▌        | 1613/10682 [13:22<1:14:46,  2.02it/s]
 15%|█▌        | 1614/10682 [13:23<1:14:42,  2.02it/s]
 15%|█▌        | 1615/10682 [13:23<1:14:42,  2.02it/s]
 15%|█▌        | 1616/10682 [13:24<1:14:42,  2.02it/s]
 15%|█▌        | 1617/10682 [13:24<1:14:43,  2.02it/s]
 15%|█▌        | 1618/10682 [13:25<1:14:44,  2.02it/s]
 15%|█▌        | 1619/10682 [13:25<1:14:47,  2.02it/s]
 15%|█▌        | 1620/10682 [13:26<1:14:58,  2.01it/s]
 15%|█▌        | 1621/10682 [13:26<1:14:51,  2.02it/s]
 15%|█▌        | 1622/10682 [13:27<1:14:50,  2.02it/s]
 15%|█▌        | 1623/10682 [13:27<1:14:41,  2.02it/s]
 15%|█▌        | 1624/10682 [13:28<1:14:42,  2.02it/s]
 15%|█▌        | 1625/10682 [13:28<1:14:36,  2.02it/s]{'loss': 4.056, 'grad_norm': 0.2571159899234772, 'learning_rate': 0.0009917685524442382, 'epoch': 0.15}
-
                                                      

 15%|█▌        | 1625/10682 [13:28<1:14:36,  2.02it/s]
 15%|█▌        | 1626/10682 [13:29<1:14:46,  2.02it/s]
 15%|█▌        | 1627/10682 [13:29<1:14:37,  2.02it/s]
 15%|█▌        | 1628/10682 [13:30<1:14:40,  2.02it/s]
 15%|█▌        | 1629/10682 [13:30<1:14:41,  2.02it/s]
 15%|█▌        | 1630/10682 [13:31<1:14:46,  2.02it/s]
 15%|█▌        | 1631/10682 [13:31<1:14:40,  2.02it/s]
 15%|█▌        | 1632/10682 [13:32<1:14:44,  2.02it/s]
 15%|█▌        | 1633/10682 [13:32<1:14:47,  2.02it/s]
 15%|█▌        | 1634/10682 [13:33<1:14:47,  2.02it/s]
 15%|█▌        | 1635/10682 [13:33<1:14:45,  2.02it/s]
 15%|█▌        | 1636/10682 [13:34<1:14:45,  2.02it/s]
 15%|█▌        | 1637/10682 [13:34<1:14:41,  2.02it/s]
 15%|█▌        | 1638/10682 [13:35<1:14:48,  2.01it/s]
 15%|█▌        | 1639/10682 [13:35<1:14:47,  2.02it/s]
 15%|█▌        | 1640/10682 [13:36<1:14:39,  2.02it/s]
 15%|█▌        | 1641/10682 [13:36<1:14:38,  2.02it/s]
 15%|█▌        | 1642/10682 [13:37<1:14:33,  2.02it/s]
 15%|█▌        | 1643/10682 [13:37<1:14:34,  2.02it/s]
 15%|█▌        | 1644/10682 [13:38<1:14:30,  2.02it/s]
 15%|█▌        | 1645/10682 [13:38<1:14:29,  2.02it/s]
 15%|█▌        | 1646/10682 [13:39<1:14:30,  2.02it/s]
 15%|█▌        | 1647/10682 [13:39<1:14:31,  2.02it/s]
 15%|█▌        | 1648/10682 [13:40<1:14:33,  2.02it/s]
 15%|█▌        | 1649/10682 [13:40<1:14:34,  2.02it/s]
 15%|█▌        | 1650/10682 [13:41<1:14:33,  2.02it/s]
                                                      {'loss': 4.0376, 'grad_norm': 0.27991271018981934, 'learning_rate': 0.0009910139472784174, 'epoch': 0.15}
-
 15%|█▌        | 1650/10682 [13:41<1:14:33,  2.02it/s]
 15%|█▌        | 1651/10682 [13:41<1:14:40,  2.02it/s]
 15%|█▌        | 1652/10682 [13:42<1:14:39,  2.02it/s]
 15%|█▌        | 1653/10682 [13:42<1:14:33,  2.02it/s]
 15%|█▌        | 1654/10682 [13:43<1:14:32,  2.02it/s]
 15%|█▌        | 1655/10682 [13:43<1:14:30,  2.02it/s]
 16%|█▌        | 1656/10682 [13:44<1:14:31,  2.02it/s]
 16%|█▌        | 1657/10682 [13:44<1:14:27,  2.02it/s]
 16%|█▌        | 1658/10682 [13:45<1:14:33,  2.02it/s]
 16%|█▌        | 1659/10682 [13:45<1:14:30,  2.02it/s]
 16%|█▌        | 1660/10682 [13:46<1:14:30,  2.02it/s]
 16%|█▌        | 1661/10682 [13:46<1:14:25,  2.02it/s]
 16%|█▌        | 1662/10682 [13:47<1:14:29,  2.02it/s]
 16%|█▌        | 1663/10682 [13:47<1:14:24,  2.02it/s]
 16%|█▌        | 1664/10682 [13:48<1:14:24,  2.02it/s]
 16%|█▌        | 1665/10682 [13:48<1:14:21,  2.02it/s]
 16%|█▌        | 1666/10682 [13:49<1:14:28,  2.02it/s]
 16%|█▌        | 1667/10682 [13:49<1:14:20,  2.02it/s]
 16%|█▌        | 1668/10682 [13:50<1:14:21,  2.02it/s]
 16%|█▌        | 1669/10682 [13:50<1:14:19,  2.02it/s]
 16%|█▌        | 1670/10682 [13:51<1:14:24,  2.02it/s]
 16%|█▌        | 1671/10682 [13:51<1:14:20,  2.02it/s]
 16%|█▌        | 1672/10682 [13:52<1:14:21,  2.02it/s]
 16%|█▌        | 1673/10682 [13:52<1:14:16,  2.02it/s]
 16%|█▌        | 1674/10682 [13:53<1:14:19,  2.02it/s]
 16%|█▌        | 1675/10682 [13:53<1:14:17,  2.02it/s]
                                                      {'loss': 4.0246, 'grad_norm': 0.24152375757694244, 'learning_rate': 0.0009902265663133602, 'epoch': 0.16}
-
 16%|█▌        | 1675/10682 [13:53<1:14:17,  2.02it/s]
 16%|█▌        | 1676/10682 [13:54<1:14:32,  2.01it/s]
 16%|█▌        | 1677/10682 [13:54<1:14:27,  2.02it/s]
 16%|█▌        | 1678/10682 [13:55<1:14:30,  2.01it/s]
 16%|█▌        | 1679/10682 [13:55<1:14:31,  2.01it/s]
 16%|█▌        | 1680/10682 [13:56<1:14:25,  2.02it/s]
 16%|█▌        | 1681/10682 [13:56<1:14:23,  2.02it/s]
 16%|█▌        | 1682/10682 [13:57<1:14:18,  2.02it/s]
 16%|█▌        | 1683/10682 [13:57<1:14:20,  2.02it/s]
 16%|█▌        | 1684/10682 [13:58<1:14:15,  2.02it/s]
 16%|█▌        | 1685/10682 [13:58<1:14:13,  2.02it/s]
 16%|█▌        | 1686/10682 [13:59<1:14:14,  2.02it/s]
 16%|█▌        | 1687/10682 [13:59<1:14:09,  2.02it/s]
 16%|█▌        | 1688/10682 [14:00<1:14:11,  2.02it/s]
 16%|█▌        | 1689/10682 [14:00<1:14:13,  2.02it/s]
 16%|█▌        | 1690/10682 [14:01<1:14:11,  2.02it/s]
 16%|█▌        | 1691/10682 [14:01<1:14:14,  2.02it/s]
 16%|█▌        | 1692/10682 [14:02<1:14:12,  2.02it/s]
 16%|█▌        | 1693/10682 [14:02<1:14:07,  2.02it/s]
 16%|█▌        | 1694/10682 [14:03<1:14:07,  2.02it/s]
 16%|█▌        | 1695/10682 [14:03<1:14:00,  2.02it/s]
 16%|█▌        | 1696/10682 [14:04<1:14:02,  2.02it/s]
 16%|█▌        | 1697/10682 [14:04<1:14:01,  2.02it/s]
 16%|█▌        | 1698/10682 [14:05<1:14:05,  2.02it/s]
 16%|█▌        | 1699/10682 [14:05<1:14:06,  2.02it/s]
 16%|█▌        | 1700/10682 [14:06<1:14:06,  2.02it/s]
                                                      
{'loss': 4.0253, 'grad_norm': 0.23957741260528564, 'learning_rate': 0.0009894064621077375, 'epoch': 0.16}
-
 16%|█▌        | 1700/10682 [14:06<1:14:06,  2.02it/s]
 16%|█▌        | 1701/10682 [14:06<1:14:21,  2.01it/s]
 16%|█▌        | 1702/10682 [14:07<1:14:15,  2.02it/s]
 16%|█▌        | 1703/10682 [14:07<1:14:14,  2.02it/s]
 16%|█▌        | 1704/10682 [14:08<1:14:11,  2.02it/s]
 16%|█▌        | 1705/10682 [14:08<1:14:14,  2.02it/s]
 16%|█▌        | 1706/10682 [14:09<1:14:04,  2.02it/s]
 16%|█▌        | 1707/10682 [14:09<1:14:08,  2.02it/s]
 16%|█▌        | 1708/10682 [14:10<1:13:59,  2.02it/s]
 16%|█▌        | 1709/10682 [14:10<1:14:13,  2.02it/s]
 16%|█▌        | 1710/10682 [14:11<1:14:10,  2.02it/s]
 16%|█▌        | 1711/10682 [14:11<1:14:03,  2.02it/s]
 16%|█▌        | 1712/10682 [14:12<1:14:00,  2.02it/s]
 16%|█▌        | 1713/10682 [14:12<1:14:02,  2.02it/s]
 16%|█▌        | 1714/10682 [14:13<1:14:03,  2.02it/s]
 16%|█▌        | 1715/10682 [14:13<1:14:05,  2.02it/s]
 16%|█▌        | 1716/10682 [14:14<1:14:04,  2.02it/s]
 16%|█▌        | 1717/10682 [14:14<1:14:08,  2.02it/s]
 16%|█▌        | 1718/10682 [14:15<1:14:06,  2.02it/s]
 16%|█▌        | 1719/10682 [14:15<1:14:04,  2.02it/s]
 16%|█▌        | 1720/10682 [14:16<1:14:03,  2.02it/s]
 16%|█▌        | 1721/10682 [14:16<1:13:56,  2.02it/s]
 16%|█▌        | 1722/10682 [14:16<1:13:55,  2.02it/s]
 16%|█▌        | 1723/10682 [14:17<1:13:53,  2.02it/s]
 16%|█▌        | 1724/10682 [14:17<1:13:58,  2.02it/s]
 16%|█▌        | 1725/10682 [14:18<1:13:52,  2.02it/s]{'loss': 4.0153, 'grad_norm': 0.25565871596336365, 'learning_rate': 0.0009885536894045377, 'epoch': 0.16}
                                                      
-
 16%|█▌        | 1725/10682 [14:18<1:13:52,  2.02it/s]
 16%|█▌        | 1726/10682 [14:18<1:14:01,  2.02it/s]
 16%|█▌        | 1727/10682 [14:19<1:13:53,  2.02it/s]
 16%|█▌        | 1728/10682 [14:19<1:13:55,  2.02it/s]
 16%|█▌        | 1729/10682 [14:20<1:14:06,  2.01it/s]
 16%|█▌        | 1730/10682 [14:20<1:14:06,  2.01it/s]
 16%|█▌        | 1731/10682 [14:21<1:14:05,  2.01it/s]
 16%|█▌        | 1732/10682 [14:21<1:14:05,  2.01it/s]
 16%|█▌        | 1733/10682 [14:22<1:13:57,  2.02it/s]
 16%|█▌        | 1734/10682 [14:22<1:13:57,  2.02it/s]
 16%|█▌        | 1735/10682 [14:23<1:13:50,  2.02it/s]
 16%|█▋        | 1736/10682 [14:23<1:13:54,  2.02it/s]
 16%|█▋        | 1737/10682 [14:24<1:13:53,  2.02it/s]
 16%|█▋        | 1738/10682 [14:24<1:13:55,  2.02it/s]
 16%|█▋        | 1739/10682 [14:25<1:13:52,  2.02it/s]
 16%|█▋        | 1740/10682 [14:25<1:13:52,  2.02it/s]
 16%|█▋        | 1741/10682 [14:26<1:13:47,  2.02it/s]
 16%|█▋        | 1742/10682 [14:26<1:13:55,  2.02it/s]
 16%|█▋        | 1743/10682 [14:27<1:13:46,  2.02it/s]
 16%|█▋        | 1744/10682 [14:27<1:13:48,  2.02it/s]
 16%|█▋        | 1745/10682 [14:28<1:13:45,  2.02it/s]
 16%|█▋        | 1746/10682 [14:28<1:13:50,  2.02it/s]
 16%|█▋        | 1747/10682 [14:29<1:13:44,  2.02it/s]
 16%|█▋        | 1748/10682 [14:29<1:13:47,  2.02it/s]
 16%|█▋        | 1749/10682 [14:30<1:13:48,  2.02it/s]
 16%|█▋        | 1750/10682 [14:30<1:13:45,  2.02it/s]
                                                      {'loss': 4.0058, 'grad_norm': 0.24486158788204193, 'learning_rate': 0.0009876683051274124, 'epoch': 0.16}
-
 16%|█▋        | 1750/10682 [14:30<1:13:45,  2.02it/s]
 16%|█▋        | 1751/10682 [14:31<1:14:00,  2.01it/s]
 16%|█▋        | 1752/10682 [14:31<1:13:54,  2.01it/s]
 16%|█▋        | 1753/10682 [14:32<1:13:49,  2.02it/s]
 16%|█▋        | 1754/10682 [14:32<1:13:45,  2.02it/s]
 16%|█▋        | 1755/10682 [14:33<1:13:45,  2.02it/s]
 16%|█▋        | 1756/10682 [14:34<1:26:50,  1.71it/s]
 16%|█▋        | 1757/10682 [14:34<1:22:50,  1.80it/s]
 16%|█▋        | 1758/10682 [14:35<1:20:04,  1.86it/s]
 16%|█▋        | 1759/10682 [14:35<1:18:06,  1.90it/s]
 16%|█▋        | 1760/10682 [14:36<1:16:55,  1.93it/s]
 16%|█▋        | 1761/10682 [14:36<1:15:54,  1.96it/s]
 16%|█▋        | 1762/10682 [14:37<1:15:10,  1.98it/s]
 17%|█▋        | 1763/10682 [14:37<1:26:39,  1.72it/s]
 17%|█▋        | 1764/10682 [14:38<1:22:40,  1.80it/s]
 17%|█▋        | 1765/10682 [14:38<1:19:57,  1.86it/s]
 17%|█▋        | 1766/10682 [14:39<1:17:56,  1.91it/s]
 17%|█▋        | 1767/10682 [14:39<1:16:43,  1.94it/s]
 17%|█▋        | 1768/10682 [14:40<1:15:44,  1.96it/s]
 17%|█▋        | 1769/10682 [14:40<1:15:04,  1.98it/s]
 17%|█▋        | 1770/10682 [14:41<1:14:36,  1.99it/s]
 17%|█▋        | 1771/10682 [14:41<1:14:17,  2.00it/s]
 17%|█▋        | 1772/10682 [14:42<1:13:57,  2.01it/s]
 17%|█▋        | 1773/10682 [14:42<1:13:50,  2.01it/s]
 17%|█▋        | 1774/10682 [14:43<1:13:47,  2.01it/s]
 17%|█▋        | 1775/10682 [14:43<1:13:40,  2.01it/s]
                                                      {'loss': 4.0031, 'grad_norm': 0.23441751301288605, 'learning_rate': 0.0009867503683768772, 'epoch': 0.17}
-

 17%|█▋        | 1775/10682 [14:43<1:13:40,  2.01it/s]
 17%|█▋        | 1776/10682 [14:44<1:13:45,  2.01it/s]
 17%|█▋        | 1777/10682 [14:44<1:13:36,  2.02it/s]
 17%|█▋        | 1778/10682 [14:45<1:13:31,  2.02it/s]
 17%|█▋        | 1779/10682 [14:45<1:13:38,  2.01it/s]
 17%|█▋        | 1780/10682 [14:46<1:13:38,  2.01it/s]
 17%|█▋        | 1781/10682 [14:46<1:13:30,  2.02it/s]
 17%|█▋        | 1782/10682 [14:47<1:13:30,  2.02it/s]
 17%|█▋        | 1783/10682 [14:47<1:13:24,  2.02it/s]
 17%|█▋        | 1784/10682 [14:48<1:13:24,  2.02it/s]
 17%|█▋        | 1785/10682 [14:48<1:13:17,  2.02it/s]
 17%|█▋        | 1786/10682 [14:49<1:13:21,  2.02it/s]
 17%|█▋        | 1787/10682 [14:49<1:13:16,  2.02it/s]
 17%|█▋        | 1788/10682 [14:50<1:13:20,  2.02it/s]
 17%|█▋        | 1789/10682 [14:50<1:13:13,  2.02it/s]
 17%|█▋        | 1790/10682 [14:51<1:13:20,  2.02it/s]
 17%|█▋        | 1791/10682 [14:51<1:13:21,  2.02it/s]
 17%|█▋        | 1792/10682 [14:52<1:13:22,  2.02it/s]
 17%|█▋        | 1793/10682 [14:52<1:13:14,  2.02it/s]
 17%|█▋        | 1794/10682 [14:53<1:13:16,  2.02it/s]
 17%|█▋        | 1795/10682 [14:53<1:13:11,  2.02it/s]
 17%|█▋        | 1796/10682 [14:54<1:13:14,  2.02it/s]
 17%|█▋        | 1797/10682 [14:54<1:13:13,  2.02it/s]
 17%|█▋        | 1798/10682 [14:55<1:13:16,  2.02it/s]
 17%|█▋        | 1799/10682 [14:55<1:13:11,  2.02it/s]
 17%|█▋        | 1800/10682 [14:56<1:13:21,  2.02it/s]
                                                      {'loss': 3.9877, 'grad_norm': 0.2605009078979492, 'learning_rate': 0.0009857999404263656, 'epoch': 0.17}
-
 17%|█▋        | 1800/10682 [14:56<1:13:21,  2.02it/s]
 17%|█▋        | 1801/10682 [14:56<1:13:31,  2.01it/s]
 17%|█▋        | 1802/10682 [14:57<1:13:23,  2.02it/s]
 17%|█▋        | 1803/10682 [14:57<1:13:15,  2.02it/s]
 17%|█▋        | 1804/10682 [14:58<1:13:13,  2.02it/s]
 17%|█▋        | 1805/10682 [14:58<1:13:10,  2.02it/s]
 17%|█▋        | 1806/10682 [14:59<1:13:13,  2.02it/s]
 17%|█▋        | 1807/10682 [14:59<1:13:05,  2.02it/s]
 17%|█▋        | 1808/10682 [15:00<1:13:10,  2.02it/s]
 17%|█▋        | 1809/10682 [15:00<1:13:10,  2.02it/s]
 17%|█▋        | 1810/10682 [15:01<1:13:17,  2.02it/s]
 17%|█▋        | 1811/10682 [15:01<1:13:11,  2.02it/s]
 17%|█▋        | 1812/10682 [15:02<1:13:10,  2.02it/s]
 17%|█▋        | 1813/10682 [15:02<1:13:08,  2.02it/s]
 17%|█▋        | 1814/10682 [15:03<1:13:11,  2.02it/s]
 17%|█▋        | 1815/10682 [15:03<1:13:07,  2.02it/s]
 17%|█▋        | 1816/10682 [15:04<1:13:07,  2.02it/s]
 17%|█▋        | 1817/10682 [15:04<1:13:05,  2.02it/s]
 17%|█▋        | 1818/10682 [15:05<1:13:10,  2.02it/s]
 17%|█▋        | 1819/10682 [15:05<1:13:05,  2.02it/s]
 17%|█▋        | 1820/10682 [15:06<1:13:14,  2.02it/s]
 17%|█▋        | 1821/10682 [15:06<1:13:11,  2.02it/s]
 17%|█▋        | 1822/10682 [15:07<1:13:07,  2.02it/s]
 17%|█▋        | 1823/10682 [15:07<1:13:06,  2.02it/s]
 17%|█▋        | 1824/10682 [15:08<1:13:00,  2.02it/s]
 17%|█▋        | 1825/10682 [15:08<1:13:02,  2.02it/s]
                                                      
{'loss': 3.9843, 'grad_norm': 0.22341132164001465, 'learning_rate': 0.0009848170847181403, 'epoch': 0.17}
-
 17%|█▋        | 1825/10682 [15:08<1:13:02,  2.02it/s]
 17%|█▋        | 1826/10682 [15:09<1:13:05,  2.02it/s]
 17%|█▋        | 1827/10682 [15:09<1:13:01,  2.02it/s]
 17%|█▋        | 1828/10682 [15:10<1:12:59,  2.02it/s]
 17%|█▋        | 1829/10682 [15:10<1:13:03,  2.02it/s]
 17%|█▋        | 1830/10682 [15:11<1:13:07,  2.02it/s]
 17%|█▋        | 1831/10682 [15:11<1:13:05,  2.02it/s]
 17%|█▋        | 1832/10682 [15:12<1:13:09,  2.02it/s]
 17%|█▋        | 1833/10682 [15:12<1:13:09,  2.02it/s]
 17%|█▋        | 1834/10682 [15:13<1:12:58,  2.02it/s]
 17%|█▋        | 1835/10682 [15:13<1:12:59,  2.02it/s]
 17%|█▋        | 1836/10682 [15:14<1:12:59,  2.02it/s]
 17%|█▋        | 1837/10682 [15:14<1:12:59,  2.02it/s]
 17%|█▋        | 1838/10682 [15:15<1:13:04,  2.02it/s]
 17%|█▋        | 1839/10682 [15:15<1:13:00,  2.02it/s]
 17%|█▋        | 1840/10682 [15:15<1:13:05,  2.02it/s]
 17%|█▋        | 1841/10682 [15:16<1:12:59,  2.02it/s]
 17%|█▋        | 1842/10682 [15:16<1:12:59,  2.02it/s]
 17%|█▋        | 1843/10682 [15:17<1:12:56,  2.02it/s]
 17%|█▋        | 1844/10682 [15:17<1:13:00,  2.02it/s]
 17%|█▋        | 1845/10682 [15:18<1:12:53,  2.02it/s]
 17%|█▋        | 1846/10682 [15:18<1:12:58,  2.02it/s]
 17%|█▋        | 1847/10682 [15:19<1:12:55,  2.02it/s]
 17%|█▋        | 1848/10682 [15:19<1:13:00,  2.02it/s]
 17%|█▋        | 1849/10682 [15:20<1:12:58,  2.02it/s]
 17%|█▋        | 1850/10682 [15:20<1:13:01,  2.02it/s]
                                                      {'loss': 3.9796, 'grad_norm': 0.23130911588668823, 'learning_rate': 0.0009838018668590573, 'epoch': 0.17}
-
 17%|█▋        | 1850/10682 [15:20<1:13:01,  2.02it/s]
 17%|█▋        | 1851/10682 [15:21<1:13:09,  2.01it/s]
 17%|█▋        | 1852/10682 [15:21<1:12:59,  2.02it/s]
 17%|█▋        | 1853/10682 [15:22<1:12:57,  2.02it/s]
 17%|█▋        | 1854/10682 [15:22<1:12:53,  2.02it/s]
 17%|█▋        | 1855/10682 [15:23<1:12:50,  2.02it/s]
 17%|█▋        | 1856/10682 [15:23<1:12:48,  2.02it/s]
 17%|█▋        | 1857/10682 [15:24<1:12:46,  2.02it/s]
 17%|█▋        | 1858/10682 [15:24<1:12:51,  2.02it/s]
 17%|█▋        | 1859/10682 [15:25<1:12:49,  2.02it/s]
 17%|█▋        | 1860/10682 [15:25<1:12:50,  2.02it/s]
 17%|█▋        | 1861/10682 [15:26<1:12:49,  2.02it/s]
 17%|█▋        | 1862/10682 [15:26<1:12:50,  2.02it/s]
 17%|█▋        | 1863/10682 [15:27<1:12:49,  2.02it/s]
 17%|█▋        | 1864/10682 [15:27<1:12:44,  2.02it/s]
 17%|█▋        | 1865/10682 [15:28<1:12:39,  2.02it/s]
 17%|█▋        | 1866/10682 [15:28<1:12:40,  2.02it/s]
 17%|█▋        | 1867/10682 [15:29<1:12:40,  2.02it/s]
 17%|█▋        | 1868/10682 [15:29<1:12:45,  2.02it/s]
 17%|█▋        | 1869/10682 [15:30<1:12:41,  2.02it/s]
 18%|█▊        | 1870/10682 [15:30<1:12:54,  2.01it/s]
 18%|█▊        | 1871/10682 [15:31<1:12:45,  2.02it/s]
 18%|█▊        | 1872/10682 [15:31<1:12:53,  2.01it/s]
 18%|█▊        | 1873/10682 [15:32<1:12:51,  2.02it/s]
 18%|█▊        | 1874/10682 [15:32<1:12:49,  2.02it/s]
 18%|█▊        | 1875/10682 [15:33<1:12:46,  2.02it/s]
                                                      
{'loss': 3.965, 'grad_norm': 0.23937085270881653, 'learning_rate': 0.000982754354616187, 'epoch': 0.18}
-
 18%|█▊        | 1875/10682 [15:33<1:12:46,  2.02it/s]
 18%|█▊        | 1876/10682 [15:33<1:12:54,  2.01it/s]
 18%|█▊        | 1877/10682 [15:34<1:12:47,  2.02it/s]
 18%|█▊        | 1878/10682 [15:34<1:12:49,  2.01it/s]
 18%|█▊        | 1879/10682 [15:35<1:12:43,  2.02it/s]
 18%|█▊        | 1880/10682 [15:35<1:12:45,  2.02it/s]
 18%|█▊        | 1881/10682 [15:36<1:12:39,  2.02it/s]
 18%|█▊        | 1882/10682 [15:36<1:12:39,  2.02it/s]
 18%|█▊        | 1883/10682 [15:37<1:12:34,  2.02it/s]
 18%|█▊        | 1884/10682 [15:37<1:12:38,  2.02it/s]
 18%|█▊        | 1885/10682 [15:38<1:12:34,  2.02it/s]
 18%|█▊        | 1886/10682 [15:38<1:12:41,  2.02it/s]
 18%|█▊        | 1887/10682 [15:39<1:12:40,  2.02it/s]
 18%|█▊        | 1888/10682 [15:39<1:12:35,  2.02it/s]
 18%|█▊        | 1889/10682 [15:40<1:12:37,  2.02it/s]
 18%|█▊        | 1890/10682 [15:40<1:12:37,  2.02it/s]
 18%|█▊        | 1891/10682 [15:41<1:12:44,  2.01it/s]
 18%|█▊        | 1892/10682 [15:41<1:12:39,  2.02it/s]
 18%|█▊        | 1893/10682 [15:42<1:12:37,  2.02it/s]
 18%|█▊        | 1894/10682 [15:42<1:12:32,  2.02it/s]
 18%|█▊        | 1895/10682 [15:43<1:12:30,  2.02it/s]
 18%|█▊        | 1896/10682 [15:43<1:12:30,  2.02it/s]
 18%|█▊        | 1897/10682 [15:44<1:12:31,  2.02it/s]
 18%|█▊        | 1898/10682 [15:44<1:12:29,  2.02it/s]
 18%|█▊        | 1899/10682 [15:45<1:12:36,  2.02it/s]
 18%|█▊        | 1900/10682 [15:45<1:12:33,  2.02it/s]{'loss': 3.9638, 'grad_norm': 0.2289176881313324, 'learning_rate': 0.0009816746179122904, 'epoch': 0.18}
                                                      
-
 18%|█▊        | 1900/10682 [15:45<1:12:33,  2.02it/s]
 18%|█▊        | 1901/10682 [15:46<1:12:38,  2.01it/s]
 18%|█▊        | 1902/10682 [15:46<1:12:41,  2.01it/s]
 18%|█▊        | 1903/10682 [15:47<1:12:34,  2.02it/s]
 18%|█▊        | 1904/10682 [15:47<1:12:29,  2.02it/s]
 18%|█▊        | 1905/10682 [15:48<1:12:24,  2.02it/s]
 18%|█▊        | 1906/10682 [15:48<1:12:24,  2.02it/s]
 18%|█▊        | 1907/10682 [15:49<1:12:25,  2.02it/s]
 18%|█▊        | 1908/10682 [15:49<1:12:29,  2.02it/s]
 18%|█▊        | 1909/10682 [15:50<1:12:25,  2.02it/s]
 18%|█▊        | 1910/10682 [15:50<1:12:27,  2.02it/s]
 18%|█▊        | 1911/10682 [15:51<1:12:23,  2.02it/s]
 18%|█▊        | 1912/10682 [15:51<1:12:23,  2.02it/s]
 18%|█▊        | 1913/10682 [15:52<1:12:16,  2.02it/s]
 18%|█▊        | 1914/10682 [15:52<1:12:16,  2.02it/s]
 18%|█▊        | 1915/10682 [15:53<1:12:14,  2.02it/s]
 18%|█▊        | 1916/10682 [15:53<1:12:17,  2.02it/s]
 18%|█▊        | 1917/10682 [15:54<1:12:17,  2.02it/s]
 18%|█▊        | 1918/10682 [15:54<1:12:20,  2.02it/s]
 18%|█▊        | 1919/10682 [15:55<1:12:30,  2.01it/s]
 18%|█▊        | 1920/10682 [15:55<1:12:23,  2.02it/s]
 18%|█▊        | 1921/10682 [15:56<1:12:21,  2.02it/s]
 18%|█▊        | 1922/10682 [15:56<1:12:18,  2.02it/s]
 18%|█▊        | 1923/10682 [15:57<1:12:20,  2.02it/s]
 18%|█▊        | 1924/10682 [15:57<1:12:15,  2.02it/s]
 18%|█▊        | 1925/10682 [15:58<1:12:20,  2.02it/s]{'loss': 3.9528, 'grad_norm': 0.23266488313674927, 'learning_rate': 0.0009805627288211524, 'epoch': 0.18}
                                                      
-
 18%|█▊        | 1925/10682 [15:58<1:12:20,  2.02it/s]
 18%|█▊        | 1926/10682 [15:58<1:12:23,  2.02it/s]
 18%|█▊        | 1927/10682 [15:59<1:12:22,  2.02it/s]
 18%|█▊        | 1928/10682 [15:59<1:12:18,  2.02it/s]
 18%|█▊        | 1929/10682 [16:00<1:12:18,  2.02it/s]
 18%|█▊        | 1930/10682 [16:00<1:12:19,  2.02it/s]
 18%|█▊        | 1931/10682 [16:01<1:12:12,  2.02it/s]
 18%|█▊        | 1932/10682 [16:01<1:12:14,  2.02it/s]
 18%|█▊        | 1933/10682 [16:02<1:12:10,  2.02it/s]
 18%|█▊        | 1934/10682 [16:02<1:12:14,  2.02it/s]
 18%|█▊        | 1935/10682 [16:03<1:12:09,  2.02it/s]
 18%|█▊        | 1936/10682 [16:03<1:12:08,  2.02it/s]
 18%|█▊        | 1937/10682 [16:04<1:12:11,  2.02it/s]
 18%|█▊        | 1938/10682 [16:04<1:12:11,  2.02it/s]
 18%|█▊        | 1939/10682 [16:05<1:12:17,  2.02it/s]
 18%|█▊        | 1940/10682 [16:05<1:12:18,  2.01it/s]
 18%|█▊        | 1941/10682 [16:06<1:12:09,  2.02it/s]
 18%|█▊        | 1942/10682 [16:06<1:12:11,  2.02it/s]
 18%|█▊        | 1943/10682 [16:07<1:12:14,  2.02it/s]
 18%|█▊        | 1944/10682 [16:07<1:12:15,  2.02it/s]
 18%|█▊        | 1945/10682 [16:08<1:12:06,  2.02it/s]
 18%|█▊        | 1946/10682 [16:08<1:12:04,  2.02it/s]
 18%|█▊        | 1947/10682 [16:09<1:12:07,  2.02it/s]
 18%|█▊        | 1948/10682 [16:09<1:12:08,  2.02it/s]
 18%|█▊        | 1949/10682 [16:10<1:12:12,  2.02it/s]
 18%|█▊        | 1950/10682 [16:10<1:12:09,  2.02it/s]
                                                      
{'loss': 3.9532, 'grad_norm': 0.23597806692123413, 'learning_rate': 0.0009794187615627694, 'epoch': 0.18}
-
 18%|█▊        | 1950/10682 [16:10<1:12:09,  2.02it/s]
 18%|█▊        | 1951/10682 [16:11<1:12:12,  2.02it/s]
 18%|█▊        | 1952/10682 [16:11<1:12:06,  2.02it/s]
 18%|█▊        | 1953/10682 [16:11<1:12:03,  2.02it/s]
 18%|█▊        | 1954/10682 [16:12<1:12:04,  2.02it/s]
 18%|█▊        | 1955/10682 [16:12<1:12:02,  2.02it/s]
 18%|█▊        | 1956/10682 [16:13<1:12:02,  2.02it/s]
 18%|█▊        | 1957/10682 [16:13<1:12:04,  2.02it/s]
 18%|█▊        | 1958/10682 [16:14<1:12:01,  2.02it/s]
 18%|█▊        | 1959/10682 [16:14<1:12:14,  2.01it/s]
 18%|█▊        | 1960/10682 [16:15<1:12:05,  2.02it/s]
 18%|█▊        | 1961/10682 [16:15<1:11:59,  2.02it/s]
 18%|█▊        | 1962/10682 [16:16<1:11:58,  2.02it/s]
 18%|█▊        | 1963/10682 [16:16<1:11:57,  2.02it/s]
 18%|█▊        | 1964/10682 [16:17<1:12:00,  2.02it/s]
 18%|█▊        | 1965/10682 [16:17<1:12:01,  2.02it/s]
 18%|█▊        | 1966/10682 [16:18<1:11:57,  2.02it/s]
 18%|█▊        | 1967/10682 [16:18<1:11:59,  2.02it/s]
 18%|█▊        | 1968/10682 [16:19<1:11:55,  2.02it/s]
 18%|█▊        | 1969/10682 [16:19<1:11:59,  2.02it/s]
 18%|█▊        | 1970/10682 [16:20<1:11:53,  2.02it/s]
 18%|█▊        | 1971/10682 [16:20<1:11:53,  2.02it/s]
 18%|█▊        | 1972/10682 [16:21<1:11:50,  2.02it/s]
 18%|█▊        | 1973/10682 [16:21<1:11:51,  2.02it/s]
 18%|█▊        | 1974/10682 [16:22<1:11:48,  2.02it/s]
 18%|█▊        | 1975/10682 [16:22<1:11:50,  2.02it/s]
                                                      {'loss': 3.9405, 'grad_norm': 0.239703968167305, 'learning_rate': 0.0009782427924983968, 'epoch': 0.18}
-
 18%|█▊        | 1975/10682 [16:22<1:11:50,  2.02it/s]
 18%|█▊        | 1976/10682 [16:23<1:11:57,  2.02it/s]
 19%|█▊        | 1977/10682 [16:23<1:11:51,  2.02it/s]
 19%|█▊        | 1978/10682 [16:24<1:11:50,  2.02it/s]
 19%|█▊        | 1979/10682 [16:24<1:11:51,  2.02it/s]
 19%|█▊        | 1980/10682 [16:25<1:11:51,  2.02it/s]
 19%|█▊        | 1981/10682 [16:25<1:11:49,  2.02it/s]
 19%|█▊        | 1982/10682 [16:26<1:11:43,  2.02it/s]
 19%|█▊        | 1983/10682 [16:26<1:11:44,  2.02it/s]
 19%|█▊        | 1984/10682 [16:27<1:11:38,  2.02it/s]
 19%|█▊        | 1985/10682 [16:27<1:11:41,  2.02it/s]
 19%|█▊        | 1986/10682 [16:28<1:11:40,  2.02it/s]
 19%|█▊        | 1987/10682 [16:28<1:11:43,  2.02it/s]
 19%|█▊        | 1988/10682 [16:29<1:11:38,  2.02it/s]
 19%|█▊        | 1989/10682 [16:29<1:11:42,  2.02it/s]
 19%|█▊        | 1990/10682 [16:30<1:11:37,  2.02it/s]
 19%|█▊        | 1991/10682 [16:30<1:11:42,  2.02it/s]
 19%|█▊        | 1992/10682 [16:31<1:11:39,  2.02it/s]
 19%|█▊        | 1993/10682 [16:31<1:11:42,  2.02it/s]
 19%|█▊        | 1994/10682 [16:32<1:11:38,  2.02it/s]
 19%|█▊        | 1995/10682 [16:32<1:11:38,  2.02it/s]
 19%|█▊        | 1996/10682 [16:33<1:11:37,  2.02it/s]
 19%|█▊        | 1997/10682 [16:33<1:11:42,  2.02it/s]
 19%|█▊        | 1998/10682 [16:34<1:11:43,  2.02it/s]
 19%|█▊        | 1999/10682 [16:34<1:11:39,  2.02it/s]
 19%|█▊        | 2000/10682 [16:35<1:11:39,  2.02it/s]{'loss': 3.9298, 'grad_norm': 0.2278049737215042, 'learning_rate': 0.00097703490012545, 'epoch': 0.19}
                                                      
-
 19%|█▊        | 2000/10682 [16:35<1:11:39,  2.02it/s]
 19%|█▊        | 2001/10682 [16:35<1:11:47,  2.02it/s]
 19%|█▊        | 2002/10682 [16:36<1:11:45,  2.02it/s]
 19%|█▉        | 2003/10682 [16:36<1:11:41,  2.02it/s]
 19%|█▉        | 2004/10682 [16:37<1:11:39,  2.02it/s]
 19%|█▉        | 2005/10682 [16:37<1:11:34,  2.02it/s]
 19%|█▉        | 2006/10682 [16:38<1:11:38,  2.02it/s]
 19%|█▉        | 2007/10682 [16:38<1:11:40,  2.02it/s]
 19%|█▉        | 2008/10682 [16:39<1:11:34,  2.02it/s]
 19%|█▉        | 2009/10682 [16:39<1:11:32,  2.02it/s]
 19%|█▉        | 2010/10682 [16:40<1:11:30,  2.02it/s]
 19%|█▉        | 2011/10682 [16:40<1:11:28,  2.02it/s]
 19%|█▉        | 2012/10682 [16:41<1:11:30,  2.02it/s]
 19%|█▉        | 2013/10682 [16:41<1:11:31,  2.02it/s]
 19%|█▉        | 2014/10682 [16:42<1:11:35,  2.02it/s]
 19%|█▉        | 2015/10682 [16:42<1:11:31,  2.02it/s]
 19%|█▉        | 2016/10682 [16:43<1:11:31,  2.02it/s]
 19%|█▉        | 2017/10682 [16:43<1:11:27,  2.02it/s]
 19%|█▉        | 2018/10682 [16:44<1:11:36,  2.02it/s]
 19%|█▉        | 2019/10682 [16:44<1:11:35,  2.02it/s]
 19%|█▉        | 2020/10682 [16:45<1:11:31,  2.02it/s]
 19%|█▉        | 2021/10682 [16:45<1:11:28,  2.02it/s]
 19%|█▉        | 2022/10682 [16:46<1:11:24,  2.02it/s]
 19%|█▉        | 2023/10682 [16:46<1:11:24,  2.02it/s]
 19%|█▉        | 2024/10682 [16:47<1:11:25,  2.02it/s]
 19%|█▉        | 2025/10682 [16:47<1:11:22,  2.02it/s]{'loss': 3.9328, 'grad_norm': 0.24783878028392792, 'learning_rate': 0.0009757951650722661, 'epoch': 0.19}
                                                      
-
 19%|█▉        | 2025/10682 [16:47<1:11:22,  2.02it/s]
 19%|█▉        | 2026/10682 [16:48<1:11:30,  2.02it/s]
 19%|█▉        | 2027/10682 [16:48<1:11:27,  2.02it/s]
 19%|█▉        | 2028/10682 [16:49<1:11:27,  2.02it/s]
 19%|█▉        | 2029/10682 [16:49<1:11:25,  2.02it/s]
 19%|█▉        | 2030/10682 [16:50<1:11:23,  2.02it/s]
 19%|█▉        | 2031/10682 [16:50<1:11:20,  2.02it/s]
 19%|█▉        | 2032/10682 [16:51<1:11:23,  2.02it/s]
 19%|█▉        | 2033/10682 [16:51<1:11:20,  2.02it/s]
 19%|█▉        | 2034/10682 [16:52<1:11:23,  2.02it/s]
 19%|█▉        | 2035/10682 [16:52<1:11:19,  2.02it/s]
 19%|█▉        | 2036/10682 [16:53<1:11:23,  2.02it/s]
 19%|█▉        | 2037/10682 [16:53<1:11:19,  2.02it/s]
 19%|█▉        | 2038/10682 [16:54<1:11:24,  2.02it/s]
 19%|█▉        | 2039/10682 [16:54<1:11:24,  2.02it/s]
 19%|█▉        | 2040/10682 [16:55<1:11:19,  2.02it/s]
 19%|█▉        | 2041/10682 [16:55<1:11:17,  2.02it/s]
 19%|█▉        | 2042/10682 [16:56<1:11:13,  2.02it/s]
 19%|█▉        | 2043/10682 [16:56<1:11:13,  2.02it/s]
 19%|█▉        | 2044/10682 [16:57<1:11:12,  2.02it/s]
 19%|█▉        | 2045/10682 [16:57<1:11:12,  2.02it/s]
 19%|█▉        | 2046/10682 [16:58<1:11:17,  2.02it/s]
 19%|█▉        | 2047/10682 [16:58<1:11:21,  2.02it/s]
 19%|█▉        | 2048/10682 [16:59<1:11:20,  2.02it/s]
 19%|█▉        | 2049/10682 [16:59<1:11:20,  2.02it/s]
 19%|█▉        | 2050/10682 [17:00<1:11:15,  2.02it/s]
                                                      {'loss': 3.9247, 'grad_norm': 0.2235223650932312, 'learning_rate': 0.0009745236700927205, 'epoch': 0.19}
-
 19%|█▉        | 2050/10682 [17:00<1:11:15,  2.02it/s]
 19%|█▉        | 2051/10682 [17:00<1:11:23,  2.02it/s]
 19%|█▉        | 2052/10682 [17:01<1:11:18,  2.02it/s]
 19%|█▉        | 2053/10682 [17:01<1:11:17,  2.02it/s]
 19%|█▉        | 2054/10682 [17:02<1:11:10,  2.02it/s]
 19%|█▉        | 2055/10682 [17:02<1:11:11,  2.02it/s]
 19%|█▉        | 2056/10682 [17:02<1:11:10,  2.02it/s]
 19%|█▉        | 2057/10682 [17:03<1:11:11,  2.02it/s]
 19%|█▉        | 2058/10682 [17:03<1:11:11,  2.02it/s]
 19%|█▉        | 2059/10682 [17:04<1:11:08,  2.02it/s]
 19%|█▉        | 2060/10682 [17:04<1:11:08,  2.02it/s]
 19%|█▉        | 2061/10682 [17:05<1:11:10,  2.02it/s]
 19%|█▉        | 2062/10682 [17:05<1:11:16,  2.02it/s]
 19%|█▉        | 2063/10682 [17:06<1:11:12,  2.02it/s]
 19%|█▉        | 2064/10682 [17:06<1:11:11,  2.02it/s]
 19%|█▉        | 2065/10682 [17:07<1:11:07,  2.02it/s]
 19%|█▉        | 2066/10682 [17:07<1:11:15,  2.02it/s]
 19%|█▉        | 2067/10682 [17:08<1:11:08,  2.02it/s]
 19%|█▉        | 2068/10682 [17:08<1:11:08,  2.02it/s]
 19%|█▉        | 2069/10682 [17:09<1:11:07,  2.02it/s]
 19%|█▉        | 2070/10682 [17:09<1:11:06,  2.02it/s]
 19%|█▉        | 2071/10682 [17:10<1:11:03,  2.02it/s]
 19%|█▉        | 2072/10682 [17:10<1:11:03,  2.02it/s]
 19%|█▉        | 2073/10682 [17:11<1:10:56,  2.02it/s]
 19%|█▉        | 2074/10682 [17:11<1:10:57,  2.02it/s]
 19%|█▉        | 2075/10682 [17:12<1:10:59,  2.02it/s]
                                                      {'loss': 3.9237, 'grad_norm': 0.22847244143486023, 'learning_rate': 0.0009732205000607044, 'epoch': 0.19}
-
 19%|█▉        | 2075/10682 [17:12<1:10:59,  2.02it/s]
 19%|█▉        | 2076/10682 [17:12<1:11:09,  2.02it/s]
 19%|█▉        | 2077/10682 [17:13<1:11:04,  2.02it/s]
 19%|█▉        | 2078/10682 [17:13<1:11:04,  2.02it/s]
 19%|█▉        | 2079/10682 [17:14<1:10:58,  2.02it/s]
 19%|█▉        | 2080/10682 [17:14<1:11:01,  2.02it/s]
 19%|█▉        | 2081/10682 [17:15<1:10:55,  2.02it/s]
 19%|█▉        | 2082/10682 [17:15<1:10:56,  2.02it/s]
 20%|█▉        | 2083/10682 [17:16<1:10:52,  2.02it/s]
 20%|█▉        | 2084/10682 [17:16<1:10:53,  2.02it/s]
 20%|█▉        | 2085/10682 [17:17<1:10:57,  2.02it/s]
 20%|█▉        | 2086/10682 [17:17<1:11:01,  2.02it/s]
 20%|█▉        | 2087/10682 [17:18<1:10:57,  2.02it/s]
 20%|█▉        | 2088/10682 [17:18<1:10:58,  2.02it/s]
 20%|█▉        | 2089/10682 [17:19<1:10:55,  2.02it/s]
 20%|█▉        | 2090/10682 [17:19<1:11:01,  2.02it/s]
 20%|█▉        | 2091/10682 [17:20<1:10:55,  2.02it/s]
 20%|█▉        | 2092/10682 [17:20<1:10:57,  2.02it/s]
 20%|█▉        | 2093/10682 [17:21<1:10:53,  2.02it/s]
 20%|█▉        | 2094/10682 [17:21<1:10:51,  2.02it/s]
 20%|█▉        | 2095/10682 [17:22<1:10:53,  2.02it/s]
 20%|█▉        | 2096/10682 [17:22<1:10:54,  2.02it/s]
 20%|█▉        | 2097/10682 [17:23<1:10:52,  2.02it/s]
 20%|█▉        | 2098/10682 [17:23<1:10:57,  2.02it/s]
 20%|█▉        | 2099/10682 [17:24<1:10:51,  2.02it/s]
 20%|█▉        | 2100/10682 [17:24<1:10:55,  2.02it/s]{'loss': 3.9056, 'grad_norm': 0.236755833029747, 'learning_rate': 0.000971885741964458, 'epoch': 0.2}
                                                      
-
 20%|█▉        | 2100/10682 [17:24<1:10:55,  2.02it/s]
 20%|█▉        | 2101/10682 [17:25<1:11:05,  2.01it/s]
 20%|█▉        | 2102/10682 [17:25<1:11:02,  2.01it/s]
 20%|█▉        | 2103/10682 [17:26<1:10:59,  2.01it/s]
 20%|█▉        | 2104/10682 [17:26<1:10:53,  2.02it/s]
 20%|█▉        | 2105/10682 [17:27<1:10:55,  2.02it/s]
 20%|█▉        | 2106/10682 [17:27<1:10:50,  2.02it/s]
 20%|█▉        | 2107/10682 [17:28<1:10:55,  2.02it/s]
 20%|█▉        | 2108/10682 [17:28<1:10:51,  2.02it/s]
 20%|█▉        | 2109/10682 [17:29<1:10:50,  2.02it/s]
 20%|█▉        | 2110/10682 [17:29<1:10:48,  2.02it/s]
 20%|█▉        | 2111/10682 [17:30<1:10:44,  2.02it/s]
 20%|█▉        | 2112/10682 [17:30<1:10:44,  2.02it/s]
 20%|█▉        | 2113/10682 [17:31<1:10:47,  2.02it/s]
 20%|█▉        | 2114/10682 [17:31<1:10:47,  2.02it/s]
 20%|█▉        | 2115/10682 [17:32<1:10:46,  2.02it/s]
 20%|█▉        | 2116/10682 [17:32<1:10:41,  2.02it/s]
 20%|█▉        | 2117/10682 [17:33<1:10:42,  2.02it/s]
 20%|█▉        | 2118/10682 [17:33<1:10:43,  2.02it/s]
 20%|█▉        | 2119/10682 [17:34<1:10:39,  2.02it/s]
 20%|█▉        | 2120/10682 [17:34<1:10:40,  2.02it/s]
 20%|█▉        | 2121/10682 [17:35<1:10:36,  2.02it/s]
 20%|█▉        | 2122/10682 [17:35<1:10:35,  2.02it/s]
 20%|█▉        | 2123/10682 [17:36<1:10:35,  2.02it/s]
 20%|█▉        | 2124/10682 [17:36<1:10:35,  2.02it/s]
 20%|█▉        | 2125/10682 [17:37<1:10:35,  2.02it/s]{'loss': 3.9067, 'grad_norm': 0.22438713908195496, 'learning_rate': 0.0009705194849007652, 'epoch': 0.2}
                                                      
-
 20%|█▉        | 2125/10682 [17:37<1:10:35,  2.02it/s]
 20%|█▉        | 2126/10682 [17:37<1:10:42,  2.02it/s]
 20%|█▉        | 2127/10682 [17:38<1:10:39,  2.02it/s]
 20%|█▉        | 2128/10682 [17:38<1:10:39,  2.02it/s]
 20%|█▉        | 2129/10682 [17:39<1:10:34,  2.02it/s]
 20%|█▉        | 2130/10682 [17:39<1:10:36,  2.02it/s]
 20%|█▉        | 2131/10682 [17:40<1:10:33,  2.02it/s]
 20%|█▉        | 2132/10682 [17:40<1:10:34,  2.02it/s]
 20%|█▉        | 2133/10682 [17:41<1:10:32,  2.02it/s]
 20%|█▉        | 2134/10682 [17:41<1:10:35,  2.02it/s]
 20%|█▉        | 2135/10682 [17:42<1:10:30,  2.02it/s]
 20%|█▉        | 2136/10682 [17:42<1:10:33,  2.02it/s]
 20%|██        | 2137/10682 [17:43<1:10:32,  2.02it/s]
 20%|██        | 2138/10682 [17:43<1:10:45,  2.01it/s]
 20%|██        | 2139/10682 [17:44<1:10:37,  2.02it/s]
 20%|██        | 2140/10682 [17:44<1:10:34,  2.02it/s]
 20%|██        | 2141/10682 [17:45<1:10:32,  2.02it/s]
 20%|██        | 2142/10682 [17:45<1:10:30,  2.02it/s]
 20%|██        | 2143/10682 [17:46<1:10:27,  2.02it/s]
 20%|██        | 2144/10682 [17:46<1:10:28,  2.02it/s]
 20%|██        | 2145/10682 [17:47<1:10:28,  2.02it/s]
 20%|██        | 2146/10682 [17:47<1:10:40,  2.01it/s]
 20%|██        | 2147/10682 [17:48<1:10:33,  2.02it/s]
 20%|██        | 2148/10682 [17:48<1:10:37,  2.01it/s]
 20%|██        | 2149/10682 [17:49<1:10:27,  2.02it/s]
 20%|██        | 2150/10682 [17:49<1:10:28,  2.02it/s]
                                                      {'loss': 3.9195, 'grad_norm': 0.24850378930568695, 'learning_rate': 0.0009691218200690053, 'epoch': 0.2}
-

 20%|██        | 2150/10682 [17:49<1:10:28,  2.02it/s]
 20%|██        | 2151/10682 [17:50<1:10:29,  2.02it/s]
 20%|██        | 2152/10682 [17:50<1:10:29,  2.02it/s]
 20%|██        | 2153/10682 [17:51<1:10:21,  2.02it/s]
 20%|██        | 2154/10682 [17:51<1:10:19,  2.02it/s]
 20%|██        | 2155/10682 [17:52<1:10:21,  2.02it/s]
 20%|██        | 2156/10682 [17:52<1:10:27,  2.02it/s]
 20%|██        | 2157/10682 [17:53<1:10:32,  2.01it/s]
 20%|██        | 2158/10682 [17:53<1:10:25,  2.02it/s]
 20%|██        | 2159/10682 [17:54<1:10:28,  2.02it/s]
 20%|██        | 2160/10682 [17:54<1:10:22,  2.02it/s]
 20%|██        | 2161/10682 [17:55<1:10:22,  2.02it/s]
 20%|██        | 2162/10682 [17:55<1:10:17,  2.02it/s]
 20%|██        | 2163/10682 [17:56<1:10:15,  2.02it/s]
 20%|██        | 2164/10682 [17:56<1:10:14,  2.02it/s]
 20%|██        | 2165/10682 [17:57<1:10:20,  2.02it/s]
 20%|██        | 2166/10682 [17:57<1:10:23,  2.02it/s]
 20%|██        | 2167/10682 [17:58<1:10:25,  2.02it/s]
 20%|██        | 2168/10682 [17:58<1:10:18,  2.02it/s]
 20%|██        | 2169/10682 [17:58<1:10:18,  2.02it/s]
 20%|██        | 2170/10682 [17:59<1:10:17,  2.02it/s]
 20%|██        | 2171/10682 [17:59<1:10:19,  2.02it/s]
 20%|██        | 2172/10682 [18:00<1:10:10,  2.02it/s]
 20%|██        | 2173/10682 [18:00<1:10:11,  2.02it/s]
 20%|██        | 2174/10682 [18:01<1:10:07,  2.02it/s]
 20%|██        | 2175/10682 [18:01<1:10:15,  2.02it/s]
                                                      {'loss': 3.8942, 'grad_norm': 0.26835235953330994, 'learning_rate': 0.0009676928407650656, 'epoch': 0.2}
-
 20%|██        | 2175/10682 [18:01<1:10:15,  2.02it/s]
 20%|██        | 2176/10682 [18:02<1:10:33,  2.01it/s]
 20%|██        | 2177/10682 [18:02<1:10:23,  2.01it/s]
 20%|██        | 2178/10682 [18:03<1:10:19,  2.02it/s]
 20%|██        | 2179/10682 [18:03<1:10:13,  2.02it/s]
 20%|██        | 2180/10682 [18:04<1:10:15,  2.02it/s]
 20%|██        | 2181/10682 [18:04<1:10:09,  2.02it/s]
 20%|██        | 2182/10682 [18:05<1:10:09,  2.02it/s]
 20%|██        | 2183/10682 [18:05<1:10:03,  2.02it/s]
 20%|██        | 2184/10682 [18:06<1:10:09,  2.02it/s]
 20%|██        | 2185/10682 [18:06<1:10:07,  2.02it/s]
 20%|██        | 2186/10682 [18:07<1:10:12,  2.02it/s]
 20%|██        | 2187/10682 [18:07<1:10:13,  2.02it/s]
 20%|██        | 2188/10682 [18:08<1:10:10,  2.02it/s]
 20%|██        | 2189/10682 [18:08<1:10:08,  2.02it/s]
 21%|██        | 2190/10682 [18:09<1:10:02,  2.02it/s]
 21%|██        | 2191/10682 [18:09<1:10:03,  2.02it/s]
 21%|██        | 2192/10682 [18:10<1:09:58,  2.02it/s]
 21%|██        | 2193/10682 [18:10<1:09:57,  2.02it/s]
 21%|██        | 2194/10682 [18:11<1:09:56,  2.02it/s]
 21%|██        | 2195/10682 [18:11<1:09:58,  2.02it/s]
 21%|██        | 2196/10682 [18:12<1:10:03,  2.02it/s]
 21%|██        | 2197/10682 [18:12<1:10:04,  2.02it/s]
 21%|██        | 2198/10682 [18:13<1:10:02,  2.02it/s]
 21%|██        | 2199/10682 [18:13<1:10:04,  2.02it/s]
 21%|██        | 2200/10682 [18:14<1:09:57,  2.02it/s]
                                                      {'loss': 3.8925, 'grad_norm': 0.22390270233154297, 'learning_rate': 0.0009662326423751136, 'epoch': 0.21}
-
 21%|██        | 2200/10682 [18:14<1:09:57,  2.02it/s]
 21%|██        | 2201/10682 [18:14<1:10:02,  2.02it/s]
 21%|██        | 2202/10682 [18:15<1:09:54,  2.02it/s]
 21%|██        | 2203/10682 [18:15<1:09:52,  2.02it/s]
 21%|██        | 2204/10682 [18:16<1:09:55,  2.02it/s]
 21%|██        | 2205/10682 [18:16<1:10:01,  2.02it/s]
 21%|██        | 2206/10682 [18:17<1:10:01,  2.02it/s]
 21%|██        | 2207/10682 [18:17<1:10:01,  2.02it/s]
 21%|██        | 2208/10682 [18:18<1:09:57,  2.02it/s]
 21%|██        | 2209/10682 [18:18<1:10:00,  2.02it/s]
 21%|██        | 2210/10682 [18:19<1:10:02,  2.02it/s]
 21%|██        | 2211/10682 [18:19<1:09:59,  2.02it/s]
 21%|██        | 2212/10682 [18:20<1:10:01,  2.02it/s]
 21%|██        | 2213/10682 [18:20<1:09:58,  2.02it/s]
 21%|██        | 2214/10682 [18:21<1:09:54,  2.02it/s]
 21%|██        | 2215/10682 [18:21<1:09:53,  2.02it/s]
 21%|██        | 2216/10682 [18:22<1:09:51,  2.02it/s]
 21%|██        | 2217/10682 [18:22<1:09:53,  2.02it/s]
 21%|██        | 2218/10682 [18:23<1:09:49,  2.02it/s]
 21%|██        | 2219/10682 [18:23<1:09:54,  2.02it/s]
 21%|██        | 2220/10682 [18:24<1:09:49,  2.02it/s]
 21%|██        | 2221/10682 [18:24<1:09:51,  2.02it/s]
 21%|██        | 2222/10682 [18:25<1:09:48,  2.02it/s]
 21%|██        | 2223/10682 [18:25<1:09:48,  2.02it/s]
 21%|██        | 2224/10682 [18:26<1:09:48,  2.02it/s]
 21%|██        | 2225/10682 [18:26<1:09:49,  2.02it/s]{'loss': 3.8801, 'grad_norm': 0.22843162715435028, 'learning_rate': 0.000964741322369231, 'epoch': 0.21}
-
                                                      

 21%|██        | 2225/10682 [18:26<1:09:49,  2.02it/s]
 21%|██        | 2226/10682 [18:27<1:09:55,  2.02it/s]
 21%|██        | 2227/10682 [18:27<1:09:55,  2.02it/s]
 21%|██        | 2228/10682 [18:28<1:09:54,  2.02it/s]
 21%|██        | 2229/10682 [18:28<1:09:50,  2.02it/s]
 21%|██        | 2230/10682 [18:29<1:09:46,  2.02it/s]
 21%|██        | 2231/10682 [18:29<1:09:42,  2.02it/s]
 21%|██        | 2232/10682 [18:30<1:09:44,  2.02it/s]
 21%|██        | 2233/10682 [18:30<1:09:42,  2.02it/s]
 21%|██        | 2234/10682 [18:31<1:09:41,  2.02it/s]
 21%|██        | 2235/10682 [18:31<1:09:44,  2.02it/s]
 21%|██        | 2236/10682 [18:32<1:09:52,  2.01it/s]
 21%|██        | 2237/10682 [18:32<1:09:45,  2.02it/s]
 21%|██        | 2238/10682 [18:33<1:09:43,  2.02it/s]
 21%|██        | 2239/10682 [18:33<1:09:40,  2.02it/s]
 21%|██        | 2240/10682 [18:34<1:09:38,  2.02it/s]
 21%|██        | 2241/10682 [18:34<1:09:39,  2.02it/s]
 21%|██        | 2242/10682 [18:35<1:09:36,  2.02it/s]
 21%|██        | 2243/10682 [18:35<1:09:35,  2.02it/s]
 21%|██        | 2244/10682 [18:36<1:09:41,  2.02it/s]
 21%|██        | 2245/10682 [18:36<1:09:45,  2.02it/s]
 21%|██        | 2246/10682 [18:37<1:09:45,  2.02it/s]
 21%|██        | 2247/10682 [18:37<1:09:41,  2.02it/s]
 21%|██        | 2248/10682 [18:38<1:09:45,  2.01it/s]
 21%|██        | 2249/10682 [18:38<1:09:34,  2.02it/s]
 21%|██        | 2250/10682 [18:39<1:09:37,  2.02it/s]
                                                      {'loss': 3.875, 'grad_norm': 0.22622725367546082, 'learning_rate': 0.0009632189802949055, 'epoch': 0.21}
-
 21%|██        | 2250/10682 [18:39<1:09:37,  2.02it/s]
 21%|██        | 2251/10682 [18:39<1:09:39,  2.02it/s]
 21%|██        | 2252/10682 [18:40<1:09:40,  2.02it/s]
 21%|██        | 2253/10682 [18:40<1:09:34,  2.02it/s]
 21%|██        | 2254/10682 [18:41<1:09:34,  2.02it/s]
 21%|██        | 2255/10682 [18:41<1:09:31,  2.02it/s]
 21%|██        | 2256/10682 [18:42<1:09:33,  2.02it/s]
 21%|██        | 2257/10682 [18:42<1:09:30,  2.02it/s]
 21%|██        | 2258/10682 [18:43<1:09:32,  2.02it/s]
 21%|██        | 2259/10682 [18:43<1:09:26,  2.02it/s]
 21%|██        | 2260/10682 [18:44<1:09:30,  2.02it/s]
 21%|██        | 2261/10682 [18:44<1:09:25,  2.02it/s]
 21%|██        | 2262/10682 [18:45<1:09:29,  2.02it/s]
 21%|██        | 2263/10682 [18:45<1:09:28,  2.02it/s]
 21%|██        | 2264/10682 [18:46<1:09:32,  2.02it/s]
 21%|██        | 2265/10682 [18:46<1:09:31,  2.02it/s]
 21%|██        | 2266/10682 [18:47<1:09:32,  2.02it/s]
 21%|██        | 2267/10682 [18:47<1:09:27,  2.02it/s]
 21%|██        | 2268/10682 [18:48<1:09:29,  2.02it/s]
 21%|██        | 2269/10682 [18:48<1:09:21,  2.02it/s]
 21%|██▏       | 2270/10682 [18:49<1:09:23,  2.02it/s]
 21%|██▏       | 2271/10682 [18:49<1:09:20,  2.02it/s]
 21%|██▏       | 2272/10682 [18:50<1:09:23,  2.02it/s]
 21%|██▏       | 2273/10682 [18:50<1:09:23,  2.02it/s]
 21%|██▏       | 2274/10682 [18:51<1:09:23,  2.02it/s]
 21%|██▏       | 2275/10682 [18:51<1:09:23,  2.02it/s]
                                                      
{'loss': 3.875, 'grad_norm': 0.2146376073360443, 'learning_rate': 0.0009616657177703876, 'epoch': 0.21}
-
 21%|██▏       | 2275/10682 [18:51<1:09:23,  2.02it/s]
 21%|██▏       | 2276/10682 [18:51<1:09:32,  2.01it/s]
 21%|██▏       | 2277/10682 [18:52<1:09:23,  2.02it/s]
 21%|██▏       | 2278/10682 [18:52<1:09:20,  2.02it/s]
 21%|██▏       | 2279/10682 [18:53<1:09:22,  2.02it/s]
 21%|██▏       | 2280/10682 [18:53<1:09:15,  2.02it/s]
 21%|██▏       | 2281/10682 [18:54<1:09:19,  2.02it/s]
 21%|██▏       | 2282/10682 [18:54<1:09:16,  2.02it/s]
 21%|██▏       | 2283/10682 [18:55<1:09:20,  2.02it/s]
 21%|██▏       | 2284/10682 [18:55<1:09:17,  2.02it/s]
 21%|██▏       | 2285/10682 [18:56<1:09:18,  2.02it/s]
 21%|██▏       | 2286/10682 [18:56<1:09:12,  2.02it/s]
 21%|██▏       | 2287/10682 [18:57<1:09:15,  2.02it/s]
 21%|██▏       | 2288/10682 [18:57<1:09:13,  2.02it/s]
 21%|██▏       | 2289/10682 [18:58<1:09:13,  2.02it/s]
 21%|██▏       | 2290/10682 [18:58<1:09:10,  2.02it/s]
 21%|██▏       | 2291/10682 [18:59<1:09:09,  2.02it/s]
 21%|██▏       | 2292/10682 [18:59<1:09:11,  2.02it/s]
 21%|██▏       | 2293/10682 [19:00<1:09:14,  2.02it/s]
 21%|██▏       | 2294/10682 [19:00<1:09:12,  2.02it/s]
 21%|██▏       | 2295/10682 [19:01<1:09:17,  2.02it/s]
 21%|██▏       | 2296/10682 [19:01<1:09:12,  2.02it/s]
 22%|██▏       | 2297/10682 [19:02<1:09:19,  2.02it/s]
 22%|██▏       | 2298/10682 [19:02<1:09:17,  2.02it/s]
 22%|██▏       | 2299/10682 [19:03<1:09:14,  2.02it/s]
 22%|██▏       | 2300/10682 [19:03<1:09:13,  2.02it/s]
                                                      
{'loss': 3.8746, 'grad_norm': 0.22618240118026733, 'learning_rate': 0.0009600816384779067, 'epoch': 0.22}
-
 22%|██▏       | 2300/10682 [19:03<1:09:13,  2.02it/s]
 22%|██▏       | 2301/10682 [19:04<1:09:16,  2.02it/s]
 22%|██▏       | 2302/10682 [19:04<1:09:12,  2.02it/s]
 22%|██▏       | 2303/10682 [19:05<1:09:16,  2.02it/s]
 22%|██▏       | 2304/10682 [19:05<1:09:21,  2.01it/s]
 22%|██▏       | 2305/10682 [19:06<1:09:24,  2.01it/s]
 22%|██▏       | 2306/10682 [19:06<1:09:22,  2.01it/s]
 22%|██▏       | 2307/10682 [19:07<1:09:23,  2.01it/s]
 22%|██▏       | 2308/10682 [19:07<1:09:16,  2.01it/s]
 22%|██▏       | 2309/10682 [19:08<1:09:19,  2.01it/s]
 22%|██▏       | 2310/10682 [19:08<1:09:10,  2.02it/s]
 22%|██▏       | 2311/10682 [19:09<1:09:12,  2.02it/s]
 22%|██▏       | 2312/10682 [19:09<1:09:05,  2.02it/s]
 22%|██▏       | 2313/10682 [19:10<1:09:09,  2.02it/s]
 22%|██▏       | 2314/10682 [19:10<1:09:04,  2.02it/s]
 22%|██▏       | 2315/10682 [19:11<1:09:10,  2.02it/s]
 22%|██▏       | 2316/10682 [19:11<1:09:13,  2.01it/s]
 22%|██▏       | 2317/10682 [19:12<1:09:06,  2.02it/s]
 22%|██▏       | 2318/10682 [19:12<1:09:07,  2.02it/s]
 22%|██▏       | 2319/10682 [19:13<1:09:02,  2.02it/s]
 22%|██▏       | 2320/10682 [19:13<1:09:02,  2.02it/s]
 22%|██▏       | 2321/10682 [19:14<1:08:58,  2.02it/s]
 22%|██▏       | 2322/10682 [19:14<1:08:56,  2.02it/s]
 22%|██▏       | 2323/10682 [19:15<1:08:58,  2.02it/s]
 22%|██▏       | 2324/10682 [19:15<1:09:01,  2.02it/s]
 22%|██▏       | 2325/10682 [19:16<1:09:01,  2.02it/s]
                                                      {'loss': 3.8752, 'grad_norm': 0.23058968782424927, 'learning_rate': 0.00095846684815675, 'epoch': 0.22}
-
 22%|██▏       | 2325/10682 [19:16<1:09:01,  2.02it/s]
 22%|██▏       | 2326/10682 [19:16<1:09:11,  2.01it/s]
 22%|██▏       | 2327/10682 [19:17<1:09:07,  2.01it/s]
 22%|██▏       | 2328/10682 [19:17<1:09:02,  2.02it/s]
 22%|██▏       | 2329/10682 [19:18<1:09:01,  2.02it/s]
 22%|██▏       | 2330/10682 [19:18<1:08:53,  2.02it/s]
 22%|██▏       | 2331/10682 [19:19<1:08:51,  2.02it/s]
 22%|██▏       | 2332/10682 [19:19<1:08:51,  2.02it/s]
 22%|██▏       | 2333/10682 [19:20<1:08:57,  2.02it/s]
 22%|██▏       | 2334/10682 [19:20<1:08:54,  2.02it/s]
 22%|██▏       | 2335/10682 [19:21<1:08:58,  2.02it/s]
 22%|██▏       | 2336/10682 [19:21<1:08:53,  2.02it/s]
 22%|██▏       | 2337/10682 [19:22<1:08:57,  2.02it/s]
 22%|██▏       | 2338/10682 [19:22<1:08:57,  2.02it/s]
 22%|██▏       | 2339/10682 [19:23<1:08:54,  2.02it/s]
 22%|██▏       | 2340/10682 [19:23<1:08:49,  2.02it/s]
 22%|██▏       | 2341/10682 [19:24<1:08:51,  2.02it/s]
 22%|██▏       | 2342/10682 [19:24<1:08:53,  2.02it/s]
 22%|██▏       | 2343/10682 [19:25<1:08:49,  2.02it/s]
 22%|██▏       | 2344/10682 [19:25<1:08:54,  2.02it/s]
 22%|██▏       | 2345/10682 [19:26<1:08:49,  2.02it/s]
 22%|██▏       | 2346/10682 [19:26<1:08:47,  2.02it/s]
 22%|██▏       | 2347/10682 [19:27<1:08:48,  2.02it/s]
 22%|██▏       | 2348/10682 [19:27<1:08:46,  2.02it/s]
 22%|██▏       | 2349/10682 [19:28<1:08:41,  2.02it/s]
 22%|██▏       | 2350/10682 [19:28<1:08:38,  2.02it/s]
                                                      
{'loss': 3.8629, 'grad_norm': 0.22423987090587616, 'learning_rate': 0.0009568214545962046, 'epoch': 0.22}
-
 22%|██▏       | 2350/10682 [19:28<1:08:38,  2.02it/s]
 22%|██▏       | 2351/10682 [19:29<1:08:56,  2.01it/s]
 22%|██▏       | 2352/10682 [19:29<1:08:54,  2.01it/s]
 22%|██▏       | 2353/10682 [19:30<1:08:48,  2.02it/s]
 22%|██▏       | 2354/10682 [19:30<1:08:47,  2.02it/s]
 22%|██▏       | 2355/10682 [19:31<1:08:45,  2.02it/s]
 22%|██▏       | 2356/10682 [19:31<1:08:46,  2.02it/s]
 22%|██▏       | 2357/10682 [19:32<1:08:45,  2.02it/s]
 22%|██▏       | 2358/10682 [19:32<1:08:45,  2.02it/s]
 22%|██▏       | 2359/10682 [19:33<1:08:43,  2.02it/s]
 22%|██▏       | 2360/10682 [19:33<1:08:42,  2.02it/s]
 22%|██▏       | 2361/10682 [19:34<1:08:42,  2.02it/s]
 22%|██▏       | 2362/10682 [19:34<1:08:41,  2.02it/s]
 22%|██▏       | 2363/10682 [19:35<1:08:40,  2.02it/s]
 22%|██▏       | 2364/10682 [19:35<1:08:43,  2.02it/s]
 22%|██▏       | 2365/10682 [19:36<1:08:44,  2.02it/s]
 22%|██▏       | 2366/10682 [19:36<1:08:39,  2.02it/s]
 22%|██▏       | 2367/10682 [19:37<1:08:37,  2.02it/s]
 22%|██▏       | 2368/10682 [19:37<1:08:36,  2.02it/s]
 22%|██▏       | 2369/10682 [19:38<1:08:34,  2.02it/s]
 22%|██▏       | 2370/10682 [19:38<1:08:36,  2.02it/s]
 22%|██▏       | 2371/10682 [19:39<1:08:28,  2.02it/s]
 22%|██▏       | 2372/10682 [19:39<1:08:34,  2.02it/s]
 22%|██▏       | 2373/10682 [19:40<1:08:27,  2.02it/s]
 22%|██▏       | 2374/10682 [19:40<1:08:31,  2.02it/s]
 22%|██▏       | 2375/10682 [19:41<1:08:28,  2.02it/s]
                                                      {'loss': 3.8619, 'grad_norm': 0.23539182543754578, 'learning_rate': 0.0009551455676283627, 'epoch': 0.22}
-
 22%|██▏       | 2375/10682 [19:41<1:08:28,  2.02it/s]
 22%|██▏       | 2376/10682 [19:41<1:08:38,  2.02it/s]
 22%|██▏       | 2377/10682 [19:42<1:08:34,  2.02it/s]
 22%|██▏       | 2378/10682 [19:42<1:08:33,  2.02it/s]
 22%|██▏       | 2379/10682 [19:43<1:08:33,  2.02it/s]
 22%|██▏       | 2380/10682 [19:43<1:08:33,  2.02it/s]
 22%|██▏       | 2381/10682 [19:44<1:08:28,  2.02it/s]
 22%|██▏       | 2382/10682 [19:44<1:08:33,  2.02it/s]
 22%|██▏       | 2383/10682 [19:45<1:08:31,  2.02it/s]
 22%|██▏       | 2384/10682 [19:45<1:08:35,  2.02it/s]
 22%|██▏       | 2385/10682 [19:46<1:08:42,  2.01it/s]
 22%|██▏       | 2386/10682 [19:46<1:08:34,  2.02it/s]
 22%|██▏       | 2387/10682 [19:46<1:08:34,  2.02it/s]
 22%|██▏       | 2388/10682 [19:47<1:08:29,  2.02it/s]
 22%|██▏       | 2389/10682 [19:47<1:08:32,  2.02it/s]
 22%|██▏       | 2390/10682 [19:48<1:08:27,  2.02it/s]
 22%|██▏       | 2391/10682 [19:48<1:08:27,  2.02it/s]
 22%|██▏       | 2392/10682 [19:49<1:08:27,  2.02it/s]
 22%|██▏       | 2393/10682 [19:49<1:08:26,  2.02it/s]
 22%|██▏       | 2394/10682 [19:50<1:08:29,  2.02it/s]
 22%|██▏       | 2395/10682 [19:50<1:08:27,  2.02it/s]
 22%|██▏       | 2396/10682 [19:51<1:08:22,  2.02it/s]
 22%|██▏       | 2397/10682 [19:51<1:08:25,  2.02it/s]
 22%|██▏       | 2398/10682 [19:52<1:08:23,  2.02it/s]
 22%|██▏       | 2399/10682 [19:52<1:08:23,  2.02it/s]
 22%|██▏       | 2400/10682 [19:53<1:08:22,  2.02it/s]{'loss': 3.857, 'grad_norm': 0.22913986444473267, 'learning_rate': 0.0009534392991207893, 'epoch': 0.22}
                                                      
-
 22%|██▏       | 2400/10682 [19:53<1:08:22,  2.02it/s]
 22%|██▏       | 2401/10682 [19:53<1:08:27,  2.02it/s]
 22%|██▏       | 2402/10682 [19:54<1:08:27,  2.02it/s]
 22%|██▏       | 2403/10682 [19:54<1:08:21,  2.02it/s]
 23%|██▎       | 2404/10682 [19:55<1:08:26,  2.02it/s]
 23%|██▎       | 2405/10682 [19:55<1:08:23,  2.02it/s]
 23%|██▎       | 2406/10682 [19:56<1:08:26,  2.02it/s]
 23%|██▎       | 2407/10682 [19:56<1:08:24,  2.02it/s]
 23%|██▎       | 2408/10682 [19:57<1:08:18,  2.02it/s]
 23%|██▎       | 2409/10682 [19:57<1:08:17,  2.02it/s]
 23%|██▎       | 2410/10682 [19:58<1:08:19,  2.02it/s]
 23%|██▎       | 2411/10682 [19:58<1:08:15,  2.02it/s]
 23%|██▎       | 2412/10682 [19:59<1:08:23,  2.02it/s]
 23%|██▎       | 2413/10682 [19:59<1:08:16,  2.02it/s]
 23%|██▎       | 2414/10682 [20:00<1:08:17,  2.02it/s]
 23%|██▎       | 2415/10682 [20:00<1:08:15,  2.02it/s]
 23%|██▎       | 2416/10682 [20:01<1:08:15,  2.02it/s]
 23%|██▎       | 2417/10682 [20:01<1:08:16,  2.02it/s]
 23%|██▎       | 2418/10682 [20:02<1:08:15,  2.02it/s]
 23%|██▎       | 2419/10682 [20:02<1:08:14,  2.02it/s]
 23%|██▎       | 2420/10682 [20:03<1:08:12,  2.02it/s]
 23%|██▎       | 2421/10682 [20:03<1:08:11,  2.02it/s]
 23%|██▎       | 2422/10682 [20:04<1:08:11,  2.02it/s]
 23%|██▎       | 2423/10682 [20:04<1:08:09,  2.02it/s]
 23%|██▎       | 2424/10682 [20:05<1:08:12,  2.02it/s]
 23%|██▎       | 2425/10682 [20:05<1:08:10,  2.02it/s]
                                                      {'loss': 3.8559, 'grad_norm': 0.23068368434906006, 'learning_rate': 0.0009517027629690559, 'epoch': 0.23}
-
 23%|██▎       | 2425/10682 [20:05<1:08:10,  2.02it/s]
 23%|██▎       | 2426/10682 [20:06<1:08:14,  2.02it/s]
 23%|██▎       | 2427/10682 [20:06<1:08:11,  2.02it/s]
 23%|██▎       | 2428/10682 [20:07<1:08:04,  2.02it/s]
 23%|██▎       | 2429/10682 [20:07<1:08:06,  2.02it/s]
 23%|██▎       | 2430/10682 [20:08<1:08:05,  2.02it/s]
 23%|██▎       | 2431/10682 [20:08<1:08:07,  2.02it/s]
 23%|██▎       | 2432/10682 [20:09<1:08:02,  2.02it/s]
 23%|██▎       | 2433/10682 [20:09<1:08:06,  2.02it/s]
 23%|██▎       | 2434/10682 [20:10<1:08:08,  2.02it/s]
 23%|██▎       | 2435/10682 [20:10<1:08:04,  2.02it/s]
 23%|██▎       | 2436/10682 [20:11<1:08:05,  2.02it/s]
 23%|██▎       | 2437/10682 [20:11<1:08:07,  2.02it/s]
 23%|██▎       | 2438/10682 [20:12<1:08:05,  2.02it/s]
 23%|██▎       | 2439/10682 [20:12<1:08:01,  2.02it/s]
 23%|██▎       | 2440/10682 [20:13<1:08:02,  2.02it/s]
 23%|██▎       | 2441/10682 [20:13<1:08:03,  2.02it/s]
 23%|██▎       | 2442/10682 [20:14<1:08:05,  2.02it/s]
 23%|██▎       | 2443/10682 [20:14<1:08:11,  2.01it/s]
 23%|██▎       | 2444/10682 [20:15<1:08:08,  2.02it/s]
 23%|██▎       | 2445/10682 [20:15<1:08:02,  2.02it/s]
 23%|██▎       | 2446/10682 [20:16<1:07:58,  2.02it/s]
 23%|██▎       | 2447/10682 [20:16<1:08:05,  2.02it/s]
 23%|██▎       | 2448/10682 [20:17<1:08:00,  2.02it/s]
 23%|██▎       | 2449/10682 [20:17<1:07:59,  2.02it/s]
 23%|██▎       | 2450/10682 [20:18<1:07:53,  2.02it/s]
                                                      {'loss': 3.845, 'grad_norm': 0.22878892719745636, 'learning_rate': 0.0009499360750891371, 'epoch': 0.23}
-

 23%|██▎       | 2450/10682 [20:18<1:07:53,  2.02it/s]
 23%|██▎       | 2451/10682 [20:18<1:08:10,  2.01it/s]
 23%|██▎       | 2452/10682 [20:19<1:08:06,  2.01it/s]
 23%|██▎       | 2453/10682 [20:19<1:08:07,  2.01it/s]
 23%|██▎       | 2454/10682 [20:20<1:08:02,  2.02it/s]
 23%|██▎       | 2455/10682 [20:20<1:07:56,  2.02it/s]
 23%|██▎       | 2456/10682 [20:21<1:07:57,  2.02it/s]
 23%|██▎       | 2457/10682 [20:21<1:07:54,  2.02it/s]
 23%|██▎       | 2458/10682 [20:22<1:07:54,  2.02it/s]
 23%|██▎       | 2459/10682 [20:22<1:07:49,  2.02it/s]
 23%|██▎       | 2460/10682 [20:23<1:07:52,  2.02it/s]
 23%|██▎       | 2461/10682 [20:23<1:07:51,  2.02it/s]
 23%|██▎       | 2462/10682 [20:24<1:07:58,  2.02it/s]
 23%|██▎       | 2463/10682 [20:24<1:07:56,  2.02it/s]
 23%|██▎       | 2464/10682 [20:25<1:07:55,  2.02it/s]
 23%|██▎       | 2465/10682 [20:25<1:07:51,  2.02it/s]
 23%|██▎       | 2466/10682 [20:26<1:07:50,  2.02it/s]
 23%|██▎       | 2467/10682 [20:26<1:07:47,  2.02it/s]
 23%|██▎       | 2468/10682 [20:27<1:07:48,  2.02it/s]
 23%|██▎       | 2469/10682 [20:27<1:07:46,  2.02it/s]
 23%|██▎       | 2470/10682 [20:28<1:07:44,  2.02it/s]
 23%|██▎       | 2471/10682 [20:28<1:19:34,  1.72it/s]
 23%|██▎       | 2472/10682 [20:29<1:16:04,  1.80it/s]
 23%|██▎       | 2473/10682 [20:29<1:13:34,  1.86it/s]
 23%|██▎       | 2474/10682 [20:30<1:11:45,  1.91it/s]
 23%|██▎       | 2475/10682 [20:30<1:10:37,  1.94it/s]
                                                      {'loss': 3.8367, 'grad_norm': 0.2495122104883194, 'learning_rate': 0.0009481393534096735, 'epoch': 0.23}
-
 23%|██▎       | 2475/10682 [20:30<1:10:37,  1.94it/s]
 23%|██▎       | 2476/10682 [20:31<1:09:46,  1.96it/s]
 23%|██▎       | 2477/10682 [20:31<1:09:13,  1.98it/s]
 23%|██▎       | 2478/10682 [20:32<1:08:40,  1.99it/s]
 23%|██▎       | 2479/10682 [20:32<1:08:25,  2.00it/s]
 23%|██▎       | 2480/10682 [20:33<1:08:08,  2.01it/s]
 23%|██▎       | 2481/10682 [20:33<1:08:04,  2.01it/s]
 23%|██▎       | 2482/10682 [20:34<1:07:57,  2.01it/s]
 23%|██▎       | 2483/10682 [20:35<1:18:56,  1.73it/s]
 23%|██▎       | 2484/10682 [20:35<1:15:35,  1.81it/s]
 23%|██▎       | 2485/10682 [20:36<1:13:10,  1.87it/s]
 23%|██▎       | 2486/10682 [20:36<1:11:31,  1.91it/s]
 23%|██▎       | 2487/10682 [20:37<1:10:10,  1.95it/s]
 23%|██▎       | 2488/10682 [20:37<1:09:23,  1.97it/s]
 23%|██▎       | 2489/10682 [20:38<1:08:46,  1.99it/s]
 23%|██▎       | 2490/10682 [20:38<1:08:28,  1.99it/s]
 23%|██▎       | 2491/10682 [20:39<1:08:10,  2.00it/s]
 23%|██▎       | 2492/10682 [20:39<1:08:02,  2.01it/s]
 23%|██▎       | 2493/10682 [20:40<1:07:51,  2.01it/s]
 23%|██▎       | 2494/10682 [20:40<1:07:49,  2.01it/s]
 23%|██▎       | 2495/10682 [20:41<1:07:42,  2.02it/s]
 23%|██▎       | 2496/10682 [20:41<1:07:47,  2.01it/s]
 23%|██▎       | 2497/10682 [20:42<1:07:40,  2.02it/s]
 23%|██▎       | 2498/10682 [20:42<1:07:38,  2.02it/s]
 23%|██▎       | 2499/10682 [20:43<1:07:33,  2.02it/s]
 23%|██▎       | 2500/10682 [20:43<1:07:27,  2.02it/s]
                                                      {'loss': 3.8343, 'grad_norm': 0.20996138453483582, 'learning_rate': 0.0009463127178640994, 'epoch': 0.23}
-
 23%|██▎       | 2500/10682 [20:43<1:07:27,  2.02it/s]
 23%|██▎       | 2501/10682 [20:44<1:07:39,  2.02it/s]
 23%|██▎       | 2502/10682 [20:44<1:07:44,  2.01it/s]
 23%|██▎       | 2503/10682 [20:45<1:07:45,  2.01it/s]
 23%|██▎       | 2504/10682 [20:45<1:07:45,  2.01it/s]
 23%|██▎       | 2505/10682 [20:46<1:07:40,  2.01it/s]
 23%|██▎       | 2506/10682 [20:46<1:07:30,  2.02it/s]
 23%|██▎       | 2507/10682 [20:47<1:07:29,  2.02it/s]
 23%|██▎       | 2508/10682 [20:47<1:07:24,  2.02it/s]
 23%|██▎       | 2509/10682 [20:48<1:07:28,  2.02it/s]
 23%|██▎       | 2510/10682 [20:48<1:07:22,  2.02it/s]
 24%|██▎       | 2511/10682 [20:48<1:07:23,  2.02it/s]
 24%|██▎       | 2512/10682 [20:49<1:07:22,  2.02it/s]
 24%|██▎       | 2513/10682 [20:49<1:07:25,  2.02it/s]
 24%|██▎       | 2514/10682 [20:50<1:07:23,  2.02it/s]
 24%|██▎       | 2515/10682 [20:50<1:07:20,  2.02it/s]
 24%|██▎       | 2516/10682 [20:51<1:07:26,  2.02it/s]
 24%|██▎       | 2517/10682 [20:51<1:07:19,  2.02it/s]
 24%|██▎       | 2518/10682 [20:52<1:07:19,  2.02it/s]
 24%|██▎       | 2519/10682 [20:52<1:07:16,  2.02it/s]
 24%|██▎       | 2520/10682 [20:53<1:07:15,  2.02it/s]
 24%|██▎       | 2521/10682 [20:53<1:07:11,  2.02it/s]
 24%|██▎       | 2522/10682 [20:54<1:07:17,  2.02it/s]
 24%|██▎       | 2523/10682 [20:54<1:07:18,  2.02it/s]
 24%|██▎       | 2524/10682 [20:55<1:07:19,  2.02it/s]
 24%|██▎       | 2525/10682 [20:55<1:07:16,  2.02it/s]
                                                      
{'loss': 3.8282, 'grad_norm': 0.24043868482112885, 'learning_rate': 0.0009444562903826375, 'epoch': 0.24}
-
 24%|██▎       | 2525/10682 [20:55<1:07:16,  2.02it/s]
 24%|██▎       | 2526/10682 [20:56<1:07:23,  2.02it/s]
 24%|██▎       | 2527/10682 [20:56<1:07:22,  2.02it/s]
 24%|██▎       | 2528/10682 [20:57<1:07:21,  2.02it/s]
 24%|██▎       | 2529/10682 [20:57<1:07:12,  2.02it/s]
 24%|██▎       | 2530/10682 [20:58<1:07:09,  2.02it/s]
 24%|██▎       | 2531/10682 [20:58<1:07:07,  2.02it/s]
 24%|██▎       | 2532/10682 [20:59<1:07:10,  2.02it/s]
 24%|██▎       | 2533/10682 [20:59<1:07:04,  2.02it/s]
 24%|██▎       | 2534/10682 [21:00<1:07:09,  2.02it/s]
 24%|██▎       | 2535/10682 [21:00<1:07:06,  2.02it/s]
 24%|██▎       | 2536/10682 [21:01<1:07:11,  2.02it/s]
 24%|██▍       | 2537/10682 [21:01<1:07:07,  2.02it/s]
 24%|██▍       | 2538/10682 [21:02<1:07:10,  2.02it/s]
 24%|██▍       | 2539/10682 [21:02<1:07:10,  2.02it/s]
 24%|██▍       | 2540/10682 [21:03<1:07:12,  2.02it/s]
 24%|██▍       | 2541/10682 [21:03<1:07:13,  2.02it/s]
 24%|██▍       | 2542/10682 [21:04<1:07:15,  2.02it/s]
 24%|██▍       | 2543/10682 [21:04<1:07:18,  2.02it/s]
 24%|██▍       | 2544/10682 [21:05<1:07:17,  2.02it/s]
 24%|██▍       | 2545/10682 [21:05<1:07:13,  2.02it/s]
 24%|██▍       | 2546/10682 [21:06<1:07:16,  2.02it/s]
 24%|██▍       | 2547/10682 [21:06<1:07:12,  2.02it/s]
 24%|██▍       | 2548/10682 [21:07<1:07:07,  2.02it/s]
 24%|██▍       | 2549/10682 [21:07<1:07:04,  2.02it/s]
 24%|██▍       | 2550/10682 [21:08<1:06:59,  2.02it/s]{'loss': 3.8229, 'grad_norm': 0.22086229920387268, 'learning_rate': 0.0009425701948841596, 'epoch': 0.24}
                                                      
-
 24%|██▍       | 2550/10682 [21:08<1:06:59,  2.02it/s]
 24%|██▍       | 2551/10682 [21:08<1:07:09,  2.02it/s]
 24%|██▍       | 2552/10682 [21:09<1:07:09,  2.02it/s]
 24%|██▍       | 2553/10682 [21:09<1:07:12,  2.02it/s]
 24%|██▍       | 2554/10682 [21:10<1:07:12,  2.02it/s]
 24%|██▍       | 2555/10682 [21:10<1:07:06,  2.02it/s]
 24%|██▍       | 2556/10682 [21:11<1:07:09,  2.02it/s]
 24%|██▍       | 2557/10682 [21:11<1:07:02,  2.02it/s]
 24%|██▍       | 2558/10682 [21:12<1:07:03,  2.02it/s]
 24%|██▍       | 2559/10682 [21:12<1:07:00,  2.02it/s]
 24%|██▍       | 2560/10682 [21:13<1:07:02,  2.02it/s]
 24%|██▍       | 2561/10682 [21:13<1:07:00,  2.02it/s]
 24%|██▍       | 2562/10682 [21:14<1:07:03,  2.02it/s]
 24%|██▍       | 2563/10682 [21:14<1:06:57,  2.02it/s]
 24%|██▍       | 2564/10682 [21:15<1:06:58,  2.02it/s]
 24%|██▍       | 2565/10682 [21:15<1:06:55,  2.02it/s]
 24%|██▍       | 2566/10682 [21:16<1:06:56,  2.02it/s]
 24%|██▍       | 2567/10682 [21:16<1:06:53,  2.02it/s]
 24%|██▍       | 2568/10682 [21:17<1:06:53,  2.02it/s]
 24%|██▍       | 2569/10682 [21:17<1:06:54,  2.02it/s]
 24%|██▍       | 2570/10682 [21:18<1:06:56,  2.02it/s]
 24%|██▍       | 2571/10682 [21:18<1:06:51,  2.02it/s]
 24%|██▍       | 2572/10682 [21:19<1:06:53,  2.02it/s]
 24%|██▍       | 2573/10682 [21:19<1:06:51,  2.02it/s]
 24%|██▍       | 2574/10682 [21:20<1:06:53,  2.02it/s]
 24%|██▍       | 2575/10682 [21:20<1:06:49,  2.02it/s]
                                                      {'loss': 3.8268, 'grad_norm': 0.22181791067123413, 'learning_rate': 0.0009406545572679153, 'epoch': 0.24}
-
 24%|██▍       | 2575/10682 [21:20<1:06:49,  2.02it/s]
 24%|██▍       | 2576/10682 [21:21<1:06:54,  2.02it/s]
 24%|██▍       | 2577/10682 [21:21<1:06:49,  2.02it/s]
 24%|██▍       | 2578/10682 [21:22<1:06:50,  2.02it/s]
 24%|██▍       | 2579/10682 [21:22<1:06:47,  2.02it/s]
 24%|██▍       | 2580/10682 [21:23<1:06:49,  2.02it/s]
 24%|██▍       | 2581/10682 [21:23<1:06:49,  2.02it/s]
 24%|██▍       | 2582/10682 [21:24<1:06:50,  2.02it/s]
 24%|██▍       | 2583/10682 [21:24<1:06:46,  2.02it/s]
 24%|██▍       | 2584/10682 [21:25<1:06:49,  2.02it/s]
 24%|██▍       | 2585/10682 [21:25<1:06:45,  2.02it/s]
 24%|██▍       | 2586/10682 [21:26<1:06:47,  2.02it/s]
 24%|██▍       | 2587/10682 [21:26<1:06:43,  2.02it/s]
 24%|██▍       | 2588/10682 [21:27<1:06:51,  2.02it/s]
 24%|██▍       | 2589/10682 [21:27<1:06:42,  2.02it/s]
 24%|██▍       | 2590/10682 [21:28<1:06:45,  2.02it/s]
 24%|██▍       | 2591/10682 [21:28<1:06:40,  2.02it/s]
 24%|██▍       | 2592/10682 [21:29<1:06:46,  2.02it/s]
 24%|██▍       | 2593/10682 [21:29<1:06:45,  2.02it/s]
 24%|██▍       | 2594/10682 [21:30<1:06:47,  2.02it/s]
 24%|██▍       | 2595/10682 [21:30<1:06:42,  2.02it/s]
 24%|██▍       | 2596/10682 [21:31<1:06:45,  2.02it/s]
 24%|██▍       | 2597/10682 [21:31<1:06:39,  2.02it/s]
 24%|██▍       | 2598/10682 [21:32<1:06:39,  2.02it/s]
 24%|██▍       | 2599/10682 [21:32<1:06:37,  2.02it/s]
 24%|██▍       | 2600/10682 [21:33<1:06:42,  2.02it/s]
                                                      
{'loss': 3.815, 'grad_norm': 0.21949683129787445, 'learning_rate': 0.0009387095054051274, 'epoch': 0.24}
-
 24%|██▍       | 2600/10682 [21:33<1:06:42,  2.02it/s]
 24%|██▍       | 2601/10682 [21:33<1:06:51,  2.01it/s]
 24%|██▍       | 2602/10682 [21:34<1:06:49,  2.02it/s]
 24%|██▍       | 2603/10682 [21:34<1:06:48,  2.02it/s]
 24%|██▍       | 2604/10682 [21:35<1:06:39,  2.02it/s]
 24%|██▍       | 2605/10682 [21:35<1:06:44,  2.02it/s]
 24%|██▍       | 2606/10682 [21:36<1:06:36,  2.02it/s]
 24%|██▍       | 2607/10682 [21:36<1:06:37,  2.02it/s]
 24%|██▍       | 2608/10682 [21:37<1:06:32,  2.02it/s]
 24%|██▍       | 2609/10682 [21:37<1:06:34,  2.02it/s]
 24%|██▍       | 2610/10682 [21:38<1:06:30,  2.02it/s]
 24%|██▍       | 2611/10682 [21:38<1:06:36,  2.02it/s]
 24%|██▍       | 2612/10682 [21:38<1:06:38,  2.02it/s]
 24%|██▍       | 2613/10682 [21:39<1:06:43,  2.02it/s]
 24%|██▍       | 2614/10682 [21:39<1:06:44,  2.01it/s]
 24%|██▍       | 2615/10682 [21:40<1:06:38,  2.02it/s]
 24%|██▍       | 2616/10682 [21:40<1:06:38,  2.02it/s]
 24%|██▍       | 2617/10682 [21:41<1:06:35,  2.02it/s]
 25%|██▍       | 2618/10682 [21:41<1:06:33,  2.02it/s]
 25%|██▍       | 2619/10682 [21:42<1:06:29,  2.02it/s]
 25%|██▍       | 2620/10682 [21:42<1:06:33,  2.02it/s]
 25%|██▍       | 2621/10682 [21:43<1:06:29,  2.02it/s]
 25%|██▍       | 2622/10682 [21:43<1:06:33,  2.02it/s]
 25%|██▍       | 2623/10682 [21:44<1:06:31,  2.02it/s]
 25%|██▍       | 2624/10682 [21:44<1:06:29,  2.02it/s]
 25%|██▍       | 2625/10682 [21:45<1:06:32,  2.02it/s]
                                                      
{'loss': 3.8144, 'grad_norm': 0.25074854493141174, 'learning_rate': 0.0009367351691304569, 'epoch': 0.25}
-
 25%|██▍       | 2625/10682 [21:45<1:06:32,  2.02it/s]
 25%|██▍       | 2626/10682 [21:45<1:06:34,  2.02it/s]
 25%|██▍       | 2627/10682 [21:46<1:06:30,  2.02it/s]
 25%|██▍       | 2628/10682 [21:46<1:06:28,  2.02it/s]
 25%|██▍       | 2629/10682 [21:47<1:06:20,  2.02it/s]
 25%|██▍       | 2630/10682 [21:47<1:06:24,  2.02it/s]
 25%|██▍       | 2631/10682 [21:48<1:06:20,  2.02it/s]
 25%|██▍       | 2632/10682 [21:48<1:06:30,  2.02it/s]
 25%|██▍       | 2633/10682 [21:49<1:06:27,  2.02it/s]
 25%|██▍       | 2634/10682 [21:49<1:06:25,  2.02it/s]
 25%|██▍       | 2635/10682 [21:50<1:06:23,  2.02it/s]
 25%|██▍       | 2636/10682 [21:50<1:06:19,  2.02it/s]
 25%|██▍       | 2637/10682 [21:51<1:06:18,  2.02it/s]
 25%|██▍       | 2638/10682 [21:51<1:06:21,  2.02it/s]
 25%|██▍       | 2639/10682 [21:52<1:06:21,  2.02it/s]
 25%|██▍       | 2640/10682 [21:52<1:06:29,  2.02it/s]
 25%|██▍       | 2641/10682 [21:53<1:06:23,  2.02it/s]
 25%|██▍       | 2642/10682 [21:53<1:06:22,  2.02it/s]
 25%|██▍       | 2643/10682 [21:54<1:06:20,  2.02it/s]
 25%|██▍       | 2644/10682 [21:54<1:06:22,  2.02it/s]
 25%|██▍       | 2645/10682 [21:55<1:06:20,  2.02it/s]
 25%|██▍       | 2646/10682 [21:55<1:06:25,  2.02it/s]
 25%|██▍       | 2647/10682 [21:56<1:06:16,  2.02it/s]
 25%|██▍       | 2648/10682 [21:56<1:06:14,  2.02it/s]
 25%|██▍       | 2649/10682 [21:57<1:06:10,  2.02it/s]
 25%|██▍       | 2650/10682 [21:57<1:06:06,  2.02it/s]
                                                      
{'loss': 3.8155, 'grad_norm': 0.22620360553264618, 'learning_rate': 0.0009347316802333359, 'epoch': 0.25}
-
 25%|██▍       | 2650/10682 [21:57<1:06:06,  2.02it/s]
 25%|██▍       | 2651/10682 [21:58<1:06:19,  2.02it/s]
 25%|██▍       | 2652/10682 [21:58<1:06:19,  2.02it/s]
 25%|██▍       | 2653/10682 [21:59<1:06:14,  2.02it/s]
 25%|██▍       | 2654/10682 [21:59<1:06:14,  2.02it/s]
 25%|██▍       | 2655/10682 [22:00<1:06:14,  2.02it/s]
 25%|██▍       | 2656/10682 [22:00<1:06:15,  2.02it/s]
 25%|██▍       | 2657/10682 [22:01<1:06:10,  2.02it/s]
 25%|██▍       | 2658/10682 [22:01<1:06:16,  2.02it/s]
 25%|██▍       | 2659/10682 [22:02<1:06:16,  2.02it/s]
 25%|██▍       | 2660/10682 [22:02<1:06:14,  2.02it/s]
 25%|██▍       | 2661/10682 [22:03<1:06:15,  2.02it/s]
 25%|██▍       | 2662/10682 [22:03<1:06:12,  2.02it/s]
 25%|██▍       | 2663/10682 [22:04<1:06:08,  2.02it/s]
 25%|██▍       | 2664/10682 [22:04<1:06:11,  2.02it/s]
 25%|██▍       | 2665/10682 [22:05<1:06:09,  2.02it/s]
 25%|██▍       | 2666/10682 [22:05<1:06:07,  2.02it/s]
 25%|██▍       | 2667/10682 [22:06<1:06:03,  2.02it/s]
 25%|██▍       | 2668/10682 [22:06<1:06:05,  2.02it/s]
 25%|██▍       | 2669/10682 [22:07<1:06:00,  2.02it/s]
 25%|██▍       | 2670/10682 [22:07<1:06:04,  2.02it/s]
 25%|██▌       | 2671/10682 [22:08<1:06:02,  2.02it/s]
 25%|██▌       | 2672/10682 [22:08<1:06:03,  2.02it/s]
 25%|██▌       | 2673/10682 [22:09<1:06:03,  2.02it/s]
 25%|██▌       | 2674/10682 [22:09<1:06:06,  2.02it/s]
 25%|██▌       | 2675/10682 [22:10<1:06:09,  2.02it/s]
                                                      {'loss': 3.8182, 'grad_norm': 0.22346873581409454, 'learning_rate': 0.0009326991724491712, 'epoch': 0.25}
-
 25%|██▌       | 2675/10682 [22:10<1:06:09,  2.02it/s]
 25%|██▌       | 2676/10682 [22:10<1:06:11,  2.02it/s]
 25%|██▌       | 2677/10682 [22:11<1:06:08,  2.02it/s]
 25%|██▌       | 2678/10682 [22:11<1:06:03,  2.02it/s]
 25%|██▌       | 2679/10682 [22:12<1:06:05,  2.02it/s]
 25%|██▌       | 2680/10682 [22:12<1:06:04,  2.02it/s]
 25%|██▌       | 2681/10682 [22:13<1:06:03,  2.02it/s]
 25%|██▌       | 2682/10682 [22:13<1:06:08,  2.02it/s]
 25%|██▌       | 2683/10682 [22:14<1:06:02,  2.02it/s]
 25%|██▌       | 2684/10682 [22:14<1:06:05,  2.02it/s]
 25%|██▌       | 2685/10682 [22:15<1:06:00,  2.02it/s]
 25%|██▌       | 2686/10682 [22:15<1:06:02,  2.02it/s]
 25%|██▌       | 2687/10682 [22:16<1:06:00,  2.02it/s]
 25%|██▌       | 2688/10682 [22:16<1:06:02,  2.02it/s]
 25%|██▌       | 2689/10682 [22:17<1:05:59,  2.02it/s]
 25%|██▌       | 2690/10682 [22:17<1:06:00,  2.02it/s]
 25%|██▌       | 2691/10682 [22:18<1:05:55,  2.02it/s]
 25%|██▌       | 2692/10682 [22:18<1:05:56,  2.02it/s]
 25%|██▌       | 2693/10682 [22:19<1:05:58,  2.02it/s]
 25%|██▌       | 2694/10682 [22:19<1:06:04,  2.01it/s]
 25%|██▌       | 2695/10682 [22:20<1:05:56,  2.02it/s]
 25%|██▌       | 2696/10682 [22:20<1:05:53,  2.02it/s]
 25%|██▌       | 2697/10682 [22:21<1:05:52,  2.02it/s]
 25%|██▌       | 2698/10682 [22:21<1:05:51,  2.02it/s]
 25%|██▌       | 2699/10682 [22:22<1:05:49,  2.02it/s]
 25%|██▌       | 2700/10682 [22:22<1:05:54,  2.02it/s]
                                                      
{'loss': 3.8034, 'grad_norm': 0.22391316294670105, 'learning_rate': 0.0009306377814504165, 'epoch': 0.25}
-
 25%|██▌       | 2700/10682 [22:22<1:05:54,  2.02it/s]
 25%|██▌       | 2701/10682 [22:23<1:06:14,  2.01it/s]
 25%|██▌       | 2702/10682 [22:23<1:06:02,  2.01it/s]
 25%|██▌       | 2703/10682 [22:24<1:06:00,  2.01it/s]
 25%|██▌       | 2704/10682 [22:24<1:05:52,  2.02it/s]
 25%|██▌       | 2705/10682 [22:25<1:05:52,  2.02it/s]
 25%|██▌       | 2706/10682 [22:25<1:05:49,  2.02it/s]
 25%|██▌       | 2707/10682 [22:26<1:05:50,  2.02it/s]
 25%|██▌       | 2708/10682 [22:26<1:05:45,  2.02it/s]
 25%|██▌       | 2709/10682 [22:27<1:05:51,  2.02it/s]
 25%|██▌       | 2710/10682 [22:27<1:05:47,  2.02it/s]
 25%|██▌       | 2711/10682 [22:28<1:05:54,  2.02it/s]
 25%|██▌       | 2712/10682 [22:28<1:05:50,  2.02it/s]
 25%|██▌       | 2713/10682 [22:29<1:05:50,  2.02it/s]
 25%|██▌       | 2714/10682 [22:29<1:05:49,  2.02it/s]
 25%|██▌       | 2715/10682 [22:30<1:05:48,  2.02it/s]
 25%|██▌       | 2716/10682 [22:30<1:05:46,  2.02it/s]
 25%|██▌       | 2717/10682 [22:31<1:05:46,  2.02it/s]
 25%|██▌       | 2718/10682 [22:31<1:05:43,  2.02it/s]
 25%|██▌       | 2719/10682 [22:31<1:05:43,  2.02it/s]
 25%|██▌       | 2720/10682 [22:32<1:05:42,  2.02it/s]
 25%|██▌       | 2721/10682 [22:32<1:05:43,  2.02it/s]
 25%|██▌       | 2722/10682 [22:33<1:05:49,  2.02it/s]
 25%|██▌       | 2723/10682 [22:33<1:05:41,  2.02it/s]
 26%|██▌       | 2724/10682 [22:34<1:05:41,  2.02it/s]
 26%|██▌       | 2725/10682 [22:34<1:05:39,  2.02it/s]
                                                      {'loss': 3.8009, 'grad_norm': 0.22859042882919312, 'learning_rate': 0.0009285476448375166, 'epoch': 0.26}
-

 26%|██▌       | 2725/10682 [22:34<1:05:39,  2.02it/s]
 26%|██▌       | 2726/10682 [22:35<1:05:45,  2.02it/s]
 26%|██▌       | 2727/10682 [22:35<1:05:41,  2.02it/s]
 26%|██▌       | 2728/10682 [22:36<1:05:40,  2.02it/s]
 26%|██▌       | 2729/10682 [22:36<1:05:40,  2.02it/s]
 26%|██▌       | 2730/10682 [22:37<1:05:41,  2.02it/s]
 26%|██▌       | 2731/10682 [22:37<1:05:39,  2.02it/s]
 26%|██▌       | 2732/10682 [22:38<1:05:38,  2.02it/s]
 26%|██▌       | 2733/10682 [22:38<1:05:35,  2.02it/s]
 26%|██▌       | 2734/10682 [22:39<1:05:36,  2.02it/s]
 26%|██▌       | 2735/10682 [22:39<1:05:35,  2.02it/s]
 26%|██▌       | 2736/10682 [22:40<1:05:36,  2.02it/s]
 26%|██▌       | 2737/10682 [22:40<1:05:30,  2.02it/s]
 26%|██▌       | 2738/10682 [22:41<1:05:33,  2.02it/s]
 26%|██▌       | 2739/10682 [22:41<1:05:32,  2.02it/s]
 26%|██▌       | 2740/10682 [22:42<1:05:34,  2.02it/s]
 26%|██▌       | 2741/10682 [22:42<1:05:37,  2.02it/s]
 26%|██▌       | 2742/10682 [22:43<1:05:32,  2.02it/s]
 26%|██▌       | 2743/10682 [22:43<1:05:29,  2.02it/s]
 26%|██▌       | 2744/10682 [22:44<1:05:26,  2.02it/s]
 26%|██▌       | 2745/10682 [22:44<1:05:24,  2.02it/s]
 26%|██▌       | 2746/10682 [22:45<1:05:23,  2.02it/s]
 26%|██▌       | 2747/10682 [22:45<1:05:26,  2.02it/s]
 26%|██▌       | 2748/10682 [22:46<1:05:26,  2.02it/s]
 26%|██▌       | 2749/10682 [22:46<1:05:28,  2.02it/s]
 26%|██▌       | 2750/10682 [22:47<1:05:28,  2.02it/s]
                                                      {'loss': 3.8091, 'grad_norm': 0.22909915447235107, 'learning_rate': 0.0009264289021297221, 'epoch': 0.26}
-
 26%|██▌       | 2750/10682 [22:47<1:05:28,  2.02it/s]
 26%|██▌       | 2751/10682 [22:47<1:05:43,  2.01it/s]
 26%|██▌       | 2752/10682 [22:48<1:05:33,  2.02it/s]
 26%|██▌       | 2753/10682 [22:48<1:05:31,  2.02it/s]
 26%|██▌       | 2754/10682 [22:49<1:05:27,  2.02it/s]
 26%|██▌       | 2755/10682 [22:49<1:05:29,  2.02it/s]
 26%|██▌       | 2756/10682 [22:50<1:05:24,  2.02it/s]
 26%|██▌       | 2757/10682 [22:50<1:05:28,  2.02it/s]
 26%|██▌       | 2758/10682 [22:51<1:05:26,  2.02it/s]
 26%|██▌       | 2759/10682 [22:51<1:05:30,  2.02it/s]
 26%|██▌       | 2760/10682 [22:52<1:05:36,  2.01it/s]
 26%|██▌       | 2761/10682 [22:52<1:05:30,  2.02it/s]
 26%|██▌       | 2762/10682 [22:53<1:05:35,  2.01it/s]
 26%|██▌       | 2763/10682 [22:53<1:05:26,  2.02it/s]
 26%|██▌       | 2764/10682 [22:54<1:05:26,  2.02it/s]
 26%|██▌       | 2765/10682 [22:54<1:05:25,  2.02it/s]
 26%|██▌       | 2766/10682 [22:55<1:05:24,  2.02it/s]
 26%|██▌       | 2767/10682 [22:55<1:05:19,  2.02it/s]
 26%|██▌       | 2768/10682 [22:56<1:05:25,  2.02it/s]
 26%|██▌       | 2769/10682 [22:56<1:05:20,  2.02it/s]
 26%|██▌       | 2770/10682 [22:57<1:05:21,  2.02it/s]
 26%|██▌       | 2771/10682 [22:57<1:05:17,  2.02it/s]
 26%|██▌       | 2772/10682 [22:58<1:05:18,  2.02it/s]
 26%|██▌       | 2773/10682 [22:58<1:05:18,  2.02it/s]
 26%|██▌       | 2774/10682 [22:59<1:05:20,  2.02it/s]
 26%|██▌       | 2775/10682 [22:59<1:05:14,  2.02it/s]
                                                      {'loss': 3.7978, 'grad_norm': 0.22783538699150085, 'learning_rate': 0.0009242816947557767, 'epoch': 0.26}
-
 26%|██▌       | 2775/10682 [22:59<1:05:14,  2.02it/s]
 26%|██▌       | 2776/10682 [23:00<1:05:17,  2.02it/s]
 26%|██▌       | 2777/10682 [23:00<1:05:13,  2.02it/s]
 26%|██▌       | 2778/10682 [23:01<1:05:15,  2.02it/s]
 26%|██▌       | 2779/10682 [23:01<1:05:15,  2.02it/s]
 26%|██▌       | 2780/10682 [23:02<1:05:17,  2.02it/s]
 26%|██▌       | 2781/10682 [23:02<1:05:17,  2.02it/s]
 26%|██▌       | 2782/10682 [23:03<1:05:20,  2.02it/s]
 26%|██▌       | 2783/10682 [23:03<1:05:21,  2.01it/s]
 26%|██▌       | 2784/10682 [23:04<1:05:17,  2.02it/s]
 26%|██▌       | 2785/10682 [23:04<1:05:15,  2.02it/s]
 26%|██▌       | 2786/10682 [23:05<1:05:12,  2.02it/s]
 26%|██▌       | 2787/10682 [23:05<1:05:04,  2.02it/s]
 26%|██▌       | 2788/10682 [23:06<1:05:04,  2.02it/s]
 26%|██▌       | 2789/10682 [23:06<1:05:06,  2.02it/s]
 26%|██▌       | 2790/10682 [23:07<1:05:13,  2.02it/s]
 26%|██▌       | 2791/10682 [23:07<1:05:13,  2.02it/s]
 26%|██▌       | 2792/10682 [23:08<1:05:08,  2.02it/s]
 26%|██▌       | 2793/10682 [23:08<1:05:08,  2.02it/s]
 26%|██▌       | 2794/10682 [23:09<1:05:06,  2.02it/s]
 26%|██▌       | 2795/10682 [23:09<1:05:07,  2.02it/s]
 26%|██▌       | 2796/10682 [23:10<1:05:06,  2.02it/s]
 26%|██▌       | 2797/10682 [23:10<1:05:05,  2.02it/s]
 26%|██▌       | 2798/10682 [23:11<1:05:10,  2.02it/s]
 26%|██▌       | 2799/10682 [23:11<1:05:06,  2.02it/s]
 26%|██▌       | 2800/10682 [23:12<1:05:06,  2.02it/s]
                                                      {'loss': 3.7843, 'grad_norm': 0.21136480569839478, 'learning_rate': 0.0009221061660444758, 'epoch': 0.26}
-
 26%|██▌       | 2800/10682 [23:12<1:05:06,  2.02it/s]
 26%|██▌       | 2801/10682 [23:12<1:05:16,  2.01it/s]
 26%|██▌       | 2802/10682 [23:13<1:05:13,  2.01it/s]
 26%|██▌       | 2803/10682 [23:13<1:05:09,  2.02it/s]
 26%|██▌       | 2804/10682 [23:14<1:05:08,  2.02it/s]
 26%|██▋       | 2805/10682 [23:14<1:04:58,  2.02it/s]
 26%|██▋       | 2806/10682 [23:15<1:04:59,  2.02it/s]
 26%|██▋       | 2807/10682 [23:15<1:04:59,  2.02it/s]
 26%|██▋       | 2808/10682 [23:16<1:05:02,  2.02it/s]
 26%|██▋       | 2809/10682 [23:16<1:05:02,  2.02it/s]
 26%|██▋       | 2810/10682 [23:17<1:05:04,  2.02it/s]
 26%|██▋       | 2811/10682 [23:17<1:05:07,  2.01it/s]
 26%|██▋       | 2812/10682 [23:18<1:05:02,  2.02it/s]
 26%|██▋       | 2813/10682 [23:18<1:05:02,  2.02it/s]
 26%|██▋       | 2814/10682 [23:19<1:04:59,  2.02it/s]
 26%|██▋       | 2815/10682 [23:19<1:04:56,  2.02it/s]
 26%|██▋       | 2816/10682 [23:20<1:04:55,  2.02it/s]
 26%|██▋       | 2817/10682 [23:20<1:04:52,  2.02it/s]
 26%|██▋       | 2818/10682 [23:21<1:04:57,  2.02it/s]
 26%|██▋       | 2819/10682 [23:21<1:04:50,  2.02it/s]
 26%|██▋       | 2820/10682 [23:22<1:04:52,  2.02it/s]
 26%|██▋       | 2821/10682 [23:22<1:04:48,  2.02it/s]
 26%|██▋       | 2822/10682 [23:23<1:04:53,  2.02it/s]
 26%|██▋       | 2823/10682 [23:23<1:04:54,  2.02it/s]
 26%|██▋       | 2824/10682 [23:24<1:04:53,  2.02it/s]
 26%|██▋       | 2825/10682 [23:24<1:04:48,  2.02it/s]
                                                      {'loss': 3.7809, 'grad_norm': 0.2135409414768219, 'learning_rate': 0.0009199024612151008, 'epoch': 0.26}
-
 26%|██▋       | 2825/10682 [23:24<1:04:48,  2.02it/s]
 26%|██▋       | 2826/10682 [23:25<1:04:56,  2.02it/s]
 26%|██▋       | 2827/10682 [23:25<1:04:52,  2.02it/s]
 26%|██▋       | 2828/10682 [23:26<1:04:54,  2.02it/s]
 26%|██▋       | 2829/10682 [23:26<1:04:48,  2.02it/s]
 26%|██▋       | 2830/10682 [23:27<1:04:54,  2.02it/s]
 27%|██▋       | 2831/10682 [23:27<1:04:56,  2.02it/s]
 27%|██▋       | 2832/10682 [23:27<1:04:56,  2.01it/s]
 27%|██▋       | 2833/10682 [23:28<1:04:48,  2.02it/s]
 27%|██▋       | 2834/10682 [23:28<1:04:51,  2.02it/s]
 27%|██▋       | 2835/10682 [23:29<1:04:47,  2.02it/s]
 27%|██▋       | 2836/10682 [23:29<1:04:48,  2.02it/s]
 27%|██▋       | 2837/10682 [23:30<1:04:44,  2.02it/s]
 27%|██▋       | 2838/10682 [23:30<1:04:45,  2.02it/s]
 27%|██▋       | 2839/10682 [23:31<1:04:44,  2.02it/s]
 27%|██▋       | 2840/10682 [23:31<1:04:45,  2.02it/s]
 27%|██▋       | 2841/10682 [23:32<1:04:43,  2.02it/s]
 27%|██▋       | 2842/10682 [23:32<1:04:44,  2.02it/s]
 27%|██▋       | 2843/10682 [23:33<1:04:42,  2.02it/s]
 27%|██▋       | 2844/10682 [23:33<1:04:40,  2.02it/s]
 27%|██▋       | 2845/10682 [23:34<1:04:40,  2.02it/s]
 27%|██▋       | 2846/10682 [23:34<1:04:37,  2.02it/s]
 27%|██▋       | 2847/10682 [23:35<1:04:37,  2.02it/s]
 27%|██▋       | 2848/10682 [23:35<1:04:37,  2.02it/s]
 27%|██▋       | 2849/10682 [23:36<1:04:46,  2.02it/s]
 27%|██▋       | 2850/10682 [23:36<1:04:41,  2.02it/s]
                                                      {'loss': 3.786, 'grad_norm': 0.22953881323337555, 'learning_rate': 0.0009176707273677235, 'epoch': 0.27}
-
 27%|██▋       | 2850/10682 [23:36<1:04:41,  2.02it/s]
 27%|██▋       | 2851/10682 [23:37<1:04:47,  2.01it/s]
 27%|██▋       | 2852/10682 [23:37<1:04:43,  2.02it/s]
 27%|██▋       | 2853/10682 [23:38<1:04:41,  2.02it/s]
 27%|██▋       | 2854/10682 [23:38<1:04:35,  2.02it/s]
 27%|██▋       | 2855/10682 [23:39<1:04:34,  2.02it/s]
 27%|██▋       | 2856/10682 [23:39<1:04:37,  2.02it/s]
 27%|██▋       | 2857/10682 [23:40<1:04:38,  2.02it/s]
 27%|██▋       | 2858/10682 [23:40<1:04:37,  2.02it/s]
 27%|██▋       | 2859/10682 [23:41<1:04:34,  2.02it/s]
 27%|██▋       | 2860/10682 [23:41<1:04:36,  2.02it/s]
 27%|██▋       | 2861/10682 [23:42<1:04:34,  2.02it/s]
 27%|██▋       | 2862/10682 [23:42<1:04:37,  2.02it/s]
 27%|██▋       | 2863/10682 [23:43<1:04:29,  2.02it/s]
 27%|██▋       | 2864/10682 [23:43<1:04:27,  2.02it/s]
 27%|██▋       | 2865/10682 [23:44<1:04:27,  2.02it/s]
 27%|██▋       | 2866/10682 [23:44<1:04:32,  2.02it/s]
 27%|██▋       | 2867/10682 [23:45<1:04:28,  2.02it/s]
 27%|██▋       | 2868/10682 [23:45<1:04:30,  2.02it/s]
 27%|██▋       | 2869/10682 [23:46<1:04:29,  2.02it/s]
 27%|██▋       | 2870/10682 [23:46<1:04:30,  2.02it/s]
 27%|██▋       | 2871/10682 [23:47<1:04:30,  2.02it/s]
 27%|██▋       | 2872/10682 [23:47<1:04:31,  2.02it/s]
 27%|██▋       | 2873/10682 [23:48<1:04:26,  2.02it/s]
 27%|██▋       | 2874/10682 [23:48<1:04:30,  2.02it/s]
 27%|██▋       | 2875/10682 [23:49<1:04:28,  2.02it/s]
                                                      {'loss': 3.7714, 'grad_norm': 0.22673234343528748, 'learning_rate': 0.0009154111134733883, 'epoch': 0.27}
-
 27%|██▋       | 2875/10682 [23:49<1:04:28,  2.02it/s]
 27%|██▋       | 2876/10682 [23:49<1:04:28,  2.02it/s]
 27%|██▋       | 2877/10682 [23:50<1:04:30,  2.02it/s]
 27%|██▋       | 2878/10682 [23:50<1:04:24,  2.02it/s]
 27%|██▋       | 2879/10682 [23:51<1:04:25,  2.02it/s]
 27%|██▋       | 2880/10682 [23:51<1:04:20,  2.02it/s]
 27%|██▋       | 2881/10682 [23:52<1:04:19,  2.02it/s]
 27%|██▋       | 2882/10682 [23:52<1:04:22,  2.02it/s]
 27%|██▋       | 2883/10682 [23:53<1:04:20,  2.02it/s]
 27%|██▋       | 2884/10682 [23:53<1:04:15,  2.02it/s]
 27%|██▋       | 2885/10682 [23:54<1:04:14,  2.02it/s]
 27%|██▋       | 2886/10682 [23:54<1:04:14,  2.02it/s]
 27%|██▋       | 2887/10682 [23:55<1:04:16,  2.02it/s]
 27%|██▋       | 2888/10682 [23:55<1:04:11,  2.02it/s]
 27%|██▋       | 2889/10682 [23:56<1:04:15,  2.02it/s]
 27%|██▋       | 2890/10682 [23:56<1:04:13,  2.02it/s]
 27%|██▋       | 2891/10682 [23:57<1:04:30,  2.01it/s]
 27%|██▋       | 2892/10682 [23:57<1:04:24,  2.02it/s]
 27%|██▋       | 2893/10682 [23:58<1:04:22,  2.02it/s]
 27%|██▋       | 2894/10682 [23:58<1:04:15,  2.02it/s]
 27%|██▋       | 2895/10682 [23:59<1:04:18,  2.02it/s]
 27%|██▋       | 2896/10682 [23:59<1:04:10,  2.02it/s]
 27%|██▋       | 2897/10682 [24:00<1:04:16,  2.02it/s]
 27%|██▋       | 2898/10682 [24:00<1:04:12,  2.02it/s]
 27%|██▋       | 2899/10682 [24:01<1:04:16,  2.02it/s]
 27%|██▋       | 2900/10682 [24:01<1:04:20,  2.02it/s]
                                                      {'loss': 3.7819, 'grad_norm': 0.20602120459079742, 'learning_rate': 0.0009131237703641679, 'epoch': 0.27}
-
 27%|██▋       | 2900/10682 [24:01<1:04:20,  2.02it/s]
 27%|██▋       | 2901/10682 [24:02<1:04:25,  2.01it/s]
 27%|██▋       | 2902/10682 [24:02<1:04:19,  2.02it/s]
 27%|██▋       | 2903/10682 [24:03<1:04:14,  2.02it/s]
 27%|██▋       | 2904/10682 [24:03<1:04:07,  2.02it/s]
 27%|██▋       | 2905/10682 [24:04<1:04:10,  2.02it/s]
 27%|██▋       | 2906/10682 [24:04<1:04:07,  2.02it/s]
 27%|██▋       | 2907/10682 [24:05<1:04:15,  2.02it/s]
 27%|██▋       | 2908/10682 [24:05<1:04:12,  2.02it/s]
 27%|██▋       | 2909/10682 [24:06<1:04:14,  2.02it/s]
 27%|██▋       | 2910/10682 [24:06<1:04:12,  2.02it/s]
 27%|██▋       | 2911/10682 [24:07<1:04:12,  2.02it/s]
 27%|██▋       | 2912/10682 [24:07<1:04:12,  2.02it/s]
 27%|██▋       | 2913/10682 [24:08<1:04:09,  2.02it/s]
 27%|██▋       | 2914/10682 [24:08<1:04:11,  2.02it/s]
 27%|██▋       | 2915/10682 [24:09<1:04:11,  2.02it/s]
 27%|██▋       | 2916/10682 [24:09<1:04:12,  2.02it/s]
 27%|██▋       | 2917/10682 [24:10<1:04:15,  2.01it/s]
 27%|██▋       | 2918/10682 [24:10<1:04:12,  2.02it/s]
 27%|██▋       | 2919/10682 [24:11<1:04:18,  2.01it/s]
 27%|██▋       | 2920/10682 [24:11<1:04:11,  2.02it/s]
 27%|██▋       | 2921/10682 [24:12<1:04:08,  2.02it/s]
 27%|██▋       | 2922/10682 [24:12<1:04:06,  2.02it/s]
 27%|██▋       | 2923/10682 [24:13<1:04:09,  2.02it/s]
 27%|██▋       | 2924/10682 [24:13<1:04:05,  2.02it/s]
 27%|██▋       | 2925/10682 [24:14<1:04:09,  2.02it/s]
                                                      
