juyongjiang commited on
Commit
80a6a02
1 Parent(s): 3daab96

upload model checkpoint

Browse files
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
README.md CHANGED
@@ -1,3 +1,76 @@
1
- ---
2
- license: gemma
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ tags:
4
+ - alignment-handbook
5
+ - generated_from_trainer
6
+ datasets:
7
+ - llama-duo/synth_summarize_dataset_dedup
8
+ base_model: google/gemma-7b
9
+ model-index:
10
+ - name: gemma7b-summarize-claude3sonnet-64k
11
+ results: []
12
+ ---
13
+
14
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
15
+ should probably proofread and complete it, then remove this comment. -->
16
+
17
+ # gemma7b-summarize-claude3sonnet-64k
18
+
19
+ This model is a fine-tuned version of [google/gemma-7b](https://huggingface.co/google/gemma-7b) on the llama-duo/synth_summarize_dataset_dedup dataset.
20
+ It achieves the following results on the evaluation set:
21
+ - Loss: 2.5547
22
+
23
+ ## Model description
24
+
25
+ More information needed
26
+
27
+ ## Intended uses & limitations
28
+
29
+ More information needed
30
+
31
+ ## Training and evaluation data
32
+
33
+ More information needed
34
+
35
+ ## Training procedure
36
+
37
+ ### Training hyperparameters
38
+
39
+ The following hyperparameters were used during training:
40
+ - learning_rate: 0.0002
41
+ - train_batch_size: 4
42
+ - eval_batch_size: 2
43
+ - seed: 42
44
+ - distributed_type: multi-GPU
45
+ - num_devices: 8
46
+ - gradient_accumulation_steps: 2
47
+ - total_train_batch_size: 64
48
+ - total_eval_batch_size: 16
49
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
50
+ - lr_scheduler_type: cosine
51
+ - lr_scheduler_warmup_ratio: 0.1
52
+ - num_epochs: 10
53
+
54
+ ### Training results
55
+
56
+ | Training Loss | Epoch | Step | Validation Loss |
57
+ |:-------------:|:------:|:----:|:---------------:|
58
+ | 1.1216 | 0.9967 | 150 | 2.5805 |
59
+ | 0.9828 | 2.0 | 301 | 2.5169 |
60
+ | 0.9157 | 2.9967 | 451 | 2.4836 |
61
+ | 0.8753 | 4.0 | 602 | 2.5011 |
62
+ | 0.8334 | 4.9967 | 752 | 2.4945 |
63
+ | 0.796 | 6.0 | 903 | 2.5317 |
64
+ | 0.7745 | 6.9967 | 1053 | 2.5436 |
65
+ | 0.7582 | 8.0 | 1204 | 2.5522 |
66
+ | 0.754 | 8.9967 | 1354 | 2.5504 |
67
+ | 0.7572 | 9.9668 | 1500 | 2.5547 |
68
+
69
+
70
+ ### Framework versions
71
+
72
+ - PEFT 0.10.0
73
+ - Transformers 4.40.0
74
+ - Pytorch 2.1.2+cu121
75
+ - Datasets 2.18.0
76
+ - Tokenizers 0.19.1
adapter_config.json ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "google/gemma-7b",
5
+ "bias": "none",
6
+ "fan_in_fan_out": false,
7
+ "inference_mode": true,
8
+ "init_lora_weights": true,
9
+ "layer_replication": null,
10
+ "layers_pattern": null,
11
+ "layers_to_transform": null,
12
+ "loftq_config": {},
13
+ "lora_alpha": 64,
14
+ "lora_dropout": 0.05,
15
+ "megatron_config": null,
16
+ "megatron_core": "megatron.core",
17
+ "modules_to_save": null,
18
+ "peft_type": "LORA",
19
+ "r": 32,
20
+ "rank_pattern": {},
21
+ "revision": null,
22
+ "target_modules": [
23
+ "v_proj",
24
+ "q_proj"
25
+ ],
26
+ "task_type": "CAUSAL_LM",
27
+ "use_dora": false,
28
+ "use_rslora": false
29
+ }
adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:af543f051c5eebf1124e67f209d3c059c602369b5c1de960797b5b23c7c5a483
3
+ size 25705248
all_results.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 9.966777408637874,
3
+ "eval_loss": 2.5546562671661377,
4
+ "eval_runtime": 0.2362,
5
+ "eval_samples": 25,
6
+ "eval_samples_per_second": 42.34,
7
+ "eval_steps_per_second": 4.234,
8
+ "total_flos": 4.5794490708666614e+18,
9
+ "train_loss": 1.5882705609003702,
10
+ "train_runtime": 3659.0045,
11
+ "train_samples": 63494,
12
+ "train_samples_per_second": 26.291,
13
+ "train_steps_per_second": 0.41
14
+ }
config.json ADDED
@@ -0,0 +1,44 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "google/gemma-7b",
3
+ "architectures": [
4
+ "GemmaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 2,
9
+ "eos_token_id": 1,
10
+ "head_dim": 256,
11
+ "hidden_act": "gelu",
12
+ "hidden_activation": null,
13
+ "hidden_size": 3072,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 24576,
16
+ "max_position_embeddings": 8192,
17
+ "model_type": "gemma",
18
+ "num_attention_heads": 16,
19
+ "num_hidden_layers": 28,
20
+ "num_key_value_heads": 16,
21
+ "pad_token_id": 0,
22
+ "quantization_config": {
23
+ "_load_in_4bit": true,
24
+ "_load_in_8bit": false,
25
+ "bnb_4bit_compute_dtype": "bfloat16",
26
+ "bnb_4bit_quant_storage": "uint8",
27
+ "bnb_4bit_quant_type": "nf4",
28
+ "bnb_4bit_use_double_quant": false,
29
+ "llm_int8_enable_fp32_cpu_offload": false,
30
+ "llm_int8_has_fp16_weight": false,
31
+ "llm_int8_skip_modules": null,
32
+ "llm_int8_threshold": 6.0,
33
+ "load_in_4bit": true,
34
+ "load_in_8bit": false,
35
+ "quant_method": "bitsandbytes"
36
+ },
37
+ "rms_norm_eps": 1e-06,
38
+ "rope_scaling": null,
39
+ "rope_theta": 10000.0,
40
+ "torch_dtype": "bfloat16",
41
+ "transformers_version": "4.40.0",
42
+ "use_cache": true,
43
+ "vocab_size": 256000
44
+ }
eval_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 9.966777408637874,
3
+ "eval_loss": 2.5546562671661377,
4
+ "eval_runtime": 0.2362,
5
+ "eval_samples": 25,
6
+ "eval_samples_per_second": 42.34,
7
+ "eval_steps_per_second": 4.234
8
+ }
runs/Jun13_06-31-36_gpu1-1/events.out.tfevents.1718231645.gpu1-1.449944.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:23fd17c3b5113c660586eeb95bdc67c19ef297471656bfbdd9954c69e5f42366
3
+ size 71954
runs/Jun13_06-31-36_gpu1-1/events.out.tfevents.1718235304.gpu1-1.449944.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9f62d91cfb5510b2161ecc761700b913e816534575e08294745849bd38afa518
3
+ size 359
special_tokens_map.json ADDED
@@ -0,0 +1,34 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<|im_start|>",
4
+ "<|im_end|>"
5
+ ],
6
+ "bos_token": {
7
+ "content": "<bos>",
8
+ "lstrip": false,
9
+ "normalized": false,
10
+ "rstrip": false,
11
+ "single_word": false
12
+ },
13
+ "eos_token": {
14
+ "content": "<eos>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false
19
+ },
20
+ "pad_token": {
21
+ "content": "<pad>",
22
+ "lstrip": false,
23
+ "normalized": false,
24
+ "rstrip": false,
25
+ "single_word": false
26
+ },
27
+ "unk_token": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false
33
+ }
34
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:322a5f52ab5cab196761ab397a022d6fa3a2e1418585e532bb6efb2fedd2ae94
3
+ size 17477501
tokenizer_config.json ADDED
@@ -0,0 +1,70 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": false,
3
+ "add_eos_token": false,
4
+ "added_tokens_decoder": {
5
+ "0": {
6
+ "content": "<pad>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "1": {
14
+ "content": "<eos>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "2": {
22
+ "content": "<bos>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ },
29
+ "3": {
30
+ "content": "<unk>",
31
+ "lstrip": false,
32
+ "normalized": false,
33
+ "rstrip": false,
34
+ "single_word": false,
35
+ "special": true
36
+ },
37
+ "106": {
38
+ "content": "<|im_start|>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false,
43
+ "special": true
44
+ },
45
+ "107": {
46
+ "content": "<|im_end|>",
47
+ "lstrip": false,
48
+ "normalized": false,
49
+ "rstrip": false,
50
+ "single_word": false,
51
+ "special": true
52
+ }
53
+ },
54
+ "additional_special_tokens": [
55
+ "<|im_start|>",
56
+ "<|im_end|>"
57
+ ],
58
+ "bos_token": "<bos>",
59
+ "chat_template": "{% if messages[0]['role'] == 'user' or messages[0]['role'] == 'system' %}{{ bos_token }}{% endif %}{% for message in messages %}{{ '<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n' }}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% elif messages[-1]['role'] == 'assistant' %}{{ eos_token }}{% endif %}",
60
+ "clean_up_tokenization_spaces": false,
61
+ "eos_token": "<eos>",
62
