Hugging Face's logo

arunasank
/
ofkphp8d

Safetensors
Model card Files
xet
 Community
arunasank commited on
Commit
8c74892
·
verified ·
1 Parent(s): 319e4fa

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 +754 -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 sha256:7cee411d9d57324e50ce064a192cc5a858276d508611b12fc599e0c9767112e0
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 sha256:7cbc1411b64106f76b319c1751d8378931bb7d9e6f5850ad1e04995fa94ec83c
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:bb4bb8cfa128e7c600b71a13673b29164810270a00ff65dcad49e04065553af4
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 sha256:606f707044a1ed038e8e4bce1f86af3ae23112ac82d8cb349e6072d527f3b65b
3
+ size 1793214689
checkpoint-500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:acd74ad34e3c5060506b53a8a3fafbfb35d44065ee496b2abf3a0b9a3b93ba09
3
+ size 14645
checkpoint-500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9527d8453c68ef03b0e54ce001d2d4068a1f05836577d59b7d4e4a5a8aa7f0a8
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 +1,754 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.8,
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": 1.000200616195798,
14
+ "epoch": 0.016,
15
+ "grad_norm": 1776.0,
16
+ "learning_rate": 4.5000000000000003e-07,
17
+ "loss": 69.8616,
18
+ "mean_token_accuracy": 0.460323916003108,
19
+ "num_tokens": 195524.0,
20
+ "step": 10
21
+ },
22
+ {
23
+ "entropy": 0.9949515227228403,
24
+ "epoch": 0.032,
25
+ "grad_norm": 1624.0,
26
+ "learning_rate": 9.500000000000001e-07,
27
+ "loss": 68.3351,
28
+ "mean_token_accuracy": 0.4633398361504078,
29
+ "num_tokens": 390903.0,
30
+ "step": 20
31
+ },
32
+ {
33
+ "entropy": 1.0845801826566457,
34
+ "epoch": 0.048,
35
+ "grad_norm": 1728.0,
36
+ "learning_rate": 1.45e-06,
37
+ "loss": 63.0644,
38
+ "mean_token_accuracy": 0.46960638258606197,
39
+ "num_tokens": 589868.0,
40
+ "step": 30
41
+ },
42
+ {
43
+ "entropy": 1.247272988036275,
44
+ "epoch": 0.064,
45
+ "grad_norm": 488.0,
46
+ "learning_rate": 1.9500000000000004e-06,
47
+ "loss": 55.5844,
48
+ "mean_token_accuracy": 0.480246701836586,
49
+ "num_tokens": 791190.0,
50
+ "step": 40
51
+ },
52
+ {
53
+ "entropy": 1.4826370410621166,
54
+ "epoch": 0.08,
55
+ "grad_norm": 288.0,
56
+ "learning_rate": 2.4500000000000003e-06,
57
+ "loss": 50.1588,
58
+ "mean_token_accuracy": 0.48712102621793746,
59
+ "num_tokens": 989860.0,
60
+ "step": 50
61
+ },
62
+ {
63
+ "entropy": 1.814505684375763,
64
+ "epoch": 0.096,
65
+ "grad_norm": 131.0,
66
+ "learning_rate": 2.95e-06,
67
+ "loss": 45.3237,
68
+ "mean_token_accuracy": 0.48977465480566024,
69
+ "num_tokens": 1181777.0,
70
+ "step": 60
71
+ },
72
+ {
73
+ "entropy": 2.17904861420393,
74
+ "epoch": 0.112,
75
+ "grad_norm": 67.5,
76
+ "learning_rate": 3.45e-06,
77
+ "loss": 40.572,
78
