Hugging Face's logo

roonbug
/
h79e5321

Safetensors
Model card Files
xet
 Community
roonbug commited on
Commit
cc7f49f
·
verified ·
1 Parent(s): 6e182c3

Upload folder using huggingface_hub

Browse files
Files changed (20) hide show
  1. .gitattributes +1 -0
  2. checkpoint-500/added_tokens.json +3 -0
  3. checkpoint-500/chat_template.jinja +47 -0
  4. checkpoint-500/config.json +110 -0
  5. checkpoint-500/generation_config.json +13 -0
  6. checkpoint-500/model-00001-of-00005.safetensors +3 -0
  7. checkpoint-500/model-00002-of-00005.safetensors +3 -0
  8. checkpoint-500/model-00003-of-00005.safetensors +3 -0
  9. checkpoint-500/model-00004-of-00005.safetensors +3 -0
  10. checkpoint-500/model-00005-of-00005.safetensors +3 -0
  11. checkpoint-500/model.safetensors.index.json +0 -0
  12. checkpoint-500/optimizer.pt +3 -0
  13. checkpoint-500/rng_state.pth +3 -0
  14. checkpoint-500/scheduler.pt +3 -0
  15. checkpoint-500/special_tokens_map.json +33 -0
  16. checkpoint-500/tokenizer.json +3 -0
  17. checkpoint-500/tokenizer.model +3 -0
  18. checkpoint-500/tokenizer_config.json +0 -0
  19. checkpoint-500/trainer_state.json +644 -0
  20. checkpoint-500/training_args.bin +3 -0
.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
+ checkpoint-500/tokenizer.json filter=lfs diff=lfs merge=lfs -text
checkpoint-500/added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "<image_soft_token>": 262144
3
+ }
checkpoint-500/chat_template.jinja ADDED
@@ -0,0 +1,47 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {{ bos_token }}
2
+ {%- if messages[0]['role'] == 'system' -%}
3
+ {%- if messages[0]['content'] is string -%}
4
+ {%- set first_user_prefix = messages[0]['content'] + '
5
+
6
+ ' -%}
7
+ {%- else -%}
8
+ {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
9
+
10
+ ' -%}
11
+ {%- endif -%}
12
+ {%- set loop_messages = messages[1:] -%}
13
+ {%- else -%}
14
+ {%- set first_user_prefix = "" -%}
15
+ {%- set loop_messages = messages -%}
16
+ {%- endif -%}
17
+ {%- for message in loop_messages -%}
18
+ {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
19
+ {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
20
+ {%- endif -%}
21
+ {%- if (message['role'] == 'assistant') -%}
22
+ {%- set role = "model" -%}
23
+ {%- else -%}
24
+ {%- set role = message['role'] -%}
25
+ {%- endif -%}
26
+ {{ '<start_of_turn>' + role + '
27
+ ' + (first_user_prefix if loop.first else "") }}
28
+ {%- if message['content'] is string -%}
29
+ {{ message['content'] | trim }}
30
+ {%- elif message['content'] is iterable -%}
31
+ {%- for item in message['content'] -%}
32
+ {%- if item['type'] == 'image' -%}
33
+ {{ '<start_of_image>' }}
34
+ {%- elif item['type'] == 'text' -%}
35
+ {{ item['text'] | trim }}
36
+ {%- endif -%}
37
+ {%- endfor -%}
38
+ {%- else -%}
39
+ {{ raise_exception("Invalid content type") }}
40
+ {%- endif -%}
41
+ {{ '<end_of_turn>
42
+ ' }}
43
+ {%- endfor -%}
44
+ {%- if add_generation_prompt -%}
45
+ {{'<start_of_turn>model
46
+ '}}
47
+ {%- endif -%}
checkpoint-500/config.json ADDED
@@ -0,0 +1,110 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Gemma3ForConditionalGeneration"
4
+ ],
5
+ "boi_token_index": 255999,
6
+ "bos_token_id": 2,
7
+ "dtype": "bfloat16",
8
+ "eoi_token_index": 256000,
9
+ "eos_token_id": 1,
10
+ "image_token_index": 262144,