{'loss': 3.7717, 'grad_norm': 0.21566370129585266, 'learning_rate': 0.0009108088507230948, 'epoch': 0.27}
-
 27%|██▋       | 2925/10682 [24:14<1:04:09,  2.02it/s]
 27%|██▋       | 2926/10682 [24:14<1:04:06,  2.02it/s]
 27%|██▋       | 2927/10682 [24:15<1:04:07,  2.02it/s]
 27%|██▋       | 2928/10682 [24:15<1:04:05,  2.02it/s]
 27%|██▋       | 2929/10682 [24:16<1:04:02,  2.02it/s]
 27%|██▋       | 2930/10682 [24:16<1:04:08,  2.01it/s]
 27%|██▋       | 2931/10682 [24:17<1:04:01,  2.02it/s]
 27%|██▋       | 2932/10682 [24:17<1:04:00,  2.02it/s]
 27%|██▋       | 2933/10682 [24:18<1:04:04,  2.02it/s]
 27%|██▋       | 2934/10682 [24:18<1:04:04,  2.02it/s]
 27%|██▋       | 2935/10682 [24:19<1:03:59,  2.02it/s]
 27%|██▋       | 2936/10682 [24:19<1:04:01,  2.02it/s]
 27%|██▋       | 2937/10682 [24:20<1:03:56,  2.02it/s]
 28%|██▊       | 2938/10682 [24:20<1:03:57,  2.02it/s]
 28%|██▊       | 2939/10682 [24:21<1:03:59,  2.02it/s]
 28%|██▊       | 2940/10682 [24:21<1:03:59,  2.02it/s]
 28%|██▊       | 2941/10682 [24:22<1:03:59,  2.02it/s]
 28%|██▊       | 2942/10682 [24:22<1:03:55,  2.02it/s]
 28%|██▊       | 2943/10682 [24:22<1:03:52,  2.02it/s]
 28%|██▊       | 2944/10682 [24:23<1:03:48,  2.02it/s]
 28%|██▊       | 2945/10682 [24:23<1:03:47,  2.02it/s]
 28%|██▊       | 2946/10682 [24:24<1:03:49,  2.02it/s]
 28%|██▊       | 2947/10682 [24:24<1:03:48,  2.02it/s]
 28%|██▊       | 2948/10682 [24:25<1:03:53,  2.02it/s]
 28%|██▊       | 2949/10682 [24:25<1:03:50,  2.02it/s]
 28%|██▊       | 2950/10682 [24:26<1:03:50,  2.02it/s]
                                                      {'loss': 3.7627, 'grad_norm': 0.2339254468679428, 'learning_rate': 0.0009084665090739699, 'epoch': 0.28}
-
 28%|██▊       | 2950/10682 [24:26<1:03:50,  2.02it/s]
 28%|██▊       | 2951/10682 [24:26<1:03:55,  2.02it/s]
 28%|██▊       | 2952/10682 [24:27<1:03:50,  2.02it/s]
 28%|██▊       | 2953/10682 [24:27<1:03:46,  2.02it/s]
 28%|██▊       | 2954/10682 [24:28<1:03:48,  2.02it/s]
 28%|██▊       | 2955/10682 [24:28<1:03:43,  2.02it/s]
 28%|██▊       | 2956/10682 [24:29<1:03:49,  2.02it/s]
 28%|██▊       | 2957/10682 [24:29<1:03:43,  2.02it/s]
 28%|██▊       | 2958/10682 [24:30<1:03:49,  2.02it/s]
 28%|██▊       | 2959/10682 [24:30<1:03:44,  2.02it/s]
 28%|██▊       | 2960/10682 [24:31<1:03:44,  2.02it/s]
 28%|██▊       | 2961/10682 [24:31<1:03:43,  2.02it/s]
 28%|██▊       | 2962/10682 [24:32<1:03:43,  2.02it/s]
 28%|██▊       | 2963/10682 [24:32<1:03:43,  2.02it/s]
 28%|██▊       | 2964/10682 [24:33<1:03:41,  2.02it/s]
 28%|██▊       | 2965/10682 [24:33<1:03:42,  2.02it/s]
 28%|██▊       | 2966/10682 [24:34<1:03:42,  2.02it/s]
 28%|██▊       | 2967/10682 [24:34<1:03:39,  2.02it/s]
 28%|██▊       | 2968/10682 [24:35<1:03:39,  2.02it/s]
 28%|██▊       | 2969/10682 [24:35<1:03:41,  2.02it/s]
 28%|██▊       | 2970/10682 [24:36<1:03:38,  2.02it/s]
 28%|██▊       | 2971/10682 [24:36<1:03:40,  2.02it/s]
 28%|██▊       | 2972/10682 [24:37<1:03:37,  2.02it/s]
 28%|██▊       | 2973/10682 [24:37<1:03:37,  2.02it/s]
 28%|██▊       | 2974/10682 [24:38<1:03:32,  2.02it/s]
 28%|██▊       | 2975/10682 [24:38<1:03:34,  2.02it/s]
                                                      {'loss': 3.7683, 'grad_norm': 0.2100386917591095, 'learning_rate': 0.0009060969017710476, 'epoch': 0.28}
-
 28%|██▊       | 2975/10682 [24:38<1:03:34,  2.02it/s]
 28%|██▊       | 2976/10682 [24:39<1:03:47,  2.01it/s]
 28%|██▊       | 2977/10682 [24:39<1:03:46,  2.01it/s]
 28%|██▊       | 2978/10682 [24:40<1:03:42,  2.02it/s]
 28%|██▊       | 2979/10682 [24:40<1:03:38,  2.02it/s]
 28%|██▊       | 2980/10682 [24:41<1:03:38,  2.02it/s]
 28%|██▊       | 2981/10682 [24:41<1:03:37,  2.02it/s]
 28%|██▊       | 2982/10682 [24:42<1:03:34,  2.02it/s]
 28%|██▊       | 2983/10682 [24:42<1:03:34,  2.02it/s]
 28%|██▊       | 2984/10682 [24:43<1:03:35,  2.02it/s]
 28%|██▊       | 2985/10682 [24:43<1:03:32,  2.02it/s]
 28%|██▊       | 2986/10682 [24:44<1:03:32,  2.02it/s]
 28%|██▊       | 2987/10682 [24:44<1:03:33,  2.02it/s]
 28%|██▊       | 2988/10682 [24:45<1:03:34,  2.02it/s]
 28%|██▊       | 2989/10682 [24:45<1:03:36,  2.02it/s]
 28%|██▊       | 2990/10682 [24:46<1:03:29,  2.02it/s]
 28%|██▊       | 2991/10682 [24:46<1:03:29,  2.02it/s]
 28%|██▊       | 2992/10682 [24:47<1:03:26,  2.02it/s]
 28%|██▊       | 2993/10682 [24:47<1:03:25,  2.02it/s]
 28%|██▊       | 2994/10682 [24:48<1:03:22,  2.02it/s]
 28%|██▊       | 2995/10682 [24:48<1:03:23,  2.02it/s]
 28%|██▊       | 2996/10682 [24:49<1:03:27,  2.02it/s]
 28%|██▊       | 2997/10682 [24:49<1:03:26,  2.02it/s]
 28%|██▊       | 2998/10682 [24:50<1:03:29,  2.02it/s]
 28%|██▊       | 2999/10682 [24:50<1:03:30,  2.02it/s]
 28%|██▊       | 3000/10682 [24:51<1:03:25,  2.02it/s]
                                                      {'loss': 3.7634, 'grad_norm': 0.22977186739444733, 'learning_rate': 0.0009037001869885992, 'epoch': 0.28}
-
 28%|██▊       | 3000/10682 [24:51<1:03:25,  2.02it/s]
 28%|██▊       | 3001/10682 [24:51<1:03:27,  2.02it/s]
 28%|██▊       | 3002/10682 [24:52<1:03:24,  2.02it/s]
 28%|██▊       | 3003/10682 [24:52<1:03:23,  2.02it/s]
 28%|██▊       | 3004/10682 [24:53<1:03:24,  2.02it/s]
 28%|██▊       | 3005/10682 [24:53<1:03:21,  2.02it/s]
 28%|██▊       | 3006/10682 [24:54<1:03:26,  2.02it/s]
 28%|██▊       | 3007/10682 [24:54<1:03:27,  2.02it/s]
 28%|██▊       | 3008/10682 [24:55<1:03:28,  2.02it/s]
 28%|██▊       | 3009/10682 [24:55<1:03:25,  2.02it/s]
 28%|██▊       | 3010/10682 [24:56<1:03:26,  2.02it/s]
 28%|██▊       | 3011/10682 [24:56<1:03:20,  2.02it/s]
 28%|██▊       | 3012/10682 [24:57<1:03:20,  2.02it/s]
 28%|██▊       | 3013/10682 [24:57<1:03:16,  2.02it/s]
 28%|██▊       | 3014/10682 [24:58<1:03:17,  2.02it/s]
 28%|██▊       | 3015/10682 [24:58<1:03:14,  2.02it/s]
 28%|██▊       | 3016/10682 [24:59<1:03:14,  2.02it/s]
 28%|██▊       | 3017/10682 [24:59<1:03:13,  2.02it/s]
 28%|██▊       | 3018/10682 [25:00<1:03:17,  2.02it/s]
 28%|██▊       | 3019/10682 [25:00<1:03:17,  2.02it/s]
 28%|██▊       | 3020/10682 [25:01<1:03:16,  2.02it/s]
 28%|██▊       | 3021/10682 [25:01<1:03:15,  2.02it/s]
 28%|██▊       | 3022/10682 [25:02<1:03:12,  2.02it/s]
 28%|██▊       | 3023/10682 [25:02<1:03:11,  2.02it/s]
 28%|██▊       | 3024/10682 [25:03<1:03:12,  2.02it/s]
 28%|██▊       | 3025/10682 [25:03<1:03:12,  2.02it/s]
                                                      {'loss': 3.7638, 'grad_norm': 0.2221212387084961, 'learning_rate': 0.0009012765247103541, 'epoch': 0.28}
-
 28%|██▊       | 3025/10682 [25:03<1:03:12,  2.02it/s]
 28%|██▊       | 3026/10682 [25:04<1:03:19,  2.02it/s]
 28%|██▊       | 3027/10682 [25:04<1:03:19,  2.01it/s]
 28%|██▊       | 3028/10682 [25:05<1:03:10,  2.02it/s]
 28%|██▊       | 3029/10682 [25:05<1:03:09,  2.02it/s]
 28%|██▊       | 3030/10682 [25:06<1:03:08,  2.02it/s]
 28%|██▊       | 3031/10682 [25:06<1:03:09,  2.02it/s]
 28%|██▊       | 3032/10682 [25:07<1:03:05,  2.02it/s]
 28%|██▊       | 3033/10682 [25:07<1:03:07,  2.02it/s]
 28%|██▊       | 3034/10682 [25:08<1:03:06,  2.02it/s]
 28%|██▊       | 3035/10682 [25:08<1:03:09,  2.02it/s]
 28%|██▊       | 3036/10682 [25:09<1:03:07,  2.02it/s]
 28%|██▊       | 3037/10682 [25:09<1:03:10,  2.02it/s]
 28%|██▊       | 3038/10682 [25:10<1:03:10,  2.02it/s]
 28%|██▊       | 3039/10682 [25:10<1:03:08,  2.02it/s]
 28%|██▊       | 3040/10682 [25:11<1:03:06,  2.02it/s]
 28%|██▊       | 3041/10682 [25:11<1:03:06,  2.02it/s]
 28%|██▊       | 3042/10682 [25:12<1:03:06,  2.02it/s]
 28%|██▊       | 3043/10682 [25:12<1:03:04,  2.02it/s]
 28%|██▊       | 3044/10682 [25:13<1:03:01,  2.02it/s]
 29%|██▊       | 3045/10682 [25:13<1:03:06,  2.02it/s]
 29%|██▊       | 3046/10682 [25:14<1:03:00,  2.02it/s]
 29%|██▊       | 3047/10682 [25:14<1:03:00,  2.02it/s]
 29%|██▊       | 3048/10682 [25:15<1:03:06,  2.02it/s]
 29%|██▊       | 3049/10682 [25:15<1:03:00,  2.02it/s]
 29%|██▊       | 3050/10682 [25:16<1:02:57,  2.02it/s]
                                                      
{'loss': 3.7688, 'grad_norm': 0.2285774052143097, 'learning_rate': 0.000898826076718821, 'epoch': 0.29}
-
 29%|██▊       | 3050/10682 [25:16<1:02:57,  2.02it/s]
 29%|██▊       | 3051/10682 [25:16<1:03:01,  2.02it/s]
 29%|██▊       | 3052/10682 [25:16<1:03:00,  2.02it/s]
 29%|██▊       | 3053/10682 [25:17<1:03:00,  2.02it/s]
 29%|██▊       | 3054/10682 [25:17<1:02:55,  2.02it/s]
 29%|██▊       | 3055/10682 [25:18<1:03:03,  2.02it/s]
 29%|██▊       | 3056/10682 [25:18<1:03:00,  2.02it/s]
 29%|██▊       | 3057/10682 [25:19<1:02:57,  2.02it/s]
 29%|██▊       | 3058/10682 [25:19<1:02:58,  2.02it/s]
 29%|██▊       | 3059/10682 [25:20<1:02:56,  2.02it/s]
 29%|██▊       | 3060/10682 [25:20<1:02:58,  2.02it/s]
 29%|██▊       | 3061/10682 [25:21<1:02:56,  2.02it/s]
 29%|██▊       | 3062/10682 [25:21<1:02:59,  2.02it/s]
 29%|██▊       | 3063/10682 [25:22<1:02:58,  2.02it/s]
 29%|██▊       | 3064/10682 [25:22<1:02:57,  2.02it/s]
 29%|██▊       | 3065/10682 [25:23<1:02:54,  2.02it/s]
 29%|██▊       | 3066/10682 [25:23<1:02:52,  2.02it/s]
 29%|██▊       | 3067/10682 [25:24<1:02:50,  2.02it/s]
 29%|██▊       | 3068/10682 [25:24<1:02:55,  2.02it/s]
 29%|██▊       | 3069/10682 [25:25<1:02:53,  2.02it/s]
 29%|██▊       | 3070/10682 [25:25<1:02:50,  2.02it/s]
 29%|██▊       | 3071/10682 [25:26<1:02:49,  2.02it/s]
 29%|██▉       | 3072/10682 [25:26<1:02:48,  2.02it/s]
 29%|██▉       | 3073/10682 [25:27<1:02:44,  2.02it/s]
 29%|██▉       | 3074/10682 [25:27<1:02:44,  2.02it/s]
 29%|██▉       | 3075/10682 [25:28<1:02:43,  2.02it/s]
                                                      
{'loss': 3.7616, 'grad_norm': 0.20944838225841522, 'learning_rate': 0.0008963490065844889, 'epoch': 0.29}
-
 29%|██▉       | 3075/10682 [25:28<1:02:43,  2.02it/s]
 29%|██▉       | 3076/10682 [25:28<1:02:51,  2.02it/s]
 29%|██▉       | 3077/10682 [25:29<1:02:46,  2.02it/s]
 29%|██▉       | 3078/10682 [25:29<1:02:47,  2.02it/s]
 29%|██▉       | 3079/10682 [25:30<1:02:41,  2.02it/s]
 29%|██▉       | 3080/10682 [25:30<1:02:49,  2.02it/s]
 29%|██▉       | 3081/10682 [25:31<1:02:45,  2.02it/s]
 29%|██▉       | 3082/10682 [25:31<1:02:43,  2.02it/s]
 29%|██▉       | 3083/10682 [25:32<1:02:38,  2.02it/s]
 29%|██▉       | 3084/10682 [25:32<1:02:44,  2.02it/s]
 29%|██▉       | 3085/10682 [25:33<1:02:45,  2.02it/s]
 29%|██▉       | 3086/10682 [25:33<1:02:46,  2.02it/s]
 29%|██▉       | 3087/10682 [25:34<1:02:48,  2.02it/s]
 29%|██▉       | 3088/10682 [25:34<1:02:49,  2.01it/s]
 29%|██▉       | 3089/10682 [25:35<1:02:44,  2.02it/s]
 29%|██▉       | 3090/10682 [25:35<1:02:42,  2.02it/s]
 29%|██▉       | 3091/10682 [25:36<1:02:40,  2.02it/s]
 29%|██▉       | 3092/10682 [25:36<1:02:41,  2.02it/s]
 29%|██▉       | 3093/10682 [25:37<1:02:38,  2.02it/s]
 29%|██▉       | 3094/10682 [25:37<1:02:38,  2.02it/s]
 29%|██▉       | 3095/10682 [25:38<1:02:35,  2.02it/s]
 29%|██▉       | 3096/10682 [25:38<1:02:34,  2.02it/s]
 29%|██▉       | 3097/10682 [25:39<1:02:36,  2.02it/s]
 29%|██▉       | 3098/10682 [25:39<1:02:41,  2.02it/s]
 29%|██▉       | 3099/10682 [25:40<1:02:37,  2.02it/s]
 29%|██▉       | 3100/10682 [25:40<1:02:33,  2.02it/s]
                                                      
{'loss': 3.7455, 'grad_norm': 0.22968703508377075, 'learning_rate': 0.0008938454796549082, 'epoch': 0.29}
-
 29%|██▉       | 3100/10682 [25:40<1:02:33,  2.02it/s]
 29%|██▉       | 3101/10682 [25:41<1:02:37,  2.02it/s]
 29%|██▉       | 3102/10682 [25:41<1:02:37,  2.02it/s]
 29%|██▉       | 3103/10682 [25:42<1:02:34,  2.02it/s]
 29%|██▉       | 3104/10682 [25:42<1:02:37,  2.02it/s]
 29%|██▉       | 3105/10682 [25:43<1:02:38,  2.02it/s]
 29%|██▉       | 3106/10682 [25:43<1:02:49,  2.01it/s]
 29%|██▉       | 3107/10682 [25:44<1:02:45,  2.01it/s]
 29%|██▉       | 3108/10682 [25:44<1:02:41,  2.01it/s]
 29%|██▉       | 3109/10682 [25:45<1:02:38,  2.01it/s]
 29%|██▉       | 3110/10682 [25:45<1:02:33,  2.02it/s]
 29%|██▉       | 3111/10682 [25:46<1:02:32,  2.02it/s]
 29%|██▉       | 3112/10682 [25:46<1:02:28,  2.02it/s]
 29%|██▉       | 3113/10682 [25:47<1:02:33,  2.02it/s]
 29%|██▉       | 3114/10682 [25:47<1:02:40,  2.01it/s]
 29%|██▉       | 3115/10682 [25:48<1:02:34,  2.02it/s]
 29%|██▉       | 3116/10682 [25:48<1:02:33,  2.02it/s]
 29%|██▉       | 3117/10682 [25:49<1:02:29,  2.02it/s]
 29%|██▉       | 3118/10682 [25:49<1:02:30,  2.02it/s]
 29%|██▉       | 3119/10682 [25:50<1:02:26,  2.02it/s]
 29%|██▉       | 3120/10682 [25:50<1:02:27,  2.02it/s]
 29%|██▉       | 3121/10682 [25:51<1:02:23,  2.02it/s]
 29%|██▉       | 3122/10682 [25:51<1:02:27,  2.02it/s]
 29%|██▉       | 3123/10682 [25:52<1:02:21,  2.02it/s]
 29%|██▉       | 3124/10682 [25:52<1:02:32,  2.01it/s]
 29%|██▉       | 3125/10682 [25:53<1:02:28,  2.02it/s]
                                                      {'loss': 3.7466, 'grad_norm': 0.22021372616291046, 'learning_rate': 0.000891315663043654, 'epoch': 0.29}
-
 29%|██▉       | 3125/10682 [25:53<1:02:28,  2.02it/s]
 29%|██▉       | 3126/10682 [25:53<1:02:33,  2.01it/s]
 29%|██▉       | 3127/10682 [25:54<1:02:33,  2.01it/s]
 29%|██▉       | 3128/10682 [25:54<1:02:24,  2.02it/s]
 29%|██▉       | 3129/10682 [25:55<1:02:21,  2.02it/s]
 29%|██▉       | 3130/10682 [25:55<1:02:19,  2.02it/s]
 29%|██▉       | 3131/10682 [25:56<1:02:20,  2.02it/s]
 29%|██▉       | 3132/10682 [25:56<1:02:17,  2.02it/s]
 29%|██▉       | 3133/10682 [25:57<1:02:19,  2.02it/s]
 29%|██▉       | 3134/10682 [25:57<1:02:16,  2.02it/s]
 29%|██▉       | 3135/10682 [25:58<1:02:20,  2.02it/s]
 29%|██▉       | 3136/10682 [25:58<1:02:16,  2.02it/s]
 29%|██▉       | 3137/10682 [25:59<1:02:18,  2.02it/s]
 29%|██▉       | 3138/10682 [25:59<1:02:14,  2.02it/s]
 29%|██▉       | 3139/10682 [26:00<1:02:15,  2.02it/s]
 29%|██▉       | 3140/10682 [26:00<1:02:11,  2.02it/s]
 29%|██▉       | 3141/10682 [26:01<1:02:14,  2.02it/s]
 29%|██▉       | 3142/10682 [26:01<1:02:09,  2.02it/s]
 29%|██▉       | 3143/10682 [26:02<1:02:10,  2.02it/s]
 29%|██▉       | 3144/10682 [26:02<1:02:09,  2.02it/s]
 29%|██▉       | 3145/10682 [26:03<1:02:10,  2.02it/s]
 29%|██▉       | 3146/10682 [26:03<1:02:22,  2.01it/s]
 29%|██▉       | 3147/10682 [26:04<1:02:18,  2.02it/s]
 29%|██▉       | 3148/10682 [26:04<1:02:13,  2.02it/s]
 29%|██▉       | 3149/10682 [26:05<1:02:06,  2.02it/s]
 29%|██▉       | 3150/10682 [26:05<1:02:07,  2.02it/s]
                                                      
{'loss': 3.7421, 'grad_norm': 0.22233934700489044, 'learning_rate': 0.0008887597256191701, 'epoch': 0.29}
-
 29%|██▉       | 3150/10682 [26:05<1:02:07,  2.02it/s]
 29%|██▉       | 3151/10682 [26:06<1:02:08,  2.02it/s]
 30%|██▉       | 3152/10682 [26:06<1:02:08,  2.02it/s]
 30%|██▉       | 3153/10682 [26:07<1:02:04,  2.02it/s]
 30%|██▉       | 3154/10682 [26:07<1:02:10,  2.02it/s]
 30%|██▉       | 3155/10682 [26:08<1:02:04,  2.02it/s]
 30%|██▉       | 3156/10682 [26:08<1:02:06,  2.02it/s]
 30%|██▉       | 3157/10682 [26:09<1:02:04,  2.02it/s]
 30%|██▉       | 3158/10682 [26:09<1:02:04,  2.02it/s]
 30%|██▉       | 3159/10682 [26:10<1:02:02,  2.02it/s]
 30%|██▉       | 3160/10682 [26:10<1:02:02,  2.02it/s]
 30%|██▉       | 3161/10682 [26:11<1:02:02,  2.02it/s]
 30%|██▉       | 3162/10682 [26:11<1:02:07,  2.02it/s]
 30%|██▉       | 3163/10682 [26:11<1:01:59,  2.02it/s]
 30%|██▉       | 3164/10682 [26:12<1:02:04,  2.02it/s]
 30%|██▉       | 3165/10682 [26:12<1:02:02,  2.02it/s]
 30%|██▉       | 3166/10682 [26:13<1:02:03,  2.02it/s]
 30%|██▉       | 3167/10682 [26:13<1:02:01,  2.02it/s]
 30%|██▉       | 3168/10682 [26:14<1:02:01,  2.02it/s]
 30%|██▉       | 3169/10682 [26:14<1:02:01,  2.02it/s]
 30%|██▉       | 3170/10682 [26:15<1:02:02,  2.02it/s]
 30%|██▉       | 3171/10682 [26:15<1:01:56,  2.02it/s]
 30%|██▉       | 3172/10682 [26:16<1:02:00,  2.02it/s]
 30%|██▉       | 3173/10682 [26:16<1:01:56,  2.02it/s]
 30%|██▉       | 3174/10682 [26:17<1:01:59,  2.02it/s]
 30%|██▉       | 3175/10682 [26:17<1:01:56,  2.02it/s]
                                                      
{'loss': 3.7459, 'grad_norm': 0.2082512527704239, 'learning_rate': 0.0008861778379934982, 'epoch': 0.3}
-
 30%|██▉       | 3175/10682 [26:17<1:01:56,  2.02it/s]
 30%|██▉       | 3176/10682 [26:18<1:02:01,  2.02it/s]
 30%|██▉       | 3177/10682 [26:18<1:01:59,  2.02it/s]
 30%|██▉       | 3178/10682 [26:19<1:02:02,  2.02it/s]
 30%|██▉       | 3179/10682 [26:19<1:01:56,  2.02it/s]
 30%|██▉       | 3180/10682 [26:20<1:01:55,  2.02it/s]
 30%|██▉       | 3181/10682 [26:20<1:01:53,  2.02it/s]
 30%|██▉       | 3182/10682 [26:21<1:01:57,  2.02it/s]
 30%|██▉       | 3183/10682 [26:21<1:01:53,  2.02it/s]
 30%|██▉       | 3184/10682 [26:22<1:01:55,  2.02it/s]
 30%|██▉       | 3185/10682 [26:23<1:12:13,  1.73it/s]
 30%|██▉       | 3186/10682 [26:23<1:09:04,  1.81it/s]
 30%|██▉       | 3187/10682 [26:24<1:06:54,  1.87it/s]
 30%|██▉       | 3188/10682 [26:24<1:05:25,  1.91it/s]
 30%|██▉       | 3189/10682 [26:25<1:04:17,  1.94it/s]
 30%|██▉       | 3190/10682 [26:25<1:03:30,  1.97it/s]
 30%|██▉       | 3191/10682 [26:26<1:03:01,  1.98it/s]
 30%|██▉       | 3192/10682 [26:26<1:02:35,  1.99it/s]
 30%|██▉       | 3193/10682 [26:27<1:02:25,  2.00it/s]
 30%|██▉       | 3194/10682 [26:27<1:02:14,  2.01it/s]
 30%|██▉       | 3195/10682 [26:28<1:02:03,  2.01it/s]
 30%|██▉       | 3196/10682 [26:28<1:02:00,  2.01it/s]
 30%|██▉       | 3197/10682 [26:29<1:01:57,  2.01it/s]
 30%|██▉       | 3198/10682 [26:29<1:11:59,  1.73it/s]
 30%|██▉       | 3199/10682 [26:30<1:08:54,  1.81it/s]
 30%|██▉       | 3200/10682 [26:30<1:06:41,  1.87it/s]{'loss': 3.73, 'grad_norm': 0.2339678257703781, 'learning_rate': 0.000883570172510888, 'epoch': 0.3}
                                                      
-
 30%|██▉       | 3200/10682 [26:30<1:06:41,  1.87it/s]
 30%|██▉       | 3201/10682 [26:31<1:05:23,  1.91it/s]
 30%|██▉       | 3202/10682 [26:31<1:04:12,  1.94it/s]
 30%|██▉       | 3203/10682 [26:32<1:03:27,  1.96it/s]
 30%|██▉       | 3204/10682 [26:32<1:02:56,  1.98it/s]
 30%|███       | 3205/10682 [26:33<1:02:36,  1.99it/s]
 30%|███       | 3206/10682 [26:33<1:02:22,  2.00it/s]
 30%|███       | 3207/10682 [26:34<1:02:08,  2.00it/s]
 30%|███       | 3208/10682 [26:34<1:01:54,  2.01it/s]
 30%|███       | 3209/10682 [26:35<1:01:52,  2.01it/s]
 30%|███       | 3210/10682 [26:35<1:01:42,  2.02it/s]
 30%|███       | 3211/10682 [26:36<1:01:42,  2.02it/s]
 30%|███       | 3212/10682 [26:36<1:01:35,  2.02it/s]
 30%|███       | 3213/10682 [26:37<1:01:39,  2.02it/s]
 30%|███       | 3214/10682 [26:37<1:01:42,  2.02it/s]
 30%|███       | 3215/10682 [26:38<1:01:41,  2.02it/s]
 30%|███       | 3216/10682 [26:38<1:01:39,  2.02it/s]
 30%|███       | 3217/10682 [26:39<1:01:36,  2.02it/s]
 30%|███       | 3218/10682 [26:39<1:01:33,  2.02it/s]
 30%|███       | 3219/10682 [26:40<1:01:39,  2.02it/s]
 30%|███       | 3220/10682 [26:40<1:01:39,  2.02it/s]
 30%|███       | 3221/10682 [26:41<1:01:35,  2.02it/s]
 30%|███       | 3222/10682 [26:41<1:01:37,  2.02it/s]
 30%|███       | 3223/10682 [26:42<1:01:33,  2.02it/s]
 30%|███       | 3224/10682 [26:42<1:01:35,  2.02it/s]
 30%|███       | 3225/10682 [26:43<1:01:35,  2.02it/s]
                                                      {'loss': 3.7379, 'grad_norm': 0.20953932404518127, 'learning_rate': 0.0008809369032362939, 'epoch': 0.3}
-

 30%|███       | 3225/10682 [26:43<1:01:35,  2.02it/s]
 30%|███       | 3226/10682 [26:43<1:01:36,  2.02it/s]
 30%|███       | 3227/10682 [26:44<1:01:37,  2.02it/s]
 30%|███       | 3228/10682 [26:44<1:01:31,  2.02it/s]
 30%|███       | 3229/10682 [26:45<1:01:30,  2.02it/s]
 30%|███       | 3230/10682 [26:45<1:01:29,  2.02it/s]
 30%|███       | 3231/10682 [26:46<1:01:29,  2.02it/s]
 30%|███       | 3232/10682 [26:46<1:01:30,  2.02it/s]
 30%|███       | 3233/10682 [26:47<1:01:28,  2.02it/s]
 30%|███       | 3234/10682 [26:47<1:01:29,  2.02it/s]
 30%|███       | 3235/10682 [26:48<1:01:34,  2.02it/s]
 30%|███       | 3236/10682 [26:48<1:01:30,  2.02it/s]
 30%|███       | 3237/10682 [26:49<1:01:28,  2.02it/s]
 30%|███       | 3238/10682 [26:49<1:01:28,  2.02it/s]
 30%|███       | 3239/10682 [26:50<1:01:25,  2.02it/s]
 30%|███       | 3240/10682 [26:50<1:01:28,  2.02it/s]
 30%|███       | 3241/10682 [26:51<1:01:25,  2.02it/s]
 30%|███       | 3242/10682 [26:51<1:01:27,  2.02it/s]
 30%|███       | 3243/10682 [26:52<1:01:27,  2.02it/s]
 30%|███       | 3244/10682 [26:52<1:01:21,  2.02it/s]
 30%|███       | 3245/10682 [26:53<1:01:22,  2.02it/s]
 30%|███       | 3246/10682 [26:53<1:01:23,  2.02it/s]
 30%|███       | 3247/10682 [26:54<1:01:21,  2.02it/s]
 30%|███       | 3248/10682 [26:54<1:01:23,  2.02it/s]
 30%|███       | 3249/10682 [26:55<1:01:22,  2.02it/s]
 30%|███       | 3250/10682 [26:55<1:01:15,  2.02it/s]
                                                      
{'loss': 3.7412, 'grad_norm': 0.2193393111228943, 'learning_rate': 0.0008782782059437559, 'epoch': 0.3}
-
 30%|███       | 3250/10682 [26:55<1:01:15,  2.02it/s]
 30%|███       | 3251/10682 [26:56<1:01:22,  2.02it/s]
 30%|███       | 3252/10682 [26:56<1:01:18,  2.02it/s]
 30%|███       | 3253/10682 [26:57<1:01:20,  2.02it/s]
 30%|███       | 3254/10682 [26:57<1:01:12,  2.02it/s]
 30%|███       | 3255/10682 [26:58<1:01:21,  2.02it/s]
 30%|███       | 3256/10682 [26:58<1:01:20,  2.02it/s]
 30%|███       | 3257/10682 [26:59<1:01:17,  2.02it/s]
 30%|███       | 3258/10682 [26:59<1:01:12,  2.02it/s]
 31%|███       | 3259/10682 [27:00<1:01:12,  2.02it/s]
 31%|███       | 3260/10682 [27:00<1:01:08,  2.02it/s]
 31%|███       | 3261/10682 [27:01<1:01:11,  2.02it/s]
 31%|███       | 3262/10682 [27:01<1:01:11,  2.02it/s]
 31%|███       | 3263/10682 [27:02<1:01:13,  2.02it/s]
 31%|███       | 3264/10682 [27:02<1:01:09,  2.02it/s]
 31%|███       | 3265/10682 [27:03<1:01:11,  2.02it/s]
 31%|███       | 3266/10682 [27:03<1:01:13,  2.02it/s]
 31%|███       | 3267/10682 [27:04<1:01:12,  2.02it/s]
 31%|███       | 3268/10682 [27:04<1:01:12,  2.02it/s]
 31%|███       | 3269/10682 [27:05<1:01:10,  2.02it/s]
 31%|███       | 3270/10682 [27:05<1:01:14,  2.02it/s]
 31%|███       | 3271/10682 [27:06<1:01:11,  2.02it/s]
 31%|███       | 3272/10682 [27:06<1:01:13,  2.02it/s]
 31%|███       | 3273/10682 [27:07<1:01:08,  2.02it/s]
 31%|███       | 3274/10682 [27:07<1:01:12,  2.02it/s]
 31%|███       | 3275/10682 [27:08<1:01:13,  2.02it/s]
                                                      {'loss': 3.7245, 'grad_norm': 0.2172897905111313, 'learning_rate': 0.0008755942581046657, 'epoch': 0.31}
-
 31%|███       | 3275/10682 [27:08<1:01:13,  2.02it/s]
 31%|███       | 3276/10682 [27:08<1:01:17,  2.01it/s]
 31%|███       | 3277/10682 [27:09<1:01:15,  2.01it/s]
 31%|███       | 3278/10682 [27:09<1:01:13,  2.02it/s]
 31%|███       | 3279/10682 [27:09<1:01:13,  2.02it/s]
 31%|███       | 3280/10682 [27:10<1:01:06,  2.02it/s]
 31%|███       | 3281/10682 [27:10<1:01:05,  2.02it/s]
 31%|███       | 3282/10682 [27:11<1:00:59,  2.02it/s]
 31%|███       | 3283/10682 [27:11<1:01:00,  2.02it/s]
 31%|███       | 3284/10682 [27:12<1:00:55,  2.02it/s]
 31%|███       | 3285/10682 [27:12<1:00:58,  2.02it/s]
 31%|███       | 3286/10682 [27:13<1:00:58,  2.02it/s]
 31%|███       | 3287/10682 [27:13<1:00:58,  2.02it/s]
 31%|███       | 3288/10682 [27:14<1:00:58,  2.02it/s]
 31%|███       | 3289/10682 [27:14<1:01:02,  2.02it/s]
 31%|███       | 3290/10682 [27:15<1:00:56,  2.02it/s]
 31%|███       | 3291/10682 [27:15<1:01:00,  2.02it/s]
 31%|███       | 3292/10682 [27:16<1:00:55,  2.02it/s]
 31%|███       | 3293/10682 [27:16<1:00:56,  2.02it/s]
 31%|███       | 3294/10682 [27:17<1:00:51,  2.02it/s]
 31%|███       | 3295/10682 [27:17<1:01:01,  2.02it/s]
 31%|███       | 3296/10682 [27:18<1:00:54,  2.02it/s]
 31%|███       | 3297/10682 [27:18<1:00:55,  2.02it/s]
 31%|███       | 3298/10682 [27:19<1:00:52,  2.02it/s]
 31%|███       | 3299/10682 [27:19<1:00:51,  2.02it/s]
 31%|███       | 3300/10682 [27:20<1:00:49,  2.02it/s]
                                                      
{'loss': 3.7336, 'grad_norm': 0.22265179455280304, 'learning_rate': 0.0008728852388759212, 'epoch': 0.31}
-
 31%|███       | 3300/10682 [27:20<1:00:49,  2.02it/s]
 31%|███       | 3301/10682 [27:20<1:01:03,  2.01it/s]
 31%|███       | 3302/10682 [27:21<1:00:57,  2.02it/s]
 31%|███       | 3303/10682 [27:21<1:01:01,  2.02it/s]
 31%|███       | 3304/10682 [27:22<1:01:03,  2.01it/s]
 31%|███       | 3305/10682 [27:22<1:01:03,  2.01it/s]
 31%|███       | 3306/10682 [27:23<1:00:59,  2.02it/s]
 31%|███       | 3307/10682 [27:23<1:00:55,  2.02it/s]
 31%|███       | 3308/10682 [27:24<1:00:52,  2.02it/s]
 31%|███       | 3309/10682 [27:24<1:00:52,  2.02it/s]
 31%|███       | 3310/10682 [27:25<1:00:48,  2.02it/s]
 31%|███       | 3311/10682 [27:25<1:00:50,  2.02it/s]
 31%|███       | 3312/10682 [27:26<1:00:44,  2.02it/s]
 31%|███       | 3313/10682 [27:26<1:00:48,  2.02it/s]
 31%|███       | 3314/10682 [27:27<1:00:47,  2.02it/s]
 31%|███       | 3315/10682 [27:27<1:00:50,  2.02it/s]
 31%|███       | 3316/10682 [27:28<1:00:43,  2.02it/s]
 31%|███       | 3317/10682 [27:28<1:00:48,  2.02it/s]
 31%|███       | 3318/10682 [27:29<1:00:43,  2.02it/s]
 31%|███       | 3319/10682 [27:29<1:00:44,  2.02it/s]
 31%|███       | 3320/10682 [27:30<1:00:41,  2.02it/s]
 31%|███       | 3321/10682 [27:30<1:00:43,  2.02it/s]
 31%|███       | 3322/10682 [27:31<1:00:41,  2.02it/s]
 31%|███       | 3323/10682 [27:31<1:00:46,  2.02it/s]
 31%|███       | 3324/10682 [27:32<1:00:52,  2.01it/s]
 31%|███       | 3325/10682 [27:32<1:00:51,  2.01it/s]
                                                      {'loss': 3.7315, 'grad_norm': 0.21774810552597046, 'learning_rate': 0.0008701513290879668, 'epoch': 0.31}
-

 31%|███       | 3325/10682 [27:32<1:00:51,  2.01it/s]
 31%|███       | 3326/10682 [27:33<1:00:53,  2.01it/s]
 31%|███       | 3327/10682 [27:33<1:00:48,  2.02it/s]
 31%|███       | 3328/10682 [27:34<1:00:44,  2.02it/s]
 31%|███       | 3329/10682 [27:34<1:00:41,  2.02it/s]
 31%|███       | 3330/10682 [27:35<1:00:36,  2.02it/s]
 31%|███       | 3331/10682 [27:35<1:00:37,  2.02it/s]
 31%|███       | 3332/10682 [27:36<1:00:34,  2.02it/s]
 31%|███       | 3333/10682 [27:36<1:00:37,  2.02it/s]
 31%|███       | 3334/10682 [27:37<1:00:41,  2.02it/s]
 31%|███       | 3335/10682 [27:37<1:00:38,  2.02it/s]
 31%|███       | 3336/10682 [27:38<1:00:42,  2.02it/s]
 31%|███       | 3337/10682 [27:38<1:00:39,  2.02it/s]
 31%|███       | 3338/10682 [27:39<1:00:39,  2.02it/s]
 31%|███▏      | 3339/10682 [27:39<1:00:34,  2.02it/s]
 31%|███▏      | 3340/10682 [27:40<1:00:42,  2.02it/s]
 31%|███▏      | 3341/10682 [27:40<1:00:33,  2.02it/s]
 31%|███▏      | 3342/10682 [27:41<1:00:39,  2.02it/s]
 31%|███▏      | 3343/10682 [27:41<1:00:33,  2.02it/s]
 31%|███▏      | 3344/10682 [27:42<1:00:38,  2.02it/s]
 31%|███▏      | 3345/10682 [27:42<1:00:39,  2.02it/s]
 31%|███▏      | 3346/10682 [27:43<1:00:35,  2.02it/s]
 31%|███▏      | 3347/10682 [27:43<1:00:33,  2.02it/s]
 31%|███▏      | 3348/10682 [27:44<1:00:33,  2.02it/s]
 31%|███▏      | 3349/10682 [27:44<1:00:31,  2.02it/s]
 31%|███▏      | 3350/10682 [27:45<1:00:29,  2.02it/s]
                                                      {'loss': 3.7252, 'grad_norm': 0.21513047814369202, 'learning_rate': 0.0008673927112327234, 'epoch': 0.31}
-
 31%|███▏      | 3350/10682 [27:45<1:00:29,  2.02it/s]
 31%|███▏      | 3351/10682 [27:45<1:00:30,  2.02it/s]
 31%|███▏      | 3352/10682 [27:46<1:00:35,  2.02it/s]
 31%|███▏      | 3353/10682 [27:46<1:00:31,  2.02it/s]
 31%|███▏      | 3354/10682 [27:47<1:00:33,  2.02it/s]
 31%|███▏      | 3355/10682 [27:47<1:00:29,  2.02it/s]
 31%|███▏      | 3356/10682 [27:48<1:00:27,  2.02it/s]
 31%|███▏      | 3357/10682 [27:48<1:00:28,  2.02it/s]
 31%|███▏      | 3358/10682 [27:49<1:00:25,  2.02it/s]
 31%|███▏      | 3359/10682 [27:49<1:00:25,  2.02it/s]
 31%|███▏      | 3360/10682 [27:50<1:00:27,  2.02it/s]
 31%|███▏      | 3361/10682 [27:50<1:00:24,  2.02it/s]
 31%|███▏      | 3362/10682 [27:51<1:00:31,  2.02it/s]
 31%|███▏      | 3363/10682 [27:51<1:00:24,  2.02it/s]
 31%|███▏      | 3364/10682 [27:52<1:00:34,  2.01it/s]
 32%|███▏      | 3365/10682 [27:52<1:00:36,  2.01it/s]
 32%|███▏      | 3366/10682 [27:53<1:00:31,  2.01it/s]
 32%|███▏      | 3367/10682 [27:53<1:00:25,  2.02it/s]
 32%|███▏      | 3368/10682 [27:54<1:00:27,  2.02it/s]
 32%|███▏      | 3369/10682 [27:54<1:00:26,  2.02it/s]
 32%|███▏      | 3370/10682 [27:55<1:00:25,  2.02it/s]
 32%|███▏      | 3371/10682 [27:55<1:00:21,  2.02it/s]
 32%|███▏      | 3372/10682 [27:56<1:00:23,  2.02it/s]
 32%|███▏      | 3373/10682 [27:56<1:00:20,  2.02it/s]
 32%|███▏      | 3374/10682 [27:57<1:00:22,  2.02it/s]
 32%|███▏      | 3375/10682 [27:57<1:00:23,  2.02it/s]{'loss': 3.7297, 'grad_norm': 0.21629534661769867, 'learning_rate': 0.000864609569451406, 'epoch': 0.32}
                                                      
-
 32%|███▏      | 3375/10682 [27:57<1:00:23,  2.02it/s]
 32%|███▏      | 3376/10682 [27:58<1:00:28,  2.01it/s]
 32%|███▏      | 3377/10682 [27:58<1:00:24,  2.02it/s]
 32%|███▏      | 3378/10682 [27:59<1:00:19,  2.02it/s]
 32%|███▏      | 3379/10682 [27:59<1:00:17,  2.02it/s]
 32%|███▏      | 3380/10682 [28:00<1:00:15,  2.02it/s]
 32%|███▏      | 3381/10682 [28:00<1:00:10,  2.02it/s]
 32%|███▏      | 3382/10682 [28:01<1:00:13,  2.02it/s]
 32%|███▏      | 3383/10682 [28:01<1:00:12,  2.02it/s]
 32%|███▏      | 3384/10682 [28:02<1:00:12,  2.02it/s]
 32%|███▏      | 3385/10682 [28:02<1:00:14,  2.02it/s]
 32%|███▏      | 3386/10682 [28:03<1:00:13,  2.02it/s]
 32%|███▏      | 3387/10682 [28:03<1:00:17,  2.02it/s]
 32%|███▏      | 3388/10682 [28:03<1:00:18,  2.02it/s]
 32%|███▏      | 3389/10682 [28:04<1:00:14,  2.02it/s]
 32%|███▏      | 3390/10682 [28:04<1:00:12,  2.02it/s]
 32%|███▏      | 3391/10682 [28:05<1:00:08,  2.02it/s]
 32%|███▏      | 3392/10682 [28:05<1:00:11,  2.02it/s]
 32%|███▏      | 3393/10682 [28:06<1:00:08,  2.02it/s]
 32%|███▏      | 3394/10682 [28:06<1:00:09,  2.02it/s]
 32%|███▏      | 3395/10682 [28:07<1:00:08,  2.02it/s]
 32%|███▏      | 3396/10682 [28:07<1:00:04,  2.02it/s]
 32%|███▏      | 3397/10682 [28:08<1:00:05,  2.02it/s]
 32%|███▏      | 3398/10682 [28:08<1:00:00,  2.02it/s]
 32%|███▏      | 3399/10682 [28:09<1:00:02,  2.02it/s]
 32%|███▏      | 3400/10682 [28:09<1:00:00,  2.02it/s]
                                                      {'loss': 3.7185, 'grad_norm': 0.22522452473640442, 'learning_rate': 0.0008618020895222331, 'epoch': 0.32}
-
 32%|███▏      | 3400/10682 [28:09<1:00:00,  2.02it/s]
 32%|███▏      | 3401/10682 [28:10<1:00:11,  2.02it/s]
 32%|███▏      | 3402/10682 [28:10<1:00:08,  2.02it/s]
 32%|███▏      | 3403/10682 [28:11<1:00:07,  2.02it/s]
 32%|███▏      | 3404/10682 [28:11<1:00:07,  2.02it/s]
 32%|███▏      | 3405/10682 [28:12<1:00:01,  2.02it/s]
 32%|███▏      | 3406/10682 [28:12<1:00:04,  2.02it/s]
 32%|███▏      | 3407/10682 [28:13<59:59,  2.02it/s]  
 32%|███▏      | 3408/10682 [28:13<1:00:01,  2.02it/s]
 32%|███▏      | 3409/10682 [28:14<59:56,  2.02it/s]  
 32%|███▏      | 3410/10682 [28:14<59:56,  2.02it/s]
 32%|███▏      | 3411/10682 [28:15<59:53,  2.02it/s]
 32%|███▏      | 3412/10682 [28:15<59:58,  2.02it/s]
 32%|███▏      | 3413/10682 [28:16<59:58,  2.02it/s]
 32%|███▏      | 3414/10682 [28:16<59:56,  2.02it/s]
 32%|███▏      | 3415/10682 [28:17<59:54,  2.02it/s]
 32%|███▏      | 3416/10682 [28:17<59:55,  2.02it/s]
 32%|███▏      | 3417/10682 [28:18<59:51,  2.02it/s]
 32%|███▏      | 3418/10682 [28:18<1:00:03,  2.02it/s]
 32%|███▏      | 3419/10682 [28:19<59:57,  2.02it/s]  
 32%|███▏      | 3420/10682 [28:19<59:58,  2.02it/s]
 32%|███▏      | 3421/10682 [28:20<59:54,  2.02it/s]
 32%|███▏      | 3422/10682 [28:20<59:53,  2.02it/s]
 32%|███▏      | 3423/10682 [28:21<59:54,  2.02it/s]
 32%|███▏      | 3424/10682 [28:21<1:00:00,  2.02it/s]
 32%|███▏      | 3425/10682 [28:22<59:59,  2.02it/s]  {'loss': 3.712, 'grad_norm': 0.21404367685317993, 'learning_rate': 0.0008589704588480251, 'epoch': 0.32}
                                                    
-
 32%|███▏      | 3425/10682 [28:22<59:59,  2.02it/s]
 32%|███▏      | 3426/10682 [28:22<1:00:00,  2.02it/s]
 32%|███▏      | 3427/10682 [28:23<59:57,  2.02it/s]  
 32%|███▏      | 3428/10682 [28:23<59:52,  2.02it/s]
 32%|███▏      | 3429/10682 [28:24<59:51,  2.02it/s]
 32%|███▏      | 3430/10682 [28:24<59:46,  2.02it/s]
 32%|███▏      | 3431/10682 [28:25<59:50,  2.02it/s]
 32%|███▏      | 3432/10682 [28:25<59:48,  2.02it/s]
 32%|███▏      | 3433/10682 [28:26<59:49,  2.02it/s]
 32%|███▏      | 3434/10682 [28:26<59:48,  2.02it/s]
 32%|███▏      | 3435/10682 [28:27<59:49,  2.02it/s]
 32%|███▏      | 3436/10682 [28:27<59:49,  2.02it/s]
 32%|███▏      | 3437/10682 [28:28<59:47,  2.02it/s]
 32%|███▏      | 3438/10682 [28:28<59:46,  2.02it/s]
 32%|███▏      | 3439/10682 [28:29<59:45,  2.02it/s]
 32%|███▏      | 3440/10682 [28:29<59:44,  2.02it/s]
 32%|███▏      | 3441/10682 [28:30<59:42,  2.02it/s]
 32%|███▏      | 3442/10682 [28:30<59:40,  2.02it/s]
 32%|███▏      | 3443/10682 [28:31<59:48,  2.02it/s]
 32%|███▏      | 3444/10682 [28:31<59:48,  2.02it/s]
 32%|███▏      | 3445/10682 [28:32<59:45,  2.02it/s]
 32%|███▏      | 3446/10682 [28:32<59:44,  2.02it/s]
 32%|███▏      | 3447/10682 [28:33<59:41,  2.02it/s]
 32%|███▏      | 3448/10682 [28:33<59:42,  2.02it/s]
 32%|███▏      | 3449/10682 [28:34<59:39,  2.02it/s]
 32%|███▏      | 3450/10682 [28:34<59:41,  2.02it/s]
                                                    
{'loss': 3.7118, 'grad_norm': 0.22873269021511078, 'learning_rate': 0.0008561148664436948, 'epoch': 0.32}
-
 32%|███▏      | 3450/10682 [28:34<59:41,  2.02it/s]
 32%|███▏      | 3451/10682 [28:35<59:45,  2.02it/s]
 32%|███▏      | 3452/10682 [28:35<59:42,  2.02it/s]
 32%|███▏      | 3453/10682 [28:36<59:42,  2.02it/s]
 32%|███▏      | 3454/10682 [28:36<59:42,  2.02it/s]
 32%|███▏      | 3455/10682 [28:37<59:37,  2.02it/s]
 32%|███▏      | 3456/10682 [28:37<59:38,  2.02it/s]
 32%|███▏      | 3457/10682 [28:38<59:36,  2.02it/s]
 32%|███▏      | 3458/10682 [28:38<59:35,  2.02it/s]
 32%|███▏      | 3459/10682 [28:39<59:30,  2.02it/s]
 32%|███▏      | 3460/10682 [28:39<59:29,  2.02it/s]
 32%|███▏      | 3461/10682 [28:40<59:28,  2.02it/s]
 32%|███▏      | 3462/10682 [28:40<59:32,  2.02it/s]
 32%|███▏      | 3463/10682 [28:41<59:28,  2.02it/s]
 32%|███▏      | 3464/10682 [28:41<59:33,  2.02it/s]
 32%|███▏      | 3465/10682 [28:42<59:30,  2.02it/s]
 32%|███▏      | 3466/10682 [28:42<59:31,  2.02it/s]
 32%|███▏      | 3467/10682 [28:43<59:26,  2.02it/s]
 32%|███▏      | 3468/10682 [28:43<59:30,  2.02it/s]
 32%|███▏      | 3469/10682 [28:44<59:26,  2.02it/s]
 32%|███▏      | 3470/10682 [28:44<59:28,  2.02it/s]
 32%|███▏      | 3471/10682 [28:45<59:31,  2.02it/s]
 33%|███▎      | 3472/10682 [28:45<59:34,  2.02it/s]
 33%|███▎      | 3473/10682 [28:46<59:30,  2.02it/s]
 33%|███▎      | 3474/10682 [28:46<59:35,  2.02it/s]
 33%|███▎      | 3475/10682 [28:47<59:34,  2.02it/s]
                                                    
{'loss': 3.7062, 'grad_norm': 0.22672916948795319, 'learning_rate': 0.0008532355029236307, 'epoch': 0.33}
-
 33%|███▎      | 3475/10682 [28:47<59:34,  2.02it/s]
 33%|███▎      | 3476/10682 [28:47<59:43,  2.01it/s]
 33%|███▎      | 3477/10682 [28:48<59:38,  2.01it/s]
 33%|███▎      | 3478/10682 [28:48<59:38,  2.01it/s]
 33%|███▎      | 3479/10682 [28:49<59:36,  2.01it/s]
 33%|███▎      | 3480/10682 [28:49<59:38,  2.01it/s]
 33%|███▎      | 3481/10682 [28:50<59:32,  2.02it/s]
 33%|███▎      | 3482/10682 [28:50<59:33,  2.01it/s]
 33%|███▎      | 3483/10682 [28:51<59:39,  2.01it/s]
 33%|███▎      | 3484/10682 [28:51<59:41,  2.01it/s]
 33%|███▎      | 3485/10682 [28:52<59:38,  2.01it/s]
 33%|███▎      | 3486/10682 [28:52<59:31,  2.01it/s]
 33%|███▎      | 3487/10682 [28:53<59:29,  2.02it/s]
 33%|███▎      | 3488/10682 [28:53<59:27,  2.02it/s]
 33%|███▎      | 3489/10682 [28:54<59:21,  2.02it/s]
 33%|███▎      | 3490/10682 [28:54<59:25,  2.02it/s]
 33%|███▎      | 3491/10682 [28:55<59:21,  2.02it/s]
 33%|███▎      | 3492/10682 [28:55<59:23,  2.02it/s]
 33%|███▎      | 3493/10682 [28:56<59:21,  2.02it/s]
 33%|███▎      | 3494/10682 [28:56<59:25,  2.02it/s]
 33%|███▎      | 3495/10682 [28:56<59:22,  2.02it/s]
 33%|███▎      | 3496/10682 [28:57<59:21,  2.02it/s]
 33%|███▎      | 3497/10682 [28:57<59:17,  2.02it/s]
 33%|███▎      | 3498/10682 [28:58<59:16,  2.02it/s]
 33%|███▎      | 3499/10682 [28:58<59:15,  2.02it/s]
 33%|███▎      | 3500/10682 [28:59<59:15,  2.02it/s]
                                                    {'loss': 3.7082, 'grad_norm': 0.21860450506210327, 'learning_rate': 0.0008503325604889736, 'epoch': 0.33}
-

 33%|███▎      | 3500/10682 [28:59<59:15,  2.02it/s]
 33%|███▎      | 3501/10682 [28:59<59:29,  2.01it/s]
 33%|███▎      | 3502/10682 [29:00<59:28,  2.01it/s]
 33%|███▎      | 3503/10682 [29:00<59:24,  2.01it/s]
 33%|███▎      | 3504/10682 [29:01<59:22,  2.02it/s]
 33%|███▎      | 3505/10682 [29:01<59:19,  2.02it/s]
 33%|███▎      | 3506/10682 [29:02<59:18,  2.02it/s]
 33%|███▎      | 3507/10682 [29:02<59:14,  2.02it/s]
 33%|███▎      | 3508/10682 [29:03<59:11,  2.02it/s]
 33%|███▎      | 3509/10682 [29:03<59:10,  2.02it/s]
 33%|███▎      | 3510/10682 [29:04<59:17,  2.02it/s]
 33%|███▎      | 3511/10682 [29:04<59:14,  2.02it/s]
 33%|███▎      | 3512/10682 [29:05<59:18,  2.02it/s]
 33%|███▎      | 3513/10682 [29:05<59:12,  2.02it/s]
 33%|███▎      | 3514/10682 [29:06<59:12,  2.02it/s]
 33%|███▎      | 3515/10682 [29:06<59:12,  2.02it/s]
 33%|███▎      | 3516/10682 [29:07<59:11,  2.02it/s]
 33%|███▎      | 3517/10682 [29:07<59:09,  2.02it/s]
 33%|███▎      | 3518/10682 [29:08<59:08,  2.02it/s]
 33%|███▎      | 3519/10682 [29:08<59:07,  2.02it/s]
 33%|███▎      | 3520/10682 [29:09<59:11,  2.02it/s]
 33%|███▎      | 3521/10682 [29:09<59:11,  2.02it/s]
 33%|███▎      | 3522/10682 [29:10<59:17,  2.01it/s]
 33%|███▎      | 3523/10682 [29:10<59:19,  2.01it/s]
 33%|███▎      | 3524/10682 [29:11<59:15,  2.01it/s]
 33%|███▎      | 3525/10682 [29:11<59:12,  2.01it/s]
                                                    {'loss': 3.7068, 'grad_norm': 0.21592479944229126, 'learning_rate': 0.0008474062329147856, 'epoch': 0.33}
-

 33%|███▎      | 3525/10682 [29:11<59:12,  2.01it/s]
 33%|███▎      | 3526/10682 [29:12<59:15,  2.01it/s]
 33%|███▎      | 3527/10682 [29:12<59:09,  2.02it/s]
 33%|███▎      | 3528/10682 [29:13<59:14,  2.01it/s]
 33%|███▎      | 3529/10682 [29:13<59:08,  2.02it/s]
 33%|███▎      | 3530/10682 [29:14<59:12,  2.01it/s]
 33%|███▎      | 3531/10682 [29:14<59:09,  2.01it/s]
 33%|███▎      | 3532/10682 [29:15<59:10,  2.01it/s]
 33%|███▎      | 3533/10682 [29:15<59:15,  2.01it/s]
 33%|███▎      | 3534/10682 [29:16<59:06,  2.02it/s]
 33%|███▎      | 3535/10682 [29:16<59:04,  2.02it/s]
 33%|███▎      | 3536/10682 [29:17<58:59,  2.02it/s]
 33%|███▎      | 3537/10682 [29:17<58:59,  2.02it/s]
 33%|███▎      | 3538/10682 [29:18<58:54,  2.02it/s]
 33%|███▎      | 3539/10682 [29:18<58:57,  2.02it/s]
 33%|███▎      | 3540/10682 [29:19<58:55,  2.02it/s]
 33%|███▎      | 3541/10682 [29:19<58:57,  2.02it/s]
 33%|███▎      | 3542/10682 [29:20<59:01,  2.02it/s]
 33%|███▎      | 3543/10682 [29:20<58:57,  2.02it/s]
 33%|███▎      | 3544/10682 [29:21<58:59,  2.02it/s]
 33%|███▎      | 3545/10682 [29:21<58:56,  2.02it/s]
 33%|███▎      | 3546/10682 [29:22<58:53,  2.02it/s]
 33%|███▎      | 3547/10682 [29:22<58:51,  2.02it/s]
 33%|███▎      | 3548/10682 [29:23<58:55,  2.02it/s]
 33%|███▎      | 3549/10682 [29:23<58:51,  2.02it/s]
 33%|███▎      | 3550/10682 [29:24<58:49,  2.02it/s]
                                                    {'loss': 3.7107, 'grad_norm': 0.22993148863315582, 'learning_rate': 0.0008444567155371176, 'epoch': 0.33}
-
 33%|███▎      | 3550/10682 [29:24<58:49,  2.02it/s]
 33%|███▎      | 3551/10682 [29:24<58:57,  2.02it/s]
 33%|███▎      | 3552/10682 [29:25<58:58,  2.01it/s]
 33%|███▎      | 3553/10682 [29:25<59:03,  2.01it/s]
 33%|███▎      | 3554/10682 [29:26<58:55,  2.02it/s]
 33%|███▎      | 3555/10682 [29:26<58:56,  2.02it/s]
 33%|███▎      | 3556/10682 [29:27<58:48,  2.02it/s]
 33%|███▎      | 3557/10682 [29:27<58:49,  2.02it/s]
 33%|███▎      | 3558/10682 [29:28<58:45,  2.02it/s]
 33%|███▎      | 3559/10682 [29:28<58:49,  2.02it/s]
 33%|███▎      | 3560/10682 [29:29<58:46,  2.02it/s]
 33%|███▎      | 3561/10682 [29:29<58:44,  2.02it/s]
 33%|███▎      | 3562/10682 [29:30<58:44,  2.02it/s]
 33%|███▎      | 3563/10682 [29:30<58:46,  2.02it/s]
 33%|███▎      | 3564/10682 [29:31<58:42,  2.02it/s]
 33%|███▎      | 3565/10682 [29:31<58:42,  2.02it/s]
 33%|███▎      | 3566/10682 [29:32<58:46,  2.02it/s]
 33%|███▎      | 3567/10682 [29:32<58:49,  2.02it/s]
 33%|███▎      | 3568/10682 [29:33<58:43,  2.02it/s]
 33%|███▎      | 3569/10682 [29:33<58:45,  2.02it/s]
 33%|███▎      | 3570/10682 [29:34<58:39,  2.02it/s]
 33%|███▎      | 3571/10682 [29:34<58:43,  2.02it/s]
 33%|███▎      | 3572/10682 [29:35<58:38,  2.02it/s]
 33%|███▎      | 3573/10682 [29:35<58:41,  2.02it/s]
 33%|███▎      | 3574/10682 [29:36<58:37,  2.02it/s]
 33%|███▎      | 3575/10682 [29:36<58:39,  2.02it/s]
                                                    
{'loss': 3.6961, 'grad_norm': 0.20965367555618286, 'learning_rate': 0.000841484205239968, 'epoch': 0.33}
-
 33%|███▎      | 3575/10682 [29:36<58:39,  2.02it/s]
 33%|███▎      | 3576/10682 [29:37<58:45,  2.02it/s]
 33%|███▎      | 3577/10682 [29:37<58:42,  2.02it/s]
 33%|███▎      | 3578/10682 [29:38<58:37,  2.02it/s]
 34%|███▎      | 3579/10682 [29:38<58:42,  2.02it/s]
 34%|███▎      | 3580/10682 [29:39<58:37,  2.02it/s]
 34%|███▎      | 3581/10682 [29:39<58:42,  2.02it/s]
 34%|███▎      | 3582/10682 [29:40<58:38,  2.02it/s]
 34%|███▎      | 3583/10682 [29:40<58:40,  2.02it/s]
 34%|███▎      | 3584/10682 [29:41<58:35,  2.02it/s]
 34%|███▎      | 3585/10682 [29:41<58:37,  2.02it/s]
 34%|███▎      | 3586/10682 [29:42<58:32,  2.02it/s]
 34%|███▎      | 3587/10682 [29:42<58:34,  2.02it/s]
 34%|███▎      | 3588/10682 [29:43<58:30,  2.02it/s]
 34%|███▎      | 3589/10682 [29:43<58:33,  2.02it/s]
 34%|███▎      | 3590/10682 [29:44<58:29,  2.02it/s]
 34%|███▎      | 3591/10682 [29:44<58:34,  2.02it/s]
 34%|███▎      | 3592/10682 [29:45<58:31,  2.02it/s]
 34%|███▎      | 3593/10682 [29:45<58:33,  2.02it/s]
 34%|███▎      | 3594/10682 [29:46<58:29,  2.02it/s]
 34%|███▎      | 3595/10682 [29:46<58:29,  2.02it/s]
 34%|███▎      | 3596/10682 [29:47<58:28,  2.02it/s]
 34%|███▎      | 3597/10682 [29:47<58:28,  2.02it/s]
 34%|███▎      | 3598/10682 [29:48<58:26,  2.02it/s]
 34%|███▎      | 3599/10682 [29:48<58:29,  2.02it/s]
 34%|███▎      | 3600/10682 [29:49<58:27,  2.02it/s]
                                                    
{'loss': 3.6952, 'grad_norm': 0.21174922585487366, 'learning_rate': 0.0008384889004421422, 'epoch': 0.34}
-
 34%|███▎      | 3600/10682 [29:49<58:27,  2.02it/s]
 34%|███▎      | 3601/10682 [29:49<58:34,  2.01it/s]
 34%|███▎      | 3602/10682 [29:50<58:34,  2.01it/s]
 34%|███▎      | 3603/10682 [29:50<58:31,  2.02it/s]
 34%|███▎      | 3604/10682 [29:51<58:30,  2.02it/s]
 34%|███▎      | 3605/10682 [29:51<58:25,  2.02it/s]
 34%|███▍      | 3606/10682 [29:52<58:27,  2.02it/s]
 34%|███▍      | 3607/10682 [29:52<58:23,  2.02it/s]
 34%|███▍      | 3608/10682 [29:53<58:27,  2.02it/s]
 34%|███▍      | 3609/10682 [29:53<58:25,  2.02it/s]
 34%|███▍      | 3610/10682 [29:53<58:28,  2.02it/s]
 34%|███▍      | 3611/10682 [29:54<58:30,  2.01it/s]
 34%|███▍      | 3612/10682 [29:54<58:23,  2.02it/s]
 34%|███▍      | 3613/10682 [29:55<58:23,  2.02it/s]
 34%|███▍      | 3614/10682 [29:55<58:19,  2.02it/s]
 34%|███▍      | 3615/10682 [29:56<58:20,  2.02it/s]
 34%|███▍      | 3616/10682 [29:56<58:15,  2.02it/s]
 34%|███▍      | 3617/10682 [29:57<58:17,  2.02it/s]
 34%|███▍      | 3618/10682 [29:57<58:16,  2.02it/s]
 34%|███▍      | 3619/10682 [29:58<58:20,  2.02it/s]
 34%|███▍      | 3620/10682 [29:58<58:14,  2.02it/s]
 34%|███▍      | 3621/10682 [29:59<58:19,  2.02it/s]
 34%|███▍      | 3622/10682 [29:59<58:16,  2.02it/s]
 34%|███▍      | 3623/10682 [30:00<58:18,  2.02it/s]
 34%|███▍      | 3624/10682 [30:00<58:18,  2.02it/s]
 34%|███▍      | 3625/10682 [30:01<58:18,  2.02it/s]
                                                    {'loss': 3.6965, 'grad_norm': 0.21652856469154358, 'learning_rate': 0.0008354710010840066, 'epoch': 0.34}
-
 34%|███▍      | 3625/10682 [30:01<58:18,  2.02it/s]
 34%|███▍      | 3626/10682 [30:01<58:22,  2.01it/s]
 34%|███▍      | 3627/10682 [30:02<58:18,  2.02it/s]
 34%|███▍      | 3628/10682 [30:02<58:23,  2.01it/s]
 34%|███▍      | 3629/10682 [30:03<58:18,  2.02it/s]
 34%|███▍      | 3630/10682 [30:03<58:20,  2.01it/s]
 34%|███▍      | 3631/10682 [30:04<58:15,  2.02it/s]
 34%|███▍      | 3632/10682 [30:04<58:21,  2.01it/s]
 34%|███▍      | 3633/10682 [30:05<58:15,  2.02it/s]
 34%|███▍      | 3634/10682 [30:05<58:14,  2.02it/s]
 34%|███▍      | 3635/10682 [30:06<58:13,  2.02it/s]
 34%|███▍      | 3636/10682 [30:06<58:08,  2.02it/s]
 34%|███▍      | 3637/10682 [30:07<58:08,  2.02it/s]
 34%|███▍      | 3638/10682 [30:07<58:08,  2.02it/s]
 34%|███▍      | 3639/10682 [30:08<58:11,  2.02it/s]
 34%|███▍      | 3640/10682 [30:08<58:05,  2.02it/s]
 34%|███▍      | 3641/10682 [30:09<58:10,  2.02it/s]
 34%|███▍      | 3642/10682 [30:09<58:05,  2.02it/s]
 34%|███▍      | 3643/10682 [30:10<58:07,  2.02it/s]
 34%|███▍      | 3644/10682 [30:10<58:03,  2.02it/s]
 34%|███▍      | 3645/10682 [30:11<58:05,  2.02it/s]
 34%|███▍      | 3646/10682 [30:11<58:05,  2.02it/s]
 34%|███▍      | 3647/10682 [30:12<58:05,  2.02it/s]
 34%|███▍      | 3648/10682 [30:12<58:02,  2.02it/s]
 34%|███▍      | 3649/10682 [30:13<58:07,  2.02it/s]
 34%|███▍      | 3650/10682 [30:13<58:01,  2.02it/s]
                                                    
{'loss': 3.695, 'grad_norm': 0.23816394805908203, 'learning_rate': 0.0008324307086141438, 'epoch': 0.34}
-
 34%|███▍      | 3650/10682 [30:13<58:01,  2.02it/s]
 34%|███▍      | 3651/10682 [30:14<58:09,  2.01it/s]
 34%|███▍      | 3652/10682 [30:14<58:07,  2.02it/s]
 34%|███▍      | 3653/10682 [30:15<58:06,  2.02it/s]
 34%|███▍      | 3654/10682 [30:15<58:01,  2.02it/s]
 34%|███▍      | 3655/10682 [30:16<58:03,  2.02it/s]
 34%|███▍      | 3656/10682 [30:16<57:57,  2.02it/s]
 34%|███▍      | 3657/10682 [30:17<57:59,  2.02it/s]
 34%|███▍      | 3658/10682 [30:17<58:02,  2.02it/s]
 34%|███▍      | 3659/10682 [30:18<58:06,  2.01it/s]
 34%|███▍      | 3660/10682 [30:18<58:04,  2.02it/s]
 34%|███▍      | 3661/10682 [30:19<57:58,  2.02it/s]
 34%|███▍      | 3662/10682 [30:19<58:00,  2.02it/s]
 34%|███▍      | 3663/10682 [30:20<57:55,  2.02it/s]
 34%|███▍      | 3664/10682 [30:20<57:54,  2.02it/s]
 34%|███▍      | 3665/10682 [30:21<57:51,  2.02it/s]
 34%|███▍      | 3666/10682 [30:21<57:52,  2.02it/s]
 34%|███▍      | 3667/10682 [30:22<57:49,  2.02it/s]
 34%|███▍      | 3668/10682 [30:22<57:53,  2.02it/s]
 34%|███▍      | 3669/10682 [30:23<57:51,  2.02it/s]
 34%|███▍      | 3670/10682 [30:23<57:53,  2.02it/s]
 34%|███▍      | 3671/10682 [30:24<57:49,  2.02it/s]
 34%|███▍      | 3672/10682 [30:24<57:51,  2.02it/s]
 34%|███▍      | 3673/10682 [30:25<57:46,  2.02it/s]
 34%|███▍      | 3674/10682 [30:25<57:47,  2.02it/s]
 34%|███▍      | 3675/10682 [30:26<57:43,  2.02it/s]
                                                    
{'loss': 3.6845, 'grad_norm': 0.22620856761932373, 'learning_rate': 0.000829368225975904, 'epoch': 0.34}
-
 34%|███▍      | 3675/10682 [30:26<57:43,  2.02it/s]
 34%|███▍      | 3676/10682 [30:26<57:54,  2.02it/s]
 34%|███▍      | 3677/10682 [30:27<57:53,  2.02it/s]
 34%|███▍      | 3678/10682 [30:27<57:52,  2.02it/s]
 34%|███▍      | 3679/10682 [30:28<57:54,  2.02it/s]
 34%|███▍      | 3680/10682 [30:28<57:49,  2.02it/s]
 34%|███▍      | 3681/10682 [30:29<57:51,  2.02it/s]
 34%|███▍      | 3682/10682 [30:29<57:43,  2.02it/s]
 34%|███▍      | 3683/10682 [30:30<57:48,  2.02it/s]
 34%|███▍      | 3684/10682 [30:30<57:46,  2.02it/s]
 34%|███▍      | 3685/10682 [30:31<57:46,  2.02it/s]
 35%|███▍      | 3686/10682 [30:31<57:48,  2.02it/s]
 35%|███▍      | 3687/10682 [30:32<57:47,  2.02it/s]
 35%|███▍      | 3688/10682 [30:32<57:51,  2.01it/s]
 35%|███▍      | 3689/10682 [30:33<57:45,  2.02it/s]
 35%|███▍      | 3690/10682 [30:33<57:47,  2.02it/s]
 35%|███▍      | 3691/10682 [30:34<57:42,  2.02it/s]
 35%|███▍      | 3692/10682 [30:34<57:44,  2.02it/s]
 35%|███▍      | 3693/10682 [30:35<57:41,  2.02it/s]
 35%|███▍      | 3694/10682 [30:35<57:41,  2.02it/s]
 35%|███▍      | 3695/10682 [30:36<57:38,  2.02it/s]
 35%|███▍      | 3696/10682 [30:36<57:42,  2.02it/s]
 35%|███▍      | 3697/10682 [30:37<57:43,  2.02it/s]
 35%|███▍      | 3698/10682 [30:37<57:40,  2.02it/s]
 35%|███▍      | 3699/10682 [30:38<57:46,  2.01it/s]
 35%|███▍      | 3700/10682 [30:38<57:38,  2.02it/s]
                                                    {'loss': 3.6887, 'grad_norm': 0.2212211638689041, 'learning_rate': 0.0008262837575938593, 'epoch': 0.35}
-
 35%|███▍      | 3700/10682 [30:38<57:38,  2.02it/s]
 35%|███▍      | 3701/10682 [30:39<57:44,  2.01it/s]
 35%|███▍      | 3702/10682 [30:39<57:38,  2.02it/s]
 35%|███▍      | 3703/10682 [30:40<57:39,  2.02it/s]
 35%|███▍      | 3704/10682 [30:40<57:36,  2.02it/s]
 35%|███▍      | 3705/10682 [30:41<57:39,  2.02it/s]
 35%|███▍      | 3706/10682 [30:41<57:40,  2.02it/s]
 35%|███▍      | 3707/10682 [30:42<57:36,  2.02it/s]
 35%|███▍      | 3708/10682 [30:42<57:35,  2.02it/s]
 35%|███▍      | 3709/10682 [30:43<57:33,  2.02it/s]
 35%|███▍      | 3710/10682 [30:43<57:37,  2.02it/s]
 35%|███▍      | 3711/10682 [30:44<57:38,  2.02it/s]
 35%|███▍      | 3712/10682 [30:44<57:36,  2.02it/s]
 35%|███▍      | 3713/10682 [30:45<57:33,  2.02it/s]
 35%|███▍      | 3714/10682 [30:45<57:30,  2.02it/s]
 35%|███▍      | 3715/10682 [30:46<57:29,  2.02it/s]
 35%|███▍      | 3716/10682 [30:46<57:25,  2.02it/s]
 35%|███▍      | 3717/10682 [30:47<57:28,  2.02it/s]
 35%|███▍      | 3718/10682 [30:47<57:23,  2.02it/s]
 35%|███▍      | 3719/10682 [30:48<57:30,  2.02it/s]
 35%|███▍      | 3720/10682 [30:48<57:25,  2.02it/s]
 35%|███▍      | 3721/10682 [30:48<57:25,  2.02it/s]
 35%|███▍      | 3722/10682 [30:49<57:23,  2.02it/s]
 35%|███▍      | 3723/10682 [30:49<57:25,  2.02it/s]
 35%|███▍      | 3724/10682 [30:50<57:28,  2.02it/s]
 35%|███▍      | 3725/10682 [30:50<57:24,  2.02it/s]{'loss': 3.6875, 'grad_norm': 0.22687648236751556, 'learning_rate': 0.0008231775093601584, 'epoch': 0.35}
                                                    
-
 35%|███▍      | 3725/10682 [30:50<57:24,  2.02it/s]
 35%|███▍      | 3726/10682 [30:51<57:25,  2.02it/s]
 35%|███▍      | 3727/10682 [30:51<57:28,  2.02it/s]
 35%|███▍      | 3728/10682 [30:52<57:30,  2.02it/s]
 35%|███▍      | 3729/10682 [30:52<57:26,  2.02it/s]
 35%|███▍      | 3730/10682 [30:53<57:24,  2.02it/s]
 35%|███▍      | 3731/10682 [30:53<57:27,  2.02it/s]
 35%|███▍      | 3732/10682 [30:54<57:26,  2.02it/s]
 35%|███▍      | 3733/10682 [30:54<57:27,  2.02it/s]
 35%|███▍      | 3734/10682 [30:55<57:24,  2.02it/s]
 35%|███▍      | 3735/10682 [30:55<57:23,  2.02it/s]
 35%|███▍      | 3736/10682 [30:56<57:22,  2.02it/s]
 35%|███▍      | 3737/10682 [30:56<57:20,  2.02it/s]
 35%|███▍      | 3738/10682 [30:57<57:21,  2.02it/s]
 35%|███▌      | 3739/10682 [30:57<57:15,  2.02it/s]
 35%|███▌      | 3740/10682 [30:58<57:19,  2.02it/s]
 35%|███▌      | 3741/10682 [30:58<57:17,  2.02it/s]
 35%|███▌      | 3742/10682 [30:59<57:18,  2.02it/s]
 35%|███▌      | 3743/10682 [30:59<57:12,  2.02it/s]
 35%|███▌      | 3744/10682 [31:00<57:18,  2.02it/s]
 35%|███▌      | 3745/10682 [31:00<57:22,  2.01it/s]
 35%|███▌      | 3746/10682 [31:01<57:21,  2.02it/s]
 35%|███▌      | 3747/10682 [31:01<57:15,  2.02it/s]
 35%|███▌      | 3748/10682 [31:02<57:17,  2.02it/s]
 35%|███▌      | 3749/10682 [31:02<57:15,  2.02it/s]
 35%|███▌      | 3750/10682 [31:03<57:13,  2.02it/s]
                                                    {'loss': 3.6819, 'grad_norm': 0.21655499935150146, 'learning_rate': 0.0008200496886207818, 'epoch': 0.35}
-
 35%|███▌      | 3750/10682 [31:03<57:13,  2.02it/s]
 35%|███▌      | 3751/10682 [31:03<57:24,  2.01it/s]
 35%|███▌      | 3752/10682 [31:04<57:17,  2.02it/s]
 35%|███▌      | 3753/10682 [31:04<57:17,  2.02it/s]
 35%|███▌      | 3754/10682 [31:05<57:09,  2.02it/s]
 35%|███▌      | 3755/10682 [31:05<57:13,  2.02it/s]
 35%|███▌      | 3756/10682 [31:06<57:10,  2.02it/s]
 35%|███▌      | 3757/10682 [31:06<57:09,  2.02it/s]
 35%|███▌      | 3758/10682 [31:07<57:06,  2.02it/s]
 35%|███▌      | 3759/10682 [31:07<57:08,  2.02it/s]
 35%|███▌      | 3760/10682 [31:08<57:03,  2.02it/s]
 35%|███▌      | 3761/10682 [31:08<57:05,  2.02it/s]
 35%|███▌      | 3762/10682 [31:09<57:01,  2.02it/s]
 35%|███▌      | 3763/10682 [31:09<57:08,  2.02it/s]
 35%|███▌      | 3764/10682 [31:10<57:05,  2.02it/s]
 35%|███▌      | 3765/10682 [31:10<57:06,  2.02it/s]
 35%|███▌      | 3766/10682 [31:11<57:04,  2.02it/s]
 35%|███▌      | 3767/10682 [31:11<57:07,  2.02it/s]
 35%|███▌      | 3768/10682 [31:12<57:07,  2.02it/s]
 35%|███▌      | 3769/10682 [31:12<57:07,  2.02it/s]
 35%|███▌      | 3770/10682 [31:13<57:05,  2.02it/s]
 35%|███▌      | 3771/10682 [31:13<57:09,  2.02it/s]
 35%|███▌      | 3772/10682 [31:14<57:04,  2.02it/s]
 35%|███▌      | 3773/10682 [31:14<57:01,  2.02it/s]
 35%|███▌      | 3774/10682 [31:15<57:00,  2.02it/s]
 35%|███▌      | 3775/10682 [31:15<56:57,  2.02it/s]
                                                    
{'loss': 3.6746, 'grad_norm': 0.2192433923482895, 'learning_rate': 0.0008169005041617021, 'epoch': 0.35}
-
 35%|███▌      | 3775/10682 [31:15<56:57,  2.02it/s]
 35%|███▌      | 3776/10682 [31:16<57:21,  2.01it/s]
 35%|███▌      | 3777/10682 [31:16<57:15,  2.01it/s]
 35%|███▌      | 3778/10682 [31:17<57:12,  2.01it/s]
 35%|███▌      | 3779/10682 [31:17<57:07,  2.01it/s]
 35%|███▌      | 3780/10682 [31:18<57:11,  2.01it/s]
 35%|███▌      | 3781/10682 [31:18<57:03,  2.02it/s]
 35%|███▌      | 3782/10682 [31:19<57:05,  2.01it/s]
 35%|███▌      | 3783/10682 [31:19<56:58,  2.02it/s]
 35%|███▌      | 3784/10682 [31:20<57:03,  2.02it/s]
 35%|███▌      | 3785/10682 [31:20<56:57,  2.02it/s]
 35%|███▌      | 3786/10682 [31:21<57:07,  2.01it/s]
 35%|███▌      | 3787/10682 [31:21<57:01,  2.01it/s]
 35%|███▌      | 3788/10682 [31:22<56:59,  2.02it/s]
 35%|███▌      | 3789/10682 [31:22<56:57,  2.02it/s]
 35%|███▌      | 3790/10682 [31:23<56:53,  2.02it/s]
 35%|███▌      | 3791/10682 [31:23<56:56,  2.02it/s]
 35%|███▌      | 3792/10682 [31:24<56:55,  2.02it/s]
 36%|███▌      | 3793/10682 [31:24<56:54,  2.02it/s]
 36%|███▌      | 3794/10682 [31:25<56:53,  2.02it/s]
 36%|███▌      | 3795/10682 [31:25<56:55,  2.02it/s]
 36%|███▌      | 3796/10682 [31:26<56:50,  2.02it/s]
 36%|███▌      | 3797/10682 [31:26<56:50,  2.02it/s]
 36%|███▌      | 3798/10682 [31:27<56:43,  2.02it/s]
 36%|███▌      | 3799/10682 [31:27<56:46,  2.02it/s]
 36%|███▌      | 3800/10682 [31:28<56:41,  2.02it/s]
                                                    
{'loss': 3.6754, 'grad_norm': 0.21803763508796692, 'learning_rate': 0.0008137301661949477, 'epoch': 0.36}
-
 36%|███▌      | 3800/10682 [31:28<56:41,  2.02it/s]
 36%|███▌      | 3801/10682 [31:28<56:49,  2.02it/s]
 36%|███▌      | 3802/10682 [31:29<56:44,  2.02it/s]
 36%|███▌      | 3803/10682 [31:29<56:45,  2.02it/s]
 36%|███▌      | 3804/10682 [31:30<56:39,  2.02it/s]
 36%|███▌      | 3805/10682 [31:30<56:49,  2.02it/s]
 36%|███▌      | 3806/10682 [31:31<56:45,  2.02it/s]
 36%|███▌      | 3807/10682 [31:31<56:47,  2.02it/s]
 36%|███▌      | 3808/10682 [31:32<56:41,  2.02it/s]
 36%|███▌      | 3809/10682 [31:32<56:45,  2.02it/s]
 36%|███▌      | 3810/10682 [31:33<56:45,  2.02it/s]
 36%|███▌      | 3811/10682 [31:33<56:45,  2.02it/s]
 36%|███▌      | 3812/10682 [31:34<56:42,  2.02it/s]
 36%|███▌      | 3813/10682 [31:34<56:45,  2.02it/s]
 36%|███▌      | 3814/10682 [31:35<56:44,  2.02it/s]
 36%|███▌      | 3815/10682 [31:35<56:45,  2.02it/s]
 36%|███▌      | 3816/10682 [31:36<56:51,  2.01it/s]
 36%|███▌      | 3817/10682 [31:36<56:49,  2.01it/s]
 36%|███▌      | 3818/10682 [31:37<56:48,  2.01it/s]
 36%|███▌      | 3819/10682 [31:37<56:44,  2.02it/s]
 36%|███▌      | 3820/10682 [31:38<56:39,  2.02it/s]
 36%|███▌      | 3821/10682 [31:38<56:38,  2.02it/s]
 36%|███▌      | 3822/10682 [31:39<56:36,  2.02it/s]
 36%|███▌      | 3823/10682 [31:39<56:34,  2.02it/s]
 36%|███▌      | 3824/10682 [31:40<56:35,  2.02it/s]
 36%|███▌      | 3825/10682 [31:40<56:47,  2.01it/s]
                                                    
{'loss': 3.6746, 'grad_norm': 0.2136361300945282, 'learning_rate': 0.0008105388863445693, 'epoch': 0.36}
-
 36%|███▌      | 3825/10682 [31:40<56:47,  2.01it/s]
 36%|███▌      | 3826/10682 [31:41<56:50,  2.01it/s]
 36%|███▌      | 3827/10682 [31:41<56:46,  2.01it/s]
 36%|███▌      | 3828/10682 [31:42<56:41,  2.01it/s]
 36%|███▌      | 3829/10682 [31:42<56:38,  2.02it/s]
 36%|███▌      | 3830/10682 [31:43<56:37,  2.02it/s]
 36%|███▌      | 3831/10682 [31:43<56:39,  2.02it/s]
 36%|███▌      | 3832/10682 [31:44<56:38,  2.02it/s]
 36%|███▌      | 3833/10682 [31:44<56:39,  2.01it/s]
 36%|███▌      | 3834/10682 [31:45<56:37,  2.02it/s]
 36%|███▌      | 3835/10682 [31:45<56:37,  2.02it/s]
 36%|███▌      | 3836/10682 [31:46<56:37,  2.01it/s]
 36%|███▌      | 3837/10682 [31:46<56:35,  2.02it/s]
 36%|███▌      | 3838/10682 [31:46<56:31,  2.02it/s]
 36%|███▌      | 3839/10682 [31:47<56:27,  2.02it/s]
 36%|███▌      | 3840/10682 [31:47<56:27,  2.02it/s]
 36%|███▌      | 3841/10682 [31:48<56:25,  2.02it/s]
 36%|███▌      | 3842/10682 [31:48<56:26,  2.02it/s]
 36%|███▌      | 3843/10682 [31:49<56:23,  2.02it/s]
 36%|███▌      | 3844/10682 [31:49<56:30,  2.02it/s]
 36%|███▌      | 3845/10682 [31:50<56:35,  2.01it/s]
 36%|███▌      | 3846/10682 [31:50<56:34,  2.01it/s]
 36%|███▌      | 3847/10682 [31:51<56:38,  2.01it/s]
 36%|███▌      | 3848/10682 [31:51<56:32,  2.01it/s]
 36%|███▌      | 3849/10682 [31:52<56:34,  2.01it/s]
 36%|███▌      | 3850/10682 [31:52<56:29,  2.02it/s]
                                                    {'loss': 3.6736, 'grad_norm': 0.22094599902629852, 'learning_rate': 0.0008073268776325152, 'epoch': 0.36}
-

 36%|███▌      | 3850/10682 [31:52<56:29,  2.02it/s]
 36%|███▌      | 3851/10682 [31:53<56:32,  2.01it/s]
 36%|███▌      | 3852/10682 [31:53<56:27,  2.02it/s]
 36%|███▌      | 3853/10682 [31:54<56:26,  2.02it/s]
 36%|███▌      | 3854/10682 [31:54<56:24,  2.02it/s]
 36%|███▌      | 3855/10682 [31:55<56:30,  2.01it/s]
 36%|███▌      | 3856/10682 [31:55<56:31,  2.01it/s]
 36%|███▌      | 3857/10682 [31:56<56:28,  2.01it/s]
 36%|███▌      | 3858/10682 [31:56<56:26,  2.02it/s]
 36%|███▌      | 3859/10682 [31:57<56:21,  2.02it/s]
 36%|███▌      | 3860/10682 [31:57<56:20,  2.02it/s]
 36%|███▌      | 3861/10682 [31:58<56:19,  2.02it/s]
 36%|███▌      | 3862/10682 [31:58<56:23,  2.02it/s]
 36%|███▌      | 3863/10682 [31:59<56:16,  2.02it/s]
 36%|███▌      | 3864/10682 [31:59<56:20,  2.02it/s]
 36%|███▌      | 3865/10682 [32:00<56:14,  2.02it/s]
 36%|███▌      | 3866/10682 [32:00<56:13,  2.02it/s]
 36%|███▌      | 3867/10682 [32:01<56:09,  2.02it/s]
 36%|███▌      | 3868/10682 [32:01<56:14,  2.02it/s]
 36%|███▌      | 3869/10682 [32:02<56:10,  2.02it/s]
 36%|███▌      | 3870/10682 [32:02<56:14,  2.02it/s]
 36%|███▌      | 3871/10682 [32:03<56:09,  2.02it/s]
 36%|███▌      | 3872/10682 [32:03<56:15,  2.02it/s]
 36%|███▋      | 3873/10682 [32:04<56:13,  2.02it/s]
 36%|███▋      | 3874/10682 [32:04<56:15,  2.02it/s]
 36%|███▋      | 3875/10682 [32:05<56:13,  2.02it/s]
                                                    {'loss': 3.6746, 'grad_norm': 0.2228432148694992, 'learning_rate': 0.0008040943544644116, 'epoch': 0.36}
-

 36%|███▋      | 3875/10682 [32:05<56:13,  2.02it/s]
 36%|███▋      | 3876/10682 [32:05<56:15,  2.02it/s]
 36%|███▋      | 3877/10682 [32:06<56:15,  2.02it/s]
 36%|███▋      | 3878/10682 [32:06<56:10,  2.02it/s]
 36%|███▋      | 3879/10682 [32:07<56:10,  2.02it/s]
 36%|███▋      | 3880/10682 [32:07<56:07,  2.02it/s]
 36%|███▋      | 3881/10682 [32:08<56:08,  2.02it/s]
 36%|███▋      | 3882/10682 [32:08<56:04,  2.02it/s]
 36%|███▋      | 3883/10682 [32:09<56:08,  2.02it/s]
 36%|███▋      | 3884/10682 [32:09<56:07,  2.02it/s]
 36%|███▋      | 3885/10682 [32:10<56:06,  2.02it/s]
 36%|███▋      | 3886/10682 [32:10<56:03,  2.02it/s]
 36%|███▋      | 3887/10682 [32:11<56:04,  2.02it/s]
 36%|███▋      | 3888/10682 [32:11<56:01,  2.02it/s]
 36%|███▋      | 3889/10682 [32:12<56:05,  2.02it/s]
 36%|███▋      | 3890/10682 [32:12<56:07,  2.02it/s]
 36%|███▋      | 3891/10682 [32:13<56:09,  2.02it/s]
 36%|███▋      | 3892/10682 [32:13<56:09,  2.02it/s]
 36%|███▋      | 3893/10682 [32:14<56:08,  2.02it/s]
 36%|███▋      | 3894/10682 [32:14<56:09,  2.01it/s]
 36%|███▋      | 3895/10682 [32:15<56:06,  2.02it/s]
 36%|███▋      | 3896/10682 [32:15<56:08,  2.01it/s]
 36%|███▋      | 3897/10682 [32:16<56:05,  2.02it/s]
 36%|███▋      | 3898/10682 [32:16<56:03,  2.02it/s]
 37%|███▋      | 3899/10682 [32:17<56:01,  2.02it/s]
 37%|███▋      | 3900/10682 [32:17<1:05:17,  1.73it/s]{'loss': 3.6711, 'grad_norm': 0.21361783146858215, 'learning_rate': 0.00080084153261525, 'epoch': 0.37}
                                                      
-
 37%|███▋      | 3900/10682 [32:18<1:05:17,  1.73it/s]
 37%|███▋      | 3901/10682 [32:18<1:02:36,  1.81it/s]
 37%|███▋      | 3902/10682 [32:18<1:00:35,  1.86it/s]
 37%|███▋      | 3903/10682 [32:19<59:14,  1.91it/s]  
 37%|███▋      | 3904/10682 [32:19<58:09,  1.94it/s]
 37%|███▋      | 3905/10682 [32:20<57:32,  1.96it/s]
 37%|███▋      | 3906/10682 [32:20<56:59,  1.98it/s]
 37%|███▋      | 3907/10682 [32:21<56:43,  1.99it/s]
 37%|███▋      | 3908/10682 [32:21<56:24,  2.00it/s]
 37%|███▋      | 3909/10682 [32:22<56:12,  2.01it/s]
 37%|███▋      | 3910/10682 [32:22<56:09,  2.01it/s]
 37%|███▋      | 3911/10682 [32:23<56:02,  2.01it/s]
 37%|███▋      | 3912/10682 [32:23<56:00,  2.01it/s]
 37%|███▋      | 3913/10682 [32:24<55:56,  2.02it/s]
 37%|███▋      | 3914/10682 [32:24<56:02,  2.01it/s]
 37%|███▋      | 3915/10682 [32:25<56:03,  2.01it/s]
 37%|███▋      | 3916/10682 [32:25<55:59,  2.01it/s]
 37%|███▋      | 3917/10682 [32:26<55:58,  2.01it/s]
 37%|███▋      | 3918/10682 [32:27<1:05:05,  1.73it/s]
 37%|███▋      | 3919/10682 [32:27<1:02:15,  1.81it/s]
 37%|███▋      | 3920/10682 [32:28<1:00:19,  1.87it/s]
 37%|███▋      | 3921/10682 [32:28<58:58,  1.91it/s]  
 37%|███▋      | 3922/10682 [32:29<58:06,  1.94it/s]
 37%|███▋      | 3923/10682 [32:29<57:25,  1.96it/s]
 37%|███▋      | 3924/10682 [32:30<57:03,  1.97it/s]
 37%|███▋      | 3925/10682 [32:30<56:40,  1.99it/s]
                                                    
{'loss': 3.6711, 'grad_norm': 0.20562897622585297, 'learning_rate': 0.0007975686292149845, 'epoch': 0.37}
-
 37%|███▋      | 3925/10682 [32:30<56:40,  1.99it/s]
 37%|███▋      | 3926/10682 [32:31<56:30,  1.99it/s]
 37%|███▋      | 3927/10682 [32:31<56:14,  2.00it/s]
 37%|███▋      | 3928/10682 [32:32<56:01,  2.01it/s]
 37%|███▋      | 3929/10682 [32:32<55:56,  2.01it/s]
 37%|███▋      | 3930/10682 [32:33<55:51,  2.01it/s]
 37%|███▋      | 3931/10682 [32:33<55:49,  2.02it/s]
 37%|███▋      | 3932/10682 [32:34<55:48,  2.02it/s]
 37%|███▋      | 3933/10682 [32:34<55:47,  2.02it/s]
 37%|███▋      | 3934/10682 [32:35<55:45,  2.02it/s]
 37%|███▋      | 3935/10682 [32:35<55:45,  2.02it/s]
 37%|███▋      | 3936/10682 [32:36<55:41,  2.02it/s]
 37%|███▋      | 3937/10682 [32:36<55:39,  2.02it/s]
 37%|███▋      | 3938/10682 [32:37<55:40,  2.02it/s]
 37%|███▋      | 3939/10682 [32:37<55:39,  2.02it/s]
 37%|███▋      | 3940/10682 [32:38<55:37,  2.02it/s]
 37%|███▋      | 3941/10682 [32:38<55:36,  2.02it/s]
 37%|███▋      | 3942/10682 [32:39<55:37,  2.02it/s]
 37%|███▋      | 3943/10682 [32:39<55:36,  2.02it/s]
 37%|███▋      | 3944/10682 [32:40<55:36,  2.02it/s]
 37%|███▋      | 3945/10682 [32:40<55:36,  2.02it/s]
 37%|███▋      | 3946/10682 [32:41<55:37,  2.02it/s]
 37%|███▋      | 3947/10682 [32:41<55:33,  2.02it/s]
 37%|███▋      | 3948/10682 [32:42<55:32,  2.02it/s]
 37%|███▋      | 3949/10682 [32:42<55:28,  2.02it/s]
 37%|███▋      | 3950/10682 [32:43<55:31,  2.02it/s]
                                                    {'loss': 3.6613, 'grad_norm': 0.21785502135753632, 'learning_rate': 0.0007942758627340383, 'epoch': 0.37}
-
 37%|███▋      | 3950/10682 [32:43<55:31,  2.02it/s]
 37%|███▋      | 3951/10682 [32:43<55:37,  2.02it/s]
 37%|███▋      | 3952/10682 [32:44<55:39,  2.02it/s]
 37%|███▋      | 3953/10682 [32:44<55:38,  2.02it/s]
 37%|███▋      | 3954/10682 [32:45<55:34,  2.02it/s]
 37%|███▋      | 3955/10682 [32:45<55:33,  2.02it/s]
 37%|███▋      | 3956/10682 [32:46<55:32,  2.02it/s]
 37%|███▋      | 3957/10682 [32:46<55:28,  2.02it/s]
 37%|███▋      | 3958/10682 [32:47<55:28,  2.02it/s]
 37%|███▋      | 3959/10682 [32:47<55:25,  2.02it/s]
 37%|███▋      | 3960/10682 [32:47<55:28,  2.02it/s]
 37%|███▋      | 3961/10682 [32:48<55:27,  2.02it/s]
 37%|███▋      | 3962/10682 [32:48<55:24,  2.02it/s]
 37%|███▋      | 3963/10682 [32:49<55:26,  2.02it/s]
 37%|███▋      | 3964/10682 [32:49<55:25,  2.02it/s]
 37%|███▋      | 3965/10682 [32:50<55:23,  2.02it/s]
 37%|███▋      | 3966/10682 [32:50<55:23,  2.02it/s]
 37%|███▋      | 3967/10682 [32:51<55:20,  2.02it/s]
 37%|███▋      | 3968/10682 [32:51<55:21,  2.02it/s]
 37%|███▋      | 3969/10682 [32:52<55:17,  2.02it/s]
 37%|███▋      | 3970/10682 [32:52<55:20,  2.02it/s]
 37%|███▋      | 3971/10682 [32:53<55:20,  2.02it/s]
 37%|███▋      | 3972/10682 [32:53<55:19,  2.02it/s]
 37%|███▋      | 3973/10682 [32:54<55:23,  2.02it/s]
 37%|███▋      | 3974/10682 [32:54<55:21,  2.02it/s]
 37%|███▋      | 3975/10682 [32:55<55:19,  2.02it/s]
                                                    {'loss': 3.6632, 'grad_norm': 0.2148859053850174, 'learning_rate': 0.00079096345296872, 'epoch': 0.37}
-
 37%|███▋      | 3975/10682 [32:55<55:19,  2.02it/s]
 37%|███▋      | 3976/10682 [32:55<55:26,  2.02it/s]
 37%|███▋      | 3977/10682 [32:56<55:25,  2.02it/s]
 37%|███▋      | 3978/10682 [32:56<55:24,  2.02it/s]
 37%|███▋      | 3979/10682 [32:57<55:23,  2.02it/s]
 37%|███▋      | 3980/10682 [32:57<55:22,  2.02it/s]
 37%|███▋      | 3981/10682 [32:58<55:24,  2.02it/s]
 37%|███▋      | 3982/10682 [32:58<55:20,  2.02it/s]
 37%|███▋      | 3983/10682 [32:59<55:20,  2.02it/s]
 37%|███▋      | 3984/10682 [32:59<55:20,  2.02it/s]
 37%|███▋      | 3985/10682 [33:00<55:20,  2.02it/s]
 37%|███▋      | 3986/10682 [33:00<55:17,  2.02it/s]
 37%|███▋      | 3987/10682 [33:01<55:18,  2.02it/s]
 37%|███▋      | 3988/10682 [33:01<55:15,  2.02it/s]
 37%|███▋      | 3989/10682 [33:02<55:17,  2.02it/s]
 37%|███▋      | 3990/10682 [33:02<55:14,  2.02it/s]
 37%|███▋      | 3991/10682 [33:03<55:18,  2.02it/s]
 37%|███▋      | 3992/10682 [33:03<55:22,  2.01it/s]
 37%|███▋      | 3993/10682 [33:04<55:17,  2.02it/s]
 37%|███▋      | 3994/10682 [33:04<55:16,  2.02it/s]
 37%|███▋      | 3995/10682 [33:05<55:13,  2.02it/s]
 37%|███▋      | 3996/10682 [33:05<55:16,  2.02it/s]
 37%|███▋      | 3997/10682 [33:06<55:17,  2.02it/s]
 37%|███▋      | 3998/10682 [33:06<55:14,  2.02it/s]
 37%|███▋      | 3999/10682 [33:07<55:09,  2.02it/s]
 37%|███▋      | 4000/10682 [33:07<55:12,  2.02it/s]
                                                    
{'loss': 3.6679, 'grad_norm': 0.22423027455806732, 'learning_rate': 0.0007876316210265521, 'epoch': 0.37}
-
 37%|███▋      | 4000/10682 [33:07<55:12,  2.02it/s]
 37%|███▋      | 4001/10682 [33:08<55:23,  2.01it/s]
 37%|███▋      | 4002/10682 [33:08<55:16,  2.01it/s]
 37%|███▋      | 4003/10682 [33:09<55:14,  2.02it/s]
 37%|███▋      | 4004/10682 [33:09<55:10,  2.02it/s]
 37%|███▋      | 4005/10682 [33:10<55:07,  2.02it/s]
 38%|███▊      | 4006/10682 [33:10<55:06,  2.02it/s]
 38%|███▊      | 4007/10682 [33:11<55:09,  2.02it/s]
 38%|███▊      | 4008/10682 [33:11<55:04,  2.02it/s]
 38%|███▊      | 4009/10682 [33:12<55:08,  2.02it/s]
 38%|███▊      | 4010/10682 [33:12<55:04,  2.02it/s]
 38%|███▊      | 4011/10682 [33:13<55:05,  2.02it/s]
 38%|███▊      | 4012/10682 [33:13<55:01,  2.02it/s]
 38%|███▊      | 4013/10682 [33:14<55:03,  2.02it/s]
 38%|███▊      | 4014/10682 [33:14<54:59,  2.02it/s]
 38%|███▊      | 4015/10682 [33:15<55:01,  2.02it/s]
 38%|███▊      | 4016/10682 [33:15<54:57,  2.02it/s]
 38%|███▊      | 4017/10682 [33:16<54:57,  2.02it/s]
 38%|███▊      | 4018/10682 [33:16<54:57,  2.02it/s]
 38%|███▊      | 4019/10682 [33:17<54:57,  2.02it/s]
 38%|███▊      | 4020/10682 [33:17<54:59,  2.02it/s]
 38%|███▊      | 4021/10682 [33:18<54:57,  2.02it/s]
 38%|███▊      | 4022/10682 [33:18<54:56,  2.02it/s]
 38%|███▊      | 4023/10682 [33:19<54:57,  2.02it/s]
 38%|███▊      | 4024/10682 [33:19<54:55,  2.02it/s]
 38%|███▊      | 4025/10682 [33:20<54:54,  2.02it/s]
                                                    {'loss': 3.6555, 'grad_norm': 0.2103266566991806, 'learning_rate': 0.0007842805893115124, 'epoch': 0.38}
-
 38%|███▊      | 4025/10682 [33:20<54:54,  2.02it/s]
 38%|███▊      | 4026/10682 [33:20<54:59,  2.02it/s]
 38%|███▊      | 4027/10682 [33:21<54:57,  2.02it/s]
 38%|███▊      | 4028/10682 [33:21<54:56,  2.02it/s]
 38%|███▊      | 4029/10682 [33:22<54:55,  2.02it/s]
 38%|███▊      | 4030/10682 [33:22<54:52,  2.02it/s]
 38%|███▊      | 4031/10682 [33:23<54:56,  2.02it/s]
 38%|███▊      | 4032/10682 [33:23<54:53,  2.02it/s]
 38%|███▊      | 4033/10682 [33:24<54:54,  2.02it/s]
 38%|███▊      | 4034/10682 [33:24<54:52,  2.02it/s]
 38%|███▊      | 4035/10682 [33:25<54:54,  2.02it/s]
 38%|███▊      | 4036/10682 [33:25<54:49,  2.02it/s]
 38%|███▊      | 4037/10682 [33:26<54:50,  2.02it/s]
 38%|███▊      | 4038/10682 [33:26<54:48,  2.02it/s]
 38%|███▊      | 4039/10682 [33:27<54:47,  2.02it/s]
 38%|███▊      | 4040/10682 [33:27<54:48,  2.02it/s]
 38%|███▊      | 4041/10682 [33:28<54:47,  2.02it/s]
 38%|███▊      | 4042/10682 [33:28<54:47,  2.02it/s]
 38%|███▊      | 4043/10682 [33:29<54:47,  2.02it/s]
 38%|███▊      | 4044/10682 [33:29<54:46,  2.02it/s]
 38%|███▊      | 4045/10682 [33:30<54:44,  2.02it/s]
 38%|███▊      | 4046/10682 [33:30<54:45,  2.02it/s]
 38%|███▊      | 4047/10682 [33:31<54:43,  2.02it/s]
 38%|███▊      | 4048/10682 [33:31<54:43,  2.02it/s]
 38%|███▊      | 4049/10682 [33:32<54:44,  2.02it/s]
 38%|███▊      | 4050/10682 [33:32<54:50,  2.02it/s]
                                                    {'loss': 3.6508, 'grad_norm': 0.2265700250864029, 'learning_rate': 0.0007809105815091867, 'epoch': 0.38}
-
 38%|███▊      | 4050/10682 [33:32<54:50,  2.02it/s]
 38%|███▊      | 4051/10682 [33:33<54:54,  2.01it/s]
 38%|███▊      | 4052/10682 [33:33<54:52,  2.01it/s]
 38%|███▊      | 4053/10682 [33:34<54:43,  2.02it/s]
 38%|███▊      | 4054/10682 [33:34<54:43,  2.02it/s]
 38%|███▊      | 4055/10682 [33:35<54:38,  2.02it/s]
 38%|███▊      | 4056/10682 [33:35<54:40,  2.02it/s]
 38%|███▊      | 4057/10682 [33:36<54:38,  2.02it/s]
 38%|███▊      | 4058/10682 [33:36<54:40,  2.02it/s]
 38%|███▊      | 4059/10682 [33:37<54:40,  2.02it/s]
 38%|███▊      | 4060/10682 [33:37<54:41,  2.02it/s]
 38%|███▊      | 4061/10682 [33:38<54:37,  2.02it/s]
 38%|███▊      | 4062/10682 [33:38<54:40,  2.02it/s]
 38%|███▊      | 4063/10682 [33:39<54:34,  2.02it/s]
 38%|███▊      | 4064/10682 [33:39<54:34,  2.02it/s]
 38%|███▊      | 4065/10682 [33:40<54:27,  2.03it/s]
 38%|███▊      | 4066/10682 [33:40<54:30,  2.02it/s]
 38%|███▊      | 4067/10682 [33:40<54:31,  2.02it/s]
 38%|███▊      | 4068/10682 [33:41<54:35,  2.02it/s]
 38%|███▊      | 4069/10682 [33:41<54:35,  2.02it/s]
 38%|███▊      | 4070/10682 [33:42<54:35,  2.02it/s]
 38%|███▊      | 4071/10682 [33:42<54:32,  2.02it/s]
 38%|███▊      | 4072/10682 [33:43<54:33,  2.02it/s]
 38%|███▊      | 4073/10682 [33:43<54:29,  2.02it/s]
 38%|███▊      | 4074/10682 [33:44<54:34,  2.02it/s]
 38%|███▊      | 4075/10682 [33:44<54:29,  2.02it/s]
                                                    