+ "mean_token_accuracy": 0.5055354550480843,
79
+ "num_tokens": 1385513.0,
80
+ "step": 70
81
+ },
82
+ {
83
+ "entropy": 2.317786781489849,
84
+ "epoch": 0.128,
85
+ "grad_norm": 55.25,
86
+ "learning_rate": 3.95e-06,
87
+ "loss": 38.0783,
88
+ "mean_token_accuracy": 0.5195932997390628,
89
+ "num_tokens": 1582368.0,
90
+ "step": 80
91
+ },
92
+ {
93
+ "entropy": 2.2404811814427377,
94
+ "epoch": 0.144,
95
+ "grad_norm": 40.75,
96
+ "learning_rate": 4.450000000000001e-06,
97
+ "loss": 35.7283,
98
+ "mean_token_accuracy": 0.5399348795413971,
99
+ "num_tokens": 1773764.0,
100
+ "step": 90
101
+ },
102
+ {
103
+ "entropy": 2.1976587153971194,
104
+ "epoch": 0.16,
105
+ "grad_norm": 38.0,
106
+ "learning_rate": 4.95e-06,
107
+ "loss": 34.5297,
108
+ "mean_token_accuracy": 0.5478190431371331,
109
+ "num_tokens": 1970077.0,
110
+ "step": 100
111
+ },
112
+ {
113
+ "epoch": 0.16,
114
+ "eval_biology_entropy": 2.1061868534088135,
115
+ "eval_biology_loss": 2.0133121013641357,
116
+ "eval_biology_mean_token_accuracy": 0.5668603234291076,
117
+ "eval_biology_num_tokens": 1970077.0,
118
+ "eval_biology_runtime": 47.9413,
119
+ "eval_biology_samples_per_second": 10.429,
120
+ "eval_biology_steps_per_second": 2.607,
121
+ "step": 100
122
+ },
123
+ {
124
+ "epoch": 0.16,
125
+ "eval_chemistry_entropy": 1.8679178915023804,
126
+ "eval_chemistry_loss": 1.844948410987854,
127
+ "eval_chemistry_mean_token_accuracy": 0.5880863065719605,
128
+ "eval_chemistry_num_tokens": 1970077.0,
129
+ "eval_chemistry_runtime": 59.3684,
130
+ "eval_chemistry_samples_per_second": 8.422,
131
+ "eval_chemistry_steps_per_second": 2.105,
132
+ "step": 100
133
+ },
134
+ {
135
+ "epoch": 0.16,
136
+ "eval_math_entropy": 1.6290378885269166,
137
+ "eval_math_loss": 2.072597026824951,
138
+ "eval_math_mean_token_accuracy": 0.5653720505237579,
139
+ "eval_math_num_tokens": 1970077.0,
140
+ "eval_math_runtime": 60.948,
141
+ "eval_math_samples_per_second": 8.204,
142
+ "eval_math_steps_per_second": 2.051,
143
+ "step": 100
144
+ },
145
+ {
146
+ "epoch": 0.16,
147
+ "eval_physics_entropy": 1.907263669013977,
148
+ "eval_physics_loss": 2.105146884918213,
149
+ "eval_physics_mean_token_accuracy": 0.5574561676979065,
150
+ "eval_physics_num_tokens": 1970077.0,
151
+ "eval_physics_runtime": 69.3465,
152
+ "eval_physics_samples_per_second": 7.21,
153
+ "eval_physics_steps_per_second": 1.803,
154
+ "step": 100
155
+ },
156
+ {
157
+ "entropy": 2.0820947512984276,
158
+ "epoch": 0.176,
159
+ "grad_norm": 33.5,
160
+ "learning_rate": 5.450000000000001e-06,
161
+ "loss": 32.4406,
162
+ "mean_token_accuracy": 0.5621457317844033,
163
+ "num_tokens": 2168354.0,
164
+ "step": 110
165
+ },
166
+ {
167
+ "entropy": 2.0283117063343523,
168
+ "epoch": 0.192,
169
+ "grad_norm": 32.25,
170
+ "learning_rate": 5.950000000000001e-06,
171
+ "loss": 31.491,
172