11
+ "initializer_range": 0.02,
12
+ "mm_tokens_per_image": 256,
13
+ "model_type": "gemma3",
14
+ "pad_token_id": 0,
15
+ "text_config": {
16
+ "_sliding_window_pattern": 6,
17
+ "attention_bias": false,
18
+ "attention_dropout": 0.0,
19
+ "attn_logit_softcapping": null,
20
+ "dtype": "bfloat16",
21
+ "final_logit_softcapping": null,
22
+ "head_dim": 256,
23
+ "hidden_activation": "gelu_pytorch_tanh",
24
+ "hidden_size": 3840,
25
+ "initializer_range": 0.02,
26
+ "intermediate_size": 15360,
27
+ "layer_types": [
28
+ "sliding_attention",
29
+ "sliding_attention",
30
+ "sliding_attention",
31
+ "sliding_attention",
32
+ "sliding_attention",
33
+ "full_attention",
34
+ "sliding_attention",
35
+ "sliding_attention",
36
+ "sliding_attention",
37
+ "sliding_attention",
38
+ "sliding_attention",
39
+ "full_attention",
40
+ "sliding_attention",
41
+ "sliding_attention",
42
+ "sliding_attention",
43
+ "sliding_attention",
44
+ "sliding_attention",
45
+ "full_attention",
46
+ "sliding_attention",
47
+ "sliding_attention",
48
+ "sliding_attention",
49
+ "sliding_attention",
50
+ "sliding_attention",
51
+ "full_attention",
52
+ "sliding_attention",
53
+ "sliding_attention",
54
+ "sliding_attention",
55
+ "sliding_attention",
56
+ "sliding_attention",
57
+ "full_attention",
58
+ "sliding_attention",
59
+ "sliding_attention",
60
+ "sliding_attention",
61
+ "sliding_attention",
62
+ "sliding_attention",
63
+ "full_attention",
64
+ "sliding_attention",
65
+ "sliding_attention",
66
+ "sliding_attention",
67
+ "sliding_attention",
68
+ "sliding_attention",
69
+ "full_attention",
70
+ "sliding_attention",
71
+ "sliding_attention",
72
+ "sliding_attention",
73
+ "sliding_attention",
74
+ "sliding_attention",
75
+ "full_attention"
76
+ ],
77
+ "max_position_embeddings": 131072,
78
+ "model_type": "gemma3_text",
79
+ "num_attention_heads": 16,
80
+ "num_hidden_layers": 48,
81
+ "num_key_value_heads": 8,
82
+ "query_pre_attn_scalar": 256,
83
+ "rms_norm_eps": 1e-06,
84
+ "rope_local_base_freq": 10000.0,
85
+ "rope_scaling": {
86
+ "factor": 8.0,
87
+ "rope_type": "linear"
88
+ },
89
+ "rope_theta": 1000000.0,
90
+ "sliding_window": 1024,
91
+ "use_cache": true,
92
+ "vocab_size": 262208
93
+ },
94
+ "transformers_version": "4.56.1",
95
+ "vision_config": {
96
+ "attention_dropout": 0.0,
97
+ "dtype": "bfloat16",
98
+ "hidden_act": "gelu_pytorch_tanh",
99
+ "hidden_size": 1152,
100
+ "image_size": 896,
101
+ "intermediate_size": 4304,
102
+ "layer_norm_eps": 1e-06,
103
+ "model_type": "siglip_vision_model",
104
+ "num_attention_heads": 16,
105
+ "num_channels": 3,
106
+ "num_hidden_layers": 27,
107
+ "patch_size": 14,
108
+ "vision_use_head": false
109
+ }
110
+ }
checkpoint-500/generation_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 2,
3
+ "do_sample": true,
4
+ "eos_token_id": [
5
+ 1,
6
+ 1,
7
+ 106
8
+ ],
9
+ "pad_token_id": 0,
10
+ "top_k": 64,
11
+ "top_p": 0.95,
12
+ "transformers_version": "4.56.1"
13
+ }
checkpoint-500/model-00001-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4847447e92599833e8dbaa3067cd201c3bb5c052efa91f11ba891e43234f7832
3
+ size 4979902192
checkpoint-500/model-00002-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:891bd54eed03cba9ee1e705533a02a8217fcc29f356e4a1f53e5fd0d178883ad
3
+ size 4931296592