+ "legacy": null,
63
+ "model_max_length": 2048,
64
+ "pad_token": "<pad>",
65
+ "sp_model_kwargs": {},
66
+ "spaces_between_special_tokens": false,
67
+ "tokenizer_class": "GemmaTokenizer",
68
+ "unk_token": "<unk>",
69
+ "use_default_system_prompt": false
70
+ }
train_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 9.966777408637874,
3
+ "total_flos": 4.5794490708666614e+18,
4
+ "train_loss": 1.5882705609003702,
5
+ "train_runtime": 3659.0045,
6
+ "train_samples": 63494,
7
+ "train_samples_per_second": 26.291,
8
+ "train_steps_per_second": 0.41
9
+ }
trainer_state.json ADDED
@@ -0,0 +1,2217 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 9.966777408637874,
5
+ "eval_steps": 500,
6
+ "global_step": 1500,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.006644518272425249,
13
+ "grad_norm": 324.0,
14
+ "learning_rate": 1.3333333333333334e-06,
15
+ "loss": 34.1539,
16
+ "step": 1
17
+ },
18
+ {
19
+ "epoch": 0.03322259136212625,
20
+ "grad_norm": 328.0,
21
+ "learning_rate": 6.666666666666667e-06,
22
+ "loss": 34.4732,
23
+ "step": 5
24
+ },
25
+ {
26
+ "epoch": 0.0664451827242525,
27
+ "grad_norm": 132.0,
28
+ "learning_rate": 1.3333333333333333e-05,
29
+ "loss": 30.9731,
30
+ "step": 10
31
+ },
32
+ {
33
+ "epoch": 0.09966777408637874,
34
+ "grad_norm": 57.75,
35
+ "learning_rate": 2e-05,
36
+ "loss": 24.1357,
37
+ "step": 15
38
+ },
39
+ {
40
+ "epoch": 0.132890365448505,
41
+ "grad_norm": 19.125,
42
+ "learning_rate": 2.6666666666666667e-05,
43
+ "loss": 19.6743,
44
+ "step": 20
45
+ },
46
+ {
47
+ "epoch": 0.16611295681063123,
48
+ "grad_norm": 14.1875,
49
+ "learning_rate": 3.3333333333333335e-05,
50
+ "loss": 17.9465,
51
+ "step": 25
52
+ },
53
+ {
54
+ "epoch": 0.19933554817275748,
55
+ "grad_norm": 7.25,
56
+ "learning_rate": 4e-05,
57
+ "loss": 15.9561,
58
+ "step": 30
59
+ },
60
+ {
61
+ "epoch": 0.23255813953488372,
62
+ "grad_norm": 4.125,
63
+ "learning_rate": 4.666666666666667e-05,
64
+ "loss": 14.7788,
65
+ "step": 35
66
+ },
67
+ {
68
+ "epoch": 0.26578073089701,
69
+ "grad_norm": 3.484375,
70
+ "learning_rate": 5.333333333333333e-05,
71
+ "loss": 14.139,
72
+ "step": 40
73
+ },
74
+ {
75
+ "epoch": 0.29900332225913623,
76
+ "grad_norm": 4.75,
77
+ "learning_rate": 6e-05,
78
+ "loss": 13.5886,
79
+ "step": 45
80
+ },
81
+ {
82
+ "epoch": 0.33222591362126247,
83
+ "grad_norm": 6.125,
84
+ "learning_rate": 6.666666666666667e-05,
85
+ "loss": 13.0275,
86
+ "step": 50
87
+ },
88
+ {
89
+ "epoch": 0.3654485049833887,
90
+ "grad_norm": 11.5625,
91
+ "learning_rate": 7.333333333333333e-05,
92
+ "loss": 11.9071,
93
+ "step": 55
94
+ },
95
+ {
96
+ "epoch": 0.39867109634551495,
97
+ "grad_norm": 18.375,
98
+ "learning_rate": 8e-05,
99
+ "loss": 9.4575,
100
+ "step": 60
101
+ },
102
+ {
103
+ "epoch": 0.4318936877076412,
104
+ "grad_norm": 21.25,
105
+ "learning_rate": 8.666666666666667e-05,
106
+ "loss": 5.8479,
107
+ "step": 65
108
+ },
109
+ {
110
+ "epoch": 0.46511627906976744,
111
+ "grad_norm": 5.09375,
112
+ "learning_rate": 9.333333333333334e-05,
113
+ "loss": 2.6937,
114
+ "step": 70
115
+ },
116
+ {
117
+ "epoch": 0.4983388704318937,
118
+ "grad_norm": 6.46875,
119
+ "learning_rate": 0.0001,
120
+ "loss": 2.0051,
121
+ "step": 75
122
+ },
123
+ {
124
+ "epoch": 0.53156146179402,
125
+ "grad_norm": 2.78125,
126
+ "learning_rate": 0.00010666666666666667,
127
+ "loss": 1.7309,
128
+ "step": 80
129
+ },
130
+ {
131
+ "epoch": 0.5647840531561462,
132
+ "grad_norm": 0.73828125,
133
+ "learning_rate": 0.00011333333333333334,
134
+ "loss": 1.5823,
135
+ "step": 85
136
+ },
137
+ {
138
+ "epoch": 0.5980066445182725,
139
+ "grad_norm": 1.359375,
140
+ "learning_rate": 0.00012,
141
+ "loss": 1.4702,
142
+ "step": 90
143
+ },
144
+ {
145
+ "epoch": 0.6312292358803987,
146
+ "grad_norm": 0.94140625,
147
+ "learning_rate": 0.00012666666666666666,
148
+ "loss": 1.3996,
149
+ "step": 95
150
+ },
151
+ {
152
+ "epoch": 0.6644518272425249,
153
+ "grad_norm": 0.859375,
154
+ "learning_rate": 0.00013333333333333334,
155
+ "loss": 1.3389,
156
+ "step": 100
157
+ },
158
+ {
159
+ "epoch": 0.6976744186046512,
160
+ "grad_norm": 1.0390625,
161
+ "learning_rate": 0.00014,
162
+ "loss": 1.293,
163
+ "step": 105
164
+ },
165
+ {
166
+ "epoch": 0.7308970099667774,
167
+ "grad_norm": 1.2265625,
168
+ "learning_rate": 0.00014666666666666666,
169
+ "loss": 1.2656,
170
+ "step": 110
171
+ },
172
+ {
173
+ "epoch": 0.7641196013289037,
174
+ "grad_norm": 0.5703125,
175
+ "learning_rate": 0.00015333333333333334,
176
+ "loss": 1.2254,
177
+ "step": 115
178
+ },
179
+ {
180
+ "epoch": 0.7973421926910299,
181
+ "grad_norm": 1.0546875,
182
+ "learning_rate": 0.00016,
183
+ "loss": 1.2072,
184
+ "step": 120
185
+ },
186
+ {
187
+ "epoch": 0.8305647840531561,
188
+ "grad_norm": 1.28125,
189
+ "learning_rate": 0.0001666666666666667,
190
+ "loss": 1.1856,
191
+ "step": 125
192
+ },
193
+ {
194
+ "epoch": 0.8637873754152824,
195
+ "grad_norm": 1.3046875,
196
+ "learning_rate": 0.00017333333333333334,
197
+ "loss": 1.169,
198
+ "step": 130
199
+ },
200
+ {
201
+ "epoch": 0.8970099667774086,
202
+ "grad_norm": 1.3125,
203
+ "learning_rate": 0.00018,
204
+ "loss": 1.1497,
205
+ "step": 135
206
+ },
207
+ {
208
+ "epoch": 0.9302325581395349,
209
+ "grad_norm": 1.9765625,
210
+ "learning_rate": 0.0001866666666666667,
211
+ "loss": 1.131,
212
+ "step": 140
213
+ },
214
+ {
215
+ "epoch": 0.9634551495016611,
216
+ "grad_norm": 1.609375,
217
+ "learning_rate": 0.00019333333333333333,
218
+ "loss": 1.1275,
219
+ "step": 145
220
+ },
221
+ {
222
+ "epoch": 0.9966777408637874,
223
+ "grad_norm": 6.65625,
224
+ "learning_rate": 0.0002,
225
+ "loss": 1.1216,
226
+ "step": 150
227
+ },
228
+ {
229
+ "epoch": 0.9966777408637874,
230
+ "eval_loss": 2.580465793609619,
231
+ "eval_runtime": 0.2799,
232
+ "eval_samples_per_second": 35.728,
233
+ "eval_steps_per_second": 3.573,
234
+ "step": 150
235
+ },
236
+ {
237
+ "epoch": 1.0299003322259137,
238
+ "grad_norm": 1.1171875,
239
+ "learning_rate": 0.00019999323080037624,
240
+ "loss": 1.1202,
241
+ "step": 155
242
+ },
243
+ {
244
+ "epoch": 1.06312292358804,
245
+ "grad_norm": 7.21875,
246
+ "learning_rate": 0.00019997292411794618,
247
+ "loss": 1.0982,
248
+ "step": 160
249
+ },
250
+ {
251
+ "epoch": 1.0963455149501662,
252
+ "grad_norm": 0.6875,
253
+ "learning_rate": 0.0001999390827019096,
254
+ "loss": 1.1059,
255
+ "step": 165
256
+ },
257
+ {
258
+ "epoch": 1.1295681063122924,
259
+ "grad_norm": 0.74609375,
260
+ "learning_rate": 0.0001998917111338525,
261
+ "loss": 1.079,
262
+ "step": 170
263
+ },
264
+ {
265
+ "epoch": 1.1627906976744187,
266
+ "grad_norm": 1.203125,
267
+ "learning_rate": 0.00019983081582712685,
268
+ "loss": 1.0626,
269
+ "step": 175
270
+ },
271
+ {
272
+ "epoch": 1.196013289036545,
273
+ "grad_norm": 3.75,
274
+ "learning_rate": 0.00019975640502598244,
275
+ "loss": 1.0644,
276
+ "step": 180
277
+ },
278
+ {
279
+ "epoch": 1.2292358803986712,
280
+ "grad_norm": 0.85546875,
281
+ "learning_rate": 0.00019966848880445062,
282
+ "loss": 1.064,
283
+ "step": 185
284
+ },
285
+ {
286
+ "epoch": 1.2624584717607974,
287
+ "grad_norm": 1.1796875,
288
+ "learning_rate": 0.00019956707906498044,
289
+ "loss": 1.0638,
290
+ "step": 190
291
+ },
292
+ {
293
+ "epoch": 1.2956810631229236,
294
+ "grad_norm": 1.75,
295
+ "learning_rate": 0.00019945218953682734,
296
+ "loss": 1.0598,
297
+ "step": 195
298
+ },
299
+ {
300
+ "epoch": 1.3289036544850499,
301
+ "grad_norm": 1.2734375,
302
+ "learning_rate": 0.00019932383577419432,
303
+ "loss": 1.0433,
304
+ "step": 200
305
+ },
306
+ {
307
+ "epoch": 1.3621262458471761,
308
+ "grad_norm": 1.1171875,
309
+ "learning_rate": 0.00019918203515412617,
310
+ "loss": 1.0375,
311
+ "step": 205
312
+ },
313
+ {
314
+ "epoch": 1.3953488372093024,
315
+ "grad_norm": 1.1171875,
316
+ "learning_rate": 0.00019902680687415705,
317
