joshuaclymer
commited on
Commit
•
4b242c9
1
Parent(s):
075a304
Upload folder using huggingface_hub
Browse files- README.md +21 -0
- adapter_config.json +21 -0
- adapter_model.bin +3 -0
- special_tokens_map.json +24 -0
- tokenizer.json +0 -0
- tokenizer.model +3 -0
- tokenizer_config.json +32 -0
- train_args.json +123 -0
- training_args.bin +3 -0
- training_logs.json +1523 -0
README.md
ADDED
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
library_name: peft
|
3 |
+
---
|
4 |
+
## Training procedure
|
5 |
+
|
6 |
+
|
7 |
+
The following `bitsandbytes` quantization config was used during training:
|
8 |
+
- quant_method: bitsandbytes
|
9 |
+
- load_in_8bit: False
|
10 |
+
- load_in_4bit: True
|
11 |
+
- llm_int8_threshold: 6.0
|
12 |
+
- llm_int8_skip_modules: None
|
13 |
+
- llm_int8_enable_fp32_cpu_offload: False
|
14 |
+
- llm_int8_has_fp16_weight: False
|
15 |
+
- bnb_4bit_quant_type: nf4
|
16 |
+
- bnb_4bit_use_double_quant: True
|
17 |
+
- bnb_4bit_compute_dtype: float16
|
18 |
+
### Framework versions
|
19 |
+
|
20 |
+
|
21 |
+
- PEFT 0.5.0
|
adapter_config.json
ADDED
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"auto_mapping": null,
|
3 |
+
"base_model_name_or_path": "models/llama-30b",
|
4 |
+
"bias": "none",
|
5 |
+
"fan_in_fan_out": false,
|
6 |
+
"inference_mode": true,
|
7 |
+
"init_lora_weights": true,
|
8 |
+
"layers_pattern": null,
|
9 |
+
"layers_to_transform": null,
|
10 |
+
"lora_alpha": 16,
|
11 |
+
"lora_dropout": 0.0,
|
12 |
+
"modules_to_save": null,
|
13 |
+
"peft_type": "LORA",
|
14 |
+
"r": 64,
|
15 |
+
"revision": null,
|
16 |
+
"target_modules": [
|
17 |
+
"q_proj",
|
18 |
+
"v_proj"
|
19 |
+
],
|
20 |
+
"task_type": "SEQ_CLS"
|
21 |
+
}
|
adapter_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ca6c5e88c130c681ebe1d03ac5d181a07873a7ec6daba8c35b38d7ca21207f95
|
3 |
+
size 409081617
|
special_tokens_map.json
ADDED
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": {
|
3 |
+
"content": "<s>",
|
4 |
+
"lstrip": false,
|
5 |
+
"normalized": true,
|
6 |
+
"rstrip": false,
|
7 |
+
"single_word": false
|
8 |
+
},
|
9 |
+
"eos_token": {
|
10 |
+
"content": "</s>",
|
11 |
+
"lstrip": false,
|
12 |
+
"normalized": true,
|
13 |
+
"rstrip": false,
|
14 |
+
"single_word": false
|
15 |
+
},
|
16 |
+
"pad_token": "</s>",
|
17 |
+
"unk_token": {
|
18 |
+
"content": "<unk>",
|
19 |
+
"lstrip": false,
|
20 |
+
"normalized": true,
|
21 |
+
"rstrip": false,
|
22 |
+
"single_word": false
|
23 |
+
}
|
24 |
+
}
|
tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer.model
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
3 |
+
size 499723
|
tokenizer_config.json
ADDED
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": {
|
3 |
+
"__type": "AddedToken",
|
4 |
+
"content": "<s>",
|
5 |
+
"lstrip": false,
|
6 |
+
"normalized": true,
|
7 |
+
"rstrip": false,
|
8 |
+
"single_word": false
|
9 |
+
},
|
10 |
+
"clean_up_tokenization_spaces": false,
|
11 |
+
"eos_token": {
|
12 |
+
"__type": "AddedToken",
|
13 |
+
"content": "</s>",
|
14 |
+
"lstrip": false,
|
15 |
+
"normalized": true,
|
16 |
+
"rstrip": false,
|
17 |
+
"single_word": false
|
18 |
+
},
|
19 |
+
"model_max_length": 2048,
|
20 |
+
"pad_token": null,
|
21 |
+
"sp_model_kwargs": {},
|
22 |
+
"tokenizer_class": "LlamaTokenizer",
|
23 |
+
"unk_token": {
|
24 |
+
"__type": "AddedToken",
|
25 |
+
"content": "<unk>",
|
26 |
+
"lstrip": false,
|
27 |
+
"normalized": true,
|
28 |
+
"rstrip": false,
|
29 |
+
"single_word": false
|
30 |
+
},
|
31 |
+
"use_default_system_prompt": true
|
32 |
+
}
|
train_args.json
ADDED
@@ -0,0 +1,123 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"initial_model_dir": "models/llama-30b",
|
3 |
+
"distribution_id": "commonsense_qa",
|
4 |
+
"date_trained": "11/10/2023 19:58:23",
|
5 |
+
"output_dir": "models/classify_lora/llama-30b-commonsense_qa",
|
6 |
+
"overwrite_output_dir": false,
|
7 |
+
"do_train": false,
|
8 |
+
"do_eval": true,
|
9 |
+
"do_predict": false,
|
10 |
+
"evaluation_strategy": "steps",
|
11 |
+
"prediction_loss_only": false,
|
12 |
+
"per_device_train_batch_size": 8,
|
13 |
+
"per_device_eval_batch_size": 8,
|
14 |
+
"per_gpu_train_batch_size": null,
|
15 |
+
"per_gpu_eval_batch_size": null,
|
16 |
+
"gradient_accumulation_steps": 1,
|
17 |
+
"eval_accumulation_steps": null,
|
18 |
+
"eval_delay": 0,
|
19 |
+
"learning_rate": 0.0002,
|
20 |
+
"weight_decay": 0.0,
|
21 |
+
"adam_beta1": 0.9,
|
22 |
+
"adam_beta2": 0.999,
|
23 |
+
"adam_epsilon": 1e-08,
|
24 |
+
"max_grad_norm": 0.3,
|
25 |
+
"num_train_epochs": 2,
|
26 |
+
"max_steps": 100,
|
27 |
+
"lr_scheduler_type": "constant",
|
28 |
+
"warmup_ratio": 0.03,
|
29 |
+
"warmup_steps": 0,
|
30 |
+
"log_level": "passive",
|
31 |
+
"log_level_replica": "warning",
|
32 |
+
"log_on_each_node": true,
|
33 |
+
"logging_dir": "models/classify_lora/llama-30b-commonsense_qa/runs/Oct11_19-49-04_compute-permanent-node-975",
|
34 |
+
"logging_strategy": "steps",
|
35 |
+
"logging_first_step": false,
|
36 |
+
"logging_steps": 1,
|
37 |
+
"logging_nan_inf_filter": true,
|
38 |
+
"save_strategy": "steps",
|
39 |
+
"save_steps": 25,
|
40 |
+
"save_total_limit": 0,
|
41 |
+
"save_safetensors": false,
|
42 |
+
"save_on_each_node": false,
|
43 |
+
"no_cuda": false,
|
44 |
+
"use_cpu": false,
|
45 |
+
"use_mps_device": false,
|
46 |
+
"seed": 42,
|
47 |
+
"data_seed": null,
|
48 |
+
"jit_mode_eval": false,
|
49 |
+
"use_ipex": false,
|
50 |
+
"bf16": false,
|
51 |
+
"fp16": false,
|
52 |
+
"fp16_opt_level": "O1",
|
53 |
+
"half_precision_backend": "auto",
|
54 |
+
"bf16_full_eval": false,
|
55 |
+
"fp16_full_eval": false,
|
56 |
+
"tf32": null,
|
57 |
+
"local_rank": 2,
|
58 |
+
"ddp_backend": null,
|
59 |
+
"tpu_num_cores": null,
|
60 |
+
"tpu_metrics_debug": false,
|
61 |
+
"debug": [],
|
62 |
+
"dataloader_drop_last": false,
|
63 |
+
"eval_steps": 25,
|
64 |
+
"dataloader_num_workers": 0,
|
65 |
+
"past_index": -1,
|
66 |
+
"run_name": "train|models-classify_lora-llama-30b-commonsense_qa",
|
67 |
+
"disable_tqdm": false,
|
68 |
+
"remove_unused_columns": false,
|
69 |
+
"label_names": null,
|
70 |
+
"load_best_model_at_end": false,
|
71 |
+
"metric_for_best_model": "eval_commonsense_qa_score",
|
72 |
+
"greater_is_better": true,
|
73 |
+
"ignore_data_skip": false,
|
74 |
+
"sharded_ddp": [],
|
75 |
+
"fsdp": [],
|
76 |
+
"fsdp_min_num_params": 0,
|
77 |
+
"fsdp_config": {
|
78 |
+