checkpoint-500/model-00003-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:074382738827a006415efff2c0d61947f99e4e7b8deef5d3c26cefe28d843b65
3
+ size 4931296656
checkpoint-500/model-00004-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:af4207024bd7a5520e97ac6af7044291b9430a3e0a294074d4b16dfa15818265
3
+ size 4931296656
checkpoint-500/model-00005-of-00005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2fb6d7c3994507491766753d3b72d7a3134956636e4f094a7e6d2dc63602c32a
3
+ size 4601000928
checkpoint-500/model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b2f229cf3ea175e792e32523432998e0996767af1069f2155f5f590b3cb176b1
3
+ size 14573116445
checkpoint-500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:69b6509f8bbdd943c9f8984cdc19d5d17ca63b6e47cd6e28080ebf502da06ddc
3
+ size 14645
checkpoint-500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:63481e620bd287eefb582d20daf306c54c7268088fea6f19af8100945c7e3e7e
3
+ size 1465
checkpoint-500/special_tokens_map.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "boi_token": "<start_of_image>",
3
+ "bos_token": {
4
+ "content": "<bos>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false
9
+ },
10
+ "eoi_token": "<end_of_image>",
11
+ "eos_token": {
12
+ "content": "<eos>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false
17
+ },
18
+ "image_token": "<image_soft_token>",
19
+ "pad_token": {
20
+ "content": "<pad>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false
25
+ },
26
+ "unk_token": {
27
+ "content": "<unk>",
28
+ "lstrip": false,
29
+ "normalized": false,
30
+ "rstrip": false,
31
+ "single_word": false
32
+ }
33
+ }
checkpoint-500/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4667f2089529e8e7657cfb6d1c19910ae71ff5f28aa7ab2ff2763330affad795
3
+ size 33384568
checkpoint-500/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1299c11d7cf632ef3b4e11937501358ada021bbdf7c47638d13c0ee982f2e79c
3
+ size 4689074
checkpoint-500/tokenizer_config.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-500/trainer_state.json ADDED
@@ -0,0 +1,644 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.2,
6
+ "eval_steps": 100,
7
+ "global_step": 500,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "entropy": 0.4866574574261904,
14
+ "epoch": 0.004,
15
+ "grad_norm": 616.0,
16
+ "learning_rate": 1.8e-07,
17
+ "loss": 33.2311,
18
+ "mean_token_accuracy": 0.6604987328872085,
19
+ "num_tokens": 51831.0,
20
+ "step": 10
21
+ },
22
+ {
23
+ "entropy": 0.4859781404957175,
24
+ "epoch": 0.008,
25
+ "grad_norm": 676.0,
26
+ "learning_rate": 3.8e-07,
27
+ "loss": 33.1415,
28
+ "mean_token_accuracy": 0.6665693156421184,
29
+ "num_tokens": 103172.0,
30
+ "step": 20
31
+ },
32
+ {
33
+ "entropy": 0.5029121663421392,
34
+ "epoch": 0.012,
35
+ "grad_norm": 824.0,
36
+ "learning_rate": 5.800000000000001e-07,
37
+ "loss": 34.7849,
38
+ "mean_token_accuracy": 0.6541365889832378,
39
+ "num_tokens": 147915.0,
40
+ "step": 30
41
+ },
42
+ {
43
+ "entropy": 0.5095210419967771,
44
+ "epoch": 0.016,
45
+ "grad_norm": 588.0,
46
+ "learning_rate": 7.8e-07,
47
+ "loss": 33.2642,
48
+ "mean_token_accuracy": 0.6610883103683591,
49
+ "num_tokens": 195524.0,
50
+ "step": 40
51
+ },
52
+ {
53
+ "entropy": 0.5093817379325628,
54
+ "epoch": 0.02,
55
+ "grad_norm": 544.0,
56
+ "learning_rate": 9.800000000000001e-07,
57