+ "loss": 1.0293,
318
+ "step": 210
319
+ },
320
+ {
321
+ "epoch": 1.4285714285714286,
322
+ "grad_norm": 1.1640625,
323
+ "learning_rate": 0.00019885817194971117,
324
+ "loss": 1.0196,
325
+ "step": 215
326
+ },
327
+ {
328
+ "epoch": 1.4617940199335548,
329
+ "grad_norm": 1.3828125,
330
+ "learning_rate": 0.00019867615321125795,
331
+ "loss": 1.0227,
332
+ "step": 220
333
+ },
334
+ {
335
+ "epoch": 1.495016611295681,
336
+ "grad_norm": 2.703125,
337
+ "learning_rate": 0.00019848077530122083,
338
+ "loss": 1.0192,
339
+ "step": 225
340
+ },
341
+ {
342
+ "epoch": 1.5282392026578073,
343
+ "grad_norm": 2.90625,
344
+ "learning_rate": 0.00019827206467064133,
345
+ "loss": 1.0254,
346
+ "step": 230
347
+ },
348
+ {
349
+ "epoch": 1.5614617940199336,
350
+ "grad_norm": 1.90625,
351
+ "learning_rate": 0.00019805004957559793,
352
+ "loss": 1.0076,
353
+ "step": 235
354
+ },
355
+ {
356
+ "epoch": 1.5946843853820598,
357
+ "grad_norm": 1.2578125,
358
+ "learning_rate": 0.00019781476007338058,
359
+ "loss": 0.9979,
360
+ "step": 240
361
+ },
362
+ {
363
+ "epoch": 1.627906976744186,
364
+ "grad_norm": 4.1875,
365
+ "learning_rate": 0.00019756622801842143,
366
+ "loss": 0.9963,
367
+ "step": 245
368
+ },
369
+ {
370
+ "epoch": 1.6611295681063123,
371
+ "grad_norm": 2.625,
372
+ "learning_rate": 0.00019730448705798239,
373
+ "loss": 1.0017,
374
+ "step": 250
375
+ },
376
+ {
377
+ "epoch": 1.6943521594684385,
378
+ "grad_norm": 2.9375,
379
+ "learning_rate": 0.00019702957262759965,
380
+ "loss": 1.0055,
381
+ "step": 255
382
+ },
383
+ {
384
+ "epoch": 1.7275747508305648,
385
+ "grad_norm": 2.40625,
386
+ "learning_rate": 0.00019674152194628638,
387
+ "loss": 0.993,
388
+ "step": 260
389
+ },
390
+ {
391
+ "epoch": 1.760797342192691,
392
+ "grad_norm": 1.3046875,
393
+ "learning_rate": 0.0001964403740114939,
394
+ "loss": 0.9875,
395
+ "step": 265
396
+ },
397
+ {
398
+ "epoch": 1.7940199335548173,
399
+ "grad_norm": 1.2734375,
400
+ "learning_rate": 0.0001961261695938319,
401
+ "loss": 1.0015,
402
+ "step": 270
403
+ },
404
+ {
405
+ "epoch": 1.8272425249169435,
406
+ "grad_norm": 1.0,
407
+ "learning_rate": 0.0001957989512315489,
408
+ "loss": 0.9879,
409
+ "step": 275
410
+ },
411
+ {
412
+ "epoch": 1.8604651162790697,
413
+ "grad_norm": 1.8828125,
414
+ "learning_rate": 0.0001954587632247732,
415
+ "loss": 0.9846,
416
+ "step": 280
417
+ },
418
+ {
419
+ "epoch": 1.893687707641196,
420
+ "grad_norm": 1.09375,
421
+ "learning_rate": 0.00019510565162951537,
422
+ "loss": 0.9816,
423
+ "step": 285
424
+ },
425
+ {
426
+ "epoch": 1.9269102990033222,
427
+ "grad_norm": 1.15625,
428
+ "learning_rate": 0.00019473966425143292,
429
+ "loss": 0.9832,
430
+ "step": 290
431
+ },
432
+ {
433
+ "epoch": 1.9601328903654485,
434
+ "grad_norm": 1.3359375,
435
+ "learning_rate": 0.00019436085063935835,
436
+ "loss": 0.9838,
437
+ "step": 295
438
+ },
439
+ {
440
+ "epoch": 1.9933554817275747,
441
+ "grad_norm": 0.76171875,
442
+ "learning_rate": 0.00019396926207859084,
443
+ "loss": 0.9828,
444
+ "step": 300
445
+ },
446
+ {
447
+ "epoch": 2.0,
448
+ "eval_loss": 2.516935110092163,
449
+ "eval_runtime": 0.2355,
450
+ "eval_samples_per_second": 42.456,
451
+ "eval_steps_per_second": 4.246,
452
+ "step": 301
453
+ },
454
+ {
455
+ "epoch": 2.026578073089701,
456
+ "grad_norm": 1.765625,
457
+ "learning_rate": 0.00019356495158395315,
458
+ "loss": 0.9602,
459
+ "step": 305
460
+ },
461
+ {
462
+ "epoch": 2.0598006644518274,
463
+ "grad_norm": 3.375,
464
+ "learning_rate": 0.00019314797389261424,
465
+ "loss": 0.9484,
466
+ "step": 310
467
+ },
468
+ {
469
+ "epoch": 2.0930232558139537,
470
+ "grad_norm": 0.54296875,
471
+ "learning_rate": 0.00019271838545667876,
472
+ "loss": 0.9496,
473
+ "step": 315
474
+ },
475
+ {
476
+ "epoch": 2.12624584717608,
477
+ "grad_norm": 0.80859375,
478
+ "learning_rate": 0.00019227624443554425,
479
+ "loss": 0.9405,
480
+ "step": 320
481
+ },
482
+ {
483
+ "epoch": 2.159468438538206,
484
+ "grad_norm": 1.4765625,
485
+ "learning_rate": 0.00019182161068802741,
486
+ "loss": 0.9509,
487
+ "step": 325
488
+ },
489
+ {
490
+ "epoch": 2.1926910299003324,
491
+ "grad_norm": 1.3515625,
492
+ "learning_rate": 0.0001913545457642601,
493
+ "loss": 0.9532,
494
+ "step": 330
495
+ },
496
+ {
497
+ "epoch": 2.2259136212624586,
498
+ "grad_norm": 1.0234375,
499
+ "learning_rate": 0.00019087511289735644,
500
+ "loss": 0.9421,
501
+ "step": 335
502
+ },
503
+ {
504
+ "epoch": 2.259136212624585,
505
+ "grad_norm": 3.453125,
506
+ "learning_rate": 0.00019038337699485208,
507
+ "loss": 0.9347,
508
+ "step": 340
509
+ },
510
+ {
511
+ "epoch": 2.292358803986711,
512
+ "grad_norm": 1.265625,
513
+ "learning_rate": 0.0001898794046299167,
514
+ "loss": 0.9451,
515
+ "step": 345
516
+ },
517
+ {
518
+ "epoch": 2.3255813953488373,
519
+ "grad_norm": 5.25,
520
+ "learning_rate": 0.00018936326403234125,
521
+ "loss": 0.9503,
522
+ "step": 350
523
+ },
524
+ {
525
+ "epoch": 2.3588039867109636,
526
+ "grad_norm": 1.2421875,
527
+ "learning_rate": 0.00018883502507930042,
528
+ "loss": 0.9515,
529
+ "step": 355
530
+ },
531
+ {
532
+ "epoch": 2.39202657807309,
533
+ "grad_norm": 1.4375,
534
+ "learning_rate": 0.00018829475928589271,
535
+ "loss": 0.9382,
536
+ "step": 360
537
+ },
538
+ {
539
+ "epoch": 2.425249169435216,
540
+ "grad_norm": 0.82421875,
541
+ "learning_rate": 0.0001877425397954582,
542
+ "loss": 0.9309,
543
+ "step": 365
544
+ },
545
+ {
546
+ "epoch": 2.4584717607973423,
547
+ "grad_norm": 1.578125,
548
+ "learning_rate": 0.00018717844136967624,
549
+ "loss": 0.9487,
550
+ "step": 370
551
+ },
552
+ {
553
+ "epoch": 2.4916943521594686,
554
+ "grad_norm": 1.3359375,
555
+ "learning_rate": 0.00018660254037844388,
556
+ "loss": 0.9414,
557
+ "step": 375
558
+ },
559
+ {
560
+ "epoch": 2.524916943521595,
561
+ "grad_norm": 1.3125,
562
+ "learning_rate": 0.00018601491478953657,
563
+ "loss": 0.9575,
564
+ "step": 380
565
+ },
566
+ {
567
+ "epoch": 2.558139534883721,
568
+ "grad_norm": 1.90625,
569
+ "learning_rate": 0.00018541564415805258,
570
+ "loss": 0.9469,
571
+ "step": 385
572
+ },
573
+ {
574
+ "epoch": 2.5913621262458473,
575
+ "grad_norm": 8.25,
576
+ "learning_rate": 0.0001848048096156426,
577
+ "loss": 0.9246,
578
+ "step": 390
579
+ },
580
+ {
581
+ "epoch": 2.6245847176079735,
582
+ "grad_norm": 0.921875,
583
+ "learning_rate": 0.00018418249385952575,
584
+ "loss": 0.9357,
585
+ "step": 395
586
+ },
587
+ {
588
+ "epoch": 2.6578073089700998,
589
+ "grad_norm": 1.59375,
590
+ "learning_rate": 0.00018354878114129367,
591
+ "loss": 0.9264,
592
+ "step": 400
593
+ },
594
+ {
595
+ "epoch": 2.691029900332226,
596
+ "grad_norm": 2.125,
597
+ "learning_rate": 0.00018290375725550417,
598
+ "loss": 0.934,
599
+ "step": 405
600
+ },
601
+ {
602
+ "epoch": 2.7242524916943522,
603
+ "grad_norm": 5.15625,
604
+ "learning_rate": 0.00018224750952806624,
605
+ "loss": 0.9378,
606
+ "step": 410
607
+ },
608
+ {
609
+ "epoch": 2.7574750830564785,
610
+ "grad_norm": 0.66796875,
611
+ "learning_rate": 0.00018158012680441723,
612
+ "loss": 0.9325,
613
+ "step": 415
614
+ },
615
+ {
616
+ "epoch": 2.7906976744186047,
617
+ "grad_norm": 1.109375,
618
+ "learning_rate": 0.00018090169943749476,
619
+ "loss": 0.9343,
620
+ "step": 420
621
+ },
622
+ {
623
+ "epoch": 2.823920265780731,
624
+ "grad_norm": 0.68359375,
625
+ "learning_rate": 0.0001802123192755044,
626
+ "loss": 0.9322,
627
+ "step": 425
628
+ },
629
+ {
630
+ "epoch": 2.857142857142857,
631
+ "grad_norm": 1.25,
632
+ "learning_rate": 0.0001795120796494848,
633
+ "loss": 0.9203,
634
+ "step": 430
635
+ },
636
+ {
637
+ "epoch": 2.8903654485049834,
638
+ "grad_norm": 0.67578125,
639
+ "learning_rate": 0.00017880107536067218,
640
+ "loss": 0.9181,
641
+ "step": 435
642
+ },
643
+ {
644
+ "epoch": 2.9235880398671097,
645
+ "grad_norm": 0.66796875,
646