"min_num_params": 0,
|
79 |
+
"xla": false,
|
80 |
+
"xla_fsdp_grad_ckpt": false
|
81 |
+
},
|
82 |
+
"fsdp_transformer_layer_cls_to_wrap": null,
|
83 |
+
"deepspeed": "configs/ds_zero_1.json",
|
84 |
+
"label_smoothing_factor": 0.0,
|
85 |
+
"optim": "paged_adamw_32bit",
|
86 |
+
"optim_args": null,
|
87 |
+
"adafactor": false,
|
88 |
+
"group_by_length": false,
|
89 |
+
"length_column_name": "length",
|
90 |
+
"report_to": [
|
91 |
+
"wandb"
|
92 |
+
],
|
93 |
+
"ddp_find_unused_parameters": false,
|
94 |
+
"ddp_bucket_cap_mb": null,
|
95 |
+
"ddp_broadcast_buffers": null,
|
96 |
+
"dataloader_pin_memory": true,
|
97 |
+
"skip_memory_metrics": true,
|
98 |
+
"use_legacy_prediction_loop": false,
|
99 |
+
"push_to_hub": false,
|
100 |
+
"resume_from_checkpoint": null,
|
101 |
+
"hub_model_id": null,
|
102 |
+
"hub_strategy": "every_save",
|
103 |
+
"hub_token": null,
|
104 |
+
"hub_private_repo": false,
|
105 |
+
"hub_always_push": false,
|
106 |
+
"gradient_checkpointing": false,
|
107 |
+
"include_inputs_for_metrics": false,
|
108 |
+
"fp16_backend": "auto",
|
109 |
+
"push_to_hub_model_id": null,
|
110 |
+
"push_to_hub_organization": null,
|
111 |
+
"push_to_hub_token": null,
|
112 |
+
"_n_gpu": 1,
|
113 |
+
"mp_parameters": "",
|
114 |
+
"auto_find_batch_size": false,
|
115 |
+
"full_determinism": false,
|
116 |
+
"torchdynamo": null,
|
117 |
+
"ray_scope": "last",
|
118 |
+
"ddp_timeout": 1800,
|
119 |
+
"torch_compile": false,
|
120 |
+
"torch_compile_backend": null,
|
121 |
+
"torch_compile_mode": null,
|
122 |
+
"dispatch_batches": null
|
123 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f15fdcc37e2de3eea86fa6b5ff826878d92b127b8e6d461d83347684dce8c756
|
3 |
+
size 5115
|
training_logs.json
ADDED
@@ -0,0 +1,1523 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
[
|
2 |
+
{
|
3 |
+
"loss": 0.8525,
|
4 |
+
"learning_rate": 0.0002,
|
5 |
+
"epoch": 0.02,
|
6 |
+
"step": 1
|
7 |
+
},
|
8 |
+
{
|
9 |
+
"loss": 0.6634,
|
10 |
+
"learning_rate": 0.0002,
|
11 |
+
"epoch": 0.04,
|
12 |
+
"step": 2
|
13 |
+
},
|
14 |
+
{
|
15 |
+
"loss": 0.7302,
|
16 |
+
"learning_rate": 0.0002,
|
17 |
+
"epoch": 0.05,
|
18 |
+
"step": 3
|
19 |
+
},
|
20 |
+
{
|
21 |
+
"loss": 0.7507,
|
22 |
+
"learning_rate": 0.0002,
|
23 |
+
"epoch": 0.07,
|
24 |
+
"step": 4
|
25 |
+
},
|
26 |
+
{
|
27 |
+
"loss": 0.7715,
|
28 |
+
"learning_rate": 0.0002,
|
29 |
+
"epoch": 0.09,
|
30 |
+
"step": 5
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"loss": 0.7064,
|
34 |
+
"learning_rate": 0.0002,
|
35 |
+
"epoch": 0.11,
|
36 |
+
"step": 6
|
37 |
+
},
|
38 |
+
{
|
39 |
+
"loss": 0.7124,
|
40 |
+
"learning_rate": 0.0002,
|
41 |
+
"epoch": 0.12,
|
42 |
+
"step": 7
|
43 |
+
},
|
44 |
+
{
|
45 |
+
"loss": 0.7223,
|
46 |
+
"learning_rate": 0.0002,
|
47 |
+
"epoch": 0.14,
|
48 |
+
"step": 8
|
49 |
+
},
|
50 |
+
{
|
51 |
+
"loss": 0.6888,
|
52 |
+
"learning_rate": 0.0002,
|
53 |
+
"epoch": 0.16,
|
54 |
+
"step": 9
|
55 |
+
},
|
56 |
+
{
|
57 |
+
"loss": 0.7404,
|
58 |
+
"learning_rate": 0.0002,
|
59 |
+
"epoch": 0.18,
|
60 |
+
"step": 10
|
61 |
+
},
|
62 |
+
{
|
63 |
+
"loss": 0.6575,
|
64 |
+
"learning_rate": 0.0002,
|
65 |
+
"epoch": 0.19,
|
66 |
+
"step": 11
|
67 |
+
},
|
68 |
+
{
|
69 |
+
"loss": 0.6896,
|
70 |
+
"learning_rate": 0.0002,
|
71 |
+
"epoch": 0.21,
|
72 |
+
"step": 12
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"loss": 0.6336,
|
76 |
+
"learning_rate": 0.0002,
|
77 |
+
"epoch": 0.23,
|
78 |
+
"step": 13
|
79 |
+
},
|
80 |
+
{
|
81 |
+
"loss": 0.6474,
|
82 |
+
"learning_rate": 0.0002,
|
83 |
+
"epoch": 0.25,
|
84 |
+
"step": 14
|
85 |
+
},
|
86 |
+
{
|
87 |
+
"loss": 0.6993,
|
88 |
+
"learning_rate": 0.0002,
|
89 |
+
"epoch": 0.26,
|
90 |
+
"step": 15
|
91 |
+
},
|
92 |
+
{
|
93 |
+
"loss": 0.6219,
|
94 |
+
"learning_rate": 0.0002,
|
95 |
+
"epoch": 0.28,
|
96 |
+
"step": 16
|
97 |
+
},
|
98 |
+
{
|
99 |
+
"loss": 0.7434,
|
100 |
+
"learning_rate": 0.0002,
|
101 |
+
"epoch": 0.3,
|
102 |
+
"step": 17
|
103 |
+
},
|
104 |
+
{
|
105 |
+
"loss": 0.7488,
|
106 |
+
"learning_rate": 0.0002,
|
107 |
+
"epoch": 0.32,
|
108 |
+
"step": 18
|
109 |
+
},
|
110 |
+
{
|
111 |
+
"loss": 0.7165,
|
112 |
+
"learning_rate": 0.0002,
|
113 |
+
"epoch": 0.33,
|
114 |
+
"step": 19
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"loss": 0.7253,
|
118 |
+
"learning_rate": 0.0002,
|
119 |
+
"epoch": 0.35,
|
120 |
+
"step": 20
|
121 |
+
},
|
122 |
+
{
|
123 |
+
"loss": 0.6785,
|
124 |
+
"learning_rate": 0.0002,
|
125 |
+
"epoch": 0.37,
|
126 |
+
"step": 21
|
127 |
+
},
|
128 |
+
{
|
129 |
+
"loss": 0.7281,
|
130 |
+
"learning_rate": 0.0002,
|
131 |
+
"epoch": 0.39,
|
132 |
+
"step": 22
|
133 |
+
},
|
134 |
+
{
|
135 |
+
"loss": 0.7451,
|
136 |
+
"learning_rate": 0.0002,
|
137 |
+
"epoch": 0.4,
|
138 |
+
"step": 23
|
139 |
+
},
|
140 |
+
{
|
141 |
+
"loss": 0.6618,
|
142 |
+
"learning_rate": 0.0002,
|
143 |
+
"epoch": 0.42,
|
144 |
+
"step": 24
|
145 |
+
},
|
146 |
+
{
|
147 |
+
"loss": 0.6452,
|
148 |
+
"learning_rate": 0.0002,
|
149 |
+
"epoch": 0.44,
|
150 |
+
"step": 25
|
151 |
+
},
|
152 |
+
{
|
153 |
+
"eval_commonsense_qa_loss": 0.7914362549781799,
|
154 |
+
"eval_commonsense_qa_score": -0.29111722111701965,
|
155 |
+
"eval_commonsense_qa_brier_score": 0.29111722111701965,
|
156 |
+
"eval_commonsense_qa_average_probability": 0.47185489535331726,
|
157 |
+
"eval_commonsense_qa_accuracy": 0.39,
|
158 |
+
"eval_commonsense_qa_probabilities": [
|
159 |
+
0.3736628592014313,
|
160 |
+
0.41951867938041687,
|
161 |
+
0.40974757075309753,
|
162 |
+
0.43108245730400085,
|
163 |
+
0.39417126774787903,
|
164 |
+
0.36827391386032104,
|
165 |
+
0.3217296004295349,
|
166 |
+
0.4527994692325592,
|
167 |
+
0.44031405448913574,
|
168 |
+
0.5761461853981018,
|
169 |
+
0.6323458552360535,
|
170 |
+
0.5307877659797668,
|
171 |
+
0.3072277307510376,
|
172 |
+
0.35649847984313965,
|
173 |
+
0.24212150275707245,
|
174 |
+
0.4968399107456207,
|
175 |
+
0.7714088559150696,
|
176 |
+
0.4102341830730438,
|
177 |
+
0.5640316605567932,
|
178 |
+
0.5279660820960999,
|
179 |
+
0.26524466276168823,
|
180 |
+
0.46642741560935974,
|
181 |
+
0.667880654335022,
|
182 |
+
0.5137766003608704,
|
183 |
+
0.7010161280632019,
|
184 |
+
0.5712583065032959,
|
185 |
+
0.5321078300476074,
|
186 |
+
0.4304255247116089,
|
187 |
+
0.25722262263298035,
|
188 |
+
0.49681854248046875,
|
189 |
+
0.39940145611763,
|
190 |
+
0.38804712891578674,