+ "loss": 32.1157,
58
+ "mean_token_accuracy": 0.6604763394221663,
59
+ "num_tokens": 243747.0,
60
+ "step": 50
61
+ },
62
+ {
63
+ "entropy": 0.5527876092121005,
64
+ "epoch": 0.024,
65
+ "grad_norm": 628.0,
66
+ "learning_rate": 1.1800000000000001e-06,
67
+ "loss": 31.9125,
68
+ "mean_token_accuracy": 0.6578346157446504,
69
+ "num_tokens": 296999.0,
70
+ "step": 60
71
+ },
72
+ {
73
+ "entropy": 0.5809885170310736,
74
+ "epoch": 0.028,
75
+ "grad_norm": 498.0,
76
+ "learning_rate": 1.3800000000000001e-06,
77
+ "loss": 32.3494,
78
+ "mean_token_accuracy": 0.6499254325404763,
79
+ "num_tokens": 342114.0,
80
+ "step": 70
81
+ },
82
+ {
83
+ "entropy": 0.6243463078513741,
84
+ "epoch": 0.032,
85
+ "grad_norm": 430.0,
86
+ "learning_rate": 1.5800000000000001e-06,
87
+ "loss": 29.4405,
88
+ "mean_token_accuracy": 0.6568279277533293,
89
+ "num_tokens": 390903.0,
90
+ "step": 80
91
+ },
92
+ {
93
+ "entropy": 0.7070688692852855,
94
+ "epoch": 0.036,
95
+ "grad_norm": 422.0,
96
+ "learning_rate": 1.7800000000000001e-06,
97
+ "loss": 27.9455,
98
+ "mean_token_accuracy": 0.6594765370711684,
99
+ "num_tokens": 440359.0,
100
+ "step": 90
101
+ },
102
+ {
103
+ "entropy": 0.7858394492417574,
104
+ "epoch": 0.04,
105
+ "grad_norm": 334.0,
106
+ "learning_rate": 1.98e-06,
107
+ "loss": 27.3289,
108
+ "mean_token_accuracy": 0.6568881323561072,
109
+ "num_tokens": 488481.0,
110
+ "step": 100
111
+ },
112
+ {
113
+ "epoch": 0.04,
114
+ "eval_biology_entropy": 0.8150426388680935,
115
+ "eval_biology_loss": 1.601250410079956,
116
+ "eval_biology_mean_token_accuracy": 0.6667410479187965,
117
+ "eval_biology_num_tokens": 488481.0,
118
+ "eval_biology_runtime": 34.4305,
119
+ "eval_biology_samples_per_second": 14.522,
120
+ "eval_biology_steps_per_second": 14.522,
121
+ "step": 100
122
+ },
123
+ {
124
+ "epoch": 0.04,
125
+ "eval_chemistry_entropy": 0.42218632949888707,
126
+ "eval_chemistry_loss": 0.7587577700614929,
127
+ "eval_chemistry_mean_token_accuracy": 0.8110456941127777,
128
+ "eval_chemistry_num_tokens": 488481.0,
129
+ "eval_chemistry_runtime": 39.1702,
130
+ "eval_chemistry_samples_per_second": 12.765,
131
+ "eval_chemistry_steps_per_second": 12.765,
132
+ "step": 100
133
+ },
134
+ {
135
+ "entropy": 0.8809404099360109,
136
+ "epoch": 0.044,
137
+ "grad_norm": 246.0,
138
+ "learning_rate": 2.1800000000000003e-06,
139
+ "loss": 25.34,
140
+ "mean_token_accuracy": 0.6623364068567753,
141
+ "num_tokens": 537471.0,
142
+ "step": 110
143
+ },
144
+ {
145
+ "entropy": 0.9853538589552044,
146
+ "epoch": 0.048,
147
+ "grad_norm": 228.0,
148
+ "learning_rate": 2.38e-06,
149
+ "loss": 23.768,
150
+ "mean_token_accuracy": 0.6667981075122953,
151
+ "num_tokens": 589868.0,
152
+ "step": 120
153
+ },
154
+ {
155
+ "entropy": 1.1170461051166058,
156
+ "epoch": 0.052,
157
+ "grad_norm": 171.0,
158
+ "learning_rate": 2.5800000000000003e-06,
159
+ "loss": 23.5554,
160
+ "mean_token_accuracy": 0.6613909218460321,
161
+ "num_tokens": 639512.0,
162
+ "step": 130
163
+ },
164
+ {
165
+ "entropy": 1.1676542518660427,
166
+ "epoch": 0.056,
167
+ "grad_norm": 151.0,
168
+ "learning_rate": 2.7800000000000005e-06,
169
+ "loss": 22.6768,
170
+ "mean_token_accuracy": 0.6657767463475466,
171
+ "num_tokens": 690754.0,
172
+ "step": 140
173
+ },
174
+ {
175
+ "entropy": 1.2725389588624239,
176