+ "learning_rate": 0.00017807940266766593,
647
+ "loss": 0.9152,
648
+ "step": 440
649
+ },
650
+ {
651
+ "epoch": 2.956810631229236,
652
+ "grad_norm": 0.57421875,
653
+ "learning_rate": 0.0001773471592733964,
654
+ "loss": 0.9193,
655
+ "step": 445
656
+ },
657
+ {
658
+ "epoch": 2.990033222591362,
659
+ "grad_norm": 0.69140625,
660
+ "learning_rate": 0.0001766044443118978,
661
+ "loss": 0.9157,
662
+ "step": 450
663
+ },
664
+ {
665
+ "epoch": 2.9966777408637872,
666
+ "eval_loss": 2.4835643768310547,
667
+ "eval_runtime": 0.2608,
668
+ "eval_samples_per_second": 38.338,
669
+ "eval_steps_per_second": 3.834,
670
+ "step": 451
671
+ },
672
+ {
673
+ "epoch": 3.0232558139534884,
674
+ "grad_norm": 1.390625,
675
+ "learning_rate": 0.00017585135833488692,
676
+ "loss": 0.9023,
677
+ "step": 455
678
+ },
679
+ {
680
+ "epoch": 3.0564784053156147,
681
+ "grad_norm": 1.5078125,
682
+ "learning_rate": 0.00017508800329814995,
683
+ "loss": 0.8957,
684
+ "step": 460
685
+ },
686
+ {
687
+ "epoch": 3.089700996677741,
688
+ "grad_norm": 1.75,
689
+ "learning_rate": 0.00017431448254773944,
690
+ "loss": 0.8963,
691
+ "step": 465
692
+ },
693
+ {
694
+ "epoch": 3.122923588039867,
695
+ "grad_norm": 1.4921875,
696
+ "learning_rate": 0.0001735309008059829,
697
+ "loss": 0.8938,
698
+ "step": 470
699
+ },
700
+ {
701
+ "epoch": 3.1561461794019934,
702
+ "grad_norm": 1.1484375,
703
+ "learning_rate": 0.00017273736415730488,
704
+ "loss": 0.8832,
705
+ "step": 475
706
+ },
707
+ {
708
+ "epoch": 3.1893687707641196,
709
+ "grad_norm": 0.734375,
710
+ "learning_rate": 0.0001719339800338651,
711
+ "loss": 0.8824,
712
+ "step": 480
713
+ },
714
+ {
715
+ "epoch": 3.222591362126246,
716
+ "grad_norm": 0.92578125,
717
+ "learning_rate": 0.00017112085720101373,
718
+ "loss": 0.8985,
719
+ "step": 485
720
+ },
721
+ {
722
+ "epoch": 3.255813953488372,
723
+ "grad_norm": 0.77734375,
724
+ "learning_rate": 0.0001702981057425662,
725
+ "loss": 0.8915,
726
+ "step": 490
727
+ },
728
+ {
729
+ "epoch": 3.2890365448504983,
730
+ "grad_norm": 1.0703125,
731
+ "learning_rate": 0.00016946583704589973,
732
+ "loss": 0.8959,
733
+ "step": 495
734
+ },
735
+ {
736
+ "epoch": 3.3222591362126246,
737
+ "grad_norm": 0.640625,
738
+ "learning_rate": 0.0001686241637868734,
739
+ "loss": 0.8932,
740
+ "step": 500
741
+ },
742
+ {
743
+ "epoch": 3.355481727574751,
744
+ "grad_norm": 0.875,
745
+ "learning_rate": 0.00016777319991457325,
746
+ "loss": 0.9034,
747
+ "step": 505
748
+ },
749
+ {
750
+ "epoch": 3.388704318936877,
751
+ "grad_norm": 1.03125,
752
+ "learning_rate": 0.00016691306063588583,
753
+ "loss": 0.8914,
754
+ "step": 510
755
+ },
756
+ {
757
+ "epoch": 3.4219269102990033,
758
+ "grad_norm": 1.0078125,
759
+ "learning_rate": 0.00016604386239990078,
760
+ "loss": 0.8968,
761
+ "step": 515
762
+ },
763
+ {
764
+ "epoch": 3.4551495016611296,
765
+ "grad_norm": 0.7109375,
766
+ "learning_rate": 0.00016516572288214552,
767
+ "loss": 0.8899,
768
+ "step": 520
769
+ },
770
+ {
771
+ "epoch": 3.488372093023256,
772
+ "grad_norm": 0.55078125,
773
+ "learning_rate": 0.00016427876096865394,
774
+ "loss": 0.888,
775
+ "step": 525
776
+ },
777
+ {
778
+ "epoch": 3.521594684385382,
779
+ "grad_norm": 1.5703125,
780
+ "learning_rate": 0.00016338309673987101,
781
+ "loss": 0.8966,
782
+ "step": 530
783
+ },
784
+ {
785
+ "epoch": 3.5548172757475083,
786
+ "grad_norm": 0.7890625,
787
+ "learning_rate": 0.000162478851454396,
788
+ "loss": 0.8802,
789
+ "step": 535
790
+ },
791
+ {
792
+ "epoch": 3.5880398671096345,
793
+ "grad_norm": 0.63671875,
794
+ "learning_rate": 0.0001615661475325658,
795
+ "loss": 0.8864,
796
+ "step": 540
797
+ },
798
+ {
799
+ "epoch": 3.6212624584717608,
800
+ "grad_norm": 1.3359375,
801
+ "learning_rate": 0.00016064510853988138,
802
+ "loss": 0.8816,
803
+ "step": 545
804
+ },
805
+ {
806
+ "epoch": 3.654485049833887,
807
+ "grad_norm": 1.484375,
808
+ "learning_rate": 0.00015971585917027862,
809
+ "loss": 0.8906,
810
+ "step": 550
811
+ },
812
+ {
813
+ "epoch": 3.6877076411960132,
814
+ "grad_norm": 1.09375,
815
+ "learning_rate": 0.00015877852522924732,
816
+ "loss": 0.8896,
817
+ "step": 555
818
+ },
819
+ {
820
+ "epoch": 3.7209302325581395,
821
+ "grad_norm": 0.73046875,
822
+ "learning_rate": 0.00015783323361679864,
823
+ "loss": 0.8806,
824
+ "step": 560
825
+ },
826
+ {
827
+ "epoch": 3.7541528239202657,
828
+ "grad_norm": 1.25,
829
+ "learning_rate": 0.00015688011231028518,
830
+ "loss": 0.8758,
831
+ "step": 565
832
+ },
833
+ {
834
+ "epoch": 3.787375415282392,
835
+ "grad_norm": 1.2109375,
836
+ "learning_rate": 0.0001559192903470747,
837
+ "loss": 0.871,
838
+ "step": 570
839
+ },
840
+ {
841
+ "epoch": 3.820598006644518,
842
+ "grad_norm": 0.7578125,
843
+ "learning_rate": 0.0001549508978070806,
844
+ "loss": 0.8882,
845
+ "step": 575
846
+ },
847
+ {
848
+ "epoch": 3.8538205980066444,
849
+ "grad_norm": 0.66015625,
850
+ "learning_rate": 0.0001539750657951513,
851
+ "loss": 0.8719,
852
+ "step": 580
853
+ },
854
+ {
855
+ "epoch": 3.8870431893687707,
856
+ "grad_norm": 0.58203125,
857
+ "learning_rate": 0.0001529919264233205,
858
+ "loss": 0.8794,
859
+ "step": 585
860
+ },
861
+ {
862
+ "epoch": 3.920265780730897,
863
+ "grad_norm": 0.82421875,
864
+ "learning_rate": 0.00015200161279292155,
865
+ "loss": 0.8787,
866
+ "step": 590
867
+ },
868
+ {
869
+ "epoch": 3.953488372093023,
870
+ "grad_norm": 0.8125,
871
+ "learning_rate": 0.00015100425897656753,
872
+ "loss": 0.873,
873
+ "step": 595
874
+ },
875
+ {
876
+ "epoch": 3.9867109634551494,
877
+ "grad_norm": 0.578125,
878
+ "learning_rate": 0.00015000000000000001,
879
+ "loss": 0.8753,
880
+ "step": 600
881
+ },
882
+ {
883
+ "epoch": 4.0,
884
+ "eval_loss": 2.5010673999786377,
885
+ "eval_runtime": 0.239,
886
+ "eval_samples_per_second": 41.842,
887
+ "eval_steps_per_second": 4.184,
888
+ "step": 602
889
+ },
890
+ {
891
+ "epoch": 4.019933554817276,
892
+ "grad_norm": 1.359375,
893
+ "learning_rate": 0.0001489889718238087,
894
+ "loss": 0.8697,
895
+ "step": 605
896
+ },
897
+ {
898
+ "epoch": 4.053156146179402,
899
+ "grad_norm": 0.9921875,
900
+ "learning_rate": 0.00014797131132502465,
901
+ "loss": 0.8496,
902
+ "step": 610
903
+ },
904
+ {
905
+ "epoch": 4.086378737541528,
906
+ "grad_norm": 1.765625,
907
+ "learning_rate": 0.00014694715627858908,
908
+ "loss": 0.8601,
909
+ "step": 615
910
+ },
911
+ {
912
+ "epoch": 4.119601328903655,
913
+ "grad_norm": 1.140625,
914
+ "learning_rate": 0.00014591664533870118,
915
+ "loss": 0.8647,
916
+ "step": 620
917
+ },
918
+ {
919
+ "epoch": 4.152823920265781,
920
+ "grad_norm": 1.140625,
921
+ "learning_rate": 0.00014487991802004623,
922
+ "loss": 0.8541,
923
+ "step": 625
924
+ },
925
+ {
926
+ "epoch": 4.186046511627907,
927
+ "grad_norm": 0.6015625,
928
+ "learning_rate": 0.00014383711467890774,
929
+ "loss": 0.8481,
930
+ "step": 630
931
+ },
932
+ {
933
+ "epoch": 4.219269102990033,
934
+ "grad_norm": 0.96875,
935
+ "learning_rate": 0.00014278837649416544,
936
+ "loss": 0.8514,
937
+ "step": 635
938
+ },
939
+ {
940
+ "epoch": 4.25249169435216,
941
+ "grad_norm": 0.79296875,
942
+ "learning_rate": 0.0001417338454481818,
943
+ "loss": 0.8498,
944
+ "step": 640
945
+ },
946
+ {
947
+ "epoch": 4.285714285714286,
948
+ "grad_norm": 0.734375,
949
+ "learning_rate": 0.00014067366430758004,
950
+ "loss": 0.8368,
951
+ "step": 645
952
+ },
953
+ {
954
+ "epoch": 4.318936877076412,
955
+ "grad_norm": 0.875,
956
+ "learning_rate": 0.0001396079766039157,
957
+ "loss": 0.8439,
958
+ "step": 650
959
+ },
960
+ {
961
+ "epoch": 4.352159468438538,
962
+ "grad_norm": 0.67578125,
963
+ "learning_rate": 0.00013853692661424484,
964
+ "loss": 0.8565,
965
+ "step": 655
966
+ },
967
+ {
968
+ "epoch": 4.385382059800665,
969
+ "grad_norm": 0.79296875,
970
+ "learning_rate": 0.00013746065934159123,
971
+ "loss": 0.8426,