|
191 |
+
0.413534939289093,
|
192 |
+
0.36740759015083313,
|
193 |
+
0.3980674743652344,
|
194 |
+
0.3686825931072235,
|
195 |
+
0.45741069316864014,
|
196 |
+
0.4793001115322113,
|
197 |
+
0.5210141539573669,
|
198 |
+
0.5084275603294373,
|
199 |
+
0.5377230644226074,
|
200 |
+
0.4399715065956116,
|
201 |
+
0.5001020431518555,
|
202 |
+
0.4058927297592163,
|
203 |
+
0.4787653684616089,
|
204 |
+
0.35228681564331055,
|
205 |
+
0.3852904140949249,
|
206 |
+
0.38124364614486694,
|
207 |
+
0.49779626727104187,
|
208 |
+
0.47046318650245667,
|
209 |
+
0.45446330308914185,
|
210 |
+
0.49293678998947144,
|
211 |
+
0.5469354391098022,
|
212 |
+
0.4111570715904236,
|
213 |
+
0.6435064077377319,
|
214 |
+
0.3415040075778961,
|
215 |
+
0.557502269744873,
|
216 |
+
0.6114356517791748,
|
217 |
+
0.4695909917354584,
|
218 |
+
0.3526020348072052,
|
219 |
+
0.6928711533546448,
|
220 |
+
0.688827633857727,
|
221 |
+
0.6876643896102905,
|
222 |
+
0.4979974031448364,
|
223 |
+
0.5267980098724365,
|
224 |
+
0.5480849146842957,
|
225 |
+
0.5159767866134644,
|
226 |
+
0.6435903310775757,
|
227 |
+
0.48293471336364746,
|
228 |
+
0.4974452555179596,
|
229 |
+
0.39486047625541687,
|
230 |
+
0.3298123776912689,
|
231 |
+
0.39702242612838745,
|
232 |
+
0.2585730254650116,
|
233 |
+
0.5650836229324341,
|
234 |
+
0.6514415144920349,
|
235 |
+
0.567798376083374,
|
236 |
+
0.5929319858551025,
|
237 |
+
0.35160502791404724,
|
238 |
+
0.39691317081451416,
|
239 |
+
0.589564859867096,
|
240 |
+
0.4809286296367645,
|
241 |
+
0.3265831768512726,
|
242 |
+
0.34647417068481445,
|
243 |
+
0.29951584339141846,
|
244 |
+
0.540679931640625,
|
245 |
+
0.477059006690979,
|
246 |
+
0.42304784059524536,
|
247 |
+
0.3258381187915802,
|
248 |
+
0.33986833691596985,
|
249 |
+
0.6030049920082092,
|
250 |
+
0.5700676441192627,
|
251 |
+
0.5453739166259766,
|
252 |
+
0.5489950180053711,
|
253 |
+
0.5006292462348938,
|
254 |
+
0.5012415647506714,
|
255 |
+
0.4897501468658447,
|
256 |
+
0.405569463968277,
|
257 |
+
0.41794145107269287,
|
258 |
+
0.6035709977149963
|
259 |
+
],
|
260 |
+
"eval_commonsense_qa_runtime": 5.072,
|
261 |
+
"eval_commonsense_qa_samples_per_second": 19.716,
|
262 |
+
"eval_commonsense_qa_steps_per_second": 0.789,
|
263 |
+
"epoch": 0.44,
|
264 |
+
"step": 25
|
265 |
+
},
|
266 |
+
{
|
267 |
+
"eval_trivia_qa_loss": 0.7058285474777222,
|
268 |
+
"eval_trivia_qa_score": -0.2509709298610687,
|
269 |
+
"eval_trivia_qa_brier_score": 0.2509709298610687,
|
270 |
+
"eval_trivia_qa_average_probability": 0.508730411529541,
|
271 |
+
"eval_trivia_qa_accuracy": 0.56,
|
272 |
+
"eval_trivia_qa_probabilities": [
|
273 |
+
0.6108027696609497,
|
274 |
+
0.47548747062683105,
|
275 |
+
0.5868535041809082,
|
276 |
+
0.32189613580703735,
|
277 |
+
0.4967602789402008,
|
278 |
+
0.5013786554336548,
|
279 |
+
0.582642138004303,
|
280 |
+
0.5476276278495789,
|
281 |
+
0.6525075435638428,
|
282 |
+
0.3721744418144226,
|
283 |
+
0.46289870142936707,
|
284 |
+
0.5139896273612976,
|
285 |
+
0.6440446376800537,
|
286 |
+
0.5662510395050049,
|
287 |
+
0.5801640748977661,
|
288 |
+
0.27492260932922363,
|
289 |
+
0.5860297679901123,
|
290 |
+
0.4890085756778717,
|
291 |
+
0.6448793411254883,
|
292 |
+
0.5299046039581299,
|
293 |
+
0.3838443160057068,
|
294 |
+
0.2839445173740387,
|
295 |
+
0.5129914879798889,
|
296 |
+
0.6394293308258057,
|
297 |
+
0.5504795908927917,
|
298 |
+
0.6178440451622009,
|
299 |
+
0.35971587896347046,
|
300 |
+
0.45417338609695435,
|
301 |
+
0.42903023958206177,
|
302 |
+
0.559572160243988,
|
303 |
+
0.5979811549186707,
|
304 |
+
0.4994199275970459,
|
305 |
+
0.5579876899719238,
|
306 |
+
0.7093907594680786,
|
307 |
+
0.4760594666004181,
|
308 |
+
0.4218994379043579,
|
309 |
+
0.5057868361473083,
|
310 |
+
0.6060559153556824,
|
311 |
+
0.5037559866905212,
|
312 |
+
0.5046707391738892,
|
313 |
+
0.42048025131225586,
|
314 |
+
0.6409589648246765,
|
315 |
+
0.5305330753326416,
|
316 |
+
0.3996846079826355,
|
317 |
+
0.4467884302139282,
|
318 |
+
0.6522667407989502,
|
319 |
+
0.4101843535900116,
|
320 |
+
0.3745155334472656,
|
321 |
+
0.5663012266159058,
|
322 |
+
0.48798033595085144,
|
323 |
+
0.6159585118293762,
|
324 |
+
0.47460001707077026,
|
325 |
+
0.6789429187774658,
|
326 |
+
0.3534046411514282,
|
327 |
+
0.5315006375312805,
|
328 |
+
0.4756757915019989,
|
329 |
+
0.6186851263046265,
|
330 |
+
0.5154322981834412,
|
331 |
+
0.37286680936813354,
|
332 |
+
0.5532978177070618,
|
333 |
+
0.45223331451416016,
|
334 |
+
0.41775307059288025,
|
335 |
+
0.5951501727104187,
|
336 |
+
0.640890896320343,
|
337 |
+
0.4332258701324463,
|
338 |
+
0.6166451573371887,
|
339 |
+
0.626221239566803,
|
340 |
+
0.545464277267456,
|
341 |
+
0.36058875918388367,
|
342 |
+
0.4321788251399994,
|
343 |
+
0.4153941571712494,
|
344 |
+
0.5391730070114136,
|
345 |
+
0.396107017993927,
|
346 |
+
0.5979107618331909,
|
347 |
+
0.5979316830635071,
|
348 |
+
0.40809130668640137,
|
349 |
+
0.6226964592933655,
|
350 |
+
0.38245290517807007,
|
351 |
+
0.587360680103302,
|
352 |
+
0.44519785046577454,
|
353 |
+
0.659287691116333,
|
354 |
+
0.570533812046051,
|
355 |
+
0.3212454915046692,
|
356 |
+
0.5290343761444092,
|
357 |
+
0.3304256200790405,
|
358 |
+
0.5141531229019165,
|
359 |
+
0.4628876745700836,
|
360 |
+
0.4035484790802002,
|
361 |
+
0.5565099716186523,
|
362 |
+
0.5873599052429199,
|
363 |
+
0.5347287058830261,
|
364 |
+
0.4785170257091522,
|
365 |
+
0.44041702151298523,
|
366 |
+
0.4931771457195282,
|
367 |
+
0.5670301914215088,
|
368 |
+
0.5181891322135925,
|
369 |
+
0.6944608688354492,
|
370 |
+
0.4809003174304962,
|
371 |
+
0.5023355484008789,
|
372 |
+
0.3813191056251526
|
373 |
+
],
|
374 |
+
"eval_trivia_qa_runtime": 6.9532,
|
375 |
+
"eval_trivia_qa_samples_per_second": 14.382,
|
376 |
+
"eval_trivia_qa_steps_per_second": 0.575,
|
377 |
+
"epoch": 0.44,
|
378 |
+
"step": 25
|
379 |
+
},
|
380 |
+
{
|
381 |
+
"loss": 0.6417,
|
382 |
+
"learning_rate": 0.0002,
|
383 |
+
"epoch": 0.46,
|
384 |
+
"step": 26
|
385 |
+
},
|
386 |
+
{
|
387 |
+
"loss": 0.696,
|
388 |
+
"learning_rate": 0.0002,
|
389 |
+
"epoch": 0.47,
|
390 |
+
"step": 27
|
391 |
+
},
|
392 |
+
{
|
393 |
+
"loss": 0.688,
|
394 |
+
"learning_rate": 0.0002,
|
395 |
+
"epoch": 0.49,
|
396 |
+
"step": 28
|
397 |
+
},
|
398 |
+
{
|
399 |
+
"loss": 0.6599,
|
400 |
+
"learning_rate": 0.0002,
|
401 |
+
"epoch": 0.51,
|
402 |
+
"step": 29
|
403 |
+
},
|