+ "epoch": 0.06,
177
+ "grad_norm": 132.0,
178
+ "learning_rate": 2.9800000000000003e-06,
179
+ "loss": 23.1722,
180
+ "mean_token_accuracy": 0.6627637758851052,
181
+ "num_tokens": 737075.0,
182
+ "step": 150
183
+ },
184
+ {
185
+ "entropy": 1.258068286627531,
186
+ "epoch": 0.064,
187
+ "grad_norm": 142.0,
188
+ "learning_rate": 3.1800000000000005e-06,
189
+ "loss": 21.9045,
190
+ "mean_token_accuracy": 0.6725248593837023,
191
+ "num_tokens": 791190.0,
192
+ "step": 160
193
+ },
194
+ {
195
+ "entropy": 1.306521786376834,
196
+ "epoch": 0.068,
197
+ "grad_norm": 112.0,
198
+ "learning_rate": 3.3800000000000007e-06,
199
+ "loss": 22.1444,
200
+ "mean_token_accuracy": 0.6677160531282424,
201
+ "num_tokens": 841344.0,
202
+ "step": 170
203
+ },
204
+ {
205
+ "entropy": 1.2990117866545916,
206
+ "epoch": 0.072,
207
+ "grad_norm": 141.0,
208
+ "learning_rate": 3.58e-06,
209
+ "loss": 22.1479,
210
+ "mean_token_accuracy": 0.6692137856036424,
211
+ "num_tokens": 890075.0,
212
+ "step": 180
213
+ },
214
+ {
215
+ "entropy": 1.2258288636803627,
216
+ "epoch": 0.076,
217
+ "grad_norm": 99.0,
218
+ "learning_rate": 3.7800000000000002e-06,
219
+ "loss": 20.8997,
220
+ "mean_token_accuracy": 0.6792997987940907,
221
+ "num_tokens": 940599.0,
222
+ "step": 190
223
+ },
224
+ {
225
+ "entropy": 1.2281328312121331,
226
+ "epoch": 0.08,
227
+ "grad_norm": 123.0,
228
+ "learning_rate": 3.980000000000001e-06,
229
+ "loss": 20.9624,
230
+ "mean_token_accuracy": 0.6809809722006321,
231
+ "num_tokens": 989860.0,
232
+ "step": 200
233
+ },
234
+ {
235
+ "epoch": 0.08,
236
+ "eval_biology_entropy": 1.2447693885564803,
237
+ "eval_biology_loss": 1.31557035446167,
238
+ "eval_biology_mean_token_accuracy": 0.6810677705407142,
239
+ "eval_biology_num_tokens": 989860.0,
240
+ "eval_biology_runtime": 34.2892,
241
+ "eval_biology_samples_per_second": 14.582,
242
+ "eval_biology_steps_per_second": 14.582,
243
+ "step": 200
244
+ },
245
+ {
246
+ "epoch": 0.08,
247
+ "eval_chemistry_entropy": 0.6061330872625112,
248
+ "eval_chemistry_loss": 0.6521626710891724,
249
+ "eval_chemistry_mean_token_accuracy": 0.8165175079107284,
250
+ "eval_chemistry_num_tokens": 989860.0,
251
+ "eval_chemistry_runtime": 39.0885,
252
+ "eval_chemistry_samples_per_second": 12.791,
253
+ "eval_chemistry_steps_per_second": 12.791,
254
+ "step": 200
255
+ },
256
+ {
257
+ "entropy": 1.2204063970595598,
258
+ "epoch": 0.084,
259
+ "grad_norm": 99.5,
260
+ "learning_rate": 4.18e-06,
261
+ "loss": 20.4735,
262
+ "mean_token_accuracy": 0.688131432235241,
263
+ "num_tokens": 1041199.0,
264
+ "step": 210
265
+ },
266
+ {
267
+ "entropy": 1.3349104586988687,
268
+ "epoch": 0.088,
269
+ "grad_norm": 111.5,
270
+ "learning_rate": 4.38e-06,
271
+ "loss": 22.5142,
272
+ "mean_token_accuracy": 0.6627503650262951,
273
+ "num_tokens": 1087099.0,
274
+ "step": 220
275
+ },
276
+ {
277
+ "entropy": 1.2557024333626032,
278
+ "epoch": 0.092,
279
+ "grad_norm": 96.5,
280
+ "learning_rate": 4.58e-06,
281
+ "loss": 20.9139,
282
+ "mean_token_accuracy": 0.6778965264558792,
283
+ "num_tokens": 1135090.0,
284
+ "step": 230
285
+ },
286
+ {
287
+ "entropy": 1.280203009955585,
288
+ "epoch": 0.096,
289
+ "grad_norm": 105.0,
290
+ "learning_rate": 4.78e-06,
291
+ "loss": 21.3579,
292
+ "mean_token_accuracy": 0.6716440364718437,
293