+ "mean_token_accuracy": 0.5688053227961063,
173
+ "num_tokens": 2365822.0,
174
+ "step": 120
175
+ },
176
+ {
177
+ "entropy": 1.973121351003647,
178
+ "epoch": 0.208,
179
+ "grad_norm": 33.75,
180
+ "learning_rate": 6.450000000000001e-06,
181
+ "loss": 30.9217,
182
+ "mean_token_accuracy": 0.5753367917612195,
183
+ "num_tokens": 2558762.0,
184
+ "step": 130
185
+ },
186
+ {
187
+ "entropy": 1.9165778771042823,
188
+ "epoch": 0.224,
189
+ "grad_norm": 30.625,
190
+ "learning_rate": 6.95e-06,
191
+ "loss": 30.1267,
192
+ "mean_token_accuracy": 0.583262656442821,
193
+ "num_tokens": 2755347.0,
194
+ "step": 140
195
+ },
196
+ {
197
+ "entropy": 1.8870679274201394,
198
+ "epoch": 0.24,
199
+ "grad_norm": 31.5,
200
+ "learning_rate": 7.450000000000001e-06,
201
+ "loss": 29.6484,
202
+ "mean_token_accuracy": 0.5872904889285564,
203
+ "num_tokens": 2947346.0,
204
+ "step": 150
205
+ },
206
+ {
207
+ "entropy": 1.8710497491061688,
208
+ "epoch": 0.256,
209
+ "grad_norm": 29.875,
210
+ "learning_rate": 7.950000000000002e-06,
211
+ "loss": 29.4624,
212
+ "mean_token_accuracy": 0.5887285262346268,
213
+ "num_tokens": 3139957.0,
214
+ "step": 160
215
+ },
216
+ {
217
+ "entropy": 1.808907500654459,
218
+ "epoch": 0.272,
219
+ "grad_norm": 29.125,
220
+ "learning_rate": 8.45e-06,
221
+ "loss": 28.7075,
222
+ "mean_token_accuracy": 0.5965064790099859,
223
+ "num_tokens": 3335951.0,
224
+ "step": 170
225
+ },
226
+ {
227
+ "entropy": 1.774143099039793,
228
+ "epoch": 0.288,
229
+ "grad_norm": 28.125,
230
+ "learning_rate": 8.95e-06,
231
+ "loss": 28.1007,
232
+ "mean_token_accuracy": 0.5996631737798452,
233
+ "num_tokens": 3539731.0,
234
+ "step": 180
235
+ },
236
+ {
237
+ "entropy": 1.7740100547671318,
238
+ "epoch": 0.304,
239
+ "grad_norm": 27.75,
240
+ "learning_rate": 9.450000000000001e-06,
241
+ "loss": 28.1122,
242
+ "mean_token_accuracy": 0.6012205760926008,
243
+ "num_tokens": 3733488.0,
244
+ "step": 190
245
+ },
246
+ {
247
+ "entropy": 1.7725927896797657,
248
+ "epoch": 0.32,
249
+ "grad_norm": 35.25,
250
+ "learning_rate": 9.950000000000001e-06,
251
+ "loss": 28.1299,
252
+ "mean_token_accuracy": 0.5997324202209711,
253
+ "num_tokens": 3920545.0,
254
+ "step": 200
255
+ },
256
+ {
257
+ "epoch": 0.32,
258
+ "eval_biology_entropy": 1.6843824968338013,
259
+ "eval_biology_loss": 1.6854687929153442,
260
+ "eval_biology_mean_token_accuracy": 0.6137065634727478,
261
+ "eval_biology_num_tokens": 3920545.0,
262
+ "eval_biology_runtime": 47.8268,
263
+ "eval_biology_samples_per_second": 10.454,
264
+ "eval_biology_steps_per_second": 2.614,
265
+ "step": 200
266
+ },
267
+ {
268
+ "epoch": 0.32,
269
+ "eval_chemistry_entropy": 1.441379102230072,
270
+ "eval_chemistry_loss": 1.4537148475646973,
271
+ "eval_chemistry_mean_token_accuracy": 0.6551240344047546,
272
+ "eval_chemistry_num_tokens": 3920545.0,
273