404 |
+
{
|
405 |
+
"loss": 0.726,
|
406 |
+
"learning_rate": 0.0002,
|
407 |
+
"epoch": 0.53,
|
408 |
+
"step": 30
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"loss": 0.6367,
|
412 |
+
"learning_rate": 0.0002,
|
413 |
+
"epoch": 0.54,
|
414 |
+
"step": 31
|
415 |
+
},
|
416 |
+
{
|
417 |
+
"loss": 0.7377,
|
418 |
+
"learning_rate": 0.0002,
|
419 |
+
"epoch": 0.56,
|
420 |
+
"step": 32
|
421 |
+
},
|
422 |
+
{
|
423 |
+
"loss": 0.5971,
|
424 |
+
"learning_rate": 0.0002,
|
425 |
+
"epoch": 0.58,
|
426 |
+
"step": 33
|
427 |
+
},
|
428 |
+
{
|
429 |
+
"loss": 0.7561,
|
430 |
+
"learning_rate": 0.0002,
|
431 |
+
"epoch": 0.6,
|
432 |
+
"step": 34
|
433 |
+
},
|
434 |
+
{
|
435 |
+
"loss": 0.6453,
|
436 |
+
"learning_rate": 0.0002,
|
437 |
+
"epoch": 0.61,
|
438 |
+
"step": 35
|
439 |
+
},
|
440 |
+
{
|
441 |
+
"loss": 0.6648,
|
442 |
+
"learning_rate": 0.0002,
|
443 |
+
"epoch": 0.63,
|
444 |
+
"step": 36
|
445 |
+
},
|
446 |
+
{
|
447 |
+
"loss": 0.6106,
|
448 |
+
"learning_rate": 0.0002,
|
449 |
+
"epoch": 0.65,
|
450 |
+
"step": 37
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"loss": 0.6421,
|
454 |
+
"learning_rate": 0.0002,
|
455 |
+
"epoch": 0.67,
|
456 |
+
"step": 38
|
457 |
+
},
|
458 |
+
{
|
459 |
+
"loss": 0.6576,
|
460 |
+
"learning_rate": 0.0002,
|
461 |
+
"epoch": 0.68,
|
462 |
+
"step": 39
|
463 |
+
},
|
464 |
+
{
|
465 |
+
"loss": 0.6426,
|
466 |
+
"learning_rate": 0.0002,
|
467 |
+
"epoch": 0.7,
|
468 |
+
"step": 40
|
469 |
+
},
|
470 |
+
{
|
471 |
+
"loss": 0.6443,
|
472 |
+
"learning_rate": 0.0002,
|
473 |
+
"epoch": 0.72,
|
474 |
+
"step": 41
|
475 |
+
},
|
476 |
+
{
|
477 |
+
"loss": 0.6842,
|
478 |
+
"learning_rate": 0.0002,
|
479 |
+
"epoch": 0.74,
|
480 |
+
"step": 42
|
481 |
+
},
|
482 |
+
{
|
483 |
+
"loss": 0.6407,
|
484 |
+
"learning_rate": 0.0002,
|
485 |
+
"epoch": 0.75,
|
486 |
+
"step": 43
|
487 |
+
},
|
488 |
+
{
|
489 |
+
"loss": 0.7051,
|
490 |
+
"learning_rate": 0.0002,
|
491 |
+
"epoch": 0.77,
|
492 |
+
"step": 44
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"loss": 0.6729,
|
496 |
+
"learning_rate": 0.0002,
|
497 |
+
"epoch": 0.79,
|
498 |
+
"step": 45
|
499 |
+
},
|
500 |
+
{
|
501 |
+
"loss": 0.6876,
|
502 |
+
"learning_rate": 0.0002,
|
503 |
+
"epoch": 0.81,
|
504 |
+
"step": 46
|
505 |
+
},
|
506 |
+
{
|
507 |
+
"loss": 0.6711,
|
508 |
+
"learning_rate": 0.0002,
|
509 |
+
"epoch": 0.82,
|
510 |
+
"step": 47
|
511 |
+
},
|
512 |
+
{
|
513 |
+
"loss": 0.6613,
|
514 |
+
"learning_rate": 0.0002,
|
515 |
+
"epoch": 0.84,
|
516 |
+
"step": 48
|
517 |
+
},
|
518 |
+
{
|
519 |
+
"loss": 0.6044,
|
520 |
+
"learning_rate": 0.0002,
|
521 |
+
"epoch": 0.86,
|
522 |
+
"step": 49
|
523 |
+
},
|
524 |
+
{
|
525 |
+
"loss": 0.5769,
|
526 |
+
"learning_rate": 0.0002,
|
527 |
+
"epoch": 0.88,
|
528 |
+
"step": 50
|
529 |
+
},
|
530 |
+
{
|
531 |
+
"eval_commonsense_qa_loss": 0.7653241157531738,
|
532 |
+
"eval_commonsense_qa_score": -0.2791268527507782,
|
533 |
+
"eval_commonsense_qa_brier_score": 0.2791268527507782,
|
534 |
+
"eval_commonsense_qa_average_probability": 0.48683643341064453,
|
535 |
+
"eval_commonsense_qa_accuracy": 0.42,
|
536 |
+
"eval_commonsense_qa_probabilities": [
|
537 |
+
0.4020081162452698,
|
538 |
+
0.48435282707214355,
|
539 |
+
0.43516698479652405,
|
540 |
+
0.4614918529987335,
|
541 |
+
0.4766950011253357,
|
542 |
+
0.4323585629463196,
|
543 |
+
0.30255478620529175,
|
544 |
+
0.29709967970848083,
|
545 |
+
0.545942485332489,
|
546 |
+
0.5383279323577881,
|
547 |
+
0.6857153177261353,
|
548 |
+
0.48378315567970276,
|
549 |
+
0.16650518774986267,
|
550 |
+
0.2533150315284729,
|
551 |
+
0.19820035994052887,
|
552 |
+
0.44554558396339417,
|
553 |
+
0.7646204233169556,
|
554 |
+
0.6391469240188599,
|
555 |
+
0.5909577012062073,
|
556 |
+
0.5142195224761963,
|
557 |
+
0.46391570568084717,
|
558 |
+
0.4665536880493164,
|
559 |
+
0.5840488076210022,
|
560 |
+
0.43470215797424316,
|
561 |
+
0.7169751524925232,
|
562 |
+
0.7758315205574036,
|
563 |
+
0.5787749886512756,
|
564 |
+
0.5941563844680786,
|
565 |
+
0.46830788254737854,
|
566 |
+
0.5470129251480103,
|
567 |
+
0.46815431118011475,
|
568 |
+
0.526231050491333,
|
569 |
+
0.5080574750900269,
|
570 |
+
0.29635000228881836,
|
571 |
+
0.47670942544937134,
|
572 |
+
0.32638877630233765,
|
573 |
+
0.4723879396915436,
|
574 |
+
0.6098498702049255,
|
575 |
+
0.4592856168746948,
|
576 |
+
0.5818286538124084,
|
577 |
+
0.5597203373908997,
|
578 |
+
0.3799402415752411,
|
579 |
+
0.3624171018600464,
|
580 |
+
0.40649276971817017,
|
581 |
+
0.42219263315200806,
|
582 |
+
0.3799329996109009,
|
583 |
+
0.3508457839488983,
|
584 |
+
0.46101444959640503,
|
585 |
+
0.4670948088169098,
|
586 |
+
0.6448565721511841,
|
587 |
+
0.4881596565246582,
|
588 |
+
0.44159600138664246,
|
589 |
+
0.6058750748634338,
|
590 |
+
0.590934157371521,
|
591 |
+
0.46451979875564575,
|
592 |
+
0.37275078892707825,
|
593 |
+
0.5106647610664368,
|
594 |
+
0.6377049684524536,
|
595 |
+
0.5288455486297607,
|
596 |
+
0.4052494764328003,
|
597 |
+
0.6817584037780762,
|
598 |
+
0.682870090007782,
|
599 |
+
0.7456340193748474,
|
600 |
+
0.4100780189037323,
|
601 |
+
0.5996410846710205,
|
602 |
+
0.46194082498550415,
|
603 |
+
0.4774899482727051,
|
604 |
+
0.6287940740585327,
|
605 |
+
0.442749559879303,
|
606 |
+
0.38106194138526917,
|
607 |
+
0.3868009150028229,
|
608 |
+
0.34707480669021606,
|
609 |
+
0.5333090424537659,
|
610 |
+
0.4054989218711853,
|
611 |
+
0.6355715990066528,
|
612 |
+
0.5217822790145874,
|
613 |
+
0.3845539689064026,
|
614 |
+
0.45426151156425476,
|
615 |
+
0.43503573536872864,
|
616 |
+
0.4364105463027954,
|
617 |
+
0.6591715812683105,
|
618 |
+
0.5792595744132996,
|
619 |
+
0.22389499843120575,
|
620 |
+
0.4098465144634247,
|
621 |
+
0.3524221181869507,
|
622 |
+
0.42123866081237793,
|
623 |
+
0.48706841468811035,
|
624 |
+
0.41214796900749207,
|
625 |
+
0.22283338010311127,
|
626 |
+
0.28352028131484985,
|
627 |
+
0.688373863697052,
|
628 |
+
0.5961286425590515,
|
629 |
+
0.6101290583610535,
|
630 |
+
0.5565056204795837,
|
631 |
+
0.5037069320678711,
|
632 |
+
0.41246306896209717,
|
633 |
+
0.6470851302146912,
|
634 |
+
0.5888364315032959,
|
635 |
+
0.4137541353702545,
|
636 |
+
0.6086077690124512
|
637 |
+
],
|
638 |
+
"eval_commonsense_qa_runtime": 5.0497,
|
639 |
+
"eval_commonsense_qa_samples_per_second": 19.803,
|
640 |
+
"eval_commonsense_qa_steps_per_second": 0.792,