{'loss': 3.6517, 'grad_norm': 0.21792380511760712, 'learning_rate': 0.0007775218225718394, 'epoch': 0.38}
-
 38%|███▊      | 4075/10682 [33:44<54:29,  2.02it/s]
 38%|███▊      | 4076/10682 [33:45<54:38,  2.01it/s]
 38%|███▊      | 4077/10682 [33:45<54:39,  2.01it/s]
 38%|███▊      | 4078/10682 [33:46<54:34,  2.02it/s]
 38%|███▊      | 4079/10682 [33:46<54:36,  2.02it/s]
 38%|███▊      | 4080/10682 [33:47<54:36,  2.01it/s]
 38%|███▊      | 4081/10682 [33:47<54:33,  2.02it/s]
 38%|███▊      | 4082/10682 [33:48<54:30,  2.02it/s]
 38%|███▊      | 4083/10682 [33:48<54:31,  2.02it/s]
 38%|███▊      | 4084/10682 [33:49<54:26,  2.02it/s]
 38%|███▊      | 4085/10682 [33:49<54:28,  2.02it/s]
 38%|███▊      | 4086/10682 [33:50<54:22,  2.02it/s]
 38%|███▊      | 4087/10682 [33:50<54:27,  2.02it/s]
 38%|███▊      | 4088/10682 [33:51<54:25,  2.02it/s]
 38%|███▊      | 4089/10682 [33:51<54:28,  2.02it/s]
 38%|███▊      | 4090/10682 [33:52<54:28,  2.02it/s]
 38%|███▊      | 4091/10682 [33:52<54:22,  2.02it/s]
 38%|███▊      | 4092/10682 [33:53<54:22,  2.02it/s]
 38%|███▊      | 4093/10682 [33:53<54:22,  2.02it/s]
 38%|███▊      | 4094/10682 [33:54<54:22,  2.02it/s]
 38%|███▊      | 4095/10682 [33:54<54:19,  2.02it/s]
 38%|███▊      | 4096/10682 [33:55<54:19,  2.02it/s]
 38%|███▊      | 4097/10682 [33:55<54:13,  2.02it/s]
 38%|███▊      | 4098/10682 [33:56<54:17,  2.02it/s]
 38%|███▊      | 4099/10682 [33:56<54:13,  2.02it/s]
 38%|███▊      | 4100/10682 [33:57<54:13,  2.02it/s]
                                                    {'loss': 3.6468, 'grad_norm': 0.2177785336971283, 'learning_rate': 0.0007741145387033967, 'epoch': 0.38}
-
 38%|███▊      | 4100/10682 [33:57<54:13,  2.02it/s]
 38%|███▊      | 4101/10682 [33:57<54:17,  2.02it/s]
 38%|███▊      | 4102/10682 [33:58<54:18,  2.02it/s]
 38%|███▊      | 4103/10682 [33:58<54:16,  2.02it/s]
 38%|███▊      | 4104/10682 [33:59<54:18,  2.02it/s]
 38%|███▊      | 4105/10682 [33:59<54:15,  2.02it/s]
 38%|███▊      | 4106/10682 [34:00<54:19,  2.02it/s]
 38%|███▊      | 4107/10682 [34:00<54:17,  2.02it/s]
 38%|███▊      | 4108/10682 [34:01<54:18,  2.02it/s]
 38%|███▊      | 4109/10682 [34:01<54:15,  2.02it/s]
 38%|███▊      | 4110/10682 [34:02<54:14,  2.02it/s]
 38%|███▊      | 4111/10682 [34:02<54:10,  2.02it/s]
 38%|███▊      | 4112/10682 [34:03<54:13,  2.02it/s]
 39%|███▊      | 4113/10682 [34:03<54:09,  2.02it/s]
 39%|███▊      | 4114/10682 [34:04<54:11,  2.02it/s]
 39%|███▊      | 4115/10682 [34:04<54:06,  2.02it/s]
 39%|███▊      | 4116/10682 [34:05<54:09,  2.02it/s]
 39%|███▊      | 4117/10682 [34:05<54:09,  2.02it/s]
 39%|███▊      | 4118/10682 [34:06<54:09,  2.02it/s]
 39%|███▊      | 4119/10682 [34:06<54:05,  2.02it/s]
 39%|███▊      | 4120/10682 [34:07<54:07,  2.02it/s]
 39%|███▊      | 4121/10682 [34:07<54:07,  2.02it/s]
 39%|███▊      | 4122/10682 [34:08<54:06,  2.02it/s]
 39%|███▊      | 4123/10682 [34:08<54:05,  2.02it/s]
 39%|███▊      | 4124/10682 [34:09<54:03,  2.02it/s]
 39%|███▊      | 4125/10682 [34:09<54:04,  2.02it/s]
                                                    
{'loss': 3.6513, 'grad_norm': 0.2168457806110382, 'learning_rate': 0.0007706889573443467, 'epoch': 0.39}
-
 39%|███▊      | 4125/10682 [34:09<54:04,  2.02it/s]
 39%|███▊      | 4126/10682 [34:10<54:13,  2.02it/s]
 39%|███▊      | 4127/10682 [34:10<54:09,  2.02it/s]
 39%|███▊      | 4128/10682 [34:11<54:09,  2.02it/s]
 39%|███▊      | 4129/10682 [34:11<54:04,  2.02it/s]
 39%|███▊      | 4130/10682 [34:12<54:07,  2.02it/s]
 39%|███▊      | 4131/10682 [34:12<54:05,  2.02it/s]
 39%|███▊      | 4132/10682 [34:13<54:07,  2.02it/s]
 39%|███▊      | 4133/10682 [34:13<54:04,  2.02it/s]
 39%|███▊      | 4134/10682 [34:14<54:06,  2.02it/s]
 39%|███▊      | 4135/10682 [34:14<54:03,  2.02it/s]
 39%|███▊      | 4136/10682 [34:15<54:06,  2.02it/s]
 39%|███▊      | 4137/10682 [34:15<54:04,  2.02it/s]
 39%|███▊      | 4138/10682 [34:16<54:08,  2.01it/s]
 39%|███▊      | 4139/10682 [34:16<54:00,  2.02it/s]
 39%|███▉      | 4140/10682 [34:17<53:59,  2.02it/s]
 39%|███▉      | 4141/10682 [34:17<53:55,  2.02it/s]
 39%|███▉      | 4142/10682 [34:18<53:56,  2.02it/s]
 39%|███▉      | 4143/10682 [34:18<53:53,  2.02it/s]
 39%|███▉      | 4144/10682 [34:19<53:54,  2.02it/s]
 39%|███▉      | 4145/10682 [34:19<53:53,  2.02it/s]
 39%|███▉      | 4146/10682 [34:20<53:53,  2.02it/s]
 39%|███▉      | 4147/10682 [34:20<53:52,  2.02it/s]
 39%|███▉      | 4148/10682 [34:21<53:54,  2.02it/s]
 39%|███▉      | 4149/10682 [34:21<53:49,  2.02it/s]
 39%|███▉      | 4150/10682 [34:22<53:52,  2.02it/s]
                                                    
{'loss': 3.6547, 'grad_norm': 0.22465631365776062, 'learning_rate': 0.0007672453071565586, 'epoch': 0.39}
-
 39%|███▉      | 4150/10682 [34:22<53:52,  2.02it/s]
 39%|███▉      | 4151/10682 [34:22<53:57,  2.02it/s]
 39%|███▉      | 4152/10682 [34:23<53:55,  2.02it/s]
 39%|███▉      | 4153/10682 [34:23<53:49,  2.02it/s]
 39%|███▉      | 4154/10682 [34:24<53:51,  2.02it/s]
 39%|███▉      | 4155/10682 [34:24<53:52,  2.02it/s]
 39%|███▉      | 4156/10682 [34:25<53:55,  2.02it/s]
 39%|███▉      | 4157/10682 [34:25<54:06,  2.01it/s]
 39%|███▉      | 4158/10682 [34:26<53:56,  2.02it/s]
 39%|███▉      | 4159/10682 [34:26<53:53,  2.02it/s]
 39%|███▉      | 4160/10682 [34:27<53:49,  2.02it/s]
 39%|███▉      | 4161/10682 [34:27<53:48,  2.02it/s]
 39%|███▉      | 4162/10682 [34:28<53:46,  2.02it/s]
 39%|███▉      | 4163/10682 [34:28<53:48,  2.02it/s]
 39%|███▉      | 4164/10682 [34:29<53:43,  2.02it/s]
 39%|███▉      | 4165/10682 [34:29<53:42,  2.02it/s]
 39%|███▉      | 4166/10682 [34:30<53:42,  2.02it/s]
 39%|███▉      | 4167/10682 [34:30<53:46,  2.02it/s]
 39%|███▉      | 4168/10682 [34:31<53:45,  2.02it/s]
 39%|███▉      | 4169/10682 [34:31<53:43,  2.02it/s]
 39%|███▉      | 4170/10682 [34:32<53:44,  2.02it/s]
 39%|███▉      | 4171/10682 [34:32<53:43,  2.02it/s]
 39%|███▉      | 4172/10682 [34:32<53:45,  2.02it/s]
 39%|███▉      | 4173/10682 [34:33<53:42,  2.02it/s]
 39%|███▉      | 4174/10682 [34:33<53:41,  2.02it/s]
 39%|███▉      | 4175/10682 [34:34<53:40,  2.02it/s]
                                                    {'loss': 3.6529, 'grad_norm': 0.21934100985527039, 'learning_rate': 0.0007637838180080183, 'epoch': 0.39}
-

 39%|███▉      | 4175/10682 [34:34<53:40,  2.02it/s]
 39%|███▉      | 4176/10682 [34:34<53:51,  2.01it/s]
 39%|███▉      | 4177/10682 [34:35<53:44,  2.02it/s]
 39%|███▉      | 4178/10682 [34:35<53:43,  2.02it/s]
 39%|███▉      | 4179/10682 [34:36<53:39,  2.02it/s]
 39%|███▉      | 4180/10682 [34:36<53:42,  2.02it/s]
 39%|███▉      | 4181/10682 [34:37<53:37,  2.02it/s]
 39%|███▉      | 4182/10682 [34:37<53:42,  2.02it/s]
 39%|███▉      | 4183/10682 [34:38<53:37,  2.02it/s]
 39%|███▉      | 4184/10682 [34:38<53:40,  2.02it/s]
 39%|███▉      | 4185/10682 [34:39<53:37,  2.02it/s]
 39%|███▉      | 4186/10682 [34:39<53:42,  2.02it/s]
 39%|███▉      | 4187/10682 [34:40<53:41,  2.02it/s]
 39%|███▉      | 4188/10682 [34:40<53:49,  2.01it/s]
 39%|███▉      | 4189/10682 [34:41<53:43,  2.01it/s]
 39%|███▉      | 4190/10682 [34:41<53:41,  2.02it/s]
 39%|███▉      | 4191/10682 [34:42<53:40,  2.02it/s]
 39%|███▉      | 4192/10682 [34:42<53:38,  2.02it/s]
 39%|███▉      | 4193/10682 [34:43<53:35,  2.02it/s]
 39%|███▉      | 4194/10682 [34:43<53:37,  2.02it/s]
 39%|███▉      | 4195/10682 [34:44<53:32,  2.02it/s]
 39%|███▉      | 4196/10682 [34:44<53:34,  2.02it/s]
 39%|███▉      | 4197/10682 [34:45<53:43,  2.01it/s]
 39%|███▉      | 4198/10682 [34:45<53:37,  2.02it/s]
 39%|███▉      | 4199/10682 [34:46<53:35,  2.02it/s]
 39%|███▉      | 4200/10682 [34:46<53:31,  2.02it/s]{'loss': 3.642, 'grad_norm': 0.21628917753696442, 'learning_rate': 0.0007603047209574854, 'epoch': 0.39}
                                                    
-
 39%|███▉      | 4200/10682 [34:46<53:31,  2.02it/s]
 39%|███▉      | 4201/10682 [34:47<53:35,  2.02it/s]
 39%|███▉      | 4202/10682 [34:47<53:31,  2.02it/s]
 39%|███▉      | 4203/10682 [34:48<53:29,  2.02it/s]
 39%|███▉      | 4204/10682 [34:48<53:29,  2.02it/s]
 39%|███▉      | 4205/10682 [34:49<53:27,  2.02it/s]
 39%|███▉      | 4206/10682 [34:49<53:27,  2.02it/s]
 39%|███▉      | 4207/10682 [34:50<53:28,  2.02it/s]
 39%|███▉      | 4208/10682 [34:50<53:24,  2.02it/s]
 39%|███▉      | 4209/10682 [34:51<53:24,  2.02it/s]
 39%|███▉      | 4210/10682 [34:51<53:22,  2.02it/s]
 39%|███▉      | 4211/10682 [34:52<53:22,  2.02it/s]
 39%|███▉      | 4212/10682 [34:52<53:21,  2.02it/s]
 39%|███▉      | 4213/10682 [34:53<53:22,  2.02it/s]
 39%|███▉      | 4214/10682 [34:53<53:22,  2.02it/s]
 39%|███▉      | 4215/10682 [34:54<53:26,  2.02it/s]
 39%|███▉      | 4216/10682 [34:54<53:30,  2.01it/s]
 39%|███▉      | 4217/10682 [34:55<53:28,  2.01it/s]
 39%|███▉      | 4218/10682 [34:55<53:28,  2.01it/s]
 39%|███▉      | 4219/10682 [34:56<53:22,  2.02it/s]
 40%|███▉      | 4220/10682 [34:56<53:20,  2.02it/s]
 40%|███▉      | 4221/10682 [34:57<53:21,  2.02it/s]
 40%|███▉      | 4222/10682 [34:57<53:17,  2.02it/s]
 40%|███▉      | 4223/10682 [34:58<53:16,  2.02it/s]
 40%|███▉      | 4224/10682 [34:58<53:21,  2.02it/s]
 40%|███▉      | 4225/10682 [34:59<53:21,  2.02it/s]
                                                    
{'loss': 3.6488, 'grad_norm': 0.22614902257919312, 'learning_rate': 0.0007568082482390682, 'epoch': 0.4}
-
 40%|███▉      | 4225/10682 [34:59<53:21,  2.02it/s]
 40%|███▉      | 4226/10682 [34:59<53:23,  2.02it/s]
 40%|███▉      | 4227/10682 [35:00<53:20,  2.02it/s]
 40%|███▉      | 4228/10682 [35:00<53:18,  2.02it/s]
 40%|███▉      | 4229/10682 [35:01<53:19,  2.02it/s]
 40%|███▉      | 4230/10682 [35:01<53:14,  2.02it/s]
 40%|███▉      | 4231/10682 [35:02<53:15,  2.02it/s]
 40%|███▉      | 4232/10682 [35:02<53:12,  2.02it/s]
 40%|███▉      | 4233/10682 [35:03<53:12,  2.02it/s]
 40%|███▉      | 4234/10682 [35:03<53:10,  2.02it/s]
 40%|███▉      | 4235/10682 [35:04<53:16,  2.02it/s]
 40%|███▉      | 4236/10682 [35:04<53:17,  2.02it/s]
 40%|███▉      | 4237/10682 [35:05<53:14,  2.02it/s]
 40%|███▉      | 4238/10682 [35:05<53:12,  2.02it/s]
 40%|███▉      | 4239/10682 [35:06<53:11,  2.02it/s]
 40%|███▉      | 4240/10682 [35:06<53:09,  2.02it/s]
 40%|███▉      | 4241/10682 [35:07<53:09,  2.02it/s]
 40%|███▉      | 4242/10682 [35:07<53:08,  2.02it/s]
 40%|███▉      | 4243/10682 [35:08<53:07,  2.02it/s]
 40%|███▉      | 4244/10682 [35:08<53:08,  2.02it/s]
 40%|███▉      | 4245/10682 [35:09<53:12,  2.02it/s]
 40%|███▉      | 4246/10682 [35:09<53:09,  2.02it/s]
 40%|███▉      | 4247/10682 [35:10<53:10,  2.02it/s]
 40%|███▉      | 4248/10682 [35:10<53:07,  2.02it/s]
 40%|███▉      | 4249/10682 [35:11<53:04,  2.02it/s]
 40%|███▉      | 4250/10682 [35:11<53:07,  2.02it/s]
                                                    
{'loss': 3.6366, 'grad_norm': 0.2132006287574768, 'learning_rate': 0.0007532946332467235, 'epoch': 0.4}
-
 40%|███▉      | 4250/10682 [35:11<53:07,  2.02it/s]
 40%|███▉      | 4251/10682 [35:12<53:06,  2.02it/s]
 40%|███▉      | 4252/10682 [35:12<53:07,  2.02it/s]
 40%|███▉      | 4253/10682 [35:13<53:05,  2.02it/s]
 40%|███▉      | 4254/10682 [35:13<53:08,  2.02it/s]
 40%|███▉      | 4255/10682 [35:14<53:04,  2.02it/s]
 40%|███▉      | 4256/10682 [35:14<53:04,  2.02it/s]
 40%|███▉      | 4257/10682 [35:15<53:02,  2.02it/s]
 40%|███▉      | 4258/10682 [35:15<53:03,  2.02it/s]
 40%|███▉      | 4259/10682 [35:16<53:00,  2.02it/s]
 40%|███▉      | 4260/10682 [35:16<53:04,  2.02it/s]
 40%|███▉      | 4261/10682 [35:17<53:05,  2.02it/s]
 40%|███▉      | 4262/10682 [35:17<53:02,  2.02it/s]
 40%|███▉      | 4263/10682 [35:18<53:06,  2.01it/s]
 40%|███▉      | 4264/10682 [35:18<53:03,  2.02it/s]
 40%|███▉      | 4265/10682 [35:19<53:10,  2.01it/s]
 40%|███▉      | 4266/10682 [35:19<53:05,  2.01it/s]
 40%|███▉      | 4267/10682 [35:20<53:06,  2.01it/s]
 40%|███▉      | 4268/10682 [35:20<53:03,  2.01it/s]
 40%|███▉      | 4269/10682 [35:21<53:00,  2.02it/s]
 40%|███▉      | 4270/10682 [35:21<53:00,  2.02it/s]
 40%|███▉      | 4271/10682 [35:22<52:57,  2.02it/s]
 40%|███▉      | 4272/10682 [35:22<53:01,  2.01it/s]
 40%|████      | 4273/10682 [35:23<52:57,  2.02it/s]
 40%|████      | 4274/10682 [35:23<52:55,  2.02it/s]
 40%|████      | 4275/10682 [35:24<52:53,  2.02it/s]{'loss': 3.6319, 'grad_norm': 0.22606593370437622, 'learning_rate': 0.0007497641105186759, 'epoch': 0.4}
                                                    
-
 40%|████      | 4275/10682 [35:24<52:53,  2.02it/s]
 40%|████      | 4276/10682 [35:24<52:56,  2.02it/s]
 40%|████      | 4277/10682 [35:25<52:52,  2.02it/s]
 40%|████      | 4278/10682 [35:25<52:53,  2.02it/s]
 40%|████      | 4279/10682 [35:26<52:49,  2.02it/s]
 40%|████      | 4280/10682 [35:26<52:49,  2.02it/s]
 40%|████      | 4281/10682 [35:27<52:48,  2.02it/s]
 40%|████      | 4282/10682 [35:27<52:49,  2.02it/s]
 40%|████      | 4283/10682 [35:28<52:47,  2.02it/s]
 40%|████      | 4284/10682 [35:28<52:53,  2.02it/s]
 40%|████      | 4285/10682 [35:29<52:51,  2.02it/s]
 40%|████      | 4286/10682 [35:29<52:50,  2.02it/s]
 40%|████      | 4287/10682 [35:29<52:49,  2.02it/s]
 40%|████      | 4288/10682 [35:30<52:48,  2.02it/s]
 40%|████      | 4289/10682 [35:30<52:49,  2.02it/s]
 40%|████      | 4290/10682 [35:31<52:45,  2.02it/s]
 40%|████      | 4291/10682 [35:31<52:42,  2.02it/s]
 40%|████      | 4292/10682 [35:32<52:41,  2.02it/s]
 40%|████      | 4293/10682 [35:32<52:39,  2.02it/s]
 40%|████      | 4294/10682 [35:33<52:42,  2.02it/s]
 40%|████      | 4295/10682 [35:33<52:40,  2.02it/s]
 40%|████      | 4296/10682 [35:34<52:41,  2.02it/s]
 40%|████      | 4297/10682 [35:34<52:37,  2.02it/s]
 40%|████      | 4298/10682 [35:35<52:42,  2.02it/s]
 40%|████      | 4299/10682 [35:35<52:43,  2.02it/s]
 40%|████      | 4300/10682 [35:36<52:44,  2.02it/s]
                                                    
{'loss': 3.6343, 'grad_norm': 0.2335442304611206, 'learning_rate': 0.0007462169157217631, 'epoch': 0.4}
-
 40%|████      | 4300/10682 [35:36<52:44,  2.02it/s]
 40%|████      | 4301/10682 [35:36<52:49,  2.01it/s]
 40%|████      | 4302/10682 [35:37<52:42,  2.02it/s]
 40%|████      | 4303/10682 [35:37<52:45,  2.02it/s]
 40%|████      | 4304/10682 [35:38<52:38,  2.02it/s]
 40%|████      | 4305/10682 [35:38<52:40,  2.02it/s]
 40%|████      | 4306/10682 [35:39<52:39,  2.02it/s]
 40%|████      | 4307/10682 [35:39<52:38,  2.02it/s]
 40%|████      | 4308/10682 [35:40<52:37,  2.02it/s]
 40%|████      | 4309/10682 [35:40<52:38,  2.02it/s]
 40%|████      | 4310/10682 [35:41<52:32,  2.02it/s]
 40%|████      | 4311/10682 [35:41<52:34,  2.02it/s]
 40%|████      | 4312/10682 [35:42<52:31,  2.02it/s]
 40%|████      | 4313/10682 [35:42<52:31,  2.02it/s]
 40%|████      | 4314/10682 [35:43<52:33,  2.02it/s]
 40%|████      | 4315/10682 [35:43<52:31,  2.02it/s]
 40%|████      | 4316/10682 [35:44<52:34,  2.02it/s]
 40%|████      | 4317/10682 [35:44<52:33,  2.02it/s]
 40%|████      | 4318/10682 [35:45<52:33,  2.02it/s]
 40%|████      | 4319/10682 [35:45<52:30,  2.02it/s]
 40%|████      | 4320/10682 [35:46<52:29,  2.02it/s]
 40%|████      | 4321/10682 [35:46<52:33,  2.02it/s]
 40%|████      | 4322/10682 [35:47<52:35,  2.02it/s]
 40%|████      | 4323/10682 [35:47<52:31,  2.02it/s]
 40%|████      | 4324/10682 [35:48<52:33,  2.02it/s]
 40%|████      | 4325/10682 [35:48<52:34,  2.02it/s]{'loss': 3.6338, 'grad_norm': 0.21274445950984955, 'learning_rate': 0.0007426532856357045, 'epoch': 0.4}
                                                    
-
 40%|████      | 4325/10682 [35:48<52:34,  2.02it/s]
 40%|████      | 4326/10682 [35:49<52:37,  2.01it/s]
 41%|████      | 4327/10682 [35:49<52:34,  2.01it/s]
 41%|████      | 4328/10682 [35:50<52:29,  2.02it/s]
 41%|████      | 4329/10682 [35:50<52:28,  2.02it/s]
 41%|████      | 4330/10682 [35:51<52:28,  2.02it/s]
 41%|████      | 4331/10682 [35:51<52:27,  2.02it/s]
 41%|████      | 4332/10682 [35:52<52:26,  2.02it/s]
 41%|████      | 4333/10682 [35:52<52:30,  2.01it/s]
 41%|████      | 4334/10682 [35:53<52:29,  2.02it/s]
 41%|████      | 4335/10682 [35:53<52:27,  2.02it/s]
 41%|████      | 4336/10682 [35:54<52:33,  2.01it/s]
 41%|████      | 4337/10682 [35:54<52:28,  2.02it/s]
 41%|████      | 4338/10682 [35:55<52:24,  2.02it/s]
 41%|████      | 4339/10682 [35:55<52:22,  2.02it/s]
 41%|████      | 4340/10682 [35:56<52:19,  2.02it/s]
 41%|████      | 4341/10682 [35:56<52:17,  2.02it/s]
 41%|████      | 4342/10682 [35:57<52:19,  2.02it/s]
 41%|████      | 4343/10682 [35:57<52:18,  2.02it/s]
 41%|████      | 4344/10682 [35:58<52:17,  2.02it/s]
 41%|████      | 4345/10682 [35:58<52:19,  2.02it/s]
 41%|████      | 4346/10682 [35:59<52:18,  2.02it/s]
 41%|████      | 4347/10682 [35:59<52:19,  2.02it/s]
 41%|████      | 4348/10682 [36:00<52:18,  2.02it/s]
 41%|████      | 4349/10682 [36:00<52:14,  2.02it/s]
 41%|████      | 4350/10682 [36:01<52:20,  2.02it/s]{'loss': 3.63, 'grad_norm': 0.22769547998905182, 'learning_rate': 0.0007390734581372948, 'epoch': 0.41}
                                                    
-
 41%|████      | 4350/10682 [36:01<52:20,  2.02it/s]
 41%|████      | 4351/10682 [36:01<52:17,  2.02it/s]
 41%|████      | 4352/10682 [36:02<52:21,  2.02it/s]
 41%|████      | 4353/10682 [36:02<52:29,  2.01it/s]
 41%|████      | 4354/10682 [36:03<52:22,  2.01it/s]
 41%|████      | 4355/10682 [36:03<52:21,  2.01it/s]
 41%|████      | 4356/10682 [36:04<52:15,  2.02it/s]
 41%|████      | 4357/10682 [36:04<52:17,  2.02it/s]
 41%|████      | 4358/10682 [36:05<52:12,  2.02it/s]
 41%|████      | 4359/10682 [36:05<52:11,  2.02it/s]
 41%|████      | 4360/10682 [36:06<52:09,  2.02it/s]
 41%|████      | 4361/10682 [36:06<52:09,  2.02it/s]
 41%|████      | 4362/10682 [36:07<52:06,  2.02it/s]
 41%|████      | 4363/10682 [36:07<52:08,  2.02it/s]
 41%|████      | 4364/10682 [36:08<52:08,  2.02it/s]
 41%|████      | 4365/10682 [36:08<52:11,  2.02it/s]
 41%|████      | 4366/10682 [36:09<52:08,  2.02it/s]
 41%|████      | 4367/10682 [36:09<52:06,  2.02it/s]
 41%|████      | 4368/10682 [36:10<52:04,  2.02it/s]
 41%|████      | 4369/10682 [36:10<52:05,  2.02it/s]
 41%|████      | 4370/10682 [36:11<52:07,  2.02it/s]
 41%|████      | 4371/10682 [36:11<52:06,  2.02it/s]
 41%|████      | 4372/10682 [36:12<52:08,  2.02it/s]
 41%|████      | 4373/10682 [36:12<52:04,  2.02it/s]
 41%|████      | 4374/10682 [36:13<52:08,  2.02it/s]
 41%|████      | 4375/10682 [36:13<52:05,  2.02it/s]
                                                    
{'loss': 3.6277, 'grad_norm': 0.21187908947467804, 'learning_rate': 0.0007354776721845277, 'epoch': 0.41}
-
 41%|████      | 4375/10682 [36:13<52:05,  2.02it/s]
 41%|████      | 4376/10682 [36:14<52:11,  2.01it/s]
 41%|████      | 4377/10682 [36:14<52:06,  2.02it/s]
 41%|████      | 4378/10682 [36:15<52:06,  2.02it/s]
 41%|████      | 4379/10682 [36:15<52:07,  2.02it/s]
 41%|████      | 4380/10682 [36:16<52:08,  2.01it/s]
 41%|████      | 4381/10682 [36:16<52:11,  2.01it/s]
 41%|████      | 4382/10682 [36:17<52:12,  2.01it/s]
 41%|████      | 4383/10682 [36:17<52:07,  2.01it/s]
 41%|████      | 4384/10682 [36:18<52:05,  2.02it/s]
 41%|████      | 4385/10682 [36:18<52:04,  2.02it/s]
 41%|████      | 4386/10682 [36:19<52:04,  2.01it/s]
 41%|████      | 4387/10682 [36:19<51:58,  2.02it/s]
 41%|████      | 4388/10682 [36:20<51:57,  2.02it/s]
 41%|████      | 4389/10682 [36:20<51:57,  2.02it/s]
 41%|████      | 4390/10682 [36:21<51:55,  2.02it/s]
 41%|████      | 4391/10682 [36:21<51:57,  2.02it/s]
 41%|████      | 4392/10682 [36:22<52:00,  2.02it/s]
 41%|████      | 4393/10682 [36:22<51:56,  2.02it/s]
 41%|████      | 4394/10682 [36:23<51:59,  2.02it/s]
 41%|████      | 4395/10682 [36:23<51:55,  2.02it/s]
 41%|████      | 4396/10682 [36:24<51:53,  2.02it/s]
 41%|████      | 4397/10682 [36:24<51:50,  2.02it/s]
 41%|████      | 4398/10682 [36:25<51:50,  2.02it/s]
 41%|████      | 4399/10682 [36:25<51:48,  2.02it/s]
 41%|████      | 4400/10682 [36:25<51:52,  2.02it/s]
                                                    {'loss': 3.6319, 'grad_norm': 0.2117864191532135, 'learning_rate': 0.0007318661678006427, 'epoch': 0.41}
-
 41%|████      | 4400/10682 [36:25<51:52,  2.02it/s]
 41%|████      | 4401/10682 [36:26<51:57,  2.01it/s]
 41%|████      | 4402/10682 [36:26<51:55,  2.02it/s]
 41%|████      | 4403/10682 [36:27<51:49,  2.02it/s]
 41%|████      | 4404/10682 [36:27<51:49,  2.02it/s]
 41%|████      | 4405/10682 [36:28<51:45,  2.02it/s]
 41%|████      | 4406/10682 [36:28<51:44,  2.02it/s]
 41%|████▏     | 4407/10682 [36:29<51:45,  2.02it/s]
 41%|████▏     | 4408/10682 [36:29<51:46,  2.02it/s]
 41%|████▏     | 4409/10682 [36:30<51:49,  2.02it/s]
 41%|████▏     | 4410/10682 [36:30<51:50,  2.02it/s]
 41%|████▏     | 4411/10682 [36:31<51:50,  2.02it/s]
 41%|████▏     | 4412/10682 [36:31<51:49,  2.02it/s]
 41%|████▏     | 4413/10682 [36:32<51:49,  2.02it/s]
 41%|████▏     | 4414/10682 [36:32<51:49,  2.02it/s]
 41%|████▏     | 4415/10682 [36:33<51:43,  2.02it/s]
 41%|████▏     | 4416/10682 [36:33<51:49,  2.02it/s]
 41%|████▏     | 4417/10682 [36:34<51:45,  2.02it/s]
 41%|████▏     | 4418/10682 [36:34<51:41,  2.02it/s]
 41%|████▏     | 4419/10682 [36:35<51:43,  2.02it/s]
 41%|████▏     | 4420/10682 [36:35<51:40,  2.02it/s]
 41%|████▏     | 4421/10682 [36:36<51:40,  2.02it/s]
 41%|████▏     | 4422/10682 [36:36<51:37,  2.02it/s]
 41%|████▏     | 4423/10682 [36:37<51:37,  2.02it/s]
 41%|████▏     | 4424/10682 [36:37<51:35,  2.02it/s]
 41%|████▏     | 4425/10682 [36:38<51:36,  2.02it/s]
                                                    
{'loss': 3.6192, 'grad_norm': 0.21489258110523224, 'learning_rate': 0.0007282391860581049, 'epoch': 0.41}
-
 41%|████▏     | 4425/10682 [36:38<51:36,  2.02it/s]
 41%|████▏     | 4426/10682 [36:38<51:37,  2.02it/s]
 41%|████▏     | 4427/10682 [36:39<51:36,  2.02it/s]
 41%|████▏     | 4428/10682 [36:39<51:37,  2.02it/s]
 41%|████▏     | 4429/10682 [36:40<51:39,  2.02it/s]
 41%|████▏     | 4430/10682 [36:40<51:42,  2.02it/s]
 41%|████▏     | 4431/10682 [36:41<51:45,  2.01it/s]
 41%|████▏     | 4432/10682 [36:41<51:40,  2.02it/s]
 41%|████▏     | 4433/10682 [36:42<51:41,  2.02it/s]
 42%|████▏     | 4434/10682 [36:42<51:37,  2.02it/s]
 42%|████▏     | 4435/10682 [36:43<51:35,  2.02it/s]
 42%|████▏     | 4436/10682 [36:43<51:32,  2.02it/s]
 42%|████▏     | 4437/10682 [36:44<51:31,  2.02it/s]
 42%|████▏     | 4438/10682 [36:44<51:32,  2.02it/s]
 42%|████▏     | 4439/10682 [36:45<51:32,  2.02it/s]
 42%|████▏     | 4440/10682 [36:45<51:34,  2.02it/s]
 42%|████▏     | 4441/10682 [36:46<51:34,  2.02it/s]
 42%|████▏     | 4442/10682 [36:46<51:32,  2.02it/s]
 42%|████▏     | 4443/10682 [36:47<51:34,  2.02it/s]
 42%|████▏     | 4444/10682 [36:47<51:31,  2.02it/s]
 42%|████▏     | 4445/10682 [36:48<51:31,  2.02it/s]
 42%|████▏     | 4446/10682 [36:48<51:29,  2.02it/s]
 42%|████▏     | 4447/10682 [36:49<51:30,  2.02it/s]
 42%|████▏     | 4448/10682 [36:49<51:30,  2.02it/s]
 42%|████▏     | 4449/10682 [36:50<51:29,  2.02it/s]
 42%|████▏     | 4450/10682 [36:50<51:32,  2.02it/s]
                                                    
{'loss': 3.6307, 'grad_norm': 0.22268319129943848, 'learning_rate': 0.0007245969690625126, 'epoch': 0.42}
-
 42%|████▏     | 4450/10682 [36:50<51:32,  2.02it/s]
 42%|████��     | 4451/10682 [36:51<51:37,  2.01it/s]
 42%|████▏     | 4452/10682 [36:51<51:35,  2.01it/s]
 42%|████▏     | 4453/10682 [36:52<51:31,  2.01it/s]
 42%|████▏     | 4454/10682 [36:52<51:26,  2.02it/s]
 42%|████▏     | 4455/10682 [36:53<51:26,  2.02it/s]
 42%|████▏     | 4456/10682 [36:53<51:22,  2.02it/s]
 42%|████▏     | 4457/10682 [36:54<51:21,  2.02it/s]
 42%|████▏     | 4458/10682 [36:54<51:20,  2.02it/s]
 42%|████▏     | 4459/10682 [36:55<51:22,  2.02it/s]
 42%|████▏     | 4460/10682 [36:55<51:20,  2.02it/s]
 42%|████▏     | 4461/10682 [36:56<51:19,  2.02it/s]
 42%|████▏     | 4462/10682 [36:56<51:18,  2.02it/s]
 42%|████▏     | 4463/10682 [36:57<51:17,  2.02it/s]
 42%|████▏     | 4464/10682 [36:57<51:17,  2.02it/s]
 42%|████▏     | 4465/10682 [36:58<51:17,  2.02it/s]
 42%|████▏     | 4466/10682 [36:58<51:16,  2.02it/s]
 42%|████▏     | 4467/10682 [36:59<51:18,  2.02it/s]
 42%|████▏     | 4468/10682 [36:59<51:16,  2.02it/s]
 42%|████▏     | 4469/10682 [37:00<51:17,  2.02it/s]
 42%|████▏     | 4470/10682 [37:00<51:15,  2.02it/s]
 42%|████▏     | 4471/10682 [37:01<51:17,  2.02it/s]
 42%|████▏     | 4472/10682 [37:01<51:13,  2.02it/s]
 42%|████▏     | 4473/10682 [37:02<51:13,  2.02it/s]
 42%|████▏     | 4474/10682 [37:02<51:09,  2.02it/s]
 42%|████▏     | 4475/10682 [37:03<51:09,  2.02it/s]{'loss': 3.6205, 'grad_norm': 0.22710996866226196, 'learning_rate': 0.0007209397599364363, 'epoch': 0.42}
                                                    
-
 42%|████▏     | 4475/10682 [37:03<51:09,  2.02it/s]
 42%|████▏     | 4476/10682 [37:03<51:15,  2.02it/s]
 42%|████▏     | 4477/10682 [37:04<51:14,  2.02it/s]
 42%|████▏     | 4478/10682 [37:04<51:09,  2.02it/s]
 42%|████▏     | 4479/10682 [37:05<51:15,  2.02it/s]
 42%|████▏     | 4480/10682 [37:05<51:10,  2.02it/s]
 42%|████▏     | 4481/10682 [37:06<51:11,  2.02it/s]
 42%|████▏     | 4482/10682 [37:06<51:07,  2.02it/s]
 42%|████▏     | 4483/10682 [37:07<51:07,  2.02it/s]
 42%|████▏     | 4484/10682 [37:07<51:04,  2.02it/s]
 42%|████▏     | 4485/10682 [37:08<51:06,  2.02it/s]
 42%|████▏     | 4486/10682 [37:08<51:03,  2.02it/s]
 42%|████▏     | 4487/10682 [37:09<51:09,  2.02it/s]
 42%|████▏     | 4488/10682 [37:09<51:04,  2.02it/s]
 42%|████▏     | 4489/10682 [37:10<51:09,  2.02it/s]
 42%|████▏     | 4490/10682 [37:10<51:06,  2.02it/s]
 42%|████▏     | 4491/10682 [37:11<51:06,  2.02it/s]
 42%|████▏     | 4492/10682 [37:11<51:08,  2.02it/s]
 42%|████▏     | 4493/10682 [37:12<51:06,  2.02it/s]
 42%|████▏     | 4494/10682 [37:12<51:06,  2.02it/s]
 42%|████▏     | 4495/10682 [37:13<51:06,  2.02it/s]
 42%|████▏     | 4496/10682 [37:13<51:06,  2.02it/s]
 42%|████▏     | 4497/10682 [37:14<51:01,  2.02it/s]
 42%|████▏     | 4498/10682 [37:14<51:04,  2.02it/s]
 42%|████▏     | 4499/10682 [37:15<50:58,  2.02it/s]
 42%|████▏     | 4500/10682 [37:15<50:59,  2.02it/s]
                                                    {'loss': 3.6213, 'grad_norm': 0.21730172634124756, 'learning_rate': 0.0007172678028031901, 'epoch': 0.42}
-
 42%|████▏     | 4500/10682 [37:15<50:59,  2.02it/s]
 42%|████▏     | 4501/10682 [37:16<51:05,  2.02it/s]
 42%|████▏     | 4502/10682 [37:16<51:05,  2.02it/s]
 42%|████▏     | 4503/10682 [37:17<51:03,  2.02it/s]
 42%|████▏     | 4504/10682 [37:17<51:02,  2.02it/s]
 42%|████▏     | 4505/10682 [37:18<50:56,  2.02it/s]
 42%|████▏     | 4506/10682 [37:18<51:02,  2.02it/s]
 42%|████▏     | 4507/10682 [37:19<51:00,  2.02it/s]
 42%|████▏     | 4508/10682 [37:19<51:02,  2.02it/s]
 42%|████▏     | 4509/10682 [37:20<51:10,  2.01it/s]
 42%|████▏     | 4510/10682 [37:20<51:04,  2.01it/s]
 42%|████▏     | 4511/10682 [37:20<51:01,  2.02it/s]
 42%|████▏     | 4512/10682 [37:21<50:56,  2.02it/s]
 42%|████▏     | 4513/10682 [37:21<50:57,  2.02it/s]
 42%|████▏     | 4514/10682 [37:22<50:54,  2.02it/s]
 42%|████▏     | 4515/10682 [37:22<50:53,  2.02it/s]
 42%|████▏     | 4516/10682 [37:23<50:48,  2.02it/s]
 42%|████▏     | 4517/10682 [37:23<50:50,  2.02it/s]
 42%|████▏     | 4518/10682 [37:24<50:53,  2.02it/s]
 42%|████▏     | 4519/10682 [37:24<50:53,  2.02it/s]
 42%|████▏     | 4520/10682 [37:25<50:48,  2.02it/s]
 42%|████▏     | 4521/10682 [37:25<50:49,  2.02it/s]
 42%|████▏     | 4522/10682 [37:26<50:45,  2.02it/s]
 42%|████▏     | 4523/10682 [37:26<50:47,  2.02it/s]
 42%|████▏     | 4524/10682 [37:27<50:44,  2.02it/s]
 42%|████▏     | 4525/10682 [37:27<50:42,  2.02it/s]
                                                    
{'loss': 3.6196, 'grad_norm': 0.21013906598091125, 'learning_rate': 0.0007135813427705359, 'epoch': 0.42}
-
 42%|████▏     | 4525/10682 [37:27<50:42,  2.02it/s]
 42%|████▏     | 4526/10682 [37:28<50:48,  2.02it/s]
 42%|████▏     | 4527/10682 [37:28<50:50,  2.02it/s]
 42%|████▏     | 4528/10682 [37:29<50:47,  2.02it/s]
 42%|████▏     | 4529/10682 [37:29<50:47,  2.02it/s]
 42%|████▏     | 4530/10682 [37:30<50:46,  2.02it/s]
 42%|████▏     | 4531/10682 [37:30<50:49,  2.02it/s]
 42%|████▏     | 4532/10682 [37:31<50:45,  2.02it/s]
 42%|████▏     | 4533/10682 [37:31<50:43,  2.02it/s]
 42%|████▏     | 4534/10682 [37:32<50:42,  2.02it/s]
 42%|████▏     | 4535/10682 [37:32<50:44,  2.02it/s]
 42%|████▏     | 4536/10682 [37:33<50:44,  2.02it/s]
 42%|████▏     | 4537/10682 [37:33<50:50,  2.01it/s]
 42%|████▏     | 4538/10682 [37:34<50:47,  2.02it/s]
 42%|████▏     | 4539/10682 [37:34<50:47,  2.02it/s]
 43%|████▎     | 4540/10682 [37:35<50:44,  2.02it/s]
 43%|████▎     | 4541/10682 [37:35<50:41,  2.02it/s]
 43%|████▎     | 4542/10682 [37:36<50:39,  2.02it/s]
 43%|████▎     | 4543/10682 [37:36<50:38,  2.02it/s]
 43%|████▎     | 4544/10682 [37:37<50:42,  2.02it/s]
 43%|████▎     | 4545/10682 [37:37<50:41,  2.02it/s]
 43%|████▎     | 4546/10682 [37:38<50:39,  2.02it/s]
 43%|████▎     | 4547/10682 [37:38<50:41,  2.02it/s]
 43%|████▎     | 4548/10682 [37:39<50:38,  2.02it/s]
 43%|████▎     | 4549/10682 [37:39<50:38,  2.02it/s]
 43%|████▎     | 4550/10682 [37:40<50:34,  2.02it/s]
                                                    {'loss': 3.6181, 'grad_norm': 0.22290323674678802, 'learning_rate': 0.0007098806259143232, 'epoch': 0.43}
-
 43%|████▎     | 4550/10682 [37:40<50:34,  2.02it/s]
 43%|████▎     | 4551/10682 [37:40<50:39,  2.02it/s]
 43%|████▎     | 4552/10682 [37:41<50:38,  2.02it/s]
 43%|████▎     | 4553/10682 [37:41<50:36,  2.02it/s]
 43%|████▎     | 4554/10682 [37:42<50:33,  2.02it/s]
 43%|████▎     | 4555/10682 [37:42<50:34,  2.02it/s]
 43%|████▎     | 4556/10682 [37:43<50:38,  2.02it/s]
 43%|████▎     | 4557/10682 [37:43<50:36,  2.02it/s]
 43%|████▎     | 4558/10682 [37:44<50:35,  2.02it/s]
 43%|████▎     | 4559/10682 [37:44<50:31,  2.02it/s]
 43%|████▎     | 4560/10682 [37:45<50:31,  2.02it/s]
 43%|████▎     | 4561/10682 [37:45<50:27,  2.02it/s]
 43%|████▎     | 4562/10682 [37:46<50:29,  2.02it/s]
 43%|████▎     | 4563/10682 [37:46<50:25,  2.02it/s]
 43%|████▎     | 4564/10682 [37:47<50:24,  2.02it/s]
 43%|████▎     | 4565/10682 [37:47<50:26,  2.02it/s]
 43%|████▎     | 4566/10682 [37:48<50:27,  2.02it/s]
 43%|████▎     | 4567/10682 [37:48<50:26,  2.02it/s]
 43%|████▎     | 4568/10682 [37:49<50:28,  2.02it/s]
 43%|████▎     | 4569/10682 [37:49<50:24,  2.02it/s]
 43%|████▎     | 4570/10682 [37:50<50:26,  2.02it/s]
 43%|████▎     | 4571/10682 [37:50<50:24,  2.02it/s]
 43%|████▎     | 4572/10682 [37:51<50:23,  2.02it/s]
 43%|████▎     | 4573/10682 [37:51<50:20,  2.02it/s]
 43%|████▎     | 4574/10682 [37:52<50:22,  2.02it/s]
 43%|████▎     | 4575/10682 [37:52<50:20,  2.02it/s]
                                                    {'loss': 3.6139, 'grad_norm': 0.20928622782230377, 'learning_rate': 0.0007061658992620613, 'epoch': 0.43}
-
 43%|████▎     | 4575/10682 [37:52<50:20,  2.02it/s]
 43%|████▎     | 4576/10682 [37:53<50:27,  2.02it/s]
 43%|████▎     | 4577/10682 [37:53<50:30,  2.01it/s]
 43%|████▎     | 4578/10682 [37:54<50:27,  2.02it/s]
 43%|████▎     | 4579/10682 [37:54<50:25,  2.02it/s]
 43%|████▎     | 4580/10682 [37:55<50:24,  2.02it/s]
 43%|████▎     | 4581/10682 [37:55<50:26,  2.02it/s]
 43%|████▎     | 4582/10682 [37:56<50:22,  2.02it/s]
 43%|████▎     | 4583/10682 [37:56<50:20,  2.02it/s]
 43%|████▎     | 4584/10682 [37:57<50:21,  2.02it/s]
 43%|████▎     | 4585/10682 [37:57<50:22,  2.02it/s]
 43%|████▎     | 4586/10682 [37:58<50:24,  2.02it/s]
 43%|████▎     | 4587/10682 [37:58<50:20,  2.02it/s]
 43%|████▎     | 4588/10682 [37:59<50:18,  2.02it/s]
 43%|████▎     | 4589/10682 [37:59<50:18,  2.02it/s]
 43%|████▎     | 4590/10682 [38:00<50:15,  2.02it/s]
 43%|████▎     | 4591/10682 [38:00<50:16,  2.02it/s]
 43%|████▎     | 4592/10682 [38:01<50:15,  2.02it/s]
 43%|████▎     | 4593/10682 [38:01<50:13,  2.02it/s]
 43%|████▎     | 4594/10682 [38:02<50:11,  2.02it/s]
 43%|████▎     | 4595/10682 [38:02<50:10,  2.02it/s]
 43%|████▎     | 4596/10682 [38:03<50:17,  2.02it/s]
 43%|████▎     | 4597/10682 [38:03<50:16,  2.02it/s]
 43%|████▎     | 4598/10682 [38:04<50:17,  2.02it/s]
 43%|████▎     | 4599/10682 [38:04<50:12,  2.02it/s]
 43%|████▎     | 4600/10682 [38:05<50:13,  2.02it/s]
                                                    
{'loss': 3.6122, 'grad_norm': 0.21296823024749756, 'learning_rate': 0.0007024374107764318, 'epoch': 0.43}
-
 43%|████▎     | 4600/10682 [38:05<50:13,  2.02it/s]
 43%|████▎     | 4601/10682 [38:05<50:17,  2.02it/s]
 43%|████▎     | 4602/10682 [38:06<50:17,  2.02it/s]
 43%|████▎     | 4603/10682 [38:06<50:10,  2.02it/s]
 43%|████▎     | 4604/10682 [38:07<50:13,  2.02it/s]
 43%|████▎     | 4605/10682 [38:07<50:08,  2.02it/s]
 43%|████▎     | 4606/10682 [38:08<50:12,  2.02it/s]
 43%|████▎     | 4607/10682 [38:08<50:10,  2.02it/s]
 43%|████▎     | 4608/10682 [38:09<50:09,  2.02it/s]
 43%|████▎     | 4609/10682 [38:09<50:08,  2.02it/s]
 43%|████▎     | 4610/10682 [38:10<50:09,  2.02it/s]
 43%|████▎     | 4611/10682 [38:10<50:12,  2.02it/s]
 43%|████▎     | 4612/10682 [38:11<50:06,  2.02it/s]
 43%|████▎     | 4613/10682 [38:11<50:04,  2.02it/s]
 43%|████▎     | 4614/10682 [38:12<58:15,  1.74it/s]
 43%|████▎     | 4615/10682 [38:12<55:48,  1.81it/s]
 43%|████▎     | 4616/10682 [38:13<54:01,  1.87it/s]
 43%|████▎     | 4617/10682 [38:13<52:51,  1.91it/s]
 43%|████▎     | 4618/10682 [38:14<51:57,  1.95it/s]
 43%|████▎     | 4619/10682 [38:14<51:22,  1.97it/s]
 43%|████▎     | 4620/10682 [38:15<50:54,  1.98it/s]
 43%|████▎     | 4621/10682 [38:15<50:40,  1.99it/s]
 43%|████▎     | 4622/10682 [38:16<50:26,  2.00it/s]
 43%|████▎     | 4623/10682 [38:16<50:26,  2.00it/s]
 43%|████▎     | 4624/10682 [38:17<50:16,  2.01it/s]
 43%|████▎     | 4625/10682 [38:17<50:10,  2.01it/s]
                                                    {'loss': 3.6088, 'grad_norm': 0.2180296629667282, 'learning_rate': 0.0006986954093387361, 'epoch': 0.43}
-
 43%|████▎     | 4625/10682 [38:17<50:10,  2.01it/s]
 43%|████▎     | 4626/10682 [38:18<50:18,  2.01it/s]
 43%|████▎     | 4627/10682 [38:18<50:08,  2.01it/s]
 43%|████▎     | 4628/10682 [38:19<50:06,  2.01it/s]
 43%|████▎     | 4629/10682 [38:19<49:58,  2.02it/s]
 43%|████▎     | 4630/10682 [38:20<50:03,  2.01it/s]
 43%|████▎     | 4631/10682 [38:20<49:54,  2.02it/s]
 43%|████▎     | 4632/10682 [38:21<49:54,  2.02it/s]
 43%|████▎     | 4633/10682 [38:21<57:47,  1.74it/s]
 43%|████▎     | 4634/10682 [38:22<55:27,  1.82it/s]
 43%|████▎     | 4635/10682 [38:22<53:42,  1.88it/s]
 43%|████▎     | 4636/10682 [38:23<52:34,  1.92it/s]
 43%|████▎     | 4637/10682 [38:23<51:42,  1.95it/s]
 43%|████▎     | 4638/10682 [38:24<51:10,  1.97it/s]
 43%|████▎     | 4639/10682 [38:24<50:42,  1.99it/s]
 43%|████▎     | 4640/10682 [38:25<50:27,  2.00it/s]
 43%|████▎     | 4641/10682 [38:25<50:10,  2.01it/s]
 43%|████▎     | 4642/10682 [38:26<50:06,  2.01it/s]
 43%|████▎     | 4643/10682 [38:26<49:57,  2.01it/s]
 43%|████▎     | 4644/10682 [38:27<49:52,  2.02it/s]
 43%|████▎     | 4645/10682 [38:27<49:48,  2.02it/s]
 43%|████▎     | 4646/10682 [38:28<49:47,  2.02it/s]
 44%|████▎     | 4647/10682 [38:28<49:46,  2.02it/s]
 44%|████▎     | 4648/10682 [38:29<49:45,  2.02it/s]
 44%|████▎     | 4649/10682 [38:29<49:42,  2.02it/s]
 44%|████▎     | 4650/10682 [38:30<49:41,  2.02it/s]{'loss': 3.6019, 'grad_norm': 0.21513015031814575, 'learning_rate': 0.0006949401447322816, 'epoch': 0.44}
                                                    
-
 44%|████▎     | 4650/10682 [38:30<49:41,  2.02it/s]
 44%|████▎     | 4651/10682 [38:30<49:45,  2.02it/s]
 44%|████▎     | 4652/10682 [38:31<49:42,  2.02it/s]
 44%|████▎     | 4653/10682 [38:31<49:43,  2.02it/s]
 44%|████▎     | 4654/10682 [38:32<49:42,  2.02it/s]
 44%|████▎     | 4655/10682 [38:32<49:43,  2.02it/s]
 44%|████▎     | 4656/10682 [38:33<49:41,  2.02it/s]
 44%|████▎     | 4657/10682 [38:33<49:39,  2.02it/s]
 44%|████▎     | 4658/10682 [38:34<49:36,  2.02it/s]
 44%|████▎     | 4659/10682 [38:34<49:40,  2.02it/s]
 44%|████▎     | 4660/10682 [38:35<49:36,  2.02it/s]
 44%|████▎     | 4661/10682 [38:35<49:40,  2.02it/s]
 44%|████▎     | 4662/10682 [38:36<49:37,  2.02it/s]
 44%|████▎     | 4663/10682 [38:36<49:42,  2.02it/s]
 44%|████▎     | 4664/10682 [38:37<49:37,  2.02it/s]
 44%|████▎     | 4665/10682 [38:37<49:42,  2.02it/s]
 44%|████▎     | 4666/10682 [38:38<49:38,  2.02it/s]
 44%|████▎     | 4667/10682 [38:38<49:41,  2.02it/s]
 44%|████▎     | 4668/10682 [38:39<49:34,  2.02it/s]
 44%|████▎     | 4669/10682 [38:39<49:32,  2.02it/s]
 44%|████▎     | 4670/10682 [38:40<49:32,  2.02it/s]
 44%|████▎     | 4671/10682 [38:40<49:30,  2.02it/s]
 44%|████▎     | 4672/10682 [38:41<49:31,  2.02it/s]
 44%|████▎     | 4673/10682 [38:41<49:32,  2.02it/s]
 44%|████▍     | 4674/10682 [38:42<49:32,  2.02it/s]
 44%|████▍     | 4675/10682 [38:42<49:37,  2.02it/s]
                                                    {'loss': 3.6015, 'grad_norm': 0.21772845089435577, 'learning_rate': 0.0006911718676257091, 'epoch': 0.44}
-

 44%|████▍     | 4675/10682 [38:42<49:37,  2.02it/s]
 44%|████▍     | 4676/10682 [38:43<49:38,  2.02it/s]
 44%|████▍     | 4677/10682 [38:43<49:36,  2.02it/s]
 44%|████▍     | 4678/10682 [38:44<49:33,  2.02it/s]
 44%|████▍     | 4679/10682 [38:44<49:31,  2.02it/s]
 44%|████▍     | 4680/10682 [38:45<49:30,  2.02it/s]
 44%|████▍     | 4681/10682 [38:45<49:29,  2.02it/s]
 44%|████▍     | 4682/10682 [38:46<49:29,  2.02it/s]
 44%|████▍     | 4683/10682 [38:46<49:31,  2.02it/s]
 44%|████▍     | 4684/10682 [38:47<49:30,  2.02it/s]
 44%|████▍     | 4685/10682 [38:47<49:30,  2.02it/s]
 44%|████▍     | 4686/10682 [38:48<49:28,  2.02it/s]
 44%|████▍     | 4687/10682 [38:48<49:25,  2.02it/s]
 44%|████▍     | 4688/10682 [38:49<49:24,  2.02it/s]
 44%|████▍     | 4689/10682 [38:49<49:23,  2.02it/s]
 44%|████▍     | 4690/10682 [38:50<49:23,  2.02it/s]
 44%|████▍     | 4691/10682 [38:50<49:21,  2.02it/s]
 44%|████▍     | 4692/10682 [38:51<49:25,  2.02it/s]
 44%|████▍     | 4693/10682 [38:51<49:22,  2.02it/s]
 44%|████▍     | 4694/10682 [38:52<49:29,  2.02it/s]
 44%|████▍     | 4695/10682 [38:52<49:31,  2.01it/s]
 44%|████▍     | 4696/10682 [38:53<49:25,  2.02it/s]
 44%|████▍     | 4697/10682 [38:53<49:25,  2.02it/s]
 44%|████▍     | 4698/10682 [38:54<49:20,  2.02it/s]
 44%|████▍     | 4699/10682 [38:54<49:21,  2.02it/s]
 44%|████▍     | 4700/10682 [38:55<49:17,  2.02it/s]
                                                    {'loss': 3.612, 'grad_norm': 0.22059211134910583, 'learning_rate': 0.0006873908295562602, 'epoch': 0.44}
-

 44%|████▍     | 4700/10682 [38:55<49:17,  2.02it/s]
 44%|████▍     | 4701/10682 [38:55<49:23,  2.02it/s]
 44%|████▍     | 4702/10682 [38:56<49:22,  2.02it/s]
 44%|████▍     | 4703/10682 [38:56<49:24,  2.02it/s]
 44%|████▍     | 4704/10682 [38:57<49:21,  2.02it/s]
 44%|████▍     | 4705/10682 [38:57<49:22,  2.02it/s]
 44%|████▍     | 4706/10682 [38:58<49:15,  2.02it/s]
 44%|████▍     | 4707/10682 [38:58<49:17,  2.02it/s]
 44%|████▍     | 4708/10682 [38:59<49:12,  2.02it/s]
 44%|████▍     | 4709/10682 [38:59<49:13,  2.02it/s]
 44%|████▍     | 4710/10682 [39:00<49:09,  2.02it/s]
 44%|████▍     | 4711/10682 [39:00<49:09,  2.02it/s]
 44%|████▍     | 4712/10682 [39:01<49:13,  2.02it/s]
 44%|████▍     | 4713/10682 [39:01<49:16,  2.02it/s]
 44%|████▍     | 4714/10682 [39:02<49:15,  2.02it/s]
 44%|████▍     | 4715/10682 [39:02<49:14,  2.02it/s]
 44%|████▍     | 4716/10682 [39:03<49:11,  2.02it/s]
 44%|████▍     | 4717/10682 [39:03<49:10,  2.02it/s]
 44%|████▍     | 4718/10682 [39:04<49:10,  2.02it/s]
 44%|████▍     | 4719/10682 [39:04<49:07,  2.02it/s]
 44%|████▍     | 4720/10682 [39:04<49:08,  2.02it/s]
 44%|████▍     | 4721/10682 [39:05<49:09,  2.02it/s]
 44%|████▍     | 4722/10682 [39:05<49:10,  2.02it/s]
 44%|████▍     | 4723/10682 [39:06<49:08,  2.02it/s]
 44%|████▍     | 4724/10682 [39:06<49:12,  2.02it/s]
 44%|████▍     | 4725/10682 [39:07<49:08,  2.02it/s]{'loss': 3.6053, 'grad_norm': 0.21416307985782623, 'learning_rate': 0.0006835972829129873, 'epoch': 0.44}
                                                    
-
 44%|████▍     | 4725/10682 [39:07<49:08,  2.02it/s]
 44%|████▍     | 4726/10682 [39:07<49:14,  2.02it/s]
 44%|████▍     | 4727/10682 [39:08<49:08,  2.02it/s]
 44%|████▍     | 4728/10682 [39:08<49:07,  2.02it/s]
 44%|████▍     | 4729/10682 [39:09<49:04,  2.02it/s]
 44%|████▍     | 4730/10682 [39:09<49:05,  2.02it/s]
 44%|████▍     | 4731/10682 [39:10<49:06,  2.02it/s]
 44%|████▍     | 4732/10682 [39:10<49:04,  2.02it/s]
 44%|████▍     | 4733/10682 [39:11<49:08,  2.02it/s]
 44%|████▍     | 4734/10682 [39:11<49:07,  2.02it/s]
 44%|████▍     | 4735/10682 [39:12<49:09,  2.02it/s]
 44%|████▍     | 4736/10682 [39:12<49:07,  2.02it/s]
 44%|████��     | 4737/10682 [39:13<49:03,  2.02it/s]
 44%|████▍     | 4738/10682 [39:13<49:02,  2.02it/s]
 44%|████▍     | 4739/10682 [39:14<49:00,  2.02it/s]
 44%|████▍     | 4740/10682 [39:14<48:59,  2.02it/s]
 44%|████▍     | 4741/10682 [39:15<48:59,  2.02it/s]
 44%|████▍     | 4742/10682 [39:15<48:58,  2.02it/s]
 44%|████▍     | 4743/10682 [39:16<49:02,  2.02it/s]
 44%|████▍     | 4744/10682 [39:16<49:00,  2.02it/s]
 44%|████▍     | 4745/10682 [39:17<48:58,  2.02it/s]
 44%|████▍     | 4746/10682 [39:17<48:56,  2.02it/s]
 44%|████▍     | 4747/10682 [39:18<48:56,  2.02it/s]
 44%|████▍     | 4748/10682 [39:18<48:56,  2.02it/s]
 44%|████▍     | 4749/10682 [39:19<48:55,  2.02it/s]
 44%|████▍     | 4750/10682 [39:19<48:54,  2.02it/s]
                                                    {'loss': 3.5976, 'grad_norm': 0.21803192794322968, 'learning_rate': 0.000679791480919905, 'epoch': 0.44}
-

 44%|████▍     | 4750/10682 [39:19<48:54,  2.02it/s]
 44%|████▍     | 4751/10682 [39:20<48:58,  2.02it/s]
 44%|████▍     | 4752/10682 [39:20<48:56,  2.02it/s]
 44%|████▍     | 4753/10682 [39:21<48:55,  2.02it/s]
 45%|████▍     | 4754/10682 [39:21<48:53,  2.02it/s]
 45%|████▍     | 4755/10682 [39:22<48:50,  2.02it/s]
 45%|████▍     | 4756/10682 [39:22<48:54,  2.02it/s]
 45%|████▍     | 4757/10682 [39:23<48:55,  2.02it/s]
 45%|████▍     | 4758/10682 [39:23<48:55,  2.02it/s]
 45%|████▍     | 4759/10682 [39:24<48:50,  2.02it/s]
 45%|████▍     | 4760/10682 [39:24<48:50,  2.02it/s]
 45%|████▍     | 4761/10682 [39:25<48:51,  2.02it/s]
 45%|████▍     | 4762/10682 [39:25<48:53,  2.02it/s]
 45%|████▍     | 4763/10682 [39:26<48:53,  2.02it/s]
 45%|████▍     | 4764/10682 [39:26<48:51,  2.02it/s]
 45%|████▍     | 4765/10682 [39:27<48:50,  2.02it/s]
 45%|████▍     | 4766/10682 [39:27<48:48,  2.02it/s]
 45%|████▍     | 4767/10682 [39:28<48:46,  2.02it/s]
 45%|████▍     | 4768/10682 [39:28<48:45,  2.02it/s]
 45%|████▍     | 4769/10682 [39:29<48:46,  2.02it/s]
 45%|████▍     | 4770/10682 [39:29<48:47,  2.02it/s]
 45%|████▍     | 4771/10682 [39:30<48:48,  2.02it/s]
 45%|████▍     | 4772/10682 [39:30<48:46,  2.02it/s]
 45%|████▍     | 4773/10682 [39:31<48:47,  2.02it/s]
 45%|████▍     | 4774/10682 [39:31<48:44,  2.02it/s]
 45%|████▍     | 4775/10682 [39:32<48:47,  2.02it/s]
                                                    
{'loss': 3.5989, 'grad_norm': 0.21313433349132538, 'learning_rate': 0.000675973677619089, 'epoch': 0.45}
-
 45%|████▍     | 4775/10682 [39:32<48:47,  2.02it/s]
 45%|████▍     | 4776/10682 [39:32<48:45,  2.02it/s]
 45%|████▍     | 4777/10682 [39:33<48:45,  2.02it/s]
 45%|████▍     | 4778/10682 [39:33<48:40,  2.02it/s]
 45%|████▍     | 4779/10682 [39:34<48:41,  2.02it/s]
 45%|████▍     | 4780/10682 [39:34<48:40,  2.02it/s]
 45%|████▍     | 4781/10682 [39:35<48:40,  2.02it/s]
 45%|████▍     | 4782/10682 [39:35<48:40,  2.02it/s]
 45%|████▍     | 4783/10682 [39:36<48:41,  2.02it/s]
 45%|████▍     | 4784/10682 [39:36<48:37,  2.02it/s]
 45%|████▍     | 4785/10682 [39:37<48:37,  2.02it/s]
 45%|████▍     | 4786/10682 [39:37<48:33,  2.02it/s]
 45%|████▍     | 4787/10682 [39:38<48:33,  2.02it/s]
 45%|████▍     | 4788/10682 [39:38<48:35,  2.02it/s]
 45%|████▍     | 4789/10682 [39:39<48:32,  2.02it/s]
 45%|████▍     | 4790/10682 [39:39<48:34,  2.02it/s]
 45%|████▍     | 4791/10682 [39:40<48:33,  2.02it/s]
 45%|████▍     | 4792/10682 [39:40<48:35,  2.02it/s]
 45%|████▍     | 4793/10682 [39:41<48:35,  2.02it/s]
 45%|████▍     | 4794/10682 [39:41<48:36,  2.02it/s]
 45%|████▍     | 4795/10682 [39:42<48:34,  2.02it/s]
 45%|████▍     | 4796/10682 [39:42<48:34,  2.02it/s]
 45%|████▍     | 4797/10682 [39:43<48:32,  2.02it/s]
 45%|████▍     | 4798/10682 [39:43<48:33,  2.02it/s]
 45%|████▍     | 4799/10682 [39:44<48:31,  2.02it/s]
 45%|████▍     | 4800/10682 [39:44<48:31,  2.02it/s]{'loss': 3.5892, 'grad_norm': 0.22447469830513, 'learning_rate': 0.0006721441278537161, 'epoch': 0.45}
                                                    
-
 45%|████▍     | 4800/10682 [39:44<48:31,  2.02it/s]
 45%|████▍     | 4801/10682 [39:45<48:34,  2.02it/s]
 45%|████▍     | 4802/10682 [39:45<48:34,  2.02it/s]
 45%|████▍     | 4803/10682 [39:46<48:33,  2.02it/s]
 45%|████▍     | 4804/10682 [39:46<48:29,  2.02it/s]
 45%|████▍     | 4805/10682 [39:47<48:28,  2.02it/s]
 45%|████▍     | 4806/10682 [39:47<48:33,  2.02it/s]
 45%|████▌     | 4807/10682 [39:48<48:31,  2.02it/s]
 45%|████▌     | 4808/10682 [39:48<48:30,  2.02it/s]
 45%|████▌     | 4809/10682 [39:49<48:29,  2.02it/s]
 45%|████▌     | 4810/10682 [39:49<48:31,  2.02it/s]
 45%|████▌     | 4811/10682 [39:50<48:32,  2.02it/s]
 45%|████▌     | 4812/10682 [39:50<48:31,  2.02it/s]
 45%|████▌     | 4813/10682 [39:51<48:31,  2.02it/s]
 45%|████▌     | 4814/10682 [39:51<48:28,  2.02it/s]
 45%|████▌     | 4815/10682 [39:52<48:30,  2.02it/s]
 45%|████▌     | 4816/10682 [39:52<48:27,  2.02it/s]
 45%|████▌     | 4817/10682 [39:53<48:25,  2.02it/s]
 45%|████▌     | 4818/10682 [39:53<48:20,  2.02it/s]
 45%|████▌     | 4819/10682 [39:54<48:20,  2.02it/s]
 45%|████▌     | 4820/10682 [39:54<48:19,  2.02it/s]
 45%|████▌     | 4821/10682 [39:55<48:22,  2.02it/s]
 45%|████▌     | 4822/10682 [39:55<48:19,  2.02it/s]
 45%|████▌     | 4823/10682 [39:55<48:19,  2.02it/s]
 45%|████▌     | 4824/10682 [39:56<48:19,  2.02it/s]
 45%|████▌     | 4825/10682 [39:56<48:18,  2.02it/s]
                                                    {'loss': 3.5952, 'grad_norm': 0.22358663380146027, 'learning_rate': 0.0006683030872510552, 'epoch': 0.45}
-
 45%|████▌     | 4825/10682 [39:56<48:18,  2.02it/s]
 45%|████▌     | 4826/10682 [39:57<48:19,  2.02it/s]
 45%|████▌     | 4827/10682 [39:57<48:18,  2.02it/s]
 45%|████▌     | 4828/10682 [39:58<48:16,  2.02it/s]
 45%|████▌     | 4829/10682 [39:58<48:18,  2.02it/s]
 45%|████▌     | 4830/10682 [39:59<48:16,  2.02it/s]
 45%|████▌     | 4831/10682 [39:59<48:23,  2.01it/s]
 45%|████▌     | 4832/10682 [40:00<48:22,  2.02it/s]
 45%|████▌     | 4833/10682 [40:00<48:18,  2.02it/s]
 45%|████▌     | 4834/10682 [40:01<48:17,  2.02it/s]
 45%|████▌     | 4835/10682 [40:01<48:18,  2.02it/s]
 45%|████▌     | 4836/10682 [40:02<48:14,  2.02it/s]
 45%|████▌     | 4837/10682 [40:02<48:15,  2.02it/s]
 45%|████▌     | 4838/10682 [40:03<48:10,  2.02it/s]
 45%|████▌     | 4839/10682 [40:03<48:12,  2.02it/s]
 45%|████▌     | 4840/10682 [40:04<48:09,  2.02it/s]
 45%|████▌     | 4841/10682 [40:04<48:12,  2.02it/s]
 45%|████▌     | 4842/10682 [40:05<48:08,  2.02it/s]
 45%|████▌     | 4843/10682 [40:05<48:08,  2.02it/s]
 45%|████▌     | 4844/10682 [40:06<48:06,  2.02it/s]
 45%|████▌     | 4845/10682 [40:06<48:06,  2.02it/s]
 45%|████▌     | 4846/10682 [40:07<48:03,  2.02it/s]
 45%|████▌     | 4847/10682 [40:07<48:04,  2.02it/s]
 45%|████▌     | 4848/10682 [40:08<48:03,  2.02it/s]
 45%|████▌     | 4849/10682 [40:08<48:03,  2.02it/s]
 45%|████▌     | 4850/10682 [40:09<48:06,  2.02it/s]
                                                    
{'loss': 3.5924, 'grad_norm': 0.23816557228565216, 'learning_rate': 0.0006644508122054031, 'epoch': 0.45}
-
 45%|████▌     | 4850/10682 [40:09<48:06,  2.02it/s]
 45%|████▌     | 4851/10682 [40:09<48:11,  2.02it/s]
 45%|████▌     | 4852/10682 [40:10<48:08,  2.02it/s]
 45%|████▌     | 4853/10682 [40:10<48:06,  2.02it/s]
 45%|████▌     | 4854/10682 [40:11<48:08,  2.02it/s]
 45%|████▌     | 4855/10682 [40:11<48:06,  2.02it/s]
 45%|████▌     | 4856/10682 [40:12<48:04,  2.02it/s]
 45%|████▌     | 4857/10682 [40:12<48:01,  2.02it/s]
 45%|████▌     | 4858/10682 [40:13<47:59,  2.02it/s]
 45%|████▌     | 4859/10682 [40:13<48:04,  2.02it/s]
 45%|████▌     | 4860/10682 [40:14<48:06,  2.02it/s]
 46%|████▌     | 4861/10682 [40:14<48:05,  2.02it/s]
 46%|████▌     | 4862/10682 [40:15<48:04,  2.02it/s]
 46%|████▌     | 4863/10682 [40:15<48:01,  2.02it/s]
 46%|████▌     | 4864/10682 [40:16<48:01,  2.02it/s]
 46%|████▌     | 4865/10682 [40:16<47:58,  2.02it/s]
 46%|████▌     | 4866/10682 [40:17<47:56,  2.02it/s]
 46%|████▌     | 4867/10682 [40:17<47:56,  2.02it/s]
 46%|████▌     | 4868/10682 [40:18<47:53,  2.02it/s]
 46%|████▌     | 4869/10682 [40:18<47:54,  2.02it/s]
 46%|████▌     | 4870/10682 [40:19<47:54,  2.02it/s]
 46%|████▌     | 4871/10682 [40:19<47:56,  2.02it/s]
 46%|████▌     | 4872/10682 [40:20<47:53,  2.02it/s]
 46%|████▌     | 4873/10682 [40:20<47:55,  2.02it/s]
 46%|████▌     | 4874/10682 [40:21<47:51,  2.02it/s]
 46%|████▌     | 4875/10682 [40:21<47:52,  2.02it/s]
                                                    
{'loss': 3.59, 'grad_norm': 0.21047383546829224, 'learning_rate': 0.0006605875598609699, 'epoch': 0.46}
-
 46%|████▌     | 4875/10682 [40:21<47:52,  2.02it/s]
 46%|████▌     | 4876/10682 [40:22<47:53,  2.02it/s]
 46%|████▌     | 4877/10682 [40:22<47:53,  2.02it/s]
 46%|████▌     | 4878/10682 [40:23<47:51,  2.02it/s]
 46%|████▌     | 4879/10682 [40:23<47:53,  2.02it/s]
 46%|████▌     | 4880/10682 [40:24<47:54,  2.02it/s]
 46%|████▌     | 4881/10682 [40:24<47:56,  2.02it/s]
 46%|████▌     | 4882/10682 [40:25<47:54,  2.02it/s]
 46%|████▌     | 4883/10682 [40:25<47:52,  2.02it/s]
 46%|████▌     | 4884/10682 [40:26<47:49,  2.02it/s]
 46%|████▌     | 4885/10682 [40:26<47:47,  2.02it/s]
 46%|████▌     | 4886/10682 [40:27<47:48,  2.02it/s]
 46%|████▌     | 4887/10682 [40:27<47:46,  2.02it/s]
 46%|████▌     | 4888/10682 [40:28<47:47,  2.02it/s]
 46%|████▌     | 4889/10682 [40:28<47:48,  2.02it/s]
 46%|████▌     | 4890/10682 [40:29<47:52,  2.02it/s]
 46%|████▌     | 4891/10682 [40:29<47:49,  2.02it/s]
 46%|████▌     | 4892/10682 [40:30<47:47,  2.02it/s]
 46%|████▌     | 4893/10682 [40:30<47:47,  2.02it/s]
 46%|████▌     | 4894/10682 [40:31<47:45,  2.02it/s]
 46%|████▌     | 4895/10682 [40:31<47:42,  2.02it/s]
 46%|████▌     | 4896/10682 [40:32<47:40,  2.02it/s]
 46%|████▌     | 4897/10682 [40:32<47:41,  2.02it/s]
 46%|████▌     | 4898/10682 [40:33<47:43,  2.02it/s]
 46%|████▌     | 4899/10682 [40:33<47:42,  2.02it/s]
 46%|████▌     | 4900/10682 [40:34<47:40,  2.02it/s]{'loss': 3.5912, 'grad_norm': 0.21346454322338104, 'learning_rate': 0.0006567135880947134, 'epoch': 0.46}
                                                    
-
 46%|████▌     | 4900/10682 [40:34<47:40,  2.02it/s]
 46%|████▌     | 4901/10682 [40:34<47:43,  2.02it/s]
 46%|████▌     | 4902/10682 [40:35<47:43,  2.02it/s]
 46%|████▌     | 4903/10682 [40:35<47:40,  2.02it/s]
 46%|████▌     | 4904/10682 [40:36<47:38,  2.02it/s]
 46%|████▌     | 4905/10682 [40:36<47:35,  2.02it/s]
 46%|████▌     | 4906/10682 [40:37<47:36,  2.02it/s]
 46%|████▌     | 4907/10682 [40:37<47:37,  2.02it/s]
 46%|████▌     | 4908/10682 [40:38<47:42,  2.02it/s]
 46%|████▌     | 4909/10682 [40:38<47:36,  2.02it/s]
 46%|████▌     | 4910/10682 [40:39<47:41,  2.02it/s]
 46%|████▌     | 4911/10682 [40:39<47:37,  2.02it/s]
 46%|████▌     | 4912/10682 [40:40<47:37,  2.02it/s]
 46%|████▌     | 4913/10682 [40:40<47:33,  2.02it/s]
 46%|████▌     | 4914/10682 [40:41<47:32,  2.02it/s]
 46%|████▌     | 4915/10682 [40:41<47:31,  2.02it/s]
 46%|████▌     | 4916/10682 [40:42<47:34,  2.02it/s]
 46%|████▌     | 4917/10682 [40:42<47:33,  2.02it/s]
 46%|████▌     | 4918/10682 [40:43<47:32,  2.02it/s]
 46%|████▌     | 4919/10682 [40:43<47:34,  2.02it/s]
 46%|████▌     | 4920/10682 [40:44<47:34,  2.02it/s]
 46%|████▌     | 4921/10682 [40:44<47:30,  2.02it/s]
 46%|████▌     | 4922/10682 [40:45<47:31,  2.02it/s]
 46%|████▌     | 4923/10682 [40:45<47:27,  2.02it/s]
 46%|████▌     | 4924/10682 [40:45<47:29,  2.02it/s]
 46%|████▌     | 4925/10682 [40:46<47:29,  2.02it/s]
                                                    {'loss': 3.5822, 'grad_norm': 0.2335873246192932, 'learning_rate': 0.0006528291554991272, 'epoch': 0.46}
-

 46%|████▌     | 4925/10682 [40:46<47:29,  2.02it/s]
 46%|████▌     | 4926/10682 [40:46<47:32,  2.02it/s]
 46%|████▌     | 4927/10682 [40:47<47:30,  2.02it/s]
 46%|████▌     | 4928/10682 [40:47<47:30,  2.02it/s]
 46%|████▌     | 4929/10682 [40:48<47:34,  2.02it/s]
 46%|████▌     | 4930/10682 [40:48<47:32,  2.02it/s]
 46%|████▌     | 4931/10682 [40:49<47:29,  2.02it/s]
 46%|████▌     | 4932/10682 [40:49<47:29,  2.02it/s]
 46%|████▌     | 4933/10682 [40:50<47:28,  2.02it/s]
 46%|████▌     | 4934/10682 [40:50<47:26,  2.02it/s]
 46%|████▌     | 4935/10682 [40:51<47:25,  2.02it/s]
 46%|████▌     | 4936/10682 [40:51<47:24,  2.02it/s]
 46%|████▌     | 4937/10682 [40:52<47:28,  2.02it/s]
 46%|████▌     | 4938/10682 [40:52<47:27,  2.02it/s]
 46%|████▌     | 4939/10682 [40:53<47:30,  2.01it/s]
 46%|████▌     | 4940/10682 [40:53<47:27,  2.02it/s]
 46%|████▋     | 4941/10682 [40:54<47:26,  2.02it/s]
 46%|████▋     | 4942/10682 [40:54<47:22,  2.02it/s]
 46%|████▋     | 4943/10682 [40:55<47:19,  2.02it/s]
 46%|████▋     | 4944/10682 [40:55<47:20,  2.02it/s]
 46%|████▋     | 4945/10682 [40:56<47:22,  2.02it/s]
 46%|████▋     | 4946/10682 [40:56<47:22,  2.02it/s]
 46%|████▋     | 4947/10682 [40:57<47:18,  2.02it/s]
 46%|████▋     | 4948/10682 [40:57<47:20,  2.02it/s]
 46%|████▋     | 4949/10682 [40:58<47:19,  2.02it/s]
 46%|████▋     | 4950/10682 [40:58<47:20,  2.02it/s]
                                                    
{'loss': 3.5867, 'grad_norm': 0.21616508066654205, 'learning_rate': 0.0006489345213649782, 'epoch': 0.46}
-
 46%|████▋     | 4950/10682 [40:58<47:20,  2.02it/s]
 46%|████▋     | 4951/10682 [40:59<47:20,  2.02it/s]
 46%|████▋     | 4952/10682 [40:59<47:19,  2.02it/s]
 46%|████▋     | 4953/10682 [41:00<47:17,  2.02it/s]
 46%|████▋     | 4954/10682 [41:00<47:15,  2.02it/s]
 46%|████▋     | 4955/10682 [41:01<47:17,  2.02it/s]
 46%|████▋     | 4956/10682 [41:01<47:16,  2.02it/s]
 46%|████▋     | 4957/10682 [41:02<47:14,  2.02it/s]
 46%|████▋     | 4958/10682 [41:02<47:14,  2.02it/s]
 46%|████▋     | 4959/10682 [41:03<47:19,  2.02it/s]
 46%|████▋     | 4960/10682 [41:03<47:12,  2.02it/s]
 46%|████▋     | 4961/10682 [41:04<47:14,  2.02it/s]
 46%|████▋     | 4962/10682 [41:04<47:14,  2.02it/s]
 46%|████▋     | 4963/10682 [41:05<47:15,  2.02it/s]
 46%|████▋     | 4964/10682 [41:05<47:12,  2.02it/s]
 46%|████▋     | 4965/10682 [41:06<47:13,  2.02it/s]
 46%|████▋     | 4966/10682 [41:06<47:14,  2.02it/s]
 46%|████▋     | 4967/10682 [41:07<47:10,  2.02it/s]
 47%|████▋     | 4968/10682 [41:07<47:09,  2.02it/s]
 47%|████▋     | 4969/10682 [41:08<47:09,  2.02it/s]
 47%|████▋     | 4970/10682 [41:08<47:08,  2.02it/s]
 47%|████▋     | 4971/10682 [41:09<47:08,  2.02it/s]
 47%|████▋     | 4972/10682 [41:09<47:08,  2.02it/s]
 47%|████▋     | 4973/10682 [41:10<47:10,  2.02it/s]
 47%|████▋     | 4974/10682 [41:10<47:13,  2.01it/s]
 47%|████▋     | 4975/10682 [41:11<47:11,  2.02it/s]{'loss': 3.5917, 'grad_norm': 0.2209041565656662, 'learning_rate': 0.0006450299456639985, 'epoch': 0.47}
-
                                                    

 47%|████▋     | 4975/10682 [41:11<47:11,  2.02it/s]
 47%|████▋     | 4976/10682 [41:11<47:16,  2.01it/s]
 47%|████▋     | 4977/10682 [41:12<47:15,  2.01it/s]
 47%|████▋     | 4978/10682 [41:12<47:15,  2.01it/s]
 47%|████▋     | 4979/10682 [41:13<47:14,  2.01it/s]
 47%|████▋     | 4980/10682 [41:13<47:09,  2.02it/s]
 47%|████▋     | 4981/10682 [41:14<47:12,  2.01it/s]
 47%|████▋     | 4982/10682 [41:14<47:13,  2.01it/s]
 47%|████▋     | 4983/10682 [41:15<47:05,  2.02it/s]
 47%|████▋     | 4984/10682 [41:15<47:06,  2.02it/s]
 47%|████▋     | 4985/10682 [41:16<47:02,  2.02it/s]
 47%|████▋     | 4986/10682 [41:16<47:01,  2.02it/s]
 47%|████▋     | 4987/10682 [41:17<46:59,  2.02it/s]
 47%|████▋     | 4988/10682 [41:17<46:58,  2.02it/s]
 47%|████▋     | 4989/10682 [41:18<47:00,  2.02it/s]
 47%|████▋     | 4990/10682 [41:18<46:58,  2.02it/s]
 47%|████▋     | 4991/10682 [41:19<47:02,  2.02it/s]
 47%|████▋     | 4992/10682 [41:19<46:59,  2.02it/s]
 47%|████▋     | 4993/10682 [41:20<47:00,  2.02it/s]
 47%|████▋     | 4994/10682 [41:20<46:56,  2.02it/s]
 47%|████▋     | 4995/10682 [41:21<46:58,  2.02it/s]
 47%|████▋     | 4996/10682 [41:21<46:59,  2.02it/s]
 47%|████▋     | 4997/10682 [41:22<47:01,  2.01it/s]
 47%|████▋     | 4998/10682 [41:22<47:02,  2.01it/s]
 47%|████▋     | 4999/10682 [41:23<47:01,  2.01it/s]
 47%|████▋     | 5000/10682 [41:23<46:59,  2.02it/s]
                                                    {'loss': 3.5855, 'grad_norm': 0.2146376073360443, 'learning_rate': 0.0006411156890315328, 'epoch': 0.47}
-
 47%|████▋     | 5000/10682 [41:23<46:59,  2.02it/s]
 47%|████▋     | 5001/10682 [41:24<47:01,  2.01it/s]
 47%|████▋     | 5002/10682 [41:24<46:57,  2.02it/s]
 47%|████▋     | 5003/10682 [41:25<46:54,  2.02it/s]
 47%|████▋     | 5004/10682 [41:25<46:54,  2.02it/s]
 47%|████▋     | 5005/10682 [41:26<46:54,  2.02it/s]
 47%|████▋     | 5006/10682 [41:26<47:00,  2.01it/s]
 47%|████▋     | 5007/10682 [41:27<46:57,  2.01it/s]
 47%|████▋     | 5008/10682 [41:27<46:56,  2.01it/s]
 47%|████▋     | 5009/10682 [41:28<46:52,  2.02it/s]
 47%|████▋     | 5010/10682 [41:28<46:55,  2.01it/s]
 47%|████▋     | 5011/10682 [41:29<46:51,  2.02it/s]
 47%|████▋     | 5012/10682 [41:29<46:49,  2.02it/s]
 47%|████▋     | 5013/10682 [41:30<46:47,  2.02it/s]
 47%|████▋     | 5014/10682 [41:30<46:47,  2.02it/s]
 47%|████▋     | 5015/10682 [41:31<46:50,  2.02it/s]
 47%|████▋     | 5016/10682 [41:31<46:45,  2.02it/s]
 47%|████▋     | 5017/10682 [41:32<46:51,  2.01it/s]
 47%|████▋     | 5018/10682 [41:32<46:50,  2.02it/s]
 47%|████▋     | 5019/10682 [41:33<46:47,  2.02it/s]
 47%|████▋     | 5020/10682 [41:33<46:46,  2.02it/s]
 47%|████▋     | 5021/10682 [41:34<46:44,  2.02it/s]
 47%|████▋     | 5022/10682 [41:34<46:41,  2.02it/s]
 47%|████▋     | 5023/10682 [41:35<46:43,  2.02it/s]
 47%|████▋     | 5024/10682 [41:35<46:42,  2.02it/s]
 47%|████▋     | 5025/10682 [41:36<46:44,  2.02it/s]
                                                    
{'loss': 3.5723, 'grad_norm': 0.2215278446674347, 'learning_rate': 0.0006371920127491398, 'epoch': 0.47}
-
 47%|████▋     | 5025/10682 [41:36<46:44,  2.02it/s]
 47%|████▋     | 5026/10682 [41:36<46:47,  2.01it/s]
 47%|████▋     | 5027/10682 [41:37<46:45,  2.02it/s]
 47%|████▋     | 5028/10682 [41:37<46:46,  2.01it/s]
 47%|████▋     | 5029/10682 [41:38<46:40,  2.02it/s]
 47%|████▋     | 5030/10682 [41:38<46:42,  2.02it/s]
 47%|████▋     | 5031/10682 [41:39<46:39,  2.02it/s]
 47%|████▋     | 5032/10682 [41:39<46:41,  2.02it/s]
 47%|████▋     | 5033/10682 [41:40<46:39,  2.02it/s]
 47%|████▋     | 5034/10682 [41:40<46:39,  2.02it/s]
 47%|████▋     | 5035/10682 [41:41<46:38,  2.02it/s]
 47%|████▋     | 5036/10682 [41:41<46:37,  2.02it/s]
 47%|████▋     | 5037/10682 [41:42<46:42,  2.01it/s]
 47%|████▋     | 5038/10682 [41:42<46:38,  2.02it/s]
 47%|████▋     | 5039/10682 [41:43<46:37,  2.02it/s]
 47%|████▋     | 5040/10682 [41:43<46:36,  2.02it/s]
 47%|████▋     | 5041/10682 [41:43<46:36,  2.02it/s]
 47%|████▋     | 5042/10682 [41:44<46:34,  2.02it/s]
 47%|████▋     | 5043/10682 [41:44<46:33,  2.02it/s]
 47%|████▋     | 5044/10682 [41:45<46:33,  2.02it/s]
 47%|████▋     | 5045/10682 [41:45<46:34,  2.02it/s]
 47%|████▋     | 5046/10682 [41:46<46:38,  2.01it/s]
 47%|████▋     | 5047/10682 [41:46<46:37,  2.01it/s]
 47%|████▋     | 5048/10682 [41:47<46:35,  2.02it/s]
 47%|████▋     | 5049/10682 [41:47<46:30,  2.02it/s]
 47%|████▋     | 5050/10682 [41:48<46:30,  2.02it/s]
                                                    {'loss': 3.5722, 'grad_norm': 0.21945799887180328, 'learning_rate': 0.0006332591787271516, 'epoch': 0.47}
-
 47%|████▋     | 5050/10682 [41:48<46:30,  2.02it/s]
 47%|████▋     | 5051/10682 [41:48<46:31,  2.02it/s]
 47%|████▋     | 5052/10682 [41:49<46:30,  2.02it/s]
 47%|████▋     | 5053/10682 [41:49<46:32,  2.02it/s]
 47%|████▋     | 5054/10682 [41:50<46:30,  2.02it/s]
 47%|████▋     | 5055/10682 [41:50<46:31,  2.02it/s]
 47%|████▋     | 5056/10682 [41:51<46:28,  2.02it/s]
 47%|████▋     | 5057/10682 [41:51<46:28,  2.02it/s]
 47%|████▋     | 5058/10682 [41:52<46:24,  2.02it/s]
 47%|████▋     | 5059/10682 [41:52<46:26,  2.02it/s]
 47%|████▋     | 5060/10682 [41:53<46:22,  2.02it/s]
 47%|████▋     | 5061/10682 [41:53<46:25,  2.02it/s]
 47%|████▋     | 5062/10682 [41:54<46:23,  2.02it/s]
 47%|████▋     | 5063/10682 [41:54<46:23,  2.02it/s]
 47%|████▋     | 5064/10682 [41:55<46:22,  2.02it/s]
 47%|████▋     | 5065/10682 [41:55<46:22,  2.02it/s]
 47%|████▋     | 5066/10682 [41:56<46:23,  2.02it/s]
 47%|████▋     | 5067/10682 [41:56<46:24,  2.02it/s]
 47%|████▋     | 5068/10682 [41:57<46:24,  2.02it/s]
 47%|████▋     | 5069/10682 [41:57<46:27,  2.01it/s]
 47%|████▋     | 5070/10682 [41:58<46:24,  2.02it/s]
 47%|████▋     | 5071/10682 [41:58<46:20,  2.02it/s]
 47%|████▋     | 5072/10682 [41:59<46:21,  2.02it/s]
 47%|████▋     | 5073/10682 [41:59<46:20,  2.02it/s]
 48%|████▊     | 5074/10682 [42:00<46:26,  2.01it/s]
 48%|████▊     | 5075/10682 [42:00<46:25,  2.01it/s]
                                                    {'loss': 3.5799, 'grad_norm': 0.22343337535858154, 'learning_rate': 0.0006293174494871916, 'epoch': 0.48}
-
 48%|████▊     | 5075/10682 [42:00<46:25,  2.01it/s]
 48%|████▊     | 5076/10682 [42:01<46:25,  2.01it/s]
 48%|████▊     | 5077/10682 [42:01<46:19,  2.02it/s]
 48%|████▊     | 5078/10682 [42:02<46:19,  2.02it/s]
 48%|████▊     | 5079/10682 [42:02<46:19,  2.02it/s]
 48%|████▊     | 5080/10682 [42:03<46:17,  2.02it/s]
 48%|████▊     | 5081/10682 [42:03<46:16,  2.02it/s]
 48%|████▊     | 5082/10682 [42:04<46:15,  2.02it/s]
 48%|████▊     | 5083/10682 [42:04<46:15,  2.02it/s]
 48%|████▊     | 5084/10682 [42:05<46:15,  2.02it/s]
 48%|████▊     | 5085/10682 [42:05<46:13,  2.02it/s]
 48%|████▊     | 5086/10682 [42:06<46:18,  2.01it/s]
 48%|████▊     | 5087/10682 [42:06<46:18,  2.01it/s]
 48%|████▊     | 5088/10682 [42:07<46:16,  2.01it/s]
 48%|████▊     | 5089/10682 [42:07<46:15,  2.02it/s]
 48%|████▊     | 5090/10682 [42:08<46:14,  2.02it/s]
 48%|████▊     | 5091/10682 [42:08<46:13,  2.02it/s]
 48%|████▊     | 5092/10682 [42:09<46:13,  2.02it/s]
 48%|████▊     | 5093/10682 [42:09<46:13,  2.02it/s]
 48%|████▊     | 5094/10682 [42:10<46:09,  2.02it/s]
 48%|████▊     | 5095/10682 [42:10<46:07,  2.02it/s]
 48%|████▊     | 5096/10682 [42:11<46:11,  2.02it/s]
 48%|████▊     | 5097/10682 [42:11<46:10,  2.02it/s]
 48%|████▊     | 5098/10682 [42:12<46:11,  2.02it/s]
 48%|████▊     | 5099/10682 [42:12<46:08,  2.02it/s]
 48%|████▊     | 5100/10682 [42:13<46:07,  2.02it/s]
                                                    {'loss': 3.564, 'grad_norm': 0.23749935626983643, 'learning_rate': 0.0006253670881446494, 'epoch': 0.48}
-
 48%|████▊     | 5100/10682 [42:13<46:07,  2.02it/s]
 48%|████▊     | 5101/10682 [42:13<46:15,  2.01it/s]
 48%|████▊     | 5102/10682 [42:14<46:12,  2.01it/s]
 48%|████▊     | 5103/10682 [42:14<46:08,  2.01it/s]
 48%|████▊     | 5104/10682 [42:15<46:07,  2.02it/s]
 48%|████▊     | 5105/10682 [42:15<46:09,  2.01it/s]
 48%|████▊     | 5106/10682 [42:16<46:11,  2.01it/s]
 48%|████▊     | 5107/10682 [42:16<46:06,  2.02it/s]
 48%|████▊     | 5108/10682 [42:17<46:04,  2.02it/s]
 48%|████▊     | 5109/10682 [42:17<45:59,  2.02it/s]
 48%|████▊     | 5110/10682 [42:18<45:59,  2.02it/s]
 48%|████▊     | 5111/10682 [42:18<45:56,  2.02it/s]
 48%|████▊     | 5112/10682 [42:19<45:59,  2.02it/s]
 48%|████▊     | 5113/10682 [42:19<45:54,  2.02it/s]
 48%|████▊     | 5114/10682 [42:20<45:56,  2.02it/s]
 48%|████▊     | 5115/10682 [42:20<45:54,  2.02it/s]
 48%|████▊     | 5116/10682 [42:21<46:00,  2.02it/s]
 48%|████▊     | 5117/10682 [42:21<46:03,  2.01it/s]
 48%|████▊     | 5118/10682 [42:22<46:00,  2.02it/s]
 48%|████▊     | 5119/10682 [42:22<45:57,  2.02it/s]
 48%|████▊     | 5120/10682 [42:23<45:55,  2.02it/s]
 48%|████▊     | 5121/10682 [42:23<45:56,  2.02it/s]
 48%|████▊     | 5122/10682 [42:24<45:52,  2.02it/s]
 48%|████▊     | 5123/10682 [42:24<45:52,  2.02it/s]
 48%|████▊     | 5124/10682 [42:25<45:51,  2.02it/s]
 48%|████▊     | 5125/10682 [42:25<45:51,  2.02it/s]
                                                    
{'loss': 3.5662, 'grad_norm': 0.20944134891033173, 'learning_rate': 0.0006214083583911186, 'epoch': 0.48}
-
 48%|████▊     | 5125/10682 [42:25<45:51,  2.02it/s]
 48%|████▊     | 5126/10682 [42:26<45:56,  2.02it/s]
 48%|████▊     | 5127/10682 [42:26<45:54,  2.02it/s]
 48%|████▊     | 5128/10682 [42:27<45:53,  2.02it/s]
 48%|████▊     | 5129/10682 [42:27<45:50,  2.02it/s]
 48%|████▊     | 5130/10682 [42:28<45:52,  2.02it/s]
 48%|████▊     | 5131/10682 [42:28<45:49,  2.02it/s]
 48%|████▊     | 5132/10682 [42:29<45:54,  2.01it/s]
 48%|████▊     | 5133/10682 [42:29<45:52,  2.02it/s]
 48%|████▊     | 5134/10682 [42:30<45:53,  2.01it/s]
 48%|████▊     | 5135/10682 [42:30<45:50,  2.02it/s]
 48%|████▊     | 5136/10682 [42:31<45:52,  2.01it/s]
 48%|████▊     | 5137/10682 [42:31<45:48,  2.02it/s]
 48%|████▊     | 5138/10682 [42:32<45:48,  2.02it/s]
 48%|████▊     | 5139/10682 [42:32<45:47,  2.02it/s]
 48%|████▊     | 5140/10682 [42:33<45:46,  2.02it/s]
 48%|████▊     | 5141/10682 [42:33<45:47,  2.02it/s]
 48%|████▊     | 5142/10682 [42:34<45:46,  2.02it/s]
 48%|████▊     | 5143/10682 [42:34<45:45,  2.02it/s]
 48%|████▊     | 5144/10682 [42:35<45:46,  2.02it/s]
 48%|████▊     | 5145/10682 [42:35<45:48,  2.01it/s]
 48%|████▊     | 5146/10682 [42:36<45:44,  2.02it/s]
 48%|████▊     | 5147/10682 [42:36<45:43,  2.02it/s]
 48%|████▊     | 5148/10682 [42:37<45:39,  2.02it/s]
 48%|████▊     | 5149/10682 [42:37<45:38,  2.02it/s]
 48%|████▊     | 5150/10682 [42:38<45:36,  2.02it/s]
                                                    
{'loss': 3.5692, 'grad_norm': 0.2170136272907257, 'learning_rate': 0.0006174415244767953, 'epoch': 0.48}
-
 48%|████▊     | 5150/10682 [42:38<45:36,  2.02it/s]
 48%|████▊     | 5151/10682 [42:38<45:42,  2.02it/s]
 48%|████▊     | 5152/10682 [42:39<45:40,  2.02it/s]
 48%|████▊     | 5153/10682 [42:39<45:41,  2.02it/s]
 48%|████▊     | 5154/10682 [42:40<45:42,  2.02it/s]
 48%|████▊     | 5155/10682 [42:40<45:41,  2.02it/s]
 48%|████▊     | 5156/10682 [42:41<45:40,  2.02it/s]
 48%|████▊     | 5157/10682 [42:41<45:38,  2.02it/s]
 48%|████▊     | 5158/10682 [42:42<45:37,  2.02it/s]
 48%|████▊     | 5159/10682 [42:42<45:33,  2.02it/s]
 48%|████▊     | 5160/10682 [42:42<45:32,  2.02it/s]
 48%|████▊     | 5161/10682 [42:43<45:31,  2.02it/s]
 48%|████▊     | 5162/10682 [42:43<45:31,  2.02it/s]
 48%|████▊     | 5163/10682 [42:44<45:32,  2.02it/s]
 48%|████▊     | 5164/10682 [42:44<45:37,  2.02it/s]
 48%|████▊     | 5165/10682 [42:45<45:33,  2.02it/s]
 48%|████▊     | 5166/10682 [42:45<45:34,  2.02it/s]
 48%|████▊     | 5167/10682 [42:46<45:31,  2.02it/s]
 48%|████▊     | 5168/10682 [42:46<45:33,  2.02it/s]
 48%|████▊     | 5169/10682 [42:47<45:31,  2.02it/s]
 48%|████▊     | 5170/10682 [42:47<45:32,  2.02it/s]
 48%|████▊     | 5171/10682 [42:48<45:30,  2.02it/s]
 48%|████▊     | 5172/10682 [42:48<45:30,  2.02it/s]
 48%|████▊     | 5173/10682 [42:49<45:26,  2.02it/s]
 48%|████▊     | 5174/10682 [42:49<45:30,  2.02it/s]
 48%|████▊     | 5175/10682 [42:50<45:37,  2.01it/s]
                                                    
{'loss': 3.5715, 'grad_norm': 0.22715036571025848, 'learning_rate': 0.0006134668511928376, 'epoch': 0.48}
-
 48%|████▊     | 5175/10682 [42:50<45:37,  2.01it/s]
 48%|████▊     | 5176/10682 [42:50<45:46,  2.00it/s]
 48%|████▊     | 5177/10682 [42:51<45:37,  2.01it/s]
 48%|████▊     | 5178/10682 [42:51<45:34,  2.01it/s]
 48%|████▊     | 5179/10682 [42:52<45:29,  2.02it/s]
 48%|████▊     | 5180/10682 [42:52<45:31,  2.01it/s]
 49%|████▊     | 5181/10682 [42:53<45:28,  2.02it/s]
 49%|████▊     | 5182/10682 [42:53<45:28,  2.02it/s]
 49%|████▊     | 5183/10682 [42:54<45:30,  2.01it/s]
 49%|████▊     | 5184/10682 [42:54<45:27,  2.02it/s]
 49%|████▊     | 5185/10682 [42:55<45:27,  2.02it/s]
 49%|████▊     | 5186/10682 [42:55<45:24,  2.02it/s]
 49%|████▊     | 5187/10682 [42:56<45:24,  2.02it/s]
 49%|████▊     | 5188/10682 [42:56<45:20,  2.02it/s]
 49%|████▊     | 5189/10682 [42:57<45:22,  2.02it/s]
 49%|████▊     | 5190/10682 [42:57<45:19,  2.02it/s]
 49%|████▊     | 5191/10682 [42:58<45:25,  2.01it/s]
 49%|████▊     | 5192/10682 [42:58<45:23,  2.02it/s]
 49%|████▊     | 5193/10682 [42:59<45:21,  2.02it/s]
 49%|████▊     | 5194/10682 [42:59<45:22,  2.02it/s]
 49%|████▊     | 5195/10682 [43:00<45:20,  2.02it/s]
 49%|████▊     | 5196/10682 [43:00<45:21,  2.02it/s]
 49%|████▊     | 5197/10682 [43:01<45:20,  2.02it/s]
 49%|████▊     | 5198/10682 [43:01<45:20,  2.02it/s]
 49%|████▊     | 5199/10682 [43:02<45:18,  2.02it/s]
 49%|████▊     | 5200/10682 [43:02<45:18,  2.02it/s]{'loss': 3.5669, 'grad_norm': 0.20816950500011444, 'learning_rate': 0.0006094846038536922, 'epoch': 0.49}
                                                    
-
 49%|████▊     | 5200/10682 [43:02<45:18,  2.02it/s]
 49%|████▊     | 5201/10682 [43:03<45:19,  2.02it/s]
 49%|████▊     | 5202/10682 [43:03<45:17,  2.02it/s]
 49%|████▊     | 5203/10682 [43:04<45:17,  2.02it/s]
 49%|████▊     | 5204/10682 [43:04<45:17,  2.02it/s]
 49%|████▊     | 5205/10682 [43:05<45:15,  2.02it/s]
 49%|████▊     | 5206/10682 [43:05<45:13,  2.02it/s]
 49%|████▊     | 5207/10682 [43:06<45:13,  2.02it/s]
 49%|████▉     | 5208/10682 [43:06<45:12,  2.02it/s]
 49%|████▉     | 5209/10682 [43:07<45:13,  2.02it/s]
 49%|████▉     | 5210/10682 [43:07<45:13,  2.02it/s]
 49%|████▉     | 5211/10682 [43:08<45:14,  2.02it/s]
 49%|████▉     | 5212/10682 [43:08<45:11,  2.02it/s]
 49%|████▉     | 5213/10682 [43:09<45:17,  2.01it/s]
 49%|████▉     | 5214/10682 [43:09<45:13,  2.01it/s]
 49%|████▉     | 5215/10682 [43:10<45:13,  2.01it/s]
 49%|████▉     | 5216/10682 [43:10<45:08,  2.02it/s]
 49%|████▉     | 5217/10682 [43:11<45:11,  2.02it/s]
 49%|████▉     | 5218/10682 [43:11<45:09,  2.02it/s]
 49%|████▉     | 5219/10682 [43:12<45:10,  2.02it/s]
 49%|████▉     | 5220/10682 [43:12<45:07,  2.02it/s]
 49%|████▉     | 5221/10682 [43:13<45:11,  2.01it/s]
 49%|████▉     | 5222/10682 [43:13<45:13,  2.01it/s]
 49%|████▉     | 5223/10682 [43:14<45:08,  2.02it/s]
 49%|████▉     | 5224/10682 [43:14<45:09,  2.01it/s]
 49%|████▉     | 5225/10682 [43:15<45:07,  2.02it/s]
                                                    {'loss': 3.5687, 'grad_norm': 0.21576984226703644, 'learning_rate': 0.000605495048279383, 'epoch': 0.49}
-

 49%|████▉     | 5225/10682 [43:15<45:07,  2.02it/s]
 49%|████▉     | 5226/10682 [43:15<45:10,  2.01it/s]
 49%|████▉     | 5227/10682 [43:16<45:08,  2.01it/s]
 49%|████▉     | 5228/10682 [43:16<45:06,  2.02it/s]
 49%|████▉     | 5229/10682 [43:17<45:00,  2.02it/s]
 49%|████▉     | 5230/10682 [43:17<45:01,  2.02it/s]
 49%|████▉     | 5231/10682 [43:18<45:01,  2.02it/s]
 49%|████▉     | 5232/10682 [43:18<45:04,  2.02it/s]
 49%|████▉     | 5233/10682 [43:19<45:05,  2.01it/s]
 49%|████▉     | 5234/10682 [43:19<45:03,  2.01it/s]
 49%|████��     | 5235/10682 [43:20<45:02,  2.02it/s]
 49%|████▉     | 5236/10682 [43:20<45:00,  2.02it/s]
 49%|████▉     | 5237/10682 [43:21<44:58,  2.02it/s]
 49%|████▉     | 5238/10682 [43:21<45:00,  2.02it/s]
 49%|████▉     | 5239/10682 [43:22<44:55,  2.02it/s]
 49%|████▉     | 5240/10682 [43:22<44:57,  2.02it/s]
 49%|████▉     | 5241/10682 [43:23<44:55,  2.02it/s]
 49%|████▉     | 5242/10682 [43:23<44:59,  2.02it/s]
 49%|████▉     | 5243/10682 [43:24<44:52,  2.02it/s]
 49%|████▉     | 5244/10682 [43:24<44:52,  2.02it/s]
 49%|████▉     | 5245/10682 [43:25<44:50,  2.02it/s]
 49%|████▉     | 5246/10682 [43:25<44:53,  2.02it/s]
 49%|████▉     | 5247/10682 [43:26<44:49,  2.02it/s]
 49%|████▉     | 5248/10682 [43:26<44:50,  2.02it/s]
 49%|████▉     | 5249/10682 [43:27<44:46,  2.02it/s]
 49%|████▉     | 5250/10682 [43:27<44:49,  2.02it/s]
                                                    {'loss': 3.5698, 'grad_norm': 0.24666103720664978, 'learning_rate': 0.0006014984507777682, 'epoch': 0.49}
-
 49%|████▉     | 5250/10682 [43:27<44:49,  2.02it/s]
 49%|████▉     | 5251/10682 [43:28<44:54,  2.02it/s]
 49%|████▉     | 5252/10682 [43:28<44:56,  2.01it/s]
 49%|████▉     | 5253/10682 [43:29<44:56,  2.01it/s]
 49%|████▉     | 5254/10682 [43:29<44:51,  2.02it/s]
 49%|████▉     | 5255/10682 [43:30<44:52,  2.02it/s]
 49%|████▉     | 5256/10682 [43:30<44:49,  2.02it/s]
 49%|████▉     | 5257/10682 [43:31<44:49,  2.02it/s]
 49%|████▉     | 5258/10682 [43:31<44:45,  2.02it/s]
 49%|████▉     | 5259/10682 [43:32<44:47,  2.02it/s]
 49%|████▉     | 5260/10682 [43:32<44:46,  2.02it/s]
 49%|████▉     | 5261/10682 [43:33<44:49,  2.02it/s]
 49%|████▉     | 5262/10682 [43:33<44:50,  2.01it/s]
 49%|████▉     | 5263/10682 [43:34<44:46,  2.02it/s]
 49%|████▉     | 5264/10682 [43:34<44:47,  2.02it/s]
 49%|████▉     | 5265/10682 [43:35<44:41,  2.02it/s]
 49%|████▉     | 5266/10682 [43:35<44:43,  2.02it/s]
 49%|████▉     | 5267/10682 [43:36<44:39,  2.02it/s]
 49%|████▉     | 5268/10682 [43:36<44:42,  2.02it/s]
 49%|████▉     | 5269/10682 [43:37<44:41,  2.02it/s]
 49%|████▉     | 5270/10682 [43:37<44:39,  2.02it/s]
 49%|████▉     | 5271/10682 [43:38<44:40,  2.02it/s]
 49%|████▉     | 5272/10682 [43:38<44:39,  2.02it/s]
 49%|████▉     | 5273/10682 [43:39<44:40,  2.02it/s]
 49%|████▉     | 5274/10682 [43:39<44:39,  2.02it/s]
 49%|████▉     | 5275/10682 [43:40<44:39,  2.02it/s]
                                                    {'loss': 3.5552, 'grad_norm': 0.20812074840068817, 'learning_rate': 0.000597495078126763, 'epoch': 0.49}
-

 49%|████▉     | 5275/10682 [43:40<44:39,  2.02it/s]
 49%|████▉     | 5276/10682 [43:40<44:46,  2.01it/s]
 49%|████▉     | 5277/10682 [43:41<44:43,  2.01it/s]
 49%|████▉     | 5278/10682 [43:41<44:44,  2.01it/s]
 49%|████▉     | 5279/10682 [43:42<44:41,  2.01it/s]
 49%|████▉     | 5280/10682 [43:42<44:43,  2.01it/s]
 49%|████▉     | 5281/10682 [43:42<44:44,  2.01it/s]
 49%|████▉     | 5282/10682 [43:43<44:41,  2.01it/s]
 49%|████▉     | 5283/10682 [43:43<44:42,  2.01it/s]
 49%|████▉     | 5284/10682 [43:44<44:40,  2.01it/s]
 49%|████▉     | 5285/10682 [43:44<44:39,  2.01it/s]
 49%|████▉     | 5286/10682 [43:45<44:33,  2.02it/s]
 49%|████▉     | 5287/10682 [43:45<44:33,  2.02it/s]
 50%|████▉     | 5288/10682 [43:46<44:37,  2.01it/s]
 50%|████▉     | 5289/10682 [43:46<44:37,  2.01it/s]
 50%|████▉     | 5290/10682 [43:47<44:42,  2.01it/s]
 50%|████▉     | 5291/10682 [43:47<44:37,  2.01it/s]
 50%|████▉     | 5292/10682 [43:48<44:36,  2.01it/s]
 50%|████▉     | 5293/10682 [43:48<44:32,  2.02it/s]
 50%|████▉     | 5294/10682 [43:49<44:31,  2.02it/s]
 50%|████▉     | 5295/10682 [43:49<44:30,  2.02it/s]
 50%|████▉     | 5296/10682 [43:50<44:29,  2.02it/s]
 50%|████▉     | 5297/10682 [43:50<44:30,  2.02it/s]
 50%|████▉     | 5298/10682 [43:51<44:28,  2.02it/s]
 50%|████▉     | 5299/10682 [43:51<44:30,  2.02it/s]
 50%|████▉     | 5300/10682 [43:52<44:29,  2.02it/s]{'loss': 3.5591, 'grad_norm': 0.21730123460292816, 'learning_rate': 0.000593485197556533, 'epoch': 0.5}
-
                                                    

 50%|████▉     | 5300/10682 [43:52<44:29,  2.02it/s]
 50%|████▉     | 5301/10682 [43:52<44:35,  2.01it/s]
 50%|████▉     | 5302/10682 [43:53<44:30,  2.01it/s]
 50%|████▉     | 5303/10682 [43:53<44:29,  2.01it/s]
 50%|████▉     | 5304/10682 [43:54<44:25,  2.02it/s]
 50%|████▉     | 5305/10682 [43:54<44:23,  2.02it/s]
 50%|████▉     | 5306/10682 [43:55<44:24,  2.02it/s]
 50%|████▉     | 5307/10682 [43:55<44:21,  2.02it/s]
 50%|████▉     | 5308/10682 [43:56<44:25,  2.02it/s]
 50%|████▉     | 5309/10682 [43:56<44:19,  2.02it/s]
 50%|████▉     | 5310/10682 [43:57<44:22,  2.02it/s]
 50%|████▉     | 5311/10682 [43:57<44:20,  2.02it/s]
 50%|████▉     | 5312/10682 [43:58<44:22,  2.02it/s]
 50%|████▉     | 5313/10682 [43:58<44:23,  2.02it/s]
 50%|████▉     | 5314/10682 [43:59<44:24,  2.01it/s]
 50%|████▉     | 5315/10682 [43:59<44:24,  2.01it/s]
 50%|████▉     | 5316/10682 [44:00<44:19,  2.02it/s]
 50%|████▉     | 5317/10682 [44:00<44:18,  2.02it/s]
 50%|████▉     | 5318/10682 [44:01<44:17,  2.02it/s]
 50%|████▉     | 5319/10682 [44:01<44:19,  2.02it/s]
 50%|████▉     | 5320/10682 [44:02<44:21,  2.01it/s]
 50%|████▉     | 5321/10682 [44:02<44:19,  2.02it/s]
 50%|████▉     | 5322/10682 [44:03<44:22,  2.01it/s]
 50%|████▉     | 5323/10682 [44:03<44:16,  2.02it/s]
 50%|████▉     | 5324/10682 [44:04<44:17,  2.02it/s]
 50%|████▉     | 5325/10682 [44:04<44:12,  2.02it/s]{'loss': 3.5472, 'grad_norm': 0.21972304582595825, 'learning_rate': 0.0005894690767316552, 'epoch': 0.5}
-
                                                    

 50%|████▉     | 5325/10682 [44:04<44:12,  2.02it/s]
 50%|████▉     | 5326/10682 [44:05<44:15,  2.02it/s]
 50%|████▉     | 5327/10682 [44:05<44:11,  2.02it/s]
 50%|████▉     | 5328/10682 [44:06<44:15,  2.02it/s]
 50%|████▉     | 5329/10682 [44:07<51:32,  1.73it/s]
 50%|████▉     | 5330/10682 [44:07<49:28,  1.80it/s]
 50%|████▉     | 5331/10682 [44:08<47:52,  1.86it/s]
 50%|████▉     | 5332/10682 [44:08<46:46,  1.91it/s]
 50%|████▉     | 5333/10682 [44:09<45:56,  1.94it/s]
 50%|████▉     | 5334/10682 [44:09<45:28,  1.96it/s]
 50%|████▉     | 5335/10682 [44:10<45:00,  1.98it/s]
 50%|████▉     | 5336/10682 [44:10<44:45,  1.99it/s]
 50%|████▉     | 5337/10682 [44:11<44:37,  2.00it/s]
 50%|████▉     | 5338/10682 [44:11<44:27,  2.00it/s]
 50%|████▉     | 5339/10682 [44:12<44:20,  2.01it/s]
 50%|████▉     | 5340/10682 [44:12<44:16,  2.01it/s]
 50%|█████     | 5341/10682 [44:13<44:15,  2.01it/s]
 50%|█████     | 5342/10682 [44:13<44:12,  2.01it/s]
 50%|█████     | 5343/10682 [44:14<44:08,  2.02it/s]
 50%|█████     | 5344/10682 [44:14<44:07,  2.02it/s]
 50%|█████     | 5345/10682 [44:15<44:06,  2.02it/s]
 50%|█████     | 5346/10682 [44:15<44:05,  2.02it/s]
 50%|█████     | 5347/10682 [44:16<44:07,  2.01it/s]
 50%|█████     | 5348/10682 [44:16<44:05,  2.02it/s]
 50%|█████     | 5349/10682 [44:16<44:08,  2.01it/s]
 50%|█████     | 5350/10682 [44:17<44:05,  2.02it/s]
                                                    
{'loss': 3.5518, 'grad_norm': 0.2349366396665573, 'learning_rate': 0.0005854469837332512, 'epoch': 0.5}
-
 50%|█████     | 5350/10682 [44:17<44:05,  2.02it/s]
 50%|█████     | 5351/10682 [44:17<44:09,  2.01it/s]
 50%|█████     | 5352/10682 [44:18<44:08,  2.01it/s]
 50%|█████     | 5353/10682 [44:19<51:10,  1.74it/s]
 50%|█████     | 5354/10682 [44:19<49:01,  1.81it/s]
 50%|█████     | 5355/10682 [44:20<47:29,  1.87it/s]
 50%|█████     | 5356/10682 [44:20<46:26,  1.91it/s]
 50%|█████     | 5357/10682 [44:21<45:44,  1.94it/s]
 50%|█████     | 5358/10682 [44:21<45:09,  1.96it/s]
 50%|█████     | 5359/10682 [44:22<44:52,  1.98it/s]
 50%|█████     | 5360/10682 [44:22<44:37,  1.99it/s]
 50%|█████     | 5361/10682 [44:23<44:28,  1.99it/s]
 50%|█████     | 5362/10682 [44:23<44:16,  2.00it/s]
 50%|█████     | 5363/10682 [44:24<44:09,  2.01it/s]
 50%|█████     | 5364/10682 [44:24<44:04,  2.01it/s]
 50%|█████     | 5365/10682 [44:25<44:00,  2.01it/s]
 50%|█████     | 5366/10682 [44:25<43:59,  2.01it/s]
 50%|█████     | 5367/10682 [44:26<44:00,  2.01it/s]
 50%|█████     | 5368/10682 [44:26<43:56,  2.02it/s]
 50%|█████     | 5369/10682 [44:27<43:56,  2.02it/s]
 50%|█████     | 5370/10682 [44:27<43:56,  2.01it/s]
 50%|█████     | 5371/10682 [44:28<43:52,  2.02it/s]
 50%|█████     | 5372/10682 [44:28<43:53,  2.02it/s]
 50%|█████     | 5373/10682 [44:29<43:48,  2.02it/s]
 50%|█████     | 5374/10682 [44:29<43:54,  2.01it/s]
 50%|█████     | 5375/10682 [44:30<43:55,  2.01it/s]
                                                    {'loss': 3.5484, 'grad_norm': 0.22608642280101776, 'learning_rate': 0.0005814191870410933, 'epoch': 0.5}
-
 50%|█████     | 5375/10682 [44:30<43:55,  2.01it/s]
 50%|█████     | 5376/10682 [44:30<43:55,  2.01it/s]
 50%|█████     | 5377/10682 [44:31<43:55,  2.01it/s]
 50%|█████     | 5378/10682 [44:31<43:51,  2.02it/s]
 50%|█████     | 5379/10682 [44:32<43:50,  2.02it/s]
 50%|█████     | 5380/10682 [44:32<43:49,  2.02it/s]
 50%|█████     | 5381/10682 [44:33<43:55,  2.01it/s]
 50%|█████     | 5382/10682 [44:33<43:50,  2.02it/s]
 50%|█████     | 5383/10682 [44:34<43:48,  2.02it/s]
 50%|█████     | 5384/10682 [44:34<43:43,  2.02it/s]
 50%|█████     | 5385/10682 [44:35<43:42,  2.02it/s]
 50%|█████     | 5386/10682 [44:35<43:41,  2.02it/s]
 50%|█████     | 5387/10682 [44:36<43:43,  2.02it/s]
 50%|█████     | 5388/10682 [44:36<43:38,  2.02it/s]
 50%|█████     | 5389/10682 [44:37<43:40,  2.02it/s]
 50%|█████     | 5390/10682 [44:37<43:36,  2.02it/s]
 50%|█████     | 5391/10682 [44:38<43:39,  2.02it/s]
 50%|█████     | 5392/10682 [44:38<43:38,  2.02it/s]
 50%|█████     | 5393/10682 [44:39<43:40,  2.02it/s]
 50%|█████     | 5394/10682 [44:39<43:39,  2.02it/s]
 51%|█████     | 5395/10682 [44:40<43:39,  2.02it/s]
 51%|█████     | 5396/10682 [44:40<43:39,  2.02it/s]
 51%|█████     | 5397/10682 [44:41<43:38,  2.02it/s]
 51%|█████     | 5398/10682 [44:41<43:36,  2.02it/s]
 51%|█████     | 5399/10682 [44:42<43:40,  2.02it/s]
 51%|█████     | 5400/10682 [44:42<43:38,  2.02it/s]{'loss': 3.5596, 'grad_norm': 0.22293880581855774, 'learning_rate': 0.0005773859555156823, 'epoch': 0.51}
-
                                                    