+ "eval_chemistry_runtime": 59.3401,
274
+ "eval_chemistry_samples_per_second": 8.426,
275
+ "eval_chemistry_steps_per_second": 2.106,
276
+ "step": 200
277
+ },
278
+ {
279
+ "epoch": 0.32,
280
+ "eval_math_entropy": 1.3251727666854858,
281
+ "eval_math_loss": 1.6510798931121826,
282
+ "eval_math_mean_token_accuracy": 0.6343780555725098,
283
+ "eval_math_num_tokens": 3920545.0,
284
+ "eval_math_runtime": 60.8747,
285
+ "eval_math_samples_per_second": 8.214,
286
+ "eval_math_steps_per_second": 2.053,
287
+ "step": 200
288
+ },
289
+ {
290
+ "epoch": 0.32,
291
+ "eval_physics_entropy": 1.4914224824905395,
292
+ "eval_physics_loss": 1.6194021701812744,
293
+ "eval_physics_mean_token_accuracy": 0.6387534441947937,
294
+ "eval_physics_num_tokens": 3920545.0,
295
+ "eval_physics_runtime": 69.3484,
296
+ "eval_physics_samples_per_second": 7.21,
297
+ "eval_physics_steps_per_second": 1.802,
298
+ "step": 200
299
+ },
300
+ {
301
+ "entropy": 1.7085595309734345,
302
+ "epoch": 0.336,
303
+ "grad_norm": 30.375,
304
+ "learning_rate": 1.045e-05,
305
+ "loss": 27.2575,
306
+ "mean_token_accuracy": 0.6091335583478212,
307
+ "num_tokens": 4114077.0,
308
+ "step": 210
309
+ },
310
+ {
311
+ "entropy": 1.7301082991063594,
312
+ "epoch": 0.352,
313
+ "grad_norm": 31.0,
314
+ "learning_rate": 1.095e-05,
315
+ "loss": 27.6337,
316
+ "mean_token_accuracy": 0.6071538373827934,
317
+ "num_tokens": 4306949.0,
318
+ "step": 220
319
+ },
320
+ {
321
+ "entropy": 1.695884221047163,
322
+ "epoch": 0.368,
323
+ "grad_norm": 27.75,
324
+ "learning_rate": 1.145e-05,
325
+ "loss": 27.0338,
326
+ "mean_token_accuracy": 0.6128786243498325,
327
+ "num_tokens": 4504001.0,
328
+ "step": 230
329
+ },
330
+ {
331
+ "entropy": 1.6912743583321572,
332
+ "epoch": 0.384,
333
+ "grad_norm": 31.0,
334
+ "learning_rate": 1.195e-05,
335
+ "loss": 27.0143,
336
+ "mean_token_accuracy": 0.6135674986988306,
337
+ "num_tokens": 4693812.0,
338
+ "step": 240
339
+ },
340
+ {
341
+ "entropy": 1.6936707220971585,
342
+ "epoch": 0.4,
343
+ "grad_norm": 24.5,
344
+ "learning_rate": 1.2450000000000003e-05,
345
+ "loss": 26.8651,
346
+ "mean_token_accuracy": 0.6143408048897981,
347
+ "num_tokens": 4887094.0,
348
+ "step": 250
349
+ },
350
+ {
351
+ "entropy": 1.678075761348009,
352
+ "epoch": 0.416,
353
+ "grad_norm": 27.875,
354
+ "learning_rate": 1.295e-05,
355
+ "loss": 27.047,
356
+ "mean_token_accuracy": 0.6129864633083344,
357
+ "num_tokens": 5085369.0,
358
+ "step": 260
359
+ },
360
+ {
361
+ "entropy": 1.6657070152461528,
362
+ "epoch": 0.432,
363
+ "grad_norm": 29.75,
364
+ "learning_rate": 1.3450000000000002e-05,
365
+ "loss": 26.7844,
366
+ "mean_token_accuracy": 0.6148818587884307,
367
+ "num_tokens": 5271275.0,
368
+ "step": 270
369
+ },
370
+ {
371
+ "entropy": 1.6670012198388577,
372
+ "epoch": 0.448,
373
+ "grad_norm": 31.375,
374