|
641 |
+
"epoch": 0.88,
|
642 |
+
"step": 50
|
643 |
+
},
|
644 |
+
{
|
645 |
+
"eval_trivia_qa_loss": 0.6967568397521973,
|
646 |
+
"eval_trivia_qa_score": -0.24871157109737396,
|
647 |
+
"eval_trivia_qa_brier_score": 0.24871157109737396,
|
648 |
+
"eval_trivia_qa_average_probability": 0.5072451829910278,
|
649 |
+
"eval_trivia_qa_accuracy": 0.5,
|
650 |
+
"eval_trivia_qa_probabilities": [
|
651 |
+
0.5316422581672668,
|
652 |
+
0.432894766330719,
|
653 |
+
0.5510445833206177,
|
654 |
+
0.4671865999698639,
|
655 |
+
0.5174410343170166,
|
656 |
+
0.5549404621124268,
|
657 |
+
0.4334378242492676,
|
658 |
+
0.39600223302841187,
|
659 |
+
0.528852641582489,
|
660 |
+
0.42653071880340576,
|
661 |
+
0.5234686136245728,
|
662 |
+
0.42779213190078735,
|
663 |
+
0.5071923732757568,
|
664 |
+
0.4844723641872406,
|
665 |
+
0.5264027714729309,
|
666 |
+
0.439650297164917,
|
667 |
+
0.553399384021759,
|
668 |
+
0.5268545746803284,
|
669 |
+
0.42666733264923096,
|
670 |
+
0.47586703300476074,
|
671 |
+
0.4965582489967346,
|
672 |
+
0.34971049427986145,
|
673 |
+
0.6940780878067017,
|
674 |
+
0.5421137809753418,
|
675 |
+
0.48094817996025085,
|
676 |
+
0.43245548009872437,
|
677 |
+
0.5597365498542786,
|
678 |
+
0.4554421305656433,
|
679 |
+
0.4893034100532532,
|
680 |
+
0.5171124935150146,
|
681 |
+
0.5146493911743164,
|
682 |
+
0.495728462934494,
|
683 |
+
0.4713417887687683,
|
684 |
+
0.6222578287124634,
|
685 |
+
0.498948872089386,
|
686 |
+
0.40657901763916016,
|
687 |
+
0.527847409248352,
|
688 |
+
0.7099897265434265,
|
689 |
+
0.5247665047645569,
|
690 |
+
0.4212891161441803,
|
691 |
+
0.5828162431716919,
|
692 |
+
0.540741503238678,
|
693 |
+
0.5784919857978821,
|
694 |
+
0.5161374807357788,
|
695 |
+
0.4517030119895935,
|
696 |
+
0.6503217220306396,
|
697 |
+
0.5777159929275513,
|
698 |
+
0.36965277791023254,
|
699 |
+
0.4844502806663513,
|
700 |
+
0.5828574299812317,
|
701 |
+
0.6145827174186707,
|
702 |
+
0.3802032172679901,
|
703 |
+
0.7286924123764038,
|
704 |
+
0.4977825880050659,
|
705 |
+
0.5343342423439026,
|
706 |
+
0.5175969004631042,
|
707 |
+
0.48933130502700806,
|
708 |
+
0.5237868428230286,
|
709 |
+
0.39064183831214905,
|
710 |
+
0.5096392035484314,
|
711 |
+
0.5159491896629333,
|
712 |
+
0.3979681730270386,
|
713 |
+
0.6354355812072754,
|
714 |
+
0.5407276153564453,
|
715 |
+
0.4297018051147461,
|
716 |
+
0.5711097121238708,
|
717 |
+
0.6063168048858643,
|
718 |
+
0.4898560345172882,
|
719 |
+
0.43474793434143066,
|
720 |
+
0.5227797627449036,
|
721 |
+
0.4176207184791565,
|
722 |
+
0.6359185576438904,
|
723 |
+
0.4207601547241211,
|
724 |
+
0.4994364082813263,
|
725 |
+
0.5145835876464844,
|
726 |
+
0.37604933977127075,
|
727 |
+
0.5790051817893982,
|
728 |
+
0.45702680945396423,
|
729 |
+
0.48608124256134033,
|
730 |
+
0.47564682364463806,
|
731 |
+
0.6444876194000244,
|
732 |
+
0.4628547132015228,
|
733 |
+
0.4236738979816437,
|
734 |
+
0.5284913778305054,
|
735 |
+
0.3784801661968231,
|
736 |
+
0.6344289779663086,
|
737 |
+
0.523414671421051,
|
738 |
+
0.4968854784965515,
|
739 |
+
0.4976916015148163,
|
740 |
+
0.4860800504684448,
|
741 |
+
0.47529691457748413,
|
742 |
+
0.4533490836620331,
|
743 |
+
0.4189370274543762,
|
744 |
+
0.6194345355033875,
|
745 |
+
0.484923779964447,
|
746 |
+
0.5832501649856567,
|
747 |
+
0.6392970681190491,
|
748 |
+
0.509278416633606,
|
749 |
+
0.5340873003005981,
|
750 |
+
0.4593735337257385
|
751 |
+
],
|
752 |
+
"eval_trivia_qa_runtime": 6.9799,
|
753 |
+
"eval_trivia_qa_samples_per_second": 14.327,
|
754 |
+
"eval_trivia_qa_steps_per_second": 0.573,
|
755 |
+
"epoch": 0.88,
|
756 |
+
"step": 50
|
757 |
+
},
|
758 |
+
{
|
759 |
+
"loss": 0.6601,
|
760 |
+
"learning_rate": 0.0002,
|
761 |
+
"epoch": 0.89,
|
762 |
+
"step": 51
|
763 |
+
},
|
764 |
+
{
|
765 |
+
"loss": 0.6454,
|
766 |
+
"learning_rate": 0.0002,
|
767 |
+
"epoch": 0.91,
|
768 |
+
"step": 52
|
769 |
+
},
|
770 |
+
{
|
771 |
+
"loss": 0.6547,
|
772 |
+
"learning_rate": 0.0002,
|
773 |
+
"epoch": 0.93,
|
774 |
+
"step": 53
|
775 |
+
},
|
776 |
+
{
|
777 |
+
"loss": 0.7068,
|
778 |
+
"learning_rate": 0.0002,
|
779 |
+
"epoch": 0.95,
|
780 |
+
"step": 54
|
781 |
+
},
|
782 |
+
{
|
783 |
+
"loss": 0.6629,
|
784 |
+
"learning_rate": 0.0002,
|
785 |
+
"epoch": 0.96,
|
786 |
+
"step": 55
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"loss": 0.7136,
|
790 |
+
"learning_rate": 0.0002,
|
791 |
+
"epoch": 0.98,
|
792 |
+
"step": 56
|
793 |
+
},
|
794 |
+
{
|
795 |
+
"loss": 0.6437,
|
796 |
+
"learning_rate": 0.0002,
|
797 |
+
"epoch": 1.0,
|
798 |
+
"step": 57
|
799 |
+
},
|
800 |
+
{
|
801 |
+
"loss": 0.5198,
|
802 |
+
"learning_rate": 0.0002,
|
803 |
+
"epoch": 1.02,
|
804 |
+
"step": 58
|
805 |
+
},
|
806 |
+
{
|
807 |
+
"loss": 0.4605,
|
808 |
+
"learning_rate": 0.0002,
|
809 |
+
"epoch": 1.04,
|
810 |
+
"step": 59
|
811 |
+
},
|
812 |
+
{
|
813 |
+
"loss": 0.4861,
|
814 |
+
"learning_rate": 0.0002,
|
815 |
+
"epoch": 1.05,
|
816 |
+
"step": 60
|
817 |
+
},
|
818 |
+
{
|
819 |
+
"loss": 0.483,
|
820 |
+
"learning_rate": 0.0002,
|
821 |
+
"epoch": 1.07,
|
822 |
+
"step": 61
|
823 |
+
},
|
824 |
+
{
|
825 |
+
"loss": 0.5581,
|
826 |
+
"learning_rate": 0.0002,
|
827 |
+
"epoch": 1.09,
|
828 |
+
"step": 62
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"loss": 0.4564,
|
832 |
+
"learning_rate": 0.0002,
|
833 |
+
"epoch": 1.11,
|
834 |
+
"step": 63
|
835 |
+
},
|
836 |
+
{
|
837 |
+
"loss": 0.4056,
|
838 |
+
"learning_rate": 0.0002,
|
839 |
+
"epoch": 1.12,
|
840 |
+
"step": 64
|
841 |
+
},
|
842 |
+
{
|
843 |
+
"loss": 0.4167,
|
844 |
+
"learning_rate": 0.0002,
|
845 |
+
"epoch": 1.14,
|
846 |
+
"step": 65
|
847 |
+
},
|
848 |
+
{
|
849 |
+
"loss": 0.7253,
|
850 |
+
"learning_rate": 0.0002,
|
851 |
+
"epoch": 1.16,
|
852 |
+
"step": 66
|
853 |
+
},
|
854 |
+
{
|
855 |
+
"loss": 0.4609,
|
856 |
+
"learning_rate": 0.0002,
|
857 |
+
"epoch": 1.18,
|
858 |
+
"step": 67
|
859 |
+
},
|
860 |
+
{
|
861 |
+
"loss": 0.5302,
|
862 |
+
"learning_rate": 0.0002,
|
863 |
+
"epoch": 1.19,
|
864 |
+
"step": 68
|
865 |
+
},
|
866 |
+
{
|
867 |
+
"loss": 0.6734,
|
868 |
+
"learning_rate": 0.0002,
|
869 |
+
"epoch": 1.21,
|
870 |
+
"step": 69
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"loss": 0.5224,
|
874 |
+
"learning_rate": 0.0002,
|
875 |
+
"epoch": 1.23,
|
876 |
+
"step": 70
|
877 |
+
},
|
878 |
+
{
|
879 |
+
"loss": 0.4326,
|
880 |
+