 51%|█████     | 5400/10682 [44:42<43:38,  2.02it/s]
 51%|█████     | 5401/10682 [44:43<43:40,  2.02it/s]
 51%|█████     | 5402/10682 [44:43<43:35,  2.02it/s]
 51%|█████     | 5403/10682 [44:44<43:37,  2.02it/s]
 51%|█████     | 5404/10682 [44:44<43:33,  2.02it/s]
 51%|█████     | 5405/10682 [44:45<43:32,  2.02it/s]
 51%|█████     | 5406/10682 [44:45<43:32,  2.02it/s]
 51%|█████     | 5407/10682 [44:46<43:35,  2.02it/s]
 51%|█████     | 5408/10682 [44:46<43:35,  2.02it/s]
 51%|█████     | 5409/10682 [44:47<43:37,  2.01it/s]
 51%|█████     | 5410/10682 [44:47<43:34,  2.02it/s]
 51%|█████     | 5411/10682 [44:48<43:37,  2.01it/s]
 51%|█████     | 5412/10682 [44:48<43:32,  2.02it/s]
 51%|█████     | 5413/10682 [44:48<43:34,  2.01it/s]
 51%|█████     | 5414/10682 [44:49<43:29,  2.02it/s]
 51%|█████     | 5415/10682 [44:49<43:28,  2.02it/s]
 51%|█████     | 5416/10682 [44:50<43:26,  2.02it/s]
 51%|█████     | 5417/10682 [44:50<43:26,  2.02it/s]
 51%|█████     | 5418/10682 [44:51<43:28,  2.02it/s]
 51%|█████     | 5419/10682 [44:51<43:31,  2.02it/s]
 51%|█████     | 5420/10682 [44:52<43:28,  2.02it/s]
 51%|█████     | 5421/10682 [44:52<43:26,  2.02it/s]
 51%|█████     | 5422/10682 [44:53<43:26,  2.02it/s]
 51%|█████     | 5423/10682 [44:53<43:23,  2.02it/s]
 51%|█████     | 5424/10682 [44:54<43:25,  2.02it/s]
 51%|█████     | 5425/10682 [44:54<43:20,  2.02it/s]
                                                    {'loss': 3.5542, 'grad_norm': 0.23708032071590424, 'learning_rate': 0.0005733475583803009, 'epoch': 0.51}
-
 51%|█████     | 5425/10682 [44:54<43:20,  2.02it/s]
 51%|█████     | 5426/10682 [44:55<43:27,  2.02it/s]
 51%|█████     | 5427/10682 [44:55<43:23,  2.02it/s]
 51%|█████     | 5428/10682 [44:56<43:25,  2.02it/s]
 51%|█████     | 5429/10682 [44:56<43:21,  2.02it/s]
 51%|█████     | 5430/10682 [44:57<43:21,  2.02it/s]
 51%|█████     | 5431/10682 [44:57<43:22,  2.02it/s]
 51%|█████     | 5432/10682 [44:58<43:21,  2.02it/s]
 51%|█████     | 5433/10682 [44:58<43:16,  2.02it/s]
 51%|█████     | 5434/10682 [44:59<43:18,  2.02it/s]
 51%|█████     | 5435/10682 [44:59<43:24,  2.01it/s]
 51%|█████     | 5436/10682 [45:00<43:20,  2.02it/s]
 51%|█████     | 5437/10682 [45:00<43:20,  2.02it/s]
 51%|█████     | 5438/10682 [45:01<43:17,  2.02it/s]
 51%|█████     | 5439/10682 [45:01<43:17,  2.02it/s]
 51%|█████     | 5440/10682 [45:02<43:16,  2.02it/s]
 51%|█████     | 5441/10682 [45:02<43:15,  2.02it/s]
 51%|█████     | 5442/10682 [45:03<43:12,  2.02it/s]
 51%|█████     | 5443/10682 [45:03<43:13,  2.02it/s]
 51%|█████     | 5444/10682 [45:04<43:11,  2.02it/s]
 51%|█████     | 5445/10682 [45:04<43:12,  2.02it/s]
 51%|█████     | 5446/10682 [45:05<43:11,  2.02it/s]
 51%|█████     | 5447/10682 [45:05<43:14,  2.02it/s]
 51%|█████     | 5448/10682 [45:06<43:11,  2.02it/s]
 51%|█████     | 5449/10682 [45:06<43:14,  2.02it/s]
 51%|█████     | 5450/10682 [45:07<43:13,  2.02it/s]
                                                    
{'loss': 3.5528, 'grad_norm': 0.20878161489963531, 'learning_rate': 0.0005693042652030427, 'epoch': 0.51}
-
 51%|█████     | 5450/10682 [45:07<43:13,  2.02it/s]
 51%|█████     | 5451/10682 [45:07<43:13,  2.02it/s]
 51%|█████     | 5452/10682 [45:08<43:13,  2.02it/s]
 51%|█████     | 5453/10682 [45:08<43:10,  2.02it/s]
 51%|█████     | 5454/10682 [45:09<43:08,  2.02it/s]
 51%|█████     | 5455/10682 [45:09<43:11,  2.02it/s]
 51%|█████     | 5456/10682 [45:10<43:07,  2.02it/s]
 51%|█████     | 5457/10682 [45:10<43:08,  2.02it/s]
 51%|█████     | 5458/10682 [45:11<43:07,  2.02it/s]
 51%|█████     | 5459/10682 [45:11<43:04,  2.02it/s]
 51%|█████     | 5460/10682 [45:12<43:07,  2.02it/s]
 51%|█████     | 5461/10682 [45:12<43:04,  2.02it/s]
 51%|█████     | 5462/10682 [45:13<43:03,  2.02it/s]
 51%|█████     | 5463/10682 [45:13<43:03,  2.02it/s]
 51%|█████     | 5464/10682 [45:14<43:05,  2.02it/s]
 51%|█████     | 5465/10682 [45:14<43:02,  2.02it/s]
 51%|█████     | 5466/10682 [45:15<43:04,  2.02it/s]
 51%|█████     | 5467/10682 [45:15<43:06,  2.02it/s]
 51%|█████     | 5468/10682 [45:16<43:07,  2.02it/s]
 51%|█████     | 5469/10682 [45:16<43:06,  2.02it/s]
 51%|█████     | 5470/10682 [45:17<43:03,  2.02it/s]
 51%|█████     | 5471/10682 [45:17<43:03,  2.02it/s]
 51%|█████     | 5472/10682 [45:18<43:02,  2.02it/s]
 51%|█████     | 5473/10682 [45:18<42:58,  2.02it/s]
 51%|█████     | 5474/10682 [45:19<42:57,  2.02it/s]
 51%|█████▏    | 5475/10682 [45:19<43:00,  2.02it/s]
                                                    {'loss': 3.5589, 'grad_norm': 0.23211058974266052, 'learning_rate': 0.0005652563458788184, 'epoch': 0.51}
-
 51%|█████▏    | 5475/10682 [45:19<43:00,  2.02it/s]
 51%|█████▏    | 5476/10682 [45:20<43:03,  2.01it/s]
 51%|█████▏    | 5477/10682 [45:20<43:03,  2.01it/s]
 51%|█████▏    | 5478/10682 [45:21<42:58,  2.02it/s]
 51%|█████▏    | 5479/10682 [45:21<42:59,  2.02it/s]
 51%|█████▏    | 5480/10682 [45:22<42:58,  2.02it/s]
 51%|█████▏    | 5481/10682 [45:22<42:58,  2.02it/s]
 51%|█████▏    | 5482/10682 [45:23<42:54,  2.02it/s]
 51%|█████▏    | 5483/10682 [45:23<42:54,  2.02it/s]
 51%|█████▏    | 5484/10682 [45:24<42:50,  2.02it/s]
 51%|█████▏    | 5485/10682 [45:24<42:50,  2.02it/s]
 51%|█████▏    | 5486/10682 [45:25<42:52,  2.02it/s]
 51%|█████▏    | 5487/10682 [45:25<42:53,  2.02it/s]
 51%|█████▏    | 5488/10682 [45:26<42:50,  2.02it/s]
 51%|█████▏    | 5489/10682 [45:26<42:50,  2.02it/s]
 51%|█████▏    | 5490/10682 [45:27<42:50,  2.02it/s]
 51%|█████▏    | 5491/10682 [45:27<42:50,  2.02it/s]
 51%|█████▏    | 5492/10682 [45:28<42:48,  2.02it/s]
 51%|█████▏    | 5493/10682 [45:28<42:49,  2.02it/s]
 51%|█████▏    | 5494/10682 [45:29<42:46,  2.02it/s]
 51%|█████▏    | 5495/10682 [45:29<42:49,  2.02it/s]
 51%|█████▏    | 5496/10682 [45:30<42:48,  2.02it/s]
 51%|█████▏    | 5497/10682 [45:30<42:52,  2.02it/s]
 51%|█████▏    | 5498/10682 [45:31<42:48,  2.02it/s]
 51%|█████▏    | 5499/10682 [45:31<42:50,  2.02it/s]
 51%|█████▏    | 5500/10682 [45:32<42:44,  2.02it/s]
                                                    {'loss': 3.5491, 'grad_norm': 0.21205273270606995, 'learning_rate': 0.0005612040706113396, 'epoch': 0.51}
-
 51%|█████▏    | 5500/10682 [45:32<42:44,  2.02it/s]
 51%|█████▏    | 5501/10682 [45:32<42:52,  2.01it/s]
 52%|█████▏    | 5502/10682 [45:33<42:46,  2.02it/s]
 52%|█████▏    | 5503/10682 [45:33<42:44,  2.02it/s]
 52%|█████▏    | 5504/10682 [45:34<42:43,  2.02it/s]
 52%|█████▏    | 5505/10682 [45:34<42:44,  2.02it/s]
 52%|█████▏    | 5506/10682 [45:35<42:49,  2.01it/s]
 52%|█████▏    | 5507/10682 [45:35<42:49,  2.01it/s]
 52%|█████▏    | 5508/10682 [45:36<42:48,  2.01it/s]
 52%|█████▏    | 5509/10682 [45:36<42:43,  2.02it/s]
 52%|█████▏    | 5510/10682 [45:37<42:41,  2.02it/s]
 52%|█████▏    | 5511/10682 [45:37<42:39,  2.02it/s]
 52%|█████▏    | 5512/10682 [45:38<42:39,  2.02it/s]
 52%|█████▏    | 5513/10682 [45:38<42:37,  2.02it/s]
 52%|█████▏    | 5514/10682 [45:39<42:39,  2.02it/s]
 52%|█████▏    | 5515/10682 [45:39<42:35,  2.02it/s]
 52%|█████▏    | 5516/10682 [45:40<42:36,  2.02it/s]
 52%|█████▏    | 5517/10682 [45:40<42:32,  2.02it/s]
 52%|█████▏    | 5518/10682 [45:41<42:34,  2.02it/s]
 52%|█████▏    | 5519/10682 [45:41<42:31,  2.02it/s]
 52%|█████▏    | 5520/10682 [45:41<42:31,  2.02it/s]
 52%|█████▏    | 5521/10682 [45:42<42:28,  2.02it/s]
 52%|█████▏    | 5522/10682 [45:42<42:30,  2.02it/s]
 52%|█████▏    | 5523/10682 [45:43<42:30,  2.02it/s]
 52%|█████▏    | 5524/10682 [45:43<42:32,  2.02it/s]
 52%|█████▏    | 5525/10682 [45:44<42:34,  2.02it/s]
                                                    
{'loss': 3.5526, 'grad_norm': 0.21488770842552185, 'learning_rate': 0.0005571477098950832, 'epoch': 0.52}
-
 52%|█████▏    | 5525/10682 [45:44<42:34,  2.02it/s]
 52%|█████▏    | 5526/10682 [45:44<42:38,  2.02it/s]
 52%|█████▏    | 5527/10682 [45:45<42:34,  2.02it/s]
 52%|█████▏    | 5528/10682 [45:45<42:32,  2.02it/s]
 52%|█████▏    | 5529/10682 [45:46<42:31,  2.02it/s]
 52%|█████▏    | 5530/10682 [45:46<42:32,  2.02it/s]
 52%|█████▏    | 5531/10682 [45:47<42:28,  2.02it/s]
 52%|█████▏    | 5532/10682 [45:47<42:28,  2.02it/s]
 52%|█████▏    | 5533/10682 [45:48<42:28,  2.02it/s]
 52%|█████▏    | 5534/10682 [45:48<42:27,  2.02it/s]
 52%|█████▏    | 5535/10682 [45:49<42:28,  2.02it/s]
 52%|█████▏    | 5536/10682 [45:49<42:30,  2.02it/s]
 52%|█████▏    | 5537/10682 [45:50<42:28,  2.02it/s]
 52%|█████▏    | 5538/10682 [45:50<42:28,  2.02it/s]
 52%|█████▏    | 5539/10682 [45:51<42:26,  2.02it/s]
 52%|█████▏    | 5540/10682 [45:51<42:24,  2.02it/s]
 52%|█████▏    | 5541/10682 [45:52<42:24,  2.02it/s]
 52%|█████▏    | 5542/10682 [45:52<42:23,  2.02it/s]
 52%|█████▏    | 5543/10682 [45:53<42:23,  2.02it/s]
 52%|█████▏    | 5544/10682 [45:53<42:24,  2.02it/s]
 52%|█████▏    | 5545/10682 [45:54<42:23,  2.02it/s]
 52%|█████▏    | 5546/10682 [45:54<42:23,  2.02it/s]
 52%|█████▏    | 5547/10682 [45:55<42:23,  2.02it/s]
 52%|█████▏    | 5548/10682 [45:55<42:22,  2.02it/s]
 52%|█████▏    | 5549/10682 [45:56<42:21,  2.02it/s]
 52%|█████▏    | 5550/10682 [45:56<42:21,  2.02it/s]
                                                    
{'loss': 3.543, 'grad_norm': 0.22506076097488403, 'learning_rate': 0.0005530875344972343, 'epoch': 0.52}
-
 52%|█████▏    | 5550/10682 [45:56<42:21,  2.02it/s]
 52%|█████▏    | 5551/10682 [45:57<42:20,  2.02it/s]
 52%|█████▏    | 5552/10682 [45:57<42:18,  2.02it/s]
 52%|█████▏    | 5553/10682 [45:58<42:18,  2.02it/s]
 52%|█████▏    | 5554/10682 [45:58<42:17,  2.02it/s]
 52%|█████▏    | 5555/10682 [45:59<42:16,  2.02it/s]
 52%|█████▏    | 5556/10682 [45:59<42:17,  2.02it/s]
 52%|█████▏    | 5557/10682 [46:00<42:16,  2.02it/s]
 52%|█████▏    | 5558/10682 [46:00<42:14,  2.02it/s]
 52%|█████▏    | 5559/10682 [46:01<42:15,  2.02it/s]
 52%|█████▏    | 5560/10682 [46:01<42:15,  2.02it/s]
 52%|█████▏    | 5561/10682 [46:02<42:18,  2.02it/s]
 52%|█████▏    | 5562/10682 [46:02<42:14,  2.02it/s]
 52%|█████▏    | 5563/10682 [46:03<42:15,  2.02it/s]
 52%|█████▏    | 5564/10682 [46:03<42:14,  2.02it/s]
 52%|█████▏    | 5565/10682 [46:04<42:13,  2.02it/s]
 52%|█████▏    | 5566/10682 [46:04<42:12,  2.02it/s]
 52%|█████▏    | 5567/10682 [46:05<42:12,  2.02it/s]
 52%|█████▏    | 5568/10682 [46:05<42:10,  2.02it/s]
 52%|█████▏    | 5569/10682 [46:06<42:07,  2.02it/s]
 52%|█████▏    | 5570/10682 [46:06<42:09,  2.02it/s]
 52%|█████▏    | 5571/10682 [46:07<42:08,  2.02it/s]
 52%|█████▏    | 5572/10682 [46:07<42:08,  2.02it/s]
 52%|█████▏    | 5573/10682 [46:08<42:06,  2.02it/s]
 52%|█████▏    | 5574/10682 [46:08<42:07,  2.02it/s]
 52%|█████▏    | 5575/10682 [46:09<42:06,  2.02it/s]
                                                    
{'loss': 3.537, 'grad_norm': 0.23361843824386597, 'learning_rate': 0.0005490238154396129, 'epoch': 0.52}
-
 52%|█████▏    | 5575/10682 [46:09<42:06,  2.02it/s]
 52%|█████▏    | 5576/10682 [46:09<42:11,  2.02it/s]
 52%|█████▏    | 5577/10682 [46:10<42:09,  2.02it/s]
 52%|█████▏    | 5578/10682 [46:10<42:07,  2.02it/s]
 52%|█████▏    | 5579/10682 [46:11<42:03,  2.02it/s]
 52%|█████▏    | 5580/10682 [46:11<42:05,  2.02it/s]
 52%|█████▏    | 5581/10682 [46:12<42:03,  2.02it/s]
 52%|█████▏    | 5582/10682 [46:12<42:05,  2.02it/s]
 52%|█████▏    | 5583/10682 [46:13<42:03,  2.02it/s]
 52%|█████▏    | 5584/10682 [46:13<42:06,  2.02it/s]
 52%|█████▏    | 5585/10682 [46:14<42:07,  2.02it/s]
 52%|█████▏    | 5586/10682 [46:14<42:04,  2.02it/s]
 52%|█████▏    | 5587/10682 [46:15<42:03,  2.02it/s]
 52%|█████▏    | 5588/10682 [46:15<42:01,  2.02it/s]
 52%|█████▏    | 5589/10682 [46:16<42:00,  2.02it/s]
 52%|█████▏    | 5590/10682 [46:16<41:59,  2.02it/s]
 52%|█████▏    | 5591/10682 [46:17<41:56,  2.02it/s]
 52%|█████▏    | 5592/10682 [46:17<41:57,  2.02it/s]
 52%|█████▏    | 5593/10682 [46:18<41:56,  2.02it/s]
 52%|█████▏    | 5594/10682 [46:18<42:00,  2.02it/s]
 52%|█████▏    | 5595/10682 [46:19<42:02,  2.02it/s]
 52%|█████▏    | 5596/10682 [46:19<41:58,  2.02it/s]
 52%|█████▏    | 5597/10682 [46:20<41:58,  2.02it/s]
 52%|█████▏    | 5598/10682 [46:20<41:56,  2.02it/s]
 52%|█████▏    | 5599/10682 [46:21<41:56,  2.02it/s]
 52%|█████▏    | 5600/10682 [46:21<41:54,  2.02it/s]{'loss': 3.5458, 'grad_norm': 0.21791163086891174, 'learning_rate': 0.0005449568239805831, 'epoch': 0.52}
                                                    
-
 52%|█████▏    | 5600/10682 [46:21<41:54,  2.02it/s]
 52%|█████▏    | 5601/10682 [46:22<42:01,  2.02it/s]
 52%|█████▏    | 5602/10682 [46:22<42:01,  2.01it/s]
 52%|█████▏    | 5603/10682 [46:23<41:58,  2.02it/s]
 52%|█████▏    | 5604/10682 [46:23<42:00,  2.01it/s]
 52%|█████▏    | 5605/10682 [46:24<41:59,  2.02it/s]
 52%|█████▏    | 5606/10682 [46:24<42:00,  2.01it/s]
 52%|█████▏    | 5607/10682 [46:25<41:55,  2.02it/s]
 52%|█████▏    | 5608/10682 [46:25<41:53,  2.02it/s]
 53%|█████▎    | 5609/10682 [46:26<41:54,  2.02it/s]
 53%|█████▎    | 5610/10682 [46:26<41:54,  2.02it/s]
 53%|█████▎    | 5611/10682 [46:27<41:50,  2.02it/s]
 53%|█████▎    | 5612/10682 [46:27<41:51,  2.02it/s]
 53%|█████▎    | 5613/10682 [46:28<41:52,  2.02it/s]
 53%|█████▎    | 5614/10682 [46:28<41:51,  2.02it/s]
 53%|█████▎    | 5615/10682 [46:29<41:48,  2.02it/s]
 53%|█████▎    | 5616/10682 [46:29<41:50,  2.02it/s]
 53%|█████▎    | 5617/10682 [46:30<41:46,  2.02it/s]
 53%|█████▎    | 5618/10682 [46:30<41:47,  2.02it/s]
 53%|█████▎    | 5619/10682 [46:31<41:43,  2.02it/s]
 53%|█████▎    | 5620/10682 [46:31<41:45,  2.02it/s]
 53%|█████▎    | 5621/10682 [46:32<41:44,  2.02it/s]
 53%|█████▎    | 5622/10682 [46:32<41:45,  2.02it/s]
 53%|█████▎    | 5623/10682 [46:32<41:43,  2.02it/s]
 53%|█████▎    | 5624/10682 [46:33<41:43,  2.02it/s]
 53%|█████▎    | 5625/10682 [46:33<41:47,  2.02it/s]
                                                    {'loss': 3.5325, 'grad_norm': 0.20920264720916748, 'learning_rate': 0.0005408868315969458, 'epoch': 0.53}
-
 53%|█████▎    | 5625/10682 [46:33<41:47,  2.02it/s]
 53%|█████▎    | 5626/10682 [46:34<41:49,  2.01it/s]
 53%|█████▎    | 5627/10682 [46:34<41:47,  2.02it/s]
 53%|█████▎    | 5628/10682 [46:35<41:44,  2.02it/s]
 53%|█████▎    | 5629/10682 [46:35<41:41,  2.02it/s]
 53%|█████▎    | 5630/10682 [46:36<41:41,  2.02it/s]
 53%|█████▎    | 5631/10682 [46:36<41:41,  2.02it/s]
 53%|█████▎    | 5632/10682 [46:37<41:39,  2.02it/s]
 53%|█████▎    | 5633/10682 [46:37<41:41,  2.02it/s]
 53%|█████▎    | 5634/10682 [46:38<41:40,  2.02it/s]
 53%|█████▎    | 5635/10682 [46:38<41:38,  2.02it/s]
 53%|█████▎    | 5636/10682 [46:39<41:37,  2.02it/s]
 53%|█████▎    | 5637/10682 [46:39<41:34,  2.02it/s]
 53%|█████▎    | 5638/10682 [46:40<41:35,  2.02it/s]
 53%|█████▎    | 5639/10682 [46:40<41:35,  2.02it/s]
 53%|█████▎    | 5640/10682 [46:41<41:33,  2.02it/s]
 53%|█████▎    | 5641/10682 [46:41<41:31,  2.02it/s]
 53%|█████▎    | 5642/10682 [46:42<41:34,  2.02it/s]
 53%|█████▎    | 5643/10682 [46:42<41:33,  2.02it/s]
 53%|█████▎    | 5644/10682 [46:43<41:35,  2.02it/s]
 53%|█████▎    | 5645/10682 [46:43<41:35,  2.02it/s]
 53%|█████▎    | 5646/10682 [46:44<41:34,  2.02it/s]
 53%|█████▎    | 5647/10682 [46:44<41:29,  2.02it/s]
 53%|█████▎    | 5648/10682 [46:45<41:30,  2.02it/s]
 53%|█████▎    | 5649/10682 [46:45<41:27,  2.02it/s]
 53%|█████▎    | 5650/10682 [46:46<41:27,  2.02it/s]
                                                    
{'loss': 3.5449, 'grad_norm': 0.2258412092924118, 'learning_rate': 0.0005368141099658168, 'epoch': 0.53}
-
 53%|█████▎    | 5650/10682 [46:46<41:27,  2.02it/s]
 53%|█████▎    | 5651/10682 [46:46<41:27,  2.02it/s]
 53%|█████▎    | 5652/10682 [46:47<41:29,  2.02it/s]
 53%|█████▎    | 5653/10682 [46:47<41:26,  2.02it/s]
 53%|█████▎    | 5654/10682 [46:48<41:28,  2.02it/s]
 53%|█████▎    | 5655/10682 [46:48<41:26,  2.02it/s]
 53%|█████▎    | 5656/10682 [46:49<41:26,  2.02it/s]
 53%|█████▎    | 5657/10682 [46:49<41:23,  2.02it/s]
 53%|█████▎    | 5658/10682 [46:50<41:26,  2.02it/s]
 53%|█████▎    | 5659/10682 [46:50<41:23,  2.02it/s]
 53%|█████▎    | 5660/10682 [46:51<41:26,  2.02it/s]
 53%|█████▎    | 5661/10682 [46:51<41:23,  2.02it/s]
 53%|█████▎    | 5662/10682 [46:52<41:26,  2.02it/s]
 53%|█████▎    | 5663/10682 [46:52<41:26,  2.02it/s]
 53%|█████▎    | 5664/10682 [46:53<41:26,  2.02it/s]
 53%|█████▎    | 5665/10682 [46:53<41:22,  2.02it/s]
 53%|█████▎    | 5666/10682 [46:54<41:22,  2.02it/s]
 53%|█████▎    | 5667/10682 [46:54<41:19,  2.02it/s]
 53%|█████▎    | 5668/10682 [46:55<41:20,  2.02it/s]
 53%|█████▎    | 5669/10682 [46:55<41:18,  2.02it/s]
 53%|█████▎    | 5670/10682 [46:56<41:18,  2.02it/s]
 53%|█████▎    | 5671/10682 [46:56<41:17,  2.02it/s]
 53%|█████▎    | 5672/10682 [46:57<41:17,  2.02it/s]
 53%|█████▎    | 5673/10682 [46:57<41:20,  2.02it/s]
 53%|█████▎    | 5674/10682 [46:58<41:18,  2.02it/s]
 53%|█████▎    | 5675/10682 [46:58<41:15,  2.02it/s]{'loss': 3.5375, 'grad_norm': 0.22537115216255188, 'learning_rate': 0.0005327389309464938, 'epoch': 0.53}
                                                    
-
 53%|█████▎    | 5675/10682 [46:58<41:15,  2.02it/s]
 53%|█████▎    | 5676/10682 [46:59<41:19,  2.02it/s]
 53%|█████▎    | 5677/10682 [46:59<41:19,  2.02it/s]
 53%|█████▎    | 5678/10682 [47:00<41:16,  2.02it/s]
 53%|█████▎    | 5679/10682 [47:00<41:13,  2.02it/s]
 53%|█████▎    | 5680/10682 [47:01<41:18,  2.02it/s]
 53%|█████▎    | 5681/10682 [47:01<41:15,  2.02it/s]
 53%|█████▎    | 5682/10682 [47:02<41:21,  2.02it/s]
 53%|█████▎    | 5683/10682 [47:02<41:19,  2.02it/s]
 53%|█████▎    | 5684/10682 [47:03<41:15,  2.02it/s]
 53%|█████▎    | 5685/10682 [47:03<41:14,  2.02it/s]
 53%|█████▎    | 5686/10682 [47:04<41:14,  2.02it/s]
 53%|█████▎    | 5687/10682 [47:04<41:13,  2.02it/s]
 53%|█████▎    | 5688/10682 [47:05<41:13,  2.02it/s]
 53%|█████▎    | 5689/10682 [47:05<41:11,  2.02it/s]
 53%|█████▎    | 5690/10682 [47:06<41:12,  2.02it/s]
 53%|█████▎    | 5691/10682 [47:06<41:13,  2.02it/s]
 53%|█████▎    | 5692/10682 [47:07<41:14,  2.02it/s]
 53%|█████▎    | 5693/10682 [47:07<41:14,  2.02it/s]
 53%|█████▎    | 5694/10682 [47:08<41:13,  2.02it/s]
 53%|█████▎    | 5695/10682 [47:08<41:13,  2.02it/s]
 53%|█████▎    | 5696/10682 [47:09<41:10,  2.02it/s]
 53%|█████▎    | 5697/10682 [47:09<41:10,  2.02it/s]
 53%|█████▎    | 5698/10682 [47:10<41:09,  2.02it/s]
 53%|█████▎    | 5699/10682 [47:10<41:08,  2.02it/s]
 53%|█████▎    | 5700/10682 [47:11<41:07,  2.02it/s]
                                                    
{'loss': 3.5419, 'grad_norm': 0.22173942625522614, 'learning_rate': 0.0005286615665623071, 'epoch': 0.53}
-
 53%|█████▎    | 5700/10682 [47:11<41:07,  2.02it/s]
 53%|█████▎    | 5701/10682 [47:11<41:10,  2.02it/s]
 53%|█████▎    | 5702/10682 [47:12<41:11,  2.01it/s]
 53%|█████▎    | 5703/10682 [47:12<41:14,  2.01it/s]
 53%|█████▎    | 5704/10682 [47:13<41:13,  2.01it/s]
 53%|█████▎    | 5705/10682 [47:13<41:08,  2.02it/s]
 53%|█████▎    | 5706/10682 [47:14<41:07,  2.02it/s]
 53%|█████▎    | 5707/10682 [47:14<41:04,  2.02it/s]
 53%|█████▎    | 5708/10682 [47:15<41:06,  2.02it/s]
 53%|█████▎    | 5709/10682 [47:15<41:05,  2.02it/s]
 53%|█████▎    | 5710/10682 [47:16<41:03,  2.02it/s]
 53%|█████▎    | 5711/10682 [47:16<41:05,  2.02it/s]
 53%|█████▎    | 5712/10682 [47:17<41:01,  2.02it/s]
 53%|█████▎    | 5713/10682 [47:17<41:04,  2.02it/s]
 53%|█████▎    | 5714/10682 [47:18<41:02,  2.02it/s]
 54%|█████▎    | 5715/10682 [47:18<41:02,  2.02it/s]
 54%|█████▎    | 5716/10682 [47:19<41:01,  2.02it/s]
 54%|█████▎    | 5717/10682 [47:19<41:02,  2.02it/s]
 54%|█████▎    | 5718/10682 [47:20<41:01,  2.02it/s]
 54%|█████▎    | 5719/10682 [47:20<41:01,  2.02it/s]
 54%|█████▎    | 5720/10682 [47:21<41:01,  2.02it/s]
 54%|█████▎    | 5721/10682 [47:21<41:00,  2.02it/s]
 54%|█████▎    | 5722/10682 [47:22<41:03,  2.01it/s]
 54%|█████▎    | 5723/10682 [47:22<40:58,  2.02it/s]
 54%|█████▎    | 5724/10682 [47:23<40:58,  2.02it/s]
 54%|█████▎    | 5725/10682 [47:23<40:55,  2.02it/s]
                                                    {'loss': 3.5283, 'grad_norm': 0.21883758902549744, 'learning_rate': 0.0005245822889824633, 'epoch': 0.54}
-
 54%|█████▎    | 5725/10682 [47:23<40:55,  2.02it/s]
 54%|█████▎    | 5726/10682 [47:24<40:58,  2.02it/s]
 54%|█████▎    | 5727/10682 [47:24<40:57,  2.02it/s]
 54%|█████▎    | 5728/10682 [47:25<40:54,  2.02it/s]
 54%|█████▎    | 5729/10682 [47:25<40:59,  2.01it/s]
 54%|█████▎    | 5730/10682 [47:25<40:56,  2.02it/s]
 54%|█████▎    | 5731/10682 [47:26<41:00,  2.01it/s]
 54%|█████▎    | 5732/10682 [47:26<40:57,  2.01it/s]
 54%|█████▎    | 5733/10682 [47:27<40:55,  2.02it/s]
 54%|█████▎    | 5734/10682 [47:27<40:56,  2.01it/s]
 54%|█████▎    | 5735/10682 [47:28<40:51,  2.02it/s]
 54%|█████▎    | 5736/10682 [47:28<40:52,  2.02it/s]
 54%|█████▎    | 5737/10682 [47:29<40:53,  2.02it/s]
 54%|█████▎    | 5738/10682 [47:29<40:53,  2.02it/s]
 54%|█████▎    | 5739/10682 [47:30<40:50,  2.02it/s]
 54%|█████▎    | 5740/10682 [47:30<40:51,  2.02it/s]
 54%|█████▎    | 5741/10682 [47:31<40:52,  2.01it/s]
 54%|█████▍    | 5742/10682 [47:31<40:50,  2.02it/s]
 54%|█████▍    | 5743/10682 [47:32<40:51,  2.01it/s]
 54%|█████▍    | 5744/10682 [47:32<40:47,  2.02it/s]
 54%|█████▍    | 5745/10682 [47:33<40:47,  2.02it/s]
 54%|█████▍    | 5746/10682 [47:33<40:46,  2.02it/s]
 54%|█████▍    | 5747/10682 [47:34<40:46,  2.02it/s]
 54%|█████▍    | 5748/10682 [47:34<40:44,  2.02it/s]
 54%|█████▍    | 5749/10682 [47:35<40:50,  2.01it/s]
 54%|█████▍    | 5750/10682 [47:35<40:47,  2.02it/s]{'loss': 3.5366, 'grad_norm': 0.22337029874324799, 'learning_rate': 0.0005205013705038776, 'epoch': 0.54}
                                                    
-
 54%|█████▍    | 5750/10682 [47:35<40:47,  2.02it/s]
 54%|█████▍    | 5751/10682 [47:36<40:49,  2.01it/s]
 54%|█████▍    | 5752/10682 [47:36<40:46,  2.02it/s]
 54%|█████▍    | 5753/10682 [47:37<40:45,  2.02it/s]
 54%|█████▍    | 5754/10682 [47:37<40:41,  2.02it/s]
 54%|█████▍    | 5755/10682 [47:38<40:42,  2.02it/s]
 54%|█████▍    | 5756/10682 [47:38<40:37,  2.02it/s]
 54%|█████▍    | 5757/10682 [47:39<40:40,  2.02it/s]
 54%|█████▍    | 5758/10682 [47:39<40:41,  2.02it/s]
 54%|█████▍    | 5759/10682 [47:40<40:41,  2.02it/s]
 54%|█████▍    | 5760/10682 [47:40<40:37,  2.02it/s]
 54%|█████▍    | 5761/10682 [47:41<40:40,  2.02it/s]
 54%|█████▍    | 5762/10682 [47:41<40:38,  2.02it/s]
 54%|█████▍    | 5763/10682 [47:42<40:37,  2.02it/s]
 54%|█████▍    | 5764/10682 [47:42<40:39,  2.02it/s]
 54%|█████▍    | 5765/10682 [47:43<40:37,  2.02it/s]
 54%|█████▍    | 5766/10682 [47:43<40:40,  2.01it/s]
 54%|█████▍    | 5767/10682 [47:44<40:36,  2.02it/s]
 54%|█████▍    | 5768/10682 [47:44<40:40,  2.01it/s]
 54%|█████▍    | 5769/10682 [47:45<40:37,  2.02it/s]
 54%|█████▍    | 5770/10682 [47:45<40:40,  2.01it/s]
 54%|█████▍    | 5771/10682 [47:46<40:39,  2.01it/s]
 54%|█████▍    | 5772/10682 [47:46<40:40,  2.01it/s]
 54%|█████▍    | 5773/10682 [47:47<40:36,  2.01it/s]
 54%|█████▍    | 5774/10682 [47:47<40:33,  2.02it/s]
 54%|█████▍    | 5775/10682 [47:48<40:33,  2.02it/s]
                                                    
{'loss': 3.5313, 'grad_norm': 0.2217974066734314, 'learning_rate': 0.0005164190835329964, 'epoch': 0.54}
-
 54%|█████▍    | 5775/10682 [47:48<40:33,  2.02it/s]
 54%|█████▍    | 5776/10682 [47:48<40:34,  2.02it/s]
 54%|█████▍    | 5777/10682 [47:49<40:33,  2.02it/s]
 54%|█████▍    | 5778/10682 [47:49<40:32,  2.02it/s]
 54%|█████▍    | 5779/10682 [47:50<40:32,  2.02it/s]
 54%|█████▍    | 5780/10682 [47:50<40:28,  2.02it/s]
 54%|█████▍    | 5781/10682 [47:51<40:27,  2.02it/s]
 54%|█████▍    | 5782/10682 [47:51<40:28,  2.02it/s]
 54%|█████▍    | 5783/10682 [47:52<40:26,  2.02it/s]
 54%|█████▍    | 5784/10682 [47:52<40:23,  2.02it/s]
 54%|█████▍    | 5785/10682 [47:53<40:26,  2.02it/s]
 54%|█████▍    | 5786/10682 [47:53<40:26,  2.02it/s]
 54%|█████▍    | 5787/10682 [47:54<40:24,  2.02it/s]
 54%|█████▍    | 5788/10682 [47:54<40:26,  2.02it/s]
 54%|█████▍    | 5789/10682 [47:55<40:27,  2.02it/s]
 54%|█████▍    | 5790/10682 [47:55<40:27,  2.02it/s]
 54%|█████▍    | 5791/10682 [47:56<40:25,  2.02it/s]
 54%|█████▍    | 5792/10682 [47:56<40:24,  2.02it/s]
 54%|█████▍    | 5793/10682 [47:57<40:26,  2.01it/s]
 54%|█████▍    | 5794/10682 [47:57<40:25,  2.02it/s]
 54%|█████▍    | 5795/10682 [47:58<40:21,  2.02it/s]
 54%|█████▍    | 5796/10682 [47:58<40:20,  2.02it/s]
 54%|█████▍    | 5797/10682 [47:59<40:20,  2.02it/s]
 54%|█████▍    | 5798/10682 [47:59<40:20,  2.02it/s]
 54%|█████▍    | 5799/10682 [48:00<40:19,  2.02it/s]
 54%|█████▍    | 5800/10682 [48:00<40:21,  2.02it/s]
                                                    {'loss': 3.5287, 'grad_norm': 0.20533543825149536, 'learning_rate': 0.0005123357005676156, 'epoch': 0.54}
-
 54%|█████▍    | 5800/10682 [48:00<40:21,  2.02it/s]
 54%|█████▍    | 5801/10682 [48:01<40:21,  2.02it/s]
 54%|█████▍    | 5802/10682 [48:01<40:21,  2.02it/s]
 54%|█████▍    | 5803/10682 [48:02<40:16,  2.02it/s]
 54%|█████▍    | 5804/10682 [48:02<40:15,  2.02it/s]
 54%|█████▍    | 5805/10682 [48:03<40:13,  2.02it/s]
 54%|█████▍    | 5806/10682 [48:03<40:15,  2.02it/s]
 54%|█████▍    | 5807/10682 [48:04<40:15,  2.02it/s]
 54%|█████▍    | 5808/10682 [48:04<40:15,  2.02it/s]
 54%|█████▍    | 5809/10682 [48:05<40:16,  2.02it/s]
 54%|█████▍    | 5810/10682 [48:05<40:19,  2.01it/s]
 54%|█████▍    | 5811/10682 [48:06<40:19,  2.01it/s]
 54%|█████▍    | 5812/10682 [48:06<40:17,  2.01it/s]
 54%|█████▍    | 5813/10682 [48:07<40:15,  2.02it/s]
 54%|█████▍    | 5814/10682 [48:07<40:13,  2.02it/s]
 54%|█████▍    | 5815/10682 [48:08<40:11,  2.02it/s]
 54%|█████▍    | 5816/10682 [48:08<40:10,  2.02it/s]
 54%|█████▍    | 5817/10682 [48:09<40:09,  2.02it/s]
 54%|█████▍    | 5818/10682 [48:09<40:09,  2.02it/s]
 54%|█████▍    | 5819/10682 [48:10<40:08,  2.02it/s]
 54%|█████▍    | 5820/10682 [48:10<40:08,  2.02it/s]
 54%|█████▍    | 5821/10682 [48:11<40:06,  2.02it/s]
 55%|█████▍    | 5822/10682 [48:11<40:08,  2.02it/s]
 55%|█████▍    | 5823/10682 [48:12<40:04,  2.02it/s]
 55%|█████▍    | 5824/10682 [48:12<40:05,  2.02it/s]
 55%|█████▍    | 5825/10682 [48:13<40:03,  2.02it/s]
                                                    {'loss': 3.5226, 'grad_norm': 0.22263549268245697, 'learning_rate': 0.0005082514941786895, 'epoch': 0.55}
-
 55%|█████▍    | 5825/10682 [48:13<40:03,  2.02it/s]
 55%|█████▍    | 5826/10682 [48:13<40:08,  2.02it/s]
 55%|█████▍    | 5827/10682 [48:14<40:04,  2.02it/s]
 55%|█████▍    | 5828/10682 [48:14<40:05,  2.02it/s]
 55%|█████▍    | 5829/10682 [48:15<40:05,  2.02it/s]
 55%|█████▍    | 5830/10682 [48:15<40:03,  2.02it/s]
 55%|█████▍    | 5831/10682 [48:16<40:07,  2.01it/s]
 55%|█████▍    | 5832/10682 [48:16<40:05,  2.02it/s]
 55%|█████▍    | 5833/10682 [48:17<40:06,  2.01it/s]
 55%|█████▍    | 5834/10682 [48:17<40:03,  2.02it/s]
 55%|█████▍    | 5835/10682 [48:18<40:02,  2.02it/s]
 55%|█████▍    | 5836/10682 [48:18<39:59,  2.02it/s]
 55%|█████▍    | 5837/10682 [48:19<40:02,  2.02it/s]
 55%|█████▍    | 5838/10682 [48:19<40:05,  2.01it/s]
 55%|█████▍    | 5839/10682 [48:20<40:04,  2.01it/s]
 55%|█████▍    | 5840/10682 [48:20<40:03,  2.01it/s]
 55%|█████▍    | 5841/10682 [48:21<40:00,  2.02it/s]
 55%|█████▍    | 5842/10682 [48:21<40:00,  2.02it/s]
 55%|█████▍    | 5843/10682 [48:22<39:57,  2.02it/s]
 55%|█████▍    | 5844/10682 [48:22<39:57,  2.02it/s]
 55%|█████▍    | 5845/10682 [48:23<39:58,  2.02it/s]
 55%|█████▍    | 5846/10682 [48:23<39:59,  2.02it/s]
 55%|█████▍    | 5847/10682 [48:24<39:57,  2.02it/s]
 55%|█████▍    | 5848/10682 [48:24<39:57,  2.02it/s]
 55%|█████▍    | 5849/10682 [48:25<39:58,  2.02it/s]
 55%|█████▍    | 5850/10682 [48:25<39:54,  2.02it/s]{'loss': 3.5365, 'grad_norm': 0.2260356992483139, 'learning_rate': 0.0005041667369921374, 'epoch': 0.55}
                                                    
-
 55%|█████▍    | 5850/10682 [48:25<39:54,  2.02it/s]
 55%|█████▍    | 5851/10682 [48:25<40:02,  2.01it/s]
 55%|█████▍    | 5852/10682 [48:26<40:00,  2.01it/s]
 55%|█████▍    | 5853/10682 [48:26<39:58,  2.01it/s]
 55%|█████▍    | 5854/10682 [48:27<39:57,  2.01it/s]
 55%|█████▍    | 5855/10682 [48:27<39:52,  2.02it/s]
 55%|█████▍    | 5856/10682 [48:28<39:54,  2.02it/s]
 55%|█████▍    | 5857/10682 [48:28<39:54,  2.02it/s]
 55%|█████▍    | 5858/10682 [48:29<39:56,  2.01it/s]
 55%|█████▍    | 5859/10682 [48:29<39:55,  2.01it/s]
 55%|█████▍    | 5860/10682 [48:30<39:52,  2.02it/s]
 55%|█████▍    | 5861/10682 [48:30<39:54,  2.01it/s]
 55%|█████▍    | 5862/10682 [48:31<39:50,  2.02it/s]
 55%|█████▍    | 5863/10682 [48:31<39:49,  2.02it/s]
 55%|█████▍    | 5864/10682 [48:32<39:47,  2.02it/s]
 55%|█���███▍    | 5865/10682 [48:32<39:47,  2.02it/s]
 55%|█████▍    | 5866/10682 [48:33<39:46,  2.02it/s]
 55%|█████▍    | 5867/10682 [48:33<39:47,  2.02it/s]
 55%|█████▍    | 5868/10682 [48:34<39:46,  2.02it/s]
 55%|█████▍    | 5869/10682 [48:34<39:44,  2.02it/s]
 55%|█████▍    | 5870/10682 [48:35<39:47,  2.02it/s]
 55%|█████▍    | 5871/10682 [48:35<39:42,  2.02it/s]
 55%|█████▍    | 5872/10682 [48:36<39:41,  2.02it/s]
 55%|█████▍    | 5873/10682 [48:36<39:39,  2.02it/s]
 55%|█████▍    | 5874/10682 [48:37<39:41,  2.02it/s]
 55%|█████▍    | 5875/10682 [48:37<39:38,  2.02it/s]{'loss': 3.5216, 'grad_norm': 0.21838359534740448, 'learning_rate': 0.0005000817016706441, 'epoch': 0.55}
                                                    
-
 55%|█████▍    | 5875/10682 [48:37<39:38,  2.02it/s]
 55%|█████▌    | 5876/10682 [48:38<39:44,  2.02it/s]
 55%|█████▌    | 5877/10682 [48:38<39:46,  2.01it/s]
 55%|█████▌    | 5878/10682 [48:39<39:42,  2.02it/s]
 55%|█████▌    | 5879/10682 [48:39<39:41,  2.02it/s]
 55%|█████▌    | 5880/10682 [48:40<39:39,  2.02it/s]
 55%|█████▌    | 5881/10682 [48:40<39:38,  2.02it/s]
 55%|█████▌    | 5882/10682 [48:41<39:35,  2.02it/s]
 55%|█████▌    | 5883/10682 [48:41<39:31,  2.02it/s]
 55%|█████▌    | 5884/10682 [48:42<39:33,  2.02it/s]
 55%|█████▌    | 5885/10682 [48:42<39:34,  2.02it/s]
 55%|█████▌    | 5886/10682 [48:43<39:35,  2.02it/s]
 55%|█████▌    | 5887/10682 [48:43<39:33,  2.02it/s]
 55%|█████▌    | 5888/10682 [48:44<39:38,  2.02it/s]
 55%|█████▌    | 5889/10682 [48:44<39:36,  2.02it/s]
 55%|█████▌    | 5890/10682 [48:45<39:39,  2.01it/s]
 55%|█████▌    | 5891/10682 [48:45<39:36,  2.02it/s]
 55%|█████▌    | 5892/10682 [48:46<39:36,  2.02it/s]
 55%|█████▌    | 5893/10682 [48:46<39:32,  2.02it/s]
 55%|█████▌    | 5894/10682 [48:47<39:31,  2.02it/s]
 55%|█████▌    | 5895/10682 [48:47<39:33,  2.02it/s]
 55%|█████▌    | 5896/10682 [48:48<39:35,  2.01it/s]
 55%|█████▌    | 5897/10682 [48:48<39:36,  2.01it/s]
 55%|█████▌    | 5898/10682 [48:49<39:33,  2.02it/s]
 55%|█████▌    | 5899/10682 [48:49<39:32,  2.02it/s]
 55%|█████▌    | 5900/10682 [48:50<39:30,  2.02it/s]{'loss': 3.5264, 'grad_norm': 0.22030845284461975, 'learning_rate': 0.0004959966608954618, 'epoch': 0.55}
                                                    
-
 55%|█████▌    | 5900/10682 [48:50<39:30,  2.02it/s]
 55%|█████▌    | 5901/10682 [48:50<39:32,  2.01it/s]
 55%|█████▌    | 5902/10682 [48:51<39:31,  2.02it/s]
 55%|█████▌    | 5903/10682 [48:51<39:31,  2.02it/s]
 55%|█████▌    | 5904/10682 [48:52<39:26,  2.02it/s]
 55%|█████▌    | 5905/10682 [48:52<39:30,  2.02it/s]
 55%|█████▌    | 5906/10682 [48:53<39:30,  2.02it/s]
 55%|█████▌    | 5907/10682 [48:53<39:28,  2.02it/s]
 55%|█████▌    | 5908/10682 [48:54<39:29,  2.01it/s]
 55%|█████▌    | 5909/10682 [48:54<39:27,  2.02it/s]
 55%|█████▌    | 5910/10682 [48:55<39:25,  2.02it/s]
 55%|█████▌    | 5911/10682 [48:55<39:22,  2.02it/s]
 55%|█████▌    | 5912/10682 [48:56<39:21,  2.02it/s]
 55%|█████▌    | 5913/10682 [48:56<39:19,  2.02it/s]
 55%|█████▌    | 5914/10682 [48:57<39:19,  2.02it/s]
 55%|█████▌    | 5915/10682 [48:57<39:17,  2.02it/s]
 55%|█████▌    | 5916/10682 [48:58<39:22,  2.02it/s]
 55%|█████▌    | 5917/10682 [48:58<39:21,  2.02it/s]
 55%|█████▌    | 5918/10682 [48:59<39:22,  2.02it/s]
 55%|█████▌    | 5919/10682 [48:59<39:21,  2.02it/s]
 55%|█████▌    | 5920/10682 [49:00<39:22,  2.02it/s]
 55%|█████▌    | 5921/10682 [49:00<39:21,  2.02it/s]
 55%|█████▌    | 5922/10682 [49:01<39:18,  2.02it/s]
 55%|█████▌    | 5923/10682 [49:01<39:21,  2.02it/s]
 55%|█████▌    | 5924/10682 [49:02<39:18,  2.02it/s]
 55%|█████▌    | 5925/10682 [49:02<39:17,  2.02it/s]
                                                    
{'loss': 3.5198, 'grad_norm': 0.2114148586988449, 'learning_rate': 0.0004919118873482054, 'epoch': 0.55}
-
 55%|█████▌    | 5925/10682 [49:02<39:17,  2.02it/s]
 55%|█████▌    | 5926/10682 [49:03<39:25,  2.01it/s]
 55%|█████▌    | 5927/10682 [49:03<39:19,  2.02it/s]
 55%|█████▌    | 5928/10682 [49:04<39:18,  2.02it/s]
 56%|█████▌    | 5929/10682 [49:04<39:15,  2.02it/s]
 56%|█████▌    | 5930/10682 [49:05<39:15,  2.02it/s]
 56%|█████▌    | 5931/10682 [49:05<39:10,  2.02it/s]
 56%|█████▌    | 5932/10682 [49:06<39:11,  2.02it/s]
 56%|█████▌    | 5933/10682 [49:06<39:09,  2.02it/s]
 56%|█████▌    | 5934/10682 [49:07<39:12,  2.02it/s]
 56%|█████▌    | 5935/10682 [49:07<39:09,  2.02it/s]
 56%|█████▌    | 5936/10682 [49:08<39:09,  2.02it/s]
 56%|█████▌    | 5937/10682 [49:08<39:08,  2.02it/s]
 56%|█████▌    | 5938/10682 [49:09<39:08,  2.02it/s]
 56%|█████▌    | 5939/10682 [49:09<39:08,  2.02it/s]
 56%|█████▌    | 5940/10682 [49:10<39:07,  2.02it/s]
 56%|█████▌    | 5941/10682 [49:10<39:09,  2.02it/s]
 56%|█████▌    | 5942/10682 [49:11<39:07,  2.02it/s]
 56%|█████▌    | 5943/10682 [49:11<39:07,  2.02it/s]
 56%|█████▌    | 5944/10682 [49:12<39:07,  2.02it/s]
 56%|█████▌    | 5945/10682 [49:12<39:10,  2.02it/s]
 56%|█████▌    | 5946/10682 [49:13<39:08,  2.02it/s]
 56%|█████▌    | 5947/10682 [49:13<39:08,  2.02it/s]
 56%|█████▌    | 5948/10682 [49:14<39:05,  2.02it/s]
 56%|█████▌    | 5949/10682 [49:14<39:06,  2.02it/s]
 56%|█████▌    | 5950/10682 [49:15<39:05,  2.02it/s]{'loss': 3.5197, 'grad_norm': 0.2160656452178955, 'learning_rate': 0.0004878276536926525, 'epoch': 0.56}
                                                    
-
 56%|█████▌    | 5950/10682 [49:15<39:05,  2.02it/s]
 56%|█████▌    | 5951/10682 [49:15<39:10,  2.01it/s]
 56%|█████▌    | 5952/10682 [49:16<39:08,  2.01it/s]
 56%|█████▌    | 5953/10682 [49:16<39:06,  2.02it/s]
 56%|█████▌    | 5954/10682 [49:17<39:05,  2.02it/s]
 56%|█████▌    | 5955/10682 [49:17<39:05,  2.02it/s]
 56%|█████▌    | 5956/10682 [49:18<39:02,  2.02it/s]
 56%|█████▌    | 5957/10682 [49:18<39:03,  2.02it/s]
 56%|█████▌    | 5958/10682 [49:19<39:01,  2.02it/s]
 56%|█████▌    | 5959/10682 [49:19<38:58,  2.02it/s]
 56%|█████▌    | 5960/10682 [49:20<38:55,  2.02it/s]
 56%|█████▌    | 5961/10682 [49:20<38:57,  2.02it/s]
 56%|█████▌    | 5962/10682 [49:21<38:55,  2.02it/s]
 56%|█████▌    | 5963/10682 [49:21<38:59,  2.02it/s]
 56%|█████▌    | 5964/10682 [49:22<39:04,  2.01it/s]
 56%|█████▌    | 5965/10682 [49:22<39:02,  2.01it/s]
 56%|█████▌    | 5966/10682 [49:23<39:01,  2.01it/s]
 56%|█████▌    | 5967/10682 [49:23<38:59,  2.02it/s]
 56%|█████▌    | 5968/10682 [49:23<38:58,  2.02it/s]
 56%|█████▌    | 5969/10682 [49:24<38:55,  2.02it/s]
 56%|█████▌    | 5970/10682 [49:24<38:56,  2.02it/s]
 56%|█████▌    | 5971/10682 [49:25<38:55,  2.02it/s]
 56%|█████▌    | 5972/10682 [49:25<38:59,  2.01it/s]
 56%|█████▌    | 5973/10682 [49:26<38:59,  2.01it/s]
 56%|█████▌    | 5974/10682 [49:26<38:57,  2.01it/s]
 56%|█████▌    | 5975/10682 [49:27<38:57,  2.01it/s]
                                                    
{'loss': 3.5127, 'grad_norm': 0.2165766805410385, 'learning_rate': 0.0004837442325565422, 'epoch': 0.56}
-
 56%|█████▌    | 5975/10682 [49:27<38:57,  2.01it/s]
 56%|█████▌    | 5976/10682 [49:27<39:01,  2.01it/s]
 56%|█████▌    | 5977/10682 [49:28<38:58,  2.01it/s]
 56%|█████▌    | 5978/10682 [49:28<38:56,  2.01it/s]
 56%|█████▌    | 5979/10682 [49:29<38:54,  2.01it/s]
 56%|█████▌    | 5980/10682 [49:29<38:52,  2.02it/s]
 56%|█████▌    | 5981/10682 [49:30<38:52,  2.02it/s]
 56%|█████▌    | 5982/10682 [49:30<38:49,  2.02it/s]
 56%|█████▌    | 5983/10682 [49:31<38:52,  2.01it/s]
 56%|█████▌    | 5984/10682 [49:31<38:51,  2.02it/s]
 56%|█████▌    | 5985/10682 [49:32<38:55,  2.01it/s]
 56%|█████▌    | 5986/10682 [49:32<38:54,  2.01it/s]
 56%|█████▌    | 5987/10682 [49:33<38:53,  2.01it/s]
 56%|█████▌    | 5988/10682 [49:33<38:50,  2.01it/s]
 56%|█████▌    | 5989/10682 [49:34<38:47,  2.02it/s]
 56%|█████▌    | 5990/10682 [49:34<38:46,  2.02it/s]
 56%|█████▌    | 5991/10682 [49:35<38:45,  2.02it/s]
 56%|█████▌    | 5992/10682 [49:35<38:45,  2.02it/s]
 56%|█████▌    | 5993/10682 [49:36<38:44,  2.02it/s]
 56%|█████▌    | 5994/10682 [49:36<38:45,  2.02it/s]
 56%|█████▌    | 5995/10682 [49:37<38:46,  2.01it/s]
 56%|█████▌    | 5996/10682 [49:37<38:46,  2.01it/s]
 56%|█████▌    | 5997/10682 [49:38<38:47,  2.01it/s]
 56%|█████▌    | 5998/10682 [49:38<38:46,  2.01it/s]
 56%|█████▌    | 5999/10682 [49:39<38:44,  2.01it/s]
 56%|█████▌    | 6000/10682 [49:39<38:41,  2.02it/s]{'loss': 3.5187, 'grad_norm': 0.2178547978401184, 'learning_rate': 0.00047966189651337725, 'epoch': 0.56}
                                                    
-
 56%|█████▌    | 6000/10682 [49:39<38:41,  2.02it/s]
 56%|█████▌    | 6001/10682 [49:40<38:41,  2.02it/s]
 56%|█████▌    | 6002/10682 [49:40<38:39,  2.02it/s]
 56%|█████▌    | 6003/10682 [49:41<38:39,  2.02it/s]
 56%|█████▌    | 6004/10682 [49:41<38:40,  2.02it/s]
 56%|█████▌    | 6005/10682 [49:42<38:42,  2.01it/s]
 56%|█████▌    | 6006/10682 [49:42<38:42,  2.01it/s]
 56%|█████▌    | 6007/10682 [49:43<38:39,  2.02it/s]
 56%|█████▌    | 6008/10682 [49:43<38:39,  2.02it/s]
 56%|█████▋    | 6009/10682 [49:44<38:37,  2.02it/s]
 56%|█████▋    | 6010/10682 [49:44<38:36,  2.02it/s]
 56%|█████▋    | 6011/10682 [49:45<38:33,  2.02it/s]
 56%|█████▋    | 6012/10682 [49:45<38:36,  2.02it/s]
 56%|█████▋    | 6013/10682 [49:46<38:35,  2.02it/s]
 56%|█████▋    | 6014/10682 [49:46<38:34,  2.02it/s]
 56%|█████▋    | 6015/10682 [49:47<38:35,  2.02it/s]
 56%|█████▋    | 6016/10682 [49:47<38:30,  2.02it/s]
 56%|█████▋    | 6017/10682 [49:48<38:32,  2.02it/s]
 56%|█████▋    | 6018/10682 [49:48<38:28,  2.02it/s]
 56%|█████▋    | 6019/10682 [49:49<38:29,  2.02it/s]
 56%|█████▋    | 6020/10682 [49:49<38:27,  2.02it/s]
 56%|█████▋    | 6021/10682 [49:50<38:25,  2.02it/s]
 56%|█████▋    | 6022/10682 [49:50<38:29,  2.02it/s]
 56%|█████▋    | 6023/10682 [49:51<38:34,  2.01it/s]
 56%|█████▋    | 6024/10682 [49:51<38:34,  2.01it/s]
 56%|█████▋    | 6025/10682 [49:52<38:32,  2.01it/s]{'loss': 3.5121, 'grad_norm': 0.21396030485630035, 'learning_rate': 0.0004755809180642286, 'epoch': 0.56}
                                                    
-
 56%|█████▋    | 6025/10682 [49:52<38:32,  2.01it/s]
 56%|█████▋    | 6026/10682 [49:52<38:36,  2.01it/s]
 56%|█████▋    | 6027/10682 [49:53<38:32,  2.01it/s]
 56%|█████▋    | 6028/10682 [49:53<38:32,  2.01it/s]
 56%|█████▋    | 6029/10682 [49:54<38:28,  2.02it/s]
 56%|█████▋    | 6030/10682 [49:54<38:28,  2.01it/s]
 56%|█████▋    | 6031/10682 [49:55<38:27,  2.02it/s]
 56%|█████▋    | 6032/10682 [49:55<38:26,  2.02it/s]
 56%|█████▋    | 6033/10682 [49:56<38:24,  2.02it/s]
 56%|█████▋    | 6034/10682 [49:56<38:28,  2.01it/s]
 56%|█████▋    | 6035/10682 [49:57<38:23,  2.02it/s]
 57%|█████▋    | 6036/10682 [49:57<38:23,  2.02it/s]
 57%|█████▋    | 6037/10682 [49:58<38:20,  2.02it/s]
 57%|█████▋    | 6038/10682 [49:58<38:24,  2.02it/s]
 57%|█████▋    | 6039/10682 [49:59<38:22,  2.02it/s]
 57%|█████▋    | 6040/10682 [49:59<38:22,  2.02it/s]
 57%|█████▋    | 6041/10682 [50:00<38:24,  2.01it/s]
 57%|█████▋    | 6042/10682 [50:00<38:20,  2.02it/s]
 57%|█████▋    | 6043/10682 [50:01<44:43,  1.73it/s]
 57%|█████▋    | 6044/10682 [50:01<42:48,  1.81it/s]
 57%|█████▋    | 6045/10682 [50:02<41:26,  1.87it/s]
 57%|█████▋    | 6046/10682 [50:02<40:31,  1.91it/s]
 57%|█████▋    | 6047/10682 [50:03<39:52,  1.94it/s]
 57%|█████▋    | 6048/10682 [50:03<39:22,  1.96it/s]
 57%|█████▋    | 6049/10682 [50:04<39:02,  1.98it/s]
 57%|█████▋    | 6050/10682 [50:04<38:49,  1.99it/s]
                                                    
{'loss': 3.5031, 'grad_norm': 0.21698886156082153, 'learning_rate': 0.00047150156961954607, 'epoch': 0.57}
-
 57%|█████▋    | 6050/10682 [50:04<38:49,  1.99it/s]
 57%|█████▋    | 6051/10682 [50:05<38:44,  1.99it/s]
 57%|█████▋    | 6052/10682 [50:05<38:36,  2.00it/s]
 57%|█████▋    | 6053/10682 [50:06<38:27,  2.01it/s]
 57%|█████▋    | 6054/10682 [50:06<38:24,  2.01it/s]
 57%|█████▋    | 6055/10682 [50:07<38:22,  2.01it/s]
 57%|█████▋    | 6056/10682 [50:07<38:19,  2.01it/s]
 57%|█████▋    | 6057/10682 [50:08<38:17,  2.01it/s]
 57%|█████▋    | 6058/10682 [50:08<38:15,  2.01it/s]
 57%|█████▋    | 6059/10682 [50:09<38:10,  2.02it/s]
 57%|█████▋    | 6060/10682 [50:09<38:11,  2.02it/s]
 57%|█████▋    | 6061/10682 [50:10<38:08,  2.02it/s]
 57%|█████▋    | 6062/10682 [50:10<38:12,  2.02it/s]
 57%|█████▋    | 6063/10682 [50:11<38:09,  2.02it/s]
 57%|█████▋    | 6064/10682 [50:11<38:11,  2.02it/s]
 57%|█████▋    | 6065/10682 [50:12<38:06,  2.02it/s]
 57%|█████▋    | 6066/10682 [50:12<38:06,  2.02it/s]
 57%|█████▋    | 6067/10682 [50:13<38:04,  2.02it/s]
 57%|█████▋    | 6068/10682 [50:14<44:14,  1.74it/s]
 57%|█████▋    | 6069/10682 [50:14<42:24,  1.81it/s]
 57%|█████▋    | 6070/10682 [50:15<41:03,  1.87it/s]
 57%|█████▋    | 6071/10682 [50:15<40:11,  1.91it/s]
 57%|█████▋    | 6072/10682 [50:16<39:32,  1.94it/s]
 57%|█████▋    | 6073/10682 [50:16<39:09,  1.96it/s]
 57%|█████▋    | 6074/10682 [50:17<38:47,  1.98it/s]
 57%|█████▋    | 6075/10682 [50:17<38:39,  1.99it/s]{'loss': 3.5177, 'grad_norm': 0.23647558689117432, 'learning_rate': 0.0004674241234809751, 'epoch': 0.57}
                                                    
-
 57%|█████▋    | 6075/10682 [50:17<38:39,  1.99it/s]
 57%|█████▋    | 6076/10682 [50:18<38:32,  1.99it/s]
 57%|█████▋    | 6077/10682 [50:18<38:22,  2.00it/s]
 57%|█████▋    | 6078/10682 [50:19<38:14,  2.01it/s]
 57%|█████▋    | 6079/10682 [50:19<38:10,  2.01it/s]
 57%|█████▋    | 6080/10682 [50:20<38:04,  2.01it/s]
 57%|█████▋    | 6081/10682 [50:20<38:06,  2.01it/s]
 57%|█████▋    | 6082/10682 [50:21<38:04,  2.01it/s]
 57%|█████▋    | 6083/10682 [50:21<38:06,  2.01it/s]
 57%|█████▋    | 6084/10682 [50:22<38:00,  2.02it/s]
 57%|█████▋    | 6085/10682 [50:22<38:00,  2.02it/s]
 57%|█████▋    | 6086/10682 [50:23<37:57,  2.02it/s]
 57%|█████▋    | 6087/10682 [50:23<37:57,  2.02it/s]
 57%|█████▋    | 6088/10682 [50:24<37:55,  2.02it/s]
 57%|█████▋    | 6089/10682 [50:24<37:56,  2.02it/s]
 57%|█████▋    | 6090/10682 [50:25<37:54,  2.02it/s]
 57%|█████▋    | 6091/10682 [50:25<37:54,  2.02it/s]
 57%|█████▋    | 6092/10682 [50:26<37:55,  2.02it/s]
 57%|█████▋    | 6093/10682 [50:26<37:53,  2.02it/s]
 57%|█████▋    | 6094/10682 [50:27<37:52,  2.02it/s]
 57%|█████▋    | 6095/10682 [50:27<37:52,  2.02it/s]
 57%|█████▋    | 6096/10682 [50:28<37:50,  2.02it/s]
 57%|█████▋    | 6097/10682 [50:28<37:48,  2.02it/s]
 57%|█████▋    | 6098/10682 [50:29<37:48,  2.02it/s]
 57%|█████▋    | 6099/10682 [50:29<37:50,  2.02it/s]
 57%|█████▋    | 6100/10682 [50:30<37:48,  2.02it/s]{'loss': 3.5061, 'grad_norm': 0.21883301436901093, 'learning_rate': 0.0004633488518231793, 'epoch': 0.57}
                                                    
-
 57%|█████▋    | 6100/10682 [50:30<37:48,  2.02it/s]
 57%|█████▋    | 6101/10682 [50:30<37:54,  2.01it/s]
 57%|█████▋    | 6102/10682 [50:31<37:55,  2.01it/s]
 57%|█████▋    | 6103/10682 [50:31<37:53,  2.01it/s]
 57%|█████▋    | 6104/10682 [50:32<37:57,  2.01it/s]
 57%|█████▋    | 6105/10682 [50:32<37:53,  2.01it/s]
 57%|█████▋    | 6106/10682 [50:32<37:51,  2.01it/s]
 57%|█████▋    | 6107/10682 [50:33<37:47,  2.02it/s]
 57%|█████▋    | 6108/10682 [50:33<37:46,  2.02it/s]
 57%|█████▋    | 6109/10682 [50:34<37:46,  2.02it/s]
 57%|█████▋    | 6110/10682 [50:34<37:48,  2.02it/s]
 57%|█████▋    | 6111/10682 [50:35<37:48,  2.02it/s]
 57%|█████▋    | 6112/10682 [50:35<37:44,  2.02it/s]
 57%|█████▋    | 6113/10682 [50:36<37:43,  2.02it/s]
 57%|█████▋    | 6114/10682 [50:36<37:40,  2.02it/s]
 57%|█████▋    | 6115/10682 [50:37<37:40,  2.02it/s]
 57%|█████▋    | 6116/10682 [50:37<37:37,  2.02it/s]
 57%|█████▋    | 6117/10682 [50:38<37:36,  2.02it/s]
 57%|█████▋    | 6118/10682 [50:38<37:35,  2.02it/s]
 57%|█████▋    | 6119/10682 [50:39<37:36,  2.02it/s]
 57%|█████▋    | 6120/10682 [50:39<37:36,  2.02it/s]
 57%|█████▋    | 6121/10682 [50:40<37:36,  2.02it/s]
 57%|█████▋    | 6122/10682 [50:40<37:35,  2.02it/s]
 57%|█████▋    | 6123/10682 [50:41<37:34,  2.02it/s]
 57%|█████▋    | 6124/10682 [50:41<37:34,  2.02it/s]
 57%|█████▋    | 6125/10682 [50:42<37:36,  2.02it/s]{'loss': 3.5185, 'grad_norm': 0.21809537708759308, 'learning_rate': 0.00045927602667567333, 'epoch': 0.57}
                                                    
-
 57%|█████▋    | 6125/10682 [50:42<37:36,  2.02it/s]
 57%|█████▋    | 6126/10682 [50:42<37:37,  2.02it/s]
 57%|█████▋    | 6127/10682 [50:43<37:40,  2.01it/s]
 57%|█████▋    | 6128/10682 [50:43<37:39,  2.02it/s]
 57%|█████▋    | 6129/10682 [50:44<37:40,  2.01it/s]
 57%|█████▋    | 6130/10682 [50:44<37:41,  2.01it/s]
 57%|█████▋    | 6131/10682 [50:45<37:37,  2.02it/s]
 57%|█████▋    | 6132/10682 [50:45<37:40,  2.01it/s]
 57%|█████▋    | 6133/10682 [50:46<37:40,  2.01it/s]
 57%|█████▋    | 6134/10682 [50:46<37:38,  2.01it/s]
 57%|█████▋    | 6135/10682 [50:47<37:34,  2.02it/s]
 57%|█████▋    | 6136/10682 [50:47<37:32,  2.02it/s]
 57%|█████▋    | 6137/10682 [50:48<37:30,  2.02it/s]
 57%|█████▋    | 6138/10682 [50:48<37:30,  2.02it/s]
 57%|█████▋    | 6139/10682 [50:49<37:29,  2.02it/s]
 57%|█████▋    | 6140/10682 [50:49<37:32,  2.02it/s]
 57%|█████▋    | 6141/10682 [50:50<37:32,  2.02it/s]
 57%|█████▋    | 6142/10682 [50:50<37:31,  2.02it/s]
 58%|█████▊    | 6143/10682 [50:51<37:36,  2.01it/s]
 58%|█████▊    | 6144/10682 [50:51<37:32,  2.01it/s]
 58%|█████▊    | 6145/10682 [50:52<37:30,  2.02it/s]
 58%|█████▊    | 6146/10682 [50:52<37:27,  2.02it/s]
 58%|█████▊    | 6147/10682 [50:53<37:26,  2.02it/s]
 58%|█████▊    | 6148/10682 [50:53<37:25,  2.02it/s]
 58%|█████▊    | 6149/10682 [50:54<37:27,  2.02it/s]
 58%|█████▊    | 6150/10682 [50:54<37:22,  2.02it/s]
                                                    {'loss': 3.5114, 'grad_norm': 0.22755597531795502, 'learning_rate': 0.00045520591990466365, 'epoch': 0.58}
-
 58%|█████▊    | 6150/10682 [50:54<37:22,  2.02it/s]
 58%|█████▊    | 6151/10682 [50:55<37:30,  2.01it/s]
 58%|█████▊    | 6152/10682 [50:55<37:29,  2.01it/s]
 58%|█████▊    | 6153/10682 [50:56<37:29,  2.01it/s]
 58%|█████▊    | 6154/10682 [50:56<37:30,  2.01it/s]
 58%|█████▊    | 6155/10682 [50:57<37:26,  2.01it/s]
 58%|█████▊    | 6156/10682 [50:57<37:24,  2.02it/s]
 58%|█████▊    | 6157/10682 [50:58<37:22,  2.02it/s]
 58%|█████▊    | 6158/10682 [50:58<37:22,  2.02it/s]
 58%|█████▊    | 6159/10682 [50:59<37:20,  2.02it/s]
 58%|█████▊    | 6160/10682 [50:59<37:21,  2.02it/s]
 58%|█████▊    | 6161/10682 [51:00<37:20,  2.02it/s]
 58%|█████▊    | 6162/10682 [51:00<37:22,  2.02it/s]
 58%|█████▊    | 6163/10682 [51:01<37:20,  2.02it/s]
 58%|█████▊    | 6164/10682 [51:01<37:20,  2.02it/s]
 58%|█████▊    | 6165/10682 [51:02<37:17,  2.02it/s]
 58%|█████▊    | 6166/10682 [51:02<37:18,  2.02it/s]
 58%|█████▊    | 6167/10682 [51:03<37:16,  2.02it/s]
 58%|█████▊    | 6168/10682 [51:03<37:14,  2.02it/s]
 58%|█████▊    | 6169/10682 [51:04<37:17,  2.02it/s]
 58%|█████▊    | 6170/10682 [51:04<37:16,  2.02it/s]
 58%|█████▊    | 6171/10682 [51:05<37:14,  2.02it/s]
 58%|█████▊    | 6172/10682 [51:05<37:13,  2.02it/s]
 58%|█████▊    | 6173/10682 [51:06<37:11,  2.02it/s]
 58%|█████▊    | 6174/10682 [51:06<37:13,  2.02it/s]
 58%|█████▊    | 6175/10682 [51:07<37:11,  2.02it/s]
                                                    {'loss': 3.4999, 'grad_norm': 0.22326922416687012, 'learning_rate': 0.00045113880319490264, 'epoch': 0.58}
-
 58%|█████▊    | 6175/10682 [51:07<37:11,  2.02it/s]
 58%|█████▊    | 6176/10682 [51:07<37:12,  2.02it/s]
 58%|█████▊    | 6177/10682 [51:08<37:11,  2.02it/s]
 58%|█████▊    | 6178/10682 [51:08<37:09,  2.02it/s]
 58%|█████▊    | 6179/10682 [51:09<37:09,  2.02it/s]
 58%|█████▊    | 6180/10682 [51:09<37:12,  2.02it/s]
 58%|█████▊    | 6181/10682 [51:10<37:11,  2.02it/s]
 58%|█████▊    | 6182/10682 [51:10<37:12,  2.02it/s]
 58%|█████▊    | 6183/10682 [51:11<37:10,  2.02it/s]
 58%|█████▊    | 6184/10682 [51:11<37:08,  2.02it/s]
 58%|█████▊    | 6185/10682 [51:12<37:09,  2.02it/s]
 58%|█████▊    | 6186/10682 [51:12<37:07,  2.02it/s]
 58%|█████▊    | 6187/10682 [51:13<37:09,  2.02it/s]
 58%|█████▊    | 6188/10682 [51:13<37:05,  2.02it/s]
 58%|█████▊    | 6189/10682 [51:14<37:07,  2.02it/s]
 58%|█████▊    | 6190/10682 [51:14<37:05,  2.02it/s]
 58%|█████▊    | 6191/10682 [51:15<37:07,  2.02it/s]
 58%|█████▊    | 6192/10682 [51:15<37:04,  2.02it/s]
 58%|█████▊    | 6193/10682 [51:16<37:03,  2.02it/s]
 58%|█████▊    | 6194/10682 [51:16<37:00,  2.02it/s]
 58%|█████▊    | 6195/10682 [51:17<37:03,  2.02it/s]
 58%|█████▊    | 6196/10682 [51:17<37:01,  2.02it/s]
 58%|█████▊    | 6197/10682 [51:18<37:04,  2.02it/s]
 58%|█████▊    | 6198/10682 [51:18<37:05,  2.01it/s]
 58%|█████▊    | 6199/10682 [51:19<37:02,  2.02it/s]
 58%|█████▊    | 6200/10682 [51:19<37:04,  2.01it/s]
                                                    
{'loss': 3.5091, 'grad_norm': 0.21641342341899872, 'learning_rate': 0.00044707494803155137, 'epoch': 0.58}
-
 58%|█████▊    | 6200/10682 [51:19<37:04,  2.01it/s]
 58%|█████▊    | 6201/10682 [51:20<37:05,  2.01it/s]
 58%|█████▊    | 6202/10682 [51:20<37:02,  2.02it/s]
 58%|█████▊    | 6203/10682 [51:21<36:59,  2.02it/s]
 58%|█████▊    | 6204/10682 [51:21<37:00,  2.02it/s]
 58%|█████▊    | 6205/10682 [51:22<36:59,  2.02it/s]
 58%|█████▊    | 6206/10682 [51:22<37:00,  2.02it/s]
 58%|█████▊    | 6207/10682 [51:23<36:58,  2.02it/s]
 58%|█████▊    | 6208/10682 [51:23<37:00,  2.02it/s]
 58%|█████▊    | 6209/10682 [51:24<36:58,  2.02it/s]
 58%|████���▊    | 6210/10682 [51:24<37:02,  2.01it/s]
 58%|█████▊    | 6211/10682 [51:25<36:57,  2.02it/s]
 58%|█████▊    | 6212/10682 [51:25<36:55,  2.02it/s]
 58%|█████▊    | 6213/10682 [51:26<36:53,  2.02it/s]
 58%|█████▊    | 6214/10682 [51:26<36:52,  2.02it/s]
 58%|█████▊    | 6215/10682 [51:27<36:50,  2.02it/s]
 58%|█████▊    | 6216/10682 [51:27<36:50,  2.02it/s]
 58%|█████▊    | 6217/10682 [51:28<36:47,  2.02it/s]
 58%|█████▊    | 6218/10682 [51:28<36:52,  2.02it/s]
 58%|█████▊    | 6219/10682 [51:28<36:50,  2.02it/s]
 58%|█████▊    | 6220/10682 [51:29<36:54,  2.02it/s]
 58%|█████▊    | 6221/10682 [51:29<36:50,  2.02it/s]
 58%|█████▊    | 6222/10682 [51:30<36:50,  2.02it/s]
 58%|█████▊    | 6223/10682 [51:30<36:45,  2.02it/s]
 58%|█████▊    | 6224/10682 [51:31<36:44,  2.02it/s]
 58%|█████▊    | 6225/10682 [51:31<36:46,  2.02it/s]{'loss': 3.506, 'grad_norm': 0.21803636848926544, 'learning_rate': 0.00044301462568205915, 'epoch': 0.58}
                                                    
-
 58%|█████▊    | 6225/10682 [51:31<36:46,  2.02it/s]
 58%|█████▊    | 6226/10682 [51:32<36:48,  2.02it/s]
 58%|█████▊    | 6227/10682 [51:32<36:46,  2.02it/s]
 58%|█████▊    | 6228/10682 [51:33<36:46,  2.02it/s]
 58%|█████▊    | 6229/10682 [51:33<36:44,  2.02it/s]
 58%|█████▊    | 6230/10682 [51:34<36:45,  2.02it/s]
 58%|█████▊    | 6231/10682 [51:34<36:44,  2.02it/s]
 58%|█████▊    | 6232/10682 [51:35<36:47,  2.02it/s]
 58%|█████▊    | 6233/10682 [51:35<36:42,  2.02it/s]
 58%|█████▊    | 6234/10682 [51:36<36:41,  2.02it/s]
 58%|█████▊    | 6235/10682 [51:36<36:40,  2.02it/s]
 58%|█████▊    | 6236/10682 [51:37<36:41,  2.02it/s]
 58%|█████▊    | 6237/10682 [51:37<36:39,  2.02it/s]
 58%|█████▊    | 6238/10682 [51:38<36:39,  2.02it/s]
 58%|█████▊    | 6239/10682 [51:38<36:37,  2.02it/s]
 58%|█████▊    | 6240/10682 [51:39<36:41,  2.02it/s]
 58%|█████▊    | 6241/10682 [51:39<36:36,  2.02it/s]
 58%|█████▊    | 6242/10682 [51:40<36:38,  2.02it/s]
 58%|█████▊    | 6243/10682 [51:40<36:36,  2.02it/s]
 58%|█████▊    | 6244/10682 [51:41<36:35,  2.02it/s]
 58%|█████▊    | 6245/10682 [51:41<36:31,  2.02it/s]
 58%|█████▊    | 6246/10682 [51:42<36:32,  2.02it/s]
 58%|█████▊    | 6247/10682 [51:42<36:32,  2.02it/s]
 58%|█████▊    | 6248/10682 [51:43<36:37,  2.02it/s]
 59%|█████▊    | 6249/10682 [51:43<36:35,  2.02it/s]
 59%|█████▊    | 6250/10682 [51:44<36:35,  2.02it/s]
                                                    {'loss': 3.5012, 'grad_norm': 0.2222476303577423, 'learning_rate': 0.0004389581071780553, 'epoch': 0.59}
-
 59%|█████▊    | 6250/10682 [51:44<36:35,  2.02it/s]
 59%|█████▊    | 6251/10682 [51:44<36:37,  2.02it/s]
 59%|█████▊    | 6252/10682 [51:45<36:37,  2.02it/s]
 59%|█████▊    | 6253/10682 [51:45<36:39,  2.01it/s]
 59%|█████▊    | 6254/10682 [51:46<36:35,  2.02it/s]
 59%|█████▊    | 6255/10682 [51:46<36:35,  2.02it/s]
 59%|█████▊    | 6256/10682 [51:47<36:31,  2.02it/s]
 59%|█████▊    | 6257/10682 [51:47<36:34,  2.02it/s]
 59%|█████▊    | 6258/10682 [51:48<36:29,  2.02it/s]
 59%|█████▊    | 6259/10682 [51:48<36:35,  2.01it/s]
 59%|█████▊    | 6260/10682 [51:49<36:33,  2.02it/s]
 59%|█████▊    | 6261/10682 [51:49<36:33,  2.02it/s]
 59%|█████▊    | 6262/10682 [51:50<36:32,  2.02it/s]
 59%|█████▊    | 6263/10682 [51:50<36:31,  2.02it/s]
 59%|█████▊    | 6264/10682 [51:51<36:31,  2.02it/s]
 59%|█████▊    | 6265/10682 [51:51<36:30,  2.02it/s]
 59%|█████▊    | 6266/10682 [51:52<36:30,  2.02it/s]
 59%|█████▊    | 6267/10682 [51:52<36:28,  2.02it/s]
 59%|█████▊    | 6268/10682 [51:53<36:27,  2.02it/s]
 59%|█████▊    | 6269/10682 [51:53<36:24,  2.02it/s]
 59%|█████▊    | 6270/10682 [51:54<36:25,  2.02it/s]
 59%|█████▊    | 6271/10682 [51:54<36:25,  2.02it/s]
 59%|█████▊    | 6272/10682 [51:55<36:24,  2.02it/s]
 59%|█████▊    | 6273/10682 [51:55<36:21,  2.02it/s]
 59%|█████▊    | 6274/10682 [51:56<36:19,  2.02it/s]
 59%|█████▊    | 6275/10682 [51:56<36:18,  2.02it/s]
                                                    {'loss': 3.5012, 'grad_norm': 0.21941015124320984, 'learning_rate': 0.0004349056632972576, 'epoch': 0.59}
-
 59%|█████▊    | 6275/10682 [51:56<36:18,  2.02it/s]
 59%|█████▉    | 6276/10682 [51:57<36:26,  2.02it/s]
 59%|█████▉    | 6277/10682 [51:57<36:22,  2.02it/s]
 59%|█████▉    | 6278/10682 [51:58<36:23,  2.02it/s]
 59%|█████▉    | 6279/10682 [51:58<36:26,  2.01it/s]
 59%|█████▉    | 6280/10682 [51:59<36:23,  2.02it/s]
 59%|█████▉    | 6281/10682 [51:59<36:22,  2.02it/s]
 59%|█████▉    | 6282/10682 [52:00<36:22,  2.02it/s]
 59%|█████▉    | 6283/10682 [52:00<36:21,  2.02it/s]
 59%|█████▉    | 6284/10682 [52:01<36:24,  2.01it/s]
 59%|█████▉    | 6285/10682 [52:01<36:25,  2.01it/s]
 59%|█████▉    | 6286/10682 [52:02<36:26,  2.01it/s]
 59%|█████▉    | 6287/10682 [52:02<36:23,  2.01it/s]
 59%|█████▉    | 6288/10682 [52:03<36:23,  2.01it/s]
 59%|█████▉    | 6289/10682 [52:03<36:20,  2.01it/s]
 59%|█████▉    | 6290/10682 [52:04<36:18,  2.02it/s]
 59%|█████▉    | 6291/10682 [52:04<36:18,  2.02it/s]
 59%|█████▉    | 6292/10682 [52:05<36:18,  2.02it/s]
 59%|█████▉    | 6293/10682 [52:05<36:17,  2.02it/s]
 59%|█████▉    | 6294/10682 [52:06<36:13,  2.02it/s]
 59%|█████▉    | 6295/10682 [52:06<36:12,  2.02it/s]
 59%|█████▉    | 6296/10682 [52:07<36:10,  2.02it/s]
 59%|█████▉    | 6297/10682 [52:07<36:11,  2.02it/s]
 59%|█████▉    | 6298/10682 [52:08<36:13,  2.02it/s]
 59%|█████▉    | 6299/10682 [52:08<36:14,  2.02it/s]
 59%|█████▉    | 6300/10682 [52:09<36:10,  2.02it/s]
                                                    {'loss': 3.5085, 'grad_norm': 0.22536611557006836, 'learning_rate': 0.00043085756454539745, 'epoch': 0.59}
-
 59%|█████▉    | 6300/10682 [52:09<36:10,  2.02it/s]
 59%|█████▉    | 6301/10682 [52:09<36:14,  2.02it/s]
 59%|█████▉    | 6302/10682 [52:10<36:10,  2.02it/s]
 59%|█████▉    | 6303/10682 [52:10<36:09,  2.02it/s]
 59%|█████▉    | 6304/10682 [52:11<36:04,  2.02it/s]
 59%|█████▉    | 6305/10682 [52:11<36:07,  2.02it/s]
 59%|█████▉    | 6306/10682 [52:12<36:04,  2.02it/s]
 59%|█████▉    | 6307/10682 [52:12<36:04,  2.02it/s]
 59%|█████▉    | 6308/10682 [52:13<36:04,  2.02it/s]
 59%|█████▉    | 6309/10682 [52:13<36:06,  2.02it/s]
 59%|█████▉    | 6310/10682 [52:14<36:03,  2.02it/s]
 59%|█████▉    | 6311/10682 [52:14<36:04,  2.02it/s]
 59%|█████▉    | 6312/10682 [52:15<36:05,  2.02it/s]
 59%|█████▉    | 6313/10682 [52:15<36:04,  2.02it/s]
 59%|█████▉    | 6314/10682 [52:16<36:01,  2.02it/s]
 59%|█████▉    | 6315/10682 [52:16<36:01,  2.02it/s]
 59%|█████▉    | 6316/10682 [52:17<36:00,  2.02it/s]
 59%|█████▉    | 6317/10682 [52:17<36:02,  2.02it/s]
 59%|█████▉    | 6318/10682 [52:18<36:00,  2.02it/s]
 59%|█████▉    | 6319/10682 [52:18<36:00,  2.02it/s]
 59%|█████▉    | 6320/10682 [52:19<35:57,  2.02it/s]
 59%|█████▉    | 6321/10682 [52:19<35:58,  2.02it/s]
 59%|█████▉    | 6322/10682 [52:20<35:54,  2.02it/s]
 59%|█████▉    | 6323/10682 [52:20<35:56,  2.02it/s]
 59%|█████▉    | 6324/10682 [52:21<35:52,  2.02it/s]
 59%|█████▉    | 6325/10682 [52:21<35:58,  2.02it/s]
                                                    {'loss': 3.507, 'grad_norm': 0.23150929808616638, 'learning_rate': 0.0004268140811381641, 'epoch': 0.59}
-
 59%|█████▉    | 6325/10682 [52:21<35:58,  2.02it/s]
 59%|█████▉    | 6326/10682 [52:22<36:05,  2.01it/s]
 59%|█████▉    | 6327/10682 [52:22<35:59,  2.02it/s]
 59%|█████▉    | 6328/10682 [52:23<36:00,  2.01it/s]
 59%|█████▉    | 6329/10682 [52:23<35:56,  2.02it/s]
 59%|█████▉    | 6330/10682 [52:23<35:57,  2.02it/s]
 59%|█████▉    | 6331/10682 [52:24<35:53,  2.02it/s]
 59%|█████▉    | 6332/10682 [52:24<35:53,  2.02it/s]
 59%|█████▉    | 6333/10682 [52:25<35:50,  2.02it/s]
 59%|█████▉    | 6334/10682 [52:25<35:52,  2.02it/s]
 59%|█████▉    | 6335/10682 [52:26<35:52,  2.02it/s]
 59%|█████▉    | 6336/10682 [52:26<35:54,  2.02it/s]
 59%|█████▉    | 6337/10682 [52:27<35:52,  2.02it/s]
 59%|█████▉    | 6338/10682 [52:27<35:53,  2.02it/s]
 59%|█████▉    | 6339/10682 [52:28<35:52,  2.02it/s]
 59%|█████▉    | 6340/10682 [52:28<35:53,  2.02it/s]
 59%|█████▉    | 6341/10682 [52:29<35:51,  2.02it/s]
 59%|█████▉    | 6342/10682 [52:29<35:48,  2.02it/s]
 59%|█████▉    | 6343/10682 [52:30<35:48,  2.02it/s]
 59%|█████▉    | 6344/10682 [52:30<35:47,  2.02it/s]
 59%|█████▉    | 6345/10682 [52:31<35:47,  2.02it/s]
 59%|█████▉    | 6346/10682 [52:31<35:48,  2.02it/s]
 59%|█████▉    | 6347/10682 [52:32<35:46,  2.02it/s]
 59%|█████▉    | 6348/10682 [52:32<35:49,  2.02it/s]
 59%|█████▉    | 6349/10682 [52:33<35:45,  2.02it/s]
 59%|█████▉    | 6350/10682 [52:33<35:48,  2.02it/s]
                                                    
{'loss': 3.4962, 'grad_norm': 0.21703720092773438, 'learning_rate': 0.0004227754829831662, 'epoch': 0.59}
-
 59%|█████▉    | 6350/10682 [52:33<35:48,  2.02it/s]
 59%|█████▉    | 6351/10682 [52:34<35:47,  2.02it/s]
 59%|█████▉    | 6352/10682 [52:34<35:46,  2.02it/s]
 59%|█████▉    | 6353/10682 [52:35<35:42,  2.02it/s]
 59%|█████▉    | 6354/10682 [52:35<35:43,  2.02it/s]
 59%|█████▉    | 6355/10682 [52:36<35:43,  2.02it/s]
 60%|█████▉    | 6356/10682 [52:36<35:42,  2.02it/s]
 60%|█████▉    | 6357/10682 [52:37<35:44,  2.02it/s]
 60%|█████▉    | 6358/10682 [52:37<35:40,  2.02it/s]
 60%|█████▉    | 6359/10682 [52:38<35:43,  2.02it/s]
 60%|█████▉    | 6360/10682 [52:38<35:39,  2.02it/s]
 60%|█████▉    | 6361/10682 [52:39<35:40,  2.02it/s]
 60%|█████▉    | 6362/10682 [52:39<35:37,  2.02it/s]
 60%|█████▉    | 6363/10682 [52:40<35:40,  2.02it/s]
 60%|█████▉    | 6364/10682 [52:40<35:37,  2.02it/s]
 60%|█████▉    | 6365/10682 [52:41<35:38,  2.02it/s]
 60%|█████▉    | 6366/10682 [52:41<35:37,  2.02it/s]
 60%|█████▉    | 6367/10682 [52:42<35:38,  2.02it/s]
 60%|█████▉    | 6368/10682 [52:42<35:34,  2.02it/s]
 60%|█████▉    | 6369/10682 [52:43<35:37,  2.02it/s]
 60%|█████▉    | 6370/10682 [52:43<35:36,  2.02it/s]
 60%|█████▉    | 6371/10682 [52:44<35:38,  2.02it/s]
 60%|█████▉    | 6372/10682 [52:44<35:36,  2.02it/s]
 60%|█████▉    | 6373/10682 [52:45<35:37,  2.02it/s]
 60%|█████▉    | 6374/10682 [52:45<35:43,  2.01it/s]
 60%|█████▉    | 6375/10682 [52:46<35:39,  2.01it/s]{'loss': 3.491, 'grad_norm': 0.22236917912960052, 'learning_rate': 0.0004187420396619156, 'epoch': 0.6}
                                                    
-
 60%|█████▉    | 6375/10682 [52:46<35:39,  2.01it/s]
 60%|█████▉    | 6376/10682 [52:46<35:42,  2.01it/s]
 60%|█████▉    | 6377/10682 [52:47<35:42,  2.01it/s]
 60%|█████▉    | 6378/10682 [52:47<35:37,  2.01it/s]
 60%|█████▉    | 6379/10682 [52:48<35:37,  2.01it/s]
 60%|█████▉    | 6380/10682 [52:48<35:32,  2.02it/s]
 60%|█████▉    | 6381/10682 [52:49<35:32,  2.02it/s]
 60%|█████▉    | 6382/10682 [52:49<35:28,  2.02it/s]
 60%|█████▉    | 6383/10682 [52:50<35:28,  2.02it/s]
 60%|█████▉    | 6384/10682 [52:50<35:29,  2.02it/s]
 60%|█████▉    | 6385/10682 [52:51<35:28,  2.02it/s]
 60%|█████▉    | 6386/10682 [52:51<35:27,  2.02it/s]
 60%|█████▉    | 6387/10682 [52:52<35:29,  2.02it/s]
 60%|█████▉    | 6388/10682 [52:52<35:27,  2.02it/s]
 60%|█████▉    | 6389/10682 [52:53<35:27,  2.02it/s]
 60%|█████▉    | 6390/10682 [52:53<35:23,  2.02it/s]
 60%|█████▉    | 6391/10682 [52:54<35:24,  2.02it/s]
 60%|█████▉    | 6392/10682 [52:54<35:23,  2.02it/s]
 60%|█████▉    | 6393/10682 [52:55<35:24,  2.02it/s]
 60%|█████▉    | 6394/10682 [52:55<35:22,  2.02it/s]
 60%|█████▉    | 6395/10682 [52:56<35:22,  2.02it/s]
 60%|█████▉    | 6396/10682 [52:56<35:22,  2.02it/s]
 60%|█████▉    | 6397/10682 [52:57<35:21,  2.02it/s]
 60%|█████▉    | 6398/10682 [52:57<35:22,  2.02it/s]
 60%|█████▉    | 6399/10682 [52:58<35:22,  2.02it/s]
 60%|█████▉    | 6400/10682 [52:58<35:21,  2.02it/s]{'loss': 3.4976, 'grad_norm': 0.23872733116149902, 'learning_rate': 0.00041471402041183287, 'epoch': 0.6}
                                                    
-
 60%|█████▉    | 6400/10682 [52:58<35:21,  2.02it/s]
 60%|█████▉    | 6401/10682 [52:59<35:23,  2.02it/s]
 60%|█████▉    | 6402/10682 [52:59<35:22,  2.02it/s]
 60%|█████▉    | 6403/10682 [53:00<35:18,  2.02it/s]
 60%|█████▉    | 6404/10682 [53:00<35:18,  2.02it/s]
 60%|█████▉    | 6405/10682 [53:01<35:16,  2.02it/s]
 60%|█████▉    | 6406/10682 [53:01<35:17,  2.02it/s]
 60%|█████▉    | 6407/10682 [53:02<35:14,  2.02it/s]
 60%|█████▉    | 6408/10682 [53:02<35:16,  2.02it/s]
 60%|█████▉    | 6409/10682 [53:03<35:13,  2.02it/s]
 60%|██████    | 6410/10682 [53:03<35:14,  2.02it/s]
 60%|██████    | 6411/10682 [53:04<35:10,  2.02it/s]
 60%|██████    | 6412/10682 [53:04<35:12,  2.02it/s]
 60%|██████    | 6413/10682 [53:05<35:10,  2.02it/s]
 60%|██████    | 6414/10682 [53:05<35:12,  2.02it/s]
 60%|██████    | 6415/10682 [53:06<35:15,  2.02it/s]
 60%|██████    | 6416/10682 [53:06<35:13,  2.02it/s]
 60%|██████    | 6417/10682 [53:07<35:12,  2.02it/s]
 60%|██████    | 6418/10682 [53:07<35:12,  2.02it/s]
 60%|██████    | 6419/10682 [53:08<35:13,  2.02it/s]
 60%|██████    | 6420/10682 [53:08<35:10,  2.02it/s]
 60%|██████    | 6421/10682 [53:09<35:09,  2.02it/s]
 60%|██████    | 6422/10682 [53:09<35:08,  2.02it/s]
 60%|██████    | 6423/10682 [53:10<35:08,  2.02it/s]
 60%|██████    | 6424/10682 [53:10<35:07,  2.02it/s]
 60%|██████    | 6425/10682 [53:11<35:13,  2.01it/s]{'loss': 3.4916, 'grad_norm': 0.21252360939979553, 'learning_rate': 0.0004106916941082745, 'epoch': 0.6}
                                                    
-
 60%|██████    | 6425/10682 [53:11<35:13,  2.01it/s]
 60%|██████    | 6426/10682 [53:11<35:13,  2.01it/s]
 60%|██████    | 6427/10682 [53:12<35:11,  2.01it/s]
 60%|██████    | 6428/10682 [53:12<35:07,  2.02it/s]
 60%|██████    | 6429/10682 [53:13<35:08,  2.02it/s]
 60%|██████    | 6430/10682 [53:13<35:05,  2.02it/s]
 60%|██████    | 6431/10682 [53:14<35:05,  2.02it/s]
 60%|██████    | 6432/10682 [53:14<35:04,  2.02it/s]
 60%|██████    | 6433/10682 [53:15<35:07,  2.02it/s]
 60%|██████    | 6434/10682 [53:15<35:03,  2.02it/s]
 60%|██████    | 6435/10682 [53:16<35:05,  2.02it/s]
 60%|██████    | 6436/10682 [53:16<35:03,  2.02it/s]
 60%|██████    | 6437/10682 [53:17<35:02,  2.02it/s]
 60%|██████    | 6438/10682 [53:17<34:59,  2.02it/s]
 60%|██████    | 6439/10682 [53:17<35:00,  2.02it/s]
 60%|██████    | 6440/10682 [53:18<34:59,  2.02it/s]
 60%|██████    | 6441/10682 [53:18<35:01,  2.02it/s]
 60%|██████    | 6442/10682 [53:19<35:00,  2.02it/s]
 60%|██████    | 6443/10682 [53:19<35:00,  2.02it/s]
 60%|██████    | 6444/10682 [53:20<34:59,  2.02it/s]
 60%|██████    | 6445/10682 [53:20<34:59,  2.02it/s]
 60%|██████    | 6446/10682 [53:21<34:57,  2.02it/s]
 60%|██████    | 6447/10682 [53:21<34:58,  2.02it/s]
 60%|██████    | 6448/10682 [53:22<34:54,  2.02it/s]
 60%|██████    | 6449/10682 [53:22<34:58,  2.02it/s]
 60%|██████    | 6450/10682 [53:23<34:55,  2.02it/s]
                                                    {'loss': 3.4724, 'grad_norm': 0.21352419257164001, 'learning_rate': 0.0004066753292465861, 'epoch': 0.6}
-
 60%|██████    | 6450/10682 [53:23<34:55,  2.02it/s]
 60%|██████    | 6451/10682 [53:23<35:01,  2.01it/s]
 60%|██████    | 6452/10682 [53:24<35:01,  2.01it/s]
 60%|██████    | 6453/10682 [53:24<34:57,  2.02it/s]
 60%|██████    | 6454/10682 [53:25<35:00,  2.01it/s]
 60%|██████    | 6455/10682 [53:25<35:00,  2.01it/s]
 60%|██████    | 6456/10682 [53:26<34:57,  2.01it/s]
 60%|██████    | 6457/10682 [53:26<34:57,  2.01it/s]
 60%|██████    | 6458/10682 [53:27<34:53,  2.02it/s]
 60%|██████    | 6459/10682 [53:27<34:54,  2.02it/s]
 60%|██████    | 6460/10682 [53:28<34:53,  2.02it/s]
 60%|██████    | 6461/10682 [53:28<34:52,  2.02it/s]
 60%|██████    | 6462/10682 [53:29<34:51,  2.02it/s]
 61%|██████    | 6463/10682 [53:29<34:51,  2.02it/s]
 61%|██████    | 6464/10682 [53:30<34:51,  2.02it/s]
 61%|██████    | 6465/10682 [53:30<34:53,  2.01it/s]
 61%|██████    | 6466/10682 [53:31<34:53,  2.01it/s]
 61%|██████    | 6467/10682 [53:31<34:50,  2.02it/s]
 61%|██████    | 6468/10682 [53:32<34:50,  2.02it/s]
 61%|██████    | 6469/10682 [53:32<34:46,  2.02it/s]
 61%|██████    | 6470/10682 [53:33<34:50,  2.01it/s]
 61%|██████    | 6471/10682 [53:33<34:46,  2.02it/s]
 61%|██████    | 6472/10682 [53:34<34:49,  2.02it/s]
 61%|██████    | 6473/10682 [53:34<34:47,  2.02it/s]
 61%|██████    | 6474/10682 [53:35<34:48,  2.01it/s]
 61%|██████    | 6475/10682 [53:35<34:45,  2.02it/s]{'loss': 3.4876, 'grad_norm': 0.21488559246063232, 'learning_rate': 0.00040266519392417936, 'epoch': 0.61}
                                                    
-
 61%|██████    | 6475/10682 [53:35<34:45,  2.02it/s]
 61%|██████    | 6476/10682 [53:36<34:46,  2.02it/s]
 61%|██████    | 6477/10682 [53:36<34:46,  2.02it/s]
 61%|██████    | 6478/10682 [53:37<34:42,  2.02it/s]
 61%|██████    | 6479/10682 [53:37<34:42,  2.02it/s]
 61%|██████    | 6480/10682 [53:38<34:41,  2.02it/s]
 61%|██████    | 6481/10682 [53:38<34:43,  2.02it/s]
 61%|██████    | 6482/10682 [53:39<34:45,  2.01it/s]
 61%|██████    | 6483/10682 [53:39<34:44,  2.01it/s]
 61%|██████    | 6484/10682 [53:40<34:43,  2.01it/s]
 61%|██████    | 6485/10682 [53:40<34:43,  2.01it/s]
 61%|██████    | 6486/10682 [53:41<34:41,  2.02it/s]
 61%|██████    | 6487/10682 [53:41<34:39,  2.02it/s]
 61%|██████    | 6488/10682 [53:42<34:38,  2.02it/s]
 61%|██████    | 6489/10682 [53:42<34:39,  2.02it/s]
 61%|██████    | 6490/10682 [53:43<34:38,  2.02it/s]
 61%|██████    | 6491/10682 [53:43<34:37,  2.02it/s]
 61%|██████    | 6492/10682 [53:44<34:37,  2.02it/s]
 61%|██████    | 6493/10682 [53:44<34:37,  2.02it/s]
 61%|██████    | 6494/10682 [53:45<34:35,  2.02it/s]
 61%|██████    | 6495/10682 [53:45<34:34,  2.02it/s]
 61%|██████    | 6496/10682 [53:46<34:34,  2.02it/s]
 61%|██████    | 6497/10682 [53:46<34:32,  2.02it/s]
 61%|██████    | 6498/10682 [53:47<34:30,  2.02it/s]
 61%|██████    | 6499/10682 [53:47<34:31,  2.02it/s]
 61%|██████    | 6500/10682 [53:48<34:31,  2.02it/s]{'loss': 3.4826, 'grad_norm': 0.21983318030834198, 'learning_rate': 0.00039866155582263653, 'epoch': 0.61}
                                                    
-
 61%|██████    | 6500/10682 [53:48<34:31,  2.02it/s]
 61%|██████    | 6501/10682 [53:48<34:37,  2.01it/s]
 61%|██████    | 6502/10682 [53:49<34:33,  2.02it/s]
 61%|██████    | 6503/10682 [53:49<34:34,  2.01it/s]
 61%|██████    | 6504/10682 [53:50<34:32,  2.02it/s]
 61%|██████    | 6505/10682 [53:50<34:32,  2.02it/s]
 61%|██████    | 6506/10682 [53:51<34:30,  2.02it/s]
 61%|██████    | 6507/10682 [53:51<34:29,  2.02it/s]
 61%|██████    | 6508/10682 [53:52<34:27,  2.02it/s]
 61%|██████    | 6509/10682 [53:52<34:26,  2.02it/s]
 61%|██████    | 6510/10682 [53:53<34:25,  2.02it/s]
 61%|██████    | 6511/10682 [53:53<34:26,  2.02it/s]
 61%|██████    | 6512/10682 [53:54<34:25,  2.02it/s]
 61%|██████    | 6513/10682 [53:54<34:24,  2.02it/s]
 61%|██████    | 6514/10682 [53:55<34:24,  2.02it/s]
 61%|██████    | 6515/10682 [53:55<34:22,  2.02it/s]
 61%|██████    | 6516/10682 [53:56<34:23,  2.02it/s]
 61%|██████    | 6517/10682 [53:56<34:21,  2.02it/s]
 61%|██████    | 6518/10682 [53:57<34:21,  2.02it/s]
 61%|██████    | 6519/10682 [53:57<34:21,  2.02it/s]
 61%|██████    | 6520/10682 [53:58<34:21,  2.02it/s]
 61%|██████    | 6521/10682 [53:58<34:19,  2.02it/s]
 61%|██████    | 6522/10682 [53:59<34:20,  2.02it/s]
 61%|██████    | 6523/10682 [53:59<34:19,  2.02it/s]
 61%|██████    | 6524/10682 [54:00<34:20,  2.02it/s]
 61%|██████    | 6525/10682 [54:00<34:19,  2.02it/s]
                                                    {'loss': 3.4922, 'grad_norm': 0.21692056953907013, 'learning_rate': 0.0003946646821898418, 'epoch': 0.61}
-
 61%|██████    | 6525/10682 [54:00<34:19,  2.02it/s]
 61%|██████    | 6526/10682 [54:01<34:26,  2.01it/s]
 61%|██████    | 6527/10682 [54:01<34:23,  2.01it/s]
 61%|██████    | 6528/10682 [54:02<34:20,  2.02it/s]
 61%|██████    | 6529/10682 [54:02<34:19,  2.02it/s]
 61%|██████    | 6530/10682 [54:03<34:19,  2.02it/s]
 61%|██████    | 6531/10682 [54:03<34:19,  2.02it/s]
 61%|██████    | 6532/10682 [54:04<34:19,  2.02it/s]
 61%|██████    | 6533/10682 [54:04<34:20,  2.01it/s]
 61%|██████    | 6534/10682 [54:05<34:17,  2.02it/s]
 61%|██████    | 6535/10682 [54:05<34:15,  2.02it/s]
 61%|██████    | 6536/10682 [54:06<34:12,  2.02it/s]
 61%|██████    | 6537/10682 [54:06<34:13,  2.02it/s]
 61%|██████    | 6538/10682 [54:07<34:12,  2.02it/s]
 61%|██████    | 6539/10682 [54:07<34:12,  2.02it/s]
 61%|██████    | 6540/10682 [54:08<34:09,  2.02it/s]
 61%|██████    | 6541/10682 [54:08<34:10,  2.02it/s]
 61%|██████    | 6542/10682 [54:09<34:08,  2.02it/s]
 61%|██████▏   | 6543/10682 [54:09<34:08,  2.02it/s]
 61%|██████▏   | 6544/10682 [54:10<34:06,  2.02it/s]
 61%|██████▏   | 6545/10682 [54:10<34:08,  2.02it/s]
 61%|██████▏   | 6546/10682 [54:11<34:06,  2.02it/s]
 61%|██████▏   | 6547/10682 [54:11<34:07,  2.02it/s]
 61%|██████▏   | 6548/10682 [54:12<34:05,  2.02it/s]
 61%|██████▏   | 6549/10682 [54:12<34:06,  2.02it/s]
 61%|██████▏   | 6550/10682 [54:13<34:06,  2.02it/s]{'loss': 3.4799, 'grad_norm': 0.22251372039318085, 'learning_rate': 0.0003906748398221429, 'epoch': 0.61}
                                                    
-
 61%|██████▏   | 6550/10682 [54:13<34:06,  2.02it/s]
 61%|██████▏   | 6551/10682 [54:13<34:07,  2.02it/s]
 61%|██████▏   | 6552/10682 [54:14<34:10,  2.01it/s]
 61%|██████▏   | 6553/10682 [54:14<34:08,  2.02it/s]
 61%|██████▏   | 6554/10682 [54:15<34:09,  2.01it/s]
 61%|���█████▏   | 6555/10682 [54:15<34:08,  2.01it/s]
 61%|██████▏   | 6556/10682 [54:15<34:06,  2.02it/s]
 61%|██████▏   | 6557/10682 [54:16<34:05,  2.02it/s]
 61%|██████▏   | 6558/10682 [54:16<34:04,  2.02it/s]
 61%|██████▏   | 6559/10682 [54:17<34:01,  2.02it/s]
 61%|██████▏   | 6560/10682 [54:17<34:03,  2.02it/s]
 61%|██████▏   | 6561/10682 [54:18<34:05,  2.01it/s]
 61%|██████▏   | 6562/10682 [54:18<34:03,  2.02it/s]
 61%|██████▏   | 6563/10682 [54:19<34:03,  2.02it/s]
 61%|██████▏   | 6564/10682 [54:19<34:00,  2.02it/s]
 61%|██████▏   | 6565/10682 [54:20<33:59,  2.02it/s]
 61%|██████▏   | 6566/10682 [54:20<33:58,  2.02it/s]
 61%|██████▏   | 6567/10682 [54:21<33:57,  2.02it/s]
 61%|██████▏   | 6568/10682 [54:21<33:58,  2.02it/s]
 61%|██████▏   | 6569/10682 [54:22<33:56,  2.02it/s]
 62%|██████▏   | 6570/10682 [54:22<33:55,  2.02it/s]
 62%|██████▏   | 6571/10682 [54:23<33:59,  2.02it/s]
 62%|██████▏   | 6572/10682 [54:23<33:59,  2.02it/s]
 62%|██████▏   | 6573/10682 [54:24<33:59,  2.02it/s]
 62%|██████▏   | 6574/10682 [54:24<33:59,  2.01it/s]
 62%|██████▏   | 6575/10682 [54:25<33:55,  2.02it/s]
                                                    
{'loss': 3.4807, 'grad_norm': 0.21863852441310883, 'learning_rate': 0.00038669229504654146, 'epoch': 0.62}
-
 62%|██████▏   | 6575/10682 [54:25<33:55,  2.02it/s]
 62%|██████▏   | 6576/10682 [54:25<33:57,  2.01it/s]
 62%|██████▏   | 6577/10682 [54:26<33:55,  2.02it/s]
 62%|██████▏   | 6578/10682 [54:26<33:54,  2.02it/s]
 62%|██████▏   | 6579/10682 [54:27<33:52,  2.02it/s]
 62%|██████▏   | 6580/10682 [54:27<33:51,  2.02it/s]
 62%|██████▏   | 6581/10682 [54:28<33:51,  2.02it/s]
 62%|██████▏   | 6582/10682 [54:28<33:51,  2.02it/s]
 62%|██████▏   | 6583/10682 [54:29<33:50,  2.02it/s]
 62%|██████▏   | 6584/10682 [54:29<33:51,  2.02it/s]
 62%|██████▏   | 6585/10682 [54:30<33:51,  2.02it/s]
 62%|██████▏   | 6586/10682 [54:30<33:53,  2.01it/s]
 62%|██████▏   | 6587/10682 [54:31<33:51,  2.02it/s]
 62%|██████▏   | 6588/10682 [54:31<33:51,  2.02it/s]
 62%|██████▏   | 6589/10682 [54:32<33:50,  2.02it/s]
 62%|██████▏   | 6590/10682 [54:32<33:49,  2.02it/s]
 62%|██████▏   | 6591/10682 [54:33<33:50,  2.01it/s]
 62%|██████▏   | 6592/10682 [54:33<33:51,  2.01it/s]
 62%|██████▏   | 6593/10682 [54:34<33:50,  2.01it/s]
 62%|██████▏   | 6594/10682 [54:34<33:48,  2.02it/s]
 62%|██████▏   | 6595/10682 [54:35<33:47,  2.02it/s]
 62%|██████▏   | 6596/10682 [54:35<33:46,  2.02it/s]
 62%|██████▏   | 6597/10682 [54:36<33:45,  2.02it/s]
 62%|██████▏   | 6598/10682 [54:36<33:42,  2.02it/s]
 62%|██████▏   | 6599/10682 [54:37<33:43,  2.02it/s]
 62%|██████▏   | 6600/10682 [54:37<33:44,  2.02it/s]{'loss': 3.4815, 'grad_norm': 0.22285053133964539, 'learning_rate': 0.00038271731370291583, 'epoch': 0.62}
                                                    