972
+ "step": 660
973
+ },
974
+ {
975
+ "epoch": 4.4186046511627906,
976
+ "grad_norm": 0.6953125,
977
+ "learning_rate": 0.00013637932049531516,
978
+ "loss": 0.8471,
979
+ "step": 665
980
+ },
981
+ {
982
+ "epoch": 4.451827242524917,
983
+ "grad_norm": 1.28125,
984
+ "learning_rate": 0.00013529305647138687,
985
+ "loss": 0.8417,
986
+ "step": 670
987
+ },
988
+ {
989
+ "epoch": 4.485049833887043,
990
+ "grad_norm": 0.94140625,
991
+ "learning_rate": 0.00013420201433256689,
992
+ "loss": 0.8493,
993
+ "step": 675
994
+ },
995
+ {
996
+ "epoch": 4.51827242524917,
997
+ "grad_norm": 1.59375,
998
+ "learning_rate": 0.0001331063417884958,
999
+ "loss": 0.8506,
1000
+ "step": 680
1001
+ },
1002
+ {
1003
+ "epoch": 4.5514950166112955,
1004
+ "grad_norm": 0.97265625,
1005
+ "learning_rate": 0.00013200618717569714,
1006
+ "loss": 0.841,
1007
+ "step": 685
1008
+ },
1009
+ {
1010
+ "epoch": 4.584717607973422,
1011
+ "grad_norm": 0.75,
1012
+ "learning_rate": 0.00013090169943749476,
1013
+ "loss": 0.8415,
1014
+ "step": 690
1015
+ },
1016
+ {
1017
+ "epoch": 4.617940199335548,
1018
+ "grad_norm": 0.703125,
1019
+ "learning_rate": 0.0001297930281038482,
1020
+ "loss": 0.8506,
1021
+ "step": 695
1022
+ },
1023
+ {
1024
+ "epoch": 4.651162790697675,
1025
+ "grad_norm": 0.8359375,
1026
+ "learning_rate": 0.00012868032327110904,
1027
+ "loss": 0.8425,
1028
+ "step": 700
1029
+ },
1030
+ {
1031
+ "epoch": 4.6843853820598005,
1032
+ "grad_norm": 1.1953125,
1033
+ "learning_rate": 0.0001275637355816999,
1034
+ "loss": 0.8466,
1035
+ "step": 705
1036
+ },
1037
+ {
1038
+ "epoch": 4.717607973421927,
1039
+ "grad_norm": 0.51171875,
1040
+ "learning_rate": 0.00012644341620372023,
1041
+ "loss": 0.841,
1042
+ "step": 710
1043
+ },
1044
+ {
1045
+ "epoch": 4.750830564784053,
1046
+ "grad_norm": 1.7578125,
1047
+ "learning_rate": 0.0001253195168104802,
1048
+ "loss": 0.8396,
1049
+ "step": 715
1050
+ },
1051
+ {
1052
+ "epoch": 4.78405315614618,
1053
+ "grad_norm": 1.609375,
1054
+ "learning_rate": 0.00012419218955996676,
1055
+ "loss": 0.8423,
1056
+ "step": 720
1057
+ },
1058
+ {
1059
+ "epoch": 4.8172757475083055,
1060
+ "grad_norm": 0.86328125,
1061
+ "learning_rate": 0.00012306158707424403,
1062
+ "loss": 0.839,
1063
+ "step": 725
1064
+ },
1065
+ {
1066
+ "epoch": 4.850498338870432,
1067
+ "grad_norm": 0.75,
1068
+ "learning_rate": 0.00012192786241879033,
1069
+ "loss": 0.8342,
1070
+ "step": 730
1071
+ },
1072
+ {
1073
+ "epoch": 4.883720930232558,
1074
+ "grad_norm": 1.1328125,
1075
+ "learning_rate": 0.00012079116908177593,
1076
+ "loss": 0.8358,
1077
+ "step": 735
1078
+ },
1079
+ {
1080
+ "epoch": 4.916943521594685,
1081
+ "grad_norm": 1.078125,
1082
+ "learning_rate": 0.00011965166095328301,
1083
+ "loss": 0.8432,
1084
+ "step": 740
1085
+ },
1086
+ {
1087
+ "epoch": 4.95016611295681,
1088
+ "grad_norm": 0.68359375,
1089
+ "learning_rate": 0.00011850949230447145,
1090
+ "loss": 0.8368,
1091
+ "step": 745
1092
+ },
1093
+ {
1094
+ "epoch": 4.983388704318937,
1095
+ "grad_norm": 1.0078125,
1096
+ "learning_rate": 0.00011736481776669306,
1097
+ "loss": 0.8334,
1098
+ "step": 750
1099
+ },
1100
+ {
1101
+ "epoch": 4.996677740863787,
1102
+ "eval_loss": 2.4944658279418945,
1103
+ "eval_runtime": 0.2592,
1104
+ "eval_samples_per_second": 38.58,
1105
+ "eval_steps_per_second": 3.858,
1106
+ "step": 752
1107
+ },
1108
+ {
1109
+ "epoch": 5.016611295681063,
1110
+ "grad_norm": 0.7265625,
1111
+ "learning_rate": 0.00011621779231055676,
1112
+ "loss": 0.8264,
1113
+ "step": 755
1114
+ },
1115
+ {
1116
+ "epoch": 5.04983388704319,
1117
+ "grad_norm": 1.6484375,
1118
+ "learning_rate": 0.00011506857122494831,
1119
+ "loss": 0.8175,
1120
+ "step": 760
1121
+ },
1122
+ {
1123
+ "epoch": 5.083056478405315,
1124
+ "grad_norm": 0.8828125,
1125
+ "learning_rate": 0.00011391731009600654,
1126
+ "loss": 0.8207,
1127
+ "step": 765
1128
+ },
1129
+ {
1130
+ "epoch": 5.116279069767442,
1131
+ "grad_norm": 0.8359375,
1132
+ "learning_rate": 0.00011276416478605949,
1133
+ "loss": 0.8134,
1134
+ "step": 770
1135
+ },
1136
+ {
1137
+ "epoch": 5.149501661129568,
1138
+ "grad_norm": 0.8828125,
1139
+ "learning_rate": 0.00011160929141252303,
1140
+ "loss": 0.8146,
1141
+ "step": 775
1142
+ },
1143
+ {
1144
+ "epoch": 5.1827242524916945,
1145
+ "grad_norm": 1.140625,
1146
+ "learning_rate": 0.00011045284632676536,
1147
+ "loss": 0.8118,
1148
+ "step": 780
1149
+ },
1150
+ {
1151
+ "epoch": 5.21594684385382,
1152
+ "grad_norm": 3.4375,
1153
+ "learning_rate": 0.00010929498609293924,
1154
+ "loss": 0.8142,
1155
+ "step": 785
1156
+ },
1157
+ {
1158
+ "epoch": 5.249169435215947,
1159
+ "grad_norm": 1.015625,
1160
+ "learning_rate": 0.00010813586746678583,
1161
+ "loss": 0.8156,
1162
+ "step": 790
1163
+ },
1164
+ {
1165
+ "epoch": 5.282392026578073,
1166
+ "grad_norm": 2.828125,
1167
+ "learning_rate": 0.00010697564737441252,
1168
+ "loss": 0.8097,
1169
+ "step": 795
1170
+ },
1171
+ {
1172
+ "epoch": 5.3156146179401995,
1173
+ "grad_norm": 0.8828125,
1174
+ "learning_rate": 0.00010581448289104758,
1175
+ "loss": 0.8213,
1176
+ "step": 800
1177
+ },
1178
+ {
1179
+ "epoch": 5.348837209302325,
1180
+ "grad_norm": 1.3984375,
1181
+ "learning_rate": 0.0001046525312197747,
1182
+ "loss": 0.8087,
1183
+ "step": 805
1184
+ },
1185
+ {
1186
+ "epoch": 5.382059800664452,
1187
+ "grad_norm": 1.0625,
1188
+ "learning_rate": 0.00010348994967025012,
1189
+ "loss": 0.8046,
1190
+ "step": 810
1191
+ },
1192
+ {
1193
+ "epoch": 5.415282392026578,
1194
+ "grad_norm": 2.734375,
1195
+ "learning_rate": 0.00010232689563740563,
1196
+ "loss": 0.8086,
1197
+ "step": 815
1198
+ },
1199
+ {
1200
+ "epoch": 5.4485049833887045,
1201
+ "grad_norm": 0.9921875,
1202
+ "learning_rate": 0.00010116352658013973,
1203
+ "loss": 0.809,
1204
+ "step": 820
1205
+ },
1206
+ {
1207
+ "epoch": 5.48172757475083,
1208
+ "grad_norm": 1.0,
1209
+ "learning_rate": 0.0001,
1210
+ "loss": 0.8155,
1211
+ "step": 825
1212
+ },
1213
+ {
1214
+ "epoch": 5.514950166112957,
1215
+ "grad_norm": 0.73046875,
1216
+ "learning_rate": 9.883647341986032e-05,
1217
+ "loss": 0.8016,
1218
+ "step": 830
1219
+ },
1220
+ {
1221
+ "epoch": 5.548172757475083,
1222
+ "grad_norm": 0.6796875,
1223
+ "learning_rate": 9.767310436259438e-05,
1224
+ "loss": 0.8013,
1225
+ "step": 835
1226
+ },
1227
+ {
1228
+ "epoch": 5.5813953488372094,
1229
+ "grad_norm": 0.81640625,
1230
+ "learning_rate": 9.651005032974994e-05,
1231
+ "loss": 0.8123,
1232
+ "step": 840
1233
+ },
1234
+ {
1235
+ "epoch": 5.614617940199335,
1236
+ "grad_norm": 2.1875,
1237
+ "learning_rate": 9.534746878022534e-05,
1238
+ "loss": 0.8163,
1239
+ "step": 845
1240
+ },
1241
+ {
1242
+ "epoch": 5.647840531561462,
1243
+ "grad_norm": 0.72265625,
1244
+ "learning_rate": 9.418551710895243e-05,
1245
+ "loss": 0.8164,
1246
+ "step": 850
1247
+ },
1248
+ {
1249
+ "epoch": 5.681063122923588,
1250
+ "grad_norm": 1.6953125,
1251
+ "learning_rate": 9.302435262558747e-05,
1252
+ "loss": 0.7974,
1253
+ "step": 855
1254
+ },
1255
+ {
1256
+ "epoch": 5.714285714285714,
1257
+ "grad_norm": 0.76953125,
1258
+ "learning_rate": 9.186413253321418e-05,
1259
+ "loss": 0.8142,
1260
+ "step": 860
1261
+ },
1262
+ {
1263
+ "epoch": 5.74750830564784,
1264
+ "grad_norm": 1.109375,
1265
+ "learning_rate": 9.070501390706079e-05,
1266
+ "loss": 0.8026,
1267
+ "step": 865
1268
+ },
1269
+ {
1270
+ "epoch": 5.780730897009967,
1271
+ "grad_norm": 0.640625,
1272
+ "learning_rate": 8.954715367323468e-05,
1273
+ "loss": 0.8005,
1274
+ "step": 870
1275
+ },
1276
+ {
1277
+ "epoch": 5.813953488372093,
1278
+ "grad_norm": 0.85546875,
1279
+ "learning_rate": 8.839070858747697e-05,
1280
+ "loss": 0.8015,
1281
+ "step": 875
1282
+ },
1283
+ {
1284
+ "epoch": 5.847176079734219,
1285
+ "grad_norm": 0.52734375,
1286
+ "learning_rate": 8.723583521394054e-05,
1287
+ "loss": 0.7924,
1288