"learning_rate": 0.0002,
|
881 |
+
"epoch": 1.25,
|
882 |
+
"step": 71
|
883 |
+
},
|
884 |
+
{
|
885 |
+
"loss": 0.5751,
|
886 |
+
"learning_rate": 0.0002,
|
887 |
+
"epoch": 1.26,
|
888 |
+
"step": 72
|
889 |
+
},
|
890 |
+
{
|
891 |
+
"loss": 0.5653,
|
892 |
+
"learning_rate": 0.0002,
|
893 |
+
"epoch": 1.28,
|
894 |
+
"step": 73
|
895 |
+
},
|
896 |
+
{
|
897 |
+
"loss": 0.4586,
|
898 |
+
"learning_rate": 0.0002,
|
899 |
+
"epoch": 1.3,
|
900 |
+
"step": 74
|
901 |
+
},
|
902 |
+
{
|
903 |
+
"loss": 0.4466,
|
904 |
+
"learning_rate": 0.0002,
|
905 |
+
"epoch": 1.32,
|
906 |
+
"step": 75
|
907 |
+
},
|
908 |
+
{
|
909 |
+
"eval_commonsense_qa_loss": 1.0953351259231567,
|
910 |
+
"eval_commonsense_qa_score": -0.3630787134170532,
|
911 |
+
"eval_commonsense_qa_brier_score": 0.3630787134170532,
|
912 |
+
"eval_commonsense_qa_average_probability": 0.46727100014686584,
|
913 |
+
"eval_commonsense_qa_accuracy": 0.47,
|
914 |
+
"eval_commonsense_qa_probabilities": [
|
915 |
+
0.6413354277610779,
|
916 |
+
0.6861792206764221,
|
917 |
+
0.9139598608016968,
|
918 |
+
0.3129615783691406,
|
919 |
+
0.4858136773109436,
|
920 |
+
0.3958660662174225,
|
921 |
+
0.015349932946264744,
|
922 |
+
0.03121619112789631,
|
923 |
+
0.3732204735279083,
|
924 |
+
0.5287988781929016,
|
925 |
+
0.9410275220870972,
|
926 |
+
0.4859994351863861,
|
927 |
+
0.00484914006665349,
|
928 |
+
0.06006177142262459,
|
929 |
+
0.016101302579045296,
|
930 |
+
0.25591790676116943,
|
931 |
+
0.9382115006446838,
|
932 |
+
0.7453911304473877,
|
933 |
+
0.49844464659690857,
|
934 |
+
0.11961045861244202,
|
935 |
+
0.21799755096435547,
|
936 |
+
0.22540347278118134,
|
937 |
+
0.6641191840171814,
|
938 |
+
0.3415297865867615,
|
939 |
+
0.9549143314361572,
|
940 |
+
0.9577696323394775,
|
941 |
+
0.3804933726787567,
|
942 |
+
0.5659024715423584,
|
943 |
+
0.6399015188217163,
|
944 |
+
0.6054954528808594,
|
945 |
+
0.16932412981987,
|
946 |
+
0.24352750182151794,
|
947 |
+
0.5880881547927856,
|
948 |
+
0.10066776722669601,
|
949 |
+
0.2192084938287735,
|
950 |
+
0.08628320693969727,
|
951 |
+
0.33604246377944946,
|
952 |
+
0.27943921089172363,
|
953 |
+
0.3816390931606293,
|
954 |
+
0.7643809914588928,
|
955 |
+
0.7210124731063843,
|
956 |
+
0.5266180038452148,
|
957 |
+
0.204672709107399,
|
958 |
+
0.5603741407394409,
|
959 |
+
0.6822576522827148,
|
960 |
+
0.19128814339637756,
|
961 |
+
0.20498374104499817,
|
962 |
+
0.09368855506181717,
|
963 |
+
0.6106529235839844,
|
964 |
+
0.8263741135597229,
|
965 |
+
0.6096042394638062,
|
966 |
+
0.3612224757671356,
|
967 |
+
0.3293110430240631,
|
968 |
+
0.4423503577709198,
|
969 |
+
0.1176837608218193,
|
970 |
+
0.1601405143737793,
|
971 |
+
0.3888700008392334,
|
972 |
+
0.8851404786109924,
|
973 |
+
0.578056275844574,
|
974 |
+
0.2645527720451355,
|
975 |
+
0.9431173205375671,
|
976 |
+
0.8939391374588013,
|
977 |
+
0.9614054560661316,
|
978 |
+
0.15922978520393372,
|
979 |
+
0.5092063546180725,
|
980 |
+
0.7479695081710815,
|
981 |
+
0.2555106580257416,
|
982 |
+
0.9339343905448914,
|
983 |
+
0.5020483136177063,
|
984 |
+
0.2890869677066803,
|
985 |
+
0.25056713819503784,
|
986 |
+
0.07140284031629562,
|
987 |
+
0.6142622828483582,
|
988 |
+
0.6608532667160034,
|
989 |
+
0.8618139028549194,
|
990 |
+
0.8793162107467651,
|
991 |
+
0.5890896916389465,
|
992 |
+
0.8838966488838196,
|
993 |
+
0.4716782569885254,
|
994 |
+
0.398629367351532,
|
995 |
+
0.8436529040336609,
|
996 |
+
0.3973071277141571,
|
997 |
+
0.06621242314577103,
|
998 |
+
0.5546140670776367,
|
999 |
+
0.16331911087036133,
|
1000 |
+
0.606509804725647,
|
1001 |
+
0.6033067107200623,
|
1002 |
+
0.2906716763973236,
|
1003 |
+
0.009353392757475376,
|
1004 |
+
0.061065930873155594,
|
1005 |
+
0.8538415431976318,
|
1006 |
+
0.5957357883453369,
|
1007 |
+
0.7623599767684937,
|
1008 |
+
0.6353998780250549,
|
1009 |
+
0.6718450784683228,
|
1010 |
+
0.06562834978103638,
|
1011 |
+
0.4774560332298279,
|
1012 |
+
0.20875515043735504,
|
1013 |
+
0.2653869092464447,
|
1014 |
+
0.7864195108413696
|
1015 |
+
],
|
1016 |
+
"eval_commonsense_qa_runtime": 5.0892,
|
1017 |
+
"eval_commonsense_qa_samples_per_second": 19.649,
|
1018 |
+
"eval_commonsense_qa_steps_per_second": 0.786,
|
1019 |
+
"epoch": 1.32,
|
1020 |
+
"step": 75
|
1021 |
+
},
|
1022 |
+
{
|
1023 |
+
"eval_trivia_qa_loss": 0.7178796529769897,
|
1024 |
+
"eval_trivia_qa_score": -0.25323230028152466,
|
1025 |
+
"eval_trivia_qa_brier_score": 0.25323230028152466,
|
1026 |
+
"eval_trivia_qa_average_probability": 0.5258346796035767,
|
1027 |
+
"eval_trivia_qa_accuracy": 0.56,
|
1028 |
+
"eval_trivia_qa_probabilities": [
|
1029 |
+
0.6001328825950623,
|
1030 |
+
0.43153026700019836,
|
1031 |
+
0.7303735613822937,
|
1032 |
+
0.4660419523715973,
|
1033 |
+
0.6144695281982422,
|
1034 |
+
0.44857344031333923,
|
1035 |
+
0.13023461401462555,
|
1036 |
+
0.21497581899166107,
|
1037 |
+
0.7541935443878174,
|
1038 |
+
0.3930618166923523,
|
1039 |
+
0.5923050045967102,
|
1040 |
+
0.24356049299240112,
|
1041 |
+
0.5985594391822815,
|
1042 |
+
0.44693607091903687,
|
1043 |
+
0.5665600299835205,
|
1044 |
+
0.37776947021484375,
|
1045 |
+
0.6056777834892273,
|
1046 |
+
0.6600034832954407,
|
1047 |
+
0.506766676902771,
|
1048 |
+
0.3960486650466919,
|
1049 |
+
0.6760654449462891,
|
1050 |
+
0.325588583946228,
|
1051 |
+
0.84026700258255,
|
1052 |
+
0.6310024857521057,
|
1053 |
+
0.41549575328826904,
|
1054 |
+
0.48468682169914246,
|
1055 |
+
0.30185914039611816,
|
1056 |
+
0.4738370478153229,
|
1057 |
+
0.40581652522087097,
|
1058 |
+
0.538021445274353,
|
1059 |
+
0.40905657410621643,
|
1060 |
+
0.5703331232070923,
|
1061 |
+
0.5543002486228943,
|
1062 |
+
0.6837395429611206,
|
1063 |
+
0.7139797806739807,
|
1064 |
+
0.47817718982696533,
|
1065 |
+
0.5026704668998718,
|
1066 |
+
0.6653541922569275,
|
1067 |
+
0.5143362283706665,
|
1068 |
+
0.33803310990333557,
|
1069 |
+
0.5520848035812378,
|
1070 |
+
0.7011743783950806,
|
1071 |
+
0.8836812973022461,
|
1072 |
+
0.6092294454574585,
|
1073 |
+
0.2784689962863922,
|
1074 |
+
0.7969092726707458,
|
1075 |
+
0.6553284525871277,
|
1076 |
+
0.5456470251083374,
|
1077 |
+
0.44695183634757996,
|
1078 |
+
0.548007071018219,
|
1079 |
+
0.5680496692657471,
|
1080 |
+
0.41621971130371094,
|
1081 |
+
0.830963134765625,
|
1082 |
+
0.7895110845565796,
|
1083 |
+
0.4964308738708496,
|
1084 |
+
0.3645930290222168,
|
1085 |
+
0.5637221932411194,
|
1086 |
+