-
 62%|██████▏   | 6600/10682 [54:37<33:44,  2.02it/s]
 62%|██████▏   | 6601/10682 [54:38<33:45,  2.01it/s]
 62%|██████▏   | 6602/10682 [54:38<33:44,  2.02it/s]
 62%|██████▏   | 6603/10682 [54:39<33:42,  2.02it/s]
 62%|██████▏   | 6604/10682 [54:39<33:42,  2.02it/s]
 62%|██████▏   | 6605/10682 [54:40<33:40,  2.02it/s]
 62%|██████▏   | 6606/10682 [54:40<33:43,  2.01it/s]
 62%|██████▏   | 6607/10682 [54:41<33:43,  2.01it/s]
 62%|██████▏   | 6608/10682 [54:41<33:41,  2.02it/s]
 62%|██████▏   | 6609/10682 [54:42<33:42,  2.01it/s]
 62%|██████▏   | 6610/10682 [54:42<33:39,  2.02it/s]
 62%|██████▏   | 6611/10682 [54:43<33:42,  2.01it/s]
 62%|██████▏   | 6612/10682 [54:43<33:38,  2.02it/s]
 62%|██████▏   | 6613/10682 [54:44<33:41,  2.01it/s]
 62%|██████▏   | 6614/10682 [54:44<33:39,  2.01it/s]
 62%|██████▏   | 6615/10682 [54:45<33:37,  2.02it/s]
 62%|██████▏   | 6616/10682 [54:45<33:35,  2.02it/s]
 62%|██████▏   | 6617/10682 [54:46<33:35,  2.02it/s]
 62%|██████▏   | 6618/10682 [54:46<33:37,  2.01it/s]
 62%|██████▏   | 6619/10682 [54:47<33:35,  2.02it/s]
 62%|██████▏   | 6620/10682 [54:47<33:34,  2.02it/s]
 62%|██████▏   | 6621/10682 [54:48<33:35,  2.02it/s]
 62%|██████▏   | 6622/10682 [54:48<33:32,  2.02it/s]
 62%|██████▏   | 6623/10682 [54:49<33:33,  2.02it/s]
 62%|██████▏   | 6624/10682 [54:49<33:30,  2.02it/s]
 62%|██████▏   | 6625/10682 [54:50<33:33,  2.02it/s]
                                                    {'loss': 3.4849, 'grad_norm': 0.2284301370382309, 'learning_rate': 0.0003787501611262762, 'epoch': 0.62}
-
 62%|██████▏   | 6625/10682 [54:50<33:33,  2.02it/s]
 62%|██████▏   | 6626/10682 [54:50<33:40,  2.01it/s]
 62%|██████▏   | 6627/10682 [54:51<33:36,  2.01it/s]
 62%|██████▏   | 6628/10682 [54:51<33:37,  2.01it/s]
 62%|██████▏   | 6629/10682 [54:52<33:36,  2.01it/s]
 62%|██████▏   | 6630/10682 [54:52<33:36,  2.01it/s]
 62%|██████▏   | 6631/10682 [54:53<33:36,  2.01it/s]
 62%|██████▏   | 6632/10682 [54:53<33:32,  2.01it/s]
 62%|██████▏   | 6633/10682 [54:54<33:32,  2.01it/s]
 62%|██████▏   | 6634/10682 [54:54<33:28,  2.02it/s]
 62%|██████▏   | 6635/10682 [54:55<33:27,  2.02it/s]
 62%|██████▏   | 6636/10682 [54:55<33:23,  2.02it/s]
 62%|██████▏   | 6637/10682 [54:56<33:24,  2.02it/s]
 62%|██████▏   | 6638/10682 [54:56<33:24,  2.02it/s]
 62%|██████▏   | 6639/10682 [54:57<33:23,  2.02it/s]
 62%|██████▏   | 6640/10682 [54:57<33:27,  2.01it/s]
 62%|██████▏   | 6641/10682 [54:58<33:25,  2.02it/s]
 62%|██████▏   | 6642/10682 [54:58<33:26,  2.01it/s]
 62%|██████▏   | 6643/10682 [54:59<33:22,  2.02it/s]
 62%|██████▏   | 6644/10682 [54:59<33:21,  2.02it/s]
 62%|██████▏   | 6645/10682 [55:00<33:17,  2.02it/s]
 62%|██████▏   | 6646/10682 [55:00<33:18,  2.02it/s]
 62%|██████▏   | 6647/10682 [55:01<33:14,  2.02it/s]
 62%|██████▏   | 6648/10682 [55:01<33:18,  2.02it/s]
 62%|██████▏   | 6649/10682 [55:02<33:20,  2.02it/s]
 62%|██████▏   | 6650/10682 [55:02<33:20,  2.02it/s]
                                                    {'loss': 3.4764, 'grad_norm': 0.23645159602165222, 'learning_rate': 0.0003747911021290519, 'epoch': 0.62}
-
 62%|██████▏   | 6650/10682 [55:02<33:20,  2.02it/s]
 62%|██████▏   | 6651/10682 [55:03<33:24,  2.01it/s]
 62%|██████▏   | 6652/10682 [55:03<33:20,  2.01it/s]
 62%|██████▏   | 6653/10682 [55:04<33:18,  2.02it/s]
 62%|██████▏   | 6654/10682 [55:04<33:16,  2.02it/s]
 62%|██████▏   | 6655/10682 [55:05<33:16,  2.02it/s]
 62%|██████▏   | 6656/10682 [55:05<33:13,  2.02it/s]
 62%|██████▏   | 6657/10682 [55:06<33:14,  2.02it/s]
 62%|██████▏   | 6658/10682 [55:06<33:15,  2.02it/s]
 62%|██████▏   | 6659/10682 [55:07<33:14,  2.02it/s]
 62%|██████▏   | 6660/10682 [55:07<33:15,  2.02it/s]
 62%|██████▏   | 6661/10682 [55:08<33:15,  2.02it/s]
 62%|██████▏   | 6662/10682 [55:08<33:15,  2.01it/s]
 62%|██████▏   | 6663/10682 [55:09<33:11,  2.02it/s]
 62%|██████▏   | 6664/10682 [55:09<33:11,  2.02it/s]
 62%|██████▏   | 6665/10682 [55:10<33:07,  2.02it/s]
 62%|██████▏   | 6666/10682 [55:10<33:07,  2.02it/s]
 62%|██████▏   | 6667/10682 [55:11<33:08,  2.02it/s]
 62%|██████▏   | 6668/10682 [55:11<33:08,  2.02it/s]
 62%|██████▏   | 6669/10682 [55:12<33:07,  2.02it/s]
 62%|██████▏   | 6670/10682 [55:12<33:07,  2.02it/s]
 62%|██████▏   | 6671/10682 [55:13<33:05,  2.02it/s]
 62%|██████▏   | 6672/10682 [55:13<33:06,  2.02it/s]
 62%|██████▏   | 6673/10682 [55:14<33:05,  2.02it/s]
 62%|██████▏   | 6674/10682 [55:14<33:05,  2.02it/s]
 62%|██████▏   | 6675/10682 [55:15<33:05,  2.02it/s]
                                                    {'loss': 3.4781, 'grad_norm': 0.22300295531749725, 'learning_rate': 0.0003708404009834161, 'epoch': 0.62}
-
 62%|██████▏   | 6675/10682 [55:15<33:05,  2.02it/s]
 62%|██████▏   | 6676/10682 [55:15<33:07,  2.02it/s]
 63%|██████▎   | 6677/10682 [55:16<33:07,  2.02it/s]
 63%|██████▎   | 6678/10682 [55:16<33:05,  2.02it/s]
 63%|██████▎   | 6679/10682 [55:17<33:07,  2.01it/s]
 63%|██████▎   | 6680/10682 [55:17<33:07,  2.01it/s]
 63%|██████▎   | 6681/10682 [55:17<33:02,  2.02it/s]
 63%|██████▎   | 6682/10682 [55:18<33:02,  2.02it/s]
 63%|██████▎   | 6683/10682 [55:18<33:00,  2.02it/s]
 63%|██████▎   | 6684/10682 [55:19<33:00,  2.02it/s]
 63%|██████▎   | 6685/10682 [55:19<32:58,  2.02it/s]
 63%|██████▎   | 6686/10682 [55:20<32:57,  2.02it/s]
 63%|██████▎   | 6687/10682 [55:20<32:57,  2.02it/s]
 63%|██████▎   | 6688/10682 [55:21<32:55,  2.02it/s]
 63%|██████▎   | 6689/10682 [55:21<32:56,  2.02it/s]
 63%|██████▎   | 6690/10682 [55:22<32:58,  2.02it/s]
 63%|██████▎   | 6691/10682 [55:22<32:55,  2.02it/s]
 63%|██████▎   | 6692/10682 [55:23<32:56,  2.02it/s]
 63%|██████▎   | 6693/10682 [55:23<32:56,  2.02it/s]
 63%|██████▎   | 6694/10682 [55:24<32:58,  2.02it/s]
 63%|██████▎   | 6695/10682 [55:24<32:55,  2.02it/s]
 63%|██████▎   | 6696/10682 [55:25<32:55,  2.02it/s]
 63%|██████▎   | 6697/10682 [55:25<32:53,  2.02it/s]
 63%|██████▎   | 6698/10682 [55:26<32:55,  2.02it/s]
 63%|██████▎   | 6699/10682 [55:26<32:54,  2.02it/s]
 63%|██████▎   | 6700/10682 [55:27<32:54,  2.02it/s]{'loss': 3.4777, 'grad_norm': 0.2152986228466034, 'learning_rate': 0.0003668983214036445, 'epoch': 0.63}
                                                    
-
 63%|██████▎   | 6700/10682 [55:27<32:54,  2.02it/s]
 63%|██████▎   | 6701/10682 [55:27<32:59,  2.01it/s]
 63%|██████▎   | 6702/10682 [55:28<32:55,  2.02it/s]
 63%|██████▎   | 6703/10682 [55:28<32:54,  2.02it/s]
 63%|██████▎   | 6704/10682 [55:29<32:51,  2.02it/s]
 63%|██████▎   | 6705/10682 [55:29<32:50,  2.02it/s]
 63%|██████▎   | 6706/10682 [55:30<32:50,  2.02it/s]
 63%|██████▎   | 6707/10682 [55:30<32:50,  2.02it/s]
 63%|██████▎   | 6708/10682 [55:31<32:50,  2.02it/s]
 63%|██████▎   | 6709/10682 [55:31<32:52,  2.01it/s]
 63%|██████▎   | 6710/10682 [55:32<32:49,  2.02it/s]
 63%|██████▎   | 6711/10682 [55:32<32:48,  2.02it/s]
 63%|██████▎   | 6712/10682 [55:33<32:44,  2.02it/s]
 63%|██████▎   | 6713/10682 [55:33<32:45,  2.02it/s]
 63%|██████▎   | 6714/10682 [55:34<32:44,  2.02it/s]
 63%|██████▎   | 6715/10682 [55:34<32:44,  2.02it/s]
 63%|██████▎   | 6716/10682 [55:35<32:43,  2.02it/s]
 63%|██████▎   | 6717/10682 [55:35<32:45,  2.02it/s]
 63%|██████▎   | 6718/10682 [55:36<32:47,  2.01it/s]
 63%|██████▎   | 6719/10682 [55:36<32:47,  2.01it/s]
 63%|██████▎   | 6720/10682 [55:37<32:46,  2.02it/s]
 63%|██████▎   | 6721/10682 [55:37<32:46,  2.01it/s]
 63%|██████▎   | 6722/10682 [55:38<32:44,  2.02it/s]
 63%|██████▎   | 6723/10682 [55:38<32:41,  2.02it/s]
 63%|██████▎   | 6724/10682 [55:39<32:40,  2.02it/s]
 63%|██████▎   | 6725/10682 [55:39<32:37,  2.02it/s]{'loss': 3.4781, 'grad_norm': 0.21404793858528137, 'learning_rate': 0.00036296512652851323, 'epoch': 0.63}
                                                    
-
 63%|██████▎   | 6725/10682 [55:39<32:37,  2.02it/s]
 63%|██████▎   | 6726/10682 [55:40<32:44,  2.01it/s]
 63%|██████▎   | 6727/10682 [55:40<32:42,  2.02it/s]
 63%|██████▎   | 6728/10682 [55:41<32:39,  2.02it/s]
 63%|██████▎   | 6729/10682 [55:41<32:39,  2.02it/s]
 63%|██████▎   | 6730/10682 [55:42<32:38,  2.02it/s]
 63%|██████▎   | 6731/10682 [55:42<32:38,  2.02it/s]
 63%|██████▎   | 6732/10682 [55:43<32:35,  2.02it/s]
 63%|██████▎   | 6733/10682 [55:43<32:36,  2.02it/s]
 63%|██████▎   | 6734/10682 [55:44<32:34,  2.02it/s]
 63%|██████▎   | 6735/10682 [55:44<32:33,  2.02it/s]
 63%|██████▎   | 6736/10682 [55:45<32:34,  2.02it/s]
 63%|██████▎   | 6737/10682 [55:45<32:34,  2.02it/s]
 63%|██████▎   | 6738/10682 [55:46<32:34,  2.02it/s]
 63%|██████▎   | 6739/10682 [55:46<32:36,  2.02it/s]
 63%|██████▎   | 6740/10682 [55:47<32:33,  2.02it/s]
 63%|██████▎   | 6741/10682 [55:47<32:32,  2.02it/s]
 63%|██████▎   | 6742/10682 [55:48<32:28,  2.02it/s]
 63%|██████▎   | 6743/10682 [55:48<32:30,  2.02it/s]
 63%|██████▎   | 6744/10682 [55:49<32:25,  2.02it/s]
 63%|██████▎   | 6745/10682 [55:49<32:25,  2.02it/s]
 63%|██████▎   | 6746/10682 [55:50<32:26,  2.02it/s]
 63%|██████▎   | 6747/10682 [55:50<32:26,  2.02it/s]
 63%|██████▎   | 6748/10682 [55:51<32:30,  2.02it/s]
 63%|██████▎   | 6749/10682 [55:51<32:29,  2.02it/s]
 63%|██████▎   | 6750/10682 [55:52<32:28,  2.02it/s]{'loss': 3.4697, 'grad_norm': 0.21071042120456696, 'learning_rate': 0.00035904107890373196, 'epoch': 0.63}
                                                    
-
 63%|██████▎   | 6750/10682 [55:52<32:28,  2.02it/s]
 63%|██████▎   | 6751/10682 [55:52<32:31,  2.01it/s]
 63%|██████▎   | 6752/10682 [55:53<32:29,  2.02it/s]
 63%|██████▎   | 6753/10682 [55:53<32:27,  2.02it/s]
 63%|██████▎   | 6754/10682 [55:54<32:25,  2.02it/s]
 63%|██████▎   | 6755/10682 [55:54<32:24,  2.02it/s]
 63%|██████▎   | 6756/10682 [55:55<32:24,  2.02it/s]
 63%|���█████▎   | 6757/10682 [55:55<32:25,  2.02it/s]
 63%|██████▎   | 6758/10682 [55:56<37:52,  1.73it/s]
 63%|██████▎   | 6759/10682 [55:56<36:12,  1.81it/s]
 63%|██████▎   | 6760/10682 [55:57<35:05,  1.86it/s]
 63%|██████▎   | 6761/10682 [55:57<34:17,  1.91it/s]
 63%|██████▎   | 6762/10682 [55:58<33:41,  1.94it/s]
 63%|██████▎   | 6763/10682 [55:58<33:16,  1.96it/s]
 63%|██████▎   | 6764/10682 [55:59<32:59,  1.98it/s]
 63%|██████▎   | 6765/10682 [55:59<32:52,  1.99it/s]
 63%|██████▎   | 6766/10682 [56:00<32:39,  2.00it/s]
 63%|██████▎   | 6767/10682 [56:00<32:35,  2.00it/s]
 63%|██████▎   | 6768/10682 [56:01<32:29,  2.01it/s]
 63%|██████▎   | 6769/10682 [56:01<32:29,  2.01it/s]
 63%|██████▎   | 6770/10682 [56:02<32:24,  2.01it/s]
 63%|██████▎   | 6771/10682 [56:02<32:22,  2.01it/s]
 63%|██████▎   | 6772/10682 [56:03<32:22,  2.01it/s]
 63%|██████▎   | 6773/10682 [56:03<32:18,  2.02it/s]
 63%|██████▎   | 6774/10682 [56:04<32:17,  2.02it/s]
 63%|██████▎   | 6775/10682 [56:04<32:14,  2.02it/s]
                                                    
{'loss': 3.4796, 'grad_norm': 0.2233514040708542, 'learning_rate': 0.00035512644046442116, 'epoch': 0.63}
-
 63%|██████▎   | 6775/10682 [58:26<32:14,  2.02it/s]
 63%|██████▎   | 6776/10682 [58:26<46:37:17, 42.97s/it]
 63%|██████▎   | 6777/10682 [58:27<32:47:14, 30.23s/it]
 63%|██████▎   | 6778/10682 [58:27<23:06:23, 21.31s/it]
 63%|██████▎   | 6779/10682 [58:28<16:19:56, 15.06s/it]
 63%|██████▎   | 6780/10682 [58:28<11:35:25, 10.69s/it]
 63%|██████▎   | 6781/10682 [58:29<8:16:17,  7.63s/it] 
 63%|██████▎   | 6782/10682 [58:29<5:56:55,  5.49s/it]
 63%|██████▎   | 6783/10682 [58:30<4:19:26,  3.99s/it]
 64%|██████▎   | 6784/10682 [58:30<3:11:10,  2.94s/it]
 64%|██████▎   | 6785/10682 [58:31<2:23:24,  2.21s/it]
 64%|██████▎   | 6786/10682 [58:31<1:50:01,  1.69s/it]
 64%|██████▎   | 6787/10682 [58:32<1:26:38,  1.33s/it]
 64%|██████▎   | 6788/10682 [58:33<1:15:36,  1.16s/it]
 64%|██████▎   | 6789/10682 [58:33<1:02:34,  1.04it/s]
 64%|██████▎   | 6790/10682 [58:34<53:25,  1.21it/s]  
 64%|██████▎   | 6791/10682 [58:34<46:57,  1.38it/s]
 64%|██████▎   | 6792/10682 [58:35<42:31,  1.52it/s]
 64%|██████▎   | 6793/10682 [58:35<39:21,  1.65it/s]
 64%|██████▎   | 6794/10682 [58:36<37:11,  1.74it/s]
 64%|██████▎   | 6795/10682 [58:36<35:37,  1.82it/s]
 64%|██████▎   | 6796/10682 [58:37<34:34,  1.87it/s]
 64%|██████▎   | 6797/10682 [58:37<33:46,  1.92it/s]
 64%|██████▎   | 6798/10682 [58:38<33:16,  1.95it/s]
 64%|██████▎   | 6799/10682 [58:38<32:51,  1.97it/s]
 64%|██████▎   | 6800/10682 [58:39<32:34,  1.99it/s]{'loss': 3.4726, 'grad_norm': 0.21001626551151276, 'learning_rate': 0.00035122147251762554, 'epoch': 0.64}
                                                    
-
 64%|██████▎   | 6800/10682 [58:39<32:34,  1.99it/s]
 64%|██████▎   | 6801/10682 [58:39<32:25,  1.99it/s]
 64%|██████▎   | 6802/10682 [58:40<32:15,  2.00it/s]
 64%|██████▎   | 6803/10682 [58:40<32:08,  2.01it/s]
 64%|██████▎   | 6804/10682 [58:41<32:07,  2.01it/s]
 64%|██████▎   | 6805/10682 [58:41<32:01,  2.02it/s]
 64%|██████▎   | 6806/10682 [58:42<32:02,  2.02it/s]
 64%|██████▎   | 6807/10682 [58:42<31:59,  2.02it/s]
 64%|██████▎   | 6808/10682 [58:43<32:01,  2.02it/s]
 64%|██████▎   | 6809/10682 [58:43<32:00,  2.02it/s]
 64%|██████▍   | 6810/10682 [58:44<31:58,  2.02it/s]
 64%|██████▍   | 6811/10682 [58:44<31:55,  2.02it/s]
 64%|██████▍   | 6812/10682 [58:45<31:53,  2.02it/s]
 64%|██████▍   | 6813/10682 [58:45<31:51,  2.02it/s]
 64%|██████▍   | 6814/10682 [58:46<31:53,  2.02it/s]
 64%|██████▍   | 6815/10682 [58:46<31:53,  2.02it/s]
 64%|██████▍   | 6816/10682 [58:46<31:53,  2.02it/s]
 64%|██████▍   | 6817/10682 [58:47<31:50,  2.02it/s]
 64%|██████▍   | 6818/10682 [58:47<31:51,  2.02it/s]
 64%|██████▍   | 6819/10682 [58:48<31:49,  2.02it/s]
 64%|██████▍   | 6820/10682 [58:48<31:49,  2.02it/s]
 64%|██████▍   | 6821/10682 [58:49<31:47,  2.02it/s]
 64%|██████▍   | 6822/10682 [58:49<31:47,  2.02it/s]
 64%|██████▍   | 6823/10682 [58:50<31:47,  2.02it/s]
 64%|██████▍   | 6824/10682 [58:50<31:45,  2.02it/s]
 64%|██████▍   | 6825/10682 [58:51<31:46,  2.02it/s]
                                                    
{'loss': 3.4709, 'grad_norm': 0.21363143622875214, 'learning_rate': 0.00034732643572487243, 'epoch': 0.64}
-
 64%|██████▍   | 6825/10682 [58:51<31:46,  2.02it/s]
 64%|██████▍   | 6826/10682 [58:51<31:51,  2.02it/s]
 64%|██████▍   | 6827/10682 [58:52<31:51,  2.02it/s]
 64%|██████▍   | 6828/10682 [58:52<31:51,  2.02it/s]
 64%|██████▍   | 6829/10682 [58:53<31:48,  2.02it/s]
 64%|██████▍   | 6830/10682 [58:53<31:48,  2.02it/s]
 64%|██████▍   | 6831/10682 [58:54<31:47,  2.02it/s]
 64%|██████▍   | 6832/10682 [58:54<31:44,  2.02it/s]
 64%|██████▍   | 6833/10682 [58:55<31:44,  2.02it/s]
 64%|██████▍   | 6834/10682 [58:55<31:43,  2.02it/s]
 64%|██████▍   | 6835/10682 [58:56<31:44,  2.02it/s]
 64%|██████▍   | 6836/10682 [58:56<31:42,  2.02it/s]
 64%|██████▍   | 6837/10682 [58:57<31:45,  2.02it/s]
 64%|██████▍   | 6838/10682 [58:57<31:46,  2.02it/s]
 64%|██████▍   | 6839/10682 [58:58<31:42,  2.02it/s]
 64%|██████▍   | 6840/10682 [58:58<31:42,  2.02it/s]
 64%|██████▍   | 6841/10682 [58:59<31:38,  2.02it/s]
 64%|██████▍   | 6842/10682 [58:59<31:38,  2.02it/s]
 64%|██████▍   | 6843/10682 [59:00<31:38,  2.02it/s]
 64%|██████▍   | 6844/10682 [59:00<31:38,  2.02it/s]
 64%|██████▍   | 6845/10682 [59:01<31:37,  2.02it/s]
 64%|██████▍   | 6846/10682 [59:01<31:37,  2.02it/s]
 64%|██████▍   | 6847/10682 [59:02<31:37,  2.02it/s]
 64%|██████▍   | 6848/10682 [59:02<31:36,  2.02it/s]
 64%|██████▍   | 6849/10682 [59:03<31:32,  2.03it/s]
 64%|██████▍   | 6850/10682 [59:03<31:32,  2.02it/s]
                                                    {'loss': 3.4682, 'grad_norm': 0.2166104018688202, 'learning_rate': 0.000343441590084772, 'epoch': 0.64}
-
 64%|██████▍   | 6850/10682 [59:03<31:32,  2.02it/s]
 64%|██████▍   | 6851/10682 [59:04<31:33,  2.02it/s]
 64%|██████▍   | 6852/10682 [59:04<31:34,  2.02it/s]
 64%|██████▍   | 6853/10682 [59:05<31:32,  2.02it/s]
 64%|██████▍   | 6854/10682 [59:05<31:46,  2.01it/s]
 64%|██████▍   | 6855/10682 [59:06<31:46,  2.01it/s]
 64%|██████▍   | 6856/10682 [59:06<31:41,  2.01it/s]
 64%|██████▍   | 6857/10682 [59:07<31:42,  2.01it/s]
 64%|██████▍   | 6858/10682 [59:07<31:41,  2.01it/s]
 64%|██████▍   | 6859/10682 [59:08<31:37,  2.02it/s]
 64%|██████▍   | 6860/10682 [59:08<31:38,  2.01it/s]
 64%|██████▍   | 6861/10682 [59:09<31:33,  2.02it/s]
 64%|██████▍   | 6862/10682 [59:09<31:33,  2.02it/s]
 64%|██████▍   | 6863/10682 [59:10<31:31,  2.02it/s]
 64%|██████▍   | 6864/10682 [59:10<31:33,  2.02it/s]
 64%|██████▍   | 6865/10682 [59:11<31:30,  2.02it/s]
 64%|██████▍   | 6866/10682 [59:11<31:29,  2.02it/s]
 64%|██████▍   | 6867/10682 [59:12<31:30,  2.02it/s]
 64%|██████▍   | 6868/10682 [59:12<31:30,  2.02it/s]
 64%|██████▍   | 6869/10682 [59:13<31:29,  2.02it/s]
 64%|██████▍   | 6870/10682 [59:13<31:28,  2.02it/s]
 64%|██████▍   | 6871/10682 [59:14<31:26,  2.02it/s]
 64%|██████▍   | 6872/10682 [59:14<31:25,  2.02it/s]
 64%|██████▍   | 6873/10682 [59:15<31:25,  2.02it/s]
 64%|██████▍   | 6874/10682 [59:15<31:25,  2.02it/s]
 64%|██████▍   | 6875/10682 [59:16<31:27,  2.02it/s]{'loss': 3.4643, 'grad_norm': 0.218482106924057, 'learning_rate': 0.0003395671949156626, 'epoch': 0.64}
-
                                                    

 64%|██████▍   | 6875/10682 [59:16<31:27,  2.02it/s]
 64%|██████▍   | 6876/10682 [59:16<31:30,  2.01it/s]
 64%|██████▍   | 6877/10682 [59:17<31:27,  2.02it/s]
 64%|██████▍   | 6878/10682 [59:17<31:27,  2.02it/s]
 64%|██████▍   | 6879/10682 [59:18<31:23,  2.02it/s]
 64%|██████▍   | 6880/10682 [59:18<31:22,  2.02it/s]
 64%|██████▍   | 6881/10682 [59:19<31:19,  2.02it/s]
 64%|██████▍   | 6882/10682 [59:19<31:20,  2.02it/s]
 64%|██████▍   | 6883/10682 [59:20<31:19,  2.02it/s]
 64%|██████▍   | 6884/10682 [59:20<31:21,  2.02it/s]
 64%|██████▍   | 6885/10682 [59:21<31:18,  2.02it/s]
 64%|██████▍   | 6886/10682 [59:21<31:16,  2.02it/s]
 64%|██████▍   | 6887/10682 [59:22<31:17,  2.02it/s]
 64%|██████▍   | 6888/10682 [59:22<31:16,  2.02it/s]
 64%|██████▍   | 6889/10682 [59:23<31:18,  2.02it/s]
 65%|██████▍   | 6890/10682 [59:23<31:15,  2.02it/s]
 65%|██████▍   | 6891/10682 [59:24<31:16,  2.02it/s]
 65%|██████▍   | 6892/10682 [59:24<31:14,  2.02it/s]
 65%|██████▍   | 6893/10682 [59:25<31:15,  2.02it/s]
 65%|██████▍   | 6894/10682 [59:25<31:14,  2.02it/s]
 65%|██████▍   | 6895/10682 [59:26<31:15,  2.02it/s]
 65%|██████▍   | 6896/10682 [59:26<31:15,  2.02it/s]
 65%|██████▍   | 6897/10682 [59:27<31:17,  2.02it/s]
 65%|██████▍   | 6898/10682 [59:27<31:15,  2.02it/s]
 65%|██████▍   | 6899/10682 [59:28<31:12,  2.02it/s]
 65%|██████▍   | 6900/10682 [59:28<31:13,  2.02it/s]
                                                    {'loss': 3.4604, 'grad_norm': 0.2229122519493103, 'learning_rate': 0.0003357035088382997, 'epoch': 0.65}
-

 65%|██████▍   | 6900/10682 [59:28<31:13,  2.02it/s]
 65%|██████▍   | 6901/10682 [59:29<31:13,  2.02it/s]
 65%|██████▍   | 6902/10682 [59:29<31:11,  2.02it/s]
 65%|██████▍   | 6903/10682 [59:30<31:09,  2.02it/s]
 65%|██████▍   | 6904/10682 [59:30<31:12,  2.02it/s]
 65%|██████▍   | 6905/10682 [59:31<31:09,  2.02it/s]
 65%|██████▍   | 6906/10682 [59:31<31:10,  2.02it/s]
 65%|██████▍   | 6907/10682 [59:32<31:09,  2.02it/s]
 65%|██████▍   | 6908/10682 [59:32<31:06,  2.02it/s]
 65%|██████▍   | 6909/10682 [59:33<31:06,  2.02it/s]
 65%|██████▍   | 6910/10682 [59:33<31:04,  2.02it/s]
 65%|██████▍   | 6911/10682 [59:34<31:03,  2.02it/s]
 65%|██████▍   | 6912/10682 [59:34<31:02,  2.02it/s]
 65%|██████▍   | 6913/10682 [59:35<31:01,  2.02it/s]
 65%|██████▍   | 6914/10682 [59:35<31:01,  2.02it/s]
 65%|██████▍   | 6915/10682 [59:36<31:02,  2.02it/s]
 65%|██████▍   | 6916/10682 [59:36<31:03,  2.02it/s]
 65%|██████▍   | 6917/10682 [59:36<31:02,  2.02it/s]
 65%|██████▍   | 6918/10682 [59:37<31:01,  2.02it/s]
 65%|██████▍   | 6919/10682 [59:37<31:01,  2.02it/s]
 65%|██████▍   | 6920/10682 [59:38<31:00,  2.02it/s]
 65%|██████▍   | 6921/10682 [59:38<30:59,  2.02it/s]
 65%|██████▍   | 6922/10682 [59:39<31:01,  2.02it/s]
 65%|██████▍   | 6923/10682 [59:39<31:01,  2.02it/s]
 65%|██████▍   | 6924/10682 [59:40<31:00,  2.02it/s]
 65%|██████▍   | 6925/10682 [59:40<31:01,  2.02it/s]{'loss': 3.4546, 'grad_norm': 0.2172532081604004, 'learning_rate': 0.0003318507897585944, 'epoch': 0.65}
                                                    
-
 65%|██████▍   | 6925/10682 [59:40<31:01,  2.02it/s]
 65%|██████▍   | 6926/10682 [59:41<31:06,  2.01it/s]
 65%|██████▍   | 6927/10682 [59:41<31:03,  2.02it/s]
 65%|██████▍   | 6928/10682 [59:42<31:01,  2.02it/s]
 65%|██████▍   | 6929/10682 [59:42<30:58,  2.02it/s]
 65%|██████▍   | 6930/10682 [59:43<30:58,  2.02it/s]
 65%|██████▍   | 6931/10682 [59:43<30:56,  2.02it/s]
 65%|██████▍   | 6932/10682 [59:44<30:55,  2.02it/s]
 65%|██████▍   | 6933/10682 [59:44<30:54,  2.02it/s]
 65%|██████▍   | 6934/10682 [59:45<30:56,  2.02it/s]
 65%|██████▍   | 6935/10682 [59:45<30:54,  2.02it/s]
 65%|██████▍   | 6936/10682 [59:46<30:55,  2.02it/s]
 65%|██████▍   | 6937/10682 [59:46<30:51,  2.02it/s]
 65%|██████▍   | 6938/10682 [59:47<30:52,  2.02it/s]
 65%|██████▍   | 6939/10682 [59:47<30:51,  2.02it/s]
 65%|██████▍   | 6940/10682 [59:48<30:50,  2.02it/s]
 65%|██████▍   | 6941/10682 [59:48<30:48,  2.02it/s]
 65%|██████▍   | 6942/10682 [59:49<30:51,  2.02it/s]
 65%|██████▍   | 6943/10682 [59:49<30:51,  2.02it/s]
 65%|██████▌   | 6944/10682 [59:50<30:51,  2.02it/s]
 65%|██████▌   | 6945/10682 [59:50<30:49,  2.02it/s]
 65%|██████▌   | 6946/10682 [59:51<30:50,  2.02it/s]
 65%|██████▌   | 6947/10682 [59:51<30:49,  2.02it/s]
 65%|██████▌   | 6948/10682 [59:52<30:49,  2.02it/s]
 65%|██████▌   | 6949/10682 [59:52<30:47,  2.02it/s]
 65%|██████▌   | 6950/10682 [59:53<30:46,  2.02it/s]
                                                    
{'loss': 3.468, 'grad_norm': 0.21285296976566315, 'learning_rate': 0.0003280092948503962, 'epoch': 0.65}
-
 65%|██████▌   | 6950/10682 [59:53<30:46,  2.02it/s]
 65%|██████▌   | 6951/10682 [59:53<30:49,  2.02it/s]
 65%|██████▌   | 6952/10682 [59:54<30:48,  2.02it/s]
 65%|██████▌   | 6953/10682 [59:54<30:48,  2.02it/s]
 65%|██████▌   | 6954/10682 [59:55<30:46,  2.02it/s]
 65%|██████▌   | 6955/10682 [59:55<30:47,  2.02it/s]
 65%|██████▌   | 6956/10682 [59:56<30:45,  2.02it/s]
 65%|██████▌   | 6957/10682 [59:56<30:46,  2.02it/s]
 65%|██████▌   | 6958/10682 [59:57<30:43,  2.02it/s]
 65%|██████▌   | 6959/10682 [59:57<30:40,  2.02it/s]
 65%|██████▌   | 6960/10682 [59:58<30:40,  2.02it/s]
 65%|██████▌   | 6961/10682 [59:58<30:38,  2.02it/s]
 65%|██████▌   | 6962/10682 [59:59<30:41,  2.02it/s]
 65%|██████▌   | 6963/10682 [59:59<30:40,  2.02it/s]
 65%|██████▌   | 6964/10682 [1:00:00<30:41,  2.02it/s]
 65%|██████▌   | 6965/10682 [1:00:00<30:39,  2.02it/s]
 65%|██████▌   | 6966/10682 [1:00:01<30:38,  2.02it/s]
 65%|██████▌   | 6967/10682 [1:00:01<30:37,  2.02it/s]
 65%|██████▌   | 6968/10682 [1:00:02<30:38,  2.02it/s]
 65%|██████▌   | 6969/10682 [1:00:02<30:35,  2.02it/s]
 65%|██████▌   | 6970/10682 [1:00:03<30:35,  2.02it/s]
 65%|██████▌   | 6971/10682 [1:00:03<30:35,  2.02it/s]
 65%|██████▌   | 6972/10682 [1:00:04<30:36,  2.02it/s]
 65%|██████▌   | 6973/10682 [1:00:04<30:34,  2.02it/s]
 65%|██████▌   | 6974/10682 [1:00:05<30:36,  2.02it/s]
 65%|██████▌   | 6975/10682 [1:00:05<30:35,  2.02it/s]
                                                      {'loss': 3.4641, 'grad_norm': 0.22115369141101837, 'learning_rate': 0.00032417928053832723, 'epoch': 0.65}
-
 65%|██████▌   | 6975/10682 [1:00:05<30:35,  2.02it/s]
 65%|██████▌   | 6976/10682 [1:00:06<30:39,  2.01it/s]
 65%|██████▌   | 6977/10682 [1:00:06<30:37,  2.02it/s]
 65%|██████▌   | 6978/10682 [1:00:07<30:35,  2.02it/s]
 65%|██████▌   | 6979/10682 [1:00:07<30:34,  2.02it/s]
 65%|██████▌   | 6980/10682 [1:00:08<30:32,  2.02it/s]
 65%|██████▌   | 6981/10682 [1:00:08<30:33,  2.02it/s]
 65%|██████▌   | 6982/10682 [1:00:09<30:30,  2.02it/s]
 65%|██████▌   | 6983/10682 [1:00:09<30:34,  2.02it/s]
 65%|██████▌   | 6984/10682 [1:00:10<30:30,  2.02it/s]
 65%|██████▌   | 6985/10682 [1:00:10<30:34,  2.02it/s]
 65%|██████▌   | 6986/10682 [1:00:11<30:35,  2.01it/s]
 65%|██████▌   | 6987/10682 [1:00:11<30:33,  2.02it/s]
 65%|██████▌   | 6988/10682 [1:00:12<30:32,  2.02it/s]
 65%|██████▌   | 6989/10682 [1:00:12<30:28,  2.02it/s]
 65%|██████▌   | 6990/10682 [1:00:13<30:29,  2.02it/s]
 65%|██████▌   | 6991/10682 [1:00:13<30:26,  2.02it/s]
 65%|██████▌   | 6992/10682 [1:00:14<30:26,  2.02it/s]
 65%|██████▌   | 6993/10682 [1:00:14<30:25,  2.02it/s]
 65%|██████▌   | 6994/10682 [1:00:15<30:28,  2.02it/s]
 65%|██████▌   | 6995/10682 [1:00:15<30:27,  2.02it/s]
 65%|██████▌   | 6996/10682 [1:00:16<30:30,  2.01it/s]
 66%|██████▌   | 6997/10682 [1:00:16<30:29,  2.01it/s]
 66%|██████▌   | 6998/10682 [1:00:17<30:26,  2.02it/s]
 66%|██████▌   | 6999/10682 [1:00:17<30:25,  2.02it/s]
 66%|██████▌   | 7000/10682 [1:00:18<30:22,  2.02it/s]{'loss': 3.4623, 'grad_norm': 0.23284398019313812, 'learning_rate': 0.0003203610024806651, 'epoch': 0.66}
                                                      
-
 66%|██████▌   | 7000/10682 [1:00:18<30:22,  2.02it/s]
 66%|██████▌   | 7001/10682 [1:00:18<30:29,  2.01it/s]
 66%|██████▌   | 7002/10682 [1:00:19<30:25,  2.02it/s]
 66%|██████▌   | 7003/10682 [1:00:19<30:25,  2.02it/s]
 66%|██████▌   | 7004/10682 [1:00:20<30:24,  2.02it/s]
 66%|██████▌   | 7005/10682 [1:00:20<30:23,  2.02it/s]
 66%|██████▌   | 7006/10682 [1:00:21<30:22,  2.02it/s]
 66%|██████▌   | 7007/10682 [1:00:21<30:21,  2.02it/s]
 66%|██████▌   | 7008/10682 [1:00:22<30:18,  2.02it/s]
 66%|██████▌   | 7009/10682 [1:00:22<30:18,  2.02it/s]
 66%|██████▌   | 7010/10682 [1:00:23<30:14,  2.02it/s]
 66%|██████▌   | 7011/10682 [1:00:23<30:19,  2.02it/s]
 66%|██████▌   | 7012/10682 [1:00:24<30:19,  2.02it/s]
 66%|██████▌   | 7013/10682 [1:00:24<30:16,  2.02it/s]
 66%|██████▌   | 7014/10682 [1:00:25<30:15,  2.02it/s]
 66%|██████▌   | 7015/10682 [1:00:25<30:15,  2.02it/s]
 66%|██████▌   | 7016/10682 [1:00:26<30:15,  2.02it/s]
 66%|██████▌   | 7017/10682 [1:00:26<30:14,  2.02it/s]
 66%|██████▌   | 7018/10682 [1:00:27<30:13,  2.02it/s]
 66%|██████▌   | 7019/10682 [1:00:27<30:10,  2.02it/s]
 66%|██████▌   | 7020/10682 [1:00:28<30:12,  2.02it/s]
 66%|██████▌   | 7021/10682 [1:00:28<30:11,  2.02it/s]
 66%|██████▌   | 7022/10682 [1:00:28<30:12,  2.02it/s]
 66%|██████▌   | 7023/10682 [1:00:29<30:09,  2.02it/s]
 66%|██████▌   | 7024/10682 [1:00:29<30:10,  2.02it/s]
 66%|██████▌   | 7025/10682 [1:00:30<30:07,  2.02it/s]
                                                      {'loss': 3.4643, 'grad_norm': 0.22449393570423126, 'learning_rate': 0.0003165547155522783, 'epoch': 0.66}
-
 66%|██████▌   | 7025/10682 [1:00:30<30:07,  2.02it/s]
 66%|██████▌   | 7026/10682 [1:00:30<30:12,  2.02it/s]
 66%|██████▌   | 7027/10682 [1:00:31<30:09,  2.02it/s]
 66%|██████▌   | 7028/10682 [1:00:31<30:09,  2.02it/s]
 66%|██████▌   | 7029/10682 [1:00:32<30:07,  2.02it/s]
 66%|██████▌   | 7030/10682 [1:00:32<30:07,  2.02it/s]
 66%|██████▌   | 7031/10682 [1:00:33<30:05,  2.02it/s]
 66%|██████▌   | 7032/10682 [1:00:33<30:09,  2.02it/s]
 66%|██████▌   | 7033/10682 [1:00:34<30:05,  2.02it/s]
 66%|██████▌   | 7034/10682 [1:00:34<30:05,  2.02it/s]
 66%|██████▌   | 7035/10682 [1:00:35<30:08,  2.02it/s]
 66%|██████▌   | 7036/10682 [1:00:35<30:06,  2.02it/s]
 66%|██████▌   | 7037/10682 [1:00:36<30:06,  2.02it/s]
 66%|██████▌   | 7038/10682 [1:00:36<30:02,  2.02it/s]
 66%|██████▌   | 7039/10682 [1:00:37<30:01,  2.02it/s]
 66%|██████▌   | 7040/10682 [1:00:37<30:02,  2.02it/s]
 66%|██████▌   | 7041/10682 [1:00:38<30:02,  2.02it/s]
 66%|██████▌   | 7042/10682 [1:00:38<30:02,  2.02it/s]
 66%|██████▌   | 7043/10682 [1:00:39<30:01,  2.02it/s]
 66%|██████▌   | 7044/10682 [1:00:39<30:01,  2.02it/s]
 66%|██████▌   | 7045/10682 [1:00:40<30:00,  2.02it/s]
 66%|██████▌   | 7046/10682 [1:00:40<30:00,  2.02it/s]
 66%|██████▌   | 7047/10682 [1:00:41<30:02,  2.02it/s]
 66%|██████▌   | 7048/10682 [1:00:41<30:00,  2.02it/s]
 66%|██████▌   | 7049/10682 [1:00:42<29:58,  2.02it/s]
 66%|██████▌   | 7050/10682 [1:00:42<29:58,  2.02it/s]
                                                      {'loss': 3.4617, 'grad_norm': 0.22690068185329437, 'learning_rate': 0.0003127606738276115, 'epoch': 0.66}
-
 66%|██████▌   | 7050/10682 [1:00:42<29:58,  2.02it/s]
 66%|██████▌   | 7051/10682 [1:00:43<30:00,  2.02it/s]
 66%|██████▌   | 7052/10682 [1:00:43<29:58,  2.02it/s]
 66%|██████▌   | 7053/10682 [1:00:44<29:57,  2.02it/s]
 66%|██████▌   | 7054/10682 [1:00:44<29:58,  2.02it/s]
 66%|██████▌   | 7055/10682 [1:00:45<29:56,  2.02it/s]
 66%|██████▌   | 7056/10682 [1:00:45<29:55,  2.02it/s]
 66%|██████▌   | 7057/10682 [1:00:46<29:55,  2.02it/s]
 66%|██████▌   | 7058/10682 [1:00:46<29:54,  2.02it/s]
 66%|██████▌   | 7059/10682 [1:00:47<29:54,  2.02it/s]
 66%|██████▌   | 7060/10682 [1:00:47<29:52,  2.02it/s]
 66%|██████▌   | 7061/10682 [1:00:48<29:52,  2.02it/s]
 66%|██████▌   | 7062/10682 [1:00:48<29:51,  2.02it/s]
 66%|██████▌   | 7063/10682 [1:00:49<29:51,  2.02it/s]
 66%|██████▌   | 7064/10682 [1:00:49<29:49,  2.02it/s]
 66%|██████▌   | 7065/10682 [1:00:50<29:49,  2.02it/s]
 66%|██████▌   | 7066/10682 [1:00:50<29:49,  2.02it/s]
 66%|██████▌   | 7067/10682 [1:00:51<29:49,  2.02it/s]
 66%|██████▌   | 7068/10682 [1:00:51<29:45,  2.02it/s]
 66%|██████▌   | 7069/10682 [1:00:52<29:47,  2.02it/s]
 66%|██████▌   | 7070/10682 [1:00:52<29:45,  2.02it/s]
 66%|██████▌   | 7071/10682 [1:00:53<29:47,  2.02it/s]
 66%|██████▌   | 7072/10682 [1:00:53<29:45,  2.02it/s]
 66%|██████▌   | 7073/10682 [1:00:54<29:46,  2.02it/s]
 66%|██████▌   | 7074/10682 [1:00:54<29:45,  2.02it/s]
 66%|██████▌   | 7075/10682 [1:00:55<29:46,  2.02it/s]
                                                      {'loss': 3.4576, 'grad_norm': 0.2203853577375412, 'learning_rate': 0.00030897913056372755, 'epoch': 0.66}
-
 66%|██████▌   | 7075/10682 [1:00:55<29:46,  2.02it/s]
 66%|██████▌   | 7076/10682 [1:00:55<29:48,  2.02it/s]
 66%|██████▋   | 7077/10682 [1:00:56<29:47,  2.02it/s]
 66%|██████▋   | 7078/10682 [1:00:56<29:41,  2.02it/s]
 66%|██████▋   | 7079/10682 [1:00:57<29:41,  2.02it/s]
 66%|██████▋   | 7080/10682 [1:00:57<29:42,  2.02it/s]
 66%|██████▋   | 7081/10682 [1:00:58<29:41,  2.02it/s]
 66%|██████▋   | 7082/10682 [1:00:58<29:40,  2.02it/s]
 66%|██████▋   | 7083/10682 [1:00:59<29:42,  2.02it/s]
 66%|██████▋   | 7084/10682 [1:00:59<29:40,  2.02it/s]
 66%|██████▋   | 7085/10682 [1:01:00<29:42,  2.02it/s]
 66%|██████▋   | 7086/10682 [1:01:00<29:42,  2.02it/s]
 66%|██████▋   | 7087/10682 [1:01:01<29:41,  2.02it/s]
 66%|██████▋   | 7088/10682 [1:01:01<29:39,  2.02it/s]
 66%|██████▋   | 7089/10682 [1:01:02<29:38,  2.02it/s]
 66%|██████▋   | 7090/10682 [1:01:02<29:36,  2.02it/s]
 66%|██████▋   | 7091/10682 [1:01:03<29:36,  2.02it/s]
 66%|██████▋   | 7092/10682 [1:01:03<29:34,  2.02it/s]
 66%|██████▋   | 7093/10682 [1:01:04<29:34,  2.02it/s]
 66%|██████▋   | 7094/10682 [1:01:04<29:33,  2.02it/s]
 66%|██████▋   | 7095/10682 [1:01:05<29:34,  2.02it/s]
 66%|██████▋   | 7096/10682 [1:01:05<29:34,  2.02it/s]
 66%|██████▋   | 7097/10682 [1:01:06<29:35,  2.02it/s]
 66%|██████▋   | 7098/10682 [1:01:06<29:33,  2.02it/s]
 66%|██████▋   | 7099/10682 [1:01:07<29:31,  2.02it/s]
 66%|██████▋   | 7100/10682 [1:01:07<29:32,  2.02it/s]
                                                      
{'loss': 3.4601, 'grad_norm': 0.2171563059091568, 'learning_rate': 0.00030521033818340054, 'epoch': 0.66}
-
 66%|██████▋   | 7100/10682 [1:01:07<29:32,  2.02it/s]
 66%|██████▋   | 7101/10682 [1:01:08<29:35,  2.02it/s]
 66%|██████▋   | 7102/10682 [1:01:08<29:35,  2.02it/s]
 66%|██████▋   | 7103/10682 [1:01:09<29:33,  2.02it/s]
 67%|██████▋   | 7104/10682 [1:01:09<29:34,  2.02it/s]
 67%|██████▋   | 7105/10682 [1:01:10<29:33,  2.02it/s]
 67%|██████▋   | 7106/10682 [1:01:10<29:32,  2.02it/s]
 67%|██████▋   | 7107/10682 [1:01:11<29:31,  2.02it/s]
 67%|██████▋   | 7108/10682 [1:01:11<29:30,  2.02it/s]
 67%|██████▋   | 7109/10682 [1:01:12<29:27,  2.02it/s]
 67%|██████▋   | 7110/10682 [1:01:12<29:29,  2.02it/s]
 67%|██████▋   | 7111/10682 [1:01:13<29:28,  2.02it/s]
 67%|██████▋   | 7112/10682 [1:01:13<29:29,  2.02it/s]
 67%|██████▋   | 7113/10682 [1:01:14<29:27,  2.02it/s]
 67%|██████▋   | 7114/10682 [1:01:14<29:30,  2.02it/s]
 67%|██████▋   | 7115/10682 [1:01:15<29:30,  2.01it/s]
 67%|██████▋   | 7116/10682 [1:01:15<29:28,  2.02it/s]
 67%|██████▋   | 7117/10682 [1:01:16<29:30,  2.01it/s]
 67%|██████▋   | 7118/10682 [1:01:16<29:28,  2.02it/s]
 67%|██████▋   | 7119/10682 [1:01:17<29:28,  2.01it/s]
 67%|██████▋   | 7120/10682 [1:01:17<29:26,  2.02it/s]
 67%|██████▋   | 7121/10682 [1:01:18<29:24,  2.02it/s]
 67%|██████▋   | 7122/10682 [1:01:18<29:23,  2.02it/s]
 67%|██████▋   | 7123/10682 [1:01:19<29:22,  2.02it/s]
 67%|██████▋   | 7124/10682 [1:01:19<29:20,  2.02it/s]
 67%|██████▋   | 7125/10682 [1:01:20<29:20,  2.02it/s]
                                                      
{'loss': 3.448, 'grad_norm': 0.2165171056985855, 'learning_rate': 0.0003014545482582672, 'epoch': 0.67}
-
 67%|██████▋   | 7125/10682 [1:01:20<29:20,  2.02it/s]
 67%|██████▋   | 7126/10682 [1:01:20<29:25,  2.01it/s]
 67%|██████▋   | 7127/10682 [1:01:20<29:23,  2.02it/s]
 67%|██████▋   | 7128/10682 [1:01:21<29:24,  2.01it/s]
 67%|██████▋   | 7129/10682 [1:01:21<29:21,  2.02it/s]
 67%|██████▋   | 7130/10682 [1:01:22<29:22,  2.01it/s]
 67%|██████▋   | 7131/10682 [1:01:22<29:23,  2.01it/s]
 67%|██████▋   | 7132/10682 [1:01:23<29:22,  2.01it/s]
 67%|██████▋   | 7133/10682 [1:01:23<29:20,  2.02it/s]
 67%|██████▋   | 7134/10682 [1:01:24<29:18,  2.02it/s]
 67%|██████▋   | 7135/10682 [1:01:24<29:19,  2.02it/s]
 67%|██████▋   | 7136/10682 [1:01:25<29:18,  2.02it/s]
 67%|██████▋   | 7137/10682 [1:01:25<29:19,  2.01it/s]
 67%|██████▋   | 7138/10682 [1:01:26<29:17,  2.02it/s]
 67%|██████▋   | 7139/10682 [1:01:26<29:14,  2.02it/s]
 67%|██████▋   | 7140/10682 [1:01:27<29:13,  2.02it/s]
 67%|██████▋   | 7141/10682 [1:01:27<29:11,  2.02it/s]
 67%|██████▋   | 7142/10682 [1:01:28<29:12,  2.02it/s]
 67%|██████▋   | 7143/10682 [1:01:28<29:13,  2.02it/s]
 67%|██████▋   | 7144/10682 [1:01:29<29:14,  2.02it/s]
 67%|██████▋   | 7145/10682 [1:01:29<29:13,  2.02it/s]
 67%|██████▋   | 7146/10682 [1:01:30<29:11,  2.02it/s]
 67%|██████▋   | 7147/10682 [1:01:30<29:11,  2.02it/s]
 67%|██████▋   | 7148/10682 [1:01:31<29:08,  2.02it/s]
 67%|██████▋   | 7149/10682 [1:01:31<29:07,  2.02it/s]
 67%|██████▋   | 7150/10682 [1:01:32<29:07,  2.02it/s]
                                                      {'loss': 3.4591, 'grad_norm': 0.22385798394680023, 'learning_rate': 0.0002977120114920343, 'epoch': 0.67}
-
 67%|██████▋   | 7150/10682 [1:01:32<29:07,  2.02it/s]
 67%|██████▋   | 7151/10682 [1:01:32<29:11,  2.02it/s]
 67%|██████▋   | 7152/10682 [1:01:33<29:10,  2.02it/s]
 67%|██████▋   | 7153/10682 [1:01:33<29:10,  2.02it/s]
 67%|██████▋   | 7154/10682 [1:01:34<29:07,  2.02it/s]
 67%|██████▋   | 7155/10682 [1:01:34<29:09,  2.02it/s]
 67%|██████▋   | 7156/10682 [1:01:35<29:06,  2.02it/s]
 67%|██████▋   | 7157/10682 [1:01:35<29:07,  2.02it/s]
 67%|██████▋   | 7158/10682 [1:01:36<29:04,  2.02it/s]
 67%|██████▋   | 7159/10682 [1:01:36<29:05,  2.02it/s]
 67%|██████▋   | 7160/10682 [1:01:37<29:02,  2.02it/s]
 67%|██████▋   | 7161/10682 [1:01:37<29:03,  2.02it/s]
 67%|██████▋   | 7162/10682 [1:01:38<29:02,  2.02it/s]
 67%|██████▋   | 7163/10682 [1:01:38<29:03,  2.02it/s]
 67%|██████▋   | 7164/10682 [1:01:39<29:01,  2.02it/s]
 67%|██████▋   | 7165/10682 [1:01:39<29:03,  2.02it/s]
 67%|██████▋   | 7166/10682 [1:01:40<29:02,  2.02it/s]
 67%|██████▋   | 7167/10682 [1:01:40<29:00,  2.02it/s]
 67%|██████▋   | 7168/10682 [1:01:41<28:58,  2.02it/s]
 67%|██████▋   | 7169/10682 [1:01:41<28:59,  2.02it/s]
 67%|██████▋   | 7170/10682 [1:01:42<29:02,  2.02it/s]
 67%|██████▋   | 7171/10682 [1:01:42<28:59,  2.02it/s]
 67%|██████▋   | 7172/10682 [1:01:43<28:59,  2.02it/s]
 67%|██████▋   | 7173/10682 [1:01:43<28:56,  2.02it/s]
 67%|██████▋   | 7174/10682 [1:01:44<28:57,  2.02it/s]
 67%|██████▋   | 7175/10682 [1:01:44<28:56,  2.02it/s]
                                                      {'loss': 3.4639, 'grad_norm': 0.22371283173561096, 'learning_rate': 0.00029398297770374336, 'epoch': 0.67}
-
 67%|██████▋   | 7175/10682 [1:01:44<28:56,  2.02it/s]
 67%|██████▋   | 7176/10682 [1:01:45<28:58,  2.02it/s]
 67%|██████▋   | 7177/10682 [1:01:45<28:57,  2.02it/s]
 67%|██████▋   | 7178/10682 [1:01:46<28:56,  2.02it/s]
 67%|██████▋   | 7179/10682 [1:01:46<28:56,  2.02it/s]
 67%|██████▋   | 7180/10682 [1:01:47<28:54,  2.02it/s]
 67%|██████▋   | 7181/10682 [1:01:47<28:52,  2.02it/s]
 67%|██████▋   | 7182/10682 [1:01:48<28:52,  2.02it/s]
 67%|██████▋   | 7183/10682 [1:01:48<28:51,  2.02it/s]
 67%|██████▋   | 7184/10682 [1:01:49<28:53,  2.02it/s]
 67%|██████▋   | 7185/10682 [1:01:49<28:52,  2.02it/s]
 67%|██████▋   | 7186/10682 [1:01:50<28:51,  2.02it/s]
 67%|██████▋   | 7187/10682 [1:01:50<28:53,  2.02it/s]
 67%|██████▋   | 7188/10682 [1:01:51<28:51,  2.02it/s]
 67%|██████▋   | 7189/10682 [1:01:51<28:52,  2.02it/s]
 67%|██████▋   | 7190/10682 [1:01:52<28:50,  2.02it/s]
 67%|██████▋   | 7191/10682 [1:01:52<28:49,  2.02it/s]
 67%|██████▋   | 7192/10682 [1:01:53<28:50,  2.02it/s]
 67%|██████▋   | 7193/10682 [1:01:53<28:50,  2.02it/s]
 67%|██████▋   | 7194/10682 [1:01:54<28:52,  2.01it/s]
 67%|██████▋   | 7195/10682 [1:01:54<28:50,  2.02it/s]
 67%|██████▋   | 7196/10682 [1:01:55<28:49,  2.02it/s]
 67%|██████▋   | 7197/10682 [1:01:55<28:47,  2.02it/s]
 67%|██████▋   | 7198/10682 [1:01:56<28:46,  2.02it/s]
 67%|██████▋   | 7199/10682 [1:01:56<28:42,  2.02it/s]
 67%|██████▋   | 7200/10682 [1:01:57<28:43,  2.02it/s]
                                                      {'loss': 3.448, 'grad_norm': 0.21843619644641876, 'learning_rate': 0.0002902676958110948, 'epoch': 0.67}
-
 67%|██████▋   | 7200/10682 [1:01:57<28:43,  2.02it/s]
 67%|██████▋   | 7201/10682 [1:01:57<28:47,  2.02it/s]
 67%|██████▋   | 7202/10682 [1:01:58<28:45,  2.02it/s]
 67%|██████▋   | 7203/10682 [1:01:58<28:45,  2.02it/s]
 67%|██████▋   | 7204/10682 [1:01:59<28:44,  2.02it/s]
 67%|██████▋   | 7205/10682 [1:01:59<28:45,  2.02it/s]
 67%|██████▋   | 7206/10682 [1:02:00<28:43,  2.02it/s]
 67%|██████▋   | 7207/10682 [1:02:00<28:44,  2.02it/s]
 67%|██████▋   | 7208/10682 [1:02:01<28:41,  2.02it/s]
 67%|██████▋   | 7209/10682 [1:02:01<28:39,  2.02it/s]
 67%|██████▋   | 7210/10682 [1:02:02<28:37,  2.02it/s]
 68%|██████▊   | 7211/10682 [1:02:02<28:39,  2.02it/s]
 68%|██████▊   | 7212/10682 [1:02:03<28:39,  2.02it/s]
 68%|██████▊   | 7213/10682 [1:02:03<28:37,  2.02it/s]
 68%|██████▊   | 7214/10682 [1:02:04<28:37,  2.02it/s]
 68%|██████▊   | 7215/10682 [1:02:04<28:36,  2.02it/s]
 68%|██████▊   | 7216/10682 [1:02:05<28:37,  2.02it/s]
 68%|██████▊   | 7217/10682 [1:02:05<28:36,  2.02it/s]
 68%|██████▊   | 7218/10682 [1:02:06<28:35,  2.02it/s]
 68%|██████▊   | 7219/10682 [1:02:06<28:34,  2.02it/s]
 68%|██████▊   | 7220/10682 [1:02:07<28:35,  2.02it/s]
 68%|██████▊   | 7221/10682 [1:02:07<28:32,  2.02it/s]
 68%|██████▊   | 7222/10682 [1:02:08<28:35,  2.02it/s]
 68%|██████▊   | 7223/10682 [1:02:08<28:33,  2.02it/s]
 68%|██████▊   | 7224/10682 [1:02:09<28:34,  2.02it/s]
 68%|██████▊   | 7225/10682 [1:02:09<28:34,  2.02it/s]
                                                      
{'loss': 3.4479, 'grad_norm': 0.21397678554058075, 'learning_rate': 0.0002865664138138335, 'epoch': 0.68}
-
 68%|██████▊   | 7225/10682 [1:02:09<28:34,  2.02it/s]
 68%|██████▊   | 7226/10682 [1:02:10<28:36,  2.01it/s]
 68%|██████▊   | 7227/10682 [1:02:10<28:31,  2.02it/s]
 68%|██████▊   | 7228/10682 [1:02:11<28:30,  2.02it/s]
 68%|██████▊   | 7229/10682 [1:02:11<28:26,  2.02it/s]
 68%|██████▊   | 7230/10682 [1:02:12<28:26,  2.02it/s]
 68%|██████▊   | 7231/10682 [1:02:12<28:26,  2.02it/s]
 68%|██████▊   | 7232/10682 [1:02:13<28:27,  2.02it/s]
 68%|██████▊   | 7233/10682 [1:02:13<28:26,  2.02it/s]
 68%|██████▊   | 7234/10682 [1:02:14<28:29,  2.02it/s]
 68%|██████▊   | 7235/10682 [1:02:14<28:30,  2.02it/s]
 68%|██████▊   | 7236/10682 [1:02:15<28:30,  2.01it/s]
 68%|██████▊   | 7237/10682 [1:02:15<28:28,  2.02it/s]
 68%|██████▊   | 7238/10682 [1:02:15<28:28,  2.02it/s]
 68%|██████▊   | 7239/10682 [1:02:16<28:24,  2.02it/s]
 68%|██████▊   | 7240/10682 [1:02:16<28:24,  2.02it/s]
 68%|██████▊   | 7241/10682 [1:02:17<28:24,  2.02it/s]
 68%|██████▊   | 7242/10682 [1:02:17<28:28,  2.01it/s]
 68%|██████▊   | 7243/10682 [1:02:18<28:27,  2.01it/s]
 68%|██████▊   | 7244/10682 [1:02:18<28:27,  2.01it/s]
 68%|██████▊   | 7245/10682 [1:02:19<28:30,  2.01it/s]
 68%|██████▊   | 7246/10682 [1:02:19<28:25,  2.01it/s]
 68%|██████▊   | 7247/10682 [1:02:20<28:27,  2.01it/s]
 68%|██████▊   | 7248/10682 [1:02:20<28:22,  2.02it/s]
 68%|██████▊   | 7249/10682 [1:02:21<28:22,  2.02it/s]
 68%|██████▊   | 7250/10682 [1:02:21<28:21,  2.02it/s]{'loss': 3.455, 'grad_norm': 0.23159068822860718, 'learning_rate': 0.0002828793787771929, 'epoch': 0.68}
                                                      
-
 68%|██████▊   | 7250/10682 [1:02:21<28:21,  2.02it/s]
 68%|██████▊   | 7251/10682 [1:02:22<28:24,  2.01it/s]
 68%|██████▊   | 7252/10682 [1:02:22<28:25,  2.01it/s]
 68%|██████▊   | 7253/10682 [1:02:23<28:21,  2.02it/s]
 68%|██████▊   | 7254/10682 [1:02:23<28:24,  2.01it/s]
 68%|██████▊   | 7255/10682 [1:02:24<28:24,  2.01it/s]
 68%|██████▊   | 7256/10682 [1:02:24<28:22,  2.01it/s]
 68%|██████▊   | 7257/10682 [1:02:25<28:21,  2.01it/s]
 68%|██████▊   | 7258/10682 [1:02:25<28:19,  2.02it/s]
 68%|██████▊   | 7259/10682 [1:02:26<28:19,  2.01it/s]
 68%|██████▊   | 7260/10682 [1:02:26<28:16,  2.02it/s]
 68%|██████▊   | 7261/10682 [1:02:27<28:15,  2.02it/s]
 68%|██████▊   | 7262/10682 [1:02:27<28:13,  2.02it/s]
 68%|██████▊   | 7263/10682 [1:02:28<28:14,  2.02it/s]
 68%|██████▊   | 7264/10682 [1:02:28<28:16,  2.02it/s]
 68%|██████▊   | 7265/10682 [1:02:29<28:15,  2.02it/s]
 68%|██████▊   | 7266/10682 [1:02:29<28:16,  2.01it/s]
 68%|██████▊   | 7267/10682 [1:02:30<28:14,  2.02it/s]
 68%|██████▊   | 7268/10682 [1:02:30<28:11,  2.02it/s]
 68%|██████▊   | 7269/10682 [1:02:31<28:09,  2.02it/s]
 68%|██████▊   | 7270/10682 [1:02:31<28:10,  2.02it/s]
 68%|██████▊   | 7271/10682 [1:02:32<28:07,  2.02it/s]
 68%|██████▊   | 7272/10682 [1:02:32<28:07,  2.02it/s]
 68%|██████▊   | 7273/10682 [1:02:33<28:08,  2.02it/s]
 68%|██████▊   | 7274/10682 [1:02:33<28:11,  2.01it/s]
 68%|██████▊   | 7275/10682 [1:02:34<28:12,  2.01it/s]
                                                      {'loss': 3.4582, 'grad_norm': 0.23988227546215057, 'learning_rate': 0.0002792068368154043, 'epoch': 0.68}
-
 68%|██████▊   | 7275/10682 [1:02:34<28:12,  2.01it/s]
 68%|██████▊   | 7276/10682 [1:02:34<28:13,  2.01it/s]
 68%|██████▊   | 7277/10682 [1:02:35<28:11,  2.01it/s]
 68%|██████▊   | 7278/10682 [1:02:35<28:10,  2.01it/s]
 68%|██████▊   | 7279/10682 [1:02:36<28:10,  2.01it/s]
 68%|██████▊   | 7280/10682 [1:02:36<28:09,  2.01it/s]
 68%|██████▊   | 7281/10682 [1:02:37<28:05,  2.02it/s]
 68%|██████▊   | 7282/10682 [1:02:37<28:05,  2.02it/s]
 68%|██████▊   | 7283/10682 [1:02:38<28:03,  2.02it/s]
 68%|██████▊   | 7284/10682 [1:02:38<28:04,  2.02it/s]
 68%|██████▊   | 7285/10682 [1:02:39<28:05,  2.02it/s]
 68%|█��████▊   | 7286/10682 [1:02:39<28:02,  2.02it/s]
 68%|██████▊   | 7287/10682 [1:02:40<28:01,  2.02it/s]
 68%|██████▊   | 7288/10682 [1:02:40<28:02,  2.02it/s]
 68%|██████▊   | 7289/10682 [1:02:41<28:02,  2.02it/s]
 68%|██████▊   | 7290/10682 [1:02:41<28:02,  2.02it/s]
 68%|██████▊   | 7291/10682 [1:02:42<27:59,  2.02it/s]
 68%|██████▊   | 7292/10682 [1:02:42<28:00,  2.02it/s]
 68%|██████▊   | 7293/10682 [1:02:43<27:58,  2.02it/s]
 68%|██████▊   | 7294/10682 [1:02:43<28:02,  2.01it/s]
 68%|██████▊   | 7295/10682 [1:02:44<28:00,  2.02it/s]
 68%|██████▊   | 7296/10682 [1:02:44<28:00,  2.02it/s]
 68%|██████▊   | 7297/10682 [1:02:45<27:57,  2.02it/s]
 68%|██████▊   | 7298/10682 [1:02:45<27:58,  2.02it/s]
 68%|██████▊   | 7299/10682 [1:02:46<27:54,  2.02it/s]
 68%|██████▊   | 7300/10682 [1:02:46<27:56,  2.02it/s]
                                                      {'loss': 3.4543, 'grad_norm': 0.22589965164661407, 'learning_rate': 0.0002755490330752679, 'epoch': 0.68}
-
 68%|██████▊   | 7300/10682 [1:02:46<27:56,  2.02it/s]
 68%|██████▊   | 7301/10682 [1:02:47<27:56,  2.02it/s]
 68%|██████▊   | 7302/10682 [1:02:47<27:56,  2.02it/s]
 68%|██████▊   | 7303/10682 [1:02:48<27:55,  2.02it/s]
 68%|██████▊   | 7304/10682 [1:02:48<27:55,  2.02it/s]
 68%|██████▊   | 7305/10682 [1:02:49<27:55,  2.02it/s]
 68%|██████▊   | 7306/10682 [1:02:49<27:52,  2.02it/s]
 68%|██████▊   | 7307/10682 [1:02:50<27:53,  2.02it/s]
 68%|██████▊   | 7308/10682 [1:02:50<27:52,  2.02it/s]
 68%|██████▊   | 7309/10682 [1:02:51<27:52,  2.02it/s]
 68%|██████▊   | 7310/10682 [1:02:51<27:50,  2.02it/s]
 68%|██████▊   | 7311/10682 [1:02:52<27:49,  2.02it/s]
 68%|██████▊   | 7312/10682 [1:02:52<27:49,  2.02it/s]
 68%|██████▊   | 7313/10682 [1:02:53<27:49,  2.02it/s]
 68%|██████▊   | 7314/10682 [1:02:53<27:51,  2.01it/s]
 68%|██████▊   | 7315/10682 [1:02:54<27:49,  2.02it/s]
 68%|██████▊   | 7316/10682 [1:02:54<27:51,  2.01it/s]
 68%|██████▊   | 7317/10682 [1:02:55<27:53,  2.01it/s]
 69%|██████▊   | 7318/10682 [1:02:55<27:52,  2.01it/s]
 69%|██████▊   | 7319/10682 [1:02:56<27:51,  2.01it/s]
 69%|██████▊   | 7320/10682 [1:02:56<27:49,  2.01it/s]
 69%|██████▊   | 7321/10682 [1:02:57<27:50,  2.01it/s]
 69%|██████▊   | 7322/10682 [1:02:57<27:48,  2.01it/s]
 69%|██████▊   | 7323/10682 [1:02:58<27:47,  2.01it/s]
 69%|██████▊   | 7324/10682 [1:02:58<27:52,  2.01it/s]
 69%|██████▊   | 7325/10682 [1:02:59<27:50,  2.01it/s]
                                                      
{'loss': 3.4452, 'grad_norm': 0.24658824503421783, 'learning_rate': 0.0002719062117197888, 'epoch': 0.69}
-
 69%|██████▊   | 7325/10682 [1:02:59<27:50,  2.01it/s]
 69%|██████▊   | 7326/10682 [1:02:59<27:52,  2.01it/s]
 69%|██████▊   | 7327/10682 [1:03:00<27:50,  2.01it/s]
 69%|██████▊   | 7328/10682 [1:03:00<27:45,  2.01it/s]
 69%|██████▊   | 7329/10682 [1:03:01<27:44,  2.01it/s]
 69%|██████▊   | 7330/10682 [1:03:01<27:43,  2.01it/s]
 69%|██████▊   | 7331/10682 [1:03:02<27:42,  2.02it/s]
 69%|██████▊   | 7332/10682 [1:03:02<27:46,  2.01it/s]
 69%|██████▊   | 7333/10682 [1:03:03<27:42,  2.01it/s]
 69%|██████▊   | 7334/10682 [1:03:03<27:44,  2.01it/s]
 69%|██████▊   | 7335/10682 [1:03:04<27:42,  2.01it/s]
 69%|██████▊   | 7336/10682 [1:03:04<27:41,  2.01it/s]
 69%|██████▊   | 7337/10682 [1:03:05<27:38,  2.02it/s]
 69%|██████▊   | 7338/10682 [1:03:05<27:37,  2.02it/s]
 69%|██████▊   | 7339/10682 [1:03:06<27:34,  2.02it/s]
 69%|██████▊   | 7340/10682 [1:03:06<27:34,  2.02it/s]
 69%|██████▊   | 7341/10682 [1:03:07<27:33,  2.02it/s]
 69%|██████▊   | 7342/10682 [1:03:07<27:33,  2.02it/s]
 69%|██████▊   | 7343/10682 [1:03:08<27:30,  2.02it/s]
 69%|██████▉   | 7344/10682 [1:03:08<27:34,  2.02it/s]
 69%|██████▉   | 7345/10682 [1:03:09<27:32,  2.02it/s]
 69%|██████▉   | 7346/10682 [1:03:09<27:32,  2.02it/s]
 69%|██████▉   | 7347/10682 [1:03:10<27:30,  2.02it/s]
 69%|██████▉   | 7348/10682 [1:03:10<27:30,  2.02it/s]
 69%|██████▉   | 7349/10682 [1:03:11<27:32,  2.02it/s]
 69%|██████▉   | 7350/10682 [1:03:11<27:31,  2.02it/s]{'loss': 3.4488, 'grad_norm': 0.21183805167675018, 'learning_rate': 0.00026827861591187944, 'epoch': 0.69}
                                                      
-
 69%|██████▉   | 7350/10682 [1:03:11<27:31,  2.02it/s]
 69%|██████▉   | 7351/10682 [1:03:12<27:32,  2.02it/s]
 69%|██████▉   | 7352/10682 [1:03:12<27:30,  2.02it/s]
 69%|██████▉   | 7353/10682 [1:03:13<27:30,  2.02it/s]
 69%|██████▉   | 7354/10682 [1:03:13<27:32,  2.01it/s]
 69%|██████▉   | 7355/10682 [1:03:14<27:32,  2.01it/s]
 69%|██████▉   | 7356/10682 [1:03:14<27:29,  2.02it/s]
 69%|██████▉   | 7357/10682 [1:03:15<27:29,  2.02it/s]
 69%|██████▉   | 7358/10682 [1:03:15<27:29,  2.01it/s]
 69%|██████▉   | 7359/10682 [1:03:16<27:27,  2.02it/s]
 69%|██████▉   | 7360/10682 [1:03:16<27:26,  2.02it/s]
 69%|██████▉   | 7361/10682 [1:03:17<27:25,  2.02it/s]
 69%|██████▉   | 7362/10682 [1:03:17<27:25,  2.02it/s]
 69%|██████▉   | 7363/10682 [1:03:18<27:24,  2.02it/s]
 69%|██████▉   | 7364/10682 [1:03:18<27:25,  2.02it/s]
 69%|██████▉   | 7365/10682 [1:03:18<27:25,  2.02it/s]
 69%|██████▉   | 7366/10682 [1:03:19<27:23,  2.02it/s]
 69%|██████▉   | 7367/10682 [1:03:19<27:23,  2.02it/s]
 69%|██████▉   | 7368/10682 [1:03:20<27:22,  2.02it/s]
 69%|██████▉   | 7369/10682 [1:03:20<27:22,  2.02it/s]
 69%|██████▉   | 7370/10682 [1:03:21<27:22,  2.02it/s]
 69%|██████▉   | 7371/10682 [1:03:21<27:24,  2.01it/s]
 69%|██████▉   | 7372/10682 [1:03:22<27:23,  2.01it/s]
 69%|██████▉   | 7373/10682 [1:03:22<27:21,  2.02it/s]
 69%|██████▉   | 7374/10682 [1:03:23<27:23,  2.01it/s]
 69%|██████▉   | 7375/10682 [1:03:23<27:23,  2.01it/s]
                                                      
{'loss': 3.4474, 'grad_norm': 0.2247314304113388, 'learning_rate': 0.00026466648779812794, 'epoch': 0.69}
-
 69%|██████▉   | 7375/10682 [1:03:23<27:23,  2.01it/s]
 69%|██████▉   | 7376/10682 [1:03:24<27:22,  2.01it/s]
 69%|██████▉   | 7377/10682 [1:03:24<27:21,  2.01it/s]
 69%|██████▉   | 7378/10682 [1:03:25<27:20,  2.01it/s]
 69%|██████▉   | 7379/10682 [1:03:25<27:22,  2.01it/s]
 69%|██████▉   | 7380/10682 [1:03:26<27:19,  2.01it/s]
 69%|██████▉   | 7381/10682 [1:03:26<27:21,  2.01it/s]
 69%|██████▉   | 7382/10682 [1:03:27<27:21,  2.01it/s]
 69%|██████▉   | 7383/10682 [1:03:27<27:18,  2.01it/s]
 69%|██████▉   | 7384/10682 [1:03:28<27:18,  2.01it/s]
 69%|██████▉   | 7385/10682 [1:03:28<27:16,  2.01it/s]
 69%|██████▉   | 7386/10682 [1:03:29<27:16,  2.01it/s]
 69%|██████▉   | 7387/10682 [1:03:29<27:12,  2.02it/s]
 69%|██████▉   | 7388/10682 [1:03:30<27:14,  2.02it/s]
 69%|██████▉   | 7389/10682 [1:03:30<27:10,  2.02it/s]
 69%|██████▉   | 7390/10682 [1:03:31<27:10,  2.02it/s]
 69%|██████▉   | 7391/10682 [1:03:31<27:10,  2.02it/s]
 69%|██████▉   | 7392/10682 [1:03:32<27:09,  2.02it/s]
 69%|██████▉   | 7393/10682 [1:03:32<27:07,  2.02it/s]
 69%|██████▉   | 7394/10682 [1:03:33<27:12,  2.01it/s]
 69%|██████▉   | 7395/10682 [1:03:33<27:11,  2.01it/s]
 69%|██████▉   | 7396/10682 [1:03:34<27:10,  2.02it/s]
 69%|██████▉   | 7397/10682 [1:03:34<27:09,  2.02it/s]
 69%|██████▉   | 7398/10682 [1:03:35<27:08,  2.02it/s]
 69%|██████▉   | 7399/10682 [1:03:35<27:07,  2.02it/s]
 69%|██████▉   | 7400/10682 [1:03:36<27:04,  2.02it/s]
                                                      
{'loss': 3.4359, 'grad_norm': 0.2222198247909546, 'learning_rate': 0.0002610700684926339, 'epoch': 0.69}
-
 69%|██████▉   | 7400/10682 [1:03:36<27:04,  2.02it/s]
 69%|██████▉   | 7401/10682 [1:03:36<27:06,  2.02it/s]
 69%|██████▉   | 7402/10682 [1:03:37<27:05,  2.02it/s]
 69%|██████▉   | 7403/10682 [1:03:37<27:05,  2.02it/s]
 69%|██████▉   | 7404/10682 [1:03:38<27:06,  2.02it/s]
 69%|██████▉   | 7405/10682 [1:03:38<27:06,  2.01it/s]
 69%|██████▉   | 7406/10682 [1:03:39<27:05,  2.02it/s]
 69%|██████▉   | 7407/10682 [1:03:39<27:02,  2.02it/s]
 69%|██████▉   | 7408/10682 [1:03:40<27:02,  2.02it/s]
 69%|██████▉   | 7409/10682 [1:03:40<26:59,  2.02it/s]
 69%|██████▉   | 7410/10682 [1:03:41<26:58,  2.02it/s]
 69%|██████▉   | 7411/10682 [1:03:41<26:59,  2.02it/s]
 69%|██████▉   | 7412/10682 [1:03:42<26:59,  2.02it/s]
 69%|██████▉   | 7413/10682 [1:03:42<26:57,  2.02it/s]
 69%|██████▉   | 7414/10682 [1:03:43<27:01,  2.02it/s]
 69%|██████▉   | 7415/10682 [1:03:43<27:00,  2.02it/s]
 69%|██████▉   | 7416/10682 [1:03:44<27:00,  2.02it/s]
 69%|██████▉   | 7417/10682 [1:03:44<26:58,  2.02it/s]
 69%|██████▉   | 7418/10682 [1:03:45<26:59,  2.02it/s]
 69%|██████▉   | 7419/10682 [1:03:45<26:57,  2.02it/s]
 69%|██████▉   | 7420/10682 [1:03:46<26:55,  2.02it/s]
 69%|██████▉   | 7421/10682 [1:03:46<26:54,  2.02it/s]
 69%|██████▉   | 7422/10682 [1:03:47<26:55,  2.02it/s]
 69%|██████▉   | 7423/10682 [1:03:47<26:57,  2.02it/s]
 70%|██████▉   | 7424/10682 [1:03:48<26:55,  2.02it/s]
 70%|██████▉   | 7425/10682 [1:03:48<26:55,  2.02it/s]
                                                      
{'loss': 3.4397, 'grad_norm': 0.22158439457416534, 'learning_rate': 0.00025748959806091423, 'epoch': 0.7}
-
 70%|██████▉   | 7425/10682 [1:03:48<26:55,  2.02it/s]
 70%|██████▉   | 7426/10682 [1:03:49<26:57,  2.01it/s]
 70%|██████▉   | 7427/10682 [1:03:49<26:56,  2.01it/s]
 70%|██████▉   | 7428/10682 [1:03:50<26:54,  2.01it/s]
 70%|██████▉   | 7429/10682 [1:03:50<26:54,  2.01it/s]
 70%|██████▉   | 7430/10682 [1:03:51<26:54,  2.01it/s]
 70%|██████▉   | 7431/10682 [1:03:51<26:52,  2.02it/s]
 70%|██████▉   | 7432/10682 [1:03:52<26:54,  2.01it/s]
 70%|██████▉   | 7433/10682 [1:03:52<26:52,  2.02it/s]
 70%|██████▉   | 7434/10682 [1:03:53<26:52,  2.01it/s]
 70%|██████▉   | 7435/10682 [1:03:53<26:50,  2.02it/s]
 70%|██████▉   | 7436/10682 [1:03:54<26:51,  2.01it/s]
 70%|██████▉   | 7437/10682 [1:03:54<26:49,  2.02it/s]
 70%|██████▉   | 7438/10682 [1:03:55<26:47,  2.02it/s]
 70%|██████▉   | 7439/10682 [1:03:55<26:47,  2.02it/s]
 70%|██████▉   | 7440/10682 [1:03:56<26:48,  2.02it/s]
 70%|██████▉   | 7441/10682 [1:03:56<26:46,  2.02it/s]
 70%|██████▉   | 7442/10682 [1:03:57<26:48,  2.01it/s]
 70%|██████▉   | 7443/10682 [1:03:57<26:47,  2.01it/s]
 70%|██████▉   | 7444/10682 [1:03:58<26:45,  2.02it/s]
 70%|██████▉   | 7445/10682 [1:03:58<26:44,  2.02it/s]
 70%|██████▉   | 7446/10682 [1:03:59<26:44,  2.02it/s]
 70%|██████▉   | 7447/10682 [1:03:59<26:44,  2.02it/s]
 70%|██████▉   | 7448/10682 [1:04:00<26:41,  2.02it/s]
 70%|██████▉   | 7449/10682 [1:04:00<26:42,  2.02it/s]
 70%|██████▉   | 7450/10682 [1:04:01<26:41,  2.02it/s]{'loss': 3.4354, 'grad_norm': 0.2178167849779129, 'learning_rate': 0.00025392531550387867, 'epoch': 0.7}
                                                      
-
 70%|██████▉   | 7450/10682 [1:04:01<26:41,  2.02it/s]
 70%|██████▉   | 7451/10682 [1:04:01<26:42,  2.02it/s]
 70%|██████▉   | 7452/10682 [1:04:02<26:41,  2.02it/s]
 70%|██████▉   | 7453/10682 [1:04:02<26:40,  2.02it/s]
 70%|██████▉   | 7454/10682 [1:04:03<26:41,  2.02it/s]
 70%|██████▉   | 7455/10682 [1:04:03<26:41,  2.01it/s]
 70%|██████▉   | 7456/10682 [1:04:04<26:40,  2.02it/s]
 70%|██████▉   | 7457/10682 [1:04:04<26:42,  2.01it/s]
 70%|██████▉   | 7458/10682 [1:04:05<26:41,  2.01it/s]
 70%|██████▉   | 7459/10682 [1:04:05<26:40,  2.01it/s]
 70%|██████▉   | 7460/10682 [1:04:06<26:39,  2.01it/s]
 70%|██████▉   | 7461/10682 [1:04:06<26:38,  2.01it/s]
 70%|██████▉   | 7462/10682 [1:04:07<26:37,  2.02it/s]
 70%|██████▉   | 7463/10682 [1:04:07<26:36,  2.02it/s]
 70%|██████▉   | 7464/10682 [1:04:08<26:36,  2.02it/s]
 70%|██████▉   | 7465/10682 [1:04:08<26:34,  2.02it/s]
 70%|██████▉   | 7466/10682 [1:04:09<26:34,  2.02it/s]
 70%|██████▉   | 7467/10682 [1:04:09<26:34,  2.02it/s]
 70%|██████▉   | 7468/10682 [1:04:10<26:34,  2.02it/s]
 70%|██████▉   | 7469/10682 [1:04:10<26:32,  2.02it/s]
 70%|██████▉   | 7470/10682 [1:04:11<26:32,  2.02it/s]
 70%|██████▉   | 7471/10682 [1:04:11<26:30,  2.02it/s]
 70%|██████▉   | 7472/10682 [1:04:12<30:54,  1.73it/s]
 70%|██████▉   | 7473/10682 [1:04:12<29:37,  1.81it/s]
 70%|██████▉   | 7474/10682 [1:04:13<28:42,  1.86it/s]
 70%|██████▉   | 7475/10682 [1:04:13<28:01,  1.91it/s]
                                                      {'loss': 3.4481, 'grad_norm': 0.2123980075120926, 'learning_rate': 0.0002503774587418758, 'epoch': 0.7}
-
 70%|██████▉   | 7475/10682 [1:04:13<28:01,  1.91it/s]
 70%|██████▉   | 7476/10682 [1:04:14<27:36,  1.94it/s]
 70%|██████▉   | 7477/10682 [1:04:14<27:15,  1.96it/s]
 70%|███████   | 7478/10682 [1:04:15<27:01,  1.98it/s]
 70%|███████   | 7479/10682 [1:04:15<26:51,  1.99it/s]
 70%|███████   | 7480/10682 [1:04:16<26:44,  2.00it/s]
 70%|███████   | 7481/10682 [1:04:16<26:39,  2.00it/s]
 70%|███████   | 7482/10682 [1:04:17<26:34,  2.01it/s]
 70%|███████   | 7483/10682 [1:04:17<26:33,  2.01it/s]
 70%|███████   | 7484/10682 [1:04:18<26:31,  2.01it/s]
 70%|███████   | 7485/10682 [1:04:18<26:28,  2.01it/s]
 70%|███████   | 7486/10682 [1:04:19<26:25,  2.02it/s]
 70%|███████   | 7487/10682 [1:04:19<26:26,  2.01it/s]
 70%|███████   | 7488/10682 [1:04:20<26:23,  2.02it/s]
 70%|███████   | 7489/10682 [1:04:20<26:23,  2.02it/s]
 70%|███████   | 7490/10682 [1:04:21<26:20,  2.02it/s]
 70%|███████   | 7491/10682 [1:04:21<26:21,  2.02it/s]
 70%|███████   | 7492/10682 [1:04:22<26:20,  2.02it/s]
 70%|███████   | 7493/10682 [1:04:22<26:20,  2.02it/s]
 70%|███████   | 7494/10682 [1:04:23<26:23,  2.01it/s]
 70%|███████   | 7495/10682 [1:04:23<26:20,  2.02it/s]
 70%|███████   | 7496/10682 [1:04:24<26:20,  2.02it/s]
 70%|███████   | 7497/10682 [1:04:24<26:21,  2.01it/s]
 70%|███████   | 7498/10682 [1:04:25<26:20,  2.01it/s]
 70%|███████   | 7499/10682 [1:04:25<26:16,  2.02it/s]
 70%|███████   | 7500/10682 [1:04:26<26:17,  2.02it/s]
                                                      {'loss': 3.4318, 'grad_norm': 0.2146175354719162, 'learning_rate': 0.00024684626459881156, 'epoch': 0.7}
-
 70%|███████   | 7500/10682 [1:04:26<26:17,  2.02it/s]
 70%|███████   | 7501/10682 [1:04:26<26:17,  2.02it/s]
 70%|███████   | 7502/10682 [1:04:27<26:19,  2.01it/s]
 70%|███████   | 7503/10682 [1:04:27<30:38,  1.73it/s]
 70%|███████   | 7504/10682 [1:04:28<29:22,  1.80it/s]
 70%|███████   | 7505/10682 [1:04:28<28:25,  1.86it/s]
 70%|███████   | 7506/10682 [1:04:29<27:48,  1.90it/s]
 70%|███████   | 7507/10682 [1:04:29<27:16,  1.94it/s]
 70%|███████   | 7508/10682 [1:04:30<26:57,  1.96it/s]
 70%|███████   | 7509/10682 [1:04:30<26:43,  1.98it/s]
 70%|███████   | 7510/10682 [1:04:31<26:33,  1.99it/s]
 70%|███████   | 7511/10682 [1:04:31<26:27,  2.00it/s]
 70%|███████   | 7512/10682 [1:04:32<26:22,  2.00it/s]
 70%|███████   | 7513/10682 [1:04:32<26:18,  2.01it/s]
 70%|███████   | 7514/10682 [1:04:33<26:17,  2.01it/s]
 70%|███████   | 7515/10682 [1:04:33<26:16,  2.01it/s]
 70%|███████   | 7516/10682 [1:04:34<26:14,  2.01it/s]
 70%|███████   | 7517/10682 [1:04:34<26:14,  2.01it/s]
 70%|███████   | 7518/10682 [1:04:35<26:15,  2.01it/s]
 70%|███████   | 7519/10682 [1:04:35<26:12,  2.01it/s]
 70%|███████   | 7520/10682 [1:04:36<26:09,  2.01it/s]
 70%|███████   | 7521/10682 [1:04:36<26:07,  2.02it/s]
 70%|███████   | 7522/10682 [1:04:37<26:08,  2.02it/s]
 70%|███████   | 7523/10682 [1:04:37<26:06,  2.02it/s]
 70%|███████   | 7524/10682 [1:04:38<26:07,  2.01it/s]
 70%|███████   | 7525/10682 [1:04:38<26:04,  2.02it/s]{'loss': 3.4432, 'grad_norm': 0.21388407051563263, 'learning_rate': 0.00024333196878634163, 'epoch': 0.7}
                                                      
-
 70%|███████   | 7525/10682 [1:04:38<26:04,  2.02it/s]
 70%|███████   | 7526/10682 [1:04:39<26:11,  2.01it/s]
 70%|███████   | 7527/10682 [1:04:39<26:08,  2.01it/s]
 70%|███████   | 7528/10682 [1:04:40<26:05,  2.02it/s]
 70%|███████   | 7529/10682 [1:04:40<26:03,  2.02it/s]
 70%|███████   | 7530/10682 [1:04:41<26:01,  2.02it/s]
 71%|███████   | 7531/10682 [1:04:41<25:59,  2.02it/s]
 71%|███████   | 7532/10682 [1:04:42<26:00,  2.02it/s]
 71%|███████   | 7533/10682 [1:04:42<26:01,  2.02it/s]
 71%|███████   | 7534/10682 [1:04:43<26:00,  2.02it/s]
 71%|███████   | 7535/10682 [1:04:43<26:01,  2.01it/s]
 71%|███████   | 7536/10682 [1:04:44<25:58,  2.02it/s]
 71%|███████   | 7537/10682 [1:04:44<25:59,  2.02it/s]
 71%|███████   | 7538/10682 [1:04:45<25:55,  2.02it/s]
 71%|███████   | 7539/10682 [1:04:45<25:56,  2.02it/s]
 71%|███████   | 7540/10682 [1:04:46<25:55,  2.02it/s]
 71%|███████   | 7541/10682 [1:04:46<25:56,  2.02it/s]
 71%|███████   | 7542/10682 [1:04:47<25:55,  2.02it/s]
 71%|███████   | 7543/10682 [1:04:47<25:55,  2.02it/s]
 71%|███████   | 7544/10682 [1:04:48<25:57,  2.01it/s]
 71%|███████   | 7545/10682 [1:04:48<25:55,  2.02it/s]
 71%|███████   | 7546/10682 [1:04:49<25:55,  2.02it/s]
 71%|███████   | 7547/10682 [1:04:49<25:54,  2.02it/s]
 71%|███████   | 7548/10682 [1:04:50<25:53,  2.02it/s]
 71%|███████   | 7549/10682 [1:04:50<25:52,  2.02it/s]
 71%|███████   | 7550/10682 [1:04:51<25:51,  2.02it/s]
                                                      
{'loss': 3.4476, 'grad_norm': 0.21357175707817078, 'learning_rate': 0.00023983480588813622, 'epoch': 0.71}
-
 71%|███████   | 7550/10682 [1:04:51<25:51,  2.02it/s]
 71%|███████   | 7551/10682 [1:04:51<25:53,  2.02it/s]
 71%|███████   | 7552/10682 [1:04:52<25:53,  2.01it/s]
 71%|███████   | 7553/10682 [1:04:52<25:54,  2.01it/s]
 71%|███████   | 7554/10682 [1:04:53<25:55,  2.01it/s]
 71%|███████   | 7555/10682 [1:04:53<25:53,  2.01it/s]
 71%|███████   | 7556/10682 [1:04:54<25:53,  2.01it/s]
 71%|███████   | 7557/10682 [1:04:54<25:50,  2.02it/s]
 71%|███████   | 7558/10682 [1:04:55<25:51,  2.01it/s]
 71%|███████   | 7559/10682 [1:04:55<25:54,  2.01it/s]
 71%|███████   | 7560/10682 [1:04:56<25:51,  2.01it/s]
 71%|███████   | 7561/10682 [1:04:56<25:50,  2.01it/s]
 71%|███████   | 7562/10682 [1:04:57<25:48,  2.02it/s]
 71%|███████   | 7563/10682 [1:04:57<25:50,  2.01it/s]
 71%|███████   | 7564/10682 [1:04:58<25:47,  2.02it/s]
 71%|███████   | 7565/10682 [1:04:58<25:47,  2.01it/s]
 71%|███████   | 7566/10682 [1:04:59<25:46,  2.01it/s]
 71%|███████   | 7567/10682 [1:04:59<25:46,  2.01it/s]
 71%|███████   | 7568/10682 [1:05:00<25:44,  2.02it/s]
 71%|███████   | 7569/10682 [1:05:00<25:44,  2.02it/s]
 71%|███████   | 7570/10682 [1:05:01<25:44,  2.02it/s]
 71%|███████   | 7571/10682 [1:05:01<25:42,  2.02it/s]
 71%|███████   | 7572/10682 [1:05:02<25:42,  2.02it/s]
 71%|███████   | 7573/10682 [1:05:02<25:40,  2.02it/s]
 71%|███████   | 7574/10682 [1:05:03<25:43,  2.01it/s]
 71%|███████   | 7575/10682 [1:05:03<25:40,  2.02it/s]
                                                      
{'loss': 3.4339, 'grad_norm': 0.22650697827339172, 'learning_rate': 0.00023635500934422239, 'epoch': 0.71}
-
 71%|███████   | 7575/10682 [1:05:03<25:40,  2.02it/s]
 71%|███████   | 7576/10682 [1:05:04<25:43,  2.01it/s]
 71%|███████   | 7577/10682 [1:05:04<25:43,  2.01it/s]
 71%|███████   | 7578/10682 [1:05:05<25:40,  2.02it/s]
 71%|███████   | 7579/10682 [1:05:05<25:40,  2.01it/s]
 71%|███████   | 7580/10682 [1:05:06<25:38,  2.02it/s]
 71%|███████   | 7581/10682 [1:05:06<25:36,  2.02it/s]
 71%|███████   | 7582/10682 [1:05:07<25:35,  2.02it/s]
 71%|███████   | 7583/10682 [1:05:07<25:36,  2.02it/s]
 71%|███████   | 7584/10682 [1:05:08<25:35,  2.02it/s]
 71%|███████   | 7585/10682 [1:05:08<25:36,  2.02it/s]
 71%|███████   | 7586/10682 [1:05:09<25:36,  2.02it/s]
 71%|███████   | 7587/10682 [1:05:09<25:34,  2.02it/s]
 71%|███████   | 7588/10682 [1:05:10<25:33,  2.02it/s]
 71%|███████   | 7589/10682 [1:05:10<25:31,  2.02it/s]
 71%|███████   | 7590/10682 [1:05:11<25:32,  2.02it/s]
 71%|███████   | 7591/10682 [1:05:11<25:31,  2.02it/s]
 71%|███████   | 7592/10682 [1:05:12<25:32,  2.02it/s]
 71%|███████   | 7593/10682 [1:05:12<25:30,  2.02it/s]
 71%|███████   | 7594/10682 [1:05:13<25:32,  2.02it/s]
 71%|███████   | 7595/10682 [1:05:13<25:31,  2.02it/s]
 71%|███████   | 7596/10682 [1:05:14<25:29,  2.02it/s]
 71%|███████   | 7597/10682 [1:05:14<25:31,  2.01it/s]
 71%|███████   | 7598/10682 [1:05:15<25:29,  2.02it/s]
 71%|███████   | 7599/10682 [1:05:15<25:27,  2.02it/s]
 71%|███████   | 7600/10682 [1:05:16<25:25,  2.02it/s]
                                                      {'loss': 3.4435, 'grad_norm': 0.22191022336483002, 'learning_rate': 0.00023289281143540065, 'epoch': 0.71}
-

 71%|███████   | 7600/10682 [1:05:16<25:25,  2.02it/s]
 71%|███████   | 7601/10682 [1:05:16<25:29,  2.01it/s]
 71%|███████   | 7602/10682 [1:05:17<25:27,  2.02it/s]
 71%|███████   | 7603/10682 [1:05:17<25:28,  2.01it/s]
 71%|███████   | 7604/10682 [1:05:18<25:32,  2.01it/s]
 71%|███████   | 7605/10682 [1:05:18<25:29,  2.01it/s]
 71%|███████   | 7606/10682 [1:05:19<25:27,  2.01it/s]
 71%|███████   | 7607/10682 [1:05:19<25:25,  2.02it/s]
 71%|███████   | 7608/10682 [1:05:20<25:24,  2.02it/s]
 71%|███████   | 7609/10682 [1:05:20<25:24,  2.02it/s]
 71%|███████   | 7610/10682 [1:05:21<25:24,  2.02it/s]
 71%|███████▏  | 7611/10682 [1:05:21<25:24,  2.01it/s]
 71%|███████▏  | 7612/10682 [1:05:22<25:22,  2.02it/s]
 71%|███████��  | 7613/10682 [1:05:22<25:23,  2.01it/s]
 71%|███████▏  | 7614/10682 [1:05:23<25:20,  2.02it/s]
 71%|███████▏  | 7615/10682 [1:05:23<25:23,  2.01it/s]
 71%|███████▏  | 7616/10682 [1:05:24<25:22,  2.01it/s]
 71%|███████▏  | 7617/10682 [1:05:24<25:21,  2.01it/s]
 71%|███████▏  | 7618/10682 [1:05:25<25:22,  2.01it/s]
 71%|███████▏  | 7619/10682 [1:05:25<25:21,  2.01it/s]
 71%|███████▏  | 7620/10682 [1:05:26<25:20,  2.01it/s]
 71%|███████▏  | 7621/10682 [1:05:26<25:20,  2.01it/s]
 71%|███████▏  | 7622/10682 [1:05:27<25:20,  2.01it/s]
 71%|███████▏  | 7623/10682 [1:05:27<25:20,  2.01it/s]
 71%|███████▏  | 7624/10682 [1:05:28<25:19,  2.01it/s]
 71%|███████▏  | 7625/10682 [1:05:28<25:19,  2.01it/s]
                                                      {'loss': 3.4229, 'grad_norm': 0.21966253221035004, 'learning_rate': 0.00022944844326774121, 'epoch': 0.71}
-
 71%|███████▏  | 7625/10682 [1:05:28<25:19,  2.01it/s]
 71%|███████▏  | 7626/10682 [1:05:29<25:22,  2.01it/s]
 71%|███████▏  | 7627/10682 [1:05:29<25:19,  2.01it/s]
 71%|███████▏  | 7628/10682 [1:05:30<25:20,  2.01it/s]
 71%|███████▏  | 7629/10682 [1:05:30<25:19,  2.01it/s]
 71%|███████▏  | 7630/10682 [1:05:31<25:16,  2.01it/s]
 71%|███████▏  | 7631/10682 [1:05:31<25:16,  2.01it/s]
 71%|███████▏  | 7632/10682 [1:05:32<25:13,  2.02it/s]
 71%|███████▏  | 7633/10682 [1:05:32<25:13,  2.01it/s]
 71%|███████▏  | 7634/10682 [1:05:33<25:11,  2.02it/s]
 71%|███████▏  | 7635/10682 [1:05:33<25:12,  2.01it/s]
 71%|███████▏  | 7636/10682 [1:05:34<25:10,  2.02it/s]
 71%|███████▏  | 7637/10682 [1:05:34<25:11,  2.01it/s]
 72%|███████▏  | 7638/10682 [1:05:34<25:10,  2.02it/s]
 72%|███████▏  | 7639/10682 [1:05:35<25:09,  2.02it/s]
 72%|███████▏  | 7640/10682 [1:05:35<25:08,  2.02it/s]
 72%|███████▏  | 7641/10682 [1:05:36<25:09,  2.02it/s]
 72%|███████▏  | 7642/10682 [1:05:36<25:07,  2.02it/s]
 72%|███████▏  | 7643/10682 [1:05:37<25:09,  2.01it/s]
 72%|███████▏  | 7644/10682 [1:05:37<25:08,  2.01it/s]
 72%|███████▏  | 7645/10682 [1:05:38<25:07,  2.02it/s]
 72%|███████▏  | 7646/10682 [1:05:38<25:09,  2.01it/s]
 72%|███████▏  | 7647/10682 [1:05:39<25:06,  2.01it/s]
 72%|███████▏  | 7648/10682 [1:05:39<25:05,  2.02it/s]
 72%|███████▏  | 7649/10682 [1:05:40<25:03,  2.02it/s]
 72%|███████▏  | 7650/10682 [1:05:40<25:05,  2.01it/s]
                                                      
{'loss': 3.437, 'grad_norm': 0.21285399794578552, 'learning_rate': 0.00022602213475715589, 'epoch': 0.72}
-
 72%|███████▏  | 7650/10682 [1:05:40<25:05,  2.01it/s]
 72%|███████▏  | 7651/10682 [1:05:41<25:07,  2.01it/s]
 72%|███████▏  | 7652/10682 [1:05:41<25:07,  2.01it/s]
 72%|███████▏  | 7653/10682 [1:05:42<25:07,  2.01it/s]
 72%|███████▏  | 7654/10682 [1:05:42<25:06,  2.01it/s]
 72%|███████▏  | 7655/10682 [1:05:43<25:04,  2.01it/s]
 72%|███████▏  | 7656/10682 [1:05:43<25:05,  2.01it/s]
 72%|███████▏  | 7657/10682 [1:05:44<25:02,  2.01it/s]
 72%|███████▏  | 7658/10682 [1:05:44<25:01,  2.01it/s]
 72%|███████▏  | 7659/10682 [1:05:45<25:00,  2.01it/s]
 72%|███████▏  | 7660/10682 [1:05:45<25:00,  2.01it/s]
 72%|███████▏  | 7661/10682 [1:05:46<25:00,  2.01it/s]
 72%|███████▏  | 7662/10682 [1:05:46<24:57,  2.02it/s]
 72%|███████▏  | 7663/10682 [1:05:47<25:01,  2.01it/s]
 72%|███████▏  | 7664/10682 [1:05:47<24:59,  2.01it/s]
 72%|███████▏  | 7665/10682 [1:05:48<24:58,  2.01it/s]
 72%|███████▏  | 7666/10682 [1:05:48<24:56,  2.01it/s]
 72%|███████▏  | 7667/10682 [1:05:49<24:55,  2.02it/s]
 72%|███████▏  | 7668/10682 [1:05:49<24:54,  2.02it/s]
 72%|███████▏  | 7669/10682 [1:05:50<24:53,  2.02it/s]
 72%|███████▏  | 7670/10682 [1:05:50<24:55,  2.01it/s]
 72%|███████▏  | 7671/10682 [1:05:51<24:53,  2.02it/s]
 72%|███████▏  | 7672/10682 [1:05:51<24:53,  2.02it/s]
 72%|███████▏  | 7673/10682 [1:05:52<24:55,  2.01it/s]
 72%|███████▏  | 7674/10682 [1:05:52<24:53,  2.01it/s]
 72%|███████▏  | 7675/10682 [1:05:53<24:54,  2.01it/s]
                                                      
{'loss': 3.4172, 'grad_norm': 0.21401026844978333, 'learning_rate': 0.0002226141146140523, 'epoch': 0.72}
-
 72%|███████▏  | 7675/10682 [1:05:53<24:54,  2.01it/s]
 72%|███████▏  | 7676/10682 [1:05:53<24:54,  2.01it/s]
 72%|███████▏  | 7677/10682 [1:05:54<24:54,  2.01it/s]
 72%|███████▏  | 7678/10682 [1:05:54<24:54,  2.01it/s]
 72%|███████▏  | 7679/10682 [1:05:55<24:53,  2.01it/s]
 72%|███████▏  | 7680/10682 [1:05:55<24:51,  2.01it/s]
 72%|███████▏  | 7681/10682 [1:05:56<24:50,  2.01it/s]
 72%|███████▏  | 7682/10682 [1:05:56<24:50,  2.01it/s]
 72%|███████▏  | 7683/10682 [1:05:57<24:51,  2.01it/s]
 72%|███████▏  | 7684/10682 [1:05:57<24:52,  2.01it/s]
 72%|███████▏  | 7685/10682 [1:05:58<24:52,  2.01it/s]
 72%|███████▏  | 7686/10682 [1:05:58<24:50,  2.01it/s]
 72%|███████▏  | 7687/10682 [1:05:59<24:48,  2.01it/s]
 72%|███████▏  | 7688/10682 [1:05:59<24:48,  2.01it/s]
 72%|███████▏  | 7689/10682 [1:06:00<24:45,  2.01it/s]
 72%|███████▏  | 7690/10682 [1:06:00<24:44,  2.02it/s]
 72%|███████▏  | 7691/10682 [1:06:01<24:45,  2.01it/s]
 72%|███████▏  | 7692/10682 [1:06:01<24:42,  2.02it/s]
 72%|███████▏  | 7693/10682 [1:06:02<24:44,  2.01it/s]
 72%|███████▏  | 7694/10682 [1:06:02<24:42,  2.01it/s]
 72%|███████▏  | 7695/10682 [1:06:03<24:42,  2.01it/s]
 72%|███████▏  | 7696/10682 [1:06:03<24:42,  2.01it/s]
 72%|███████▏  | 7697/10682 [1:06:04<24:43,  2.01it/s]
 72%|███████▏  | 7698/10682 [1:06:04<24:40,  2.02it/s]
 72%|███████▏  | 7699/10682 [1:06:05<24:41,  2.01it/s]
 72%|███████▏  | 7700/10682 [1:06:05<24:40,  2.01it/s]
                                                      
{'loss': 3.432, 'grad_norm': 0.22154217958450317, 'learning_rate': 0.00021922461032806601, 'epoch': 0.72}
-
 72%|███████▏  | 7700/10682 [1:06:05<24:40,  2.01it/s]
 72%|███████▏  | 7701/10682 [1:06:06<24:42,  2.01it/s]
 72%|███████▏  | 7702/10682 [1:06:06<24:40,  2.01it/s]
 72%|███████▏  | 7703/10682 [1:06:07<24:40,  2.01it/s]
 72%|███████▏  | 7704/10682 [1:06:07<24:40,  2.01it/s]
 72%|███████▏  | 7705/10682 [1:06:08<24:38,  2.01it/s]
 72%|███████▏  | 7706/10682 [1:06:08<24:38,  2.01it/s]
 72%|███████▏  | 7707/10682 [1:06:09<24:36,  2.02it/s]
 72%|███████▏  | 7708/10682 [1:06:09<24:36,  2.01it/s]
 72%|███████▏  | 7709/10682 [1:06:10<24:36,  2.01it/s]
 72%|███████▏  | 7710/10682 [1:06:10<24:34,  2.02it/s]
 72%|███████▏  | 7711/10682 [1:06:11<24:35,  2.01it/s]
 72%|███████▏  | 7712/10682 [1:06:11<24:31,  2.02it/s]
 72%|███████▏  | 7713/10682 [1:06:12<24:33,  2.02it/s]
 72%|███████▏  | 7714/10682 [1:06:12<24:30,  2.02it/s]
 72%|███████▏  | 7715/10682 [1:06:13<24:32,  2.01it/s]
 72%|███████▏  | 7716/10682 [1:06:13<24:32,  2.01it/s]
 72%|███████▏  | 7717/10682 [1:06:14<24:31,  2.01it/s]
 72%|███████▏  | 7718/10682 [1:06:14<24:30,  2.02it/s]
 72%|███████▏  | 7719/10682 [1:06:15<24:29,  2.02it/s]
 72%|███████▏  | 7720/10682 [1:06:15<24:29,  2.02it/s]
 72%|███████▏  | 7721/10682 [1:06:16<24:30,  2.01it/s]
 72%|███████▏  | 7722/10682 [1:06:16<24:27,  2.02it/s]
 72%|███████▏  | 7723/10682 [1:06:17<24:27,  2.02it/s]
 72%|███████▏  | 7724/10682 [1:06:17<24:28,  2.01it/s]
 72%|███████▏  | 7725/10682 [1:06:18<24:27,  2.01it/s]{'loss': 3.4316, 'grad_norm': 0.2137116640806198, 'learning_rate': 0.0002158538481528759, 'epoch': 0.72}
                                                      
-
 72%|███████▏  | 7725/10682 [1:06:18<24:27,  2.01it/s]
 72%|███████▏  | 7726/10682 [1:06:18<24:28,  2.01it/s]
 72%|███████▏  | 7727/10682 [1:06:19<24:28,  2.01it/s]
 72%|███████▏  | 7728/10682 [1:06:19<24:26,  2.01it/s]
 72%|███████▏  | 7729/10682 [1:06:20<24:25,  2.02it/s]
 72%|███████▏  | 7730/10682 [1:06:20<24:24,  2.02it/s]
 72%|███████▏  | 7731/10682 [1:06:21<24:24,  2.02it/s]
 72%|███████▏  | 7732/10682 [1:06:21<24:24,  2.01it/s]
 72%|███████▏  | 7733/10682 [1:06:22<24:23,  2.01it/s]
 72%|███████▏  | 7734/10682 [1:06:22<24:23,  2.01it/s]
 72%|███████▏  | 7735/10682 [1:06:23<24:21,  2.02it/s]
 72%|███████▏  | 7736/10682 [1:06:23<24:21,  2.02it/s]
 72%|███████▏  | 7737/10682 [1:06:24<24:21,  2.01it/s]
 72%|███████▏  | 7738/10682 [1:06:24<24:19,  2.02it/s]
 72%|███████▏  | 7739/10682 [1:06:25<24:19,  2.02it/s]
 72%|███████▏  | 7740/10682 [1:06:25<24:19,  2.02it/s]
 72%|█��█████▏  | 7741/10682 [1:06:26<24:21,  2.01it/s]
 72%|███████▏  | 7742/10682 [1:06:26<24:23,  2.01it/s]
 72%|███████▏  | 7743/10682 [1:06:27<24:24,  2.01it/s]
 72%|███████▏  | 7744/10682 [1:06:27<24:22,  2.01it/s]
 73%|███████▎  | 7745/10682 [1:06:28<24:20,  2.01it/s]
 73%|███████▎  | 7746/10682 [1:06:28<24:17,  2.01it/s]
 73%|███████▎  | 7747/10682 [1:06:29<24:17,  2.01it/s]
 73%|███████▎  | 7748/10682 [1:06:29<24:15,  2.02it/s]
 73%|███████▎  | 7749/10682 [1:06:30<24:15,  2.02it/s]
 73%|███████▎  | 7750/10682 [1:06:30<24:14,  2.02it/s]
                                                      
{'loss': 3.4305, 'grad_norm': 0.20959575474262238, 'learning_rate': 0.00021250205309110155, 'epoch': 0.73}
-
 73%|███████▎  | 7750/10682 [1:06:30<24:14,  2.02it/s]
 73%|███████▎  | 7751/10682 [1:06:31<24:17,  2.01it/s]
 73%|███████▎  | 7752/10682 [1:06:31<24:14,  2.01it/s]
 73%|███████▎  | 7753/10682 [1:06:32<24:11,  2.02it/s]
 73%|███████▎  | 7754/10682 [1:06:32<24:13,  2.01it/s]
 73%|███████▎  | 7755/10682 [1:06:33<24:11,  2.02it/s]
 73%|███████▎  | 7756/10682 [1:06:33<24:12,  2.01it/s]
 73%|███████▎  | 7757/10682 [1:06:34<24:12,  2.01it/s]
 73%|███████▎  | 7758/10682 [1:06:34<24:10,  2.02it/s]
 73%|███████▎  | 7759/10682 [1:06:35<24:09,  2.02it/s]
 73%|███████▎  | 7760/10682 [1:06:35<24:08,  2.02it/s]
 73%|███████▎  | 7761/10682 [1:06:36<24:08,  2.02it/s]
 73%|███████▎  | 7762/10682 [1:06:36<24:06,  2.02it/s]
 73%|███████▎  | 7763/10682 [1:06:37<24:07,  2.02it/s]
 73%|███████▎  | 7764/10682 [1:06:37<24:05,  2.02it/s]
 73%|███████▎  | 7765/10682 [1:06:38<24:06,  2.02it/s]
 73%|███████▎  | 7766/10682 [1:06:38<24:03,  2.02it/s]
 73%|███████▎  | 7767/10682 [1:06:39<24:04,  2.02it/s]
 73%|███████▎  | 7768/10682 [1:06:39<24:01,  2.02it/s]
 73%|███████▎  | 7769/10682 [1:06:40<24:04,  2.02it/s]
 73%|███████▎  | 7770/10682 [1:06:40<24:00,  2.02it/s]
 73%|███████▎  | 7771/10682 [1:06:41<24:00,  2.02it/s]
 73%|███████▎  | 7772/10682 [1:06:41<24:00,  2.02it/s]
 73%|███████▎  | 7773/10682 [1:06:42<24:00,  2.02it/s]
 73%|███████▎  | 7774/10682 [1:06:42<23:59,  2.02it/s]
 73%|███████▎  | 7775/10682 [1:06:43<23:59,  2.02it/s]
                                                      
{'loss': 3.4242, 'grad_norm': 0.22165252268314362, 'learning_rate': 0.00020916944887928359, 'epoch': 0.73}
-
 73%|███████▎  | 7775/10682 [1:06:43<23:59,  2.02it/s]
 73%|███████▎  | 7776/10682 [1:06:43<24:02,  2.02it/s]
 73%|███████▎  | 7777/10682 [1:06:43<24:00,  2.02it/s]
 73%|███████▎  | 7778/10682 [1:06:44<24:02,  2.01it/s]
 73%|███████▎  | 7779/10682 [1:06:44<24:00,  2.02it/s]
 73%|███████▎  | 7780/10682 [1:06:45<23:59,  2.02it/s]
 73%|███████▎  | 7781/10682 [1:06:45<23:59,  2.02it/s]
 73%|███████▎  | 7782/10682 [1:06:46<24:00,  2.01it/s]
 73%|███████▎  | 7783/10682 [1:06:46<24:02,  2.01it/s]
 73%|███████▎  | 7784/10682 [1:06:47<23:59,  2.01it/s]
 73%|███████▎  | 7785/10682 [1:06:47<23:58,  2.01it/s]
 73%|███████▎  | 7786/10682 [1:06:48<23:56,  2.02it/s]
 73%|███████▎  | 7787/10682 [1:06:48<23:56,  2.02it/s]
 73%|███████▎  | 7788/10682 [1:06:49<23:56,  2.01it/s]
 73%|███████▎  | 7789/10682 [1:06:49<23:56,  2.01it/s]
 73%|███████▎  | 7790/10682 [1:06:50<23:55,  2.01it/s]
 73%|███████▎  | 7791/10682 [1:06:50<23:55,  2.01it/s]
 73%|███████▎  | 7792/10682 [1:06:51<23:54,  2.02it/s]
 73%|███████▎  | 7793/10682 [1:06:51<23:54,  2.01it/s]
 73%|███████▎  | 7794/10682 [1:06:52<23:53,  2.01it/s]
 73%|███████▎  | 7795/10682 [1:06:52<23:52,  2.02it/s]
 73%|███████▎  | 7796/10682 [1:06:53<23:52,  2.02it/s]
 73%|███████▎  | 7797/10682 [1:06:53<23:51,  2.02it/s]
 73%|███████▎  | 7798/10682 [1:06:54<23:54,  2.01it/s]
 73%|███████▎  | 7799/10682 [1:06:54<23:52,  2.01it/s]
 73%|███████▎  | 7800/10682 [1:06:55<23:52,  2.01it/s]
                                                      {'loss': 3.4281, 'grad_norm': 0.21701517701148987, 'learning_rate': 0.00020585625797294927, 'epoch': 0.73}
-
 73%|███████▎  | 7800/10682 [1:06:55<23:52,  2.01it/s]
 73%|███████▎  | 7801/10682 [1:06:55<23:55,  2.01it/s]
 73%|███████▎  | 7802/10682 [1:06:56<23:52,  2.01it/s]
 73%|███████▎  | 7803/10682 [1:06:56<23:51,  2.01it/s]
 73%|███████▎  | 7804/10682 [1:06:57<23:49,  2.01it/s]
 73%|███████▎  | 7805/10682 [1:06:57<23:47,  2.02it/s]
 73%|███████▎  | 7806/10682 [1:06:58<23:45,  2.02it/s]
 73%|███████▎  | 7807/10682 [1:06:58<23:46,  2.02it/s]
 73%|███████▎  | 7808/10682 [1:06:59<23:47,  2.01it/s]
 73%|███████▎  | 7809/10682 [1:06:59<23:44,  2.02it/s]
 73%|███████▎  | 7810/10682 [1:07:00<23:44,  2.02it/s]
 73%|███████▎  | 7811/10682 [1:07:00<23:43,  2.02it/s]
 73%|███████▎  | 7812/10682 [1:07:01<23:42,  2.02it/s]
 73%|███████▎  | 7813/10682 [1:07:01<23:43,  2.01it/s]
 73%|███████▎  | 7814/10682 [1:07:02<23:42,  2.02it/s]
 73%|███████▎  | 7815/10682 [1:07:02<23:42,  2.02it/s]
 73%|███████▎  | 7816/10682 [1:07:03<23:40,  2.02it/s]
 73%|███████▎  | 7817/10682 [1:07:03<23:40,  2.02it/s]
 73%|███████▎  | 7818/10682 [1:07:04<23:38,  2.02it/s]
 73%|███████▎  | 7819/10682 [1:07:04<23:39,  2.02it/s]
 73%|███████▎  | 7820/10682 [1:07:05<23:36,  2.02it/s]
 73%|███████▎  | 7821/10682 [1:07:05<23:36,  2.02it/s]
 73%|███████▎  | 7822/10682 [1:07:06<23:36,  2.02it/s]
 73%|███████▎  | 7823/10682 [1:07:06<23:37,  2.02it/s]
 73%|███████▎  | 7824/10682 [1:07:07<23:35,  2.02it/s]
 73%|███████▎  | 7825/10682 [1:07:07<23:33,  2.02it/s]
                                                      
{'loss': 3.4305, 'grad_norm': 0.22064314782619476, 'learning_rate': 0.00020256270153176371, 'epoch': 0.73}
-
 73%|███████▎  | 7825/10682 [1:07:07<23:33,  2.02it/s]
 73%|███████▎  | 7826/10682 [1:07:08<23:38,  2.01it/s]
 73%|███████▎  | 7827/10682 [1:07:08<23:35,  2.02it/s]
 73%|███████▎  | 7828/10682 [1:07:09<23:36,  2.01it/s]
 73%|███████▎  | 7829/10682 [1:07:09<23:33,  2.02it/s]
 73%|███████▎  | 7830/10682 [1:07:10<23:33,  2.02it/s]
 73%|███████▎  | 7831/10682 [1:07:10<23:33,  2.02it/s]
 73%|███████▎  | 7832/10682 [1:07:11<23:33,  2.02it/s]
 73%|███████▎  | 7833/10682 [1:07:11<23:34,  2.01it/s]
 73%|███████▎  | 7834/10682 [1:07:12<23:34,  2.01it/s]
 73%|███████▎  | 7835/10682 [1:07:12<23:33,  2.01it/s]
 73%|███████▎  | 7836/10682 [1:07:13<23:31,  2.02it/s]
 73%|███████▎  | 7837/10682 [1:07:13<23:31,  2.02it/s]
 73%|███████▎  | 7838/10682 [1:07:14<23:30,  2.02it/s]
 73%|███████▎  | 7839/10682 [1:07:14<23:27,  2.02it/s]
 73%|███████▎  | 7840/10682 [1:07:15<23:28,  2.02it/s]
 73%|███████▎  | 7841/10682 [1:07:15<23:32,  2.01it/s]
 73%|███████▎  | 7842/10682 [1:07:16<23:30,  2.01it/s]
 73%|███████▎  | 7843/10682 [1:07:16<23:29,  2.01it/s]
 73%|███████▎  | 7844/10682 [1:07:17<23:27,  2.02it/s]
 73%|███████▎  | 7845/10682 [1:07:17<23:28,  2.01it/s]
 73%|███████▎  | 7846/10682 [1:07:18<23:25,  2.02it/s]
 73%|███████▎  | 7847/10682 [1:07:18<23:25,  2.02it/s]
 73%|███████▎  | 7848/10682 [1:07:19<23:26,  2.02it/s]
 73%|███████▎  | 7849/10682 [1:07:19<23:25,  2.02it/s]
 73%|███████▎  | 7850/10682 [1:07:20<23:24,  2.02it/s]
                                                      