+ "step": 880
1289
+ },
1290
+ {
1291
+ "epoch": 5.880398671096345,
1292
+ "grad_norm": 0.59765625,
1293
+ "learning_rate": 8.608268990399349e-05,
1294
+ "loss": 0.812,
1295
+ "step": 885
1296
+ },
1297
+ {
1298
+ "epoch": 5.913621262458472,
1299
+ "grad_norm": 0.70703125,
1300
+ "learning_rate": 8.49314287750517e-05,
1301
+ "loss": 0.7969,
1302
+ "step": 890
1303
+ },
1304
+ {
1305
+ "epoch": 5.946843853820598,
1306
+ "grad_norm": 0.74609375,
1307
+ "learning_rate": 8.378220768944327e-05,
1308
+ "loss": 0.7965,
1309
+ "step": 895
1310
+ },
1311
+ {
1312
+ "epoch": 5.980066445182724,
1313
+ "grad_norm": 2.015625,
1314
+ "learning_rate": 8.263518223330697e-05,
1315
+ "loss": 0.796,
1316
+ "step": 900
1317
+ },
1318
+ {
1319
+ "epoch": 6.0,
1320
+ "eval_loss": 2.531708240509033,
1321
+ "eval_runtime": 0.239,
1322
+ "eval_samples_per_second": 41.85,
1323
+ "eval_steps_per_second": 4.185,
1324
+ "step": 903
1325
+ },
1326
+ {
1327
+ "epoch": 6.01328903654485,
1328
+ "grad_norm": 0.482421875,
1329
+ "learning_rate": 8.149050769552856e-05,
1330
+ "loss": 0.7892,
1331
+ "step": 905
1332
+ },
1333
+ {
1334
+ "epoch": 6.046511627906977,
1335
+ "grad_norm": 0.65234375,
1336
+ "learning_rate": 8.034833904671698e-05,
1337
+ "loss": 0.7792,
1338
+ "step": 910
1339
+ },
1340
+ {
1341
+ "epoch": 6.079734219269103,
1342
+ "grad_norm": 0.7578125,
1343
+ "learning_rate": 7.920883091822408e-05,
1344
+ "loss": 0.7814,
1345
+ "step": 915
1346
+ },
1347
+ {
1348
+ "epoch": 6.112956810631229,
1349
+ "grad_norm": 0.484375,
1350
+ "learning_rate": 7.807213758120966e-05,
1351
+ "loss": 0.7822,
1352
+ "step": 920
1353
+ },
1354
+ {
1355
+ "epoch": 6.146179401993355,
1356
+ "grad_norm": 0.80859375,
1357
+ "learning_rate": 7.693841292575598e-05,
1358
+ "loss": 0.7749,
1359
+ "step": 925
1360
+ },
1361
+ {
1362
+ "epoch": 6.179401993355482,
1363
+ "grad_norm": 0.81640625,
1364
+ "learning_rate": 7.580781044003324e-05,
1365
+ "loss": 0.7821,
1366
+ "step": 930
1367
+ },
1368
+ {
1369
+ "epoch": 6.212624584717608,
1370
+ "grad_norm": 5.34375,
1371
+ "learning_rate": 7.468048318951983e-05,
1372
+ "loss": 0.7872,
1373
+ "step": 935
1374
+ },
1375
+ {
1376
+ "epoch": 6.245847176079734,
1377
+ "grad_norm": 2.21875,
1378
+ "learning_rate": 7.35565837962798e-05,
1379
+ "loss": 0.7855,
1380
+ "step": 940
1381
+ },
1382
+ {
1383
+ "epoch": 6.27906976744186,
1384
+ "grad_norm": 3.28125,
1385
+ "learning_rate": 7.243626441830009e-05,
1386
+ "loss": 0.7763,
1387
+ "step": 945
1388
+ },
1389
+ {
1390
+ "epoch": 6.312292358803987,
1391
+ "grad_norm": 0.62890625,
1392
+ "learning_rate": 7.131967672889101e-05,
1393
+ "loss": 0.7901,
1394
+ "step": 950
1395
+ },
1396
+ {
1397
+ "epoch": 6.3455149501661126,
1398
+ "grad_norm": 0.9765625,
1399
+ "learning_rate": 7.02069718961518e-05,
1400
+ "loss": 0.7814,
1401
+ "step": 955
1402
+ },
1403
+ {
1404
+ "epoch": 6.378737541528239,
1405
+ "grad_norm": 0.8203125,
1406
+ "learning_rate": 6.909830056250527e-05,
1407
+ "loss": 0.7752,
1408
+ "step": 960
1409
+ },
1410
+ {
1411
+ "epoch": 6.411960132890365,
1412
+ "grad_norm": 0.92578125,
1413
+ "learning_rate": 6.799381282430284e-05,
1414
+ "loss": 0.7782,
1415
+ "step": 965
1416
+ },
1417
+ {
1418
+ "epoch": 6.445182724252492,
1419
+ "grad_norm": 0.91796875,
1420
+ "learning_rate": 6.68936582115042e-05,
1421
+ "loss": 0.7748,
1422
+ "step": 970
1423
+ },
1424
+ {
1425
+ "epoch": 6.4784053156146175,
1426
+ "grad_norm": 1.1328125,
1427
+ "learning_rate": 6.579798566743314e-05,
1428
+ "loss": 0.7815,
1429
+ "step": 975
1430
+ },
1431
+ {
1432
+ "epoch": 6.511627906976744,
1433
+ "grad_norm": 3.734375,
1434
+ "learning_rate": 6.470694352861312e-05,
1435
+ "loss": 0.7747,
1436
+ "step": 980
1437
+ },
1438
+ {
1439
+ "epoch": 6.544850498338871,
1440
+ "grad_norm": 0.6015625,
1441
+ "learning_rate": 6.362067950468489e-05,
1442
+ "loss": 0.785,
1443
+ "step": 985
1444
+ },
1445
+ {
1446
+ "epoch": 6.578073089700997,
1447
+ "grad_norm": 0.73828125,
1448
+ "learning_rate": 6.25393406584088e-05,
1449
+ "loss": 0.7716,
1450
+ "step": 990
1451
+ },
1452
+ {
1453
+ "epoch": 6.6112956810631225,
1454
+ "grad_norm": 0.79296875,
1455
+ "learning_rate": 6.146307338575519e-05,
1456
+ "loss": 0.7723,
1457
+ "step": 995
1458
+ },
1459
+ {
1460
+ "epoch": 6.644518272425249,
1461
+ "grad_norm": 0.69921875,
1462
+ "learning_rate": 6.039202339608432e-05,
1463
+ "loss": 0.7745,
1464
+ "step": 1000
1465
+ },
1466
+ {
1467
+ "epoch": 6.677740863787376,
1468
+ "grad_norm": 1.96875,
1469
+ "learning_rate": 5.9326335692419995e-05,
1470
+ "loss": 0.7848,
1471
+ "step": 1005
1472
+ },
1473
+ {
1474
+ "epoch": 6.710963455149502,
1475
+ "grad_norm": 0.734375,
1476
+ "learning_rate": 5.8266154551818216e-05,
1477
+ "loss": 0.7797,
1478
+ "step": 1010
1479
+ },
1480
+ {
1481
+ "epoch": 6.7441860465116275,
1482
+ "grad_norm": 0.474609375,
1483
+ "learning_rate": 5.72116235058346e-05,
1484
+ "loss": 0.7714,
1485
+ "step": 1015
1486
+ },
1487
+ {
1488
+ "epoch": 6.777408637873754,
1489
+ "grad_norm": 0.478515625,
1490
+ "learning_rate": 5.616288532109225e-05,
1491
+ "loss": 0.7716,
1492
+ "step": 1020
1493
+ },
1494
+ {
1495
+ "epoch": 6.810631229235881,
1496
+ "grad_norm": 0.494140625,
1497
+ "learning_rate": 5.5120081979953785e-05,
1498
+ "loss": 0.7807,
1499
+ "step": 1025
1500
+ },
1501
+ {
1502
+ "epoch": 6.843853820598007,
1503
+ "grad_norm": 0.65234375,
1504
+ "learning_rate": 5.4083354661298814e-05,
1505
+ "loss": 0.7647,
1506
+ "step": 1030
1507
+ },
1508
+ {
1509
+ "epoch": 6.877076411960132,
1510
+ "grad_norm": 0.6328125,
1511
+ "learning_rate": 5.305284372141095e-05,
1512
+ "loss": 0.7755,
1513
+ "step": 1035
1514
+ },
1515
+ {
1516
+ "epoch": 6.910299003322259,
1517
+ "grad_norm": 0.4765625,
1518
+ "learning_rate": 5.2028688674975415e-05,
1519
+ "loss": 0.7738,
1520
+ "step": 1040
1521
+ },
1522
+ {
1523
+ "epoch": 6.943521594684386,
1524
+ "grad_norm": 0.5625,
1525
+ "learning_rate": 5.101102817619131e-05,
1526
+ "loss": 0.7765,
1527
+ "step": 1045
1528
+ },
1529
+ {
1530
+ "epoch": 6.976744186046512,
1531
+ "grad_norm": 0.70703125,
1532
+ "learning_rate": 5.000000000000002e-05,
1533
+ "loss": 0.7745,
1534
+ "step": 1050
1535
+ },
1536
+ {
1537
+ "epoch": 6.996677740863787,
1538
+ "eval_loss": 2.5435612201690674,
1539
+ "eval_runtime": 0.2585,
1540
+ "eval_samples_per_second": 38.679,
1541
+ "eval_steps_per_second": 3.868,
1542
+ "step": 1053
1543
+ },
1544
+ {
1545
+ "epoch": 7.009966777408638,
1546
+ "grad_norm": 0.53125,
1547
+ "learning_rate": 4.899574102343247e-05,
1548
+ "loss": 0.771,
1549
+ "step": 1055
1550
+ },
1551
+ {
1552
+ "epoch": 7.043189368770764,
1553
+ "grad_norm": 0.640625,
1554
+ "learning_rate": 4.799838720707846e-05,
1555
+ "loss": 0.7653,
1556
+ "step": 1060
1557
+ },
1558
+ {
1559
+ "epoch": 7.076411960132891,
1560
+ "grad_norm": 0.52734375,
1561
+ "learning_rate": 4.700807357667952e-05,
1562
+ "loss": 0.7644,
1563
+ "step": 1065
1564
+ },
1565
+ {
1566
+ "epoch": 7.1096345514950166,
1567
+ "grad_norm": 0.490234375,
1568
+ "learning_rate": 4.6024934204848745e-05,
1569
+ "loss": 0.7632,
1570
+ "step": 1070
1571
+ },
1572
+ {
1573
+ "epoch": 7.142857142857143,
1574
+ "grad_norm": 0.55859375,
1575
+ "learning_rate": 4.50491021929194e-05,
1576
+ "loss": 0.7686,
1577
+ "step": 1075
1578
+ },
1579
+ {
1580
+ "epoch": 7.176079734219269,
1581
+ "grad_norm": 0.46484375,
1582
+ "learning_rate": 4.4080709652925336e-05,
1583
+ "loss": 0.7549,
1584
+ "step": 1080
1585
+ },
1586
+ {
1587
+ "epoch": 7.209302325581396,
1588
+ "grad_norm": 0.58203125,
1589
+ "learning_rate": 4.3119887689714844e-05,
1590
+ "loss": 0.7626,
1591
+ "step": 1085
1592
+ },
1593
+ {
1594
+ "epoch": 7.2425249169435215,
1595
+ "grad_norm": 0.5546875,
1596
+ "learning_rate": 4.216676638320135e-05,
1597
+ "loss": 0.7588,
1598
+ "step": 1090
1599
+ },
1600
+ {
1601
+ "epoch": 7.275747508305648,
1602