+ "learning_rate": 1.3950000000000002e-05,
375
+ "loss": 26.617,
376
+ "mean_token_accuracy": 0.6159537024796009,
377
+ "num_tokens": 5460559.0,
378
+ "step": 280
379
+ },
380
+ {
381
+ "entropy": 1.6152651645243168,
382
+ "epoch": 0.464,
383
+ "grad_norm": 32.25,
384
+ "learning_rate": 1.4450000000000002e-05,
385
+ "loss": 25.6477,
386
+ "mean_token_accuracy": 0.6256863068789243,
387
+ "num_tokens": 5653809.0,
388
+ "step": 290
389
+ },
390
+ {
391
+ "entropy": 1.624342393130064,
392
+ "epoch": 0.48,
393
+ "grad_norm": 25.75,
394
+ "learning_rate": 1.4950000000000003e-05,
395
+ "loss": 26.2019,
396
+ "mean_token_accuracy": 0.6199633304029704,
397
+ "num_tokens": 5850176.0,
398
+ "step": 300
399
+ },
400
+ {
401
+ "epoch": 0.48,
402
+ "eval_biology_entropy": 1.5827035818099975,
403
+ "eval_biology_loss": 1.570924997329712,
404
+ "eval_biology_mean_token_accuracy": 0.6328578324317932,
405
+ "eval_biology_num_tokens": 5850176.0,
406
+ "eval_biology_runtime": 47.7211,
407
+ "eval_biology_samples_per_second": 10.478,
408
+ "eval_biology_steps_per_second": 2.619,
409
+ "step": 300
410
+ },
411
+ {
412
+ "epoch": 0.48,
413
+ "eval_chemistry_entropy": 1.3193063869476318,
414
+ "eval_chemistry_loss": 1.3243473768234253,
415
+ "eval_chemistry_mean_token_accuracy": 0.6796774597167968,
416
+ "eval_chemistry_num_tokens": 5850176.0,
417
+ "eval_chemistry_runtime": 59.1912,
418
+ "eval_chemistry_samples_per_second": 8.447,
419
+ "eval_chemistry_steps_per_second": 2.112,
420
+ "step": 300
421
+ },
422
+ {
423
+ "epoch": 0.48,
424
+ "eval_math_entropy": 1.2024567556381225,
425
+ "eval_math_loss": 1.5285311937332153,
426
+ "eval_math_mean_token_accuracy": 0.6541934962272644,
427
+ "eval_math_num_tokens": 5850176.0,
428
+ "eval_math_runtime": 60.6728,
429
+ "eval_math_samples_per_second": 8.241,
430
+ "eval_math_steps_per_second": 2.06,
431
+ "step": 300
432
+ },
433
+ {
434
+ "epoch": 0.48,
435
+ "eval_physics_entropy": 1.3696330366134644,
436
+ "eval_physics_loss": 1.4624738693237305,
437
+ "eval_physics_mean_token_accuracy": 0.6671717586517334,
438
+ "eval_physics_num_tokens": 5850176.0,
439
+ "eval_physics_runtime": 69.1435,
440
+ "eval_physics_samples_per_second": 7.231,
441
+ "eval_physics_steps_per_second": 1.808,
442
+ "step": 300
443
+ },
444
+ {
445
+ "entropy": 1.543294757604599,
446
+ "epoch": 0.496,
447
+ "grad_norm": 24.125,
448
+ "learning_rate": 1.545e-05,
449
+ "loss": 24.7301,
450
+ "mean_token_accuracy": 0.6344853390008212,
451
+ "num_tokens": 6046503.0,
452
+ "step": 310
453
+ },
454
+ {
455
+ "entropy": 1.6137450806796552,
456
+ "epoch": 0.512,
457
+ "grad_norm": 30.0,
458
+ "learning_rate": 1.595e-05,
459
+ "loss": 25.8005,
460
+ "mean_token_accuracy": 0.624148428812623,
461
+ "num_tokens": 6240456.0,
462
+ "step": 320
463
+ },
464
+ {
465
+ "entropy": 1.5901939049363136,
466
+ "epoch": 0.528,
467