{'loss': 3.4229, 'grad_norm': 0.21936888992786407, 'learning_rate': 0.00019928899940476624, 'epoch': 0.73}
-
 73%|███████▎  | 7850/10682 [1:07:20<23:24,  2.02it/s]
 73%|███████▎  | 7851/10682 [1:07:20<23:29,  2.01it/s]
 74%|███████▎  | 7852/10682 [1:07:21<23:27,  2.01it/s]
 74%|███████▎  | 7853/10682 [1:07:21<23:26,  2.01it/s]
 74%|███████▎  | 7854/10682 [1:07:22<23:25,  2.01it/s]
 74%|███████▎  | 7855/10682 [1:07:22<23:24,  2.01it/s]
 74%|███████▎  | 7856/10682 [1:07:23<23:23,  2.01it/s]
 74%|███████▎  | 7857/10682 [1:07:23<23:23,  2.01it/s]
 74%|███████▎  | 7858/10682 [1:07:24<23:22,  2.01it/s]
 74%|███████▎  | 7859/10682 [1:07:24<23:20,  2.02it/s]
 74%|███████▎  | 7860/10682 [1:07:25<23:19,  2.02it/s]
 74%|███████▎  | 7861/10682 [1:07:25<23:21,  2.01it/s]
 74%|███████▎  | 7862/10682 [1:07:26<23:20,  2.01it/s]
 74%|███████▎  | 7863/10682 [1:07:26<23:20,  2.01it/s]
 74%|███████▎  | 7864/10682 [1:07:27<23:17,  2.02it/s]
 74%|███████▎  | 7865/10682 [1:07:27<23:15,  2.02it/s]
 74%|███████▎  | 7866/10682 [1:07:28<23:15,  2.02it/s]
 74%|███████▎  | 7867/10682 [1:07:28<23:15,  2.02it/s]
 74%|███████▎  | 7868/10682 [1:07:29<23:13,  2.02it/s]
 74%|███████▎  | 7869/10682 [1:07:29<23:13,  2.02it/s]
 74%|███████▎  | 7870/10682 [1:07:30<23:13,  2.02it/s]
 74%|███████▎  | 7871/10682 [1:07:30<23:15,  2.01it/s]
 74%|███████▎  | 7872/10682 [1:07:31<23:14,  2.02it/s]
 74%|███████▎  | 7873/10682 [1:07:31<23:13,  2.02it/s]
 74%|███████▎  | 7874/10682 [1:07:32<23:11,  2.02it/s]
 74%|███████▎  | 7875/10682 [1:07:32<23:10,  2.02it/s]
                                                      
{'loss': 3.4183, 'grad_norm': 0.22054912149906158, 'learning_rate': 0.00019603537011569566, 'epoch': 0.74}
-
 74%|███████▎  | 7875/10682 [1:07:32<23:10,  2.02it/s]
 74%|███████▎  | 7876/10682 [1:07:33<23:11,  2.02it/s]
 74%|███████▎  | 7877/10682 [1:07:33<23:10,  2.02it/s]
 74%|███████▍  | 7878/10682 [1:07:34<23:09,  2.02it/s]
 74%|███████▍  | 7879/10682 [1:07:34<23:09,  2.02it/s]
 74%|███████▍  | 7880/10682 [1:07:35<23:09,  2.02it/s]
 74%|███████▍  | 7881/10682 [1:07:35<23:08,  2.02it/s]
 74%|███████▍  | 7882/10682 [1:07:36<23:12,  2.01it/s]
 74%|███████▍  | 7883/10682 [1:07:36<23:12,  2.01it/s]
 74%|███████▍  | 7884/10682 [1:07:37<23:10,  2.01it/s]
 74%|███████▍  | 7885/10682 [1:07:37<23:11,  2.01it/s]
 74%|███████▍  | 7886/10682 [1:07:38<23:10,  2.01it/s]
 74%|███████▍  | 7887/10682 [1:07:38<23:08,  2.01it/s]
 74%|███████▍  | 7888/10682 [1:07:39<23:07,  2.01it/s]
 74%|███████▍  | 7889/10682 [1:07:39<23:06,  2.01it/s]
 74%|███████▍  | 7890/10682 [1:07:40<23:05,  2.01it/s]
 74%|███████▍  | 7891/10682 [1:07:40<23:04,  2.02it/s]
 74%|███████▍  | 7892/10682 [1:07:41<23:04,  2.02it/s]
 74%|███████▍  | 7893/10682 [1:07:41<23:03,  2.02it/s]
 74%|███████▍  | 7894/10682 [1:07:42<23:04,  2.01it/s]
 74%|███████▍  | 7895/10682 [1:07:42<23:02,  2.02it/s]
 74%|███████▍  | 7896/10682 [1:07:43<23:03,  2.01it/s]
 74%|███████▍  | 7897/10682 [1:07:43<23:02,  2.01it/s]
 74%|███████▍  | 7898/10682 [1:07:44<22:59,  2.02it/s]
 74%|███████▍  | 7899/10682 [1:07:44<23:00,  2.02it/s]
 74%|███████▍  | 7900/10682 [1:07:45<22:58,  2.02it/s]
                                                      {'loss': 3.4272, 'grad_norm': 0.21948343515396118, 'learning_rate': 0.0001928020308484042, 'epoch': 0.74}
-
 74%|███████▍  | 7900/10682 [1:07:45<22:58,  2.02it/s]
 74%|███████▍  | 7901/10682 [1:07:45<23:00,  2.01it/s]
 74%|███████▍  | 7902/10682 [1:07:46<22:59,  2.02it/s]
 74%|███████▍  | 7903/10682 [1:07:46<22:59,  2.01it/s]
 74%|███████▍  | 7904/10682 [1:07:47<22:57,  2.02it/s]
 74%|███████▍  | 7905/10682 [1:07:47<22:58,  2.01it/s]
 74%|███████▍  | 7906/10682 [1:07:48<22:56,  2.02it/s]
 74%|███████▍  | 7907/10682 [1:07:48<22:55,  2.02it/s]
 74%|███████▍  | 7908/10682 [1:07:48<22:54,  2.02it/s]
 74%|███████▍  | 7909/10682 [1:07:49<22:53,  2.02it/s]
 74%|███████▍  | 7910/10682 [1:07:49<22:55,  2.02it/s]
 74%|███████▍  | 7911/10682 [1:07:50<22:54,  2.02it/s]
 74%|███████▍  | 7912/10682 [1:07:50<22:54,  2.02it/s]
 74%|███████▍  | 7913/10682 [1:07:51<22:53,  2.02it/s]
 74%|███████▍  | 7914/10682 [1:07:51<22:52,  2.02it/s]
 74%|███████▍  | 7915/10682 [1:07:52<22:50,  2.02it/s]
 74%|███████▍  | 7916/10682 [1:07:52<22:50,  2.02it/s]
 74%|███████▍  | 7917/10682 [1:07:53<22:48,  2.02it/s]
 74%|███████▍  | 7918/10682 [1:07:53<22:50,  2.02it/s]
 74%|███████▍  | 7919/10682 [1:07:54<22:49,  2.02it/s]
 74%|███████▍  | 7920/10682 [1:07:54<22:48,  2.02it/s]
 74%|███████▍  | 7921/10682 [1:07:55<22:49,  2.02it/s]
 74%|███████▍  | 7922/10682 [1:07:55<22:49,  2.02it/s]
 74%|███████▍  | 7923/10682 [1:07:56<22:50,  2.01it/s]
 74%|███████▍  | 7924/10682 [1:07:56<22:48,  2.02it/s]
 74%|███████▍  | 7925/10682 [1:07:57<22:48,  2.01it/s]{'loss': 3.4219, 'grad_norm': 0.22219917178153992, 'learning_rate': 0.00018958919743235897, 'epoch': 0.74}
                                                      
-
 74%|███████▍  | 7925/10682 [1:07:57<22:48,  2.01it/s]
 74%|███████▍  | 7926/10682 [1:07:57<22:47,  2.01it/s]
 74%|███████▍  | 7927/10682 [1:07:58<22:47,  2.01it/s]
 74%|███████▍  | 7928/10682 [1:07:58<22:44,  2.02it/s]
 74%|███████▍  | 7929/10682 [1:07:59<22:44,  2.02it/s]
 74%|███████▍  | 7930/10682 [1:07:59<22:43,  2.02it/s]
 74%|███████▍  | 7931/10682 [1:08:00<22:45,  2.01it/s]
 74%|███████▍  | 7932/10682 [1:08:00<22:44,  2.02it/s]
 74%|███████▍  | 7933/10682 [1:08:01<22:46,  2.01it/s]
 74%|███████▍  | 7934/10682 [1:08:01<22:44,  2.01it/s]
 74%|███████▍  | 7935/10682 [1:08:02<22:43,  2.01it/s]
 74%|███████▍  | 7936/10682 [1:08:02<22:42,  2.02it/s]
 74%|███████▍  | 7937/10682 [1:08:03<22:40,  2.02it/s]
 74%|███████▍  | 7938/10682 [1:08:03<22:39,  2.02it/s]
 74%|███████▍  | 7939/10682 [1:08:04<22:41,  2.01it/s]
 74%|███████▍  | 7940/10682 [1:08:04<22:39,  2.02it/s]
 74%|███████▍  | 7941/10682 [1:08:05<22:41,  2.01it/s]
 74%|███████▍  | 7942/10682 [1:08:05<22:39,  2.01it/s]
 74%|███████▍  | 7943/10682 [1:08:06<22:39,  2.02it/s]
 74%|███████▍  | 7944/10682 [1:08:06<22:37,  2.02it/s]
 74%|███████▍  | 7945/10682 [1:08:07<22:37,  2.02it/s]
 74%|███████▍  | 7946/10682 [1:08:07<22:36,  2.02it/s]
 74%|███████▍  | 7947/10682 [1:08:08<22:36,  2.02it/s]
 74%|███████▍  | 7948/10682 [1:08:08<22:33,  2.02it/s]
 74%|███████▍  | 7949/10682 [1:08:09<22:34,  2.02it/s]
 74%|███████▍  | 7950/10682 [1:08:09<22:34,  2.02it/s]{'loss': 3.4102, 'grad_norm': 0.22933071851730347, 'learning_rate': 0.0001863970843282357, 'epoch': 0.74}
                                                      
-
 74%|███████▍  | 7950/10682 [1:08:09<22:34,  2.02it/s]
 74%|███████▍  | 7951/10682 [1:08:10<22:35,  2.02it/s]
 74%|███████▍  | 7952/10682 [1:08:10<22:36,  2.01it/s]
 74%|███████▍  | 7953/10682 [1:08:11<22:35,  2.01it/s]
 74%|███████▍  | 7954/10682 [1:08:11<22:36,  2.01it/s]
 74%|███████▍  | 7955/10682 [1:08:12<22:35,  2.01it/s]
 74%|███████▍  | 7956/10682 [1:08:12<22:32,  2.02it/s]
 74%|███████▍  | 7957/10682 [1:08:13<22:32,  2.01it/s]
 74%|███████▍  | 7958/10682 [1:08:13<22:30,  2.02it/s]
 75%|███████▍  | 7959/10682 [1:08:14<22:29,  2.02it/s]
 75%|███████▍  | 7960/10682 [1:08:14<22:30,  2.02it/s]
 75%|███████▍  | 7961/10682 [1:08:15<22:29,  2.02it/s]
 75%|███████▍  | 7962/10682 [1:08:15<22:31,  2.01it/s]
 75%|███████▍  | 7963/10682 [1:08:16<22:30,  2.01it/s]
 75%|███████▍  | 7964/10682 [1:08:16<22:28,  2.02it/s]
 75%|███████▍  | 7965/10682 [1:08:17<22:28,  2.02it/s]
 75%|███████▍  | 7966/10682 [1:08:17<22:28,  2.01it/s]
 75%|███████▍  | 7967/10682 [1:08:18<22:26,  2.02it/s]
 75%|███████▍  | 7968/10682 [1:08:18<22:26,  2.02it/s]
 75%|███████▍  | 7969/10682 [1:08:19<22:25,  2.02it/s]
 75%|███████▍  | 7970/10682 [1:08:19<22:26,  2.01it/s]
 75%|███████▍  | 7971/10682 [1:08:20<22:26,  2.01it/s]
 75%|███████▍  | 7972/10682 [1:08:20<22:26,  2.01it/s]
 75%|███████▍  | 7973/10682 [1:08:21<22:25,  2.01it/s]
 75%|███████▍  | 7974/10682 [1:08:21<22:24,  2.01it/s]
 75%|███████▍  | 7975/10682 [1:08:22<22:23,  2.01it/s]
                                                      {'loss': 3.4166, 'grad_norm': 0.22176644206047058, 'learning_rate': 0.00018322590461360383, 'epoch': 0.75}
-
 75%|███████▍  | 7975/10682 [1:08:22<22:23,  2.01it/s]
 75%|███████▍  | 7976/10682 [1:08:22<22:27,  2.01it/s]
 75%|███████▍  | 7977/10682 [1:08:23<22:23,  2.01it/s]
 75%|███████▍  | 7978/10682 [1:08:23<22:22,  2.01it/s]
 75%|███████▍  | 7979/10682 [1:08:24<22:20,  2.02it/s]
 75%|███████▍  | 7980/10682 [1:08:24<22:20,  2.02it/s]
 75%|███████▍  | 7981/10682 [1:08:25<22:17,  2.02it/s]
 75%|███████▍  | 7982/10682 [1:08:25<22:20,  2.01it/s]
 75%|███████▍  | 7983/10682 [1:08:26<22:19,  2.01it/s]
 75%|███████▍  | 7984/10682 [1:08:26<22:18,  2.02it/s]
 75%|███████▍  | 7985/10682 [1:08:27<22:17,  2.02it/s]
 75%|███████▍  | 7986/10682 [1:08:27<22:16,  2.02it/s]
 75%|███████▍  | 7987/10682 [1:08:28<22:15,  2.02it/s]
 75%|███████▍  | 7988/10682 [1:08:28<22:14,  2.02it/s]
 75%|███████▍  | 7989/10682 [1:08:29<22:15,  2.02it/s]
 75%|███████▍  | 7990/10682 [1:08:29<22:13,  2.02it/s]
 75%|███████▍  | 7991/10682 [1:08:30<22:15,  2.01it/s]
 75%|███████▍  | 7992/10682 [1:08:30<22:14,  2.02it/s]
 75%|███████▍  | 7993/10682 [1:08:31<22:16,  2.01it/s]
 75%|███████▍  | 7994/10682 [1:08:31<22:17,  2.01it/s]
 75%|███████▍  | 7995/10682 [1:08:32<22:13,  2.01it/s]
 75%|███████▍  | 7996/10682 [1:08:32<22:12,  2.02it/s]
 75%|███████▍  | 7997/10682 [1:08:33<22:11,  2.02it/s]
 75%|███████▍  | 7998/10682 [1:08:33<22:13,  2.01it/s]
 75%|███████▍  | 7999/10682 [1:08:34<22:11,  2.02it/s]
 75%|███████▍  | 8000/10682 [1:08:34<22:11,  2.01it/s]
                                                      
{'loss': 3.4227, 'grad_norm': 0.2237834334373474, 'learning_rate': 0.00018007586996870206, 'epoch': 0.75}
-
 75%|███████▍  | 8000/10682 [1:08:34<22:11,  2.01it/s]
 75%|███████▍  | 8001/10682 [1:08:35<22:14,  2.01it/s]
 75%|███████▍  | 8002/10682 [1:08:35<22:12,  2.01it/s]
 75%|███████▍  | 8003/10682 [1:08:36<22:12,  2.01it/s]
 75%|███████▍  | 8004/10682 [1:08:36<22:09,  2.01it/s]
 75%|███████▍  | 8005/10682 [1:08:37<22:08,  2.02it/s]
 75%|███████▍  | 8006/10682 [1:08:37<22:08,  2.01it/s]
 75%|███████▍  | 8007/10682 [1:08:38<22:06,  2.02it/s]
 75%|███████▍  | 8008/10682 [1:08:38<22:05,  2.02it/s]
 75%|███████▍  | 8009/10682 [1:08:39<22:04,  2.02it/s]
 75%|███████▍  | 8010/10682 [1:08:39<22:03,  2.02it/s]
 75%|███████▍  | 8011/10682 [1:08:40<22:02,  2.02it/s]
 75%|███████▌  | 8012/10682 [1:08:40<22:04,  2.02it/s]
 75%|███████▌  | 8013/10682 [1:08:41<22:02,  2.02it/s]
 75%|███████▌  | 8014/10682 [1:08:41<22:02,  2.02it/s]
 75%|███████▌  | 8015/10682 [1:08:42<22:00,  2.02it/s]
 75%|███████▌  | 8016/10682 [1:08:42<22:00,  2.02it/s]
 75%|███████▌  | 8017/10682 [1:08:43<21:59,  2.02it/s]
 75%|███████▌  | 8018/10682 [1:08:43<21:58,  2.02it/s]
 75%|███████▌  | 8019/10682 [1:08:44<21:57,  2.02it/s]
 75%|███████▌  | 8020/10682 [1:08:44<21:59,  2.02it/s]
 75%|███████▌  | 8021/10682 [1:08:45<21:58,  2.02it/s]
 75%|███████▌  | 8022/10682 [1:08:45<21:58,  2.02it/s]
 75%|███████▌  | 8023/10682 [1:08:46<21:58,  2.02it/s]
 75%|███████▌  | 8024/10682 [1:08:46<21:58,  2.02it/s]
 75%|███████▌  | 8025/10682 [1:08:47<22:00,  2.01it/s]
                                                      
{'loss': 3.4217, 'grad_norm': 0.22052119672298431, 'learning_rate': 0.00017694719066230924, 'epoch': 0.75}
-
 75%|███████▌  | 8025/10682 [1:08:47<22:00,  2.01it/s]
 75%|███████▌  | 8026/10682 [1:08:47<22:02,  2.01it/s]
 75%|███████▌  | 8027/10682 [1:08:48<21:58,  2.01it/s]
 75%|███████▌  | 8028/10682 [1:08:48<21:58,  2.01it/s]
 75%|███████▌  | 8029/10682 [1:08:49<21:56,  2.02it/s]
 75%|███████▌  | 8030/10682 [1:08:49<21:57,  2.01it/s]
 75%|███████▌  | 8031/10682 [1:08:50<21:55,  2.01it/s]
 75%|███████▌  | 8032/10682 [1:08:50<21:55,  2.02it/s]
 75%|███████▌  | 8033/10682 [1:08:51<21:54,  2.02it/s]
 75%|███████▌  | 8034/10682 [1:08:51<21:53,  2.02it/s]
 75%|███████▌  | 8035/10682 [1:08:52<21:52,  2.02it/s]
 75%|███████▌  | 8036/10682 [1:08:52<21:50,  2.02it/s]
 75%|███████▌  | 8037/10682 [1:08:52<21:51,  2.02it/s]
 75%|███████▌  | 8038/10682 [1:08:53<21:51,  2.02it/s]
 75%|███████▌  | 8039/10682 [1:08:53<21:49,  2.02it/s]
 75%|███████▌  | 8040/10682 [1:08:54<21:48,  2.02it/s]
 75%|███████▌  | 8041/10682 [1:08:54<21:49,  2.02it/s]
 75%|███████▌  | 8042/10682 [1:08:55<21:47,  2.02it/s]
 75%|███████▌  | 8043/10682 [1:08:55<21:49,  2.02it/s]
 75%|███████▌  | 8044/10682 [1:08:56<21:47,  2.02it/s]
 75%|███████▌  | 8045/10682 [1:08:56<21:48,  2.02it/s]
 75%|███████▌  | 8046/10682 [1:08:57<21:45,  2.02it/s]
 75%|███████▌  | 8047/10682 [1:08:57<21:46,  2.02it/s]
 75%|███████▌  | 8048/10682 [1:08:58<21:44,  2.02it/s]
 75%|███████▌  | 8049/10682 [1:08:58<21:44,  2.02it/s]
 75%|███████▌  | 8050/10682 [1:08:59<21:43,  2.02it/s]
                                                      {'loss': 3.4116, 'grad_norm': 0.22158852219581604, 'learning_rate': 0.00017384007553770858, 'epoch': 0.75}
-
 75%|███████▌  | 8050/10682 [1:08:59<21:43,  2.02it/s]
 75%|███████▌  | 8051/10682 [1:08:59<21:44,  2.02it/s]
 75%|███████▌  | 8052/10682 [1:09:00<21:44,  2.02it/s]
 75%|███████▌  | 8053/10682 [1:09:00<21:44,  2.02it/s]
 75%|███████▌  | 8054/10682 [1:09:01<21:42,  2.02it/s]
 75%|███████▌  | 8055/10682 [1:09:01<21:43,  2.02it/s]
 75%|███████▌  | 8056/10682 [1:09:02<21:41,  2.02it/s]
 75%|███████▌  | 8057/10682 [1:09:02<21:40,  2.02it/s]
 75%|███████▌  | 8058/10682 [1:09:03<21:39,  2.02it/s]
 75%|███████▌  | 8059/10682 [1:09:03<21:38,  2.02it/s]
 75%|███████▌  | 8060/10682 [1:09:04<21:38,  2.02it/s]
 75%|███████▌  | 8061/10682 [1:09:04<21:38,  2.02it/s]
 75%|███████▌  | 8062/10682 [1:09:05<21:40,  2.02it/s]
 75%|███████▌  | 8063/10682 [1:09:05<21:38,  2.02it/s]
 75%|███████▌  | 8064/10682 [1:09:06<21:37,  2.02it/s]
 76%|███████▌  | 8065/10682 [1:09:06<21:35,  2.02it/s]
 76%|███████▌  | 8066/10682 [1:09:07<21:34,  2.02it/s]
 76%|███████▌  | 8067/10682 [1:09:07<21:33,  2.02it/s]
 76%|███████▌  | 8068/10682 [1:09:08<21:32,  2.02it/s]
 76%|███████▌  | 8069/10682 [1:09:08<21:30,  2.02it/s]
 76%|███████▌  | 8070/10682 [1:09:09<21:31,  2.02it/s]
 76%|███████▌  | 8071/10682 [1:09:09<21:30,  2.02it/s]
 76%|███████▌  | 8072/10682 [1:09:10<21:30,  2.02it/s]
 76%|███████▌  | 8073/10682 [1:09:10<21:31,  2.02it/s]
 76%|███████▌  | 8074/10682 [1:09:11<21:31,  2.02it/s]
 76%|███████▌  | 8075/10682 [1:09:11<21:29,  2.02it/s]{'loss': 3.3987, 'grad_norm': 0.2157202959060669, 'learning_rate': 0.00017075473199874692, 'epoch': 0.76}
-
                                                      

 76%|███████▌  | 8075/10682 [1:09:11<21:29,  2.02it/s]
 76%|███████▌  | 8076/10682 [1:09:12<21:33,  2.01it/s]
 76%|███████▌  | 8077/10682 [1:09:12<21:31,  2.02it/s]
 76%|███████▌  | 8078/10682 [1:09:13<21:31,  2.02it/s]
 76%|███████▌  | 8079/10682 [1:09:13<21:28,  2.02it/s]
 76%|███████▌  | 8080/10682 [1:09:14<21:29,  2.02it/s]
 76%|███████▌  | 8081/10682 [1:09:14<21:27,  2.02it/s]
 76%|███████▌  | 8082/10682 [1:09:15<21:27,  2.02it/s]
 76%|███████▌  | 8083/10682 [1:09:15<21:26,  2.02it/s]
 76%|███████▌  | 8084/10682 [1:09:16<21:26,  2.02it/s]
 76%|███████▌  | 8085/10682 [1:09:16<21:25,  2.02it/s]
 76%|███████▌  | 8086/10682 [1:09:17<21:24,  2.02it/s]
 76%|███████▌  | 8087/10682 [1:09:17<21:22,  2.02it/s]
 76%|███████▌  | 8088/10682 [1:09:18<21:23,  2.02it/s]
 76%|███████▌  | 8089/10682 [1:09:18<21:22,  2.02it/s]
 76%|███████▌  | 8090/10682 [1:09:19<21:24,  2.02it/s]
 76%|███████▌  | 8091/10682 [1:09:19<21:23,  2.02it/s]
 76%|███████▌  | 8092/10682 [1:09:20<21:25,  2.02it/s]
 76%|███████▌  | 8093/10682 [1:09:20<21:25,  2.01it/s]
 76%|███████▌  | 8094/10682 [1:09:21<21:23,  2.02it/s]
 76%|███████▌  | 8095/10682 [1:09:21<21:24,  2.01it/s]
 76%|███████▌  | 8096/10682 [1:09:22<21:23,  2.01it/s]
 76%|███████▌  | 8097/10682 [1:09:22<21:21,  2.02it/s]
 76%|███████▌  | 8098/10682 [1:09:23<21:21,  2.02it/s]
 76%|███████▌  | 8099/10682 [1:09:23<21:20,  2.02it/s]
 76%|███████▌  | 8100/10682 [1:09:24<21:20,  2.02it/s]{'loss': 3.4131, 'grad_norm': 0.21081994473934174, 'learning_rate': 0.00016769136599599017, 'epoch': 0.76}
-
                                                      

 76%|███████▌  | 8100/10682 [1:09:24<21:20,  2.02it/s]
 76%|███████▌  | 8101/10682 [1:09:24<21:21,  2.01it/s]
 76%|███████▌  | 8102/10682 [1:09:25<21:19,  2.02it/s]
 76%|███████▌  | 8103/10682 [1:09:25<21:20,  2.01it/s]
 76%|███████▌  | 8104/10682 [1:09:26<21:20,  2.01it/s]
 76%|███████▌  | 8105/10682 [1:09:26<21:19,  2.01it/s]
 76%|███████▌  | 8106/10682 [1:09:27<21:17,  2.02it/s]
 76%|███████▌  | 8107/10682 [1:09:27<21:17,  2.02it/s]
 76%|███████▌  | 8108/10682 [1:09:28<21:16,  2.02it/s]
 76%|███████▌  | 8109/10682 [1:09:28<21:16,  2.02it/s]
 76%|███████▌  | 8110/10682 [1:09:29<21:16,  2.02it/s]
 76%|███████▌  | 8111/10682 [1:09:29<21:16,  2.01it/s]
 76%|███████▌  | 8112/10682 [1:09:30<21:15,  2.01it/s]
 76%|███████▌  | 8113/10682 [1:09:30<21:13,  2.02it/s]
 76%|███████▌  | 8114/10682 [1:09:31<21:13,  2.02it/s]
 76%|███████▌  | 8115/10682 [1:09:31<21:13,  2.02it/s]
 76%|███████▌  | 8116/10682 [1:09:32<21:12,  2.02it/s]
 76%|███████▌  | 8117/10682 [1:09:32<21:10,  2.02it/s]
 76%|███████▌  | 8118/10682 [1:09:33<21:09,  2.02it/s]
 76%|███████▌  | 8119/10682 [1:09:33<21:09,  2.02it/s]
 76%|███████▌  | 8120/10682 [1:09:34<21:08,  2.02it/s]
 76%|███████▌  | 8121/10682 [1:09:34<21:10,  2.02it/s]
 76%|███████▌  | 8122/10682 [1:09:35<21:11,  2.01it/s]
 76%|███████▌  | 8123/10682 [1:09:35<21:08,  2.02it/s]
 76%|███��███▌  | 8124/10682 [1:09:36<21:09,  2.02it/s]
 76%|███████▌  | 8125/10682 [1:09:36<21:07,  2.02it/s]
                                                      
{'loss': 3.4091, 'grad_norm': 0.22090715169906616, 'learning_rate': 0.0001646501820129766, 'epoch': 0.76}
-
 76%|███████▌  | 8125/10682 [1:09:36<21:07,  2.02it/s]
 76%|███████▌  | 8126/10682 [1:09:37<21:08,  2.01it/s]
 76%|███████▌  | 8127/10682 [1:09:37<21:06,  2.02it/s]
 76%|███████▌  | 8128/10682 [1:09:38<21:06,  2.02it/s]
 76%|███████▌  | 8129/10682 [1:09:38<21:06,  2.02it/s]
 76%|███████▌  | 8130/10682 [1:09:39<21:06,  2.02it/s]
 76%|███████▌  | 8131/10682 [1:09:39<21:06,  2.01it/s]
 76%|███████▌  | 8132/10682 [1:09:40<21:04,  2.02it/s]
 76%|███████▌  | 8133/10682 [1:09:40<21:05,  2.01it/s]
 76%|███████▌  | 8134/10682 [1:09:41<21:04,  2.02it/s]
 76%|███████▌  | 8135/10682 [1:09:41<21:02,  2.02it/s]
 76%|███████▌  | 8136/10682 [1:09:42<20:59,  2.02it/s]
 76%|███████▌  | 8137/10682 [1:09:42<21:00,  2.02it/s]
 76%|███████▌  | 8138/10682 [1:09:43<20:57,  2.02it/s]
 76%|███████▌  | 8139/10682 [1:09:43<20:58,  2.02it/s]
 76%|███████▌  | 8140/10682 [1:09:44<20:57,  2.02it/s]
 76%|███████▌  | 8141/10682 [1:09:44<20:59,  2.02it/s]
 76%|███████▌  | 8142/10682 [1:09:45<20:59,  2.02it/s]
 76%|███████▌  | 8143/10682 [1:09:45<20:59,  2.02it/s]
 76%|███████▌  | 8144/10682 [1:09:46<20:58,  2.02it/s]
 76%|███████▌  | 8145/10682 [1:09:46<20:57,  2.02it/s]
 76%|███████▋  | 8146/10682 [1:09:47<20:58,  2.02it/s]
 76%|███████▋  | 8147/10682 [1:09:47<20:56,  2.02it/s]
 76%|███████▋  | 8148/10682 [1:09:48<20:56,  2.02it/s]
 76%|███████▋  | 8149/10682 [1:09:48<20:54,  2.02it/s]
 76%|███████▋  | 8150/10682 [1:09:49<20:55,  2.02it/s]{'loss': 3.4222, 'grad_norm': 0.2151668667793274, 'learning_rate': 0.00016163138305256598, 'epoch': 0.76}
                                                      
-
 76%|███████▋  | 8150/10682 [1:09:49<20:55,  2.02it/s]
 76%|███████▋  | 8151/10682 [1:09:49<20:55,  2.02it/s]
 76%|███████▋  | 8152/10682 [1:09:49<20:54,  2.02it/s]
 76%|███████▋  | 8153/10682 [1:09:50<20:53,  2.02it/s]
 76%|███████▋  | 8154/10682 [1:09:50<20:52,  2.02it/s]
 76%|███████▋  | 8155/10682 [1:09:51<20:51,  2.02it/s]
 76%|███████▋  | 8156/10682 [1:09:51<20:50,  2.02it/s]
 76%|███████▋  | 8157/10682 [1:09:52<20:50,  2.02it/s]
 76%|███████▋  | 8158/10682 [1:09:52<20:49,  2.02it/s]
 76%|███████▋  | 8159/10682 [1:09:53<20:48,  2.02it/s]
 76%|███████▋  | 8160/10682 [1:09:53<20:48,  2.02it/s]
 76%|███████▋  | 8161/10682 [1:09:54<20:48,  2.02it/s]
 76%|███████▋  | 8162/10682 [1:09:54<20:50,  2.01it/s]
 76%|███████▋  | 8163/10682 [1:09:55<20:51,  2.01it/s]
 76%|███████▋  | 8164/10682 [1:09:55<20:49,  2.02it/s]
 76%|███████▋  | 8165/10682 [1:09:56<20:47,  2.02it/s]
 76%|███████▋  | 8166/10682 [1:09:56<20:47,  2.02it/s]
 76%|███████▋  | 8167/10682 [1:09:57<20:46,  2.02it/s]
 76%|███████▋  | 8168/10682 [1:09:57<20:45,  2.02it/s]
 76%|███████▋  | 8169/10682 [1:09:58<20:45,  2.02it/s]
 76%|███████▋  | 8170/10682 [1:09:58<20:46,  2.01it/s]
 76%|███████▋  | 8171/10682 [1:09:59<20:45,  2.02it/s]
 77%|███████▋  | 8172/10682 [1:09:59<20:47,  2.01it/s]
 77%|███████▋  | 8173/10682 [1:10:00<20:47,  2.01it/s]
 77%|███████▋  | 8174/10682 [1:10:00<20:46,  2.01it/s]
 77%|███████▋  | 8175/10682 [1:10:01<20:46,  2.01it/s]
                                                      {'loss': 3.407, 'grad_norm': 0.22389106452465057, 'learning_rate': 0.00015863517062339038, 'epoch': 0.77}
-

 77%|███████▋  | 8175/10682 [1:10:01<20:46,  2.01it/s]
 77%|███████▋  | 8176/10682 [1:10:01<20:45,  2.01it/s]
 77%|███████▋  | 8177/10682 [1:10:02<20:43,  2.01it/s]
 77%|███████▋  | 8178/10682 [1:10:02<20:41,  2.02it/s]
 77%|███████▋  | 8179/10682 [1:10:03<20:40,  2.02it/s]
 77%|███████▋  | 8180/10682 [1:10:03<20:38,  2.02it/s]
 77%|███████▋  | 8181/10682 [1:10:04<20:39,  2.02it/s]
 77%|███████▋  | 8182/10682 [1:10:04<20:39,  2.02it/s]
 77%|███████▋  | 8183/10682 [1:10:05<20:39,  2.02it/s]
 77%|███████▋  | 8184/10682 [1:10:05<20:37,  2.02it/s]
 77%|███████▋  | 8185/10682 [1:10:06<20:39,  2.01it/s]
 77%|███████▋  | 8186/10682 [1:10:06<20:37,  2.02it/s]
 77%|███████▋  | 8187/10682 [1:10:07<23:57,  1.74it/s]
 77%|███████▋  | 8188/10682 [1:10:08<22:58,  1.81it/s]
 77%|███████▋  | 8189/10682 [1:10:08<22:14,  1.87it/s]
 77%|███████▋  | 8190/10682 [1:10:09<21:44,  1.91it/s]
 77%|███████▋  | 8191/10682 [1:10:09<21:23,  1.94it/s]
 77%|███████▋  | 8192/10682 [1:10:10<21:09,  1.96it/s]
 77%|███████▋  | 8193/10682 [1:10:10<21:00,  1.98it/s]
 77%|███████▋  | 8194/10682 [1:10:11<20:52,  1.99it/s]
 77%|███████▋  | 8195/10682 [1:10:11<20:46,  2.00it/s]
 77%|███████▋  | 8196/10682 [1:10:12<20:40,  2.00it/s]
 77%|███████▋  | 8197/10682 [1:10:12<20:37,  2.01it/s]
 77%|███████▋  | 8198/10682 [1:10:13<20:34,  2.01it/s]
 77%|███████▋  | 8199/10682 [1:10:13<20:33,  2.01it/s]
 77%|███████▋  | 8200/10682 [1:10:14<20:30,  2.02it/s]{'loss': 3.407, 'grad_norm': 0.21616776287555695, 'learning_rate': 0.00015566174472640188, 'epoch': 0.77}
                                                      
-
 77%|███████▋  | 8200/10682 [1:10:14<20:30,  2.02it/s]
 77%|███████▋  | 8201/10682 [1:10:14<20:32,  2.01it/s]
 77%|███████▋  | 8202/10682 [1:10:15<20:30,  2.02it/s]
 77%|███████▋  | 8203/10682 [1:10:15<20:28,  2.02it/s]
 77%|███████▋  | 8204/10682 [1:10:16<20:28,  2.02it/s]
 77%|███████▋  | 8205/10682 [1:10:16<20:27,  2.02it/s]
 77%|███████▋  | 8206/10682 [1:10:17<20:27,  2.02it/s]
 77%|███████▋  | 8207/10682 [1:10:17<20:26,  2.02it/s]
 77%|███████▋  | 8208/10682 [1:10:18<20:25,  2.02it/s]
 77%|███████▋  | 8209/10682 [1:10:18<20:24,  2.02it/s]
 77%|███████▋  | 8210/10682 [1:10:19<20:23,  2.02it/s]
 77%|███████▋  | 8211/10682 [1:10:19<20:22,  2.02it/s]
 77%|███████▋  | 8212/10682 [1:10:20<20:21,  2.02it/s]
 77%|███████▋  | 8213/10682 [1:10:20<20:23,  2.02it/s]
 77%|███████▋  | 8214/10682 [1:10:21<20:24,  2.02it/s]
 77%|███████▋  | 8215/10682 [1:10:21<20:24,  2.02it/s]
 77%|███████▋  | 8216/10682 [1:10:21<20:23,  2.02it/s]
 77%|███████▋  | 8217/10682 [1:10:22<20:21,  2.02it/s]
 77%|███████▋  | 8218/10682 [1:10:22<20:22,  2.02it/s]
 77%|███████▋  | 8219/10682 [1:10:23<20:21,  2.02it/s]
 77%|███████▋  | 8220/10682 [1:10:23<20:18,  2.02it/s]
 77%|███████▋  | 8221/10682 [1:10:24<20:17,  2.02it/s]
 77%|███████▋  | 8222/10682 [1:10:24<20:17,  2.02it/s]
 77%|███████▋  | 8223/10682 [1:10:25<23:44,  1.73it/s]
 77%|███████▋  | 8224/10682 [1:10:26<22:44,  1.80it/s]
 77%|███████▋  | 8225/10682 [1:10:26<22:02,  1.86it/s]
                                                      {'loss': 3.4203, 'grad_norm': 0.21189911663532257, 'learning_rate': 0.0001527113038415231, 'epoch': 0.77}
-
 77%|███████▋  | 8225/10682 [1:10:26<22:02,  1.86it/s]
 77%|███████▋  | 8226/10682 [1:10:27<21:33,  1.90it/s]
 77%|███████▋  | 8227/10682 [1:10:27<21:10,  1.93it/s]
 77%|███████▋  | 8228/10682 [1:10:28<20:54,  1.96it/s]
 77%|███████▋  | 8229/10682 [1:10:28<20:42,  1.97it/s]
 77%|███████▋  | 8230/10682 [1:10:29<20:34,  1.99it/s]
 77%|███████▋  | 8231/10682 [1:10:29<20:28,  2.00it/s]
 77%|███████▋  | 8232/10682 [1:10:30<20:22,  2.00it/s]
 77%|███████▋  | 8233/10682 [1:10:30<20:20,  2.01it/s]
 77%|███████▋  | 8234/10682 [1:10:31<20:19,  2.01it/s]
 77%|███████▋  | 8235/10682 [1:10:31<20:17,  2.01it/s]
 77%|███████▋  | 8236/10682 [1:10:32<20:13,  2.02it/s]
 77%|███████▋  | 8237/10682 [1:10:32<20:13,  2.02it/s]
 77%|███████▋  | 8238/10682 [1:10:33<20:11,  2.02it/s]
 77%|███████▋  | 8239/10682 [1:10:33<20:10,  2.02it/s]
 77%|███████▋  | 8240/10682 [1:10:34<20:08,  2.02it/s]
 77%|███████▋  | 8241/10682 [1:10:34<20:08,  2.02it/s]
 77%|███████▋  | 8242/10682 [1:10:35<20:07,  2.02it/s]
 77%|███████▋  | 8243/10682 [1:10:35<20:06,  2.02it/s]
 77%|███████▋  | 8244/10682 [1:10:36<20:07,  2.02it/s]
 77%|███████▋  | 8245/10682 [1:10:36<20:06,  2.02it/s]
 77%|███████▋  | 8246/10682 [1:10:37<20:05,  2.02it/s]
 77%|███████▋  | 8247/10682 [1:10:37<20:03,  2.02it/s]
 77%|███████▋  | 8248/10682 [1:10:38<20:03,  2.02it/s]
 77%|███████▋  | 8249/10682 [1:10:38<20:04,  2.02it/s]
 77%|███████▋  | 8250/10682 [1:10:39<20:04,  2.02it/s]
                                                      {'loss': 3.4106, 'grad_norm': 0.2258141189813614, 'learning_rate': 0.00014978404491439802, 'epoch': 0.77}
-
 77%|███████▋  | 8250/10682 [1:10:39<20:04,  2.02it/s]
 77%|███████▋  | 8251/10682 [1:10:39<20:04,  2.02it/s]
 77%|███████▋  | 8252/10682 [1:10:40<20:03,  2.02it/s]
 77%|███████▋  | 8253/10682 [1:10:40<20:01,  2.02it/s]
 77%|███████▋  | 8254/10682 [1:10:41<20:01,  2.02it/s]
 77%|███████▋  | 8255/10682 [1:10:41<20:01,  2.02it/s]
 77%|███████▋  | 8256/10682 [1:10:42<20:00,  2.02it/s]
 77%|███████▋  | 8257/10682 [1:10:42<20:00,  2.02it/s]
 77%|███████▋  | 8258/10682 [1:10:43<19:58,  2.02it/s]
 77%|███████▋  | 8259/10682 [1:10:43<19:59,  2.02it/s]
 77%|███████▋  | 8260/10682 [1:10:44<19:58,  2.02it/s]
 77%|███████▋  | 8261/10682 [1:10:44<19:58,  2.02it/s]
 77%|███████▋  | 8262/10682 [1:10:45<19:58,  2.02it/s]
 77%|███████▋  | 8263/10682 [1:10:45<19:59,  2.02it/s]
 77%|███████▋  | 8264/10682 [1:10:46<19:59,  2.02it/s]
 77%|███████▋  | 8265/10682 [1:10:46<19:58,  2.02it/s]
 77%|███████▋  | 8266/10682 [1:10:47<19:57,  2.02it/s]
 77%|███████▋  | 8267/10682 [1:10:47<19:56,  2.02it/s]
 77%|███████▋  | 8268/10682 [1:10:48<19:54,  2.02it/s]
 77%|███████▋  | 8269/10682 [1:10:48<19:54,  2.02it/s]
 77%|███████▋  | 8270/10682 [1:10:49<19:55,  2.02it/s]
 77%|███████▋  | 8271/10682 [1:10:49<19:54,  2.02it/s]
 77%|███████▋  | 8272/10682 [1:10:50<19:53,  2.02it/s]
 77%|███████▋  | 8273/10682 [1:10:50<19:52,  2.02it/s]
 77%|███████▋  | 8274/10682 [1:10:51<19:52,  2.02it/s]
 77%|███████▋  | 8275/10682 [1:10:51<19:51,  2.02it/s]
                                                      
{'loss': 3.4161, 'grad_norm': 0.21790149807929993, 'learning_rate': 0.00014688016334324605, 'epoch': 0.77}
-
 77%|███████▋  | 8275/10682 [1:10:51<19:51,  2.02it/s]
 77%|███████▋  | 8276/10682 [1:10:52<19:53,  2.02it/s]
 77%|███████▋  | 8277/10682 [1:10:52<19:52,  2.02it/s]
 77%|███████▋  | 8278/10682 [1:10:52<19:51,  2.02it/s]
 78%|███████▊  | 8279/10682 [1:10:53<19:50,  2.02it/s]
 78%|███████▊  | 8280/10682 [1:10:53<19:48,  2.02it/s]
 78%|███████▊  | 8281/10682 [1:10:54<19:48,  2.02it/s]
 78%|███████▊  | 8282/10682 [1:10:54<19:47,  2.02it/s]
 78%|███████▊  | 8283/10682 [1:10:55<19:48,  2.02it/s]
 78%|███████▊  | 8284/10682 [1:10:55<19:48,  2.02it/s]
 78%|███████▊  | 8285/10682 [1:10:56<19:47,  2.02it/s]
 78%|███████▊  | 8286/10682 [1:10:56<19:47,  2.02it/s]
 78%|███████▊  | 8287/10682 [1:10:57<19:45,  2.02it/s]
 78%|███████▊  | 8288/10682 [1:10:57<19:46,  2.02it/s]
 78%|███████▊  | 8289/10682 [1:10:58<19:43,  2.02it/s]
 78%|███████▊  | 8290/10682 [1:10:58<19:44,  2.02it/s]
 78%|███████▊  | 8291/10682 [1:10:59<19:44,  2.02it/s]
 78%|███████▊  | 8292/10682 [1:10:59<19:44,  2.02it/s]
 78%|███████▊  | 8293/10682 [1:11:00<19:44,  2.02it/s]
 78%|███████▊  | 8294/10682 [1:11:00<19:42,  2.02it/s]
 78%|███████▊  | 8295/10682 [1:11:01<19:41,  2.02it/s]
 78%|███████▊  | 8296/10682 [1:11:01<19:40,  2.02it/s]
 78%|███████▊  | 8297/10682 [1:11:02<19:39,  2.02it/s]
 78%|███████▊  | 8298/10682 [1:11:02<19:38,  2.02it/s]
 78%|███████▊  | 8299/10682 [1:11:03<19:36,  2.03it/s]
 78%|███████▊  | 8300/10682 [1:11:03<19:37,  2.02it/s]{'loss': 3.4068, 'grad_norm': 0.21814647316932678, 'learning_rate': 0.00014399985296581835, 'epoch': 0.78}
                                                      
-
 78%|███████▊  | 8300/10682 [1:11:03<19:37,  2.02it/s]
 78%|███████▊  | 8301/10682 [1:11:04<19:38,  2.02it/s]
 78%|███████▊  | 8302/10682 [1:11:04<19:39,  2.02it/s]
 78%|███████▊  | 8303/10682 [1:11:05<19:38,  2.02it/s]
 78%|███████▊  | 8304/10682 [1:11:05<19:37,  2.02it/s]
 78%|███████▊  | 8305/10682 [1:11:06<19:36,  2.02it/s]
 78%|███████▊  | 8306/10682 [1:11:06<19:36,  2.02it/s]
 78%|███████▊  | 8307/10682 [1:11:07<19:34,  2.02it/s]
 78%|███████▊  | 8308/10682 [1:11:07<19:34,  2.02it/s]
 78%|███████▊  | 8309/10682 [1:11:08<19:33,  2.02it/s]
 78%|███████▊  | 8310/10682 [1:11:08<19:34,  2.02it/s]
 78%|███████▊  | 8311/10682 [1:11:09<19:32,  2.02it/s]
 78%|███████▊  | 8312/10682 [1:11:09<19:32,  2.02it/s]
 78%|███████▊  | 8313/10682 [1:11:10<19:31,  2.02it/s]
 78%|██���████▊  | 8314/10682 [1:11:10<19:31,  2.02it/s]
 78%|███████▊  | 8315/10682 [1:11:11<19:32,  2.02it/s]
 78%|███████▊  | 8316/10682 [1:11:11<19:32,  2.02it/s]
 78%|███████▊  | 8317/10682 [1:11:12<19:30,  2.02it/s]
 78%|███████▊  | 8318/10682 [1:11:12<19:30,  2.02it/s]
 78%|███████▊  | 8319/10682 [1:11:13<19:30,  2.02it/s]
 78%|███████▊  | 8320/10682 [1:11:13<19:29,  2.02it/s]
 78%|███████▊  | 8321/10682 [1:11:14<19:29,  2.02it/s]
 78%|███████▊  | 8322/10682 [1:11:14<19:29,  2.02it/s]
 78%|███████▊  | 8323/10682 [1:11:15<19:29,  2.02it/s]
 78%|███████▊  | 8324/10682 [1:11:15<19:30,  2.01it/s]
 78%|███████▊  | 8325/10682 [1:11:16<19:29,  2.02it/s]{'loss': 3.4108, 'grad_norm': 0.22119268774986267, 'learning_rate': 0.00014114330604645943, 'epoch': 0.78}
                                                      
-
 78%|███████▊  | 8325/10682 [1:11:16<19:29,  2.02it/s]
 78%|███████▊  | 8326/10682 [1:11:16<19:29,  2.01it/s]
 78%|███████▊  | 8327/10682 [1:11:17<19:28,  2.02it/s]
 78%|███████▊  | 8328/10682 [1:11:17<19:26,  2.02it/s]
 78%|███████▊  | 8329/10682 [1:11:18<19:25,  2.02it/s]
 78%|███████▊  | 8330/10682 [1:11:18<19:25,  2.02it/s]
 78%|███████▊  | 8331/10682 [1:11:19<19:24,  2.02it/s]
 78%|███████▊  | 8332/10682 [1:11:19<19:24,  2.02it/s]
 78%|███████▊  | 8333/10682 [1:11:20<19:22,  2.02it/s]
 78%|███████▊  | 8334/10682 [1:11:20<19:21,  2.02it/s]
 78%|███████▊  | 8335/10682 [1:11:21<19:20,  2.02it/s]
 78%|███████▊  | 8336/10682 [1:11:21<19:19,  2.02it/s]
 78%|███████▊  | 8337/10682 [1:11:22<19:19,  2.02it/s]
 78%|███████▊  | 8338/10682 [1:11:22<19:20,  2.02it/s]
 78%|███████▊  | 8339/10682 [1:11:23<19:18,  2.02it/s]
 78%|███████▊  | 8340/10682 [1:11:23<19:20,  2.02it/s]
 78%|███████▊  | 8341/10682 [1:11:24<19:19,  2.02it/s]
 78%|███████▊  | 8342/10682 [1:11:24<19:19,  2.02it/s]
 78%|███████▊  | 8343/10682 [1:11:25<19:20,  2.02it/s]
 78%|███████▊  | 8344/10682 [1:11:25<19:19,  2.02it/s]
 78%|███████▊  | 8345/10682 [1:11:26<19:19,  2.02it/s]
 78%|███████▊  | 8346/10682 [1:11:26<19:17,  2.02it/s]
 78%|███████▊  | 8347/10682 [1:11:27<19:17,  2.02it/s]
 78%|███████▊  | 8348/10682 [1:11:27<19:15,  2.02it/s]
 78%|███████▊  | 8349/10682 [1:11:28<19:14,  2.02it/s]
 78%|███████▊  | 8350/10682 [1:11:28<19:14,  2.02it/s]
                                                      {'loss': 3.3957, 'grad_norm': 0.21989817917346954, 'learning_rate': 0.00013831071326327282, 'epoch': 0.78}
-

 78%|███████▊  | 8350/10682 [1:11:28<19:14,  2.02it/s]
 78%|███████▊  | 8351/10682 [1:11:29<19:15,  2.02it/s]
 78%|███████▊  | 8352/10682 [1:11:29<19:13,  2.02it/s]
 78%|███████▊  | 8353/10682 [1:11:30<19:13,  2.02it/s]
 78%|███████▊  | 8354/10682 [1:11:30<19:13,  2.02it/s]
 78%|███████▊  | 8355/10682 [1:11:31<19:12,  2.02it/s]
 78%|███████▊  | 8356/10682 [1:11:31<19:13,  2.02it/s]
 78%|███████▊  | 8357/10682 [1:11:32<19:13,  2.02it/s]
 78%|███████▊  | 8358/10682 [1:11:32<19:11,  2.02it/s]
 78%|███████▊  | 8359/10682 [1:11:33<19:11,  2.02it/s]
 78%|███████▊  | 8360/10682 [1:11:33<19:12,  2.01it/s]
 78%|███████▊  | 8361/10682 [1:11:34<19:11,  2.02it/s]
 78%|███████▊  | 8362/10682 [1:11:34<19:12,  2.01it/s]
 78%|███████▊  | 8363/10682 [1:11:35<19:10,  2.01it/s]
 78%|███████▊  | 8364/10682 [1:11:35<19:09,  2.02it/s]
 78%|███████▊  | 8365/10682 [1:11:36<19:07,  2.02it/s]
 78%|███████▊  | 8366/10682 [1:11:36<19:07,  2.02it/s]
 78%|███████▊  | 8367/10682 [1:11:37<19:05,  2.02it/s]
 78%|███████▊  | 8368/10682 [1:11:37<19:06,  2.02it/s]
 78%|███████▊  | 8369/10682 [1:11:38<19:04,  2.02it/s]
 78%|███████▊  | 8370/10682 [1:11:38<19:06,  2.02it/s]
 78%|███████▊  | 8371/10682 [1:11:39<19:04,  2.02it/s]
 78%|███████▊  | 8372/10682 [1:11:39<19:03,  2.02it/s]
 78%|███████▊  | 8373/10682 [1:11:40<19:03,  2.02it/s]
 78%|███████▊  | 8374/10682 [1:11:40<19:01,  2.02it/s]
 78%|███████▊  | 8375/10682 [1:11:41<19:03,  2.02it/s]
                                                      {'loss': 3.41, 'grad_norm': 0.22019170224666595, 'learning_rate': 0.0001355022636953933, 'epoch': 0.78}
-
 78%|███████▊  | 8375/10682 [1:11:41<19:03,  2.02it/s]
 78%|███████��  | 8376/10682 [1:11:41<19:03,  2.02it/s]
 78%|███████▊  | 8377/10682 [1:11:42<19:02,  2.02it/s]
 78%|███████▊  | 8378/10682 [1:11:42<19:00,  2.02it/s]
 78%|███████▊  | 8379/10682 [1:11:43<19:00,  2.02it/s]
 78%|███████▊  | 8380/10682 [1:11:43<18:59,  2.02it/s]
 78%|███████▊  | 8381/10682 [1:11:44<19:00,  2.02it/s]
 78%|███████▊  | 8382/10682 [1:11:44<18:59,  2.02it/s]
 78%|███████▊  | 8383/10682 [1:11:45<18:58,  2.02it/s]
 78%|███████▊  | 8384/10682 [1:11:45<18:56,  2.02it/s]
 78%|███████▊  | 8385/10682 [1:11:45<18:56,  2.02it/s]
 79%|███████▊  | 8386/10682 [1:11:46<18:55,  2.02it/s]
 79%|███████▊  | 8387/10682 [1:11:46<18:56,  2.02it/s]
 79%|███████▊  | 8388/10682 [1:11:47<18:55,  2.02it/s]
 79%|███████▊  | 8389/10682 [1:11:47<18:55,  2.02it/s]
 79%|███████▊  | 8390/10682 [1:11:48<18:54,  2.02it/s]
 79%|███████▊  | 8391/10682 [1:11:48<18:54,  2.02it/s]
 79%|███████▊  | 8392/10682 [1:11:49<18:53,  2.02it/s]
 79%|███████▊  | 8393/10682 [1:11:49<18:54,  2.02it/s]
 79%|███████▊  | 8394/10682 [1:11:50<18:54,  2.02it/s]
 79%|███████▊  | 8395/10682 [1:11:50<18:53,  2.02it/s]
 79%|███████▊  | 8396/10682 [1:11:51<18:53,  2.02it/s]
 79%|███████▊  | 8397/10682 [1:11:51<18:51,  2.02it/s]
 79%|███████▊  | 8398/10682 [1:11:52<18:50,  2.02it/s]
 79%|███████▊  | 8399/10682 [1:11:52<18:51,  2.02it/s]
 79%|███████▊  | 8400/10682 [1:11:53<18:50,  2.02it/s]{'loss': 3.4116, 'grad_norm': 0.2242874652147293, 'learning_rate': 0.0001327181448103661, 'epoch': 0.79}
                                                      
-
 79%|███████▊  | 8400/10682 [1:11:53<18:50,  2.02it/s]
 79%|███████▊  | 8401/10682 [1:11:53<18:51,  2.02it/s]
 79%|███████▊  | 8402/10682 [1:11:54<18:50,  2.02it/s]
 79%|███████▊  | 8403/10682 [1:11:54<18:50,  2.02it/s]
 79%|███████▊  | 8404/10682 [1:11:55<18:50,  2.01it/s]
 79%|███████▊  | 8405/10682 [1:11:55<18:49,  2.02it/s]
 79%|███████▊  | 8406/10682 [1:11:56<18:47,  2.02it/s]
 79%|███████▊  | 8407/10682 [1:11:56<18:48,  2.02it/s]
 79%|███████▊  | 8408/10682 [1:11:57<18:46,  2.02it/s]
 79%|███████▊  | 8409/10682 [1:11:57<18:47,  2.02it/s]
 79%|███████▊  | 8410/10682 [1:11:58<18:45,  2.02it/s]
 79%|███████▊  | 8411/10682 [1:11:58<18:44,  2.02it/s]
 79%|███████▊  | 8412/10682 [1:11:59<18:43,  2.02it/s]
 79%|███████▉  | 8413/10682 [1:11:59<18:42,  2.02it/s]
 79%|███████▉  | 8414/10682 [1:12:00<18:41,  2.02it/s]
 79%|███████▉  | 8415/10682 [1:12:00<18:42,  2.02it/s]
 79%|███████▉  | 8416/10682 [1:12:01<18:40,  2.02it/s]
 79%|███████▉  | 8417/10682 [1:12:01<18:41,  2.02it/s]
 79%|███████▉  | 8418/10682 [1:12:02<18:40,  2.02it/s]
 79%|███████▉  | 8419/10682 [1:12:02<18:40,  2.02it/s]
 79%|███████▉  | 8420/10682 [1:12:03<18:39,  2.02it/s]
 79%|███████▉  | 8421/10682 [1:12:03<18:40,  2.02it/s]
 79%|███████▉  | 8422/10682 [1:12:04<18:39,  2.02it/s]
 79%|███████▉  | 8423/10682 [1:12:04<18:40,  2.02it/s]
 79%|███████▉  | 8424/10682 [1:12:05<18:39,  2.02it/s]
 79%|███████▉  | 8425/10682 [1:12:05<18:40,  2.01it/s]
                                                      
{'loss': 3.4037, 'grad_norm': 0.2135806381702423, 'learning_rate': 0.00012995854245163207, 'epoch': 0.79}
-
 79%|███████▉  | 8425/10682 [1:12:05<18:40,  2.01it/s]
 79%|███████▉  | 8426/10682 [1:12:06<18:40,  2.01it/s]
 79%|███████▉  | 8427/10682 [1:12:06<18:37,  2.02it/s]
 79%|███████▉  | 8428/10682 [1:12:07<18:38,  2.01it/s]
 79%|███████▉  | 8429/10682 [1:12:07<18:35,  2.02it/s]
 79%|███████▉  | 8430/10682 [1:12:08<18:36,  2.02it/s]
 79%|███████▉  | 8431/10682 [1:12:08<18:33,  2.02it/s]
 79%|███████▉  | 8432/10682 [1:12:09<18:37,  2.01it/s]
 79%|███████▉  | 8433/10682 [1:12:09<18:35,  2.02it/s]
 79%|███████▉  | 8434/10682 [1:12:10<18:34,  2.02it/s]
 79%|███████▉  | 8435/10682 [1:12:10<18:33,  2.02it/s]
 79%|███████▉  | 8436/10682 [1:12:11<18:31,  2.02it/s]
 79%|███████▉  | 8437/10682 [1:12:11<18:30,  2.02it/s]
 79%|███████▉  | 8438/10682 [1:12:12<18:30,  2.02it/s]
 79%|███████▉  | 8439/10682 [1:12:12<18:27,  2.03it/s]
 79%|███████▉  | 8440/10682 [1:12:13<18:29,  2.02it/s]
 79%|███████▉  | 8441/10682 [1:12:13<18:30,  2.02it/s]
 79%|███████▉  | 8442/10682 [1:12:14<18:29,  2.02it/s]
 79%|███████▉  | 8443/10682 [1:12:14<18:29,  2.02it/s]
 79%|███████▉  | 8444/10682 [1:12:15<18:27,  2.02it/s]
 79%|███████▉  | 8445/10682 [1:12:15<18:26,  2.02it/s]
 79%|███████▉  | 8446/10682 [1:12:16<18:27,  2.02it/s]
 79%|███████▉  | 8447/10682 [1:12:16<18:25,  2.02it/s]
 79%|███████▉  | 8448/10682 [1:12:17<18:25,  2.02it/s]
 79%|███████▉  | 8449/10682 [1:12:17<18:24,  2.02it/s]
 79%|███████▉  | 8450/10682 [1:12:18<18:24,  2.02it/s]{'loss': 3.4065, 'grad_norm': 0.2191745489835739, 'learning_rate': 0.0001272236408261237, 'epoch': 0.79}
                                                      
-
 79%|███████▉  | 8450/10682 [1:12:18<18:24,  2.02it/s]
 79%|███████▉  | 8451/10682 [1:12:18<18:26,  2.02it/s]
 79%|███████▉  | 8452/10682 [1:12:19<18:25,  2.02it/s]
 79%|███████▉  | 8453/10682 [1:12:19<18:26,  2.01it/s]
 79%|███████▉  | 8454/10682 [1:12:20<18:25,  2.02it/s]
 79%|███████▉  | 8455/10682 [1:12:20<18:24,  2.02it/s]
 79%|███████▉  | 8456/10682 [1:12:21<18:22,  2.02it/s]
 79%|███████▉  | 8457/10682 [1:12:21<18:22,  2.02it/s]
 79%|███████▉  | 8458/10682 [1:12:22<18:21,  2.02it/s]
 79%|███████▉  | 8459/10682 [1:12:22<18:20,  2.02it/s]
 79%|███████▉  | 8460/10682 [1:12:23<18:20,  2.02it/s]
 79%|███████▉  | 8461/10682 [1:12:23<18:19,  2.02it/s]
 79%|███████▉  | 8462/10682 [1:12:24<18:19,  2.02it/s]
 79%|███████▉  | 8463/10682 [1:12:24<18:20,  2.02it/s]
 79%|███████▉  | 8464/10682 [1:12:25<18:18,  2.02it/s]
 79%|███████▉  | 8465/10682 [1:12:25<18:17,  2.02it/s]
 79%|███████▉  | 8466/10682 [1:12:26<18:16,  2.02it/s]
 79%|███████▉  | 8467/10682 [1:12:26<18:15,  2.02it/s]
 79%|███████▉  | 8468/10682 [1:12:27<18:16,  2.02it/s]
 79%|███████▉  | 8469/10682 [1:12:27<18:14,  2.02it/s]
 79%|███████▉  | 8470/10682 [1:12:28<18:15,  2.02it/s]
 79%|███████▉  | 8471/10682 [1:12:28<18:13,  2.02it/s]
 79%|███████▉  | 8472/10682 [1:12:29<18:13,  2.02it/s]
 79%|███████▉  | 8473/10682 [1:12:29<18:12,  2.02it/s]
 79%|███████▉  | 8474/10682 [1:12:30<18:14,  2.02it/s]
 79%|███████▉  | 8475/10682 [1:12:30<18:13,  2.02it/s]
                                                      {'loss': 3.4058, 'grad_norm': 0.22625406086444855, 'learning_rate': 0.00012451362249196797, 'epoch': 0.79}
-
 79%|███████▉  | 8475/10682 [1:12:30<18:13,  2.02it/s]
 79%|███████▉  | 8476/10682 [1:12:31<18:15,  2.01it/s]
 79%|███████▉  | 8477/10682 [1:12:31<18:14,  2.02it/s]
 79%|███████▉  | 8478/10682 [1:12:32<18:11,  2.02it/s]
 79%|███████▉  | 8479/10682 [1:12:32<18:11,  2.02it/s]
 79%|███████▉  | 8480/10682 [1:12:33<18:10,  2.02it/s]
 79%|███████▉  | 8481/10682 [1:12:33<18:10,  2.02it/s]
 79%|███████▉  | 8482/10682 [1:12:34<18:10,  2.02it/s]
 79%|███████▉  | 8483/10682 [1:12:34<18:09,  2.02it/s]
 79%|███████▉  | 8484/10682 [1:12:35<18:08,  2.02it/s]
 79%|███████▉  | 8485/10682 [1:12:35<18:07,  2.02it/s]
 79%|███████▉  | 8486/10682 [1:12:36<18:06,  2.02it/s]
 79%|███████▉  | 8487/10682 [1:12:36<18:06,  2.02it/s]
 79%|███████▉  | 8488/10682 [1:12:37<18:06,  2.02it/s]
 79%|███████▉  | 8489/10682 [1:12:37<18:05,  2.02it/s]
 79%|███████▉  | 8490/10682 [1:12:38<18:04,  2.02it/s]
 79%|███████▉  | 8491/10682 [1:12:38<18:04,  2.02it/s]
 79%|███████▉  | 8492/10682 [1:12:38<18:03,  2.02it/s]
 80%|███████▉  | 8493/10682 [1:12:39<18:04,  2.02it/s]
 80%|███████▉  | 8494/10682 [1:12:39<18:02,  2.02it/s]
 80%|███████▉  | 8495/10682 [1:12:40<18:03,  2.02it/s]
 80%|███████▉  | 8496/10682 [1:12:40<18:02,  2.02it/s]
 80%|███████▉  | 8497/10682 [1:12:41<18:02,  2.02it/s]
 80%|███████▉  | 8498/10682 [1:12:41<18:00,  2.02it/s]
 80%|███████▉  | 8499/10682 [1:12:42<18:01,  2.02it/s]
 80%|███████▉  | 8500/10682 [1:12:42<17:59,  2.02it/s]
                                                      
{'loss': 3.3879, 'grad_norm': 0.21326370537281036, 'learning_rate': 0.00012182866834630096, 'epoch': 0.8}
-
 80%|███████▉  | 8500/10682 [1:12:42<17:59,  2.02it/s]
 80%|███████▉  | 8501/10682 [1:12:43<18:01,  2.02it/s]
 80%|███████▉  | 8502/10682 [1:12:43<18:00,  2.02it/s]
 80%|███████▉  | 8503/10682 [1:12:44<18:00,  2.02it/s]
 80%|███████▉  | 8504/10682 [1:12:44<17:58,  2.02it/s]
 80%|███████▉  | 8505/10682 [1:12:45<17:57,  2.02it/s]
 80%|███████▉  | 8506/10682 [1:12:45<17:56,  2.02it/s]
 80%|███████▉  | 8507/10682 [1:12:46<17:55,  2.02it/s]
 80%|███████▉  | 8508/10682 [1:12:46<17:54,  2.02it/s]
 80%|███████▉  | 8509/10682 [1:12:47<17:54,  2.02it/s]
 80%|███████▉  | 8510/10682 [1:12:47<17:56,  2.02it/s]
 80%|███████▉  | 8511/10682 [1:12:48<17:56,  2.02it/s]
 80%|███████▉  | 8512/10682 [1:12:48<17:56,  2.02it/s]
 80%|███████▉  | 8513/10682 [1:12:49<17:54,  2.02it/s]
 80%|███████▉  | 8514/10682 [1:12:49<17:54,  2.02it/s]
 80%|███████▉  | 8515/10682 [1:12:50<17:53,  2.02it/s]
 80%|███████▉  | 8516/10682 [1:12:50<17:52,  2.02it/s]
 80%|███████▉  | 8517/10682 [1:12:51<17:50,  2.02it/s]
 80%|███████▉  | 8518/10682 [1:12:51<17:50,  2.02it/s]
 80%|███████▉  | 8519/10682 [1:12:52<17:49,  2.02it/s]
 80%|███████▉  | 8520/10682 [1:12:52<17:50,  2.02it/s]
 80%|███████▉  | 8521/10682 [1:12:53<17:48,  2.02it/s]
 80%|███████▉  | 8522/10682 [1:12:53<17:48,  2.02it/s]
 80%|███████▉  | 8523/10682 [1:12:54<17:47,  2.02it/s]
 80%|███████▉  | 8524/10682 [1:12:54<17:48,  2.02it/s]
 80%|███████▉  | 8525/10682 [1:12:55<17:46,  2.02it/s]
                                                      
{'loss': 3.3904, 'grad_norm': 0.21883653104305267, 'learning_rate': 0.00011916895761319264, 'epoch': 0.8}
-
 80%|███████▉  | 8525/10682 [1:12:55<17:46,  2.02it/s]
 80%|███████▉  | 8526/10682 [1:12:55<17:48,  2.02it/s]
 80%|███████▉  | 8527/10682 [1:12:56<17:46,  2.02it/s]
 80%|███████▉  | 8528/10682 [1:12:56<17:46,  2.02it/s]
 80%|███████▉  | 8529/10682 [1:12:57<17:45,  2.02it/s]
 80%|███████▉  | 8530/10682 [1:12:57<17:46,  2.02it/s]
 80%|███████▉  | 8531/10682 [1:12:58<17:44,  2.02it/s]
 80%|███████▉  | 8532/10682 [1:12:58<17:46,  2.02it/s]
 80%|███████▉  | 8533/10682 [1:12:59<17:45,  2.02it/s]
 80%|███████▉  | 8534/10682 [1:12:59<17:45,  2.02it/s]
 80%|███████▉  | 8535/10682 [1:13:00<17:44,  2.02it/s]
 80%|███████▉  | 8536/10682 [1:13:00<17:42,  2.02it/s]
 80%|███████▉  | 8537/10682 [1:13:01<17:41,  2.02it/s]
 80%|███████▉  | 8538/10682 [1:13:01<17:41,  2.02it/s]
 80%|███████▉  | 8539/10682 [1:13:02<17:41,  2.02it/s]
 80%|███████▉  | 8540/10682 [1:13:02<17:41,  2.02it/s]
 80%|███████▉  | 8541/10682 [1:13:03<17:42,  2.02it/s]
 80%|███████▉  | 8542/10682 [1:13:03<17:41,  2.02it/s]
 80%|███████▉  | 8543/10682 [1:13:04<17:39,  2.02it/s]
 80%|███████▉  | 8544/10682 [1:13:04<17:38,  2.02it/s]
 80%|███████▉  | 8545/10682 [1:13:05<17:39,  2.02it/s]
 80%|████████  | 8546/10682 [1:13:05<17:37,  2.02it/s]
 80%|████████  | 8547/10682 [1:13:06<17:37,  2.02it/s]
 80%|████████  | 8548/10682 [1:13:06<17:36,  2.02it/s]
 80%|████████  | 8549/10682 [1:13:07<17:35,  2.02it/s]
 80%|████████  | 8550/10682 [1:13:07<17:35,  2.02it/s]{'loss': 3.4068, 'grad_norm': 0.20940828323364258, 'learning_rate': 0.0001165346678316832, 'epoch': 0.8}
                                                      
-
 80%|████████  | 8550/10682 [1:13:07<17:35,  2.02it/s]
 80%|████████  | 8551/10682 [1:13:08<17:35,  2.02it/s]
 80%|████████  | 8552/10682 [1:13:08<17:34,  2.02it/s]
 80%|████████  | 8553/10682 [1:13:09<17:34,  2.02it/s]
 80%|████████  | 8554/10682 [1:13:09<17:32,  2.02it/s]
 80%|████████  | 8555/10682 [1:13:10<17:32,  2.02it/s]
 80%|████████  | 8556/10682 [1:13:10<17:31,  2.02it/s]
 80%|████████  | 8557/10682 [1:13:11<17:31,  2.02it/s]
 80%|████████  | 8558/10682 [1:13:11<17:30,  2.02it/s]
 80%|████████  | 8559/10682 [1:13:12<17:30,  2.02it/s]
 80%|████████  | 8560/10682 [1:13:12<17:31,  2.02it/s]
 80%|████████  | 8561/10682 [1:13:13<17:31,  2.02it/s]
 80%|████████  | 8562/10682 [1:13:13<17:32,  2.01it/s]
 80%|████████  | 8563/10682 [1:13:14<17:30,  2.02it/s]
 80%|████████  | 8564/10682 [1:13:14<17:30,  2.02it/s]
 80%|████████  | 8565/10682 [1:13:15<17:29,  2.02it/s]
 80%|████████  | 8566/10682 [1:13:15<17:28,  2.02it/s]
 80%|████████  | 8567/10682 [1:13:16<17:26,  2.02it/s]
 80%|████████  | 8568/10682 [1:13:16<17:27,  2.02it/s]
 80%|████████  | 8569/10682 [1:13:17<17:24,  2.02it/s]
 80%|████████  | 8570/10682 [1:13:17<17:25,  2.02it/s]
 80%|████████  | 8571/10682 [1:13:18<17:24,  2.02it/s]
 80%|████████  | 8572/10682 [1:13:18<17:25,  2.02it/s]
 80%|████████  | 8573/10682 [1:13:19<17:24,  2.02it/s]
 80%|████████  | 8574/10682 [1:13:19<17:23,  2.02it/s]
 80%|████████  | 8575/10682 [1:13:20<17:23,  2.02it/s]
                                                      
{'loss': 3.4041, 'grad_norm': 0.22104041278362274, 'learning_rate': 0.00011392597484393285, 'epoch': 0.8}
-
 80%|████████  | 8575/10682 [1:13:20<17:23,  2.02it/s]
 80%|████████  | 8576/10682 [1:13:20<17:25,  2.01it/s]
 80%|████████  | 8577/10682 [1:13:21<17:25,  2.01it/s]
 80%|████████  | 8578/10682 [1:13:21<17:28,  2.01it/s]
 80%|████████  | 8579/10682 [1:13:22<17:26,  2.01it/s]
 80%|████████  | 8580/10682 [1:13:22<17:27,  2.01it/s]
 80%|████████  | 8581/10682 [1:13:23<17:25,  2.01it/s]
 80%|████████  | 8582/10682 [1:13:23<17:26,  2.01it/s]
 80%|████████  | 8583/10682 [1:13:24<17:25,  2.01it/s]
 80%|████████  | 8584/10682 [1:13:24<17:24,  2.01it/s]
 80%|████████  | 8585/10682 [1:13:25<17:23,  2.01it/s]
 80%|████████  | 8586/10682 [1:13:25<17:20,  2.02it/s]
 80%|████████  | 8587/10682 [1:13:26<17:19,  2.02it/s]
 80%|████████  | 8588/10682 [1:13:26<17:17,  2.02it/s]
 80%|████████  | 8589/10682 [1:13:27<17:17,  2.02it/s]
 80%|████████  | 8590/10682 [1:13:27<17:17,  2.02it/s]
 80%|████████  | 8591/10682 [1:13:28<17:17,  2.02it/s]
 80%|████████  | 8592/10682 [1:13:28<17:17,  2.01it/s]
 80%|████████  | 8593/10682 [1:13:29<17:17,  2.01it/s]
 80%|████████  | 8594/10682 [1:13:29<17:17,  2.01it/s]
 80%|████████  | 8595/10682 [1:13:30<17:16,  2.01it/s]
 80%|████████  | 8596/10682 [1:13:30<17:16,  2.01it/s]
 80%|████████  | 8597/10682 [1:13:31<17:15,  2.01it/s]
 80%|████████  | 8598/10682 [1:13:31<17:14,  2.01it/s]
 80%|████████  | 8599/10682 [1:13:32<17:16,  2.01it/s]
 81%|████████  | 8600/10682 [1:13:32<17:15,  2.01it/s]
                                                      
{'loss': 3.4091, 'grad_norm': 0.21381522715091705, 'learning_rate': 0.00011134305278348312, 'epoch': 0.81}
-
 81%|████████  | 8600/10682 [1:13:32<17:15,  2.01it/s]
 81%|████████  | 8601/10682 [1:13:33<17:16,  2.01it/s]
 81%|████████  | 8602/10682 [1:13:33<17:17,  2.01it/s]
 81%|████████  | 8603/10682 [1:13:34<17:14,  2.01it/s]
 81%|████████  | 8604/10682 [1:13:34<17:14,  2.01it/s]
 81%|████████  | 8605/10682 [1:13:35<17:12,  2.01it/s]
 81%|████████  | 8606/10682 [1:13:35<17:11,  2.01it/s]
 81%|████████  | 8607/10682 [1:13:36<17:09,  2.02it/s]
 81%|████████  | 8608/10682 [1:13:36<17:08,  2.02it/s]
 81%|████████  | 8609/10682 [1:13:36<17:09,  2.01it/s]
 81%|████████  | 8610/10682 [1:13:37<17:09,  2.01it/s]
 81%|████████  | 8611/10682 [1:13:37<17:10,  2.01it/s]
 81%|████████  | 8612/10682 [1:13:38<17:10,  2.01it/s]
 81%|████████  | 8613/10682 [1:13:38<17:07,  2.01it/s]
 81%|████████  | 8614/10682 [1:13:39<17:08,  2.01it/s]
 81%|████████  | 8615/10682 [1:13:39<17:06,  2.01it/s]
 81%|████████  | 8616/10682 [1:13:40<17:06,  2.01it/s]
 81%|████████  | 8617/10682 [1:13:40<17:04,  2.02it/s]
 81%|████████  | 8618/10682 [1:13:41<17:03,  2.02it/s]
 81%|████████  | 8619/10682 [1:13:41<17:04,  2.01it/s]
 81%|████████  | 8620/10682 [1:13:42<17:02,  2.02it/s]
 81%|████████  | 8621/10682 [1:13:42<17:05,  2.01it/s]
 81%|████████  | 8622/10682 [1:13:43<17:02,  2.01it/s]
 81%|████████  | 8623/10682 [1:13:43<17:02,  2.01it/s]
 81%|████████  | 8624/10682 [1:13:44<17:01,  2.01it/s]
 81%|████████  | 8625/10682 [1:13:44<17:01,  2.01it/s]
                                                      
{'loss': 3.406, 'grad_norm': 0.2158418595790863, 'learning_rate': 0.00010878607406363367, 'epoch': 0.81}
-
 81%|████████  | 8625/10682 [1:13:44<17:01,  2.01it/s]
 81%|████████  | 8626/10682 [1:13:45<17:01,  2.01it/s]
 81%|████████  | 8627/10682 [1:13:45<17:00,  2.01it/s]
 81%|████████  | 8628/10682 [1:13:46<16:59,  2.01it/s]
 81%|████████  | 8629/10682 [1:13:46<16:58,  2.02it/s]
 81%|████████  | 8630/10682 [1:13:47<16:58,  2.02it/s]
 81%|████████  | 8631/10682 [1:13:47<16:57,  2.02it/s]
 81%|████████  | 8632/10682 [1:13:48<16:57,  2.01it/s]
 81%|████████  | 8633/10682 [1:13:48<16:58,  2.01it/s]
 81%|████████  | 8634/10682 [1:13:49<16:59,  2.01it/s]
 81%|████████  | 8635/10682 [1:13:49<16:58,  2.01it/s]
 81%|████████  | 8636/10682 [1:13:50<16:56,  2.01it/s]
 81%|████████  | 8637/10682 [1:13:50<16:55,  2.01it/s]
 81%|████████  | 8638/10682 [1:13:51<16:55,  2.01it/s]
 81%|████████  | 8639/10682 [1:13:51<16:56,  2.01it/s]
 81%|████████  | 8640/10682 [1:13:52<16:54,  2.01it/s]
 81%|████████  | 8641/10682 [1:13:52<16:54,  2.01it/s]
 81%|████████  | 8642/10682 [1:13:53<16:53,  2.01it/s]
 81%|████████  | 8643/10682 [1:13:53<16:53,  2.01it/s]
 81%|████████  | 8644/10682 [1:13:54<16:53,  2.01it/s]
 81%|████████  | 8645/10682 [1:13:54<16:52,  2.01it/s]
 81%|████████  | 8646/10682 [1:13:55<16:51,  2.01it/s]
 81%|████████  | 8647/10682 [1:13:55<16:50,  2.01it/s]
 81%|████████  | 8648/10682 [1:13:56<16:50,  2.01it/s]
 81%|████████  | 8649/10682 [1:13:56<16:50,  2.01it/s]
 81%|████████  | 8650/10682 [1:13:57<16:50,  2.01it/s]
                                                      {'loss': 3.3912, 'grad_norm': 0.2183406502008438, 'learning_rate': 0.00010625520936593375, 'epoch': 0.81}
-
 81%|████████  | 8650/10682 [1:13:57<16:50,  2.01it/s]
 81%|████████  | 8651/10682 [1:13:57<16:51,  2.01it/s]
 81%|████████  | 8652/10682 [1:13:58<16:50,  2.01it/s]
 81%|████████  | 8653/10682 [1:13:58<16:50,  2.01it/s]
 81%|████████  | 8654/10682 [1:13:59<16:47,  2.01it/s]
 81%|████████  | 8655/10682 [1:13:59<16:48,  2.01it/s]
 81%|████████  | 8656/10682 [1:14:00<16:47,  2.01it/s]
 81%|████████  | 8657/10682 [1:14:00<16:46,  2.01it/s]
 81%|████████  | 8658/10682 [1:14:01<16:46,  2.01it/s]
 81%|████████  | 8659/10682 [1:14:01<16:44,  2.01it/s]
 81%|████████  | 8660/10682 [1:14:02<16:46,  2.01it/s]
 81%|████████  | 8661/10682 [1:14:02<16:46,  2.01it/s]
 81%|████████  | 8662/10682 [1:14:03<16:43,  2.01it/s]
 81%|████████  | 8663/10682 [1:14:03<16:44,  2.01it/s]
 81%|████████  | 8664/10682 [1:14:04<16:43,  2.01it/s]
 81%|████████  | 8665/10682 [1:14:04<16:42,  2.01it/s]
 81%|████████  | 8666/10682 [1:14:05<16:43,  2.01it/s]
 81%|████████  | 8667/10682 [1:14:05<16:42,  2.01it/s]
 81%|████████  | 8668/10682 [1:14:06<16:42,  2.01it/s]
 81%|████████  | 8669/10682 [1:14:06<16:41,  2.01it/s]
 81%|████████  | 8670/10682 [1:14:07<16:39,  2.01it/s]
 81%|████████  | 8671/10682 [1:14:07<16:39,  2.01it/s]
 81%|████████  | 8672/10682 [1:14:08<16:37,  2.01it/s]
 81%|████████  | 8673/10682 [1:14:08<16:38,  2.01it/s]
 81%|████████  | 8674/10682 [1:14:09<16:37,  2.01it/s]
 81%|████████  | 8675/10682 [1:14:09<16:37,  2.01it/s]{'loss': 3.4007, 'grad_norm': 0.22159014642238617, 'learning_rate': 0.0001037506276287885, 'epoch': 0.81}
                                                      
-
 81%|████████  | 8675/10682 [1:14:09<16:37,  2.01it/s]
 81%|████████  | 8676/10682 [1:14:10<16:38,  2.01it/s]
 81%|████████  | 8677/10682 [1:14:10<16:37,  2.01it/s]
 81%|████████  | 8678/10682 [1:14:11<16:38,  2.01it/s]
 81%|████████  | 8679/10682 [1:14:11<16:37,  2.01it/s]
 81%|████████▏ | 8680/10682 [1:14:12<16:36,  2.01it/s]
 81%|████████▏ | 8681/10682 [1:14:12<16:36,  2.01it/s]
 81%|████████▏ | 8682/10682 [1:14:13<16:36,  2.01it/s]
 81%|████████▏ | 8683/10682 [1:14:13<16:33,  2.01it/s]
 81%|████████▏ | 8684/10682 [1:14:14<16:34,  2.01it/s]
 81%|████████▏ | 8685/10682 [1:14:14<16:35,  2.01it/s]
 81%|████████▏ | 8686/10682 [1:14:15<16:34,  2.01it/s]
 81%|████████▏ | 8687/10682 [1:14:15<16:33,  2.01it/s]
 81%|████████▏ | 8688/10682 [1:14:16<16:32,  2.01it/s]
 81%|████████▏ | 8689/10682 [1:14:16<16:31,  2.01it/s]
 81%|████████▏ | 8690/10682 [1:14:17<16:30,  2.01it/s]
 81%|████████▏ | 8691/10682 [1:14:17<16:28,  2.01it/s]
 81%|████████▏ | 8692/10682 [1:14:18<16:29,  2.01it/s]
 81%|████████▏ | 8693/10682 [1:14:18<16:29,  2.01it/s]
 81%|████████▏ | 8694/10682 [1:14:19<16:27,  2.01it/s]
 81%|████████▏ | 8695/10682 [1:14:19<16:28,  2.01it/s]
 81%|████████▏ | 8696/10682 [1:14:20<16:25,  2.02it/s]
 81%|████████▏ | 8697/10682 [1:14:20<16:26,  2.01it/s]
 81%|████████▏ | 8698/10682 [1:14:21<16:25,  2.01it/s]
 81%|████████▏ | 8699/10682 [1:14:21<16:24,  2.01it/s]
 81%|████████▏ | 8700/10682 [1:14:22<16:25,  2.01it/s]
                                                      {'loss': 3.3982, 'grad_norm': 0.2158488780260086, 'learning_rate': 0.0001012724960361826, 'epoch': 0.81}
-
 81%|████████▏ | 8700/10682 [1:14:22<16:25,  2.01it/s]
 81%|████████▏ | 8701/10682 [1:14:22<16:26,  2.01it/s]
 81%|████████▏ | 8702/10682 [1:14:23<16:25,  2.01it/s]
 81%|████████▏ | 8703/10682 [1:14:23<16:25,  2.01it/s]
 81%|████████▏ | 8704/10682 [1:14:24<16:23,  2.01it/s]
 81%|████████▏ | 8705/10682 [1:14:24<16:22,  2.01it/s]
 82%|████████▏ | 8706/10682 [1:14:25<16:20,  2.01it/s]
 82%|████████▏ | 8707/10682 [1:14:25<16:21,  2.01it/s]
 82%|████████▏ | 8708/10682 [1:14:26<16:20,  2.01it/s]
 82%|████████▏ | 8709/10682 [1:14:26<16:22,  2.01it/s]
 82%|████████▏ | 8710/10682 [1:14:27<16:22,  2.01it/s]
 82%|████████▏ | 8711/10682 [1:14:27<16:20,  2.01it/s]
 82%|████████▏ | 8712/10682 [1:14:28<16:19,  2.01it/s]
 82%|████████▏ | 8713/10682 [1:14:28<16:19,  2.01it/s]
 82%|████████▏ | 8714/10682 [1:14:29<16:17,  2.01it/s]
 82%|████████▏ | 8715/10682 [1:14:29<16:17,  2.01it/s]
 82%|████████▏ | 8716/10682 [1:14:30<16:17,  2.01it/s]
 82%|████████▏ | 8717/10682 [1:14:30<16:17,  2.01it/s]
 82%|████████▏ | 8718/10682 [1:14:31<16:17,  2.01it/s]
 82%|████████▏ | 8719/10682 [1:14:31<16:17,  2.01it/s]
 82%|████████▏ | 8720/10682 [1:14:32<16:16,  2.01it/s]
 82%|████████▏ | 8721/10682 [1:14:32<16:16,  2.01it/s]
 82%|████████▏ | 8722/10682 [1:14:33<16:14,  2.01it/s]
 82%|████████▏ | 8723/10682 [1:14:33<16:14,  2.01it/s]
 82%|████████▏ | 8724/10682 [1:14:34<16:12,  2.01it/s]
 82%|████████▏ | 8725/10682 [1:14:34<16:12,  2.01it/s]{'loss': 3.3949, 'grad_norm': 0.21897079050540924, 'learning_rate': 9.882098000652034e-05, 'epoch': 0.82}
                                                      
-
 82%|████████▏ | 8725/10682 [1:14:34<16:12,  2.01it/s]
 82%|████████▏ | 8726/10682 [1:14:35<16:13,  2.01it/s]
 82%|████████▏ | 8727/10682 [1:14:35<16:12,  2.01it/s]
 82%|████████▏ | 8728/10682 [1:14:36<16:12,  2.01it/s]
 82%|████████▏ | 8729/10682 [1:14:36<16:10,  2.01it/s]
 82%|████████▏ | 8730/10682 [1:14:37<16:11,  2.01it/s]
 82%|████████▏ | 8731/10682 [1:14:37<16:10,  2.01it/s]
 82%|████████▏ | 8732/10682 [1:14:38<16:08,  2.01it/s]
 82%|████████▏ | 8733/10682 [1:14:38<16:09,  2.01it/s]
 82%|████████▏ | 8734/10682 [1:14:39<16:09,  2.01it/s]
 82%|████████▏ | 8735/10682 [1:14:39<16:08,  2.01it/s]
 82%|████████▏ | 8736/10682 [1:14:40<16:09,  2.01it/s]
 82%|████████▏ | 8737/10682 [1:14:40<16:08,  2.01it/s]
 82%|████████▏ | 8738/10682 [1:14:41<16:08,  2.01it/s]
 82%|████████▏ | 8739/10682 [1:14:41<16:08,  2.01it/s]
 82%|████████▏ | 8740/10682 [1:14:42<16:06,  2.01it/s]
 82%|████████▏ | 8741/10682 [1:14:42<16:07,  2.01it/s]
 82%|████████▏ | 8742/10682 [1:14:43<16:06,  2.01it/s]
 82%|████████▏ | 8743/10682 [1:14:43<16:03,  2.01it/s]
 82%|████████▏ | 8744/10682 [1:14:44<16:04,  2.01it/s]
 82%|████████▏ | 8745/10682 [1:14:44<16:03,  2.01it/s]
 82%|████████▏ | 8746/10682 [1:14:45<16:02,  2.01it/s]
 82%|████████▏ | 8747/10682 [1:14:45<16:02,  2.01it/s]
 82%|████████▏ | 8748/10682 [1:14:46<16:03,  2.01it/s]
 82%|████████▏ | 8749/10682 [1:14:46<16:02,  2.01it/s]
 82%|████████▏ | 8750/10682 [1:14:47<16:02,  2.01it/s]
                                                      {'loss': 3.3944, 'grad_norm': 0.2202555537223816, 'learning_rate': 9.639624318158335e-05, 'epoch': 0.82}
-
 82%|████████▏ | 8750/10682 [1:14:47<16:02,  2.01it/s]
 82%|████████▏ | 8751/10682 [1:14:47<16:04,  2.00it/s]
 82%|████████▏ | 8752/10682 [1:14:48<16:00,  2.01it/s]
 82%|████████▏ | 8753/10682 [1:14:48<16:00,  2.01it/s]
 82%|████████▏ | 8754/10682 [1:14:49<16:00,  2.01it/s]
 82%|████████▏ | 8755/10682 [1:14:49<15:58,  2.01it/s]
 82%|████████▏ | 8756/10682 [1:14:50<15:58,  2.01it/s]
 82%|████████�� | 8757/10682 [1:14:50<15:58,  2.01it/s]
 82%|████████▏ | 8758/10682 [1:14:51<15:57,  2.01it/s]
 82%|████████▏ | 8759/10682 [1:14:51<15:57,  2.01it/s]
 82%|████████▏ | 8760/10682 [1:14:52<15:57,  2.01it/s]
 82%|████████▏ | 8761/10682 [1:14:52<15:55,  2.01it/s]
 82%|████████▏ | 8762/10682 [1:14:53<15:54,  2.01it/s]
 82%|████████▏ | 8763/10682 [1:14:53<15:54,  2.01it/s]
 82%|████████▏ | 8764/10682 [1:14:54<15:54,  2.01it/s]
 82%|████████▏ | 8765/10682 [1:14:54<15:54,  2.01it/s]
 82%|████████▏ | 8766/10682 [1:14:55<15:53,  2.01it/s]
 82%|████████▏ | 8767/10682 [1:14:55<15:54,  2.01it/s]
 82%|████████▏ | 8768/10682 [1:14:56<15:54,  2.00it/s]
 82%|████████▏ | 8769/10682 [1:14:56<15:53,  2.01it/s]
 82%|████████▏ | 8770/10682 [1:14:57<15:52,  2.01it/s]
 82%|████████▏ | 8771/10682 [1:14:57<15:51,  2.01it/s]
 82%|████████▏ | 8772/10682 [1:14:58<15:50,  2.01it/s]
 82%|████████▏ | 8773/10682 [1:14:58<15:50,  2.01it/s]
 82%|████████▏ | 8774/10682 [1:14:59<15:49,  2.01it/s]
 82%|████████▏ | 8775/10682 [1:14:59<15:49,  2.01it/s]{'loss': 3.3955, 'grad_norm': 0.21792520582675934, 'learning_rate': 9.399844741560781e-05, 'epoch': 0.82}
-
                                                      

 82%|████████▏ | 8775/10682 [1:14:59<15:49,  2.01it/s]
 82%|████████▏ | 8776/10682 [1:15:00<15:49,  2.01it/s]
 82%|████████▏ | 8777/10682 [1:15:00<15:49,  2.01it/s]
 82%|████████▏ | 8778/10682 [1:15:01<15:49,  2.01it/s]
 82%|████████▏ | 8779/10682 [1:15:01<15:48,  2.01it/s]
 82%|████████▏ | 8780/10682 [1:15:02<15:48,  2.01it/s]
 82%|████████▏ | 8781/10682 [1:15:02<15:47,  2.01it/s]
 82%|████████▏ | 8782/10682 [1:15:03<15:45,  2.01it/s]
 82%|████████▏ | 8783/10682 [1:15:03<15:44,  2.01it/s]
 82%|████████▏ | 8784/10682 [1:15:04<15:42,  2.01it/s]
 82%|████████▏ | 8785/10682 [1:15:04<15:43,  2.01it/s]
 82%|████████▏ | 8786/10682 [1:15:05<15:44,  2.01it/s]
 82%|████████▏ | 8787/10682 [1:15:05<15:44,  2.01it/s]
 82%|████████▏ | 8788/10682 [1:15:06<15:42,  2.01it/s]
 82%|████████▏ | 8789/10682 [1:15:06<15:41,  2.01it/s]
 82%|████████▏ | 8790/10682 [1:15:07<15:41,  2.01it/s]
 82%|████████▏ | 8791/10682 [1:15:07<15:39,  2.01it/s]
 82%|████████▏ | 8792/10682 [1:15:08<15:38,  2.01it/s]
 82%|████████▏ | 8793/10682 [1:15:08<15:37,  2.01it/s]
 82%|████████▏ | 8794/10682 [1:15:09<15:37,  2.01it/s]
 82%|████████▏ | 8795/10682 [1:15:09<15:38,  2.01it/s]
 82%|████████▏ | 8796/10682 [1:15:10<15:36,  2.01it/s]
 82%|████████▏ | 8797/10682 [1:15:10<15:37,  2.01it/s]
 82%|████████▏ | 8798/10682 [1:15:11<15:36,  2.01it/s]
 82%|████████▏ | 8799/10682 [1:15:11<15:33,  2.02it/s]
 82%|████████▏ | 8800/10682 [1:15:11<15:34,  2.01it/s]
                                                      {'loss': 3.4038, 'grad_norm': 0.21202631294727325, 'learning_rate': 9.162775276448015e-05, 'epoch': 0.82}
-
 82%|████████▏ | 8800/10682 [1:15:11<15:34,  2.01it/s]
 82%|████████▏ | 8801/10682 [1:15:12<15:35,  2.01it/s]
 82%|████████▏ | 8802/10682 [1:15:12<15:34,  2.01it/s]
 82%|████████▏ | 8803/10682 [1:15:13<15:32,  2.01it/s]
 82%|████████▏ | 8804/10682 [1:15:13<15:31,  2.02it/s]
 82%|████████▏ | 8805/10682 [1:15:14<15:33,  2.01it/s]
 82%|████████▏ | 8806/10682 [1:15:14<15:34,  2.01it/s]
 82%|████████▏ | 8807/10682 [1:15:15<15:33,  2.01it/s]
 82%|████████▏ | 8808/10682 [1:15:15<15:32,  2.01it/s]
 82%|████████▏ | 8809/10682 [1:15:16<15:31,  2.01it/s]
 82%|████████▏ | 8810/10682 [1:15:16<15:30,  2.01it/s]
 82%|████████▏ | 8811/10682 [1:15:17<15:30,  2.01it/s]
 82%|████████▏ | 8812/10682 [1:15:17<15:28,  2.01it/s]
 83%|████████▎ | 8813/10682 [1:15:18<15:28,  2.01it/s]
 83%|████████▎ | 8814/10682 [1:15:18<15:28,  2.01it/s]
 83%|████████▎ | 8815/10682 [1:15:19<15:27,  2.01it/s]
 83%|████████▎ | 8816/10682 [1:15:19<15:27,  2.01it/s]
 83%|████████▎ | 8817/10682 [1:15:20<15:25,  2.02it/s]
 83%|████████▎ | 8818/10682 [1:15:20<15:25,  2.01it/s]
 83%|████████▎ | 8819/10682 [1:15:21<15:25,  2.01it/s]
 83%|████████▎ | 8820/10682 [1:15:21<15:24,  2.01it/s]
 83%|��███████▎ | 8821/10682 [1:15:22<15:23,  2.01it/s]
 83%|████████▎ | 8822/10682 [1:15:22<15:23,  2.01it/s]
 83%|████████▎ | 8823/10682 [1:15:23<15:22,  2.02it/s]
 83%|████████▎ | 8824/10682 [1:15:23<15:21,  2.02it/s]
 83%|████████▎ | 8825/10682 [1:15:24<15:22,  2.01it/s]
                                                      
{'loss': 3.3976, 'grad_norm': 0.22617369890213013, 'learning_rate': 8.928431747505355e-05, 'epoch': 0.83}
-
 83%|████████▎ | 8825/10682 [1:15:24<15:22,  2.01it/s]
 83%|████████▎ | 8826/10682 [1:15:24<15:24,  2.01it/s]
 83%|████████▎ | 8827/10682 [1:15:25<15:22,  2.01it/s]
 83%|████████▎ | 8828/10682 [1:15:25<15:21,  2.01it/s]
 83%|████████▎ | 8829/10682 [1:15:26<15:20,  2.01it/s]
 83%|████████▎ | 8830/10682 [1:15:26<15:20,  2.01it/s]
 83%|████████▎ | 8831/10682 [1:15:27<15:20,  2.01it/s]
 83%|████████▎ | 8832/10682 [1:15:27<15:17,  2.02it/s]
 83%|████████▎ | 8833/10682 [1:15:28<15:17,  2.01it/s]
 83%|████████▎ | 8834/10682 [1:15:28<15:16,  2.02it/s]
 83%|████████▎ | 8835/10682 [1:15:29<15:17,  2.01it/s]
 83%|████████▎ | 8836/10682 [1:15:29<15:17,  2.01it/s]
 83%|████████▎ | 8837/10682 [1:15:30<15:15,  2.02it/s]
 83%|████████▎ | 8838/10682 [1:15:30<15:17,  2.01it/s]
 83%|████████▎ | 8839/10682 [1:15:31<15:15,  2.01it/s]
 83%|████████▎ | 8840/10682 [1:15:31<15:14,  2.01it/s]
 83%|████████▎ | 8841/10682 [1:15:32<15:14,  2.01it/s]
 83%|████████▎ | 8842/10682 [1:15:32<15:12,  2.02it/s]
 83%|████████▎ | 8843/10682 [1:15:33<15:13,  2.01it/s]
 83%|████████▎ | 8844/10682 [1:15:33<15:13,  2.01it/s]
 83%|████████▎ | 8845/10682 [1:15:34<15:12,  2.01it/s]
 83%|████████▎ | 8846/10682 [1:15:34<15:13,  2.01it/s]
 83%|████████▎ | 8847/10682 [1:15:35<15:11,  2.01it/s]
 83%|████████▎ | 8848/10682 [1:15:35<15:12,  2.01it/s]
 83%|████████▎ | 8849/10682 [1:15:36<15:10,  2.01it/s]
 83%|████████▎ | 8850/10682 [1:15:36<15:09,  2.01it/s]{'loss': 3.3908, 'grad_norm': 0.21251575648784637, 'learning_rate': 8.6968297974584e-05, 'epoch': 0.83}
                                                      
-
 83%|████████▎ | 8850/10682 [1:15:36<15:09,  2.01it/s]
 83%|████████▎ | 8851/10682 [1:15:37<15:11,  2.01it/s]
 83%|████████▎ | 8852/10682 [1:15:37<15:09,  2.01it/s]
 83%|████████▎ | 8853/10682 [1:15:38<15:09,  2.01it/s]
 83%|████████▎ | 8854/10682 [1:15:38<15:08,  2.01it/s]
 83%|████████▎ | 8855/10682 [1:15:39<15:10,  2.01it/s]
 83%|████████▎ | 8856/10682 [1:15:39<15:08,  2.01it/s]
 83%|████████▎ | 8857/10682 [1:15:40<15:06,  2.01it/s]
 83%|████████▎ | 8858/10682 [1:15:40<15:05,  2.01it/s]
 83%|████████▎ | 8859/10682 [1:15:41<15:04,  2.02it/s]
 83%|████████▎ | 8860/10682 [1:15:41<15:02,  2.02it/s]
 83%|████████▎ | 8861/10682 [1:15:42<15:02,  2.02it/s]
 83%|████████▎ | 8862/10682 [1:15:42<15:02,  2.02it/s]
 83%|████████▎ | 8863/10682 [1:15:43<15:02,  2.01it/s]
 83%|████████▎ | 8864/10682 [1:15:43<15:02,  2.01it/s]
 83%|████████▎ | 8865/10682 [1:15:44<15:03,  2.01it/s]
 83%|████████▎ | 8866/10682 [1:15:44<15:03,  2.01it/s]
 83%|████████▎ | 8867/10682 [1:15:45<15:01,  2.01it/s]
 83%|████████▎ | 8868/10682 [1:15:45<15:00,  2.01it/s]
 83%|████████▎ | 8869/10682 [1:15:46<14:59,  2.02it/s]
 83%|████████▎ | 8870/10682 [1:15:46<14:59,  2.01it/s]
 83%|████████▎ | 8871/10682 [1:15:47<14:59,  2.01it/s]
 83%|████████▎ | 8872/10682 [1:15:47<14:58,  2.02it/s]
 83%|████████▎ | 8873/10682 [1:15:48<14:58,  2.01it/s]
 83%|████████▎ | 8874/10682 [1:15:48<14:57,  2.01it/s]
 83%|████████▎ | 8875/10682 [1:15:49<14:58,  2.01it/s]
                                                      {'loss': 3.3848, 'grad_norm': 0.21532462537288666, 'learning_rate': 8.467984886028967e-05, 'epoch': 0.83}
-
 83%|████████▎ | 8875/10682 [1:15:49<14:58,  2.01it/s]
 83%|████████▎ | 8876/10682 [1:15:49<15:01,  2.00it/s]
 83%|████████▎ | 8877/10682 [1:15:50<14:59,  2.01it/s]
 83%|████████▎ | 8878/10682 [1:15:50<14:57,  2.01it/s]
 83%|████████▎ | 8879/10682 [1:15:51<14:56,  2.01it/s]
 83%|████████▎ | 8880/10682 [1:15:51<14:54,  2.01it/s]
 83%|████████▎ | 8881/10682 [1:15:52<14:54,  2.01it/s]
 83%|████████▎ | 8882/10682 [1:15:52<14:53,  2.02it/s]
 83%|████████▎ | 8883/10682 [1:15:53<14:53,  2.01it/s]
 83%|████████▎ | 8884/10682 [1:15:53<14:53,  2.01it/s]
 83%|████████▎ | 8885/10682 [1:15:54<14:52,  2.01it/s]
 83%|████████▎ | 8886/10682 [1:15:54<14:52,  2.01it/s]
 83%|████████▎ | 8887/10682 [1:15:55<14:50,  2.02it/s]
 83%|████████▎ | 8888/10682 [1:15:55<14:50,  2.01it/s]
 83%|████████▎ | 8889/10682 [1:15:56<14:48,  2.02it/s]
 83%|████████▎ | 8890/10682 [1:15:56<14:48,  2.02it/s]
 83%|████████▎ | 8891/10682 [1:15:57<14:46,  2.02it/s]
 83%|████████▎ | 8892/10682 [1:15:57<14:47,  2.02it/s]
 83%|████████▎ | 8893/10682 [1:15:58<14:47,  2.02it/s]
 83%|████████▎ | 8894/10682 [1:15:58<14:46,  2.02it/s]
 83%|████████▎ | 8895/10682 [1:15:59<14:45,  2.02it/s]
 83%|████████▎ | 8896/10682 [1:15:59<14:45,  2.02it/s]
 83%|████████▎ | 8897/10682 [1:16:00<14:46,  2.01it/s]
 83%|████████▎ | 8898/10682 [1:16:00<14:45,  2.02it/s]
 83%|████████▎ | 8899/10682 [1:16:01<14:45,  2.01it/s]
 83%|████████▎ | 8900/10682 [1:16:01<14:44,  2.01it/s]{'loss': 3.3948, 'grad_norm': 0.22366255521774292, 'learning_rate': 8.24191228890303e-05, 'epoch': 0.83}
-
                                                      

 83%|████████▎ | 8900/10682 [1:16:01<14:44,  2.01it/s]
 83%|████████▎ | 8901/10682 [1:16:02<17:12,  1.72it/s]
 83%|████████▎ | 8902/10682 [1:16:02<16:26,  1.80it/s]
 83%|████████▎ | 8903/10682 [1:16:03<15:56,  1.86it/s]
 83%|████████▎ | 8904/10682 [1:16:03<15:33,  1.90it/s]
 83%|████████▎ | 8905/10682 [1:16:04<15:17,  1.94it/s]
 83%|████████▎ | 8906/10682 [1:16:04<15:06,  1.96it/s]
 83%|████████▎ | 8907/10682 [1:16:05<14:58,  1.98it/s]
 83%|████████▎ | 8908/10682 [1:16:05<14:52,  1.99it/s]
 83%|████████▎ | 8909/10682 [1:16:06<14:48,  1.99it/s]
 83%|████████▎ | 8910/10682 [1:16:06<14:45,  2.00it/s]
 83%|████████▎ | 8911/10682 [1:16:07<14:43,  2.00it/s]
 83%|████████▎ | 8912/10682 [1:16:07<14:42,  2.01it/s]
 83%|████████▎ | 8913/10682 [1:16:08<14:40,  2.01it/s]
 83%|████████▎ | 8914/10682 [1:16:08<14:38,  2.01it/s]
 83%|████████▎ | 8915/10682 [1:16:09<14:38,  2.01it/s]
 83%|████████▎ | 8916/10682 [1:16:09<14:37,  2.01it/s]
 83%|████████▎ | 8917/10682 [1:16:10<14:36,  2.01it/s]
 83%|████████▎ | 8918/10682 [1:16:10<14:34,  2.02it/s]
 83%|████████▎ | 8919/10682 [1:16:11<14:34,  2.02it/s]
 84%|████████▎ | 8920/10682 [1:16:11<14:33,  2.02it/s]
 84%|████████▎ | 8921/10682 [1:16:12<14:34,  2.01it/s]
 84%|████████▎ | 8922/10682 [1:16:12<14:32,  2.02it/s]
 84%|████████▎ | 8923/10682 [1:16:13<14:32,  2.02it/s]
 84%|████████▎ | 8924/10682 [1:16:13<14:33,  2.01it/s]
 84%|████████▎ | 8925/10682 [1:16:14<14:31,  2.01it/s]
                                                      {'loss': 3.3943, 'grad_norm': 0.21397025883197784, 'learning_rate': 8.018627096711106e-05, 'epoch': 0.84}
-
 84%|████████▎ | 8925/10682 [1:16:14<14:31,  2.01it/s]
 84%|████████▎ | 8926/10682 [1:16:14<14:34,  2.01it/s]
 84%|████████▎ | 8927/10682 [1:16:15<14:33,  2.01it/s]
 84%|████████▎ | 8928/10682 [1:16:15<14:31,  2.01it/s]
 84%|████████▎ | 8929/10682 [1:16:16<14:30,  2.01it/s]
 84%|████████▎ | 8930/10682 [1:16:16<14:29,  2.02it/s]
 84%|████████▎ | 8931/10682 [1:16:17<14:28,  2.02it/s]
 84%|████████▎ | 8932/10682 [1:16:17<14:27,  2.02it/s]
 84%|████████▎ | 8933/10682 [1:16:18<14:28,  2.01it/s]
 84%|████████▎ | 8934/10682 [1:16:18<14:28,  2.01it/s]
 84%|████████▎ | 8935/10682 [1:16:19<14:27,  2.01it/s]
 84%|████████▎ | 8936/10682 [1:16:19<14:27,  2.01it/s]
 84%|████████▎ | 8937/10682 [1:16:20<14:26,  2.01it/s]
 84%|████████▎ | 8938/10682 [1:16:21<16:49,  1.73it/s]
 84%|████████▎ | 8939/10682 [1:16:21<16:05,  1.80it/s]
 84%|████████▎ | 8940/10682 [1:16:22<15:34,  1.86it/s]
 84%|████████▎ | 8941/10682 [1:16:22<15:13,  1.91it/s]
 84%|████████▎ | 8942/10682 [1:16:23<14:59,  1.94it/s]
 84%|████████▎ | 8943/10682 [1:16:23<14:47,  1.96it/s]
 84%|████████▎ | 8944/10682 [1:16:24<14:41,  1.97it/s]
 84%|████���███▎ | 8945/10682 [1:16:24<14:35,  1.98it/s]
 84%|████████▎ | 8946/10682 [1:16:25<14:30,  1.99it/s]
 84%|████████▍ | 8947/10682 [1:16:25<14:27,  2.00it/s]
 84%|████████▍ | 8948/10682 [1:16:26<14:23,  2.01it/s]
 84%|████████▍ | 8949/10682 [1:16:26<14:22,  2.01it/s]
 84%|████████▍ | 8950/10682 [1:16:27<14:20,  2.01it/s]{'loss': 3.3819, 'grad_norm': 0.21692153811454773, 'learning_rate': 7.798144214020909e-05, 'epoch': 0.84}
-
                                                      

 84%|████████▍ | 8950/10682 [1:16:27<14:20,  2.01it/s]
 84%|████████▍ | 8951/10682 [1:16:27<14:21,  2.01it/s]
 84%|████████▍ | 8952/10682 [1:16:28<14:20,  2.01it/s]
 84%|████████▍ | 8953/10682 [1:16:28<14:19,  2.01it/s]
 84%|████████▍ | 8954/10682 [1:16:29<14:19,  2.01it/s]
 84%|████████▍ | 8955/10682 [1:16:29<14:17,  2.01it/s]
 84%|████████▍ | 8956/10682 [1:16:30<14:17,  2.01it/s]
 84%|████████▍ | 8957/10682 [1:16:30<14:15,  2.02it/s]
 84%|████████▍ | 8958/10682 [1:16:31<14:14,  2.02it/s]
 84%|████████▍ | 8959/10682 [1:16:31<14:13,  2.02it/s]
 84%|████████▍ | 8960/10682 [1:16:32<14:13,  2.02it/s]
 84%|████████▍ | 8961/10682 [1:16:32<14:13,  2.02it/s]
 84%|████████▍ | 8962/10682 [1:16:32<14:12,  2.02it/s]
 84%|████████▍ | 8963/10682 [1:16:33<14:14,  2.01it/s]
 84%|████████▍ | 8964/10682 [1:16:33<14:12,  2.01it/s]
 84%|████████▍ | 8965/10682 [1:16:34<14:12,  2.01it/s]
 84%|████████▍ | 8966/10682 [1:16:34<14:11,  2.02it/s]
 84%|████████▍ | 8967/10682 [1:16:35<14:10,  2.02it/s]
 84%|████████▍ | 8968/10682 [1:16:35<14:10,  2.02it/s]
 84%|████████▍ | 8969/10682 [1:16:36<14:09,  2.02it/s]
 84%|████████▍ | 8970/10682 [1:16:36<14:09,  2.02it/s]
 84%|████████▍ | 8971/10682 [1:16:37<14:08,  2.02it/s]
 84%|████████▍ | 8972/10682 [1:16:37<14:07,  2.02it/s]
 84%|████████▍ | 8973/10682 [1:16:38<14:07,  2.02it/s]
 84%|████████▍ | 8974/10682 [1:16:38<14:07,  2.02it/s]
 84%|████████▍ | 8975/10682 [1:16:39<14:07,  2.01it/s]
                                                      {'loss': 3.3849, 'grad_norm': 0.2176024615764618, 'learning_rate': 7.58047835834249e-05, 'epoch': 0.84}
-
 84%|████████▍ | 8975/10682 [1:16:39<14:07,  2.01it/s]
 84%|████████▍ | 8976/10682 [1:16:39<14:07,  2.01it/s]
 84%|████████▍ | 8977/10682 [1:16:40<14:06,  2.01it/s]
 84%|████████▍ | 8978/10682 [1:16:40<14:04,  2.02it/s]
 84%|████████▍ | 8979/10682 [1:16:41<14:04,  2.02it/s]
 84%|████████▍ | 8980/10682 [1:16:41<14:03,  2.02it/s]
 84%|████████▍ | 8981/10682 [1:16:42<14:03,  2.02it/s]
 84%|████████▍ | 8982/10682 [1:16:42<14:04,  2.01it/s]
 84%|████████▍ | 8983/10682 [1:16:43<14:03,  2.01it/s]
 84%|████████▍ | 8984/10682 [1:16:43<14:03,  2.01it/s]
 84%|████████▍ | 8985/10682 [1:16:44<14:03,  2.01it/s]
 84%|████████▍ | 8986/10682 [1:16:44<14:03,  2.01it/s]
 84%|████████▍ | 8987/10682 [1:16:45<14:00,  2.02it/s]
 84%|████████▍ | 8988/10682 [1:16:45<14:00,  2.02it/s]
 84%|████████▍ | 8989/10682 [1:16:46<13:58,  2.02it/s]
 84%|████████▍ | 8990/10682 [1:16:46<13:58,  2.02it/s]
 84%|████████▍ | 8991/10682 [1:16:47<13:57,  2.02it/s]
 84%|████████▍ | 8992/10682 [1:16:47<13:57,  2.02it/s]
 84%|████████▍ | 8993/10682 [1:16:48<13:57,  2.02it/s]
 84%|████████▍ | 8994/10682 [1:16:48<13:57,  2.02it/s]
 84%|████████▍ | 8995/10682 [1:16:49<13:58,  2.01it/s]
 84%|████████▍ | 8996/10682 [1:16:49<13:57,  2.01it/s]
 84%|████████▍ | 8997/10682 [1:16:50<13:57,  2.01it/s]
 84%|████████▍ | 8998/10682 [1:16:50<13:56,  2.01it/s]
 84%|████████▍ | 8999/10682 [1:16:51<13:56,  2.01it/s]
 84%|████████▍ | 9000/10682 [1:16:51<13:56,  2.01it/s]
                                                      
{'loss': 3.3845, 'grad_norm': 0.21368226408958435, 'learning_rate': 7.365644059145782e-05, 'epoch': 0.84}
-
 84%|████████▍ | 9000/10682 [1:16:51<13:56,  2.01it/s]
 84%|████████▍ | 9001/10682 [1:16:52<13:57,  2.01it/s]
 84%|████████▍ | 9002/10682 [1:16:52<13:56,  2.01it/s]
 84%|████████▍ | 9003/10682 [1:16:53<13:55,  2.01it/s]
 84%|████████▍ | 9004/10682 [1:16:53<13:53,  2.01it/s]
 84%|████████▍ | 9005/10682 [1:16:54<13:54,  2.01it/s]
 84%|████████▍ | 9006/10682 [1:16:54<13:51,  2.01it/s]
 84%|████████▍ | 9007/10682 [1:16:55<13:50,  2.02it/s]
 84%|████████▍ | 9008/10682 [1:16:55<13:50,  2.02it/s]
 84%|████████▍ | 9009/10682 [1:16:56<13:49,  2.02it/s]
 84%|████████▍ | 9010/10682 [1:16:56<13:50,  2.01it/s]
 84%|████████▍ | 9011/10682 [1:16:57<13:48,  2.02it/s]
 84%|████████▍ | 9012/10682 [1:16:57<13:50,  2.01it/s]
 84%|████████▍ | 9013/10682 [1:16:58<13:49,  2.01it/s]
 84%|████████▍ | 9014/10682 [1:16:58<13:48,  2.01it/s]
 84%|████████▍ | 9015/10682 [1:16:59<13:47,  2.01it/s]
 84%|████████▍ | 9016/10682 [1:16:59<13:46,  2.02it/s]
 84%|████████▍ | 9017/10682 [1:17:00<13:46,  2.02it/s]
 84%|████████▍ | 9018/10682 [1:17:00<13:45,  2.02it/s]
 84%|████████▍ | 9019/10682 [1:17:01<13:44,  2.02it/s]
 84%|████████▍ | 9020/10682 [1:17:01<13:44,  2.02it/s]
 84%|████████▍ | 9021/10682 [1:17:02<13:45,  2.01it/s]
 84%|████████▍ | 9022/10682 [1:17:02<13:47,  2.01it/s]
 84%|████████▍ | 9023/10682 [1:17:03<13:47,  2.01it/s]
 84%|████████▍ | 9024/10682 [1:17:03<13:45,  2.01it/s]
 84%|████████▍ | 9025/10682 [1:17:04<13:44,  2.01it/s]
                                                      {'loss': 3.3829, 'grad_norm': 0.21331189572811127, 'learning_rate': 7.153655656890773e-05, 'epoch': 0.84}
-