+ "grad_norm": 0.5,
1603
+ "learning_rate": 4.12214747707527e-05,
1604
+ "loss": 0.7583,
1605
+ "step": 1095
1606
+ },
1607
+ {
1608
+ "epoch": 7.308970099667774,
1609
+ "grad_norm": 0.6015625,
1610
+ "learning_rate": 4.028414082972141e-05,
1611
+ "loss": 0.7529,
1612
+ "step": 1100
1613
+ },
1614
+ {
1615
+ "epoch": 7.342192691029901,
1616
+ "grad_norm": 0.72265625,
1617
+ "learning_rate": 3.935489146011869e-05,
1618
+ "loss": 0.766,
1619
+ "step": 1105
1620
+ },
1621
+ {
1622
+ "epoch": 7.3754152823920265,
1623
+ "grad_norm": 0.46875,
1624
+ "learning_rate": 3.843385246743417e-05,
1625
+ "loss": 0.7592,
1626
+ "step": 1110
1627
+ },
1628
+ {
1629
+ "epoch": 7.408637873754153,
1630
+ "grad_norm": 0.431640625,
1631
+ "learning_rate": 3.7521148545604e-05,
1632
+ "loss": 0.7645,
1633
+ "step": 1115
1634
+ },
1635
+ {
1636
+ "epoch": 7.441860465116279,
1637
+ "grad_norm": 0.455078125,
1638
+ "learning_rate": 3.661690326012897e-05,
1639
+ "loss": 0.7629,
1640
+ "step": 1120
1641
+ },
1642
+ {
1643
+ "epoch": 7.475083056478406,
1644
+ "grad_norm": 0.4765625,
1645
+ "learning_rate": 3.5721239031346066e-05,
1646
+ "loss": 0.7591,
1647
+ "step": 1125
1648
+ },
1649
+ {
1650
+ "epoch": 7.5083056478405314,
1651
+ "grad_norm": 0.71484375,
1652
+ "learning_rate": 3.483427711785449e-05,
1653
+ "loss": 0.7558,
1654
+ "step": 1130
1655
+ },
1656
+ {
1657
+ "epoch": 7.541528239202658,
1658
+ "grad_norm": 0.53515625,
1659
+ "learning_rate": 3.395613760009925e-05,
1660
+ "loss": 0.7611,
1661
+ "step": 1135
1662
+ },
1663
+ {
1664
+ "epoch": 7.574750830564784,
1665
+ "grad_norm": 0.56640625,
1666
+ "learning_rate": 3.308693936411421e-05,
1667
+ "loss": 0.7619,
1668
+ "step": 1140
1669
+ },
1670
+ {
1671
+ "epoch": 7.607973421926911,
1672
+ "grad_norm": 0.44921875,
1673
+ "learning_rate": 3.222680008542678e-05,
1674
+ "loss": 0.7585,
1675
+ "step": 1145
1676
+ },
1677
+ {
1678
+ "epoch": 7.641196013289036,
1679
+ "grad_norm": 0.490234375,
1680
+ "learning_rate": 3.137583621312665e-05,
1681
+ "loss": 0.7551,
1682
+ "step": 1150
1683
+ },
1684
+ {
1685
+ "epoch": 7.674418604651163,
1686
+ "grad_norm": 0.490234375,
1687
+ "learning_rate": 3.053416295410026e-05,
1688
+ "loss": 0.7626,
1689
+ "step": 1155
1690
+ },
1691
+ {
1692
+ "epoch": 7.707641196013289,
1693
+ "grad_norm": 0.5,
1694
+ "learning_rate": 2.9701894257433826e-05,
1695
+ "loss": 0.764,
1696
+ "step": 1160
1697
+ },
1698
+ {
1699
+ "epoch": 7.740863787375416,
1700
+ "grad_norm": 0.46875,
1701
+ "learning_rate": 2.8879142798986292e-05,
1702
+ "loss": 0.755,
1703
+ "step": 1165
1704
+ },
1705
+ {
1706
+ "epoch": 7.774086378737541,
1707
+ "grad_norm": 0.46875,
1708
+ "learning_rate": 2.8066019966134904e-05,
1709
+ "loss": 0.7563,
1710
+ "step": 1170
1711
+ },
1712
+ {
1713
+ "epoch": 7.807308970099668,
1714
+ "grad_norm": 0.451171875,
1715
+ "learning_rate": 2.7262635842695127e-05,
1716
+ "loss": 0.7688,
1717
+ "step": 1175
1718
+ },
1719
+ {
1720
+ "epoch": 7.840531561461794,
1721
+ "grad_norm": 0.546875,
1722
+ "learning_rate": 2.6469099194017143e-05,
1723
+ "loss": 0.7665,
1724
+ "step": 1180
1725
+ },
1726
+ {
1727
+ "epoch": 7.8737541528239205,
1728
+ "grad_norm": 0.4453125,
1729
+ "learning_rate": 2.5685517452260567e-05,
1730
+ "loss": 0.7664,
1731
+ "step": 1185
1732
+ },
1733
+ {
1734
+ "epoch": 7.906976744186046,
1735
+ "grad_norm": 0.443359375,
1736
+ "learning_rate": 2.491199670185008e-05,
1737
+ "loss": 0.753,
1738
+ "step": 1190
1739
+ },
1740
+ {
1741
+ "epoch": 7.940199335548173,
1742
+ "grad_norm": 0.44921875,
1743
+ "learning_rate": 2.4148641665113113e-05,
1744
+ "loss": 0.7614,
1745
+ "step": 1195
1746
+ },
1747
+ {
1748
+ "epoch": 7.973421926910299,
1749
+ "grad_norm": 0.484375,
1750
+ "learning_rate": 2.339555568810221e-05,
1751
+ "loss": 0.7582,
1752
+ "step": 1200
1753
+ },
1754
+ {
1755
+ "epoch": 8.0,
1756
+ "eval_loss": 2.5521774291992188,
1757
+ "eval_runtime": 0.24,
1758
+ "eval_samples_per_second": 41.669,
1759
+ "eval_steps_per_second": 4.167,
1760
+ "step": 1204
1761
+ },
1762
+ {
1763
+ "epoch": 8.006644518272426,
1764
+ "grad_norm": 0.423828125,
1765
+ "learning_rate": 2.265284072660362e-05,
1766
+ "loss": 0.7646,
1767
+ "step": 1205
1768
+ },
1769
+ {
1770
+ "epoch": 8.039867109634551,
1771
+ "grad_norm": 0.44140625,
1772
+ "learning_rate": 2.192059733233408e-05,
1773
+ "loss": 0.758,
1774
+ "step": 1210
1775
+ },
1776
+ {
1777
+ "epoch": 8.073089700996677,
1778
+ "grad_norm": 0.439453125,
1779
+ "learning_rate": 2.119892463932781e-05,
1780
+ "loss": 0.7566,
1781
+ "step": 1215
1782
+ },
1783
+ {
1784
+ "epoch": 8.106312292358805,
1785
+ "grad_norm": 0.451171875,
1786
+ "learning_rate": 2.0487920350515212e-05,
1787
+ "loss": 0.7551,
1788
+ "step": 1220
1789
+ },
1790
+ {
1791
+ "epoch": 8.13953488372093,
1792
+ "grad_norm": 0.4375,
1793
+ "learning_rate": 1.9787680724495617e-05,
1794
+ "loss": 0.7421,
1795
+ "step": 1225
1796
+ },
1797
+ {
1798
+ "epoch": 8.172757475083056,
1799
+ "grad_norm": 0.44921875,
1800
+ "learning_rate": 1.9098300562505266e-05,
1801
+ "loss": 0.7513,
1802
+ "step": 1230
1803
+ },
1804
+ {
1805
+ "epoch": 8.205980066445182,
1806
+ "grad_norm": 0.44921875,
1807
+ "learning_rate": 1.8419873195582814e-05,
1808
+ "loss": 0.7578,
1809
+ "step": 1235
1810
+ },
1811
+ {
1812
+ "epoch": 8.23920265780731,
1813
+ "grad_norm": 0.421875,
1814
+ "learning_rate": 1.775249047193377e-05,
1815
+ "loss": 0.7518,
1816
+ "step": 1240
1817
+ },
1818
+ {
1819
+ "epoch": 8.272425249169435,
1820
+ "grad_norm": 0.498046875,
1821
+ "learning_rate": 1.7096242744495837e-05,
1822
+ "loss": 0.7519,
1823
+ "step": 1245
1824
+ },
1825
+ {
1826
+ "epoch": 8.305647840531561,
1827
+ "grad_norm": 0.5390625,
1828
+ "learning_rate": 1.6451218858706374e-05,
1829
+ "loss": 0.7514,
1830
+ "step": 1250
1831
+ },
1832
+ {
1833
+ "epoch": 8.338870431893687,
1834
+ "grad_norm": 0.43359375,
1835
+ "learning_rate": 1.5817506140474247e-05,
1836
+ "loss": 0.7553,
1837
+ "step": 1255
1838
+ },
1839
+ {
1840
+ "epoch": 8.372093023255815,
1841
+ "grad_norm": 0.466796875,
1842
+ "learning_rate": 1.5195190384357404e-05,
1843
+ "loss": 0.7487,
1844
+ "step": 1260
1845
+ },
1846
+ {
1847
+ "epoch": 8.40531561461794,
1848
+ "grad_norm": 0.43359375,
1849
+ "learning_rate": 1.458435584194745e-05,
1850
+ "loss": 0.7518,
1851
+ "step": 1265
1852
+ },
1853
+ {
1854
+ "epoch": 8.438538205980066,
1855
+ "grad_norm": 0.4296875,
1856
+ "learning_rate": 1.3985085210463477e-05,
1857
+ "loss": 0.7487,
1858
+ "step": 1270
1859
+ },
1860
+ {
1861
+ "epoch": 8.471760797342192,
1862
+ "grad_norm": 0.423828125,
1863
+ "learning_rate": 1.339745962155613e-05,
1864
+ "loss": 0.7467,
1865
+ "step": 1275
1866
+ },
1867
+ {
1868
+ "epoch": 8.50498338870432,
1869
+ "grad_norm": 0.4296875,
1870
+ "learning_rate": 1.2821558630323772e-05,
1871
+ "loss": 0.7478,
1872
+ "step": 1280
1873
+ },
1874
+ {
1875
+ "epoch": 8.538205980066445,
1876
+ "grad_norm": 0.46875,
1877
+ "learning_rate": 1.2257460204541794e-05,
1878
+ "loss": 0.7558,
1879
+ "step": 1285
1880
+ },
1881
+ {
1882
+ "epoch": 8.571428571428571,
1883
+ "grad_norm": 0.44921875,
1884
+ "learning_rate": 1.1705240714107302e-05,
1885
+ "loss": 0.7426,
1886
+ "step": 1290
1887
+ },
1888
+ {
1889
+ "epoch": 8.604651162790697,
1890
+ "grad_norm": 0.46875,
1891
+ "learning_rate": 1.116497492069961e-05,
1892
+ "loss": 0.7411,
1893
+ "step": 1295
1894
+ },
1895
+ {
1896
+ "epoch": 8.637873754152825,
1897
+ "grad_norm": 0.44140625,
1898
+ "learning_rate": 1.0636735967658784e-05,
1899
+ "loss": 0.7524,
1900
+ "step": 1300
1901
+ },
1902
+ {
1903
+ "epoch": 8.67109634551495,
1904
+ "grad_norm": 0.453125,
1905
+ "learning_rate": 1.0120595370083318e-05,
1906
+ "loss": 0.7499,
1907
+ "step": 1305
1908
+ },
1909
+ {
1910
+ "epoch": 8.704318936877076,
1911
+ "grad_norm": 0.435546875,
1912
+ "learning_rate": 9.616623005147951e-06,
1913
+ "loss": 0.7603,