+ "grad_norm": 26.75,
468
+ "learning_rate": 1.645e-05,
469
+ "loss": 25.5556,
470
+ "mean_token_accuracy": 0.6283756781369447,
471
+ "num_tokens": 6430555.0,
472
+ "step": 330
473
+ },
474
+ {
475
+ "entropy": 1.5919260658323764,
476
+ "epoch": 0.544,
477
+ "grad_norm": 25.375,
478
+ "learning_rate": 1.6950000000000002e-05,
479
+ "loss": 25.5888,
480
+ "mean_token_accuracy": 0.6276101280003786,
481
+ "num_tokens": 6626006.0,
482
+ "step": 340
483
+ },
484
+ {
485
+ "entropy": 1.5667011991143227,
486
+ "epoch": 0.56,
487
+ "grad_norm": 27.375,
488
+ "learning_rate": 1.7450000000000004e-05,
489
+ "loss": 25.1583,
490
+ "mean_token_accuracy": 0.6304543860256672,
491
+ "num_tokens": 6820754.0,
492
+ "step": 350
493
+ },
494
+ {
495
+ "entropy": 1.548937487602234,
496
+ "epoch": 0.576,
497
+ "grad_norm": 25.0,
498
+ "learning_rate": 1.795e-05,
499
+ "loss": 24.8095,
500
+ "mean_token_accuracy": 0.6340219389647246,
501
+ "num_tokens": 7021844.0,
502
+ "step": 360
503
+ },
504
+ {
505
+ "entropy": 1.5519795384258033,
506
+ "epoch": 0.592,
507
+ "grad_norm": 25.75,
508
+ "learning_rate": 1.845e-05,
509
+ "loss": 25.0794,
510
+ "mean_token_accuracy": 0.6325468828901648,
511
+ "num_tokens": 7213951.0,
512
+ "step": 370
513
+ },
514
+ {
515
+ "entropy": 1.5391070768237114,
516
+ "epoch": 0.608,
517
+ "grad_norm": 28.25,
518
+ "learning_rate": 1.8950000000000003e-05,
519
+ "loss": 24.6819,
520
+ "mean_token_accuracy": 0.636061006039381,
521
+ "num_tokens": 7416773.0,
522
+ "step": 380
523
+ },
524
+ {
525
+ "entropy": 1.5429081372916698,
526
+ "epoch": 0.624,
527
+ "grad_norm": 27.5,
528
+ "learning_rate": 1.9450000000000002e-05,
529
+ "loss": 24.8187,
530
+ "mean_token_accuracy": 0.634391113743186,
531
+ "num_tokens": 7612843.0,
532
+ "step": 390
533
+ },
534
+ {
535
+ "entropy": 1.5399384461343288,
536
+ "epoch": 0.64,
537
+ "grad_norm": 25.875,
538
+ "learning_rate": 1.9950000000000004e-05,
539
+ "loss": 24.8584,
540
+ "mean_token_accuracy": 0.6340381380170583,
541
+ "num_tokens": 7801633.0,
542
+ "step": 400
543
+ },
544
+ {
545
+ "epoch": 0.64,
546
+ "eval_biology_entropy": 1.4887315626144408,
547
+ "eval_biology_loss": 1.4859997034072876,
548
+ "eval_biology_mean_token_accuracy": 0.6469658589363099,
549
+ "eval_biology_num_tokens": 7801633.0,
550
+ "eval_biology_runtime": 47.815,
551
+ "eval_biology_samples_per_second": 10.457,
552
+ "eval_biology_steps_per_second": 2.614,
553
+ "step": 400
554
+ },
555
+ {
556
+ "epoch": 0.64,
557
+ "eval_chemistry_entropy": 1.216160174369812,
558
+ "eval_chemistry_loss": 1.23386549949646,
559
+ "eval_chemistry_mean_token_accuracy": 0.6968556141853333,
560
+ "eval_chemistry_num_tokens": 7801633.0,
561
+ "eval_chemistry_runtime": 59.3099,
562
+ "eval_chemistry_samples_per_second": 8.43,
563
+ "eval_chemistry_steps_per_second": 2.108,
564
+ "step": 400
565
+ },
566
+ {
567