 84%|████████▍ | 9025/10682 [1:17:04<13:44,  2.01it/s]
 84%|████████▍ | 9026/10682 [1:17:04<13:45,  2.01it/s]
 85%|████████▍ | 9027/10682 [1:17:05<13:44,  2.01it/s]
 85%|████████▍ | 9028/10682 [1:17:05<13:43,  2.01it/s]
 85%|████████▍ | 9029/10682 [1:17:06<13:42,  2.01it/s]
 85%|████████▍ | 9030/10682 [1:17:06<13:42,  2.01it/s]
 85%|████████▍ | 9031/10682 [1:17:07<13:41,  2.01it/s]
 85%|████████▍ | 9032/10682 [1:17:07<13:41,  2.01it/s]
 85%|████████▍ | 9033/10682 [1:17:08<13:42,  2.00it/s]
 85%|████████▍ | 9034/10682 [1:17:08<13:40,  2.01it/s]
 85%|████████▍ | 9035/10682 [1:17:09<13:40,  2.01it/s]
 85%|████████▍ | 9036/10682 [1:17:09<13:39,  2.01it/s]
 85%|████████▍ | 9037/10682 [1:17:10<13:38,  2.01it/s]
 85%|████████▍ | 9038/10682 [1:17:10<13:38,  2.01it/s]
 85%|████████▍ | 9039/10682 [1:17:11<13:37,  2.01it/s]
 85%|████████▍ | 9040/10682 [1:17:11<13:37,  2.01it/s]
 85%|████████▍ | 9041/10682 [1:17:12<13:36,  2.01it/s]
 85%|████████▍ | 9042/10682 [1:17:12<13:35,  2.01it/s]
 85%|████████▍ | 9043/10682 [1:17:13<13:35,  2.01it/s]
 85%|████████▍ | 9044/10682 [1:17:13<13:34,  2.01it/s]
 85%|████████▍ | 9045/10682 [1:17:14<13:34,  2.01it/s]
 85%|████████▍ | 9046/10682 [1:17:14<13:32,  2.01it/s]
 85%|████████▍ | 9047/10682 [1:17:15<13:32,  2.01it/s]
 85%|████████▍ | 9048/10682 [1:17:15<13:30,  2.02it/s]
 85%|████████▍ | 9049/10682 [1:17:16<13:30,  2.01it/s]
 85%|████████▍ | 9050/10682 [1:17:16<13:30,  2.01it/s]
                                                      {'loss': 3.3963, 'grad_norm': 0.2156878262758255, 'learning_rate': 6.94452730207023e-05, 'epoch': 0.85}
-
 85%|████████▍ | 9050/10682 [1:17:16<13:30,  2.01it/s]
 85%|████████▍ | 9051/10682 [1:17:17<13:31,  2.01it/s]
 85%|████████▍ | 9052/10682 [1:17:17<13:30,  2.01it/s]
 85%|████████▍ | 9053/10682 [1:17:18<13:29,  2.01it/s]
 85%|████████▍ | 9054/10682 [1:17:18<13:29,  2.01it/s]
 85%|████████▍ | 9055/10682 [1:17:19<13:28,  2.01it/s]
 85%|████████▍ | 9056/10682 [1:17:19<13:27,  2.01it/s]
 85%|████████▍ | 9057/10682 [1:17:20<13:26,  2.02it/s]
 85%|████████▍ | 9058/10682 [1:17:20<13:25,  2.01it/s]
 85%|████████▍ | 9059/10682 [1:17:21<13:25,  2.02it/s]
 85%|████████▍ | 9060/10682 [1:17:21<13:26,  2.01it/s]
 85%|████████▍ | 9061/10682 [1:17:22<13:26,  2.01it/s]
 85%|████████▍ | 9062/10682 [1:17:22<13:25,  2.01it/s]
 85%|████████▍ | 9063/10682 [1:17:23<13:25,  2.01it/s]
 85%|████████▍ | 9064/10682 [1:17:23<13:23,  2.01it/s]
 85%|████████▍ | 9065/10682 [1:17:24<13:22,  2.01it/s]
 85%|████████▍ | 9066/10682 [1:17:24<13:21,  2.02it/s]
 85%|████████▍ | 9067/10682 [1:17:25<13:21,  2.02it/s]
 85%|████████▍ | 9068/10682 [1:17:25<13:21,  2.01it/s]
 85%|████████��� | 9069/10682 [1:17:26<13:20,  2.01it/s]
 85%|████████▍ | 9070/10682 [1:17:26<13:21,  2.01it/s]
 85%|████████▍ | 9071/10682 [1:17:27<13:20,  2.01it/s]
 85%|████████▍ | 9072/10682 [1:17:27<13:20,  2.01it/s]
 85%|████████▍ | 9073/10682 [1:17:28<13:19,  2.01it/s]
 85%|████████▍ | 9074/10682 [1:17:28<13:17,  2.02it/s]
 85%|████████▍ | 9075/10682 [1:17:29<13:17,  2.02it/s]
                                                      
{'loss': 3.3836, 'grad_norm': 0.22241434454917908, 'learning_rate': 6.738272954265156e-05, 'epoch': 0.85}
-
 85%|████████▍ | 9075/10682 [1:17:29<13:17,  2.02it/s]
 85%|████████▍ | 9076/10682 [1:17:29<13:17,  2.01it/s]
 85%|████████▍ | 9077/10682 [1:17:30<13:16,  2.01it/s]
 85%|████████▍ | 9078/10682 [1:17:30<13:17,  2.01it/s]
 85%|████████▍ | 9079/10682 [1:17:31<13:16,  2.01it/s]
 85%|████████▌ | 9080/10682 [1:17:31<13:16,  2.01it/s]
 85%|████████▌ | 9081/10682 [1:17:32<13:15,  2.01it/s]
 85%|████████▌ | 9082/10682 [1:17:32<13:14,  2.01it/s]
 85%|████████▌ | 9083/10682 [1:17:33<13:14,  2.01it/s]
 85%|████████▌ | 9084/10682 [1:17:33<13:13,  2.01it/s]
 85%|████████▌ | 9085/10682 [1:17:34<13:12,  2.01it/s]
 85%|████████▌ | 9086/10682 [1:17:34<13:11,  2.02it/s]
 85%|████████▌ | 9087/10682 [1:17:35<13:12,  2.01it/s]
 85%|████████▌ | 9088/10682 [1:17:35<13:12,  2.01it/s]
 85%|████████▌ | 9089/10682 [1:17:36<13:11,  2.01it/s]
 85%|████████▌ | 9090/10682 [1:17:36<13:11,  2.01it/s]
 85%|████████▌ | 9091/10682 [1:17:37<13:10,  2.01it/s]
 85%|████████▌ | 9092/10682 [1:17:37<13:09,  2.01it/s]
 85%|████████▌ | 9093/10682 [1:17:38<13:10,  2.01it/s]
 85%|████████▌ | 9094/10682 [1:17:38<13:08,  2.01it/s]
 85%|████████▌ | 9095/10682 [1:17:39<13:09,  2.01it/s]
 85%|████████▌ | 9096/10682 [1:17:39<13:08,  2.01it/s]
 85%|████████▌ | 9097/10682 [1:17:40<13:07,  2.01it/s]
 85%|████████▌ | 9098/10682 [1:17:40<13:07,  2.01it/s]
 85%|████████▌ | 9099/10682 [1:17:41<13:06,  2.01it/s]
 85%|████████▌ | 9100/10682 [1:17:41<13:07,  2.01it/s]
                                                      {'loss': 3.3883, 'grad_norm': 0.22165292501449585, 'learning_rate': 6.534906381212979e-05, 'epoch': 0.85}
-
 85%|████████▌ | 9100/10682 [1:17:41<13:07,  2.01it/s]
 85%|████████▌ | 9101/10682 [1:17:42<13:08,  2.01it/s]
 85%|████████▌ | 9102/10682 [1:17:42<13:06,  2.01it/s]
 85%|████████▌ | 9103/10682 [1:17:43<13:05,  2.01it/s]
 85%|████████▌ | 9104/10682 [1:17:43<13:04,  2.01it/s]
 85%|████████▌ | 9105/10682 [1:17:44<13:02,  2.01it/s]
 85%|████████▌ | 9106/10682 [1:17:44<13:02,  2.01it/s]
 85%|████████▌ | 9107/10682 [1:17:45<13:02,  2.01it/s]
 85%|████████▌ | 9108/10682 [1:17:45<13:01,  2.01it/s]
 85%|████████▌ | 9109/10682 [1:17:46<12:59,  2.02it/s]
 85%|████████▌ | 9110/10682 [1:17:46<13:00,  2.01it/s]
 85%|████████▌ | 9111/10682 [1:17:47<13:00,  2.01it/s]
 85%|████████▌ | 9112/10682 [1:17:47<13:00,  2.01it/s]
 85%|████████▌ | 9113/10682 [1:17:48<12:59,  2.01it/s]
 85%|████████▌ | 9114/10682 [1:17:48<12:58,  2.01it/s]
 85%|████████▌ | 9115/10682 [1:17:49<12:58,  2.01it/s]
 85%|████████▌ | 9116/10682 [1:17:49<12:57,  2.02it/s]
 85%|████████▌ | 9117/10682 [1:17:49<12:57,  2.01it/s]
 85%|████████▌ | 9118/10682 [1:17:50<12:56,  2.01it/s]
 85%|████████▌ | 9119/10682 [1:17:50<12:56,  2.01it/s]
 85%|████████▌ | 9120/10682 [1:17:51<12:57,  2.01it/s]
 85%|████████▌ | 9121/10682 [1:17:51<12:57,  2.01it/s]
 85%|████████▌ | 9122/10682 [1:17:52<12:56,  2.01it/s]
 85%|████████▌ | 9123/10682 [1:17:52<12:56,  2.01it/s]
 85%|████████▌ | 9124/10682 [1:17:53<12:54,  2.01it/s]
 85%|████████▌ | 9125/10682 [1:17:53<12:54,  2.01it/s]{'loss': 3.3798, 'grad_norm': 0.22148551046848297, 'learning_rate': 6.334441157888504e-05, 'epoch': 0.85}
-
                                                      

 85%|████████▌ | 9125/10682 [1:17:53<12:54,  2.01it/s]
 85%|████████▌ | 9126/10682 [1:17:54<12:56,  2.01it/s]
 85%|████████▌ | 9127/10682 [1:17:54<12:54,  2.01it/s]
 85%|████████▌ | 9128/10682 [1:17:55<12:53,  2.01it/s]
 85%|████████▌ | 9129/10682 [1:17:55<12:52,  2.01it/s]
 85%|████████▌ | 9130/10682 [1:17:56<12:51,  2.01it/s]
 85%|████████▌ | 9131/10682 [1:17:56<12:49,  2.01it/s]
 85%|████████▌ | 9132/10682 [1:17:57<12:50,  2.01it/s]
 85%|████████▌ | 9133/10682 [1:17:57<12:50,  2.01it/s]
 86%|████████▌ | 9134/10682 [1:17:58<12:48,  2.01it/s]
 86%|████████▌ | 9135/10682 [1:17:58<12:48,  2.01it/s]
 86%|████████▌ | 9136/10682 [1:17:59<12:47,  2.02it/s]
 86%|████████▌ | 9137/10682 [1:17:59<12:46,  2.01it/s]
 86%|████████▌ | 9138/10682 [1:18:00<12:46,  2.01it/s]
 86%|████████▌ | 9139/10682 [1:18:00<12:47,  2.01it/s]
 86%|████████▌ | 9140/10682 [1:18:01<12:46,  2.01it/s]
 86%|████████▌ | 9141/10682 [1:18:01<12:45,  2.01it/s]
 86%|████████▌ | 9142/10682 [1:18:02<12:43,  2.02it/s]
 86%|████████▌ | 9143/10682 [1:18:02<12:44,  2.01it/s]
 86%|████████▌ | 9144/10682 [1:18:03<12:43,  2.01it/s]
 86%|████████▌ | 9145/10682 [1:18:03<12:42,  2.02it/s]
 86%|████████▌ | 9146/10682 [1:18:04<12:42,  2.01it/s]
 86%|████████▌ | 9147/10682 [1:18:04<12:41,  2.02it/s]
 86%|████████▌ | 9148/10682 [1:18:05<12:42,  2.01it/s]
 86%|████████▌ | 9149/10682 [1:18:05<12:41,  2.01it/s]
 86%|████████▌ | 9150/10682 [1:18:06<12:40,  2.02it/s]{'loss': 3.3885, 'grad_norm': 0.22575706243515015, 'learning_rate': 6.1368906655978e-05, 'epoch': 0.86}
                                                      
-
 86%|████████▌ | 9150/10682 [1:18:06<12:40,  2.02it/s]
 86%|████████▌ | 9151/10682 [1:18:06<12:41,  2.01it/s]
 86%|████████▌ | 9152/10682 [1:18:07<12:40,  2.01it/s]
 86%|████████▌ | 9153/10682 [1:18:07<12:39,  2.01it/s]
 86%|████████▌ | 9154/10682 [1:18:08<12:39,  2.01it/s]
 86%|████████▌ | 9155/10682 [1:18:08<12:38,  2.01it/s]
 86%|████████▌ | 9156/10682 [1:18:09<12:38,  2.01it/s]
 86%|████████▌ | 9157/10682 [1:18:09<12:37,  2.01it/s]
 86%|████████▌ | 9158/10682 [1:18:10<12:38,  2.01it/s]
 86%|████████▌ | 9159/10682 [1:18:10<12:38,  2.01it/s]
 86%|████████▌ | 9160/10682 [1:18:11<12:35,  2.01it/s]
 86%|████████▌ | 9161/10682 [1:18:11<12:35,  2.01it/s]
 86%|████████▌ | 9162/10682 [1:18:12<12:35,  2.01it/s]
 86%|████████▌ | 9163/10682 [1:18:12<12:34,  2.01it/s]
 86%|████████▌ | 9164/10682 [1:18:13<12:34,  2.01it/s]
 86%|████████▌ | 9165/10682 [1:18:13<12:33,  2.01it/s]
 86%|████████▌ | 9166/10682 [1:18:14<12:33,  2.01it/s]
 86%|████████▌ | 9167/10682 [1:18:14<12:32,  2.01it/s]
 86%|████████▌ | 9168/10682 [1:18:15<12:32,  2.01it/s]
 86%|████████▌ | 9169/10682 [1:18:15<12:32,  2.01it/s]
 86%|████████▌ | 9170/10682 [1:18:16<12:32,  2.01it/s]
 86%|████████▌ | 9171/10682 [1:18:16<12:31,  2.01it/s]
 86%|████████▌ | 9172/10682 [1:18:17<12:29,  2.01it/s]
 86%|████████▌ | 9173/10682 [1:18:17<12:29,  2.01it/s]
 86%|████████▌ | 9174/10682 [1:18:18<12:29,  2.01it/s]
 86%|████████▌ | 9175/10682 [1:18:18<12:29,  2.01it/s]
                                                      
{'loss': 3.3883, 'grad_norm': 0.2147233933210373, 'learning_rate': 5.94226809108499e-05, 'epoch': 0.86}
-
 86%|████████▌ | 9175/10682 [1:18:18<12:29,  2.01it/s]
 86%|████████▌ | 9176/10682 [1:18:19<12:30,  2.01it/s]
 86%|████████▌ | 9177/10682 [1:18:19<12:29,  2.01it/s]
 86%|████████▌ | 9178/10682 [1:18:20<12:29,  2.01it/s]
 86%|████████▌ | 9179/10682 [1:18:20<12:28,  2.01it/s]
 86%|████████▌ | 9180/10682 [1:18:21<12:26,  2.01it/s]
 86%|████████▌ | 9181/10682 [1:18:21<12:26,  2.01it/s]
 86%|████████▌ | 9182/10682 [1:18:22<12:25,  2.01it/s]
 86%|████████▌ | 9183/10682 [1:18:22<12:25,  2.01it/s]
 86%|████████▌ | 9184/10682 [1:18:23<12:24,  2.01it/s]
 86%|████████▌ | 9185/10682 [1:18:23<12:24,  2.01it/s]
 86%|████████▌ | 9186/10682 [1:18:24<12:25,  2.01it/s]
 86%|████████▌ | 9187/10682 [1:18:24<12:24,  2.01it/s]
 86%|████████▌ | 9188/10682 [1:18:25<12:24,  2.01it/s]
 86%|████████▌ | 9189/10682 [1:18:25<12:23,  2.01it/s]
 86%|████████▌ | 9190/10682 [1:18:26<12:22,  2.01it/s]
 86%|████████▌ | 9191/10682 [1:18:26<12:22,  2.01it/s]
 86%|████████▌ | 9192/10682 [1:18:27<12:20,  2.01it/s]
 86%|████████▌ | 9193/10682 [1:18:27<12:20,  2.01it/s]
 86%|████████▌ | 9194/10682 [1:18:28<12:20,  2.01it/s]
 86%|████████▌ | 9195/10682 [1:18:28<12:20,  2.01it/s]
 86%|████████▌ | 9196/10682 [1:18:29<12:21,  2.00it/s]
 86%|████████▌ | 9197/10682 [1:18:29<12:21,  2.00it/s]
 86%|████████▌ | 9198/10682 [1:18:30<12:19,  2.01it/s]
 86%|████████▌ | 9199/10682 [1:18:30<12:19,  2.01it/s]
 86%|████████▌ | 9200/10682 [1:18:31<12:20,  2.00it/s]{'loss': 3.3821, 'grad_norm': 0.2127370685338974, 'learning_rate': 5.7505864256519716e-05, 'epoch': 0.86}
                                                      
-
 86%|████████▌ | 9200/10682 [1:18:31<12:20,  2.00it/s]
 86%|████████▌ | 9201/10682 [1:18:31<12:20,  2.00it/s]
 86%|████████▌ | 9202/10682 [1:18:32<12:18,  2.00it/s]
 86%|████████▌ | 9203/10682 [1:18:32<12:17,  2.01it/s]
 86%|████████▌ | 9204/10682 [1:18:33<12:15,  2.01it/s]
 86%|████████▌ | 9205/10682 [1:18:33<12:15,  2.01it/s]
 86%|████████▌ | 9206/10682 [1:18:34<12:14,  2.01it/s]
 86%|████████▌ | 9207/10682 [1:18:34<12:13,  2.01it/s]
 86%|████████▌ | 9208/10682 [1:18:35<12:13,  2.01it/s]
 86%|████████▌ | 9209/10682 [1:18:35<12:11,  2.01it/s]
 86%|████████▌ | 9210/10682 [1:18:36<12:11,  2.01it/s]
 86%|████████▌ | 9211/10682 [1:18:36<12:09,  2.02it/s]
 86%|████████▌ | 9212/10682 [1:18:37<12:10,  2.01it/s]
 86%|████████▌ | 9213/10682 [1:18:37<12:10,  2.01it/s]
 86%|████████▋ | 9214/10682 [1:18:38<12:09,  2.01it/s]
 86%|████████▋ | 9215/10682 [1:18:38<12:09,  2.01it/s]
 86%|████████▋ | 9216/10682 [1:18:39<12:09,  2.01it/s]
 86%|████████▋ | 9217/10682 [1:18:39<12:08,  2.01it/s]
 86%|████████▋ | 9218/10682 [1:18:40<12:08,  2.01it/s]
 86%|████████▋ | 9219/10682 [1:18:40<12:07,  2.01it/s]
 86%|████████▋ | 9220/10682 [1:18:41<12:06,  2.01it/s]
 86%|████████▋ | 9221/10682 [1:18:41<12:06,  2.01it/s]
 86%|████████▋ | 9222/10682 [1:18:42<12:05,  2.01it/s]
 86%|████████▋ | 9223/10682 [1:18:42<12:05,  2.01it/s]
 86%|████████▋ | 9224/10682 [1:18:43<12:06,  2.01it/s]
 86%|████████▋ | 9225/10682 [1:18:43<12:04,  2.01it/s]
                                                      {'loss': 3.3808, 'grad_norm': 0.21177856624126434, 'learning_rate': 5.561858464291258e-05, 'epoch': 0.86}
-

 86%|████████▋ | 9225/10682 [1:18:43<12:04,  2.01it/s]
 86%|████████▋ | 9226/10682 [1:18:44<12:06,  2.00it/s]
 86%|████████▋ | 9227/10682 [1:18:44<12:05,  2.01it/s]
 86%|████████▋ | 9228/10682 [1:18:45<12:03,  2.01it/s]
 86%|████████▋ | 9229/10682 [1:18:45<12:03,  2.01it/s]
 86%|████████▋ | 9230/10682 [1:18:46<12:01,  2.01it/s]
 86%|████████▋ | 9231/10682 [1:18:46<12:00,  2.01it/s]
 86%|████████▋ | 9232/10682 [1:18:47<11:59,  2.02it/s]
 86%|████████▋ | 9233/10682 [1:18:47<11:59,  2.01it/s]
 86%|████████▋ | 9234/10682 [1:18:48<12:00,  2.01it/s]
 86%|████████▋ | 9235/10682 [1:18:48<11:59,  2.01it/s]
 86%|████████▋ | 9236/10682 [1:18:49<11:59,  2.01it/s]
 86%|████████▋ | 9237/10682 [1:18:49<11:59,  2.01it/s]
 86%|████████▋ | 9238/10682 [1:18:50<11:58,  2.01it/s]
 86%|████████▋ | 9239/10682 [1:18:50<11:57,  2.01it/s]
 87%|████████▋ | 9240/10682 [1:18:51<11:57,  2.01it/s]
 87%|████████▋ | 9241/10682 [1:18:51<11:56,  2.01it/s]
 87%|████████▋ | 9242/10682 [1:18:52<11:56,  2.01it/s]
 87%|████████▋ | 9243/10682 [1:18:52<11:55,  2.01it/s]
 87%|████████▋ | 9244/10682 [1:18:53<11:54,  2.01it/s]
 87%|████████▋ | 9245/10682 [1:18:53<11:53,  2.01it/s]
 87%|████████▋ | 9246/10682 [1:18:54<11:53,  2.01it/s]
 87%|████████▋ | 9247/10682 [1:18:54<11:53,  2.01it/s]
 87%|████████▋ | 9248/10682 [1:18:55<11:52,  2.01it/s]
 87%|████████▋ | 9249/10682 [1:18:55<11:51,  2.01it/s]
 87%|████████▋ | 9250/10682 [1:18:56<11:51,  2.01it/s]{'loss': 3.3907, 'grad_norm': 0.21076208353042603, 'learning_rate': 5.3760968048319145e-05, 'epoch': 0.87}
                                                      
-
 87%|████████▋ | 9250/10682 [1:18:56<11:51,  2.01it/s]
 87%|████████▋ | 9251/10682 [1:18:56<11:51,  2.01it/s]
 87%|████████▋ | 9252/10682 [1:18:57<11:51,  2.01it/s]
 87%|████████▋ | 9253/10682 [1:18:57<11:50,  2.01it/s]
 87%|████████▋ | 9254/10682 [1:18:58<11:48,  2.01it/s]
 87%|████████▋ | 9255/10682 [1:18:58<11:48,  2.01it/s]
 87%|████████▋ | 9256/10682 [1:18:59<11:48,  2.01it/s]
 87%|████████▋ | 9257/10682 [1:18:59<11:47,  2.01it/s]
 87%|████████▋ | 9258/10682 [1:19:00<11:46,  2.01it/s]
 87%|████████▋ | 9259/10682 [1:19:00<11:46,  2.01it/s]
 87%|████████▋ | 9260/10682 [1:19:01<11:45,  2.02it/s]
 87%|████████▋ | 9261/10682 [1:19:01<11:45,  2.01it/s]
 87%|████████▋ | 9262/10682 [1:19:02<11:46,  2.01it/s]
 87%|████████▋ | 9263/10682 [1:19:02<11:46,  2.01it/s]
 87%|████████▋ | 9264/10682 [1:19:03<11:45,  2.01it/s]
 87%|████████▋ | 9265/10682 [1:19:03<11:45,  2.01it/s]
 87%|████████▋ | 9266/10682 [1:19:04<11:44,  2.01it/s]
 87%|████████▋ | 9267/10682 [1:19:04<11:43,  2.01it/s]
 87%|████████▋ | 9268/10682 [1:19:05<11:42,  2.01it/s]
 87%|████████▋ | 9269/10682 [1:19:05<11:42,  2.01it/s]
 87%|████████▋ | 9270/10682 [1:19:06<11:42,  2.01it/s]
 87%|████████▋ | 9271/10682 [1:19:06<11:41,  2.01it/s]
 87%|████████▋ | 9272/10682 [1:19:07<11:41,  2.01it/s]
 87%|████████▋ | 9273/10682 [1:19:07<11:41,  2.01it/s]
 87%|████████▋ | 9274/10682 [1:19:08<11:40,  2.01it/s]
 87%|████████▋ | 9275/10682 [1:19:08<11:41,  2.01it/s]
                                                      
{'loss': 3.3739, 'grad_norm': 0.212422713637352, 'learning_rate': 5.193313847098613e-05, 'epoch': 0.87}
-
 87%|████████▋ | 9275/10682 [1:19:08<11:41,  2.01it/s]
 87%|████████▋ | 9276/10682 [1:19:09<11:42,  2.00it/s]
 87%|████████▋ | 9277/10682 [1:19:09<11:41,  2.00it/s]
 87%|████████▋ | 9278/10682 [1:19:10<11:41,  2.00it/s]
 87%|████████▋ | 9279/10682 [1:19:10<11:39,  2.01it/s]
 87%|████████▋ | 9280/10682 [1:19:11<11:38,  2.01it/s]
 87%|████████▋ | 9281/10682 [1:19:11<11:38,  2.01it/s]
 87%|████████▋ | 9282/10682 [1:19:12<11:36,  2.01it/s]
 87%|████████▋ | 9283/10682 [1:19:12<11:36,  2.01it/s]
 87%|████████▋ | 9284/10682 [1:19:13<11:37,  2.00it/s]
 87%|████████▋ | 9285/10682 [1:19:13<11:37,  2.00it/s]
 87%|████████▋ | 9286/10682 [1:19:14<11:36,  2.01it/s]
 87%|████████▋ | 9287/10682 [1:19:14<11:35,  2.00it/s]
 87%|████████▋ | 9288/10682 [1:19:15<11:33,  2.01it/s]
 87%|████████▋ | 9289/10682 [1:19:15<11:32,  2.01it/s]
 87%|████████▋ | 9290/10682 [1:19:16<11:31,  2.01it/s]
 87%|████████▋ | 9291/10682 [1:19:16<11:30,  2.01it/s]
 87%|████████▋ | 9292/10682 [1:19:17<11:30,  2.01it/s]
 87%|████████▋ | 9293/10682 [1:19:17<11:29,  2.01it/s]
 87%|████████▋ | 9294/10682 [1:19:18<11:30,  2.01it/s]
 87%|████████▋ | 9295/10682 [1:19:18<11:29,  2.01it/s]
 87%|████████▋ | 9296/10682 [1:19:19<11:28,  2.01it/s]
 87%|████████▋ | 9297/10682 [1:19:19<11:29,  2.01it/s]
 87%|████████▋ | 9298/10682 [1:19:20<11:27,  2.01it/s]
 87%|████████▋ | 9299/10682 [1:19:20<11:27,  2.01it/s]
 87%|████████▋ | 9300/10682 [1:19:21<11:26,  2.01it/s]{'loss': 3.3831, 'grad_norm': 0.2099282294511795, 'learning_rate': 5.0135217920839137e-05, 'epoch': 0.87}
-
                                                      

 87%|████████▋ | 9300/10682 [1:19:21<11:26,  2.01it/s]
 87%|████████▋ | 9301/10682 [1:19:21<11:27,  2.01it/s]
 87%|████████▋ | 9302/10682 [1:19:22<11:27,  2.01it/s]
 87%|████████▋ | 9303/10682 [1:19:22<11:26,  2.01it/s]
 87%|████████▋ | 9304/10682 [1:19:23<11:26,  2.01it/s]
 87%|████████▋ | 9305/10682 [1:19:23<11:26,  2.01it/s]
 87%|████████▋ | 9306/10682 [1:19:24<11:26,  2.00it/s]
 87%|████████▋ | 9307/10682 [1:19:24<11:26,  2.00it/s]
 87%|████████▋ | 9308/10682 [1:19:25<11:24,  2.01it/s]
 87%|████████▋ | 9309/10682 [1:19:25<11:22,  2.01it/s]
 87%|████████▋ | 9310/10682 [1:19:25<11:22,  2.01it/s]
 87%|████████▋ | 9311/10682 [1:19:26<11:20,  2.01it/s]
 87%|████████▋ | 9312/10682 [1:19:26<11:20,  2.01it/s]
 87%|████████▋ | 9313/10682 [1:19:27<11:21,  2.01it/s]
 87%|████████▋ | 9314/10682 [1:19:27<11:20,  2.01it/s]
 87%|████████▋ | 9315/10682 [1:19:28<11:20,  2.01it/s]
 87%|████████▋ | 9316/10682 [1:19:28<11:18,  2.01it/s]
 87%|████████▋ | 9317/10682 [1:19:29<11:18,  2.01it/s]
 87%|████████▋ | 9318/10682 [1:19:29<11:16,  2.02it/s]
 87%|████████▋ | 9319/10682 [1:19:30<11:16,  2.01it/s]
 87%|████████▋ | 9320/10682 [1:19:30<11:15,  2.02it/s]
 87%|████████▋ | 9321/10682 [1:19:31<11:15,  2.01it/s]
 87%|████████▋ | 9322/10682 [1:19:31<11:15,  2.01it/s]
 87%|████████▋ | 9323/10682 [1:19:32<11:16,  2.01it/s]
 87%|████████▋ | 9324/10682 [1:19:32<11:16,  2.01it/s]
 87%|████████▋ | 9325/10682 [1:19:33<11:15,  2.01it/s]
                                                      
{'loss': 3.3893, 'grad_norm': 0.2133982628583908, 'learning_rate': 4.836732641133895e-05, 'epoch': 0.87}
-
 87%|████████▋ | 9325/10682 [1:19:33<11:15,  2.01it/s]
 87%|████████▋ | 9326/10682 [1:19:33<11:15,  2.01it/s]
 87%|████████▋ | 9327/10682 [1:19:34<11:14,  2.01it/s]
 87%|████████▋ | 9328/10682 [1:19:34<11:13,  2.01it/s]
 87%|████████▋ | 9329/10682 [1:19:35<11:12,  2.01it/s]
 87%|████████▋ | 9330/10682 [1:19:35<11:11,  2.01it/s]
 87%|████████▋ | 9331/10682 [1:19:36<11:12,  2.01it/s]
 87%|████████▋ | 9332/10682 [1:19:36<11:11,  2.01it/s]
 87%|████████▋ | 9333/10682 [1:19:37<11:11,  2.01it/s]
 87%|████████▋ | 9334/10682 [1:19:37<11:11,  2.01it/s]
 87%|████████▋ | 9335/10682 [1:19:38<11:08,  2.01it/s]
 87%|████████▋ | 9336/10682 [1:19:38<11:08,  2.01it/s]
 87%|████████▋ | 9337/10682 [1:19:39<11:09,  2.01it/s]
 87%|████████▋ | 9338/10682 [1:19:39<11:07,  2.01it/s]
 87%|████████▋ | 9339/10682 [1:19:40<11:07,  2.01it/s]
 87%|████████▋ | 9340/10682 [1:19:40<11:06,  2.01it/s]
 87%|████████▋ | 9341/10682 [1:19:41<11:06,  2.01it/s]
 87%|████████▋ | 9342/10682 [1:19:41<11:06,  2.01it/s]
 87%|████████▋ | 9343/10682 [1:19:42<11:06,  2.01it/s]
 87%|████████▋ | 9344/10682 [1:19:42<11:06,  2.01it/s]
 87%|████████▋ | 9345/10682 [1:19:43<11:06,  2.01it/s]
 87%|████████▋ | 9346/10682 [1:19:43<11:04,  2.01it/s]
 88%|████████▊ | 9347/10682 [1:19:44<11:03,  2.01it/s]
 88%|████████▊ | 9348/10682 [1:19:44<11:04,  2.01it/s]
 88%|████████▊ | 9349/10682 [1:19:45<11:03,  2.01it/s]
 88%|████████▊ | 9350/10682 [1:19:45<11:02,  2.01it/s]
                                                      {'loss': 3.3788, 'grad_norm': 0.2135440856218338, 'learning_rate': 4.662958195146971e-05, 'epoch': 0.88}
-
 88%|████████▊ | 9350/10682 [1:19:45<11:02,  2.01it/s]
 88%|████████▊ | 9351/10682 [1:19:46<11:05,  2.00it/s]
 88%|████████▊ | 9352/10682 [1:19:46<11:03,  2.01it/s]
 88%|████████▊ | 9353/10682 [1:19:47<11:02,  2.01it/s]
 88%|████████▊ | 9354/10682 [1:19:47<11:01,  2.01it/s]
 88%|████████▊ | 9355/10682 [1:19:48<11:00,  2.01it/s]
 88%|████████▊ | 9356/10682 [1:19:48<11:00,  2.01it/s]
 88%|████████▊ | 9357/10682 [1:19:49<10:59,  2.01it/s]
 88%|████████▊ | 9358/10682 [1:19:49<10:58,  2.01it/s]
 88%|████████▊ | 9359/10682 [1:19:50<10:59,  2.01it/s]
 88%|████████▊ | 9360/10682 [1:19:50<10:58,  2.01it/s]
 88%|████████▊ | 9361/10682 [1:19:51<10:57,  2.01it/s]
 88%|████████▊ | 9362/10682 [1:19:51<10:57,  2.01it/s]
 88%|████████▊ | 9363/10682 [1:19:52<10:56,  2.01it/s]
 88%|████████▊ | 9364/10682 [1:19:52<10:54,  2.01it/s]
 88%|████████▊ | 9365/10682 [1:19:53<10:56,  2.01it/s]
 88%|████████▊ | 9366/10682 [1:19:53<10:55,  2.01it/s]
 88%|████████▊ | 9367/10682 [1:19:54<10:54,  2.01it/s]
 88%|████████▊ | 9368/10682 [1:19:54<10:53,  2.01it/s]
 88%|████████▊ | 9369/10682 [1:19:55<10:52,  2.01it/s]
 88%|████████▊ | 9370/10682 [1:19:55<10:52,  2.01it/s]
 88%|████████▊ | 9371/10682 [1:19:56<10:52,  2.01it/s]
 88%|████████▊ | 9372/10682 [1:19:56<10:50,  2.01it/s]
 88%|████████▊ | 9373/10682 [1:19:57<10:50,  2.01it/s]
 88%|████████▊ | 9374/10682 [1:19:57<10:49,  2.01it/s]
 88%|████████▊ | 9375/10682 [1:19:58<10:49,  2.01it/s]{'loss': 3.378, 'grad_norm': 0.21319204568862915, 'learning_rate': 4.492210053786228e-05, 'epoch': 0.88}
-
                                                      

 88%|████████▊ | 9375/10682 [1:19:58<10:49,  2.01it/s]
 88%|████████▊ | 9376/10682 [1:19:58<10:50,  2.01it/s]
 88%|████████▊ | 9377/10682 [1:19:59<10:49,  2.01it/s]
 88%|██��█████▊ | 9378/10682 [1:19:59<10:49,  2.01it/s]
 88%|████████▊ | 9379/10682 [1:20:00<10:47,  2.01it/s]
 88%|████████▊ | 9380/10682 [1:20:00<10:47,  2.01it/s]
 88%|████████▊ | 9381/10682 [1:20:01<10:46,  2.01it/s]
 88%|████████▊ | 9382/10682 [1:20:01<10:45,  2.01it/s]
 88%|████████▊ | 9383/10682 [1:20:02<10:45,  2.01it/s]
 88%|████████▊ | 9384/10682 [1:20:02<10:44,  2.01it/s]
 88%|████████▊ | 9385/10682 [1:20:03<10:44,  2.01it/s]
 88%|████████▊ | 9386/10682 [1:20:03<10:43,  2.01it/s]
 88%|████████▊ | 9387/10682 [1:20:04<10:42,  2.02it/s]
 88%|████████▊ | 9388/10682 [1:20:04<10:42,  2.02it/s]
 88%|████████▊ | 9389/10682 [1:20:05<10:41,  2.02it/s]
 88%|████████▊ | 9390/10682 [1:20:05<10:40,  2.02it/s]
 88%|████████▊ | 9391/10682 [1:20:06<10:40,  2.02it/s]
 88%|████████▊ | 9392/10682 [1:20:06<10:40,  2.01it/s]
 88%|████████▊ | 9393/10682 [1:20:07<10:40,  2.01it/s]
 88%|████████▊ | 9394/10682 [1:20:07<10:39,  2.01it/s]
 88%|████████▊ | 9395/10682 [1:20:08<10:38,  2.02it/s]
 88%|████████▊ | 9396/10682 [1:20:08<10:38,  2.01it/s]
 88%|████████▊ | 9397/10682 [1:20:09<10:37,  2.02it/s]
 88%|████████▊ | 9398/10682 [1:20:09<10:37,  2.01it/s]
 88%|████████▊ | 9399/10682 [1:20:10<10:36,  2.02it/s]
 88%|████████▊ | 9400/10682 [1:20:10<10:35,  2.02it/s]{'loss': 3.3851, 'grad_norm': 0.2166880965232849, 'learning_rate': 4.3244996147050855e-05, 'epoch': 0.88}
-
                                                      

 88%|████████▊ | 9400/10682 [1:20:10<10:35,  2.02it/s]
 88%|████████▊ | 9401/10682 [1:20:11<10:36,  2.01it/s]
 88%|████████▊ | 9402/10682 [1:20:11<10:35,  2.01it/s]
 88%|████████▊ | 9403/10682 [1:20:12<10:34,  2.01it/s]
 88%|████████▊ | 9404/10682 [1:20:12<10:34,  2.01it/s]
 88%|████████▊ | 9405/10682 [1:20:13<10:33,  2.01it/s]
 88%|████████▊ | 9406/10682 [1:20:13<10:33,  2.01it/s]
 88%|████████▊ | 9407/10682 [1:20:14<10:32,  2.02it/s]
 88%|████████▊ | 9408/10682 [1:20:14<10:32,  2.02it/s]
 88%|████████▊ | 9409/10682 [1:20:15<10:30,  2.02it/s]
 88%|████████▊ | 9410/10682 [1:20:15<10:30,  2.02it/s]
 88%|████████▊ | 9411/10682 [1:20:16<10:30,  2.01it/s]
 88%|████████▊ | 9412/10682 [1:20:16<10:29,  2.02it/s]
 88%|████████▊ | 9413/10682 [1:20:17<10:29,  2.02it/s]
 88%|████████▊ | 9414/10682 [1:20:17<10:29,  2.01it/s]
 88%|████████▊ | 9415/10682 [1:20:18<10:29,  2.01it/s]
 88%|████████▊ | 9416/10682 [1:20:18<10:28,  2.02it/s]
 88%|████████▊ | 9417/10682 [1:20:19<10:28,  2.01it/s]
 88%|████████▊ | 9418/10682 [1:20:19<10:27,  2.01it/s]
 88%|████████▊ | 9419/10682 [1:20:20<10:26,  2.02it/s]
 88%|████████▊ | 9420/10682 [1:20:20<10:25,  2.02it/s]
 88%|████████▊ | 9421/10682 [1:20:21<10:24,  2.02it/s]
 88%|████████▊ | 9422/10682 [1:20:21<10:24,  2.02it/s]
 88%|████████▊ | 9423/10682 [1:20:22<10:23,  2.02it/s]
 88%|████████▊ | 9424/10682 [1:20:22<10:25,  2.01it/s]
 88%|████████▊ | 9425/10682 [1:20:23<10:24,  2.01it/s]{'loss': 3.3724, 'grad_norm': 0.21202267706394196, 'learning_rate': 4.1598380727865315e-05, 'epoch': 0.88}
                                                      
-
 88%|████████▊ | 9425/10682 [1:20:23<10:24,  2.01it/s]
 88%|████████▊ | 9426/10682 [1:20:23<10:23,  2.01it/s]
 88%|████████▊ | 9427/10682 [1:20:24<10:22,  2.02it/s]
 88%|████████▊ | 9428/10682 [1:20:24<10:22,  2.01it/s]
 88%|████████▊ | 9429/10682 [1:20:25<10:21,  2.02it/s]
 88%|████████▊ | 9430/10682 [1:20:25<10:21,  2.01it/s]
 88%|████████▊ | 9431/10682 [1:20:26<10:20,  2.02it/s]
 88%|████████▊ | 9432/10682 [1:20:26<10:20,  2.01it/s]
 88%|████████▊ | 9433/10682 [1:20:27<10:19,  2.02it/s]
 88%|████████▊ | 9434/10682 [1:20:27<10:18,  2.02it/s]
 88%|████████▊ | 9435/10682 [1:20:28<10:18,  2.02it/s]
 88%|████████▊ | 9436/10682 [1:20:28<10:17,  2.02it/s]
 88%|████████▊ | 9437/10682 [1:20:29<10:17,  2.02it/s]
 88%|████████▊ | 9438/10682 [1:20:29<10:16,  2.02it/s]
 88%|████████▊ | 9439/10682 [1:20:30<10:15,  2.02it/s]
 88%|████████▊ | 9440/10682 [1:20:30<10:15,  2.02it/s]
 88%|████████▊ | 9441/10682 [1:20:31<10:15,  2.02it/s]
 88%|████████▊ | 9442/10682 [1:20:31<10:15,  2.01it/s]
 88%|████████▊ | 9443/10682 [1:20:32<10:14,  2.02it/s]
 88%|████████▊ | 9444/10682 [1:20:32<10:14,  2.01it/s]
 88%|████████▊ | 9445/10682 [1:20:33<10:13,  2.02it/s]
 88%|████████▊ | 9446/10682 [1:20:33<10:13,  2.02it/s]
 88%|████████▊ | 9447/10682 [1:20:34<10:11,  2.02it/s]
 88%|████████▊ | 9448/10682 [1:20:34<10:11,  2.02it/s]
 88%|████████▊ | 9449/10682 [1:20:35<10:12,  2.01it/s]
 88%|████████▊ | 9450/10682 [1:20:35<10:11,  2.01it/s]
                                                      {'loss': 3.3815, 'grad_norm': 0.21253050863742828, 'learning_rate': 3.998236419395806e-05, 'epoch': 0.88}
-

 88%|████████▊ | 9450/10682 [1:20:35<10:11,  2.01it/s]
 88%|████████▊ | 9451/10682 [1:20:36<10:12,  2.01it/s]
 88%|████████▊ | 9452/10682 [1:20:36<10:10,  2.01it/s]
 88%|████████▊ | 9453/10682 [1:20:37<10:10,  2.01it/s]
 89%|████████▊ | 9454/10682 [1:20:37<10:08,  2.02it/s]
 89%|████████▊ | 9455/10682 [1:20:38<10:08,  2.02it/s]
 89%|████████▊ | 9456/10682 [1:20:38<10:09,  2.01it/s]
 89%|████████▊ | 9457/10682 [1:20:39<10:08,  2.01it/s]
 89%|████████▊ | 9458/10682 [1:20:39<10:08,  2.01it/s]
 89%|████████▊ | 9459/10682 [1:20:40<10:08,  2.01it/s]
 89%|████████▊ | 9460/10682 [1:20:40<10:06,  2.01it/s]
 89%|████████▊ | 9461/10682 [1:20:41<10:06,  2.01it/s]
 89%|████████▊ | 9462/10682 [1:20:41<10:04,  2.02it/s]
 89%|████████▊ | 9463/10682 [1:20:42<10:04,  2.02it/s]
 89%|████████▊ | 9464/10682 [1:20:42<10:04,  2.02it/s]
 89%|████████▊ | 9465/10682 [1:20:43<10:04,  2.01it/s]
 89%|████████▊ | 9466/10682 [1:20:43<10:04,  2.01it/s]
 89%|████████▊ | 9467/10682 [1:20:43<10:03,  2.01it/s]
 89%|████████▊ | 9468/10682 [1:20:44<10:02,  2.02it/s]
 89%|████████▊ | 9469/10682 [1:20:44<10:01,  2.02it/s]
 89%|████████▊ | 9470/10682 [1:20:45<10:00,  2.02it/s]
 89%|████████▊ | 9471/10682 [1:20:45<09:59,  2.02it/s]
 89%|████████▊ | 9472/10682 [1:20:46<09:59,  2.02it/s]
 89%|████████▊ | 9473/10682 [1:20:46<09:58,  2.02it/s]
 89%|████████▊ | 9474/10682 [1:20:47<09:59,  2.02it/s]
 89%|████████▊ | 9475/10682 [1:20:47<09:58,  2.02it/s]{'loss': 3.393, 'grad_norm': 0.21744774281978607, 'learning_rate': 3.839705441646779e-05, 'epoch': 0.89}
                                                      
-
 89%|████████▊ | 9475/10682 [1:20:47<09:58,  2.02it/s]
 89%|████████▊ | 9476/10682 [1:20:48<10:00,  2.01it/s]
 89%|████████▊ | 9477/10682 [1:20:48<09:59,  2.01it/s]
 89%|████████▊ | 9478/10682 [1:20:49<09:59,  2.01it/s]
 89%|████████▊ | 9479/10682 [1:20:49<09:57,  2.01it/s]
 89%|████████▊ | 9480/10682 [1:20:50<09:57,  2.01it/s]
 89%|████████▉ | 9481/10682 [1:20:50<09:56,  2.01it/s]
 89%|████████▉ | 9482/10682 [1:20:51<09:55,  2.01it/s]
 89%|████████▉ | 9483/10682 [1:20:51<09:54,  2.02it/s]
 89%|████████▉ | 9484/10682 [1:20:52<09:53,  2.02it/s]
 89%|████████▉ | 9485/10682 [1:20:52<09:52,  2.02it/s]
 89%|████████▉ | 9486/10682 [1:20:53<09:51,  2.02it/s]
 89%|████████▉ | 9487/10682 [1:20:53<09:52,  2.02it/s]
 89%|████████▉ | 9488/10682 [1:20:54<09:51,  2.02it/s]
 89%|████████▉ | 9489/10682 [1:20:54<09:52,  2.02it/s]
 89%|████████▉ | 9490/10682 [1:20:55<09:51,  2.02it/s]
 89%|████████▉ | 9491/10682 [1:20:55<09:50,  2.02it/s]
 89%|████████▉ | 9492/10682 [1:20:56<09:49,  2.02it/s]
 89%|████████▉ | 9493/10682 [1:20:56<09:48,  2.02it/s]
 89%|████████▉ | 9494/10682 [1:20:57<09:48,  2.02it/s]
 89%|████████▉ | 9495/10682 [1:20:57<09:48,  2.02it/s]
 89%|████████▉ | 9496/10682 [1:20:58<09:47,  2.02it/s]
 89%|████████▉ | 9497/10682 [1:20:58<09:47,  2.02it/s]
 89%|████████▉ | 9498/10682 [1:20:59<09:46,  2.02it/s]
 89%|████████▉ | 9499/10682 [1:20:59<09:46,  2.02it/s]
 89%|████████▉ | 9500/10682 [1:21:00<09:44,  2.02it/s]{'loss': 3.3842, 'grad_norm': 0.21674393117427826, 'learning_rate': 3.6842557216818006e-05, 'epoch': 0.89}
                                                      
-
 89%|████████▉ | 9500/10682 [1:21:00<09:44,  2.02it/s]
 89%|████████▉ | 9501/10682 [1:21:00<09:46,  2.01it/s]
 89%|████████▉ | 9502/10682 [1:21:01<09:46,  2.01it/s]
 89%|████████▉ | 9503/10682 [1:21:01<09:44,  2.02it/s]
 89%|████████▉ | 9504/10682 [1:21:02<09:43,  2.02it/s]
 89%|████████▉ | 9505/10682 [1:21:02<09:43,  2.02it/s]
 89%|████████▉ | 9506/10682 [1:21:03<09:42,  2.02it/s]
 89%|████████▉ | 9507/10682 [1:21:03<09:41,  2.02it/s]
 89%|████████▉ | 9508/10682 [1:21:04<09:40,  2.02it/s]
 89%|████████▉ | 9509/10682 [1:21:04<09:40,  2.02it/s]
 89%|████████▉ | 9510/10682 [1:21:05<09:40,  2.02it/s]
 89%|████████▉ | 9511/10682 [1:21:05<09:39,  2.02it/s]
 89%|████████▉ | 9512/10682 [1:21:06<09:40,  2.02it/s]
 89%|████████▉ | 9513/10682 [1:21:06<09:39,  2.02it/s]
 89%|████████▉ | 9514/10682 [1:21:07<09:39,  2.01it/s]
 89%|████████▉ | 9515/10682 [1:21:07<09:39,  2.01it/s]
 89%|████████▉ | 9516/10682 [1:21:08<09:38,  2.02it/s]
 89%|████████▉ | 9517/10682 [1:21:08<09:37,  2.02it/s]
 89%|████████▉ | 9518/10682 [1:21:09<09:37,  2.02it/s]
 89%|████████▉ | 9519/10682 [1:21:09<09:35,  2.02it/s]
 89%|████████▉ | 9520/10682 [1:21:10<09:35,  2.02it/s]
 89%|████████▉ | 9521/10682 [1:21:10<09:35,  2.02it/s]
 89%|████████▉ | 9522/10682 [1:21:11<09:34,  2.02it/s]
 89%|████████▉ | 9523/10682 [1:21:11<09:35,  2.01it/s]
 89%|████████▉ | 9524/10682 [1:21:12<09:34,  2.02it/s]
 89%|████████▉ | 9525/10682 [1:21:12<09:33,  2.02it/s]{'loss': 3.3747, 'grad_norm': 0.21040117740631104, 'learning_rate': 3.531897635965431e-05, 'epoch': 0.89}
                                                      
-
 89%|████████▉ | 9525/10682 [1:21:12<09:33,  2.02it/s]
 89%|████████▉ | 9526/10682 [1:21:13<09:34,  2.01it/s]
 89%|████████▉ | 9527/10682 [1:21:13<09:33,  2.01it/s]
 89%|████████▉ | 9528/10682 [1:21:14<09:31,  2.02it/s]
 89%|████████▉ | 9529/10682 [1:21:14<09:32,  2.02it/s]
 89%|████████▉ | 9530/10682 [1:21:15<09:31,  2.02it/s]
 89%|████████▉ | 9531/10682 [1:21:15<09:31,  2.02it/s]
 89%|████████▉ | 9532/10682 [1:21:16<09:30,  2.02it/s]
 89%|████████▉ | 9533/10682 [1:21:16<09:29,  2.02it/s]
 89%|████████▉ | 9534/10682 [1:21:17<09:29,  2.02it/s]
 89%|████████▉ | 9535/10682 [1:21:17<09:28,  2.02it/s]
 89%|████████▉ | 9536/10682 [1:21:18<09:27,  2.02it/s]
 89%|████████▉ | 9537/10682 [1:21:18<09:28,  2.02it/s]
 89%|████████▉ | 9538/10682 [1:21:19<09:27,  2.02it/s]
 89%|████████▉ | 9539/10682 [1:21:19<09:26,  2.02it/s]
 89%|████████▉ | 9540/10682 [1:21:20<09:26,  2.02it/s]
 89%|████████▉ | 9541/10682 [1:21:20<09:25,  2.02it/s]
 89%|████████▉ | 9542/10682 [1:21:21<09:25,  2.02it/s]
 89%|████████▉ | 9543/10682 [1:21:21<09:24,  2.02it/s]
 89%|████████▉ | 9544/10682 [1:21:22<09:24,  2.02it/s]
 89%|████████▉ | 9545/10682 [1:21:22<09:23,  2.02it/s]
 89%|████████▉ | 9546/10682 [1:21:23<09:23,  2.01it/s]
 89%|████████▉ | 9547/10682 [1:21:23<09:23,  2.02it/s]
 89%|████████▉ | 9548/10682 [1:21:24<09:22,  2.01it/s]
 89%|████████▉ | 9549/10682 [1:21:24<09:21,  2.02it/s]
 89%|████████▉ | 9550/10682 [1:21:25<09:21,  2.02it/s]{'loss': 3.3848, 'grad_norm': 0.21107974648475647, 'learning_rate': 3.382641354591731e-05, 'epoch': 0.89}
                                                      
-
 89%|████████▉ | 9550/10682 [1:21:25<09:21,  2.02it/s]
 89%|████████▉ | 9551/10682 [1:21:25<09:22,  2.01it/s]
 89%|████████▉ | 9552/10682 [1:21:26<09:21,  2.01it/s]
 89%|████████▉ | 9553/10682 [1:21:26<09:20,  2.01it/s]
 89%|████████▉ | 9554/10682 [1:21:27<09:19,  2.01it/s]
 89%|████████▉ | 9555/10682 [1:21:27<09:19,  2.01it/s]
 89%|████████▉ | 9556/10682 [1:21:28<09:18,  2.02it/s]
 89%|████████▉ | 9557/10682 [1:21:28<09:17,  2.02it/s]
 89%|████████▉ | 9558/10682 [1:21:29<09:17,  2.01it/s]
 89%|████████▉ | 9559/10682 [1:21:29<09:17,  2.02it/s]
 89%|████████▉ | 9560/10682 [1:21:30<09:16,  2.02it/s]
 90%|████████▉ | 9561/10682 [1:21:30<09:15,  2.02it/s]
 90%|████████▉ | 9562/10682 [1:21:31<09:15,  2.02it/s]
 90%|████████▉ | 9563/10682 [1:21:31<09:14,  2.02it/s]
 90%|████████▉ | 9564/10682 [1:21:32<09:15,  2.01it/s]
 90%|████████▉ | 9565/10682 [1:21:32<09:15,  2.01it/s]
 90%|████████▉ | 9566/10682 [1:21:33<09:14,  2.01it/s]
 90%|████████▉ | 9567/10682 [1:21:33<09:14,  2.01it/s]
 90%|████████▉ | 9568/10682 [1:21:34<09:12,  2.02it/s]
 90%|████████▉ | 9569/10682 [1:21:34<09:12,  2.02it/s]
 90%|████████▉ | 9570/10682 [1:21:35<09:11,  2.02it/s]
 90%|████████▉ | 9571/10682 [1:21:35<09:11,  2.02it/s]
 90%|████████▉ | 9572/10682 [1:21:36<09:11,  2.01it/s]
 90%|████████▉ | 9573/10682 [1:21:36<09:10,  2.02it/s]
 90%|████████▉ | 9574/10682 [1:21:37<09:10,  2.01it/s]
 90%|████████▉ | 9575/10682 [1:21:37<09:08,  2.02it/s]
                                                      
{'loss': 3.3801, 'grad_norm': 0.21092814207077026, 'learning_rate': 3.2364968406054075e-05, 'epoch': 0.9}
-
 90%|████████▉ | 9575/10682 [1:21:37<09:08,  2.02it/s]
 90%|████████▉ | 9576/10682 [1:21:38<09:09,  2.01it/s]
 90%|████████▉ | 9577/10682 [1:21:38<09:09,  2.01it/s]
 90%|████████▉ | 9578/10682 [1:21:39<09:08,  2.01it/s]
 90%|████████▉ | 9579/10682 [1:21:39<09:08,  2.01it/s]
 90%|████████▉ | 9580/10682 [1:21:40<09:07,  2.01it/s]
 90%|████████▉ | 9581/10682 [1:21:40<09:06,  2.01it/s]
 90%|████████▉ | 9582/10682 [1:21:41<09:05,  2.02it/s]
 90%|████████▉ | 9583/10682 [1:21:41<09:06,  2.01it/s]
 90%|████████▉ | 9584/10682 [1:21:42<09:05,  2.01it/s]
 90%|████████▉ | 9585/10682 [1:21:42<09:05,  2.01it/s]
 90%|████████▉ | 9586/10682 [1:21:43<09:03,  2.02it/s]
 90%|████████▉ | 9587/10682 [1:21:43<09:03,  2.01it/s]
 90%|████████▉ | 9588/10682 [1:21:44<09:02,  2.02it/s]
 90%|████████▉ | 9589/10682 [1:21:44<09:02,  2.01it/s]
 90%|████████▉ | 9590/10682 [1:21:45<09:01,  2.02it/s]
 90%|████████▉ | 9591/10682 [1:21:45<09:00,  2.02it/s]
 90%|████████▉ | 9592/10682 [1:21:45<08:59,  2.02it/s]
 90%|████████▉ | 9593/10682 [1:21:46<08:59,  2.02it/s]
 90%|████████▉ | 9594/10682 [1:21:46<08:58,  2.02it/s]
 90%|████████▉ | 9595/10682 [1:21:47<08:58,  2.02it/s]
 90%|████████▉ | 9596/10682 [1:21:47<08:58,  2.02it/s]
 90%|████████▉ | 9597/10682 [1:21:48<08:58,  2.01it/s]
 90%|████████▉ | 9598/10682 [1:21:48<08:57,  2.02it/s]
 90%|████████▉ | 9599/10682 [1:21:49<08:57,  2.02it/s]
 90%|████████▉ | 9600/10682 [1:21:49<08:56,  2.02it/s]
                                                      
{'loss': 3.3761, 'grad_norm': 0.2076018899679184, 'learning_rate': 3.093473849336781e-05, 'epoch': 0.9}
-
 90%|████████▉ | 9600/10682 [1:21:49<08:56,  2.02it/s]
 90%|████████▉ | 9601/10682 [1:21:50<08:57,  2.01it/s]
 90%|████████▉ | 9602/10682 [1:21:50<08:56,  2.01it/s]
 90%|████████▉ | 9603/10682 [1:21:51<08:56,  2.01it/s]
 90%|████████▉ | 9604/10682 [1:21:51<08:55,  2.01it/s]
 90%|████████▉ | 9605/10682 [1:21:52<08:55,  2.01it/s]
 90%|████████▉ | 9606/10682 [1:21:52<08:55,  2.01it/s]
 90%|████████▉ | 9607/10682 [1:21:53<08:55,  2.01it/s]
 90%|████████▉ | 9608/10682 [1:21:53<08:54,  2.01it/s]
 90%|████████▉ | 9609/10682 [1:21:54<08:53,  2.01it/s]
 90%|████████▉ | 9610/10682 [1:21:54<08:52,  2.01it/s]
 90%|████████▉ | 9611/10682 [1:21:55<08:51,  2.01it/s]
 90%|████████▉ | 9612/10682 [1:21:55<08:50,  2.02it/s]
 90%|████████▉ | 9613/10682 [1:21:56<08:50,  2.01it/s]
 90%|█████████ | 9614/10682 [1:21:56<08:49,  2.02it/s]
 90%|█████████ | 9615/10682 [1:21:57<08:49,  2.02it/s]
 90%|█████████ | 9616/10682 [1:21:58<10:14,  1.73it/s]
 90%|█████████ | 9617/10682 [1:21:58<09:48,  1.81it/s]
 90%|█████████ | 9618/10682 [1:21:59<09:29,  1.87it/s]
 90%|█████████ | 9619/10682 [1:21:59<09:16,  1.91it/s]
 90%|█████████ | 9620/10682 [1:22:00<09:06,  1.94it/s]
 90%|█████████ | 9621/10682 [1:22:00<09:00,  1.96it/s]
 90%|█████████ | 9622/10682 [1:22:01<08:55,  1.98it/s]
 90%|█████████ | 9623/10682 [1:22:01<08:51,  1.99it/s]
 90%|█████████ | 9624/10682 [1:22:02<08:49,  2.00it/s]
 90%|█████████ | 9625/10682 [1:22:02<08:47,  2.00it/s]{'loss': 3.3827, 'grad_norm': 0.21286582946777344, 'learning_rate': 2.9535819277506203e-05, 'epoch': 0.9}
                                                      
-
 90%|█████████ | 9625/10682 [1:22:02<08:47,  2.00it/s]
 90%|████████��� | 9626/10682 [1:22:03<08:47,  2.00it/s]
 90%|█████████ | 9627/10682 [1:22:03<08:45,  2.01it/s]
 90%|█████████ | 9628/10682 [1:22:04<08:44,  2.01it/s]
 90%|█████████ | 9629/10682 [1:22:04<08:42,  2.01it/s]
 90%|█████████ | 9630/10682 [1:22:05<08:42,  2.01it/s]
 90%|█████████ | 9631/10682 [1:22:05<08:41,  2.02it/s]
 90%|█████████ | 9632/10682 [1:22:06<08:40,  2.02it/s]
 90%|█████████ | 9633/10682 [1:22:06<08:39,  2.02it/s]
 90%|█████████ | 9634/10682 [1:22:07<08:39,  2.02it/s]
 90%|█████████ | 9635/10682 [1:22:07<08:38,  2.02it/s]
 90%|█████████ | 9636/10682 [1:22:08<08:38,  2.02it/s]
 90%|█████████ | 9637/10682 [1:22:08<08:37,  2.02it/s]
 90%|█████████ | 9638/10682 [1:22:09<08:37,  2.02it/s]
 90%|█████████ | 9639/10682 [1:22:09<08:36,  2.02it/s]
 90%|█████████ | 9640/10682 [1:22:10<08:36,  2.02it/s]
 90%|█████████ | 9641/10682 [1:22:10<08:36,  2.02it/s]
 90%|█████████ | 9642/10682 [1:22:11<08:35,  2.02it/s]
 90%|█████████ | 9643/10682 [1:22:11<08:35,  2.02it/s]
 90%|█████████ | 9644/10682 [1:22:12<08:34,  2.02it/s]
 90%|█████████ | 9645/10682 [1:22:12<08:34,  2.02it/s]
 90%|█████████ | 9646/10682 [1:22:13<08:33,  2.02it/s]
 90%|█████████ | 9647/10682 [1:22:13<08:32,  2.02it/s]
 90%|█████████ | 9648/10682 [1:22:14<08:31,  2.02it/s]
 90%|█████████ | 9649/10682 [1:22:14<08:31,  2.02it/s]
 90%|█████████ | 9650/10682 [1:22:15<08:30,  2.02it/s]{'loss': 3.3787, 'grad_norm': 0.20917737483978271, 'learning_rate': 2.8168304138088295e-05, 'epoch': 0.9}
                                                      
-
 90%|█████████ | 9650/10682 [1:22:15<08:30,  2.02it/s]
 90%|█████████ | 9651/10682 [1:22:15<08:31,  2.01it/s]
 90%|█████████ | 9652/10682 [1:22:16<08:31,  2.02it/s]
 90%|█████████ | 9653/10682 [1:22:16<08:30,  2.02it/s]
 90%|█████████ | 9654/10682 [1:22:17<08:29,  2.02it/s]
 90%|█████████ | 9655/10682 [1:22:17<08:28,  2.02it/s]
 90%|█████████ | 9656/10682 [1:22:18<08:27,  2.02it/s]
 90%|█████████ | 9657/10682 [1:22:18<08:27,  2.02it/s]
 90%|█████████ | 9658/10682 [1:22:19<09:52,  1.73it/s]
 90%|█████████ | 9659/10682 [1:22:19<09:26,  1.81it/s]
 90%|█████████ | 9660/10682 [1:22:20<09:07,  1.87it/s]
 90%|█████████ | 9661/10682 [1:22:20<08:54,  1.91it/s]
 90%|█████████ | 9662/10682 [1:22:21<08:45,  1.94it/s]
 90%|█████████ | 9663/10682 [1:22:21<08:38,  1.97it/s]
 90%|█████████ | 9664/10682 [1:22:22<08:34,  1.98it/s]
 90%|█████████ | 9665/10682 [1:22:22<08:31,  1.99it/s]
 90%|█████████ | 9666/10682 [1:22:23<08:28,  2.00it/s]
 90%|█████████ | 9667/10682 [1:22:23<08:26,  2.00it/s]
 91%|█████████ | 9668/10682 [1:22:24<08:24,  2.01it/s]
 91%|█████████ | 9669/10682 [1:22:24<08:23,  2.01it/s]
 91%|█████████ | 9670/10682 [1:22:25<08:22,  2.02it/s]
 91%|█████████ | 9671/10682 [1:22:25<08:21,  2.02it/s]
 91%|█████████ | 9672/10682 [1:22:26<08:20,  2.02it/s]
 91%|█████████ | 9673/10682 [1:22:26<08:20,  2.02it/s]
 91%|█████████ | 9674/10682 [1:22:27<08:19,  2.02it/s]
 91%|█████████ | 9675/10682 [1:22:27<08:19,  2.02it/s]{'loss': 3.3779, 'grad_norm': 0.21561506390571594, 'learning_rate': 2.6832284358471516e-05, 'epoch': 0.91}
                                                      
-
 91%|█████████ | 9675/10682 [1:22:27<08:19,  2.02it/s]
 91%|█████████ | 9676/10682 [1:22:28<08:18,  2.02it/s]
 91%|█████████ | 9677/10682 [1:22:28<08:18,  2.02it/s]
 91%|█████████ | 9678/10682 [1:22:29<08:17,  2.02it/s]
 91%|█████████ | 9679/10682 [1:22:29<08:16,  2.02it/s]
 91%|█████████ | 9680/10682 [1:22:30<08:16,  2.02it/s]
 91%|█████████ | 9681/10682 [1:22:30<08:16,  2.02it/s]
 91%|█████████ | 9682/10682 [1:22:31<08:15,  2.02it/s]
 91%|█████████ | 9683/10682 [1:22:31<08:15,  2.02it/s]
 91%|█████████ | 9684/10682 [1:22:32<08:14,  2.02it/s]
 91%|█████████ | 9685/10682 [1:22:32<08:13,  2.02it/s]
 91%|█████████ | 9686/10682 [1:22:33<08:12,  2.02it/s]
 91%|█████████ | 9687/10682 [1:22:33<08:12,  2.02it/s]
 91%|█████████ | 9688/10682 [1:22:34<08:12,  2.02it/s]
 91%|█████████ | 9689/10682 [1:22:34<08:12,  2.02it/s]
 91%|█████████ | 9690/10682 [1:22:35<08:11,  2.02it/s]
 91%|█████████ | 9691/10682 [1:22:35<08:11,  2.02it/s]
 91%|█████████ | 9692/10682 [1:22:36<08:10,  2.02it/s]
 91%|█████████ | 9693/10682 [1:22:36<08:09,  2.02it/s]
 91%|█████████ | 9694/10682 [1:22:37<08:08,  2.02it/s]
 91%|█████████ | 9695/10682 [1:22:37<08:08,  2.02it/s]
 91%|█████████ | 9696/10682 [1:22:38<08:08,  2.02it/s]
 91%|█████████ | 9697/10682 [1:22:38<08:08,  2.02it/s]
 91%|█████████ | 9698/10682 [1:22:39<08:07,  2.02it/s]
 91%|█████████ | 9699/10682 [1:22:39<08:06,  2.02it/s]
 91%|█████████ | 9700/10682 [1:22:40<08:06,  2.02it/s]{'loss': 3.3703, 'grad_norm': 0.20957434177398682, 'learning_rate': 2.5527849119658387e-05, 'epoch': 0.91}
                                                      
-
 91%|█████████ | 9700/10682 [1:22:40<08:06,  2.02it/s]
 91%|█████████ | 9701/10682 [1:22:40<08:07,  2.01it/s]
 91%|█████████ | 9702/10682 [1:22:41<08:06,  2.01it/s]
 91%|█████████ | 9703/10682 [1:22:41<08:06,  2.01it/s]
 91%|█████████ | 9704/10682 [1:22:42<08:04,  2.02it/s]
 91%|█████████ | 9705/10682 [1:22:42<08:05,  2.01it/s]
 91%|█████████ | 9706/10682 [1:22:43<08:04,  2.01it/s]
 91%|█████████ | 9707/10682 [1:22:43<08:04,  2.01it/s]
 91%|█████████ | 9708/10682 [1:22:44<08:03,  2.02it/s]
 91%|█████████ | 9709/10682 [1:22:44<08:03,  2.01it/s]
 91%|█████████ | 9710/10682 [1:22:45<08:01,  2.02it/s]
 91%|█████████ | 9711/10682 [1:22:45<08:01,  2.02it/s]
 91%|█████████ | 9712/10682 [1:22:46<08:00,  2.02it/s]
 91%|█████████ | 9713/10682 [1:22:46<08:00,  2.02it/s]
 91%|█████████ | 9714/10682 [1:22:47<07:59,  2.02it/s]
 91%|█████████ | 9715/10682 [1:22:47<07:58,  2.02it/s]
 91%|█████████ | 9716/10682 [1:22:48<07:58,  2.02it/s]
 91%|█████████ | 9717/10682 [1:22:48<07:58,  2.02it/s]
 91%|█████████ | 9718/10682 [1:22:49<07:57,  2.02it/s]
 91%|█████████ | 9719/10682 [1:22:49<07:57,  2.02it/s]
 91%|█████████ | 9720/10682 [1:22:49<07:56,  2.02it/s]
 91%|█████████ | 9721/10682 [1:22:50<07:56,  2.02it/s]
 91%|█████████ | 9722/10682 [1:22:50<07:55,  2.02it/s]
 91%|█████████ | 9723/10682 [1:22:51<07:56,  2.01it/s]
 91%|█████████ | 9724/10682 [1:22:51<07:55,  2.02it/s]
 91%|█████████ | 9725/10682 [1:22:52<07:54,  2.02it/s]{'loss': 3.3767, 'grad_norm': 0.21194203197956085, 'learning_rate': 2.4255085494343522e-05, 'epoch': 0.91}
                                                      
-
 91%|█████████ | 9725/10682 [1:22:52<07:54,  2.02it/s]
 91%|█████████ | 9726/10682 [1:22:52<07:55,  2.01it/s]
 91%|█████████ | 9727/10682 [1:22:53<07:54,  2.01it/s]
 91%|█████████ | 9728/10682 [1:22:53<07:54,  2.01it/s]
 91%|█████████ | 9729/10682 [1:22:54<07:53,  2.01it/s]
 91%|█████████ | 9730/10682 [1:22:54<07:52,  2.01it/s]
 91%|█████████ | 9731/10682 [1:22:55<07:51,  2.02it/s]
 91%|█████████ | 9732/10682 [1:22:55<07:51,  2.02it/s]
 91%|█████████ | 9733/10682 [1:22:56<07:50,  2.02it/s]
 91%|█████████ | 9734/10682 [1:22:56<07:49,  2.02it/s]
 91%|█████████ | 9735/10682 [1:22:57<07:49,  2.02it/s]
 91%|█████████ | 9736/10682 [1:22:57<07:49,  2.02it/s]
 91%|█████████ | 9737/10682 [1:22:58<07:48,  2.02it/s]
 91%|█████████ | 9738/10682 [1:22:58<07:48,  2.02it/s]
 91%|█████████ | 9739/10682 [1:22:59<07:47,  2.02it/s]
 91%|█████████ | 9740/10682 [1:22:59<07:46,  2.02it/s]
 91%|█████████ | 9741/10682 [1:23:00<07:46,  2.02it/s]
 91%|█████████ | 9742/10682 [1:23:00<07:45,  2.02it/s]
 91%|█████████ | 9743/10682 [1:23:01<07:45,  2.02it/s]
 91%|█████████ | 9744/10682 [1:23:01<07:44,  2.02it/s]
 91%|█████████ | 9745/10682 [1:23:02<07:43,  2.02it/s]
 91%|█████████ | 9746/10682 [1:23:02<07:43,  2.02it/s]
 91%|█████████ | 9747/10682 [1:23:03<07:42,  2.02it/s]
 91%|█████████▏| 9748/10682 [1:23:03<07:42,  2.02it/s]
 91%|█████████▏| 9749/10682 [1:23:04<07:41,  2.02it/s]
 91%|█████████▏| 9750/10682 [1:23:04<07:41,  2.02it/s]{'loss': 3.3774, 'grad_norm': 0.2090487778186798, 'learning_rate': 2.301407844110154e-05, 'epoch': 0.91}
                                                      
-
 91%|████████���▏| 9750/10682 [1:23:04<07:41,  2.02it/s]
 91%|█████████▏| 9751/10682 [1:23:05<07:41,  2.02it/s]
 91%|█████████▏| 9752/10682 [1:23:05<07:41,  2.02it/s]
 91%|█████████▏| 9753/10682 [1:23:06<07:40,  2.02it/s]
 91%|█████████▏| 9754/10682 [1:23:06<07:41,  2.01it/s]
 91%|█████████▏| 9755/10682 [1:23:07<07:40,  2.01it/s]
 91%|█████████▏| 9756/10682 [1:23:07<07:39,  2.02it/s]
 91%|█████████▏| 9757/10682 [1:23:08<07:38,  2.02it/s]
 91%|█████████▏| 9758/10682 [1:23:08<07:38,  2.02it/s]
 91%|█████████▏| 9759/10682 [1:23:09<07:37,  2.02it/s]
 91%|█████████▏| 9760/10682 [1:23:09<07:37,  2.02it/s]
 91%|█████████▏| 9761/10682 [1:23:10<07:36,  2.02it/s]
 91%|█████████▏| 9762/10682 [1:23:10<07:35,  2.02it/s]
 91%|█████████▏| 9763/10682 [1:23:11<07:35,  2.02it/s]
 91%|█████████▏| 9764/10682 [1:23:11<07:34,  2.02it/s]
 91%|█████████▏| 9765/10682 [1:23:12<07:34,  2.02it/s]
 91%|█████████▏| 9766/10682 [1:23:12<07:34,  2.02it/s]
 91%|█████████▏| 9767/10682 [1:23:13<07:33,  2.02it/s]
 91%|█████████▏| 9768/10682 [1:23:13<07:33,  2.02it/s]
 91%|█████████▏| 9769/10682 [1:23:14<07:32,  2.02it/s]
 91%|█████████▏| 9770/10682 [1:23:14<07:32,  2.01it/s]
 91%|█████████▏| 9771/10682 [1:23:15<07:32,  2.01it/s]
 91%|█████████▏| 9772/10682 [1:23:15<07:31,  2.01it/s]
 91%|█████████▏| 9773/10682 [1:23:16<07:31,  2.01it/s]
 91%|█████████▏| 9774/10682 [1:23:16<07:29,  2.02it/s]
 92%|█████████▏| 9775/10682 [1:23:17<07:29,  2.02it/s]
                                                      
{'loss': 3.3701, 'grad_norm': 0.21159473061561584, 'learning_rate': 2.1804910798715826e-05, 'epoch': 0.92}
-
 92%|█████████▏| 9775/10682 [1:23:17<07:29,  2.02it/s]
 92%|█████████▏| 9776/10682 [1:23:17<07:29,  2.01it/s]
 92%|█████████▏| 9777/10682 [1:23:18<07:29,  2.01it/s]
 92%|█████████▏| 9778/10682 [1:23:18<07:29,  2.01it/s]
 92%|█████████▏| 9779/10682 [1:23:19<07:28,  2.02it/s]
 92%|█████████▏| 9780/10682 [1:23:19<07:27,  2.02it/s]
 92%|█████████▏| 9781/10682 [1:23:20<07:26,  2.02it/s]
 92%|█████████▏| 9782/10682 [1:23:20<07:26,  2.02it/s]
 92%|█████████▏| 9783/10682 [1:23:21<07:25,  2.02it/s]
 92%|█████████▏| 9784/10682 [1:23:21<07:25,  2.02it/s]
 92%|█████████▏| 9785/10682 [1:23:22<07:24,  2.02it/s]
 92%|█████████▏| 9786/10682 [1:23:22<07:24,  2.01it/s]
 92%|█████████▏| 9787/10682 [1:23:23<07:25,  2.01it/s]
 92%|█████████▏| 9788/10682 [1:23:23<07:23,  2.01it/s]
 92%|█████████▏| 9789/10682 [1:23:24<07:23,  2.01it/s]
 92%|█████████▏| 9790/10682 [1:23:24<07:21,  2.02it/s]
 92%|█████████▏| 9791/10682 [1:23:25<07:21,  2.02it/s]
 92%|█████████▏| 9792/10682 [1:23:25<07:20,  2.02it/s]
 92%|█████████▏| 9793/10682 [1:23:26<07:21,  2.02it/s]
 92%|█████████▏| 9794/10682 [1:23:26<07:19,  2.02it/s]
 92%|█████████▏| 9795/10682 [1:23:27<07:19,  2.02it/s]
 92%|█████████▏| 9796/10682 [1:23:27<07:18,  2.02it/s]
 92%|█████████▏| 9797/10682 [1:23:28<07:18,  2.02it/s]
 92%|█████████▏| 9798/10682 [1:23:28<07:17,  2.02it/s]
 92%|█████████▏| 9799/10682 [1:23:29<07:18,  2.02it/s]
 92%|█████████▏| 9800/10682 [1:23:29<07:17,  2.02it/s]{'loss': 3.3804, 'grad_norm': 0.2052086591720581, 'learning_rate': 2.0627663280649135e-05, 'epoch': 0.92}
                                                      
-
 92%|█████████▏| 9800/10682 [1:23:29<07:17,  2.02it/s]
 92%|█████████▏| 9801/10682 [1:23:30<07:17,  2.02it/s]
 92%|█████████▏| 9802/10682 [1:23:30<07:16,  2.02it/s]
 92%|█████████▏| 9803/10682 [1:23:31<07:15,  2.02it/s]
 92%|█████████▏| 9804/10682 [1:23:31<07:15,  2.02it/s]
 92%|█████████▏| 9805/10682 [1:23:32<07:14,  2.02it/s]
 92%|█████████▏| 9806/10682 [1:23:32<07:14,  2.02it/s]
 92%|█████████▏| 9807/10682 [1:23:33<07:14,  2.02it/s]
 92%|█████████▏| 9808/10682 [1:23:33<07:13,  2.02it/s]
 92%|█████████▏| 9809/10682 [1:23:34<07:13,  2.02it/s]
 92%|█████████▏| 9810/10682 [1:23:34<07:12,  2.02it/s]
 92%|█████████▏| 9811/10682 [1:23:35<07:12,  2.01it/s]
 92%|█████████▏| 9812/10682 [1:23:35<07:11,  2.02it/s]
 92%|█████████▏| 9813/10682 [1:23:36<07:10,  2.02it/s]
 92%|█████████▏| 9814/10682 [1:23:36<07:10,  2.02it/s]
 92%|█████████▏| 9815/10682 [1:23:37<07:09,  2.02it/s]
 92%|█████████▏| 9816/10682 [1:23:37<07:09,  2.02it/s]
 92%|█████████▏| 9817/10682 [1:23:38<07:08,  2.02it/s]
 92%|█████████▏| 9818/10682 [1:23:38<07:08,  2.02it/s]
 92%|█████████▏| 9819/10682 [1:23:39<07:08,  2.01it/s]
 92%|█████████▏| 9820/10682 [1:23:39<07:08,  2.01it/s]
 92%|█████████▏| 9821/10682 [1:23:40<07:07,  2.01it/s]
 92%|█████████▏| 9822/10682 [1:23:40<07:06,  2.02it/s]
 92%|█████████▏| 9823/10682 [1:23:41<07:05,  2.02it/s]
 92%|█████████▏| 9824/10682 [1:23:41<07:05,  2.02it/s]
 92%|█████████▏| 9825/10682 [1:23:42<07:05,  2.01it/s]{'loss': 3.3896, 'grad_norm': 0.20912906527519226, 'learning_rate': 1.9482414469655486e-05, 'epoch': 0.92}
                                                      
-
 92%|█████████▏| 9825/10682 [1:23:42<07:05,  2.01it/s]
 92%|█████████▏| 9826/10682 [1:23:42<07:07,  2.00it/s]
 92%|█████████▏| 9827/10682 [1:23:43<07:06,  2.01it/s]
 92%|█████████▏| 9828/10682 [1:23:43<07:04,  2.01it/s]
 92%|█████████▏| 9829/10682 [1:23:44<07:03,  2.01it/s]
 92%|█████████▏| 9830/10682 [1:23:44<07:02,  2.02it/s]
 92%|█████████▏| 9831/10682 [1:23:45<07:01,  2.02it/s]
 92%|█████████▏| 9832/10682 [1:23:45<07:01,  2.02it/s]
 92%|█████████▏| 9833/10682 [1:23:46<07:00,  2.02it/s]
 92%|█████████▏| 9834/10682 [1:23:46<07:00,  2.02it/s]
 92%|█████████▏| 9835/10682 [1:23:47<07:00,  2.01it/s]
 92%|█████████▏| 9836/10682 [1:23:47<07:00,  2.01it/s]
 92%|█████████▏| 9837/10682 [1:23:48<06:59,  2.01it/s]
 92%|█████████▏| 9838/10682 [1:23:48<06:59,  2.01it/s]
 92%|█████████▏| 9839/10682 [1:23:49<06:57,  2.02it/s]
 92%|█████████▏| 9840/10682 [1:23:49<06:57,  2.02it/s]
 92%|█████████▏| 9841/10682 [1:23:50<06:56,  2.02it/s]
 92%|█████████▏| 9842/10682 [1:23:50<06:57,  2.01it/s]
 92%|█████████▏| 9843/10682 [1:23:51<06:56,  2.01it/s]
 92%|█████████▏| 9844/10682 [1:23:51<06:55,  2.02it/s]
 92%|█████████▏| 9845/10682 [1:23:51<06:54,  2.02it/s]
 92%|█████████▏| 9846/10682 [1:23:52<06:54,  2.02it/s]
 92%|█████████▏| 9847/10682 [1:23:52<06:53,  2.02it/s]
 92%|█████████▏| 9848/10682 [1:23:53<06:52,  2.02it/s]
 92%|█████████▏| 9849/10682 [1:23:53<06:52,  2.02it/s]
 92%|█████████▏| 9850/10682 [1:23:54<06:51,  2.02it/s]{'loss': 3.3789, 'grad_norm': 0.20813621580600739, 'learning_rate': 1.8369240812535104e-05, 'epoch': 0.92}
                                                      
-
 92%|█████████▏| 9850/10682 [1:23:54<06:51,  2.02it/s]
 92%|█████████▏| 9851/10682 [1:23:54<06:55,  2.00it/s]
 92%|█████████▏| 9852/10682 [1:23:55<06:53,  2.01it/s]
 92%|█████████▏| 9853/10682 [1:23:55<06:52,  2.01it/s]
 92%|█████████▏| 9854/10682 [1:23:56<06:51,  2.01it/s]
 92%|█████████▏| 9855/10682 [1:23:56<06:50,  2.01it/s]
 92%|█████████▏| 9856/10682 [1:23:57<06:49,  2.02it/s]
 92%|█████████▏| 9857/10682 [1:23:57<06:49,  2.02it/s]
 92%|█████████▏| 9858/10682 [1:23:58<06:49,  2.01it/s]
 92%|█████████▏| 9859/10682 [1:23:58<06:48,  2.01it/s]
 92%|█████████▏| 9860/10682 [1:23:59<06:47,  2.02it/s]
 92%|█████████▏| 9861/10682 [1:23:59<06:47,  2.02it/s]
 92%|█████████▏| 9862/10682 [1:24:00<06:45,  2.02it/s]
 92%|█████████▏| 9863/10682 [1:24:00<06:46,  2.02it/s]
 92%|█████████▏| 9864/10682 [1:24:01<06:44,  2.02it/s]
 92%|█████████▏| 9865/10682 [1:24:01<06:45,  2.02it/s]
 92%|█████████▏| 9866/10682 [1:24:02<06:44,  2.02it/s]
 92%|█████████▏| 9867/10682 [1:24:02<06:44,  2.02it/s]
 92%|█████████▏| 9868/10682 [1:24:03<06:43,  2.02it/s]
 92%|█████████▏| 9869/10682 [1:24:03<06:43,  2.01it/s]
 92%|█████████▏| 9870/10682 [1:24:04<06:42,  2.02it/s]
 92%|█████████▏| 9871/10682 [1:24:04<06:42,  2.01it/s]
 92%|█████████▏| 9872/10682 [1:24:05<06:42,  2.01it/s]
 92%|█████████▏| 9873/10682 [1:24:05<06:41,  2.01it/s]
 92%|███���█████▏| 9874/10682 [1:24:06<06:40,  2.02it/s]
 92%|█████████▏| 9875/10682 [1:24:06<06:40,  2.02it/s]{'loss': 3.39, 'grad_norm': 0.21077197790145874, 'learning_rate': 1.7288216615031272e-05, 'epoch': 0.92}
-
                                                      

 92%|█████████▏| 9875/10682 [1:24:06<06:40,  2.02it/s]
 92%|█████████▏| 9876/10682 [1:24:07<06:41,  2.01it/s]
 92%|█████████▏| 9877/10682 [1:24:07<06:39,  2.01it/s]
 92%|█████████▏| 9878/10682 [1:24:08<06:39,  2.01it/s]
 92%|█████████▏| 9879/10682 [1:24:08<06:38,  2.01it/s]
 92%|█████████▏| 9880/10682 [1:24:09<06:37,  2.02it/s]
 93%|█████████▎| 9881/10682 [1:24:09<06:36,  2.02it/s]
 93%|█████████▎| 9882/10682 [1:24:10<06:36,  2.02it/s]
 93%|█████████▎| 9883/10682 [1:24:10<06:36,  2.02it/s]
 93%|█████████▎| 9884/10682 [1:24:11<06:35,  2.02it/s]
 93%|█████████▎| 9885/10682 [1:24:11<06:35,  2.02it/s]
 93%|█████████▎| 9886/10682 [1:24:12<06:34,  2.02it/s]
 93%|█████████▎| 9887/10682 [1:24:12<06:34,  2.02it/s]
 93%|█████████▎| 9888/10682 [1:24:13<06:33,  2.02it/s]
 93%|█████████▎| 9889/10682 [1:24:13<06:33,  2.01it/s]
 93%|█████████▎| 9890/10682 [1:24:14<06:32,  2.02it/s]
 93%|█████████▎| 9891/10682 [1:24:14<06:32,  2.01it/s]
 93%|█████████▎| 9892/10682 [1:24:15<06:32,  2.01it/s]
 93%|█████████▎| 9893/10682 [1:24:15<06:31,  2.02it/s]
 93%|█████████▎| 9894/10682 [1:24:16<06:31,  2.01it/s]
 93%|█████████▎| 9895/10682 [1:24:16<06:31,  2.01it/s]
 93%|█████████▎| 9896/10682 [1:24:17<06:29,  2.02it/s]
 93%|█████████▎| 9897/10682 [1:24:17<06:29,  2.02it/s]
 93%|█████████▎| 9898/10682 [1:24:18<06:28,  2.02it/s]
 93%|█████████▎| 9899/10682 [1:24:18<06:28,  2.02it/s]
 93%|█████████▎| 9900/10682 [1:24:19<06:27,  2.02it/s]{'loss': 3.3713, 'grad_norm': 0.2102077156305313, 'learning_rate': 1.6239414036870183e-05, 'epoch': 0.93}
                                                      
-
 93%|█████████▎| 9900/10682 [1:24:19<06:27,  2.02it/s]
 93%|█████████▎| 9901/10682 [1:24:19<06:27,  2.01it/s]
 93%|█████████▎| 9902/10682 [1:24:20<06:26,  2.02it/s]
 93%|█████████▎| 9903/10682 [1:24:20<06:26,  2.01it/s]
 93%|█████████▎| 9904/10682 [1:24:21<06:25,  2.02it/s]
 93%|█████████▎| 9905/10682 [1:24:21<06:25,  2.02it/s]
 93%|█████████▎| 9906/10682 [1:24:22<06:24,  2.02it/s]
 93%|█████████▎| 9907/10682 [1:24:22<06:24,  2.01it/s]
 93%|█████████▎| 9908/10682 [1:24:23<06:24,  2.01it/s]
 93%|█████████▎| 9909/10682 [1:24:23<06:24,  2.01it/s]
 93%|█████████▎| 9910/10682 [1:24:24<06:23,  2.01it/s]
 93%|█████████▎| 9911/10682 [1:24:24<06:22,  2.01it/s]
 93%|█████████▎| 9912/10682 [1:24:25<06:22,  2.01it/s]
 93%|█████████▎| 9913/10682 [1:24:25<06:21,  2.01it/s]
 93%|█████████▎| 9914/10682 [1:24:26<06:21,  2.01it/s]
 93%|█████████▎| 9915/10682 [1:24:26<06:21,  2.01it/s]
 93%|█████████▎| 9916/10682 [1:24:27<06:20,  2.01it/s]
 93%|█████████▎| 9917/10682 [1:24:27<06:20,  2.01it/s]
 93%|█████████▎| 9918/10682 [1:24:28<06:19,  2.01it/s]
 93%|█████████▎| 9919/10682 [1:24:28<06:18,  2.01it/s]
 93%|█████████▎| 9920/10682 [1:24:29<06:18,  2.01it/s]
 93%|█████████▎| 9921/10682 [1:24:29<06:17,  2.02it/s]
 93%|█████████▎| 9922/10682 [1:24:30<06:16,  2.02it/s]
 93%|█████████▎| 9923/10682 [1:24:30<06:16,  2.02it/s]
 93%|█████████▎| 9924/10682 [1:24:31<06:15,  2.02it/s]
 93%|█████████▎| 9925/10682 [1:24:31<06:15,  2.02it/s]
                                                      
{'loss': 3.3715, 'grad_norm': 0.21153439581394196, 'learning_rate': 1.5222903086944684e-05, 'epoch': 0.93}
-
 93%|█████████▎| 9925/10682 [1:24:31<06:15,  2.02it/s]
 93%|█████████▎| 9926/10682 [1:24:32<06:15,  2.01it/s]
 93%|█████████▎| 9927/10682 [1:24:32<06:14,  2.02it/s]
 93%|█████████▎| 9928/10682 [1:24:33<06:14,  2.02it/s]
 93%|█████████▎| 9929/10682 [1:24:33<06:13,  2.02it/s]
 93%|█████████▎| 9930/10682 [1:24:34<06:13,  2.02it/s]
 93%|█████████▎| 9931/10682 [1:24:34<06:12,  2.01it/s]
 93%|█████████▎| 9932/10682 [1:24:35<06:12,  2.02it/s]
 93%|█████████▎| 9933/10682 [1:24:35<06:11,  2.01it/s]
 93%|█████████▎| 9934/10682 [1:24:36<06:11,  2.01it/s]
 93%|█████████▎| 9935/10682 [1:24:36<06:10,  2.01it/s]
 93%|█████████▎| 9936/10682 [1:24:37<06:10,  2.01it/s]
 93%|█████████▎| 9937/10682 [1:24:37<06:09,  2.01it/s]
 93%|█████████▎| 9938/10682 [1:24:38<06:09,  2.02it/s]
 93%|█████████▎| 9939/10682 [1:24:38<06:08,  2.01it/s]
 93%|█████████▎| 9940/10682 [1:24:39<06:08,  2.01it/s]
 93%|█████████▎| 9941/10682 [1:24:39<06:07,  2.01it/s]
 93%|█████████▎| 9942/10682 [1:24:40<06:07,  2.01it/s]
 93%|█████████▎| 9943/10682 [1:24:40<06:06,  2.02it/s]
 93%|█████████▎| 9944/10682 [1:24:41<06:06,  2.02it/s]
 93%|█████████▎| 9945/10682 [1:24:41<06:05,  2.02it/s]
 93%|█████████▎| 9946/10682 [1:24:42<06:05,  2.02it/s]
 93%|█████████▎| 9947/10682 [1:24:42<06:05,  2.01it/s]
 93%|█████████▎| 9948/10682 [1:24:43<06:04,  2.01it/s]
 93%|█████████▎| 9949/10682 [1:24:43<06:03,  2.01it/s]
 93%|█████████▎| 9950/10682 [1:24:44<06:02,  2.02it/s]
                                                      
{'loss': 3.3723, 'grad_norm': 0.20923252403736115, 'learning_rate': 1.4238751618640577e-05, 'epoch': 0.93}
-
 93%|█████████▎| 9950/10682 [1:24:44<06:02,  2.02it/s]
 93%|█████████▎| 9951/10682 [1:24:44<06:02,  2.02it/s]
 93%|█████████▎| 9952/10682 [1:24:45<06:02,  2.02it/s]
 93%|█████████▎| 9953/10682 [1:24:45<06:02,  2.01it/s]
 93%|█████████▎| 9954/10682 [1:24:46<06:01,  2.01it/s]
 93%|█████████▎| 9955/10682 [1:24:46<06:00,  2.01it/s]
 93%|█████████▎| 9956/10682 [1:24:47<06:00,  2.02it/s]
 93%|█████████▎| 9957/10682 [1:24:47<05:59,  2.01it/s]
 93%|█████████▎| 9958/10682 [1:24:48<05:59,  2.01it/s]
 93%|█████████▎| 9959/10682 [1:24:48<05:58,  2.01it/s]
 93%|█████████▎| 9960/10682 [1:24:49<05:58,  2.01it/s]
 93%|█████████▎| 9961/10682 [1:24:49<05:58,  2.01it/s]
 93%|█████████▎| 9962/10682 [1:24:50<05:57,  2.02it/s]
 93%|█████████▎| 9963/10682 [1:24:50<05:56,  2.02it/s]
 93%|█████████▎| 9964/10682 [1:24:51<05:56,  2.02it/s]
 93%|█████████▎| 9965/10682 [1:24:51<05:55,  2.02it/s]
 93%|█████████▎| 9966/10682 [1:24:52<05:55,  2.01it/s]
 93%|█████████▎| 9967/10682 [1:24:52<05:55,  2.01it/s]
 93%|█████████▎| 9968/10682 [1:24:53<05:54,  2.01it/s]
 93%|█████████▎| 9969/10682 [1:24:53<05:53,  2.02it/s]
 93%|█████████▎| 9970/10682 [1:24:54<05:52,  2.02it/s]
 93%|█████████▎| 9971/10682 [1:24:54<05:52,  2.02it/s]
 93%|█████████▎| 9972/10682 [1:24:55<05:52,  2.02it/s]
 93%|█████████▎| 9973/10682 [1:24:55<05:52,  2.01it/s]
 93%|█████████▎| 9974/10682 [1:24:56<05:52,  2.01it/s]
 93%|█████████▎| 9975/10682 [1:24:56<05:51,  2.01it/s]{'loss': 3.3776, 'grad_norm': 0.21355055272579193, 'learning_rate': 1.3287025325307511e-05, 'epoch': 0.93}
                                                      
-
 93%|█████████▎| 9975/10682 [1:24:56<05:51,  2.01it/s]
 93%|█████████▎| 9976/10682 [1:24:57<05:51,  2.01it/s]
 93%|█████████▎| 9977/10682 [1:24:57<05:50,  2.01it/s]
 93%|█████████▎| 9978/10682 [1:24:58<05:49,  2.01it/s]
 93%|█████████▎| 9979/10682 [1:24:58<05:48,  2.01it/s]
 93%|█████████▎| 9980/10682 [1:24:59<05:48,  2.01it/s]
 93%|█████████▎| 9981/10682 [1:24:59<05:48,  2.01it/s]
 93%|█████████▎| 9982/10682 [1:24:59<05:47,  2.01it/s]
 93%|█████████▎| 9983/10682 [1:25:00<05:46,  2.02it/s]
 93%|█████████▎| 9984/10682 [1:25:00<05:46,  2.01it/s]
 93%|█████████▎| 9985/10682 [1:25:01<05:45,  2.01it/s]
 93%|█████████▎| 9986/10682 [1:25:01<05:46,  2.01it/s]
 93%|█████████▎| 9987/10682 [1:25:02<05:45,  2.01it/s]
 94%|█████████▎| 9988/10682 [1:25:02<05:44,  2.01it/s]
 94%|█████████▎| 9989/10682 [1:25:03<05:44,  2.01it/s]
 94%|█████████▎| 9990/10682 [1:25:03<05:43,  2.01it/s]
 94%|█████████▎| 9991/10682 [1:25:04<05:43,  2.01it/s]
 94%|█████████▎| 9992/10682 [1:25:04<05:43,  2.01it/s]
 94%|█████████▎| 9993/10682 [1:25:05<05:42,  2.01it/s]
 94%|█████████▎| 9994/10682 [1:25:05<05:41,  2.01it/s]
 94%|█████████▎| 9995/10682 [1:25:06<05:40,  2.02it/s]
 94%|█████████▎| 9996/10682 [1:25:06<05:40,  2.02it/s]
 94%|█████████▎| 9997/10682 [1:25:07<05:40,  2.01it/s]
 94%|█████████▎| 9998/10682 [1:25:07<05:40,  2.01it/s]
 94%|█████████▎| 9999/10682 [1:25:08<05:39,  2.01it/s]
 94%|█████████▎| 10000/10682 [1:25:08<05:38,  2.01it/s]{'loss': 3.3751, 'grad_norm': 0.2126685082912445, 'learning_rate': 1.2367787735873993e-05, 'epoch': 0.94}
                                                       
-
 94%|█████████▎| 10000/10682 [1:25:08<05:38,  2.01it/s]
 94%|█████████▎| 10001/10682 [1:25:09<05:38,  2.01it/s]
 94%|█████████▎| 10002/10682 [1:25:09<05:38,  2.01it/s]
 94%|█████████▎| 10003/10682 [1:25:10<05:36,  2.02it/s]
 94%|█████████▎| 10004/10682 [1:25:10<05:36,  2.01it/s]
 94%|█████████▎| 10005/10682 [1:25:11<05:36,  2.01it/s]
 94%|█████████▎| 10006/10682 [1:25:11<05:35,  2.01it/s]
 94%|█████████▎| 10007/10682 [1:25:12<05:35,  2.01it/s]
 94%|█████████▎| 10008/10682 [1:25:12<05:34,  2.01it/s]
 94%|█████████▎| 10009/10682 [1:25:13<05:34,  2.01it/s]
 94%|█████████▎| 10010/10682 [1:25:13<05:33,  2.01it/s]
 94%|█████████▎| 10011/10682 [1:25:14<05:33,  2.01it/s]
 94%|█████████▎| 10012/10682 [1:25:14<05:32,  2.02it/s]
 94%|█████████▎| 10013/10682 [1:25:15<05:32,  2.01it/s]
 94%|█████████▎| 10014/10682 [1:25:15<05:32,  2.01it/s]
 94%|█████████▍| 10015/10682 [1:25:16<05:31,  2.01it/s]
 94%|█████████▍| 10016/10682 [1:25:16<05:30,  2.01it/s]
 94%|█████████▍| 10017/10682 [1:25:17<05:30,  2.01it/s]
 94%|█████████▍| 10018/10682 [1:25:17<05:29,  2.02it/s]
 94%|█████████▍| 10019/10682 [1:25:18<05:28,  2.02it/s]
 94%|█████████▍| 10020/10682 [1:25:18<05:28,  2.02it/s]
 94%|█████████▍| 10021/10682 [1:25:19<05:28,  2.01it/s]
 94%|█████████▍| 10022/10682 [1:25:19<05:27,  2.02it/s]
 94%|█████████▍| 10023/10682 [1:25:20<05:27,  2.01it/s]
 94%|█████████▍| 10024/10682 [1:25:20<05:27,  2.01it/s]
 94%|█████████▍| 10025/10682 [1:25:21<05:27,  2.00it/s]{'loss': 3.3813, 'grad_norm': 0.21192748844623566, 'learning_rate': 1.1481100210606388e-05, 'epoch': 0.94}
-
                                                       

 94%|█████████▍| 10025/10682 [1:25:21<05:27,  2.00it/s]
 94%|█████████▍| 10026/10682 [1:25:21<05:26,  2.01it/s]
 94%|█████████▍| 10027/10682 [1:25:22<05:26,  2.01it/s]
 94%|█████████▍| 10028/10682 [1:25:22<05:25,  2.01it/s]
 94%|█████████▍| 10029/10682 [1:25:23<05:24,  2.01it/s]
 94%|█████████▍| 10030/10682 [1:25:23<05:23,  2.01it/s]
 94%|█████████▍| 10031/10682 [1:25:24<05:23,  2.01it/s]
 94%|█████████▍| 10032/10682 [1:25:24<05:22,  2.02it/s]
 94%|█████████▍| 10033/10682 [1:25:25<05:21,  2.02it/s]
 94%|█████████▍| 10034/10682 [1:25:25<05:21,  2.02it/s]
 94%|█████████▍| 10035/10682 [1:25:26<05:21,  2.01it/s]
 94%|█████████▍| 10036/10682 [1:25:26<05:20,  2.01it/s]
 94%|█████████▍| 10037/10682 [1:25:27<05:20,  2.01it/s]
 94%|█████████▍| 10038/10682 [1:25:27<05:20,  2.01it/s]
 94%|█████████▍| 10039/10682 [1:25:28<05:20,  2.01it/s]
 94%|█████████▍| 10040/10682 [1:25:28<05:19,  2.01it/s]
 94%|█████████▍| 10041/10682 [1:25:29<05:18,  2.01it/s]
 94%|█████████▍| 10042/10682 [1:25:29<05:18,  2.01it/s]
 94%|█████████▍| 10043/10682 [1:25:30<05:17,  2.01it/s]
 94%|█████████▍| 10044/10682 [1:25:30<05:17,  2.01it/s]
 94%|█████████▍| 10045/10682 [1:25:31<05:17,  2.00it/s]
 94%|█████████▍| 10046/10682 [1:25:31<05:17,  2.01it/s]
 94%|█████████▍| 10047/10682 [1:25:32<05:16,  2.01it/s]
 94%|█████████▍| 10048/10682 [1:25:32<05:15,  2.01it/s]
 94%|█████████▍| 10049/10682 [1:25:33<05:15,  2.01it/s]
 94%|█████████▍| 10050/10682 [1:25:33<05:15,  2.01it/s]{'loss': 3.3766, 'grad_norm': 0.2114092856645584, 'learning_rate': 1.0627021937013704e-05, 'epoch': 0.94}
                                                       
-
 94%|█████████▍| 10050/10682 [1:25:33<05:15,  2.01it/s]
 94%|█████████▍| 10051/10682 [1:25:34<05:14,  2.01it/s]
 94%|█████████▍| 10052/10682 [1:25:34<05:13,  2.01it/s]
 94%|█████████▍| 10053/10682 [1:25:35<05:12,  2.01it/s]
 94%|█████████▍| 10054/10682 [1:25:35<05:11,  2.01it/s]
 94%|█████████▍| 10055/10682 [1:25:36<05:11,  2.01it/s]
 94%|█████████▍| 10056/10682 [1:25:36<05:10,  2.01it/s]
 94%|█████████▍| 10057/10682 [1:25:37<05:10,  2.01it/s]
 94%|█████████▍| 10058/10682 [1:25:37<05:09,  2.01it/s]
 94%|█████████▍| 10059/10682 [1:25:38<05:09,  2.01it/s]
 94%|█████████▍| 10060/10682 [1:25:38<05:09,  2.01it/s]
 94%|█████████▍| 10061/10682 [1:25:39<05:08,  2.01it/s]
 94%|█████████▍| 10062/10682 [1:25:39<05:08,  2.01it/s]
 94%|█████████▍| 10063/10682 [1:25:40<05:07,  2.01it/s]
 94%|█████████▍| 10064/10682 [1:25:40<05:07,  2.01it/s]
 94%|█████████▍| 10065/10682 [1:25:41<05:06,  2.01it/s]
 94%|█████████▍| 10066/10682 [1:25:41<05:05,  2.01it/s]
 94%|█████████▍| 10067/10682 [1:25:42<05:05,  2.01it/s]
 94%|█████████▍| 10068/10682 [1:25:42<05:05,  2.01it/s]
 94%|█████████▍| 10069/10682 [1:25:43<05:04,  2.01it/s]
 94%|█████████▍| 10070/10682 [1:25:43<05:04,  2.01it/s]
 94%|█████████▍| 10071/10682 [1:25:44<05:03,  2.01it/s]
 94%|█████████▍| 10072/10682 [1:25:44<05:02,  2.01it/s]
 94%|█████████▍| 10073/10682 [1:25:45<05:01,  2.02it/s]
 94%|█████████▍| 10074/10682 [1:25:45<05:01,  2.02it/s]
 94%|█████████▍| 10075/10682 [1:25:46<05:01,  2.01it/s]{'loss': 3.372, 'grad_norm': 0.21050767600536346, 'learning_rate': 9.805609925895964e-06, 'epoch': 0.94}
-
                                                       

 94%|█████████▍| 10075/10682 [1:25:46<05:01,  2.01it/s]
 94%|█████████▍| 10076/10682 [1:25:46<05:02,  2.01it/s]
 94%|█████████▍| 10077/10682 [1:25:47<05:01,  2.01it/s]
 94%|█████████▍| 10078/10682 [1:25:47<05:00,  2.01it/s]
 94%|█████████▍| 10079/10682 [1:25:48<04:59,  2.01it/s]
 94%|█████████▍| 10080/10682 [1:25:48<04:59,  2.01it/s]
 94%|█████████▍| 10081/10682 [1:25:49<04:58,  2.01it/s]
 94%|█████████▍| 10082/10682 [1:25:49<04:57,  2.01it/s]
 94%|█████████▍| 10083/10682 [1:25:50<04:57,  2.02it/s]
 94%|█████████▍| 10084/10682 [1:25:50<04:56,  2.02it/s]
 94%|█████████▍| 10085/10682 [1:25:51<04:56,  2.02it/s]
 94%|█████████▍| 10086/10682 [1:25:51<04:56,  2.01it/s]
 94%|█████████▍| 10087/10682 [1:25:52<04:55,  2.01it/s]
 94%|█████████▍| 10088/10682 [1:25:52<04:54,  2.01it/s]
 94%|█████████▍| 10089/10682 [1:25:53<04:54,  2.02it/s]
 94%|█████████▍| 10090/10682 [1:25:53<04:53,  2.02it/s]
 94%|█████████▍| 10091/10682 [1:25:54<04:52,  2.02it/s]
 94%|█████████▍| 10092/10682 [1:25:54<04:52,  2.02it/s]
 94%|█████████▍| 10093/10682 [1:25:55<04:51,  2.02it/s]
 94%|█████████▍| 10094/10682 [1:25:55<04:51,  2.01it/s]
 95%|█████████▍| 10095/10682 [1:25:56<04:51,  2.01it/s]
 95%|█████████▍| 10096/10682 [1:25:56<04:50,  2.02it/s]
 95%|█████████▍| 10097/10682 [1:25:57<04:50,  2.01it/s]
 95%|█████████▍| 10098/10682 [1:25:57<04:49,  2.02it/s]
 95%|█████████▍| 10099/10682 [1:25:58<04:48,  2.02it/s]
 95%|█████████▍| 10100/10682 [1:25:58<04:48,  2.02it/s]{'loss': 3.3698, 'grad_norm': 0.20959755778312683, 'learning_rate': 9.01691900753926e-06, 'epoch': 0.95}
                                                       
-
 95%|█████████▍| 10100/10682 [1:25:58<04:48,  2.02it/s]
 95%|█████████▍| 10101/10682 [1:25:59<04:48,  2.02it/s]
 95%|█████████▍| 10102/10682 [1:25:59<04:48,  2.01it/s]
 95%|█████████▍| 10103/10682 [1:26:00<04:47,  2.01it/s]
 95%|█████████▍| 10104/10682 [1:26:00<04:46,  2.01it/s]
 95%|█████████▍| 10105/10682 [1:26:01<04:46,  2.02it/s]
 95%|█████████▍| 10106/10682 [1:26:01<04:45,  2.02it/s]
 95%|█████████▍| 10107/10682 [1:26:02<04:45,  2.02it/s]
 95%|█████████▍| 10108/10682 [1:26:02<04:45,  2.01it/s]
 95%|█████████▍| 10109/10682 [1:26:03<04:44,  2.01it/s]
 95%|█████████▍| 10110/10682 [1:26:03<04:44,  2.01it/s]
 95%|█████████▍| 10111/10682 [1:26:04<04:43,  2.01it/s]
 95%|█████████▍| 10112/10682 [1:26:04<04:43,  2.01it/s]
 95%|█████████▍| 10113/10682 [1:26:05<04:42,  2.01it/s]
 95%|█████████▍| 10114/10682 [1:26:05<04:42,  2.01it/s]
 95%|█████████▍| 10115/10682 [1:26:06<04:41,  2.02it/s]
 95%|█████████▍| 10116/10682 [1:26:06<04:40,  2.02it/s]
 95%|█████████▍| 10117/10682 [1:26:07<04:40,  2.02it/s]
 95%|█████████▍| 10118/10682 [1:26:07<04:39,  2.02it/s]
 95%|█████████▍| 10119/10682 [1:26:08<04:38,  2.02it/s]
 95%|█████████▍| 10120/10682 [1:26:08<04:39,  2.01it/s]
 95%|█████████▍| 10121/10682 [1:26:09<04:38,  2.01it/s]
 95%|█████████▍| 10122/10682 [1:26:09<04:37,  2.02it/s]
 95%|█████████▍| 10123/10682 [1:26:10<04:37,  2.02it/s]
 95%|█████████▍| 10124/10682 [1:26:10<04:36,  2.02it/s]
 95%|█████████▍| 10125/10682 [1:26:11<04:36,  2.02it/s]{'loss': 3.3676, 'grad_norm': 0.2096906453371048, 'learning_rate': 8.261001828055447e-06, 'epoch': 0.95}
                                                       
-
 95%|█████████▍| 10125/10682 [1:26:11<04:36,  2.02it/s]
 95%|█████████▍| 10126/10682 [1:26:11<04:35,  2.01it/s]
 95%|█████████▍| 10127/10682 [1:26:12<04:35,  2.01it/s]
 95%|█████████▍| 10128/10682 [1:26:12<04:34,  2.02it/s]
 95%|█████████▍| 10129/10682 [1:26:13<04:34,  2.02it/s]
 95%|█████████▍| 10130/10682 [1:26:13<04:33,  2.02it/s]
 95%|█████████▍| 10131/10682 [1:26:14<04:33,  2.02it/s]
 95%|█████████▍| 10132/10682 [1:26:14<04:32,  2.02it/s]
 95%|█████████▍| 10133/10682 [1:26:14<04:32,  2.02it/s]
 95%|█████████▍| 10134/10682 [1:26:15<04:31,  2.02it/s]
 95%|█████████▍| 10135/10682 [1:26:15<04:31,  2.02it/s]
 95%|█████████▍| 10136/10682 [1:26:16<04:31,  2.01it/s]
 95%|█████████▍| 10137/10682 [1:26:16<04:30,  2.01it/s]
 95%|█████████▍| 10138/10682 [1:26:17<04:29,  2.01it/s]
 95%|█████████▍| 10139/10682 [1:26:17<04:29,  2.01it/s]
 95%|█████████▍| 10140/10682 [1:26:18<04:28,  2.02it/s]
 95%|█████████▍| 10141/10682 [1:26:18<04:28,  2.02it/s]
 95%|█████████▍| 10142/10682 [1:26:19<04:27,  2.02it/s]
 95%|█████████▍| 10143/10682 [1:26:19<04:27,  2.02it/s]
 95%|█████████▍| 10144/10682 [1:26:20<04:26,  2.02it/s]
 95%|█████████▍| 10145/10682 [1:26:20<04:26,  2.02it/s]
 95%|█████████▍| 10146/10682 [1:26:21<04:25,  2.02it/s]
 95%|█████████▍| 10147/10682 [1:26:21<04:25,  2.02it/s]
 95%|█████████▌| 10148/10682 [1:26:22<04:24,  2.02it/s]
 95%|█████████▌| 10149/10682 [1:26:22<04:24,  2.02it/s]
 95%|█████████▌| 10150/10682 [1:26:23<04:23,  2.02it/s]{'loss': 3.382, 'grad_norm': 0.20925013720989227, 'learning_rate': 7.537908845868024e-06, 'epoch': 0.95}
                                                       
-
 95%|█████████▌| 10150/10682 [1:26:23<04:23,  2.02it/s]
 95%|█████████▌| 10151/10682 [1:26:23<04:23,  2.01it/s]
 95%|█████████▌| 10152/10682 [1:26:24<04:23,  2.01it/s]
 95%|█████████▌| 10153/10682 [1:26:24<04:22,  2.01it/s]
 95%|█████████▌| 10154/10682 [1:26:25<04:21,  2.02it/s]
 95%|█████████▌| 10155/10682 [1:26:25<04:21,  2.02it/s]
 95%|█████████▌| 10156/10682 [1:26:26<04:20,  2.02it/s]
 95%|█████████▌| 10157/10682 [1:26:26<04:19,  2.02it/s]
 95%|█████████▌| 10158/10682 [1:26:27<04:19,  2.02it/s]
 95%|█████████▌| 10159/10682 [1:26:27<04:19,  2.02it/s]
 95%|█████████▌| 10160/10682 [1:26:28<04:18,  2.02it/s]
 95%|█████████▌| 10161/10682 [1:26:28<04:19,  2.01it/s]
 95%|█████████▌| 10162/10682 [1:26:29<04:18,  2.01it/s]
 95%|█████████▌| 10163/10682 [1:26:29<04:17,  2.01it/s]
 95%|█████████▌| 10164/10682 [1:26:30<04:16,  2.02it/s]
 95%|█████████▌| 10165/10682 [1:26:30<04:16,  2.02it/s]
 95%|█████████▌| 10166/10682 [1:26:31<04:15,  2.02it/s]
 95%|█████████▌| 10167/10682 [1:26:31<04:15,  2.02it/s]
 95%|█████████▌| 10168/10682 [1:26:32<04:14,  2.02it/s]
 95%|█████████▌| 10169/10682 [1:26:32<04:14,  2.02it/s]
 95%|█████████▌| 10170/10682 [1:26:33<04:14,  2.02it/s]
 95%|█████████▌| 10171/10682 [1:26:33<04:13,  2.02it/s]
 95%|█████████▌| 10172/10682 [1:26:34<04:12,  2.02it/s]
 95%|█████████▌| 10173/10682 [1:26:34<04:12,  2.01it/s]
 95%|█████████▌| 10174/10682 [1:26:35<04:11,  2.02it/s]
 95%|█████████▌| 10175/10682 [1:26:35<04:11,  2.02it/s]
                                                       
{'loss': 3.3708, 'grad_norm': 0.2130783051252365, 'learning_rate': 6.847688328344037e-06, 'epoch': 0.95}
-
 95%|█████████▌| 10175/10682 [1:26:35<04:11,  2.02it/s]
 95%|█████████▌| 10176/10682 [1:26:36<04:11,  2.02it/s]
 95%|█████████▌| 10177/10682 [1:26:36<04:10,  2.01it/s]
 95%|█████████▌| 10178/10682 [1:26:37<04:09,  2.02it/s]
 95%|█████████▌| 10179/10682 [1:26:37<04:09,  2.02it/s]
 95%|█████████▌| 10180/10682 [1:26:38<04:08,  2.02it/s]
 95%|█████████▌| 10181/10682 [1:26:38<04:08,  2.02it/s]
 95%|█████████▌| 10182/10682 [1:26:39<04:07,  2.02it/s]
 95%|█████████▌| 10183/10682 [1:26:39<04:07,  2.02it/s]
 95%|█████████▌| 10184/10682 [1:26:40<04:07,  2.01it/s]
 95%|█████████▌| 10185/10682 [1:26:40<04:06,  2.02it/s]
 95%|█████████▌| 10186/10682 [1:26:41<04:05,  2.02it/s]
 95%|█████████▌| 10187/10682 [1:26:41<04:05,  2.02it/s]
 95%|█████████▌| 10188/10682 [1:26:42<04:04,  2.02it/s]
 95%|█████████▌| 10189/10682 [1:26:42<04:04,  2.02it/s]
 95%|█████████▌| 10190/10682 [1:26:43<04:03,  2.02it/s]
 95%|█████████▌| 10191/10682 [1:26:43<04:03,  2.02it/s]
 95%|█████████▌| 10192/10682 [1:26:44<04:02,  2.02it/s]
 95%|█████████▌| 10193/10682 [1:26:44<04:02,  2.01it/s]
 95%|█████████▌| 10194/10682 [1:26:45<04:02,  2.02it/s]
 95%|█████████▌| 10195/10682 [1:26:45<04:01,  2.01it/s]
 95%|█████████▌| 10196/10682 [1:26:46<04:01,  2.02it/s]
 95%|█████████▌| 10197/10682 [1:26:46<04:00,  2.02it/s]
 95%|█████████▌| 10198/10682 [1:26:47<03:59,  2.02it/s]
 95%|█████████▌| 10199/10682 [1:26:47<03:59,  2.02it/s]
 95%|█████████▌| 10200/10682 [1:26:48<03:59,  2.01it/s]{'loss': 3.3744, 'grad_norm': 0.2096262127161026, 'learning_rate': 6.190386348572108e-06, 'epoch': 0.95}
                                                       
-
 95%|█████████▌| 10200/10682 [1:26:48<03:59,  2.01it/s]
 95%|█████████▌| 10201/10682 [1:26:48<03:59,  2.01it/s]
 96%|█████████▌| 10202/10682 [1:26:49<03:59,  2.01it/s]
 96%|█████████▌| 10203/10682 [1:26:49<03:58,  2.01it/s]
 96%|█████████▌| 10204/10682 [1:26:50<03:57,  2.01it/s]
 96%|█████████▌| 10205/10682 [1:26:50<03:56,  2.01it/s]
 96%|█████████▌| 10206/10682 [1:26:51<03:55,  2.02it/s]
 96%|█████████▌| 10207/10682 [1:26:51<03:55,  2.01it/s]
 96%|█████████▌| 10208/10682 [1:26:52<03:54,  2.02it/s]
 96%|█████████▌| 10209/10682 [1:26:52<03:54,  2.01it/s]
 96%|█████████▌| 10210/10682 [1:26:53<03:54,  2.01it/s]
 96%|█████████▌| 10211/10682 [1:26:53<03:53,  2.02it/s]
 96%|█████████▌| 10212/10682 [1:26:54<03:52,  2.02it/s]
 96%|█████████▌| 10213/10682 [1:26:54<03:52,  2.02it/s]
 96%|█████████▌| 10214/10682 [1:26:55<03:52,  2.02it/s]
 96%|█████████▌| 10215/10682 [1:26:55<03:51,  2.02it/s]
 96%|█████████▌| 10216/10682 [1:26:56<03:50,  2.02it/s]
 96%|█████████▌| 10217/10682 [1:26:56<03:50,  2.02it/s]
 96%|█████████▌| 10218/10682 [1:26:57<03:49,  2.02it/s]
 96%|█████████▌| 10219/10682 [1:26:57<03:49,  2.02it/s]
 96%|█████████▌| 10220/10682 [1:26:58<03:49,  2.02it/s]
 96%|█████████▌| 10221/10682 [1:26:58<03:48,  2.02it/s]
 96%|█████████▌| 10222/10682 [1:26:59<03:48,  2.02it/s]
 96%|█████████▌| 10223/10682 [1:26:59<03:47,  2.02it/s]
 96%|█████████▌| 10224/10682 [1:27:00<03:47,  2.01it/s]
 96%|█████████▌| 10225/10682 [1:27:00<03:47,  2.01it/s]
                                                       