+ "num_tokens": 1181777.0,
294
+ "step": 240
295
+ },
296
+ {
297
+ "entropy": 1.276487385854125,
298
+ "epoch": 0.1,
299
+ "grad_norm": 100.5,
300
+ "learning_rate": 4.980000000000001e-06,
301
+ "loss": 20.8968,
302
+ "mean_token_accuracy": 0.6772379480302334,
303
+ "num_tokens": 1229242.0,
304
+ "step": 250
305
+ },
306
+ {
307
+ "entropy": 1.2585501682013274,
308
+ "epoch": 0.104,
309
+ "grad_norm": 85.0,
310
+ "learning_rate": 5.18e-06,
311
+ "loss": 20.7316,
312
+ "mean_token_accuracy": 0.679420518502593,
313
+ "num_tokens": 1283881.0,
314
+ "step": 260
315
+ },
316
+ {
317
+ "entropy": 1.248391578346491,
318
+ "epoch": 0.108,
319
+ "grad_norm": 91.0,
320
+ "learning_rate": 5.380000000000001e-06,
321
+ "loss": 20.4527,
322
+ "mean_token_accuracy": 0.682051457837224,
323
+ "num_tokens": 1332507.0,
324
+ "step": 270
325
+ },
326
+ {
327
+ "entropy": 1.2360890079289675,
328
+ "epoch": 0.112,
329
+ "grad_norm": 84.5,
330
+ "learning_rate": 5.580000000000001e-06,
331
+ "loss": 20.2916,
332
+ "mean_token_accuracy": 0.6859552793204784,
333
+ "num_tokens": 1385513.0,
334
+ "step": 280
335
+ },
336
+ {
337
+ "entropy": 1.2920903146266938,
338
+ "epoch": 0.116,
339
+ "grad_norm": 90.5,
340
+ "learning_rate": 5.78e-06,
341
+ "loss": 20.9268,
342
+ "mean_token_accuracy": 0.673469840362668,
343
+ "num_tokens": 1432251.0,
344
+ "step": 290
345
+ },
346
+ {
347
+ "entropy": 1.2169338904321194,
348
+ "epoch": 0.12,
349
+ "grad_norm": 87.0,
350
+ "learning_rate": 5.98e-06,
351
+ "loss": 19.6752,
352
+ "mean_token_accuracy": 0.6875041898339986,
353
+ "num_tokens": 1483667.0,
354
+ "step": 300
355
+ },
356
+ {
357
+ "epoch": 0.12,
358
+ "eval_biology_entropy": 1.2245318589806558,
359
+ "eval_biology_loss": 1.2560091018676758,
360
+ "eval_biology_mean_token_accuracy": 0.6886654210686683,
361
+ "eval_biology_num_tokens": 1483667.0,
362
+ "eval_biology_runtime": 34.2037,
363
+ "eval_biology_samples_per_second": 14.618,
364
+ "eval_biology_steps_per_second": 14.618,
365
+ "step": 300
366
+ },
367
+ {
368
+ "epoch": 0.12,
369
+ "eval_chemistry_entropy": 0.6325840967297554,
370
+ "eval_chemistry_loss": 0.6409750580787659,
371
+ "eval_chemistry_mean_token_accuracy": 0.8179453414678574,
372
+ "eval_chemistry_num_tokens": 1483667.0,
373
+ "eval_chemistry_runtime": 39.0169,
374
+ "eval_chemistry_samples_per_second": 12.815,
375
+ "eval_chemistry_steps_per_second": 12.815,
376
+ "step": 300
377
+ },
378
+ {
379
+ "entropy": 1.2429257260635496,
380
+ "epoch": 0.124,
381
+ "grad_norm": 79.5,
382
+ "learning_rate": 6.18e-06,
383
+ "loss": 20.6172,
384
+ "mean_token_accuracy": 0.6801770402118563,
385
+ "num_tokens": 1534286.0,
386
+ "step": 310
387
+ },
388
+ {
389
+ "entropy": 1.2787830816581844,
390
+ "epoch": 0.128,
391
+ "grad_norm": 84.5,
392
+ "learning_rate": 6.380000000000001e-06,
393
+ "loss": 20.5543,
394
+ "mean_token_accuracy": 0.6807437939569354,
395
+ "num_tokens": 1582368.0,
396
+ "step": 320
397
+ },
398
+ {
399
+ "entropy": 1.2588950738310813,
400
+ "epoch": 0.132,
401
+ "grad_norm": 89.0,
402
+ "learning_rate": 6.5800000000000005e-06,
403
+ "loss": 20.3303,
404
+ "mean_token_accuracy": 0.6814694199711084,
405
+ "num_tokens": 1630878.0,
406
+ "step": 330
407
+ },
408
+ {
409
+ "entropy": 1.2704457983374595,
410
+ "epoch": 0.136,
411
+ "grad_norm": 83.0,
412