+ "epoch": 0.64,
568
+ "eval_math_entropy": 1.0936446270942688,
569
+ "eval_math_loss": 1.4483542442321777,
570
+ "eval_math_mean_token_accuracy": 0.6711465301513672,
571
+ "eval_math_num_tokens": 7801633.0,
572
+ "eval_math_runtime": 60.6,
573
+ "eval_math_samples_per_second": 8.251,
574
+ "eval_math_steps_per_second": 2.063,
575
+ "step": 400
576
+ },
577
+ {
578
+ "epoch": 0.64,
579
+ "eval_physics_entropy": 1.2649442830085755,
580
+ "eval_physics_loss": 1.3500477075576782,
581
+ "eval_physics_mean_token_accuracy": 0.6875980172157288,
582
+ "eval_physics_num_tokens": 7801633.0,
583
+ "eval_physics_runtime": 68.83,
584
+ "eval_physics_samples_per_second": 7.264,
585
+ "eval_physics_steps_per_second": 1.816,
586
+ "step": 400
587
+ },
588
+ {
589
+ "entropy": 1.5343165121972562,
590
+ "epoch": 0.656,
591
+ "grad_norm": 26.375,
592
+ "learning_rate": 1.9950000000000004e-05,
593
+ "loss": 24.7076,
594
+ "mean_token_accuracy": 0.6344317115843296,
595
+ "num_tokens": 7995843.0,
596
+ "step": 410
597
+ },
598
+ {
599
+ "entropy": 1.5481104589998722,
600
+ "epoch": 0.672,
601
+ "grad_norm": 24.75,
602
+ "learning_rate": 1.9894444444444446e-05,
603
+ "loss": 24.7524,
604
+ "mean_token_accuracy": 0.6355006828904152,
605
+ "num_tokens": 8183103.0,
606
+ "step": 420
607
+ },
608
+ {
609
+ "entropy": 1.5163978930562734,
610
+ "epoch": 0.688,
611
+ "grad_norm": 22.75,
612
+ "learning_rate": 1.983888888888889e-05,
613
+ "loss": 24.5063,
614
+ "mean_token_accuracy": 0.6392664317041635,
615
+ "num_tokens": 8385976.0,
616
+ "step": 430
617
+ },
618
+ {
619
+ "entropy": 1.5027333579957485,
620
+ "epoch": 0.704,
621
+ "grad_norm": 21.875,
622
+ "learning_rate": 1.9783333333333334e-05,
623
+ "loss": 24.2448,
624
+ "mean_token_accuracy": 0.641432274878025,
625
+ "num_tokens": 8578431.0,
626
+ "step": 440
627
+ },
628
+ {
629
+ "entropy": 1.4510110400617122,
630
+ "epoch": 0.72,
631
+ "grad_norm": 22.625,
632
+ "learning_rate": 1.972777777777778e-05,
633
+ "loss": 23.3137,
634
+ "mean_token_accuracy": 0.6509528819471597,
635
+ "num_tokens": 8781342.0,
636
+ "step": 450
637
+ },
638
+ {
639
+ "entropy": 1.471246886998415,
640
+ "epoch": 0.736,
641
+ "grad_norm": 22.0,
642
+ "learning_rate": 1.9672222222222222e-05,
643
+ "loss": 23.5955,
644
+ "mean_token_accuracy": 0.6463174737989903,
645
+ "num_tokens": 8977918.0,
646
+ "step": 460
647
+ },
648
+ {
649
+ "entropy": 1.4814870670437812,
650
+ "epoch": 0.752,
651
+ "grad_norm": 22.25,
652
+ "learning_rate": 1.9616666666666668e-05,
653
+ "loss": 23.9505,
654
+ "mean_token_accuracy": 0.6442026626318693,
655
+ "num_tokens": 9169322.0,
656
+ "step": 470
657
+ },
658
+ {
659
+ "entropy": 1.4392801743000745,
660
+ "epoch": 0.768,
661
+ "grad_norm": 24.125,
662
+ "learning_rate": 1.9561111111111114e-05,
663
+ "loss": 23.1521,
664
+ "mean_token_accuracy": 0.6515602272003889,
665