0.44506263732910156,
|
1087 |
+
0.2878129184246063,
|
1088 |
+
0.5505443811416626,
|
1089 |
+
0.494486927986145,
|
1090 |
+
0.36506953835487366,
|
1091 |
+
0.8300395011901855,
|
1092 |
+
0.7011266946792603,
|
1093 |
+
0.348209947347641,
|
1094 |
+
0.626815140247345,
|
1095 |
+
0.7101113796234131,
|
1096 |
+
0.6857610940933228,
|
1097 |
+
0.4028257131576538,
|
1098 |
+
0.6243658065795898,
|
1099 |
+
0.15364059805870056,
|
1100 |
+
0.7464989423751831,
|
1101 |
+
0.5398872494697571,
|
1102 |
+
0.4613312780857086,
|
1103 |
+
0.40898871421813965,
|
1104 |
+
0.2252740114927292,
|
1105 |
+
0.7652521133422852,
|
1106 |
+
0.2925339937210083,
|
1107 |
+
0.4388081431388855,
|
1108 |
+
0.5058109164237976,
|
1109 |
+
0.8158372044563293,
|
1110 |
+
0.7546953558921814,
|
1111 |
+
0.2097131758928299,
|
1112 |
+
0.5340847373008728,
|
1113 |
+
0.25506791472435,
|
1114 |
+
0.537187397480011,
|
1115 |
+
0.592171311378479,
|
1116 |
+
0.5829126238822937,
|
1117 |
+
0.7708747982978821,
|
1118 |
+
0.43876922130584717,
|
1119 |
+
0.45161038637161255,
|
1120 |
+
0.4637928605079651,
|
1121 |
+
0.32175856828689575,
|
1122 |
+
0.6439905762672424,
|
1123 |
+
0.35225629806518555,
|
1124 |
+
0.5903196930885315,
|
1125 |
+
0.8497107625007629,
|
1126 |
+
0.530751645565033,
|
1127 |
+
0.5443508625030518,
|
1128 |
+
0.38598567247390747
|
1129 |
+
],
|
1130 |
+
"eval_trivia_qa_runtime": 6.9586,
|
1131 |
+
"eval_trivia_qa_samples_per_second": 14.371,
|
1132 |
+
"eval_trivia_qa_steps_per_second": 0.575,
|
1133 |
+
"epoch": 1.32,
|
1134 |
+
"step": 75
|
1135 |
+
},
|
1136 |
+
{
|
1137 |
+
"loss": 0.6169,
|
1138 |
+
"learning_rate": 0.0002,
|
1139 |
+
"epoch": 1.33,
|
1140 |
+
"step": 76
|
1141 |
+
},
|
1142 |
+
{
|
1143 |
+
"loss": 0.4915,
|
1144 |
+
"learning_rate": 0.0002,
|
1145 |
+
"epoch": 1.35,
|
1146 |
+
"step": 77
|
1147 |
+
},
|
1148 |
+
{
|
1149 |
+
"loss": 0.3843,
|
1150 |
+
"learning_rate": 0.0002,
|
1151 |
+
"epoch": 1.37,
|
1152 |
+
"step": 78
|
1153 |
+
},
|
1154 |
+
{
|
1155 |
+
"loss": 0.3157,
|
1156 |
+
"learning_rate": 0.0002,
|
1157 |
+
"epoch": 1.39,
|
1158 |
+
"step": 79
|
1159 |
+
},
|
1160 |
+
{
|
1161 |
+
"loss": 0.4288,
|
1162 |
+
"learning_rate": 0.0002,
|
1163 |
+
"epoch": 1.4,
|
1164 |
+
"step": 80
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"loss": 0.4981,
|
1168 |
+
"learning_rate": 0.0002,
|
1169 |
+
"epoch": 1.42,
|
1170 |
+
"step": 81
|
1171 |
+
},
|
1172 |
+
{
|
1173 |
+
"loss": 0.3743,
|
1174 |
+
"learning_rate": 0.0002,
|
1175 |
+
"epoch": 1.44,
|
1176 |
+
"step": 82
|
1177 |
+
},
|
1178 |
+
{
|
1179 |
+
"loss": 0.4731,
|
1180 |
+
"learning_rate": 0.0002,
|
1181 |
+
"epoch": 1.46,
|
1182 |
+
"step": 83
|
1183 |
+
},
|
1184 |
+
{
|
1185 |
+
"loss": 0.5496,
|
1186 |
+
"learning_rate": 0.0002,
|
1187 |
+
"epoch": 1.47,
|
1188 |
+
"step": 84
|
1189 |
+
},
|
1190 |
+
{
|
1191 |
+
"loss": 0.5248,
|
1192 |
+
"learning_rate": 0.0002,
|
1193 |
+
"epoch": 1.49,
|
1194 |
+
"step": 85
|
1195 |
+
},
|
1196 |
+
{
|
1197 |
+
"loss": 0.3161,
|
1198 |
+
"learning_rate": 0.0002,
|
1199 |
+
"epoch": 1.51,
|
1200 |
+
"step": 86
|
1201 |
+
},
|
1202 |
+
{
|
1203 |
+
"loss": 0.4111,
|
1204 |
+
"learning_rate": 0.0002,
|
1205 |
+
"epoch": 1.53,
|
1206 |
+
"step": 87
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"loss": 0.6771,
|
1210 |
+
"learning_rate": 0.0002,
|
1211 |
+
"epoch": 1.54,
|
1212 |
+
"step": 88
|
1213 |
+
},
|
1214 |
+
{
|
1215 |
+
"loss": 0.3828,
|
1216 |
+
"learning_rate": 0.0002,
|
1217 |
+
"epoch": 1.56,
|
1218 |
+
"step": 89
|
1219 |
+
},
|
1220 |
+
{
|
1221 |
+
"loss": 0.5683,
|
1222 |
+
"learning_rate": 0.0002,
|
1223 |
+
"epoch": 1.58,
|
1224 |
+
"step": 90
|
1225 |
+
},
|
1226 |
+
{
|
1227 |
+
"loss": 0.3922,
|
1228 |
+
"learning_rate": 0.0002,
|
1229 |
+
"epoch": 1.6,
|
1230 |
+
"step": 91
|
1231 |
+
},
|
1232 |
+
{
|
1233 |
+
"loss": 0.3031,
|
1234 |
+
"learning_rate": 0.0002,
|
1235 |
+
"epoch": 1.61,
|
1236 |
+
"step": 92
|
1237 |
+
},
|
1238 |
+
{
|
1239 |
+
"loss": 0.4393,
|
1240 |
+
"learning_rate": 0.0002,
|
1241 |
+
"epoch": 1.63,
|
1242 |
+
"step": 93
|
1243 |
+
},
|
1244 |
+
{
|
1245 |
+
"loss": 0.5812,
|
1246 |
+
"learning_rate": 0.0002,
|
1247 |
+
"epoch": 1.65,
|
1248 |
+
"step": 94
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"loss": 0.3824,
|
1252 |
+
"learning_rate": 0.0002,
|
1253 |
+
"epoch": 1.67,
|
1254 |
+
"step": 95
|
1255 |
+
},
|
1256 |
+
{
|
1257 |
+
"loss": 0.438,
|
1258 |
+
"learning_rate": 0.0002,
|
1259 |
+
"epoch": 1.68,
|
1260 |
+
"step": 96
|
1261 |
+
},
|
1262 |
+
{
|
1263 |
+
"loss": 0.4176,
|
1264 |
+
"learning_rate": 0.0002,
|
1265 |
+
"epoch": 1.7,
|
1266 |
+
"step": 97
|
1267 |
+
},
|
1268 |
+
{
|
1269 |
+
"loss": 0.6267,
|
1270 |
+
"learning_rate": 0.0002,
|
1271 |
+
"epoch": 1.72,
|
1272 |
+
"step": 98
|
1273 |
+
},
|
1274 |
+
{
|
1275 |
+
"loss": 0.3414,
|
1276 |
+
"learning_rate": 0.0002,
|
1277 |
+
"epoch": 1.74,
|
1278 |
+
"step": 99
|
1279 |
+
},
|
1280 |
+
{
|
1281 |
+
"loss": 0.347,
|
1282 |
+
"learning_rate": 0.0002,
|
1283 |
+
"epoch": 1.75,
|
1284 |
+
"step": 100
|
1285 |
+
},
|
1286 |
+
{
|
1287 |
+
"eval_commonsense_qa_loss": 1.1523878574371338,
|
1288 |
+
"eval_commonsense_qa_score": -0.396799772977829,
|
1289 |
+
"eval_commonsense_qa_brier_score": 0.396799772977829,
|
1290 |
+
"eval_commonsense_qa_average_probability": 0.4321339428424835,
|
1291 |
+
"eval_commonsense_qa_accuracy": 0.37,
|
1292 |
+
"eval_commonsense_qa_probabilities": [
|
1293 |
+
0.5133974552154541,
|
1294 |
+
0.19955140352249146,
|
1295 |
+
0.12726318836212158,
|
1296 |
+
0.2381817102432251,
|
1297 |
+
0.3061620891094208,
|
1298 |
+
0.4511687159538269,
|
1299 |
+
0.031751181930303574,
|
1300 |
+
0.04943360388278961,
|
1301 |
+
0.0996626764535904,
|
1302 |
+
0.7441174387931824,
|
1303 |
+
0.9300025105476379,
|
1304 |
+
0.7207792401313782,
|
1305 |
+
0.00975093338638544,
|
1306 |
+
0.05811166390776634,
|
1307 |
+
0.01687975972890854,
|
1308 |
+
0.6953206658363342,
|
1309 |
+
0.8796400427818298,
|
1310 |
+
0.8870823979377747,
|
1311 |
+
0.7761392593383789,
|
1312 |
+
0.2514343857765198,
|
1313 |
+
0.7701127529144287,
|
1314 |
+
0.8666774034500122,
|
1315 |
+
0.9593339562416077,
|
1316 |
+
0.39567831158638,
|
1317 |
+
0.9251853227615356,
|
1318 |
+
0.956170380115509,
|
1319 |
+
0.5061752796173096,
|
1320 |
+
0.3293822705745697,
|
1321 |
+
0.29882335662841797,
|
1322 |
+
0.24122630059719086,
|
1323 |
+
0.23124100267887115,