+ "learning_rate": 6.780000000000001e-06,
413
+ "loss": 20.6418,
414
+ "mean_token_accuracy": 0.6803831990808249,
415
+ "num_tokens": 1677060.0,
416
+ "step": 340
417
+ },
418
+ {
419
+ "entropy": 1.2321893341839314,
420
+ "epoch": 0.14,
421
+ "grad_norm": 87.5,
422
+ "learning_rate": 6.98e-06,
423
+ "loss": 19.8202,
424
+ "mean_token_accuracy": 0.6891451958566904,
425
+ "num_tokens": 1725186.0,
426
+ "step": 350
427
+ },
428
+ {
429
+ "entropy": 1.2445756699889898,
430
+ "epoch": 0.144,
431
+ "grad_norm": 92.0,
432
+ "learning_rate": 7.180000000000001e-06,
433
+ "loss": 20.2527,
434
+ "mean_token_accuracy": 0.6785595057532191,
435
+ "num_tokens": 1773764.0,
436
+ "step": 360
437
+ },
438
+ {
439
+ "entropy": 1.246081338264048,
440
+ "epoch": 0.148,
441
+ "grad_norm": 82.0,
442
+ "learning_rate": 7.3800000000000005e-06,
443
+ "loss": 20.08,
444
+ "mean_token_accuracy": 0.6857875902205706,
445
+ "num_tokens": 1823706.0,
446
+ "step": 370
447
+ },
448
+ {
449
+ "entropy": 1.2665387976914644,
450
+ "epoch": 0.152,
451
+ "grad_norm": 95.0,
452
+ "learning_rate": 7.58e-06,
453
+ "loss": 20.6091,
454
+ "mean_token_accuracy": 0.6813197182491422,
455
+ "num_tokens": 1873167.0,
456
+ "step": 380
457
+ },
458
+ {
459
+ "entropy": 1.2548780847340821,
460
+ "epoch": 0.156,
461
+ "grad_norm": 85.0,
462
+ "learning_rate": 7.78e-06,
463
+ "loss": 20.406,
464
+ "mean_token_accuracy": 0.6822501845657826,
465
+ "num_tokens": 1921484.0,
466
+ "step": 390
467
+ },
468
+ {
469
+ "entropy": 1.2407029550522566,
470
+ "epoch": 0.16,
471
+ "grad_norm": 89.0,
472
+ "learning_rate": 7.980000000000002e-06,
473
+ "loss": 19.8013,
474
+ "mean_token_accuracy": 0.6853796949610114,
475
+ "num_tokens": 1970077.0,
476
+ "step": 400
477
+ },
478
+ {
479
+ "epoch": 0.16,
480
+ "eval_biology_entropy": 1.214753457427025,
481
+ "eval_biology_loss": 1.2285655736923218,
482
+ "eval_biology_mean_token_accuracy": 0.6923919150233269,
483
+ "eval_biology_num_tokens": 1970077.0,
484
+ "eval_biology_runtime": 34.3823,
485
+ "eval_biology_samples_per_second": 14.542,
486
+ "eval_biology_steps_per_second": 14.542,
487
+ "step": 400
488
+ },
489
+ {
490
+ "epoch": 0.16,
491
+ "eval_chemistry_entropy": 0.6530047204196453,
492
+ "eval_chemistry_loss": 0.644757866859436,
493
+ "eval_chemistry_mean_token_accuracy": 0.8174828389883041,
494
+ "eval_chemistry_num_tokens": 1970077.0,
495
+ "eval_chemistry_runtime": 39.0828,
496
+ "eval_chemistry_samples_per_second": 12.793,
497
+ "eval_chemistry_steps_per_second": 12.793,
498
+ "step": 400
499
+ },
500
+ {
501
+ "entropy": 1.2693706121295691,
502
+ "epoch": 0.164,
503
+ "grad_norm": 75.0,
504
+ "learning_rate": 8.18e-06,
505
+ "loss": 20.5934,
506
+ "mean_token_accuracy": 0.6737196171656251,
507
+ "num_tokens": 2019053.0,
508
+ "step": 410
509
+ },
510
+ {
511
+ "entropy": 1.2044812317937612,
512
+ "epoch": 0.168,
513
+ "grad_norm": 72.5,
514
+ "learning_rate": 8.380000000000001e-06,
515
+ "loss": 19.2447,
516
+ "mean_token_accuracy": 0.6942890204489232,
517
+ "num_tokens": 2066815.0,
518
+ "step": 420
519
+ },
520
+ {
521
+ "entropy": 1.219953553378582,
522
+ "epoch": 0.172,
523
+ "grad_norm": 76.5,
524
+ "learning_rate": 8.580000000000001e-06,
525
+ "loss": 20.0236,
526
+ "mean_token_accuracy": 0.6850357944145798,
527
+ "num_tokens": 2116402.0,
528
+ "step": 430
529
+ },
530
+ {