+ "num_tokens": 9368141.0,
666
+ "step": 480
667
+ },
668
+ {
669
+ "entropy": 1.4612651772797107,
670
+ "epoch": 0.784,
671
+ "grad_norm": 25.875,
672
+ "learning_rate": 1.9505555555555556e-05,
673
+ "loss": 23.5729,
674
+ "mean_token_accuracy": 0.6485392026603222,
675
+ "num_tokens": 9565236.0,
676
+ "step": 490
677
+ },
678
+ {
679
+ "entropy": 1.4798763774335384,
680
+ "epoch": 0.8,
681
+ "grad_norm": 24.0,
682
+ "learning_rate": 1.9450000000000002e-05,
683
+ "loss": 23.9628,
684
+ "mean_token_accuracy": 0.6433830507099628,
685
+ "num_tokens": 9761227.0,
686
+ "step": 500
687
+ },
688
+ {
689
+ "epoch": 0.8,
690
+ "eval_biology_entropy": 1.4178470463752746,
691
+ "eval_biology_loss": 1.4294027090072632,
692
+ "eval_biology_mean_token_accuracy": 0.6571813192367554,
693
+ "eval_biology_num_tokens": 9761227.0,
694
+ "eval_biology_runtime": 47.9552,
695
+ "eval_biology_samples_per_second": 10.426,
696
+ "eval_biology_steps_per_second": 2.607,
697
+ "step": 500
698
+ },
699
+ {
700
+ "epoch": 0.8,
701
+ "eval_chemistry_entropy": 1.1406473817825318,
702
+ "eval_chemistry_loss": 1.1737452745437622,
703
+ "eval_chemistry_mean_token_accuracy": 0.7092447648048401,
704
+ "eval_chemistry_num_tokens": 9761227.0,
705
+ "eval_chemistry_runtime": 58.9872,
706
+ "eval_chemistry_samples_per_second": 8.476,
707
+ "eval_chemistry_steps_per_second": 2.119,
708
+ "step": 500
709
+ },
710
+ {
711
+ "epoch": 0.8,
712
+ "eval_math_entropy": 1.0307734375,
713
+ "eval_math_loss": 1.3721623420715332,
714
+ "eval_math_mean_token_accuracy": 0.6881859846115113,
715
+ "eval_math_num_tokens": 9761227.0,
716
+ "eval_math_runtime": 60.7077,
717
+ "eval_math_samples_per_second": 8.236,
718
+ "eval_math_steps_per_second": 2.059,
719
+ "step": 500
720
+ },
721
+ {
722
+ "epoch": 0.8,
723
+ "eval_physics_entropy": 1.1864667072296142,
724
+ "eval_physics_loss": 1.28458571434021,
725
+ "eval_physics_mean_token_accuracy": 0.701751223564148,
726
+ "eval_physics_num_tokens": 9761227.0,
727
+ "eval_physics_runtime": 69.206,
728
+ "eval_physics_samples_per_second": 7.225,
729
+ "eval_physics_steps_per_second": 1.806,
730
+ "step": 500
731
+ }
732
+ ],
733
+ "logging_steps": 10,
734
+ "max_steps": 4000,
735
+ "num_input_tokens_seen": 0,
736
+ "num_train_epochs": 7,
737
+ "save_steps": 500,
738
+ "stateful_callbacks": {
739
+ "TrainerControl": {
740
+ "args": {
741
+ "should_epoch_stop": false,
742
+ "should_evaluate": false,
743
+ "should_log": false,
744
+ "should_save": true,
745
+ "should_training_stop": false
746
+ },
747
+ "attributes": {}
748
+ }
749
+ },
750
+ "total_flos": 1.0976936565348991e+18,
751
+ "train_batch_size": 4,
752
+ "trial_name": null,
753
+ "trial_params": null
754
+ }
checkpoint-500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6f509c0fa2866c44d94e402a1f6f1cc55637496a93973e9db665aad24cccae57
3
+ size 6417