{'loss': 3.3761, 'grad_norm': 0.21162453293800354, 'learning_rate': 5.56604678228706e-06, 'epoch': 0.96}
-
 96%|█████████▌| 10225/10682 [1:27:00<03:47,  2.01it/s]
 96%|█████████▌| 10226/10682 [1:27:01<03:47,  2.00it/s]
 96%|█████████▌| 10227/10682 [1:27:01<03:46,  2.01it/s]
 96%|█████████▌| 10228/10682 [1:27:02<03:45,  2.01it/s]
 96%|█████████▌| 10229/10682 [1:27:02<03:45,  2.01it/s]
 96%|█████████▌| 10230/10682 [1:27:03<03:44,  2.01it/s]
 96%|█████████▌| 10231/10682 [1:27:03<03:44,  2.01it/s]
 96%|█████████▌| 10232/10682 [1:27:04<03:43,  2.01it/s]
 96%|█████████▌| 10233/10682 [1:27:04<03:43,  2.01it/s]
 96%|█████████▌| 10234/10682 [1:27:05<03:42,  2.01it/s]
 96%|█████████▌| 10235/10682 [1:27:05<03:41,  2.01it/s]
 96%|█████████▌| 10236/10682 [1:27:06<03:41,  2.02it/s]
 96%|█████████▌| 10237/10682 [1:27:06<03:40,  2.02it/s]
 96%|█████████▌| 10238/10682 [1:27:07<03:40,  2.02it/s]
 96%|█████████▌| 10239/10682 [1:27:07<03:39,  2.02it/s]
 96%|█████████▌| 10240/10682 [1:27:08<03:39,  2.02it/s]
 96%|█████████▌| 10241/10682 [1:27:08<03:38,  2.02it/s]
 96%|█████████▌| 10242/10682 [1:27:09<03:38,  2.02it/s]
 96%|█████████▌| 10243/10682 [1:27:09<03:37,  2.01it/s]
 96%|█████████▌| 10244/10682 [1:27:10<03:37,  2.02it/s]
 96%|█████████▌| 10245/10682 [1:27:10<03:36,  2.01it/s]
 96%|█████████▌| 10246/10682 [1:27:11<03:36,  2.02it/s]
 96%|█████████▌| 10247/10682 [1:27:11<03:35,  2.02it/s]
 96%|█████████▌| 10248/10682 [1:27:12<03:34,  2.02it/s]
 96%|█████████▌| 10249/10682 [1:27:12<03:34,  2.02it/s]
 96%|█████████▌| 10250/10682 [1:27:13<03:33,  2.02it/s]{'loss': 3.3723, 'grad_norm': 0.20746858417987823, 'learning_rate': 4.974711304941093e-06, 'epoch': 0.96}
                                                       
-
 96%|█████████▌| 10250/10682 [1:27:13<03:33,  2.02it/s]
 96%|█████████▌| 10251/10682 [1:27:13<03:33,  2.01it/s]
 96%|█████████▌| 10252/10682 [1:27:14<03:33,  2.02it/s]
 96%|█████████▌| 10253/10682 [1:27:14<03:32,  2.02it/s]
 96%|█████████▌| 10254/10682 [1:27:15<03:32,  2.02it/s]
 96%|█████████▌| 10255/10682 [1:27:15<03:31,  2.02it/s]
 96%|█████████▌| 10256/10682 [1:27:16<03:31,  2.02it/s]
 96%|█████████▌| 10257/10682 [1:27:16<03:30,  2.02it/s]
 96%|█████████▌| 10258/10682 [1:27:17<03:30,  2.02it/s]
 96%|█████████▌| 10259/10682 [1:27:17<03:29,  2.02it/s]
 96%|█████████▌| 10260/10682 [1:27:17<03:29,  2.02it/s]
 96%|█████████▌| 10261/10682 [1:27:18<03:28,  2.02it/s]
 96%|█████████▌| 10262/10682 [1:27:18<03:28,  2.02it/s]
 96%|█████████▌| 10263/10682 [1:27:19<03:27,  2.02it/s]
 96%|█████████▌| 10264/10682 [1:27:19<03:27,  2.02it/s]
 96%|█████████▌| 10265/10682 [1:27:20<03:26,  2.02it/s]
 96%|█████████▌| 10266/10682 [1:27:20<03:26,  2.02it/s]
 96%|█████████▌| 10267/10682 [1:27:21<03:25,  2.02it/s]
 96%|█████████▌| 10268/10682 [1:27:21<03:25,  2.02it/s]
 96%|█████████▌| 10269/10682 [1:27:22<03:24,  2.02it/s]
 96%|█████████▌| 10270/10682 [1:27:22<03:24,  2.02it/s]
 96%|█████████▌| 10271/10682 [1:27:23<03:23,  2.02it/s]
 96%|█████████▌| 10272/10682 [1:27:23<03:23,  2.02it/s]
 96%|█████████▌| 10273/10682 [1:27:24<03:22,  2.02it/s]
 96%|█████████▌| 10274/10682 [1:27:24<03:22,  2.02it/s]
 96%|█████████▌| 10275/10682 [1:27:25<03:21,  2.02it/s]{'loss': 3.3691, 'grad_norm': 0.2088080495595932, 'learning_rate': 4.416419388921844e-06, 'epoch': 0.96}
                                                       
-
 96%|█████████▌| 10275/10682 [1:27:25<03:21,  2.02it/s]
 96%|█████████▌| 10276/10682 [1:27:25<03:21,  2.01it/s]
 96%|█████████▌| 10277/10682 [1:27:26<03:21,  2.01it/s]
 96%|█████████▌| 10278/10682 [1:27:26<03:20,  2.02it/s]
 96%|█████████▌| 10279/10682 [1:27:27<03:19,  2.02it/s]
 96%|█████████▌| 10280/10682 [1:27:27<03:18,  2.02it/s]
 96%|█████████▌| 10281/10682 [1:27:28<03:18,  2.02it/s]
 96%|█████████▋| 10282/10682 [1:27:28<03:18,  2.02it/s]
 96%|█████████▋| 10283/10682 [1:27:29<03:18,  2.01it/s]
 96%|█████████▋| 10284/10682 [1:27:29<03:17,  2.01it/s]
 96%|█████████▋| 10285/10682 [1:27:30<03:17,  2.01it/s]
 96%|█████████▋| 10286/10682 [1:27:30<03:16,  2.01it/s]
 96%|█████████▋| 10287/10682 [1:27:31<03:16,  2.01it/s]
 96%|█████████▋| 10288/10682 [1:27:31<03:15,  2.02it/s]
 96%|█████████▋| 10289/10682 [1:27:32<03:15,  2.01it/s]
 96%|█████████▋| 10290/10682 [1:27:32<03:14,  2.01it/s]
 96%|█████████▋| 10291/10682 [1:27:33<03:14,  2.01it/s]
 96%|█████████▋| 10292/10682 [1:27:33<03:13,  2.01it/s]
 96%|█████████▋| 10293/10682 [1:27:34<03:12,  2.02it/s]
 96%|█████████▋| 10294/10682 [1:27:34<03:12,  2.01it/s]
 96%|█████████▋| 10295/10682 [1:27:35<03:11,  2.02it/s]
 96%|█████████▋| 10296/10682 [1:27:35<03:11,  2.02it/s]
 96%|█████████▋| 10297/10682 [1:27:36<03:10,  2.02it/s]
 96%|█████████▋| 10298/10682 [1:27:36<03:10,  2.02it/s]
 96%|█████████▋| 10299/10682 [1:27:37<03:09,  2.02it/s]
 96%|█████████▋| 10300/10682 [1:27:37<03:09,  2.02it/s]{'loss': 3.3839, 'grad_norm': 0.21643024682998657, 'learning_rate': 3.891208300917604e-06, 'epoch': 0.96}
-
                                                       

 96%|█████████▋| 10300/10682 [1:27:37<03:09,  2.02it/s]
 96%|█████████▋| 10301/10682 [1:27:38<03:09,  2.01it/s]
 96%|█████████▋| 10302/10682 [1:27:38<03:08,  2.02it/s]
 96%|█████████▋| 10303/10682 [1:27:39<03:07,  2.02it/s]
 96%|█████████▋| 10304/10682 [1:27:39<03:07,  2.02it/s]
 96%|█████████▋| 10305/10682 [1:27:40<03:07,  2.02it/s]
 96%|█████████▋| 10306/10682 [1:27:40<03:06,  2.01it/s]
 96%|█████████▋| 10307/10682 [1:27:41<03:06,  2.01it/s]
 96%|█████████▋| 10308/10682 [1:27:41<03:05,  2.01it/s]
 97%|█████████▋| 10309/10682 [1:27:42<03:04,  2.02it/s]
 97%|█████████▋| 10310/10682 [1:27:42<03:04,  2.01it/s]
 97%|█████████▋| 10311/10682 [1:27:43<03:04,  2.02it/s]
 97%|█████████▋| 10312/10682 [1:27:43<03:03,  2.01it/s]
 97%|█████████▋| 10313/10682 [1:27:44<03:03,  2.01it/s]
 97%|█████████▋| 10314/10682 [1:27:44<03:02,  2.02it/s]
 97%|█████████▋| 10315/10682 [1:27:45<03:01,  2.02it/s]
 97%|█████████▋| 10316/10682 [1:27:45<03:01,  2.02it/s]
 97%|█████████▋| 10317/10682 [1:27:46<03:01,  2.01it/s]
 97%|█████████▋| 10318/10682 [1:27:46<03:00,  2.02it/s]
 97%|█████████▋| 10319/10682 [1:27:47<03:00,  2.02it/s]
 97%|█████████▋| 10320/10682 [1:27:47<02:59,  2.02it/s]
 97%|█████████▋| 10321/10682 [1:27:48<02:59,  2.02it/s]
 97%|█████████▋| 10322/10682 [1:27:48<02:58,  2.02it/s]
 97%|█████████▋| 10323/10682 [1:27:49<02:58,  2.02it/s]
 97%|█████████▋| 10324/10682 [1:27:49<02:57,  2.01it/s]
 97%|█████████▋| 10325/10682 [1:27:50<02:57,  2.01it/s]{'loss': 3.3778, 'grad_norm': 0.21007899940013885, 'learning_rate': 3.3991130994299734e-06, 'epoch': 0.97}
                                                       
-
 97%|█████████▋| 10325/10682 [1:27:50<02:57,  2.01it/s]
 97%|█████████▋| 10326/10682 [1:27:50<02:56,  2.01it/s]
 97%|█████████▋| 10327/10682 [1:27:51<02:56,  2.01it/s]
 97%|█████████▋| 10328/10682 [1:27:51<02:55,  2.01it/s]
 97%|█████████▋| 10329/10682 [1:27:52<02:55,  2.02it/s]
 97%|█████████▋| 10330/10682 [1:27:52<03:23,  1.73it/s]
 97%|█████████▋| 10331/10682 [1:27:53<03:13,  1.81it/s]
 97%|█████████▋| 10332/10682 [1:27:53<03:07,  1.86it/s]
 97%|█████████▋| 10333/10682 [1:27:54<03:02,  1.91it/s]
 97%|█████████▋| 10334/10682 [1:27:54<02:59,  1.94it/s]
 97%|█████████▋| 10335/10682 [1:27:55<02:56,  1.96it/s]
 97%|█████████▋| 10336/10682 [1:27:55<02:54,  1.98it/s]
 97%|█████████▋| 10337/10682 [1:27:56<02:53,  1.99it/s]
 97%|█████████▋| 10338/10682 [1:27:56<02:52,  2.00it/s]
 97%|█████████▋| 10339/10682 [1:27:57<02:51,  2.00it/s]
 97%|█████████▋| 10340/10682 [1:27:57<02:50,  2.01it/s]
 97%|█████████▋| 10341/10682 [1:27:58<02:49,  2.01it/s]
 97%|█████████▋| 10342/10682 [1:27:58<02:49,  2.01it/s]
 97%|█████████▋| 10343/10682 [1:27:59<02:48,  2.01it/s]
 97%|█████████▋| 10344/10682 [1:27:59<02:47,  2.01it/s]
 97%|█████████▋| 10345/10682 [1:28:00<02:47,  2.02it/s]
 97%|█████████▋| 10346/10682 [1:28:00<02:46,  2.02it/s]
 97%|█████████▋| 10347/10682 [1:28:01<02:46,  2.02it/s]
 97%|█████████▋| 10348/10682 [1:28:01<02:45,  2.02it/s]
 97%|█████████▋| 10349/10682 [1:28:02<02:44,  2.02it/s]
 97%|█████████▋| 10350/10682 [1:28:02<02:44,  2.02it/s]{'loss': 3.3712, 'grad_norm': 0.20846666395664215, 'learning_rate': 2.940166632433183e-06, 'epoch': 0.97}
                                                       
-
 97%|█████████▋| 10350/10682 [1:28:02<02:44,  2.02it/s]
 97%|█████████▋| 10351/10682 [1:28:03<02:44,  2.01it/s]
 97%|█████████▋| 10352/10682 [1:28:03<02:43,  2.02it/s]
 97%|█████████▋| 10353/10682 [1:28:04<02:43,  2.02it/s]
 97%|█████████▋| 10354/10682 [1:28:04<02:42,  2.02it/s]
 97%|█████████▋| 10355/10682 [1:28:05<02:42,  2.02it/s]
 97%|█████████▋| 10356/10682 [1:28:05<02:41,  2.02it/s]
 97%|█████████▋| 10357/10682 [1:28:06<02:41,  2.01it/s]
 97%|█████████▋| 10358/10682 [1:28:06<02:40,  2.02it/s]
 97%|█████████▋| 10359/10682 [1:28:07<02:40,  2.02it/s]
 97%|█████████▋| 10360/10682 [1:28:07<02:39,  2.02it/s]
 97%|█████████▋| 10361/10682 [1:28:08<02:39,  2.02it/s]
 97%|█████████▋| 10362/10682 [1:28:08<02:38,  2.02it/s]
 97%|█████████▋| 10363/10682 [1:28:09<02:38,  2.02it/s]
 97%|█████████▋| 10364/10682 [1:28:09<02:37,  2.02it/s]
 97%|█████████▋| 10365/10682 [1:28:10<02:37,  2.02it/s]
 97%|█████████▋| 10366/10682 [1:28:10<02:36,  2.02it/s]
 97%|█████████▋| 10367/10682 [1:28:11<02:36,  2.01it/s]
 97%|█████████▋| 10368/10682 [1:28:11<02:35,  2.02it/s]
 97%|█████████▋| 10369/10682 [1:28:12<02:35,  2.02it/s]
 97%|█████████▋| 10370/10682 [1:28:12<02:34,  2.02it/s]
 97%|█████████▋| 10371/10682 [1:28:13<02:34,  2.02it/s]
 97%|█████████▋| 10372/10682 [1:28:13<02:33,  2.02it/s]
 97%|█████████▋| 10373/10682 [1:28:14<02:58,  1.73it/s]
 97%|█████████▋| 10374/10682 [1:28:15<02:49,  1.81it/s]
 97%|█████████▋| 10375/10682 [1:28:15<02:44,  1.87it/s]
                                                       
{'loss': 3.3869, 'grad_norm': 0.21141156554222107, 'learning_rate': 2.5143995351817882e-06, 'epoch': 0.97}
-
 97%|█████████▋| 10375/10682 [1:28:15<02:44,  1.87it/s]
 97%|█████████▋| 10376/10682 [1:28:16<02:40,  1.91it/s]
 97%|█████████▋| 10377/10682 [1:28:16<02:37,  1.94it/s]
 97%|█████████▋| 10378/10682 [1:28:17<02:34,  1.96it/s]
 97%|█████████▋| 10379/10682 [1:28:17<02:33,  1.98it/s]
 97%|█████████▋| 10380/10682 [1:28:18<02:31,  1.99it/s]
 97%|█████████▋| 10381/10682 [1:28:18<02:30,  2.00it/s]
 97%|█████████▋| 10382/10682 [1:28:19<02:29,  2.01it/s]
 97%|█████████▋| 10383/10682 [1:28:19<02:28,  2.01it/s]
 97%|█████████▋| 10384/10682 [1:28:20<02:28,  2.01it/s]
 97%|█████████▋| 10385/10682 [1:28:20<02:27,  2.02it/s]
 97%|█████████▋| 10386/10682 [1:28:21<02:26,  2.01it/s]
 97%|█████████▋| 10387/10682 [1:28:21<02:26,  2.02it/s]
 97%|█████████▋| 10388/10682 [1:28:22<02:25,  2.02it/s]
 97%|█████████▋| 10389/10682 [1:28:22<02:25,  2.02it/s]
 97%|█████████▋| 10390/10682 [1:28:23<02:24,  2.02it/s]
 97%|█████████▋| 10391/10682 [1:28:23<02:24,  2.01it/s]
 97%|█████████▋| 10392/10682 [1:28:24<02:24,  2.01it/s]
 97%|█████████▋| 10393/10682 [1:28:24<02:23,  2.02it/s]
 97%|█████████▋| 10394/10682 [1:28:24<02:22,  2.02it/s]
 97%|█████████▋| 10395/10682 [1:28:25<02:22,  2.02it/s]
 97%|█████████▋| 10396/10682 [1:28:25<02:21,  2.02it/s]
 97%|█████████▋| 10397/10682 [1:28:26<02:21,  2.02it/s]
 97%|█████████▋| 10398/10682 [1:28:26<02:20,  2.02it/s]
 97%|█████████▋| 10399/10682 [1:28:27<02:20,  2.02it/s]
 97%|█████████▋| 10400/10682 [1:28:27<02:19,  2.02it/s]{'loss': 3.3781, 'grad_norm': 0.2086128443479538, 'learning_rate': 2.1218402281655835e-06, 'epoch': 0.97}
                                                       
-
 97%|█████████▋| 10400/10682 [1:28:27<02:19,  2.02it/s]
 97%|█████████▋| 10401/10682 [1:28:28<02:19,  2.01it/s]
 97%|█████████▋| 10402/10682 [1:28:28<02:18,  2.01it/s]
 97%|█████████▋| 10403/10682 [1:28:29<02:18,  2.02it/s]
 97%|█████████▋| 10404/10682 [1:28:29<02:17,  2.02it/s]
 97%|█████████▋| 10405/10682 [1:28:30<02:17,  2.02it/s]
 97%|█████████▋| 10406/10682 [1:28:30<02:17,  2.01it/s]
 97%|█████████▋| 10407/10682 [1:28:31<02:16,  2.02it/s]
 97%|█████████▋| 10408/10682 [1:28:31<02:16,  2.01it/s]
 97%|█████████▋| 10409/10682 [1:28:32<02:15,  2.01it/s]
 97%|█████████▋| 10410/10682 [1:28:32<02:15,  2.01it/s]
 97%|█████████▋| 10411/10682 [1:28:33<02:14,  2.02it/s]
 97%|█████████▋| 10412/10682 [1:28:33<02:13,  2.02it/s]
 97%|█████████▋| 10413/10682 [1:28:34<02:13,  2.02it/s]
 97%|█████████▋| 10414/10682 [1:28:34<02:12,  2.02it/s]
 98%|█████████▊| 10415/10682 [1:28:35<02:12,  2.02it/s]
 98%|█████████▊| 10416/10682 [1:28:35<02:12,  2.01it/s]
 98%|█████████▊| 10417/10682 [1:28:36<02:11,  2.01it/s]
 98%|█████████▊| 10418/10682 [1:28:36<02:10,  2.02it/s]
 98%|█████████▊| 10419/10682 [1:28:37<02:10,  2.02it/s]
 98%|█████████▊| 10420/10682 [1:28:37<02:09,  2.02it/s]
 98%|█████████▊| 10421/10682 [1:28:38<02:09,  2.02it/s]
 98%|█████████▊| 10422/10682 [1:28:38<02:09,  2.02it/s]
 98%|█████████▊| 10423/10682 [1:28:39<02:08,  2.02it/s]
 98%|█████████▊| 10424/10682 [1:28:39<02:08,  2.02it/s]
 98%|█████████▊| 10425/10682 [1:28:40<02:07,  2.02it/s]
                                                       
{'loss': 3.369, 'grad_norm': 0.20524001121520996, 'learning_rate': 1.7625149152127318e-06, 'epoch': 0.98}
-
 98%|█████████▊| 10425/10682 [1:28:40<02:07,  2.02it/s]
 98%|█████████▊| 10426/10682 [1:28:40<02:07,  2.01it/s]
 98%|█████████▊| 10427/10682 [1:28:41<02:06,  2.01it/s]
 98%|█████████▊| 10428/10682 [1:28:41<02:06,  2.01it/s]
 98%|█████████▊| 10429/10682 [1:28:42<02:05,  2.02it/s]
 98%|█████████▊| 10430/10682 [1:28:42<02:04,  2.02it/s]
 98%|█████████▊| 10431/10682 [1:28:43<02:04,  2.01it/s]
 98%|█████████▊| 10432/10682 [1:28:43<02:04,  2.01it/s]
 98%|█████████▊| 10433/10682 [1:28:44<02:03,  2.01it/s]
 98%|█████████▊| 10434/10682 [1:28:44<02:03,  2.01it/s]
 98%|█████████▊| 10435/10682 [1:28:45<02:02,  2.02it/s]
 98%|█████████▊| 10436/10682 [1:28:45<02:01,  2.02it/s]
 98%|█████████▊| 10437/10682 [1:28:46<02:01,  2.02it/s]
 98%|█████████▊| 10438/10682 [1:28:46<02:01,  2.01it/s]
 98%|█████████▊| 10439/10682 [1:28:47<02:00,  2.02it/s]
 98%|█████████▊| 10440/10682 [1:28:47<02:00,  2.01it/s]
 98%|█████████▊| 10441/10682 [1:28:48<01:59,  2.02it/s]
 98%|█████████▊| 10442/10682 [1:28:48<01:59,  2.02it/s]
 98%|█████████▊| 10443/10682 [1:28:49<01:58,  2.01it/s]
 98%|█████████▊| 10444/10682 [1:28:49<01:58,  2.02it/s]
 98%|█████████▊| 10445/10682 [1:28:50<01:57,  2.02it/s]
 98%|█████████▊| 10446/10682 [1:28:50<01:56,  2.02it/s]
 98%|█████████▊| 10447/10682 [1:28:51<01:56,  2.02it/s]
 98%|█████████▊| 10448/10682 [1:28:51<01:56,  2.02it/s]
 98%|█████████▊| 10449/10682 [1:28:52<01:55,  2.01it/s]
 98%|█████████▊| 10450/10682 [1:28:52<01:55,  2.01it/s]{'loss': 3.3687, 'grad_norm': 0.21427346765995026, 'learning_rate': 1.4364475817401635e-06, 'epoch': 0.98}
-
                                                       

 98%|█████████▊| 10450/10682 [1:28:52<01:55,  2.01it/s]
 98%|█████████▊| 10451/10682 [1:28:53<01:54,  2.01it/s]
 98%|█████████▊| 10452/10682 [1:28:53<01:54,  2.02it/s]
 98%|█████████▊| 10453/10682 [1:28:54<01:53,  2.02it/s]
 98%|█████████▊| 10454/10682 [1:28:54<01:52,  2.02it/s]
 98%|█████████▊| 10455/10682 [1:28:55<01:52,  2.02it/s]
 98%|█████████▊| 10456/10682 [1:28:55<01:52,  2.01it/s]
 98%|█████████▊| 10457/10682 [1:28:56<01:51,  2.01it/s]
 98%|█████████▊| 10458/10682 [1:28:56<01:51,  2.02it/s]
 98%|█████████▊| 10459/10682 [1:28:57<01:50,  2.02it/s]
 98%|█████████▊| 10460/10682 [1:28:57<01:50,  2.02it/s]
 98%|█████████▊| 10461/10682 [1:28:58<01:49,  2.02it/s]
 98%|█████████▊| 10462/10682 [1:28:58<01:49,  2.02it/s]
 98%|█████████▊| 10463/10682 [1:28:59<01:48,  2.02it/s]
 98%|█████████▊| 10464/10682 [1:28:59<01:48,  2.02it/s]
 98%|█████████▊| 10465/10682 [1:29:00<01:47,  2.02it/s]
 98%|█████████▊| 10466/10682 [1:29:00<01:47,  2.02it/s]
 98%|█████████▊| 10467/10682 [1:29:01<01:46,  2.02it/s]
 98%|█████████▊| 10468/10682 [1:29:01<01:46,  2.02it/s]
 98%|█████████▊| 10469/10682 [1:29:02<01:45,  2.02it/s]
 98%|█████████▊| 10470/10682 [1:29:02<01:45,  2.02it/s]
 98%|█████████▊| 10471/10682 [1:29:03<01:44,  2.02it/s]
 98%|█████████▊| 10472/10682 [1:29:03<01:44,  2.01it/s]
 98%|█████████▊| 10473/10682 [1:29:04<01:43,  2.01it/s]
 98%|█████████▊| 10474/10682 [1:29:04<01:43,  2.02it/s]
 98%|█████████▊| 10475/10682 [1:29:05<01:42,  2.02it/s]{'loss': 3.3836, 'grad_norm': 0.2093481868505478, 'learning_rate': 1.143659993153079e-06, 'epoch': 0.98}
                                                       
-
 98%|█████████▊| 10475/10682 [1:29:05<01:42,  2.02it/s]
 98%|█████████▊| 10476/10682 [1:29:05<01:42,  2.01it/s]
 98%|█████████▊| 10477/10682 [1:29:06<01:41,  2.01it/s]
 98%|█████████▊| 10478/10682 [1:29:06<01:41,  2.02it/s]
 98%|█████████▊| 10479/10682 [1:29:07<01:40,  2.01it/s]
 98%|█████████▊| 10480/10682 [1:29:07<01:40,  2.02it/s]
 98%|█████████▊| 10481/10682 [1:29:08<01:39,  2.02it/s]
 98%|█████████▊| 10482/10682 [1:29:08<01:39,  2.02it/s]
 98%|█████████▊| 10483/10682 [1:29:09<01:38,  2.02it/s]
 98%|█████████▊| 10484/10682 [1:29:09<01:38,  2.02it/s]
 98%|█████████▊| 10485/10682 [1:29:10<01:37,  2.02it/s]
 98%|█████████▊| 10486/10682 [1:29:10<01:37,  2.02it/s]
 98%|█████████▊| 10487/10682 [1:29:11<01:36,  2.02it/s]
 98%|█████████▊| 10488/10682 [1:29:11<01:36,  2.02it/s]
 98%|█████████▊| 10489/10682 [1:29:12<01:35,  2.01it/s]
 98%|█████████▊| 10490/10682 [1:29:12<01:35,  2.02it/s]
 98%|█████████▊| 10491/10682 [1:29:13<01:34,  2.02it/s]
 98%|█████████▊| 10492/10682 [1:29:13<01:34,  2.02it/s]
 98%|█████████▊| 10493/10682 [1:29:14<01:33,  2.01it/s]
 98%|█████████▊| 10494/10682 [1:29:14<01:33,  2.01it/s]
 98%|█████████▊| 10495/10682 [1:29:15<01:32,  2.01it/s]
 98%|█████████▊| 10496/10682 [1:29:15<01:32,  2.01it/s]
 98%|█████████▊| 10497/10682 [1:29:16<01:31,  2.01it/s]
 98%|█████████▊| 10498/10682 [1:29:16<01:31,  2.01it/s]
 98%|█████████▊| 10499/10682 [1:29:17<01:30,  2.02it/s]
 98%|█████████▊| 10500/10682 [1:29:17<01:30,  2.02it/s]
                                                       
{'loss': 3.3776, 'grad_norm': 0.20782579481601715, 'learning_rate': 8.841716933915555e-07, 'epoch': 0.98}
-
 98%|█████████▊| 10500/10682 [1:29:17<01:30,  2.02it/s]
 98%|█████████▊| 10501/10682 [1:29:18<01:30,  2.01it/s]
 98%|█████████▊| 10502/10682 [1:29:18<01:29,  2.01it/s]
 98%|█████████▊| 10503/10682 [1:29:19<01:28,  2.01it/s]
 98%|█████████▊| 10504/10682 [1:29:19<01:28,  2.02it/s]
 98%|█████████▊| 10505/10682 [1:29:20<01:27,  2.01it/s]
 98%|█████████▊| 10506/10682 [1:29:20<01:27,  2.01it/s]
 98%|█████████▊| 10507/10682 [1:29:21<01:26,  2.01it/s]
 98%|█████████▊| 10508/10682 [1:29:21<01:26,  2.01it/s]
 98%|█████████▊| 10509/10682 [1:29:22<01:25,  2.01it/s]
 98%|█████████▊| 10510/10682 [1:29:22<01:25,  2.01it/s]
 98%|█████████▊| 10511/10682 [1:29:23<01:24,  2.02it/s]
 98%|█████████▊| 10512/10682 [1:29:23<01:24,  2.02it/s]
 98%|█████████▊| 10513/10682 [1:29:24<01:23,  2.02it/s]
 98%|█████████▊| 10514/10682 [1:29:24<01:23,  2.02it/s]
 98%|█████████▊| 10515/10682 [1:29:25<01:22,  2.02it/s]
 98%|█████████▊| 10516/10682 [1:29:25<01:22,  2.02it/s]
 98%|█████████▊| 10517/10682 [1:29:26<01:21,  2.02it/s]
 98%|█████████▊| 10518/10682 [1:29:26<01:21,  2.02it/s]
 98%|█████████▊| 10519/10682 [1:29:27<01:20,  2.02it/s]
 98%|█████████▊| 10520/10682 [1:29:27<01:20,  2.02it/s]
 98%|█████████▊| 10521/10682 [1:29:27<01:19,  2.01it/s]
 99%|█████████▊| 10522/10682 [1:29:28<01:19,  2.02it/s]
 99%|█████████▊| 10523/10682 [1:29:28<01:18,  2.02it/s]
 99%|█████████▊| 10524/10682 [1:29:29<01:18,  2.02it/s]
 99%|█████████▊| 10525/10682 [1:29:29<01:17,  2.02it/s]{'loss': 3.3751, 'grad_norm': 0.20996831357479095, 'learning_rate': 6.580000036264244e-07, 'epoch': 0.99}
                                                       
-
 99%|█████████▊| 10525/10682 [1:29:29<01:17,  2.02it/s]
 99%|█████████▊| 10526/10682 [1:29:30<01:17,  2.01it/s]
 99%|█████████▊| 10527/10682 [1:29:30<01:17,  2.01it/s]
 99%|█████████▊| 10528/10682 [1:29:31<01:16,  2.01it/s]
 99%|█████████▊| 10529/10682 [1:29:31<01:16,  2.01it/s]
 99%|█████████▊| 10530/10682 [1:29:32<01:15,  2.01it/s]
 99%|█████████▊| 10531/10682 [1:29:32<01:15,  2.01it/s]
 99%|█████████▊| 10532/10682 [1:29:33<01:14,  2.01it/s]
 99%|█████████▊| 10533/10682 [1:29:33<01:13,  2.02it/s]
 99%|█████████▊| 10534/10682 [1:29:34<01:13,  2.02it/s]
 99%|█████████▊| 10535/10682 [1:29:34<01:12,  2.02it/s]
 99%|█████████▊| 10536/10682 [1:29:35<01:12,  2.02it/s]
 99%|█████████▊| 10537/10682 [1:29:35<01:11,  2.02it/s]
 99%|█████████▊| 10538/10682 [1:29:36<01:11,  2.02it/s]
 99%|█████████▊| 10539/10682 [1:29:36<01:10,  2.02it/s]
 99%|█████████▊| 10540/10682 [1:29:37<01:10,  2.02it/s]
 99%|█████████▊| 10541/10682 [1:29:37<01:09,  2.02it/s]
 99%|█████████▊| 10542/10682 [1:29:38<01:09,  2.02it/s]
 99%|█████████▊| 10543/10682 [1:29:38<01:08,  2.02it/s]
 99%|█████████▊| 10544/10682 [1:29:39<01:08,  2.02it/s]
 99%|█████████▊| 10545/10682 [1:29:39<01:07,  2.02it/s]
 99%|█████████▊| 10546/10682 [1:29:40<01:07,  2.01it/s]
 99%|█████████▊| 10547/10682 [1:29:40<01:06,  2.02it/s]
 99%|█████████▊| 10548/10682 [1:29:41<01:06,  2.01it/s]
 99%|█████████▉| 10549/10682 [1:29:41<01:05,  2.02it/s]
 99%|█████████▉| 10550/10682 [1:29:42<01:05,  2.02it/s]
                                                       
{'loss': 3.3682, 'grad_norm': 0.20750252902507782, 'learning_rate': 4.651600211027507e-07, 'epoch': 0.99}
-
 99%|█████████▉| 10550/10682 [1:29:42<01:05,  2.02it/s]
 99%|█████████▉| 10551/10682 [1:29:42<01:05,  2.00it/s]
 99%|█████████▉| 10552/10682 [1:29:43<01:04,  2.01it/s]
 99%|█████████▉| 10553/10682 [1:29:43<01:04,  2.01it/s]
 99%|█████████▉| 10554/10682 [1:29:44<01:03,  2.01it/s]
 99%|█████████▉| 10555/10682 [1:29:44<01:03,  2.01it/s]
 99%|█████████▉| 10556/10682 [1:29:45<01:02,  2.01it/s]
 99%|█████████▉| 10557/10682 [1:29:45<01:02,  2.01it/s]
 99%|█████████▉| 10558/10682 [1:29:46<01:01,  2.01it/s]
 99%|█████████▉| 10559/10682 [1:29:46<01:01,  2.01it/s]
 99%|█████████▉| 10560/10682 [1:29:47<01:00,  2.01it/s]
 99%|█████████▉| 10561/10682 [1:29:47<01:00,  2.01it/s]
 99%|█████████▉| 10562/10682 [1:29:48<00:59,  2.01it/s]
 99%|█████████▉| 10563/10682 [1:29:48<00:59,  2.01it/s]
 99%|█████████▉| 10564/10682 [1:29:49<00:58,  2.02it/s]
 99%|█████████▉| 10565/10682 [1:29:49<00:58,  2.02it/s]
 99%|█████████▉| 10566/10682 [1:29:50<00:57,  2.02it/s]
 99%|█████████▉| 10567/10682 [1:29:50<00:57,  2.02it/s]
 99%|█████████▉| 10568/10682 [1:29:51<00:56,  2.01it/s]
 99%|█████████▉| 10569/10682 [1:29:51<00:56,  2.02it/s]
 99%|█████████▉| 10570/10682 [1:29:52<00:55,  2.01it/s]
 99%|█████████▉| 10571/10682 [1:29:52<00:55,  2.02it/s]
 99%|█████████▉| 10572/10682 [1:29:53<00:54,  2.02it/s]
 99%|█████████▉| 10573/10682 [1:29:53<00:54,  2.02it/s]
 99%|█████████▉| 10574/10682 [1:29:54<00:53,  2.02it/s]
 99%|█████████▉| 10575/10682 [1:29:54<00:53,  2.02it/s]
                                                       
{'loss': 3.3742, 'grad_norm': 0.20899268984794617, 'learning_rate': 3.0566461813213986e-07, 'epoch': 0.99}
-
 99%|█████████▉| 10575/10682 [1:29:54<00:53,  2.02it/s]
 99%|█████████▉| 10576/10682 [1:29:55<00:52,  2.02it/s]
 99%|█████████▉| 10577/10682 [1:29:55<00:52,  2.01it/s]
 99%|█████████▉| 10578/10682 [1:29:56<00:51,  2.01it/s]
 99%|█████████▉| 10579/10682 [1:29:56<00:51,  2.01it/s]
 99%|█████████▉| 10580/10682 [1:29:57<00:50,  2.01it/s]
 99%|█████████▉| 10581/10682 [1:29:57<00:50,  2.01it/s]
 99%|█████████▉| 10582/10682 [1:29:58<00:49,  2.02it/s]
 99%|█████████▉| 10583/10682 [1:29:58<00:49,  2.02it/s]
 99%|█████████▉| 10584/10682 [1:29:59<00:48,  2.01it/s]
 99%|█████████▉| 10585/10682 [1:29:59<00:48,  2.02it/s]
 99%|█████████▉| 10586/10682 [1:30:00<00:47,  2.01it/s]
 99%|█████████▉| 10587/10682 [1:30:00<00:47,  2.01it/s]
 99%|█████████▉| 10588/10682 [1:30:01<00:46,  2.02it/s]
 99%|█████████▉| 10589/10682 [1:30:01<00:46,  2.02it/s]
 99%|█████████▉| 10590/10682 [1:30:02<00:45,  2.02it/s]
 99%|█████████▉| 10591/10682 [1:30:02<00:45,  2.02it/s]
 99%|█████████▉| 10592/10682 [1:30:03<00:44,  2.02it/s]
 99%|█████████▉| 10593/10682 [1:30:03<00:44,  2.02it/s]
 99%|█████████▉| 10594/10682 [1:30:04<00:43,  2.02it/s]
 99%|█████████▉| 10595/10682 [1:30:04<00:43,  2.02it/s]
 99%|█████████▉| 10596/10682 [1:30:05<00:42,  2.02it/s]
 99%|█████████▉| 10597/10682 [1:30:05<00:42,  2.02it/s]
 99%|█████████▉| 10598/10682 [1:30:06<00:41,  2.01it/s]
 99%|█████████▉| 10599/10682 [1:30:06<00:41,  2.02it/s]
 99%|█████████▉| 10600/10682 [1:30:07<00:40,  2.02it/s]{'loss': 3.3676, 'grad_norm': 0.20937329530715942, 'learning_rate': 1.7952444123359167e-07, 'epoch': 0.99}
                                                       
-
 99%|█████████▉| 10600/10682 [1:30:07<00:40,  2.02it/s]
 99%|█████████▉| 10601/10682 [1:30:07<00:40,  2.01it/s]
 99%|█████████▉| 10602/10682 [1:30:08<00:39,  2.01it/s]
 99%|█████████▉| 10603/10682 [1:30:08<00:39,  2.01it/s]
 99%|█████████▉| 10604/10682 [1:30:09<00:38,  2.01it/s]
 99%|█████████▉| 10605/10682 [1:30:09<00:38,  2.01it/s]
 99%|█████████▉| 10606/10682 [1:30:10<00:37,  2.02it/s]
 99%|█████████▉| 10607/10682 [1:30:10<00:37,  2.02it/s]
 99%|█████████▉| 10608/10682 [1:30:11<00:36,  2.02it/s]
 99%|█████████▉| 10609/10682 [1:30:11<00:36,  2.02it/s]
 99%|█████████▉| 10610/10682 [1:30:12<00:35,  2.02it/s]
 99%|█████████▉| 10611/10682 [1:30:12<00:35,  2.02it/s]
 99%|█████████▉| 10612/10682 [1:30:13<00:34,  2.02it/s]
 99%|█████████▉| 10613/10682 [1:30:13<00:34,  2.02it/s]
 99%|█████████▉| 10614/10682 [1:30:14<00:33,  2.02it/s]
 99%|█████████▉| 10615/10682 [1:30:14<00:33,  2.02it/s]
 99%|█████████▉| 10616/10682 [1:30:15<00:32,  2.02it/s]
 99%|█████████▉| 10617/10682 [1:30:15<00:32,  2.02it/s]
 99%|█████████▉| 10618/10682 [1:30:16<00:31,  2.02it/s]
 99%|█████████▉| 10619/10682 [1:30:16<00:31,  2.02it/s]
 99%|█████████▉| 10620/10682 [1:30:17<00:30,  2.02it/s]
 99%|█████████▉| 10621/10682 [1:30:17<00:30,  2.02it/s]
 99%|█████████▉| 10622/10682 [1:30:18<00:29,  2.02it/s]
 99%|█████████▉| 10623/10682 [1:30:18<00:29,  2.02it/s]
 99%|█████████▉| 10624/10682 [1:30:19<00:28,  2.02it/s]
 99%|█████████▉| 10625/10682 [1:30:19<00:28,  2.02it/s]
                                                       {'loss': 3.363, 'grad_norm': 0.21218766272068024, 'learning_rate': 8.674791042273533e-08, 'epoch': 0.99}
-
 99%|█████████▉| 10625/10682 [1:30:19<00:28,  2.02it/s]
 99%|█████████▉| 10626/10682 [1:30:20<00:27,  2.01it/s]
 99%|█████████▉| 10627/10682 [1:30:20<00:27,  2.01it/s]
 99%|█████████▉| 10628/10682 [1:30:21<00:26,  2.01it/s]
100%|█████████▉| 10629/10682 [1:30:21<00:26,  2.01it/s]
100%|█████████▉| 10630/10682 [1:30:22<00:25,  2.01it/s]
100%|█████████▉| 10631/10682 [1:30:22<00:25,  2.01it/s]
100%|█████████▉| 10632/10682 [1:30:23<00:24,  2.02it/s]
100%|█████████▉| 10633/10682 [1:30:23<00:24,  2.01it/s]
100%|█████████▉| 10634/10682 [1:30:24<00:23,  2.01it/s]
100%|█████████▉| 10635/10682 [1:30:24<00:23,  2.01it/s]
100%|█████████▉| 10636/10682 [1:30:25<00:22,  2.02it/s]
100%|█████████▉| 10637/10682 [1:30:25<00:22,  2.02it/s]
100%|█████████▉| 10638/10682 [1:30:26<00:21,  2.01it/s]
100%|█████████▉| 10639/10682 [1:30:26<00:21,  2.01it/s]
100%|█████████▉| 10640/10682 [1:30:27<00:20,  2.01it/s]
100%|█████████▉| 10641/10682 [1:30:27<00:20,  2.02it/s]
100%|█████████▉| 10642/10682 [1:30:28<00:19,  2.02it/s]
100%|█████████▉| 10643/10682 [1:30:28<00:19,  2.02it/s]
100%|█████████▉| 10644/10682 [1:30:29<00:18,  2.02it/s]
100%|█████████▉| 10645/10682 [1:30:29<00:18,  2.01it/s]
100%|█████████▉| 10646/10682 [1:30:30<00:17,  2.02it/s]
100%|█████████▉| 10647/10682 [1:30:30<00:17,  2.02it/s]
100%|█████████▉| 10648/10682 [1:30:31<00:16,  2.02it/s]
100%|█████████▉| 10649/10682 [1:30:31<00:16,  2.02it/s]
100%|█████████▉| 10650/10682 [1:30:32<00:15,  2.01it/s]
                                                       
{'loss': 3.3761, 'grad_norm': 0.2082681804895401, 'learning_rate': 2.7341218649834522e-08, 'epoch': 1.0}
-
100%|█████████▉| 10650/10682 [1:30:32<00:15,  2.01it/s]
100%|█████████▉| 10651/10682 [1:30:32<00:15,  2.01it/s]
100%|█████████▉| 10652/10682 [1:30:33<00:14,  2.01it/s]
100%|█████████▉| 10653/10682 [1:30:33<00:14,  2.01it/s]
100%|█████████▉| 10654/10682 [1:30:34<00:13,  2.01it/s]
100%|█████████▉| 10655/10682 [1:30:34<00:13,  2.02it/s]
100%|█████████▉| 10656/10682 [1:30:34<00:12,  2.02it/s]
100%|█████████▉| 10657/10682 [1:30:35<00:12,  2.01it/s]
100%|█████████▉| 10658/10682 [1:30:35<00:11,  2.01it/s]
100%|█████████▉| 10659/10682 [1:30:36<00:11,  2.01it/s]
100%|█████████▉| 10660/10682 [1:30:36<00:10,  2.01it/s]
100%|█████████▉| 10661/10682 [1:30:37<00:10,  2.01it/s]
100%|█████████▉| 10662/10682 [1:30:37<00:09,  2.02it/s]
100%|█████████▉| 10663/10682 [1:30:38<00:09,  2.02it/s]
100%|█████████▉| 10664/10682 [1:30:38<00:08,  2.02it/s]
100%|█████████▉| 10665/10682 [1:30:39<00:08,  2.02it/s]
100%|█████████▉| 10666/10682 [1:30:39<00:07,  2.02it/s]
100%|█████████▉| 10667/10682 [1:30:40<00:07,  2.02it/s]
100%|█████████▉| 10668/10682 [1:30:40<00:06,  2.02it/s]
100%|█████████▉| 10669/10682 [1:30:41<00:06,  2.02it/s]
100%|█████████▉| 10670/10682 [1:30:41<00:05,  2.01it/s]
100%|█████████▉| 10671/10682 [1:30:42<00:05,  2.02it/s]
100%|█████████▉| 10672/10682 [1:30:42<00:04,  2.02it/s]
100%|█████████▉| 10673/10682 [1:30:43<00:04,  2.02it/s]
100%|█████████▉| 10674/10682 [1:30:43<00:03,  2.02it/s]
100%|█████████▉| 10675/10682 [1:30:44<00:03,  2.02it/s]{'loss': 3.3754, 'grad_norm': 0.2081662118434906, 'learning_rate': 1.3083313863404555e-09, 'epoch': 1.0}
                                                       
-
100%|█████████▉| 10675/10682 [1:30:44<00:03,  2.02it/s]
100%|█████████▉| 10676/10682 [1:30:44<00:02,  2.01it/s]
100%|█████████▉| 10677/10682 [1:30:45<00:02,  2.02it/s]
100%|█████████▉| 10678/10682 [1:30:45<00:01,  2.01it/s]
100%|█████████▉| 10679/10682 [1:30:46<00:01,  2.02it/s]
100%|█████████▉| 10680/10682 [1:30:46<00:00,  2.02it/s]
100%|█████████▉| 10681/10682 [1:30:47<00:00,  2.02it/s]
100%|██████████| 10682/10682 [1:30:48<00:00,  1.36it/s]
                                                       
{'train_runtime': 5460.5538, 'train_samples_per_second': 2003.145, 'train_steps_per_second': 1.956, 'train_loss': 3.7958678581410634, 'epoch': 1.0}
-
100%|██████████| 10682/10682 [1:31:00<00:00,  1.36it/s]
100%|██████████| 10682/10682 [1:31:00<00:00,  1.96it/s]
-Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
+
  0%|          | 1/10682 [00:05<16:44:48,  5.64s/it]
  0%|          | 2/10682 [00:07<10:23:57,  3.51s/it]
  0%|          | 3/10682 [00:09<7:30:27,  2.53s/it] 
  0%|          | 4/10682 [00:10<5:56:33,  2.00s/it]
  0%|          | 5/10682 [00:11<4:55:06,  1.66s/it]
  0%|          | 6/10682 [00:12<4:12:00,  1.42s/it]
  0%|          | 7/10682 [00:13<3:39:27,  1.23s/it]
  0%|          | 8/10682 [00:14<3:25:53,  1.16s/it]
  0%|          | 9/10682 [00:14<3:03:08,  1.03s/it]
  0%|          | 10/10682 [00:15<2:45:50,  1.07it/s]
  0%|          | 11/10682 [00:16<2:31:25,  1.17it/s]
  0%|          | 12/10682 [00:16<2:21:35,  1.26it/s]
  0%|          | 13/10682 [00:17<2:15:16,  1.31it/s]
  0%|          | 14/10682 [00:18<2:13:30,  1.33it/s]
  0%|          | 15/10682 [00:19<2:12:25,  1.34it/s]
  0%|          | 16/10682 [00:19<2:06:37,  1.40it/s]
  0%|          | 17/10682 [00:20<1:58:32,  1.50it/s]
  0%|          | 18/10682 [00:20<1:52:17,  1.58it/s]
  0%|          | 19/10682 [00:21<1:48:12,  1.64it/s]
  0%|          | 20/10682 [00:21<1:46:42,  1.67it/s]
  0%|          | 21/10682 [00:22<1:47:47,  1.65it/s]
  0%|          | 22/10682 [00:23<1:49:27,  1.62it/s]
  0%|          | 23/10682 [00:23<1:48:05,  1.64it/s]
  0%|          | 24/10682 [00:24<1:46:20,  1.67it/s]
  0%|          | 25/10682 [00:24<1:45:02,  1.69it/s]
                                                    
{'loss': 10.6537, 'grad_norm': 1.41543710231781, 'learning_rate': 2.3386342376052384e-05, 'epoch': 0.03}
+
  0%|          | 25/10682 [00:24<1:45:02,  1.69it/s]
  0%|          | 26/10682 [00:25<1:43:15,  1.72it/s]
  0%|          | 27/10682 [00:25<1:39:36,  1.78it/s]
  0%|          | 28/10682 [00:26<1:39:21,  1.79it/s]
  0%|          | 29/10682 [00:27<1:39:05,  1.79it/s]
  0%|          | 30/10682 [00:27<1:36:12,  1.85it/s]
  0%|          | 31/10682 [00:28<1:36:57,  1.83it/s]
  0%|          | 32/10682 [00:28<1:35:48,  1.85it/s]
  0%|          | 33/10682 [00:29<1:34:50,  1.87it/s]
  0%|          | 34/10682 [00:29<1:33:54,  1.89it/s]
  0%|          | 35/10682 [00:30<1:33:59,  1.89it/s]
  0%|          | 36/10682 [00:30<1:32:50,  1.91it/s]
  0%|          | 37/10682 [00:31<1:32:06,  1.93it/s]
  0%|          | 38/10682 [00:31<1:31:45,  1.93it/s]
  0%|          | 39/10682 [00:32<1:31:37,  1.94it/s]
  0%|          | 40/10682 [00:32<1:31:30,  1.94it/s]
  0%|          | 41/10682 [00:33<1:31:35,  1.94it/s]
  0%|          | 42/10682 [00:33<1:30:52,  1.95it/s]
  0%|          | 43/10682 [00:34<1:30:55,  1.95it/s]
  0%|          | 44/10682 [00:34<1:30:16,  1.96it/s]
  0%|          | 45/10682 [00:35<1:30:16,  1.96it/s]
  0%|          | 46/10682 [00:35<1:30:08,  1.97it/s]
  0%|          | 47/10682 [00:36<1:29:40,  1.98it/s]
  0%|          | 48/10682 [00:36<1:29:20,  1.98it/s]
  0%|          | 49/10682 [00:37<1:29:01,  1.99it/s]
  0%|          | 50/10682 [00:37<1:28:42,  2.00it/s]
                                                    
{'loss': 9.9361, 'grad_norm': 1.3332717418670654, 'learning_rate': 4.677268475210477e-05, 'epoch': 0.07}
+
  0%|          | 50/10682 [00:37<1:28:42,  2.00it/s]
  0%|          | 51/10682 [00:38<1:28:37,  2.00it/s]
  0%|          | 52/10682 [00:38<1:28:24,  2.00it/s]
  0%|          | 53/10682 [00:39<1:28:17,  2.01it/s]
  1%|          | 54/10682 [00:39<1:28:05,  2.01it/s]
  1%|          | 55/10682 [00:40<1:28:28,  2.00it/s]
  1%|          | 56/10682 [00:40<1:28:17,  2.01it/s]
  1%|          | 57/10682 [00:41<1:28:12,  2.01it/s]
  1%|          | 58/10682 [00:41<1:28:07,  2.01it/s]
  1%|          | 59/10682 [00:42<1:28:15,  2.01it/s]
  1%|          | 60/10682 [00:42<1:27:57,  2.01it/s]
  1%|          | 61/10682 [00:43<1:27:56,  2.01it/s]
  1%|          | 62/10682 [00:43<1:27:51,  2.01it/s]
  1%|          | 63/10682 [00:44<1:27:48,  2.02it/s]
  1%|          | 64/10682 [00:44<1:27:47,  2.02it/s]
  1%|          | 65/10682 [00:45<1:27:46,  2.02it/s]
  1%|          | 66/10682 [00:45<1:27:44,  2.02it/s]
  1%|          | 67/10682 [00:46<1:27:45,  2.02it/s]
  1%|          | 68/10682 [00:46<1:27:41,  2.02it/s]
  1%|          | 69/10682 [00:47<1:27:40,  2.02it/s]
  1%|          | 70/10682 [00:47<1:27:36,  2.02it/s]
  1%|          | 71/10682 [00:48<1:27:43,  2.02it/s]
  1%|          | 72/10682 [00:48<1:27:38,  2.02it/s]
  1%|          | 73/10682 [00:49<1:27:39,  2.02it/s]
  1%|          | 74/10682 [00:49<1:27:32,  2.02it/s]
  1%|          | 75/10682 [00:50<1:27:36,  2.02it/s]
                                                    
{'loss': 9.2222, 'grad_norm': 1.162742018699646, 'learning_rate': 7.015902712815715e-05, 'epoch': 0.1}
+
  1%|          | 75/10682 [00:50<1:27:36,  2.02it/s]
  1%|          | 76/10682 [00:50<1:27:34,  2.02it/s]
  1%|          | 77/10682 [00:51<1:27:33,  2.02it/s]
  1%|          | 78/10682 [00:51<1:27:28,  2.02it/s]
  1%|          | 79/10682 [00:52<1:27:24,  2.02it/s]
  1%|          | 80/10682 [00:52<1:27:20,  2.02it/s]
  1%|          | 81/10682 [00:53<1:27:18,  2.02it/s]
  1%|          | 82/10682 [00:53<1:27:16,  2.02it/s]
  1%|          | 83/10682 [00:54<1:27:10,  2.03it/s]
  1%|          | 84/10682 [00:54<1:27:05,  2.03it/s]
  1%|          | 85/10682 [00:55<1:27:07,  2.03it/s]
  1%|          | 86/10682 [00:55<1:27:10,  2.03it/s]
  1%|          | 87/10682 [00:56<1:27:09,  2.03it/s]
  1%|          | 88/10682 [00:56<1:27:05,  2.03it/s]
  1%|          | 89/10682 [00:57<1:27:08,  2.03it/s]
  1%|          | 90/10682 [00:57<1:27:10,  2.03it/s]
  1%|          | 91/10682 [00:58<1:27:15,  2.02it/s]
  1%|          | 92/10682 [00:58<1:33:43,  1.88it/s]
  1%|          | 93/10682 [00:59<1:31:45,  1.92it/s]
  1%|          | 94/10682 [00:59<1:30:22,  1.95it/s]
  1%|          | 95/10682 [01:00<1:29:19,  1.98it/s]
  1%|          | 96/10682 [01:00<1:28:39,  1.99it/s]
  1%|          | 97/10682 [01:01<1:28:11,  2.00it/s]
  1%|          | 98/10682 [01:01<1:27:52,  2.01it/s]
  1%|          | 99/10682 [01:02<1:27:41,  2.01it/s]
  1%|          | 100/10682 [01:02<1:27:29,  2.02it/s]
                                                     
{'loss': 8.4379, 'grad_norm': 0.8147327899932861, 'learning_rate': 9.354536950420954e-05, 'epoch': 0.13}
+
  1%|          | 100/10682 [01:02<1:27:29,  2.02it/s]
  1%|          | 101/10682 [01:03<1:27:33,  2.01it/s]
  1%|          | 102/10682 [01:03<1:27:30,  2.02it/s]
  1%|          | 103/10682 [01:04<1:27:25,  2.02it/s]
  1%|          | 104/10682 [01:04<1:27:23,  2.02it/s]
  1%|          | 105/10682 [01:05<1:27:15,  2.02it/s]
  1%|          | 106/10682 [01:05<1:27:11,  2.02it/s]
  1%|          | 107/10682 [01:06<1:27:11,  2.02it/s]
  1%|          | 108/10682 [01:06<1:27:06,  2.02it/s]
  1%|          | 109/10682 [01:07<1:27:05,  2.02it/s]
  1%|          | 110/10682 [01:07<1:27:05,  2.02it/s]
  1%|          | 111/10682 [01:08<1:27:04,  2.02it/s]
  1%|          | 112/10682 [01:08<1:26:59,  2.03it/s]
  1%|          | 113/10682 [01:09<1:27:01,  2.02it/s]
  1%|          | 114/10682 [01:09<1:26:59,  2.02it/s]
  1%|          | 115/10682 [01:10<1:27:00,  2.02it/s]
  1%|          | 116/10682 [01:10<1:27:05,  2.02it/s]
  1%|          | 117/10682 [01:11<1:27:04,  2.02it/s]
  1%|          | 118/10682 [01:11<1:27:02,  2.02it/s]
  1%|          | 119/10682 [01:12<1:27:00,  2.02it/s]
  1%|          | 120/10682 [01:12<1:26:59,  2.02it/s]
  1%|          | 121/10682 [01:13<1:26:58,  2.02it/s]
  1%|          | 122/10682 [01:13<1:26:57,  2.02it/s]
  1%|          | 123/10682 [01:14<1:26:57,  2.02it/s]
  1%|          | 124/10682 [01:14<1:26:56,  2.02it/s]
  1%|          | 125/10682 [01:15<1:26:54,  2.02it/s]
                                                     {'loss': 7.784, 'grad_norm': 0.4891585409641266, 'learning_rate': 0.00011693171188026193, 'epoch': 0.16}
+
  1%|          | 125/10682 [01:15<1:26:54,  2.02it/s]
  1%|          | 126/10682 [01:15<1:27:05,  2.02it/s]
  1%|          | 127/10682 [01:16<1:27:05,  2.02it/s]
  1%|          | 128/10682 [01:16<1:27:00,  2.02it/s]
  1%|          | 129/10682 [01:17<1:27:02,  2.02it/s]
  1%|          | 130/10682 [01:17<1:26:55,  2.02it/s]
  1%|          | 131/10682 [01:18<1:26:59,  2.02it/s]
  1%|          | 132/10682 [01:18<1:26:55,  2.02it/s]
  1%|          | 133/10682 [01:19<1:27:00,  2.02it/s]
  1%|▏         | 134/10682 [01:19<1:26:55,  2.02it/s]
  1%|▏         | 135/10682 [01:20<1:26:57,  2.02it/s]
  1%|▏         | 136/10682 [01:20<1:26:54,  2.02it/s]
  1%|▏         | 137/10682 [01:21<1:26:53,  2.02it/s]
  1%|▏         | 138/10682 [01:21<1:26:49,  2.02it/s]
  1%|▏         | 139/10682 [01:21<1:26:55,  2.02it/s]
  1%|▏         | 140/10682 [01:22<1:26:50,  2.02it/s]
  1%|▏         | 141/10682 [01:22<1:26:52,  2.02it/s]
  1%|▏         | 142/10682 [01:23<1:26:47,  2.02it/s]
  1%|▏         | 143/10682 [01:23<1:26:47,  2.02it/s]
  1%|▏         | 144/10682 [01:24<1:26:48,  2.02it/s]
  1%|▏         | 145/10682 [01:24<1:26:54,  2.02it/s]
  1%|▏         | 146/10682 [01:25<1:26:47,  2.02it/s]
  1%|▏         | 147/10682 [01:25<1:26:48,  2.02it/s]
  1%|▏         | 148/10682 [01:26<1:26:42,  2.02it/s]
  1%|▏         | 149/10682 [01:26<1:26:45,  2.02it/s]
  1%|▏         | 150/10682 [01:27<1:26:42,  2.02it/s]{'loss': 7.3167, 'grad_norm': 0.41034939885139465, 'learning_rate': 0.0001403180542563143, 'epoch': 0.2}
+
                                                     

  1%|▏         | 150/10682 [01:27<1:26:42,  2.02it/s]
  1%|▏         | 151/10682 [01:27<1:26:46,  2.02it/s]
  1%|▏         | 152/10682 [01:28<1:26:46,  2.02it/s]
  1%|▏         | 153/10682 [01:28<1:26:47,  2.02it/s]
  1%|▏         | 154/10682 [01:29<1:26:41,  2.02it/s]
  1%|▏         | 155/10682 [01:29<1:26:38,  2.03it/s]
  1%|▏         | 156/10682 [01:30<1:26:45,  2.02it/s]
  1%|▏         | 157/10682 [01:30<1:26:43,  2.02it/s]
  1%|▏         | 158/10682 [01:31<1:26:42,  2.02it/s]
  1%|▏         | 159/10682 [01:31<1:26:40,  2.02it/s]
  1%|▏         | 160/10682 [01:32<1:26:40,  2.02it/s]
  2%|▏         | 161/10682 [01:32<1:26:39,  2.02it/s]
  2%|▏         | 162/10682 [01:33<1:26:39,  2.02it/s]
  2%|▏         | 163/10682 [01:33<1:26:39,  2.02it/s]
  2%|▏         | 164/10682 [01:34<1:26:36,  2.02it/s]
  2%|▏         | 165/10682 [01:34<1:26:35,  2.02it/s]
  2%|▏         | 166/10682 [01:35<1:26:38,  2.02it/s]
  2%|▏         | 167/10682 [01:35<1:26:39,  2.02it/s]
  2%|▏         | 168/10682 [01:36<1:26:37,  2.02it/s]
  2%|▏         | 169/10682 [01:36<1:26:36,  2.02it/s]
  2%|▏         | 170/10682 [01:37<1:26:33,  2.02it/s]
  2%|▏         | 171/10682 [01:37<1:26:34,  2.02it/s]
  2%|▏         | 172/10682 [01:38<1:26:29,  2.03it/s]
  2%|▏         | 173/10682 [01:38<1:26:31,  2.02it/s]
  2%|▏         | 174/10682 [01:39<1:26:28,  2.03it/s]
  2%|▏         | 175/10682 [01:39<1:26:30,  2.02it/s]{'loss': 6.9165, 'grad_norm': 0.8497403860092163, 'learning_rate': 0.00016370439663236668, 'epoch': 0.23}
+
                                                     

  2%|▏         | 175/10682 [01:39<1:26:30,  2.02it/s]
  2%|▏         | 176/10682 [01:40<1:26:44,  2.02it/s]
  2%|▏         | 177/10682 [01:40<1:26:41,  2.02it/s]
  2%|▏         | 178/10682 [01:41<1:26:42,  2.02it/s]
  2%|▏         | 179/10682 [01:41<1:26:32,  2.02it/s]
  2%|▏         | 180/10682 [01:42<1:26:40,  2.02it/s]
  2%|▏         | 181/10682 [01:42<1:26:31,  2.02it/s]
  2%|▏         | 182/10682 [01:43<1:26:29,  2.02it/s]
  2%|▏         | 183/10682 [01:43<1:26:24,  2.03it/s]
  2%|▏         | 184/10682 [01:44<1:26:24,  2.02it/s]
  2%|▏         | 185/10682 [01:44<1:26:26,  2.02it/s]
  2%|▏         | 186/10682 [01:45<1:26:22,  2.03it/s]
  2%|▏         | 187/10682 [01:45<1:26:25,  2.02it/s]
  2%|▏         | 188/10682 [01:46<1:26:25,  2.02it/s]
  2%|▏         | 189/10682 [01:46<1:26:28,  2.02it/s]
  2%|▏         | 190/10682 [01:47<1:26:22,  2.02it/s]
  2%|▏         | 191/10682 [01:47<1:26:28,  2.02it/s]
  2%|▏         | 192/10682 [01:48<1:26:20,  2.03it/s]
  2%|▏         | 193/10682 [01:48<1:26:20,  2.02it/s]
  2%|▏         | 194/10682 [01:49<1:26:19,  2.02it/s]
  2%|▏         | 195/10682 [01:49<1:26:21,  2.02it/s]
  2%|▏         | 196/10682 [01:50<1:26:20,  2.02it/s]
  2%|▏         | 197/10682 [01:50<1:26:24,  2.02it/s]
  2%|▏         | 198/10682 [01:51<1:26:22,  2.02it/s]
  2%|▏         | 199/10682 [01:51<1:26:20,  2.02it/s]
  2%|▏         | 200/10682 [01:52<1:26:17,  2.02it/s]
                                                     {'loss': 6.593, 'grad_norm': 0.6057019233703613, 'learning_rate': 0.00018709073900841907, 'epoch': 0.26}
+
  2%|▏         | 200/10682 [01:52<1:26:17,  2.02it/s]
  2%|▏         | 201/10682 [01:52<1:26:24,  2.02it/s]
  2%|▏         | 202/10682 [01:53<1:26:21,  2.02it/s]
  2%|▏         | 203/10682 [01:53<1:26:22,  2.02it/s]
  2%|▏         | 204/10682 [01:54<1:26:15,  2.02it/s]
  2%|▏         | 205/10682 [01:54<1:26:19,  2.02it/s]
  2%|▏         | 206/10682 [01:55<1:26:12,  2.03it/s]
  2%|▏         | 207/10682 [01:55<1:26:15,  2.02it/s]
  2%|▏         | 208/10682 [01:56<1:26:14,  2.02it/s]
  2%|▏         | 209/10682 [01:56<1:26:17,  2.02it/s]
  2%|▏         | 210/10682 [01:57<1:26:15,  2.02it/s]
  2%|▏         | 211/10682 [01:57<1:26:14,  2.02it/s]
  2%|▏         | 212/10682 [01:58<1:26:11,  2.02it/s]
  2%|▏         | 213/10682 [01:58<1:26:14,  2.02it/s]
  2%|▏         | 214/10682 [01:59<1:26:10,  2.02it/s]
  2%|▏         | 215/10682 [01:59<1:26:10,  2.02it/s]
  2%|▏         | 216/10682 [02:00<1:26:08,  2.03it/s]
  2%|▏         | 217/10682 [02:00<1:26:13,  2.02it/s]
  2%|▏         | 218/10682 [02:01<1:26:10,  2.02it/s]
  2%|▏         | 219/10682 [02:01<1:26:09,  2.02it/s]
  2%|▏         | 220/10682 [02:02<1:26:09,  2.02it/s]
  2%|▏         | 221/10682 [02:02<1:26:10,  2.02it/s]
  2%|▏         | 222/10682 [02:03<1:26:09,  2.02it/s]
  2%|▏         | 223/10682 [02:03<1:26:07,  2.02it/s]
  2%|▏         | 224/10682 [02:04<1:26:08,  2.02it/s]
  2%|▏         | 225/10682 [02:04<1:26:07,  2.02it/s]{'loss': 6.329, 'grad_norm': 0.678991436958313, 'learning_rate': 0.00021047708138447147, 'epoch': 0.29}
+
                                                     

  2%|▏         | 225/10682 [02:04<1:26:07,  2.02it/s]
  2%|▏         | 226/10682 [02:04<1:26:13,  2.02it/s]
  2%|▏         | 227/10682 [02:05<1:26:09,  2.02it/s]
  2%|▏         | 228/10682 [02:05<1:26:04,  2.02it/s]
  2%|▏         | 229/10682 [02:06<1:26:06,  2.02it/s]
  2%|▏         | 230/10682 [02:06<1:26:03,  2.02it/s]
  2%|▏         | 231/10682 [02:07<1:26:03,  2.02it/s]
  2%|▏         | 232/10682 [02:07<1:25:59,  2.03it/s]
  2%|▏         | 233/10682 [02:08<1:26:03,  2.02it/s]
  2%|▏         | 234/10682 [02:08<1:25:57,  2.03it/s]
  2%|▏         | 235/10682 [02:09<1:25:57,  2.03it/s]
  2%|▏         | 236/10682 [02:09<1:25:57,  2.03it/s]
  2%|▏         | 237/10682 [02:10<1:25:57,  2.03it/s]
  2%|▏         | 238/10682 [02:10<1:25:57,  2.02it/s]
  2%|▏         | 239/10682 [02:11<1:25:54,  2.03it/s]
  2%|▏         | 240/10682 [02:11<1:25:57,  2.02it/s]
  2%|▏         | 241/10682 [02:12<1:25:57,  2.02it/s]
  2%|▏         | 242/10682 [02:12<1:25:59,  2.02it/s]
  2%|▏         | 243/10682 [02:13<1:25:58,  2.02it/s]
  2%|▏         | 244/10682 [02:13<1:25:57,  2.02it/s]
  2%|▏         | 245/10682 [02:14<1:25:56,  2.02it/s]
  2%|▏         | 246/10682 [02:14<1:25:57,  2.02it/s]
  2%|▏         | 247/10682 [02:15<1:25:56,  2.02it/s]
  2%|▏         | 248/10682 [02:15<1:26:00,  2.02it/s]
  2%|▏         | 249/10682 [02:16<1:25:55,  2.02it/s]
  2%|▏         | 250/10682 [02:16<1:25:53,  2.02it/s]
                                                     {'loss': 6.1214, 'grad_norm': 0.848291277885437, 'learning_rate': 0.00023386342376052386, 'epoch': 0.33}
+

  2%|▏         | 250/10682 [02:16<1:25:53,  2.02it/s]
  2%|▏         | 251/10682 [02:17<1:25:54,  2.02it/s]
  2%|▏         | 252/10682 [02:17<1:25:54,  2.02it/s]
  2%|▏         | 253/10682 [02:18<1:25:53,  2.02it/s]
  2%|▏         | 254/10682 [02:18<1:25:53,  2.02it/s]
  2%|▏         | 255/10682 [02:19<1:25:47,  2.03it/s]
  2%|▏         | 256/10682 [02:19<1:25:47,  2.03it/s]
  2%|▏         | 257/10682 [02:20<1:25:44,  2.03it/s]
  2%|▏         | 258/10682 [02:20<1:25:46,  2.03it/s]
  2%|▏         | 259/10682 [02:21<1:25:46,  2.03it/s]
  2%|▏         | 260/10682 [02:21<1:25:45,  2.03it/s]
  2%|▏         | 261/10682 [02:22<1:25:46,  2.02it/s]
  2%|▏         | 262/10682 [02:22<1:25:47,  2.02it/s]
  2%|▏         | 263/10682 [02:23<1:25:47,  2.02it/s]
  2%|▏         | 264/10682 [02:23<1:25:44,  2.03it/s]
  2%|▏         | 265/10682 [02:24<1:25:47,  2.02it/s]
  2%|▏         | 266/10682 [02:24<1:25:42,  2.03it/s]
  2%|▏         | 267/10682 [02:25<1:25:43,  2.02it/s]
  3%|▎         | 268/10682 [02:25<1:25:42,  2.03it/s]
  3%|▎         | 269/10682 [02:26<1:25:40,  2.03it/s]
  3%|▎         | 270/10682 [02:26<1:25:41,  2.03it/s]
  3%|▎         | 271/10682 [02:27<1:25:37,  2.03it/s]
  3%|▎         | 272/10682 [02:27<1:25:39,  2.03it/s]
  3%|▎         | 273/10682 [02:28<1:25:38,  2.03it/s]
  3%|▎         | 274/10682 [02:28<1:25:36,  2.03it/s]
  3%|▎         | 275/10682 [02:29<1:25:33,  2.03it/s]{'loss': 5.955, 'grad_norm': 0.8124875426292419, 'learning_rate': 0.00025724976613657625, 'epoch': 0.36}
+
                                                     

  3%|▎         | 275/10682 [02:29<1:25:33,  2.03it/s]
  3%|▎         | 276/10682 [02:29<1:25:38,  2.03it/s]
  3%|▎         | 277/10682 [02:30<1:25:38,  2.03it/s]
  3%|▎         | 278/10682 [02:30<1:25:37,  2.03it/s]
  3%|▎         | 279/10682 [02:31<1:25:36,  2.03it/s]
  3%|▎         | 280/10682 [02:31<1:25:36,  2.02it/s]
  3%|▎         | 281/10682 [02:32<1:25:37,  2.02it/s]
  3%|▎         | 282/10682 [02:32<1:25:35,  2.03it/s]
  3%|▎         | 283/10682 [02:33<1:25:38,  2.02it/s]
  3%|▎         | 284/10682 [02:33<1:25:33,  2.03it/s]
  3%|▎         | 285/10682 [02:34<1:25:36,  2.02it/s]
  3%|▎         | 286/10682 [02:34<1:25:32,  2.03it/s]
  3%|▎         | 287/10682 [02:35<1:25:32,  2.03it/s]
  3%|▎         | 288/10682 [02:35<1:25:31,  2.03it/s]
  3%|▎         | 289/10682 [02:36<1:25:33,  2.02it/s]
  3%|▎         | 290/10682 [02:36<1:25:29,  2.03it/s]
  3%|▎         | 291/10682 [02:37<1:25:34,  2.02it/s]
  3%|▎         | 292/10682 [02:37<1:25:28,  2.03it/s]
  3%|▎         | 293/10682 [02:38<1:25:31,  2.02it/s]
  3%|▎         | 294/10682 [02:38<1:25:27,  2.03it/s]
  3%|▎         | 295/10682 [02:39<1:25:24,  2.03it/s]
  3%|▎         | 296/10682 [02:39<1:25:25,  2.03it/s]
  3%|▎         | 297/10682 [02:40<1:25:23,  2.03it/s]
  3%|▎         | 298/10682 [02:40<1:25:27,  2.03it/s]
  3%|▎         | 299/10682 [02:41<1:25:25,  2.03it/s]
  3%|▎         | 300/10682 [02:41<1:25:30,  2.02it/s]{'loss': 5.8037, 'grad_norm': 0.545015811920166, 'learning_rate': 0.0002806361085126286, 'epoch': 0.39}
+
                                                     

  3%|▎         | 300/10682 [02:41<1:25:30,  2.02it/s]
  3%|▎         | 301/10682 [02:42<1:25:32,  2.02it/s]
  3%|▎         | 302/10682 [02:42<1:25:33,  2.02it/s]
  3%|▎         | 303/10682 [02:43<1:25:33,  2.02it/s]
  3%|▎         | 304/10682 [02:43<1:25:33,  2.02it/s]
  3%|▎         | 305/10682 [02:44<1:25:33,  2.02it/s]
  3%|▎         | 306/10682 [02:44<1:25:31,  2.02it/s]
  3%|▎         | 307/10682 [02:44<1:25:26,  2.02it/s]
  3%|▎         | 308/10682 [02:45<1:25:33,  2.02it/s]
  3%|▎         | 309/10682 [02:45<1:25:25,  2.02it/s]
  3%|▎         | 310/10682 [02:46<1:25:24,  2.02it/s]
  3%|▎         | 311/10682 [02:46<1:25:20,  2.03it/s]
  3%|▎         | 312/10682 [02:47<1:25:19,  2.03it/s]
  3%|▎         | 313/10682 [02:47<1:25:23,  2.02it/s]
  3%|▎         | 314/10682 [02:48<1:25:21,  2.02it/s]
  3%|▎         | 315/10682 [02:48<1:25:21,  2.02it/s]
  3%|▎         | 316/10682 [02:49<1:25:16,  2.03it/s]
  3%|▎         | 317/10682 [02:49<1:25:13,  2.03it/s]
  3%|▎         | 318/10682 [02:50<1:25:15,  2.03it/s]
  3%|▎         | 319/10682 [02:50<1:25:12,  2.03it/s]
  3%|▎         | 320/10682 [02:51<1:25:11,  2.03it/s]
  3%|▎         | 321/10682 [02:51<1:25:13,  2.03it/s]
  3%|▎         | 322/10682 [02:52<1:25:08,  2.03it/s]
  3%|▎         | 323/10682 [02:52<1:25:10,  2.03it/s]
  3%|▎         | 324/10682 [02:53<1:25:07,  2.03it/s]
  3%|▎         | 325/10682 [02:53<1:25:09,  2.03it/s]{'loss': 5.681, 'grad_norm': 0.6341584324836731, 'learning_rate': 0.00030402245088868103, 'epoch': 0.43}
+
                                                     

  3%|▎         | 325/10682 [02:53<1:25:09,  2.03it/s]
  3%|▎         | 326/10682 [02:54<1:25:14,  2.02it/s]
  3%|▎         | 327/10682 [02:54<1:25:12,  2.03it/s]
  3%|▎         | 328/10682 [02:55<1:25:13,  2.02it/s]
  3%|▎         | 329/10682 [02:55<1:25:11,  2.03it/s]
  3%|▎         | 330/10682 [02:56<1:25:13,  2.02it/s]
  3%|▎         | 331/10682 [02:56<1:25:14,  2.02it/s]
  3%|▎         | 332/10682 [02:57<1:25:13,  2.02it/s]
  3%|▎         | 333/10682 [02:57<1:25:14,  2.02it/s]
  3%|▎         | 334/10682 [02:58<1:25:14,  2.02it/s]
  3%|▎         | 335/10682 [02:58<1:25:11,  2.02it/s]
  3%|▎         | 336/10682 [02:59<1:25:10,  2.02it/s]
  3%|▎         | 337/10682 [02:59<1:25:08,  2.03it/s]
  3%|▎         | 338/10682 [03:00<1:25:12,  2.02it/s]
  3%|▎         | 339/10682 [03:00<1:25:07,  2.03it/s]
  3%|▎         | 340/10682 [03:01<1:25:10,  2.02it/s]
  3%|▎         | 341/10682 [03:01<1:25:04,  2.03it/s]
  3%|▎         | 342/10682 [03:02<1:25:02,  2.03it/s]
  3%|▎         | 343/10682 [03:02<1:25:01,  2.03it/s]
  3%|▎         | 344/10682 [03:03<1:24:59,  2.03it/s]
  3%|▎         | 345/10682 [03:03<1:25:02,  2.03it/s]
  3%|▎         | 346/10682 [03:04<1:25:01,  2.03it/s]
  3%|▎         | 347/10682 [03:04<1:25:04,  2.02it/s]
  3%|▎         | 348/10682 [03:05<1:24:59,  2.03it/s]
  3%|▎         | 349/10682 [03:05<1:25:03,  2.02it/s]
  3%|▎         | 350/10682 [03:06<1:25:00,  2.03it/s]{'loss': 5.562, 'grad_norm': 0.6854812502861023, 'learning_rate': 0.00032740879326473337, 'epoch': 0.46}
+
                                                     

  3%|▎         | 350/10682 [03:06<1:25:00,  2.03it/s]
  3%|▎         | 351/10682 [03:06<1:25:10,  2.02it/s]
  3%|▎         | 352/10682 [03:07<1:25:05,  2.02it/s]
  3%|▎         | 353/10682 [03:07<1:25:06,  2.02it/s]
  3%|▎         | 354/10682 [03:08<1:25:05,  2.02it/s]
  3%|▎         | 355/10682 [03:08<1:25:08,  2.02it/s]
  3%|▎         | 356/10682 [03:09<1:25:03,  2.02it/s]
  3%|▎         | 357/10682 [03:09<1:25:05,  2.02it/s]
  3%|▎         | 358/10682 [03:10<1:25:01,  2.02it/s]
  3%|▎         | 359/10682 [03:10<1:25:02,  2.02it/s]
  3%|▎         | 360/10682 [03:11<1:25:01,  2.02it/s]
  3%|▎         | 361/10682 [03:11<1:25:03,  2.02it/s]
  3%|▎         | 362/10682 [03:12<1:25:00,  2.02it/s]
  3%|▎         | 363/10682 [03:12<1:25:04,  2.02it/s]
  3%|▎         | 364/10682 [03:13<1:24:59,  2.02it/s]
  3%|▎         | 365/10682 [03:13<1:25:02,  2.02it/s]
  3%|▎         | 366/10682 [03:14<1:24:58,  2.02it/s]
  3%|▎         | 367/10682 [03:14<1:24:59,  2.02it/s]
  3%|▎         | 368/10682 [03:15<1:24:55,  2.02it/s]
  3%|▎         | 369/10682 [03:15<1:24:59,  2.02it/s]
  3%|▎         | 370/10682 [03:16<1:25:01,  2.02it/s]
  3%|▎         | 371/10682 [03:16<1:25:01,  2.02it/s]
  3%|▎         | 372/10682 [03:17<1:24:56,  2.02it/s]
  3%|▎         | 373/10682 [03:17<1:24:55,  2.02it/s]
  4%|▎         | 374/10682 [03:18<1:24:51,  2.02it/s]
  4%|▎         | 375/10682 [03:18<1:24:48,  2.03it/s]
                                                     
{'loss': 5.4639, 'grad_norm': 0.7402798533439636, 'learning_rate': 0.0003507951356407858, 'epoch': 0.49}
+
  4%|▎         | 375/10682 [03:18<1:24:48,  2.03it/s]
  4%|▎         | 376/10682 [03:19<1:24:55,  2.02it/s]
  4%|▎         | 377/10682 [03:19<1:24:55,  2.02it/s]
  4%|▎         | 378/10682 [03:20<1:24:56,  2.02it/s]
  4%|▎         | 379/10682 [03:20<1:24:53,  2.02it/s]
  4%|▎         | 380/10682 [03:21<1:24:48,  2.02it/s]
  4%|▎         | 381/10682 [03:21<1:24:51,  2.02it/s]
  4%|▎         | 382/10682 [03:22<1:24:49,  2.02it/s]
  4%|▎         | 383/10682 [03:22<1:24:48,  2.02it/s]
  4%|▎         | 384/10682 [03:23<1:24:49,  2.02it/s]
  4%|▎         | 385/10682 [03:23<1:24:49,  2.02it/s]
  4%|▎         | 386/10682 [03:24<1:24:45,  2.02it/s]
  4%|▎         | 387/10682 [03:24<1:24:43,  2.03it/s]
  4%|▎         | 388/10682 [03:25<1:24:44,  2.02it/s]
  4%|▎         | 389/10682 [03:25<1:24:39,  2.03it/s]
  4%|▎         | 390/10682 [03:26<1:24:39,  2.03it/s]
  4%|▎         | 391/10682 [03:26<1:24:36,  2.03it/s]
  4%|▎         | 392/10682 [03:26<1:24:34,  2.03it/s]
  4%|▎         | 393/10682 [03:27<1:24:37,  2.03it/s]
  4%|▎         | 394/10682 [03:27<1:24:35,  2.03it/s]
  4%|▎         | 395/10682 [03:28<1:24:41,  2.02it/s]
  4%|▎         | 396/10682 [03:28<1:24:39,  2.03it/s]
  4%|▎         | 397/10682 [03:29<1:24:44,  2.02it/s]
  4%|▎         | 398/10682 [03:29<1:24:37,  2.03it/s]
  4%|▎         | 399/10682 [03:30<1:24:42,  2.02it/s]
  4%|▎         | 400/10682 [03:30<1:24:33,  2.03it/s]{'loss': 5.3738, 'grad_norm': 0.6768883466720581, 'learning_rate': 0.00037418147801683815, 'epoch': 0.52}
+
                                                     

  4%|▎         | 400/10682 [03:30<1:24:33,  2.03it/s]
  4%|▍         | 401/10682 [03:31<1:24:40,  2.02it/s]
  4%|▍         | 402/10682 [03:31<1:24:35,  2.03it/s]
  4%|▍         | 403/10682 [03:32<1:24:36,  2.02it/s]
  4%|▍         | 404/10682 [03:32<1:24:38,  2.02it/s]
  4%|▍         | 405/10682 [03:33<1:24:33,  2.03it/s]
  4%|▍         | 406/10682 [03:33<1:24:43,  2.02it/s]
  4%|▍         | 407/10682 [03:34<1:24:37,  2.02it/s]
  4%|▍         | 408/10682 [03:34<1:24:39,  2.02it/s]
  4%|▍         | 409/10682 [03:35<1:24:31,  2.03it/s]
  4%|▍         | 410/10682 [03:35<1:24:35,  2.02it/s]
  4%|▍         | 411/10682 [03:36<1:24:31,  2.03it/s]
  4%|▍         | 412/10682 [03:36<1:24:30,  2.03it/s]
  4%|▍         | 413/10682 [03:37<1:24:30,  2.03it/s]
  4%|▍         | 414/10682 [03:37<1:24:25,  2.03it/s]
  4%|▍         | 415/10682 [03:38<1:24:28,  2.03it/s]
  4%|▍         | 416/10682 [03:38<1:24:24,  2.03it/s]
  4%|▍         | 417/10682 [03:39<1:24:30,  2.02it/s]
  4%|▍         | 418/10682 [03:39<1:24:23,  2.03it/s]
  4%|▍         | 419/10682 [03:40<1:24:27,  2.03it/s]
  4%|▍         | 420/10682 [03:40<1:24:24,  2.03it/s]
  4%|▍         | 421/10682 [03:41<1:24:27,  2.02it/s]
  4%|▍         | 422/10682 [03:41<1:24:23,  2.03it/s]
  4%|▍         | 423/10682 [03:42<1:24:31,  2.02it/s]
  4%|▍         | 424/10682 [03:42<1:24:26,  2.02it/s]
  4%|▍         | 425/10682 [03:43<1:24:24,  2.03it/s]{'loss': 5.2876, 'grad_norm': 0.893210232257843, 'learning_rate': 0.0003975678203928906, 'epoch': 0.56}
                                                     
+
  4%|▍         | 425/10682 [03:43<1:24:24,  2.03it/s]
  4%|▍         | 426/10682 [03:43<1:24:32,  2.02it/s]
  4%|▍         | 427/10682 [03:44<1:24:34,  2.02it/s]
  4%|▍         | 428/10682 [03:44<1:24:30,  2.02it/s]
  4%|▍         | 429/10682 [03:45<1:24:29,  2.02it/s]
  4%|▍         | 430/10682 [03:45<1:24:29,  2.02it/s]
  4%|▍         | 431/10682 [03:46<1:24:29,  2.02it/s]
  4%|▍         | 432/10682 [03:46<1:24:26,  2.02it/s]
  4%|▍         | 433/10682 [03:47<1:24:26,  2.02it/s]
  4%|▍         | 434/10682 [03:47<1:24:29,  2.02it/s]
  4%|▍         | 435/10682 [03:48<1:24:26,  2.02it/s]
  4%|▍         | 436/10682 [03:48<1:24:20,  2.02it/s]
  4%|▍         | 437/10682 [03:49<1:24:25,  2.02it/s]
  4%|▍         | 438/10682 [03:49<1:24:21,  2.02it/s]
  4%|▍         | 439/10682 [03:50<1:24:25,  2.02it/s]
  4%|▍         | 440/10682 [03:50<1:24:17,  2.03it/s]
  4%|▍         | 441/10682 [03:51<1:24:19,  2.02it/s]
  4%|▍         | 442/10682 [03:51<1:24:16,  2.03it/s]
  4%|▍         | 443/10682 [03:52<1:24:21,  2.02it/s]
  4%|▍         | 444/10682 [03:52<1:24:18,  2.02it/s]
  4%|▍         | 445/10682 [03:53<1:24:16,  2.02it/s]
  4%|▍         | 446/10682 [03:53<1:24:19,  2.02it/s]
  4%|▍         | 447/10682 [03:54<1:24:19,  2.02it/s]
  4%|▍         | 448/10682 [03:54<1:24:21,  2.02it/s]
  4%|▍         | 449/10682 [03:55<1:24:20,  2.02it/s]
  4%|▍         | 450/10682 [03:55<1:24:20,  2.02it/s]{'loss': 5.2079, 'grad_norm': 0.6004838347434998, 'learning_rate': 0.00042095416276894293, 'epoch': 0.59}
+
                                                     

  4%|▍         | 450/10682 [03:55<1:24:20,  2.02it/s]
  4%|▍         | 451/10682 [03:56<1:24:22,  2.02it/s]
  4%|▍         | 452/10682 [03:56<1:24:21,  2.02it/s]
  4%|▍         | 453/10682 [03:57<1:24:19,  2.02it/s]
  4%|▍         | 454/10682 [03:57<1:24:17,  2.02it/s]
  4%|▍         | 455/10682 [03:58<1:24:18,  2.02it/s]
  4%|▍         | 456/10682 [03:58<1:24:12,  2.02it/s]
  4%|▍         | 457/10682 [03:59<1:24:12,  2.02it/s]
  4%|▍         | 458/10682 [03:59<1:24:05,  2.03it/s]
  4%|▍         | 459/10682 [04:00<1:24:09,  2.02it/s]
  4%|▍         | 460/10682 [04:00<1:24:06,  2.03it/s]
  4%|▍         | 461/10682 [04:01<1:24:00,  2.03it/s]
  4%|▍         | 462/10682 [04:01<1:24:03,  2.03it/s]
  4%|▍         | 463/10682 [04:02<1:23:59,  2.03it/s]
  4%|▍         | 464/10682 [04:02<1:24:07,  2.02it/s]
  4%|▍         | 465/10682 [04:03<1:24:04,  2.03it/s]
  4%|▍         | 466/10682 [04:03<1:24:11,  2.02it/s]
  4%|▍         | 467/10682 [04:04<1:24:08,  2.02it/s]
  4%|▍         | 468/10682 [04:04<1:24:11,  2.02it/s]
  4%|▍         | 469/10682 [04:05<1:24:08,  2.02it/s]
  4%|▍         | 470/10682 [04:05<1:24:09,  2.02it/s]
  4%|▍         | 471/10682 [04:06<1:24:04,  2.02it/s]
  4%|▍         | 472/10682 [04:06<1:24:01,  2.03it/s]
  4%|▍         | 473/10682 [04:07<1:23:57,  2.03it/s]
  4%|▍         | 474/10682 [04:07<1:24:00,  2.03it/s]
  4%|▍         | 475/10682 [04:07<1:24:00,  2.02it/s]{'loss': 5.1415, 'grad_norm': 0.5068067908287048, 'learning_rate': 0.0004443405051449954, 'epoch': 0.62}
                                                     
+
  4%|▍         | 475/10682 [04:07<1:24:00,  2.02it/s]
  4%|▍         | 476/10682 [04:08<1:24:07,  2.02it/s]
  4%|▍         | 477/10682 [04:08<1:23:59,  2.03it/s]
  4%|▍         | 478/10682 [04:09<1:24:00,  2.02it/s]
  4%|▍         | 479/10682 [04:09<1:23:59,  2.02it/s]
  4%|▍         | 480/10682 [04:10<1:23:55,  2.03it/s]
  5%|▍         | 481/10682 [04:10<1:23:56,  2.03it/s]
  5%|▍         | 482/10682 [04:11<1:23:51,  2.03it/s]
  5%|▍         | 483/10682 [04:11<1:23:56,  2.02it/s]
  5%|▍         | 484/10682 [04:12<1:24:03,  2.02it/s]
  5%|▍         | 485/10682 [04:12<1:24:05,  2.02it/s]
  5%|▍         | 486/10682 [04:13<1:24:00,  2.02it/s]
  5%|▍         | 487/10682 [04:13<1:24:03,  2.02it/s]
  5%|▍         | 488/10682 [04:14<1:23:58,  2.02it/s]
  5%|▍         | 489/10682 [04:14<1:24:00,  2.02it/s]
  5%|▍         | 490/10682 [04:15<1:23:56,  2.02it/s]
  5%|▍         | 491/10682 [04:15<1:23:56,  2.02it/s]
  5%|▍         | 492/10682 [04:16<1:23:52,  2.02it/s]
  5%|▍         | 493/10682 [04:16<1:23:50,  2.03it/s]
  5%|▍         | 494/10682 [04:17<1:23:45,  2.03it/s]
  5%|▍         | 495/10682 [04:17<1:23:43,  2.03it/s]
  5%|▍         | 496/10682 [04:18<1:23:47,  2.03it/s]
  5%|▍         | 497/10682 [04:18<1:23:45,  2.03it/s]
  5%|▍         | 498/10682 [04:19<1:23:47,  2.03it/s]
  5%|▍         | 499/10682 [04:19<1:23:46,  2.03it/s]
  5%|▍         | 500/10682 [04:20<1:23:49,  2.02it/s]
                                                     {'loss': 5.0668, 'grad_norm': 0.6643111705780029, 'learning_rate': 0.0004677268475210477, 'epoch': 0.65}
+
  5%|▍         | 500/10682 [04:20<1:23:49,  2.02it/s]
  5%|▍         | 501/10682 [04:20<1:23:50,  2.02it/s]
  5%|▍         | 502/10682 [04:21<1:23:56,  2.02it/s]
  5%|▍         | 503/10682 [04:21<1:23:50,  2.02it/s]
  5%|▍         | 504/10682 [04:22<1:23:56,  2.02it/s]
  5%|▍         | 505/10682 [04:22<1:23:50,  2.02it/s]
  5%|▍         | 506/10682 [04:23<1:23:52,  2.02it/s]
  5%|▍         | 507/10682 [04:23<1:23:48,  2.02it/s]
  5%|▍         | 508/10682 [04:24<1:23:52,  2.02it/s]
  5%|▍         | 509/10682 [04:24<1:23:46,  2.02it/s]
  5%|▍         | 510/10682 [04:25<1:23:47,  2.02it/s]
  5%|▍         | 511/10682 [04:25<1:23:44,  2.02it/s]
  5%|▍         | 512/10682 [04:26<1:23:45,  2.02it/s]
  5%|▍         | 513/10682 [04:26<1:23:41,  2.03it/s]
  5%|▍         | 514/10682 [04:27<1:23:41,  2.02it/s]
  5%|▍         | 515/10682 [04:27<1:23:38,  2.03it/s]
  5%|▍         | 516/10682 [04:28<1:23:36,  2.03it/s]
  5%|▍         | 517/10682 [04:28<1:23:40,  2.02it/s]
  5%|▍         | 518/10682 [04:29<1:23:34,  2.03it/s]
  5%|▍         | 519/10682 [04:29<1:23:34,  2.03it/s]
  5%|▍         | 520/10682 [04:30<1:23:28,  2.03it/s]
  5%|▍         | 521/10682 [04:30<1:23:32,  2.03it/s]
  5%|▍         | 522/10682 [04:31<1:23:31,  2.03it/s]
  5%|▍         | 523/10682 [04:31<1:23:31,  2.03it/s]
  5%|▍         | 524/10682 [04:32<1:23:32,  2.03it/s]
  5%|▍         | 525/10682 [04:32<1:23:33,  2.03it/s]
                                                     {'loss': 4.9989, 'grad_norm': 0.5118005275726318, 'learning_rate': 0.0004911131898971, 'epoch': 0.69}
+
  5%|▍         | 525/10682 [04:32<1:23:33,  2.03it/s]
  5%|▍         | 526/10682 [04:33<1:23:37,  2.02it/s]
  5%|▍         | 527/10682 [04:33<1:23:31,  2.03it/s]
  5%|▍         | 528/10682 [04:34<1:23:33,  2.03it/s]
  5%|▍         | 529/10682 [04:34<1:23:31,  2.03it/s]
  5%|▍         | 530/10682 [04:35<1:23:38,  2.02it/s]
  5%|▍         | 531/10682 [04:35<1:23:32,  2.03it/s]
  5%|▍         | 532/10682 [04:36<1:23:36,  2.02it/s]
  5%|▍         | 533/10682 [04:36<1:23:33,  2.02it/s]
  5%|▍         | 534/10682 [04:37<1:23:35,  2.02it/s]
  5%|▌         | 535/10682 [04:37<1:23:31,  2.02it/s]
  5%|▌         | 536/10682 [04:38<1:23:27,  2.03it/s]
  5%|▌         | 537/10682 [04:38<1:23:29,  2.03it/s]
  5%|▌         | 538/10682 [04:39<1:23:29,  2.02it/s]
  5%|▌         | 539/10682 [04:39<1:23:33,  2.02it/s]
  5%|▌         | 540/10682 [04:40<1:23:31,  2.02it/s]
  5%|▌         | 541/10682 [04:40<1:23:32,  2.02it/s]
  5%|▌         | 542/10682 [04:41<1:23:31,  2.02it/s]
  5%|▌         | 543/10682 [04:41<1:23:34,  2.02it/s]
  5%|▌         | 544/10682 [04:42<1:23:31,  2.02it/s]
  5%|▌         | 545/10682 [04:42<1:23:35,  2.02it/s]
  5%|▌         | 546/10682 [04:43<1:23:27,  2.02it/s]
  5%|▌         | 547/10682 [04:43<1:23:29,  2.02it/s]
  5%|▌         | 548/10682 [04:44<1:23:23,  2.03it/s]
  5%|▌         | 549/10682 [04:44<1:23:25,  2.02it/s]
  5%|▌         | 550/10682 [04:45<1:23:26,  2.02it/s]{'loss': 4.948, 'grad_norm': 0.5056236982345581, 'learning_rate': 0.0005144995322731525, 'epoch': 0.72}
+
                                                     

  5%|▌         | 550/10682 [04:45<1:23:26,  2.02it/s]
  5%|▌         | 551/10682 [04:45<1:23:32,  2.02it/s]
  5%|▌         | 552/10682 [04:46<1:23:26,  2.02it/s]
  5%|▌         | 553/10682 [04:46<1:23:25,  2.02it/s]
  5%|▌         | 554/10682 [04:47<1:23:21,  2.03it/s]
  5%|▌         | 555/10682 [04:47<1:23:17,  2.03it/s]
  5%|▌         | 556/10682 [04:48<1:23:20,  2.03it/s]
  5%|▌         | 557/10682 [04:48<1:23:16,  2.03it/s]
  5%|▌         | 558/10682 [04:48<1:23:21,  2.02it/s]
  5%|▌         | 559/10682 [04:49<1:23:15,  2.03it/s]
  5%|▌         | 560/10682 [04:49<1:23:15,  2.03it/s]
  5%|▌         | 561/10682 [04:50<1:23:14,  2.03it/s]
  5%|▌         | 562/10682 [04:50<1:23:14,  2.03it/s]
  5%|▌         | 563/10682 [04:51<1:23:15,  2.03it/s]
  5%|▌         | 564/10682 [04:51<1:23:16,  2.02it/s]
  5%|▌         | 565/10682 [04:52<1:23:20,  2.02it/s]
  5%|▌         | 566/10682 [04:52<1:23:14,  2.03it/s]
  5%|▌         | 567/10682 [04:53<1:23:20,  2.02it/s]
  5%|▌         | 568/10682 [04:53<1:23:14,  2.02it/s]
  5%|▌         | 569/10682 [04:54<1:23:20,  2.02it/s]
  5%|▌         | 570/10682 [04:54<1:23:14,  2.02it/s]
  5%|▌         | 571/10682 [04:55<1:23:16,  2.02it/s]
  5%|▌         | 572/10682 [04:55<1:23:12,  2.03it/s]
  5%|▌         | 573/10682 [04:56<1:23:14,  2.02it/s]
  5%|▌         | 574/10682 [04:56<1:23:08,  2.03it/s]
  5%|▌         | 575/10682 [04:57<1:23:05,  2.03it/s]{'loss': 4.885, 'grad_norm': 0.5478827953338623, 'learning_rate': 0.0005378858746492049, 'epoch': 0.75}
+
                                                     

  5%|▌         | 575/10682 [04:57<1:23:05,  2.03it/s]
  5%|▌         | 576/10682 [04:57<1:23:10,  2.02it/s]
  5%|▌         | 577/10682 [04:58<1:23:05,  2.03it/s]
  5%|▌         | 578/10682 [04:58<1:23:09,  2.02it/s]
  5%|▌         | 579/10682 [04:59<1:23:04,  2.03it/s]
  5%|▌         | 580/10682 [04:59<1:23:09,  2.02it/s]
  5%|▌         | 581/10682 [05:00<1:23:07,  2.03it/s]
  5%|▌         | 582/10682 [05:00<1:23:07,  2.03it/s]
  5%|▌         | 583/10682 [05:01<1:23:06,  2.03it/s]
  5%|▌         | 584/10682 [05:01<1:22:58,  2.03it/s]
  5%|▌         | 585/10682 [05:02<1:23:00,  2.03it/s]
  5%|▌         | 586/10682 [05:02<1:23:00,  2.03it/s]
  5%|▌         | 587/10682 [05:03<1:22:58,  2.03it/s]
  6%|▌         | 588/10682 [05:03<1:23:01,  2.03it/s]
  6%|▌         | 589/10682 [05:04<1:23:03,  2.03it/s]
  6%|▌         | 590/10682 [05:04<1:23:02,  2.03it/s]
  6%|▌         | 591/10682 [05:05<1:23:04,  2.02it/s]
  6%|▌         | 592/10682 [05:05<1:23:02,  2.03it/s]
  6%|▌         | 593/10682 [05:06<1:23:01,  2.03it/s]
  6%|▌         | 594/10682 [05:06<1:23:01,  2.03it/s]
  6%|▌         | 595/10682 [05:07<1:23:03,  2.02it/s]
  6%|▌         | 596/10682 [05:07<1:23:01,  2.02it/s]
  6%|▌         | 597/10682 [05:08<1:22:56,  2.03it/s]
  6%|▌         | 598/10682 [05:08<1:22:55,  2.03it/s]
  6%|▌         | 599/10682 [05:09<1:22:51,  2.03it/s]
  6%|▌         | 600/10682 [05:09<1:22:55,  2.03it/s]
                                                     
{'loss': 4.8392, 'grad_norm': 0.5483365654945374, 'learning_rate': 0.0005612722170252572, 'epoch': 0.79}
+
  6%|▌         | 600/10682 [05:09<1:22:55,  2.03it/s]
  6%|▌         | 601/10682 [05:10<1:23:01,  2.02it/s]
  6%|▌         | 602/10682 [05:10<1:22:58,  2.02it/s]
  6%|▌         | 603/10682 [05:11<1:22:56,  2.03it/s]
  6%|▌         | 604/10682 [05:11<1:22:57,  2.02it/s]
  6%|▌         | 605/10682 [05:12<1:22:56,  2.02it/s]
  6%|▌         | 606/10682 [05:12<1:22:55,  2.03it/s]
  6%|▌         | 607/10682 [05:13<1:22:55,  2.03it/s]
  6%|▌         | 608/10682 [05:13<1:22:53,  2.03it/s]
  6%|▌         | 609/10682 [05:14<1:30:00,  1.87it/s]
  6%|▌         | 610/10682 [05:14<1:27:47,  1.91it/s]
  6%|▌         | 611/10682 [05:15<1:26:21,  1.94it/s]
  6%|▌         | 612/10682 [05:15<1:25:14,  1.97it/s]
  6%|▌         | 613/10682 [05:16<1:24:36,  1.98it/s]
  6%|▌         | 614/10682 [05:16<1:24:02,  2.00it/s]
  6%|▌         | 615/10682 [05:17<1:23:42,  2.00it/s]
  6%|▌         | 616/10682 [05:17<1:30:28,  1.85it/s]
  6%|▌         | 617/10682 [05:18<1:28:15,  1.90it/s]
  6%|▌         | 618/10682 [05:18<1:26:39,  1.94it/s]
  6%|▌         | 619/10682 [05:19<1:25:32,  1.96it/s]
  6%|▌         | 620/10682 [05:19<1:24:48,  1.98it/s]
  6%|▌         | 621/10682 [05:20<1:24:08,  1.99it/s]
  6%|▌         | 622/10682 [05:20<1:23:45,  2.00it/s]
  6%|▌         | 623/10682 [05:21<1:23:25,  2.01it/s]
  6%|▌         | 624/10682 [05:21<1:23:17,  2.01it/s]
  6%|▌         | 625/10682 [05:22<1:23:04,  2.02it/s]
                                                     {'loss': 4.7939, 'grad_norm': 0.7277975082397461, 'learning_rate': 0.0005846585594013096, 'epoch': 0.82}
+

  6%|▌         | 625/10682 [05:22<1:23:04,  2.02it/s]
  6%|▌         | 626/10682 [05:22<1:23:06,  2.02it/s]
  6%|▌         | 627/10682 [05:23<1:22:57,  2.02it/s]
  6%|▌         | 628/10682 [05:23<1:22:55,  2.02it/s]
  6%|▌         | 629/10682 [05:24<1:22:56,  2.02it/s]
  6%|▌         | 630/10682 [05:24<1:22:56,  2.02it/s]
  6%|▌         | 631/10682 [05:25<1:22:50,  2.02it/s]
  6%|▌         | 632/10682 [05:25<1:22:54,  2.02it/s]
  6%|▌         | 633/10682 [05:26<1:22:48,  2.02it/s]
  6%|▌         | 634/10682 [05:26<1:22:52,  2.02it/s]
  6%|▌         | 635/10682 [05:27<1:22:45,  2.02it/s]
  6%|▌         | 636/10682 [05:27<1:22:49,  2.02it/s]
  6%|▌         | 637/10682 [05:28<1:22:43,  2.02it/s]
  6%|▌         | 638/10682 [05:28<1:22:46,  2.02it/s]
  6%|▌         | 639/10682 [05:29<1:22:39,  2.03it/s]
  6%|▌         | 640/10682 [05:29<1:22:40,  2.02it/s]
  6%|▌         | 641/10682 [05:30<1:22:37,  2.03it/s]
  6%|▌         | 642/10682 [05:30<1:22:38,  2.02it/s]
  6%|▌         | 643/10682 [05:31<1:22:39,  2.02it/s]
  6%|▌         | 644/10682 [05:31<1:22:41,  2.02it/s]
  6%|▌         | 645/10682 [05:32<1:22:40,  2.02it/s]
  6%|▌         | 646/10682 [05:32<1:22:37,  2.02it/s]
  6%|▌         | 647/10682 [05:33<1:22:36,  2.02it/s]
  6%|▌         | 648/10682 [05:33<1:22:39,  2.02it/s]
  6%|▌         | 649/10682 [05:34<1:22:41,  2.02it/s]
  6%|▌         | 650/10682 [05:34<1:22:35,  2.02it/s]
                                                     {'loss': 4.7435, 'grad_norm': 0.5552908778190613, 'learning_rate': 0.0006080449017773621, 'epoch': 0.85}
+

  6%|▌         | 650/10682 [05:34<1:22:35,  2.02it/s]
  6%|▌         | 651/10682 [05:35<1:22:39,  2.02it/s]
  6%|▌         | 652/10682 [05:35<1:22:35,  2.02it/s]
  6%|▌         | 653/10682 [05:36<1:22:38,  2.02it/s]
  6%|▌         | 654/10682 [05:36<1:22:29,  2.03it/s]
  6%|▌         | 655/10682 [05:37<1:22:33,  2.02it/s]
  6%|▌         | 656/10682 [05:37<1:22:28,  2.03it/s]
  6%|▌         | 657/10682 [05:38<1:22:33,  2.02it/s]
  6%|▌         | 658/10682 [05:38<1:22:27,  2.03it/s]
  6%|▌         | 659/10682 [05:39<1:22:32,  2.02it/s]
  6%|▌         | 660/10682 [05:39<1:22:29,  2.02it/s]
  6%|▌         | 661/10682 [05:40<1:22:34,  2.02it/s]
  6%|▌         | 662/10682 [05:40<1:22:34,  2.02it/s]
  6%|▌         | 663/10682 [05:41<1:22:34,  2.02it/s]
  6%|▌         | 664/10682 [05:41<1:22:30,  2.02it/s]
  6%|▌         | 665/10682 [05:42<1:22:25,  2.03it/s]
  6%|▌         | 666/10682 [05:42<1:22:29,  2.02it/s]
  6%|▌         | 667/10682 [05:43<1:22:27,  2.02it/s]
  6%|▋         | 668/10682 [05:43<1:22:28,  2.02it/s]
  6%|▋         | 669/10682 [05:44<1:22:27,  2.02it/s]
  6%|▋         | 670/10682 [05:44<1:22:27,  2.02it/s]
  6%|▋         | 671/10682 [05:45<1:22:23,  2.03it/s]
  6%|▋         | 672/10682 [05:45<1:22:27,  2.02it/s]
  6%|▋         | 673/10682 [05:46<1:22:23,  2.02it/s]
  6%|▋         | 674/10682 [05:46<1:22:27,  2.02it/s]
  6%|▋         | 675/10682 [05:47<1:22:20,  2.03it/s]{'loss': 4.7025, 'grad_norm': 0.5710030198097229, 'learning_rate': 0.0006314312441534145, 'epoch': 0.88}
+
                                                     

  6%|▋         | 675/10682 [05:47<1:22:20,  2.03it/s]
  6%|▋         | 676/10682 [05:47<1:22:26,  2.02it/s]
  6%|▋         | 677/10682 [05:48<1:22:23,  2.02it/s]
  6%|▋         | 678/10682 [05:48<1:22:23,  2.02it/s]
  6%|▋         | 679/10682 [05:49<1:22:19,  2.02it/s]
  6%|▋         | 680/10682 [05:49<1:22:18,  2.03it/s]
  6%|▋         | 681/10682 [05:50<1:22:19,  2.02it/s]
  6%|▋         | 682/10682 [05:50<1:22:15,  2.03it/s]
  6%|▋         | 683/10682 [05:51<1:22:20,  2.02it/s]
  6%|▋         | 684/10682 [05:51<1:22:16,  2.03it/s]
  6%|▋         | 685/10682 [05:52<1:22:24,  2.02it/s]
  6%|▋         | 686/10682 [05:52<1:22:16,  2.02it/s]
  6%|▋         | 687/10682 [05:52<1:22:24,  2.02it/s]
  6%|▋         | 688/10682 [05:53<1:22:19,  2.02it/s]
  6%|▋         | 689/10682 [05:53<1:22:22,  2.02it/s]
  6%|▋         | 690/10682 [05:54<1:22:15,  2.02it/s]
  6%|▋         | 691/10682 [05:54<1:22:14,  2.02it/s]
  6%|▋         | 692/10682 [05:55<1:22:12,  2.03it/s]
  6%|▋         | 693/10682 [05:55<1:22:18,  2.02it/s]
  6%|▋         | 694/10682 [05:56<1:22:17,  2.02it/s]
  7%|▋         | 695/10682 [05:56<1:22:16,  2.02it/s]
  7%|▋         | 696/10682 [05:57<1:22:16,  2.02it/s]
  7%|▋         | 697/10682 [05:57<1:22:14,  2.02it/s]
  7%|▋         | 698/10682 [05:58<1:22:11,  2.02it/s]
  7%|▋         | 699/10682 [05:58<1:22:18,  2.02it/s]
  7%|▋         | 700/10682 [05:59<1:22:15,  2.02it/s]{'loss': 4.6602, 'grad_norm': 0.5751403570175171, 'learning_rate': 0.0006548175865294667, 'epoch': 0.92}
+
                                                     

  7%|▋         | 700/10682 [05:59<1:22:15,  2.02it/s]
  7%|▋         | 701/10682 [05:59<1:22:21,  2.02it/s]
  7%|▋         | 702/10682 [06:00<1:22:18,  2.02it/s]
  7%|▋         | 703/10682 [06:00<1:22:18,  2.02it/s]
  7%|▋         | 704/10682 [06:01<1:22:14,  2.02it/s]
  7%|▋         | 705/10682 [06:01<1:22:15,  2.02it/s]
  7%|▋         | 706/10682 [06:02<1:22:11,  2.02it/s]
  7%|▋         | 707/10682 [06:02<1:22:12,  2.02it/s]
  7%|▋         | 708/10682 [06:03<1:22:08,  2.02it/s]
  7%|▋         | 709/10682 [06:03<1:22:06,  2.02it/s]
  7%|▋         | 710/10682 [06:04<1:22:08,  2.02it/s]
  7%|▋         | 711/10682 [06:04<1:22:07,  2.02it/s]
  7%|▋         | 712/10682 [06:05<1:22:04,  2.02it/s]
  7%|▋         | 713/10682 [06:05<1:22:01,  2.03it/s]
  7%|▋         | 714/10682 [06:06<1:22:03,  2.02it/s]
  7%|▋         | 715/10682 [06:06<1:21:59,  2.03it/s]
  7%|▋         | 716/10682 [06:07<1:22:01,  2.03it/s]
  7%|▋         | 717/10682 [06:07<1:22:01,  2.02it/s]
  7%|▋         | 718/10682 [06:08<1:22:02,  2.02it/s]
  7%|▋         | 719/10682 [06:08<1:22:05,  2.02it/s]
  7%|▋         | 720/10682 [06:09<1:22:01,  2.02it/s]
  7%|▋         | 721/10682 [06:09<1:22:04,  2.02it/s]
  7%|▋         | 722/10682 [06:10<1:21:59,  2.02it/s]
  7%|▋         | 723/10682 [06:10<1:22:04,  2.02it/s]
  7%|▋         | 724/10682 [06:11<1:21:57,  2.02it/s]
  7%|▋         | 725/10682 [06:11<1:22:00,  2.02it/s]{'loss': 4.6299, 'grad_norm': 0.4613557755947113, 'learning_rate': 0.0006782039289055192, 'epoch': 0.95}
+
                                                     

  7%|▋         | 725/10682 [06:11<1:22:00,  2.02it/s]
  7%|▋         | 726/10682 [06:12<1:22:03,  2.02it/s]
  7%|▋         | 727/10682 [06:12<1:22:08,  2.02it/s]
  7%|▋         | 728/10682 [06:13<1:22:05,  2.02it/s]
  7%|▋         | 729/10682 [06:13<1:22:06,  2.02it/s]
  7%|▋         | 730/10682 [06:14<1:22:01,  2.02it/s]
  7%|▋         | 731/10682 [06:14<1:21:59,  2.02it/s]
  7%|▋         | 732/10682 [06:15<1:21:58,  2.02it/s]
  7%|▋         | 733/10682 [06:15<1:22:00,  2.02it/s]
  7%|▋         | 734/10682 [06:16<1:22:02,  2.02it/s]
  7%|▋         | 735/10682 [06:16<1:22:02,  2.02it/s]
  7%|▋         | 736/10682 [06:17<1:22:01,  2.02it/s]
  7%|▋         | 737/10682 [06:17<1:22:03,  2.02it/s]
  7%|▋         | 738/10682 [06:18<1:22:03,  2.02it/s]
  7%|▋         | 739/10682 [06:18<1:21:59,  2.02it/s]
  7%|▋         | 740/10682 [06:19<1:21:56,  2.02it/s]
  7%|▋         | 741/10682 [06:19<1:21:51,  2.02it/s]
  7%|▋         | 742/10682 [06:20<1:21:52,  2.02it/s]
  7%|▋         | 743/10682 [06:20<1:21:48,  2.02it/s]
  7%|▋         | 744/10682 [06:21<1:21:53,  2.02it/s]
  7%|▋         | 745/10682 [06:21<1:21:52,  2.02it/s]
  7%|▋         | 746/10682 [06:22<1:21:52,  2.02it/s]
  7%|▋         | 747/10682 [06:22<1:21:47,  2.02it/s]
  7%|▋         | 748/10682 [06:23<1:21:47,  2.02it/s]
  7%|▋         | 749/10682 [06:23<1:21:46,  2.02it/s]
  7%|▋         | 750/10682 [06:24<1:21:43,  2.03it/s]{'loss': 4.5874, 'grad_norm': 0.4585312008857727, 'learning_rate': 0.0007015902712815716, 'epoch': 0.98}
+
                                                     

  7%|▋         | 750/10682 [06:24<1:21:43,  2.03it/s]
  7%|▋         | 751/10682 [06:24<1:21:50,  2.02it/s]
  7%|▋         | 752/10682 [06:25<1:21:47,  2.02it/s]
  7%|▋         | 753/10682 [06:25<1:21:45,  2.02it/s]
  7%|▋         | 754/10682 [06:26<1:21:40,  2.03it/s]
  7%|▋         | 755/10682 [06:26<1:21:43,  2.02it/s]
  7%|▋         | 756/10682 [06:27<1:21:39,  2.03it/s]
  7%|▋         | 757/10682 [06:27<1:21:43,  2.02it/s]
  7%|▋         | 758/10682 [06:28<1:21:37,  2.03it/s]
  7%|▋         | 759/10682 [06:28<1:21:42,  2.02it/s]
  7%|▋         | 760/10682 [06:29<1:21:39,  2.03it/s]
  7%|▋         | 761/10682 [06:29<1:21:41,  2.02it/s]
  7%|▋         | 762/10682 [06:30<1:21:43,  2.02it/s]
  7%|▋         | 763/10682 [06:30<1:23:50,  1.97it/s]
  7%|▋         | 764/10682 [06:42<10:42:35,  3.89s/it]
  7%|▋         | 765/10682 [06:42<7:54:19,  2.87s/it] 
  7%|▋         | 766/10682 [06:43<5:56:45,  2.16s/it]
  7%|▋         | 767/10682 [06:43<4:34:16,  1.66s/it]
  7%|▋         | 768/10682 [06:44<3:36:26,  1.31s/it]
  7%|▋         | 769/10682 [06:44<2:55:58,  1.07s/it]
  7%|▋         | 770/10682 [06:45<2:27:48,  1.12it/s]
  7%|▋         | 771/10682 [06:45<2:08:09,  1.29it/s]
  7%|▋         | 772/10682 [06:46<1:54:22,  1.44it/s]
  7%|▋         | 773/10682 [06:46<1:44:36,  1.58it/s]
  7%|▋         | 774/10682 [06:47<1:37:44,  1.69it/s]
  7%|▋         | 775/10682 [06:47<1:32:51,  1.78it/s]
                                                     
{'loss': 4.5319, 'grad_norm': 0.4785366952419281, 'learning_rate': 0.0007249766136576241, 'epoch': 1.02}
+
  7%|▋         | 775/10682 [06:47<1:32:51,  1.78it/s]
  7%|▋         | 776/10682 [06:48<1:29:47,  1.84it/s]
  7%|▋         | 777/10682 [06:48<1:27:17,  1.89it/s]
  7%|▋         | 778/10682 [06:49<1:25:40,  1.93it/s]
  7%|▋         | 779/10682 [06:49<1:24:29,  1.95it/s]
  7%|▋         | 780/10682 [06:50<1:23:45,  1.97it/s]
  7%|▋         | 781/10682 [06:50<1:23:01,  1.99it/s]
  7%|▋         | 782/10682 [06:51<1:22:36,  2.00it/s]
  7%|▋         | 783/10682 [06:51<1:22:14,  2.01it/s]
  7%|▋         | 784/10682 [06:52<1:22:07,  2.01it/s]
  7%|▋         | 785/10682 [06:52<1:22:09,  2.01it/s]
  7%|▋         | 786/10682 [06:53<1:22:02,  2.01it/s]
  7%|▋         | 787/10682 [06:53<1:21:51,  2.01it/s]
  7%|▋         | 788/10682 [06:54<1:21:47,  2.02it/s]
\ No newline at end of file