531
+ "entropy": 1.1823513347655534,
532
+ "epoch": 0.176,
533
+ "grad_norm": 74.0,
534
+ "learning_rate": 8.78e-06,
535
+ "loss": 19.1286,
536
+ "mean_token_accuracy": 0.6948613569140434,
537
+ "num_tokens": 2168354.0,
538
+ "step": 440
539
+ },
540
+ {
541
+ "entropy": 1.2351123623549938,
542
+ "epoch": 0.18,
543
+ "grad_norm": 75.0,
544
+ "learning_rate": 8.98e-06,
545
+ "loss": 19.847,
546
+ "mean_token_accuracy": 0.6850900270044804,
547
+ "num_tokens": 2219158.0,
548
+ "step": 450
549
+ },
550
+ {
551
+ "entropy": 1.2182675819844007,
552
+ "epoch": 0.184,
553
+ "grad_norm": 83.0,
554
+ "learning_rate": 9.180000000000002e-06,
555
+ "loss": 19.9054,
556
+ "mean_token_accuracy": 0.6874489855021239,
557
+ "num_tokens": 2270617.0,
558
+ "step": 460
559
+ },
560
+ {
561
+ "entropy": 1.2002981722354888,
562
+ "epoch": 0.188,
563
+ "grad_norm": 74.5,
564
+ "learning_rate": 9.38e-06,
565
+ "loss": 18.9485,
566
+ "mean_token_accuracy": 0.6958082247525453,
567
+ "num_tokens": 2322034.0,
568
+ "step": 470
569
+ },
570
+ {
571
+ "entropy": 1.2613376142457127,
572
+ "epoch": 0.192,
573
+ "grad_norm": 98.5,
574
+ "learning_rate": 9.58e-06,
575
+ "loss": 20.7735,
576
+ "mean_token_accuracy": 0.6739435123279691,
577
+ "num_tokens": 2365822.0,
578
+ "step": 480
579
+ },
580
+ {
581
+ "entropy": 1.205817231722176,
582
+ "epoch": 0.196,
583
+ "grad_norm": 89.5,
584
+ "learning_rate": 9.780000000000001e-06,
585
+ "loss": 19.1434,
586
+ "mean_token_accuracy": 0.6935921791940928,
587
+ "num_tokens": 2413853.0,
588
+ "step": 490
589
+ },
590
+ {
591
+ "entropy": 1.2515719644725323,
592
+ "epoch": 0.2,
593
+ "grad_norm": 90.5,
594
+ "learning_rate": 9.980000000000001e-06,
595
+ "loss": 20.4938,
596
+ "mean_token_accuracy": 0.6777064654976129,
597
+ "num_tokens": 2459876.0,
598
+ "step": 500
599
+ },
600
+ {
601
+ "epoch": 0.2,
602
+ "eval_biology_entropy": 1.1872086516022682,
603
+ "eval_biology_loss": 1.2085970640182495,
604
+ "eval_biology_mean_token_accuracy": 0.6949169620871544,
605
+ "eval_biology_num_tokens": 2459876.0,
606
+ "eval_biology_runtime": 34.4,
607
+ "eval_biology_samples_per_second": 14.535,
608
+ "eval_biology_steps_per_second": 14.535,
609
+ "step": 500
610
+ },
611
+ {
612
+ "epoch": 0.2,
613
+ "eval_chemistry_entropy": 0.6488757311999798,
614
+ "eval_chemistry_loss": 0.6450773477554321,
615
+ "eval_chemistry_mean_token_accuracy": 0.8178821201324463,
616
+ "eval_chemistry_num_tokens": 2459876.0,
617
+ "eval_chemistry_runtime": 39.2328,
618
+ "eval_chemistry_samples_per_second": 12.744,
619
+ "eval_chemistry_steps_per_second": 12.744,
620
+ "step": 500
621
+ }
622
+ ],
623
+ "logging_steps": 10,
624
+ "max_steps": 10000,
625
+ "num_input_tokens_seen": 0,
626
+ "num_train_epochs": 4,
627
+ "save_steps": 500,
628
+ "stateful_callbacks": {
629
+ "TrainerControl": {
630
+ "args": {
631
+ "should_epoch_stop": false,
632
+ "should_evaluate": false,
633
+ "should_log": false,
634
+ "should_save": true,
635
+ "should_training_stop": false
636
+ },
637
+ "attributes": {}
638
+ }
639
+ },
640
+ "total_flos": 1.6494542652385037e+17,
641
+ "train_batch_size": 1,
642
+ "trial_name": null,
643
+ "trial_params": null
644
+ }
checkpoint-500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:80b5712b7c29bfc74fdd744346ebce13b8126ec518d448ca496e5a80dda84b00
3
+ size 6481