|
1324 |
+
0.27979356050491333,
|
1325 |
+
0.7157328724861145,
|
1326 |
+
0.08215536922216415,
|
1327 |
+
0.899575412273407,
|
1328 |
+
0.2522304952144623,
|
1329 |
+
0.3705737888813019,
|
1330 |
+
0.41751977801322937,
|
1331 |
+
0.28646501898765564,
|
1332 |
+
0.7097967267036438,
|
1333 |
+
0.5926461219787598,
|
1334 |
+
0.5077208280563354,
|
1335 |
+
0.1959858238697052,
|
1336 |
+
0.39983221888542175,
|
1337 |
+
0.27750301361083984,
|
1338 |
+
0.40524882078170776,
|
1339 |
+
0.383944571018219,
|
1340 |
+
0.677962064743042,
|
1341 |
+
0.36190852522850037,
|
1342 |
+
0.9803575873374939,
|
1343 |
+
0.7406795620918274,
|
1344 |
+
0.16878190636634827,
|
1345 |
+
0.12593944370746613,
|
1346 |
+
0.17838409543037415,
|
1347 |
+
0.3623591661453247,
|
1348 |
+
0.13244767487049103,
|
1349 |
+
0.3018617331981659,
|
1350 |
+
0.8912862539291382,
|
1351 |
+
0.5757622122764587,
|
1352 |
+
0.3221542537212372,
|
1353 |
+
0.871790885925293,
|
1354 |
+
0.8111267685890198,
|
1355 |
+
0.7601278424263,
|
1356 |
+
0.2897421419620514,
|
1357 |
+
0.7055788040161133,
|
1358 |
+
0.5388709306716919,
|
1359 |
+
0.40989163517951965,
|
1360 |
+
0.4253596067428589,
|
1361 |
+
0.0980086550116539,
|
1362 |
+
0.38305872678756714,
|
1363 |
+
0.3864794969558716,
|
1364 |
+
0.19545888900756836,
|
1365 |
+
0.6149375438690186,
|
1366 |
+
0.4320893883705139,
|
1367 |
+
0.8303354382514954,
|
1368 |
+
0.24265244603157043,
|
1369 |
+
0.13059014081954956,
|
1370 |
+
0.2409209907054901,
|
1371 |
+
0.16863232851028442,
|
1372 |
+
0.040597084909677505,
|
1373 |
+
0.37653404474258423,
|
1374 |
+
0.1172540932893753,
|
1375 |
+
0.21813638508319855,
|
1376 |
+
0.34870657324790955,
|
1377 |
+
0.1515800803899765,
|
1378 |
+
0.12445370852947235,
|
1379 |
+
0.37387892603874207,
|
1380 |
+
0.2328016608953476,
|
1381 |
+
0.5161333680152893,
|
1382 |
+
0.2136731743812561,
|
1383 |
+
0.8099603056907654,
|
1384 |
+
0.08686374127864838,
|
1385 |
+
0.3295922577381134,
|
1386 |
+
0.4922294318675995,
|
1387 |
+
0.5317037105560303,
|
1388 |
+
0.21833769977092743,
|
1389 |
+
0.40150973200798035,
|
1390 |
+
0.6546261310577393,
|
1391 |
+
0.5003038048744202,
|
1392 |
+
0.8764225840568542
|
1393 |
+
],
|
1394 |
+
"eval_commonsense_qa_runtime": 5.0489,
|
1395 |
+
"eval_commonsense_qa_samples_per_second": 19.806,
|
1396 |
+
"eval_commonsense_qa_steps_per_second": 0.792,
|
1397 |
+
"epoch": 1.75,
|
1398 |
+
"step": 100
|
1399 |
+
},
|
1400 |
+
{
|
1401 |
+
"eval_trivia_qa_loss": 0.6934054493904114,
|
1402 |
+
"eval_trivia_qa_score": -0.24079853296279907,
|
1403 |
+
"eval_trivia_qa_brier_score": 0.24079853296279907,
|
1404 |
+
"eval_trivia_qa_average_probability": 0.5418477654457092,
|
1405 |
+
"eval_trivia_qa_accuracy": 0.65,
|
1406 |
+
"eval_trivia_qa_probabilities": [
|
1407 |
+
0.6125680208206177,
|
1408 |
+
0.5168354511260986,
|
1409 |
+
0.8149361610412598,
|
1410 |
+
0.6358698010444641,
|
1411 |
+
0.6971873641014099,
|
1412 |
+
0.43288537859916687,
|
1413 |
+
0.07869447767734528,
|
1414 |
+
0.3557625114917755,
|
1415 |
+
0.5231017470359802,
|
1416 |
+
0.6738269925117493,
|
1417 |
+
0.6492495536804199,
|
1418 |
+
0.2834418714046478,
|
1419 |
+
0.32658106088638306,
|
1420 |
+
0.5844005942344666,
|
1421 |
+
0.6492470502853394,
|
1422 |
+
0.7314375042915344,
|
1423 |
+
0.6042511463165283,
|
1424 |
+
0.6276711821556091,
|
1425 |
+
0.576631486415863,
|
1426 |
+
0.386127233505249,
|
1427 |
+
0.6349095702171326,
|
1428 |
+
0.5288182497024536,
|
1429 |
+
0.729836642742157,
|
1430 |
+
0.9058018326759338,
|
1431 |
+
0.5136997103691101,
|
1432 |
+
0.5128685832023621,
|
1433 |
+
0.3874431252479553,
|
1434 |
+
0.5315768122673035,
|
1435 |
+
0.5235913991928101,
|
1436 |
+
0.3415122330188751,
|
1437 |
+
0.6356922388076782,
|
1438 |
+
0.6705965995788574,
|
1439 |
+
0.6255914568901062,
|
1440 |
+
0.5862753987312317,
|
1441 |
+
0.7883551716804504,
|
1442 |
+
0.44843554496765137,
|
1443 |
+
0.7369633913040161,
|
1444 |
+
0.661395788192749,
|
1445 |
+
0.4538457691669464,
|
1446 |
+
0.30895760655403137,
|
1447 |
+
0.7524657249450684,
|
1448 |
+
0.7141547799110413,
|
1449 |
+
0.7967507839202881,
|
1450 |
+
0.5967062711715698,
|
1451 |
+
0.43285977840423584,
|
1452 |
+
0.7750150561332703,
|
1453 |
+
0.5969040393829346,
|
1454 |
+
0.4807274043560028,
|
1455 |
+
0.5844646096229553,
|
1456 |
+
0.33085906505584717,
|
1457 |
+
0.3553105592727661,
|
1458 |
+
0.43021368980407715,
|
1459 |
+
0.6605834364891052,
|
1460 |
+
0.7358336448669434,
|
1461 |
+
0.5644637942314148,
|
1462 |
+
0.49804893136024475,
|
1463 |
+
0.2971603572368622,
|
1464 |
+
0.5024909377098083,
|
1465 |
+
0.3082197606563568,
|
1466 |
+
0.707648515701294,
|
1467 |
+
0.4596785306930542,
|
1468 |
+
0.28572654724121094,
|
1469 |
+
0.7529811859130859,
|
1470 |
+
0.7141095995903015,
|
1471 |
+
0.38188186287879944,
|
1472 |
+
0.5586990714073181,
|
1473 |
+
0.7269975543022156,
|
1474 |
+
0.5075417757034302,
|
1475 |
+
0.48148077726364136,
|
1476 |
+
0.6394890546798706,
|
1477 |
+
0.09881781041622162,
|
1478 |
+
0.7859747409820557,
|
1479 |
+
0.5184196829795837,
|
1480 |
+
0.19931602478027344,
|
1481 |
+
0.6356649398803711,
|
1482 |
+
0.21822792291641235,
|
1483 |
+
0.6228802800178528,
|
1484 |
+
0.13568221032619476,
|
1485 |
+
0.3445551097393036,
|
1486 |
+
0.5373251438140869,
|
1487 |
+
0.830248236656189,
|
1488 |
+
0.7611830830574036,
|
1489 |
+
0.4023577868938446,
|
1490 |
+
0.537299394607544,
|
1491 |
+
0.2600027620792389,
|
1492 |
+
0.6060169339179993,
|
1493 |
+
0.6491380333900452,
|
1494 |
+
0.4489743113517761,
|
1495 |
+
0.7347836494445801,
|
1496 |
+
0.3445013165473938,
|
1497 |
+
0.5936704874038696,
|
1498 |
+
0.6110560297966003,
|
1499 |
+
0.36522528529167175,
|
1500 |
+
0.6722674369812012,
|
1501 |
+
0.40544307231903076,
|
1502 |
+
0.5686450004577637,
|
1503 |
+
0.9408103823661804,
|
1504 |
+
0.5205425024032593,
|
1505 |
+
0.5648188591003418,
|
1506 |
+
0.35458648204803467
|
1507 |
+
],
|
1508 |
+
"eval_trivia_qa_runtime": 6.9847,
|
1509 |
+
"eval_trivia_qa_samples_per_second": 14.317,
|
1510 |
+
"eval_trivia_qa_steps_per_second": 0.573,
|
1511 |
+
"epoch": 1.75,
|
1512 |
+
"step": 100
|
1513 |
+
},
|
1514 |
+
{
|
1515 |
+
"train_runtime": 538.1504,
|
1516 |
+
"train_samples_per_second": 5.946,
|
1517 |
+
"train_steps_per_second": 0.186,
|
1518 |
+
"total_flos": 0.0,
|
1519 |
+
"train_loss": 0.593766241967678,
|
1520 |
+
"epoch": 1.75,
|
1521 |
+
"step": 100
|
1522 |
+
}
|
1523 |
+
]]
|