1914
+ "step": 1310
1915
+ },
1916
+ {
1917
+ "epoch": 8.737541528239202,
1918
+ "grad_norm": 0.44140625,
1919
+ "learning_rate": 9.124887102643575e-06,
1920
+ "loss": 0.7563,
1921
+ "step": 1315
1922
+ },
1923
+ {
1924
+ "epoch": 8.77076411960133,
1925
+ "grad_norm": 0.4296875,
1926
+ "learning_rate": 8.645454235739903e-06,
1927
+ "loss": 0.7594,
1928
+ "step": 1320
1929
+ },
1930
+ {
1931
+ "epoch": 8.803986710963455,
1932
+ "grad_norm": 0.431640625,
1933
+ "learning_rate": 8.178389311972612e-06,
1934
+ "loss": 0.7648,
1935
+ "step": 1325
1936
+ },
1937
+ {
1938
+ "epoch": 8.837209302325581,
1939
+ "grad_norm": 0.443359375,
1940
+ "learning_rate": 7.72375556445577e-06,
1941
+ "loss": 0.7555,
1942
+ "step": 1330
1943
+ },
1944
+ {
1945
+ "epoch": 8.870431893687707,
1946
+ "grad_norm": 0.44140625,
1947
+ "learning_rate": 7.281614543321269e-06,
1948
+ "loss": 0.7461,
1949
+ "step": 1335
1950
+ },
1951
+ {
1952
+ "epoch": 8.903654485049834,
1953
+ "grad_norm": 0.470703125,
1954
+ "learning_rate": 6.852026107385756e-06,
1955
+ "loss": 0.7606,
1956
+ "step": 1340
1957
+ },
1958
+ {
1959
+ "epoch": 8.93687707641196,
1960
+ "grad_norm": 0.435546875,
1961
+ "learning_rate": 6.435048416046863e-06,
1962
+ "loss": 0.7598,
1963
+ "step": 1345
1964
+ },
1965
+ {
1966
+ "epoch": 8.970099667774086,
1967
+ "grad_norm": 0.439453125,
1968
+ "learning_rate": 6.030737921409169e-06,
1969
+ "loss": 0.754,
1970
+ "step": 1350
1971
+ },
1972
+ {
1973
+ "epoch": 8.996677740863788,
1974
+ "eval_loss": 2.5503978729248047,
1975
+ "eval_runtime": 0.2601,
1976
+ "eval_samples_per_second": 38.445,
1977
+ "eval_steps_per_second": 3.845,
1978
+ "step": 1354
1979
+ },
1980
+ {
1981
+ "epoch": 9.003322259136212,
1982
+ "grad_norm": 0.4375,
1983
+ "learning_rate": 5.639149360641649e-06,
1984
+ "loss": 0.7546,
1985
+ "step": 1355
1986
+ },
1987
+ {
1988
+ "epoch": 9.03654485049834,
1989
+ "grad_norm": 0.439453125,
1990
+ "learning_rate": 5.26033574856708e-06,
1991
+ "loss": 0.7562,
1992
+ "step": 1360
1993
+ },
1994
+ {
1995
+ "epoch": 9.069767441860465,
1996
+ "grad_norm": 0.419921875,
1997
+ "learning_rate": 4.8943483704846475e-06,
1998
+ "loss": 0.7522,
1999
+ "step": 1365
2000
+ },
2001
+ {
2002
+ "epoch": 9.102990033222591,
2003
+ "grad_norm": 0.427734375,
2004
+ "learning_rate": 4.541236775226809e-06,
2005
+ "loss": 0.7522,
2006
+ "step": 1370
2007
+ },
2008
+ {
2009
+ "epoch": 9.136212624584717,
2010
+ "grad_norm": 0.427734375,
2011
+ "learning_rate": 4.20104876845111e-06,
2012
+ "loss": 0.7509,
2013
+ "step": 1375
2014
+ },
2015
+ {
2016
+ "epoch": 9.169435215946844,
2017
+ "grad_norm": 0.41796875,
2018
+ "learning_rate": 3.873830406168111e-06,
2019
+ "loss": 0.7444,
2020
+ "step": 1380
2021
+ },
2022
+ {
2023
+ "epoch": 9.20265780730897,
2024
+ "grad_norm": 0.453125,
2025
+ "learning_rate": 3.5596259885061102e-06,
2026
+ "loss": 0.7561,
2027
+ "step": 1385
2028
+ },
2029
+ {
2030
+ "epoch": 9.235880398671096,
2031
+ "grad_norm": 0.443359375,
2032
+ "learning_rate": 3.2584780537136207e-06,
2033
+ "loss": 0.7502,
2034
+ "step": 1390
2035
+ },
2036
+ {
2037
+ "epoch": 9.269102990033222,
2038
+ "grad_norm": 0.4921875,
2039
+ "learning_rate": 2.970427372400353e-06,
2040
+ "loss": 0.7546,
2041
+ "step": 1395
2042
+ },
2043
+ {
2044
+ "epoch": 9.30232558139535,
2045
+ "grad_norm": 0.45703125,
2046
+ "learning_rate": 2.6955129420176196e-06,
2047
+ "loss": 0.7506,
2048
+ "step": 1400
2049
+ },
2050
+ {
2051
+ "epoch": 9.335548172757475,
2052
+ "grad_norm": 0.4296875,
2053
+ "learning_rate": 2.433771981578581e-06,
2054
+ "loss": 0.7531,
2055
+ "step": 1405
2056
+ },
2057
+ {
2058
+ "epoch": 9.368770764119601,
2059
+ "grad_norm": 0.427734375,
2060
+ "learning_rate": 2.1852399266194314e-06,
2061
+ "loss": 0.75,
2062
+ "step": 1410
2063
+ },
2064
+ {
2065
+ "epoch": 9.401993355481727,
2066
+ "grad_norm": 0.4765625,
2067
+ "learning_rate": 1.9499504244020693e-06,
2068
+ "loss": 0.7449,
2069
+ "step": 1415
2070
+ },
2071
+ {
2072
+ "epoch": 9.435215946843854,
2073
+ "grad_norm": 0.416015625,
2074
+ "learning_rate": 1.7279353293586765e-06,
2075
+ "loss": 0.765,
2076
+ "step": 1420
2077
+ },
2078
+ {
2079
+ "epoch": 9.46843853820598,
2080
+ "grad_norm": 0.44921875,
2081
+ "learning_rate": 1.5192246987791981e-06,
2082
+ "loss": 0.7472,
2083
+ "step": 1425
2084
+ },
2085
+ {
2086
+ "epoch": 9.501661129568106,
2087
+ "grad_norm": 0.431640625,
2088
+ "learning_rate": 1.323846788742078e-06,
2089
+ "loss": 0.7461,
2090
+ "step": 1430
2091
+ },
2092
+ {
2093
+ "epoch": 9.534883720930232,
2094
+ "grad_norm": 0.443359375,
2095
+ "learning_rate": 1.14182805028884e-06,
2096
+ "loss": 0.7501,
2097
+ "step": 1435
2098
+ },
2099
+ {
2100
+ "epoch": 9.56810631229236,
2101
+ "grad_norm": 0.43359375,
2102
+ "learning_rate": 9.731931258429638e-07,
2103
+ "loss": 0.7501,
2104
+ "step": 1440
2105
+ },
2106
+ {
2107
+ "epoch": 9.601328903654485,
2108
+ "grad_norm": 0.41796875,
2109
+ "learning_rate": 8.17964845873831e-07,
2110
+ "loss": 0.7511,
2111
+ "step": 1445
2112
+ },
2113
+ {
2114
+ "epoch": 9.634551495016611,
2115
+ "grad_norm": 0.427734375,
2116
+ "learning_rate": 6.761642258056978e-07,
2117
+ "loss": 0.7556,
2118
+ "step": 1450
2119
+ },
2120
+ {
2121
+ "epoch": 9.667774086378738,
2122
+ "grad_norm": 0.42578125,
2123
+ "learning_rate": 5.478104631726711e-07,
2124
+ "loss": 0.751,
2125
+ "step": 1455
2126
+ },
2127
+ {
2128
+ "epoch": 9.700996677740864,
2129
+ "grad_norm": 0.421875,
2130
+ "learning_rate": 4.329209350195651e-07,
2131
+ "loss": 0.7598,
2132
+ "step": 1460
2133
+ },
2134
+ {
2135
+ "epoch": 9.73421926910299,
2136
+ "grad_norm": 0.4375,
2137
+ "learning_rate": 3.315111955493944e-07,
2138
+ "loss": 0.7572,
2139
+ "step": 1465
2140
+ },
2141
+ {
2142
+ "epoch": 9.767441860465116,
2143
+ "grad_norm": 0.46484375,
2144
+ "learning_rate": 2.4359497401758024e-07,
2145
+ "loss": 0.7478,
2146
+ "step": 1470
2147
+ },
2148
+ {
2149
+ "epoch": 9.800664451827242,
2150
+ "grad_norm": 0.419921875,
2151
+ "learning_rate": 1.6918417287318245e-07,
2152
+ "loss": 0.749,
2153
+ "step": 1475
2154
+ },
2155
+ {
2156
+ "epoch": 9.83388704318937,
2157
+ "grad_norm": 0.44921875,
2158
+ "learning_rate": 1.0828886614754341e-07,
2159
+ "loss": 0.7488,
2160
+ "step": 1480
2161
+ },
2162
+ {
2163
+ "epoch": 9.867109634551495,
2164
+ "grad_norm": 0.4609375,
2165
+ "learning_rate": 6.09172980904238e-08,
2166
+ "loss": 0.7407,
2167
+ "step": 1485
2168
+ },
2169
+ {
2170
+ "epoch": 9.90033222591362,
2171
+ "grad_norm": 0.43359375,
2172
+ "learning_rate": 2.7075882053828605e-08,
2173
+ "loss": 0.7491,
2174
+ "step": 1490
2175
+ },
2176
+ {
2177
+ "epoch": 9.933554817275748,
2178
+ "grad_norm": 0.447265625,
2179
+ "learning_rate": 6.769199623779532e-09,
2180
+ "loss": 0.7417,
2181
+ "step": 1495
2182
+ },
2183
+ {
2184
+ "epoch": 9.966777408637874,
2185
+ "grad_norm": 0.435546875,
2186
+ "learning_rate": 0.0,
2187
+ "loss": 0.7572,
2188
+ "step": 1500
2189
+ },
2190
+ {
2191
+ "epoch": 9.966777408637874,
2192
+ "eval_loss": 2.5546562671661377,
2193
+ "eval_runtime": 0.2333,
2194
+ "eval_samples_per_second": 42.867,
2195
+ "eval_steps_per_second": 4.287,
2196
+ "step": 1500
2197
+ },
2198
+ {
2199
+ "epoch": 9.966777408637874,
2200
+ "step": 1500,
2201
+ "total_flos": 4.5794490708666614e+18,
2202
+ "train_loss": 1.5882705609003702,
2203
+ "train_runtime": 3659.0045,
2204
+ "train_samples_per_second": 26.291,
2205
+ "train_steps_per_second": 0.41
2206
+ }
2207
+ ],
2208
+ "logging_steps": 5,
2209
+ "max_steps": 1500,
2210
+ "num_input_tokens_seen": 0,
2211
+ "num_train_epochs": 10,
2212
+ "save_steps": 100,
2213
+ "total_flos": 4.5794490708666614e+18,
2214
+ "train_batch_size": 4,
2215
+ "trial_name": null,
2216
+ "trial_params": null
2217
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:259949fe184f41eaa8cac44a9bfa362aab2bed7016354254f4c4ab87a6e99cd7
3
+ size 5176