Hugging Face's logo

roonbug
/
7do5jtsy

Safetensors
Model card Files
xet
 Community
roonbug commited on
Commit
298773e
·
verified ·
1 Parent(s): 5f4de53

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 sha256:7cee411d9d57324e50ce064a192cc5a858276d508611b12fc599e0c9767112e0
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:b274fd8c51100d2e17b2a5e3647304dbf3678ed9ca245166fdbf0e4ab4ea79a3
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 sha256:102843317bde32b5a55d9699b4b61572cfc9e45b506dd234fdf06baf9e05c720
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 sha256:0048c2ca2ff60c29c4db466f595da22864c07a96941f2b7d809d1370f6cbc319
3
+ size 7286558019
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": 0.8901746967842439,
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": 0.8410565290600062,
14
+ "epoch": 0.017803493935684877,
15
+ "grad_norm": 41.25,
16
+ "learning_rate": 1.8e-07,
17
+ "loss": 14.231,
18
+ "mean_token_accuracy": 0.766436281055212,
19
+ "num_tokens": 257189.0,
20
+ "step": 10
21
+ },
22
+ {
23
+ "entropy": 0.8121558612212538,
24
+ "epoch": 0.035606987871369754,
25
+ "grad_norm": 43.25,
26
+ "learning_rate": 3.8e-07,
27
+ "loss": 13.7028,
28
+ "mean_token_accuracy": 0.7704860582947731,
29
+ "num_tokens": 528285.0,
30
+ "step": 20
31
+ },
32
+ {
33
+ "entropy": 0.8321617901325226,
34
+ "epoch": 0.053410481807054634,
35
+ "grad_norm": 38.75,
36
+ "learning_rate": 5.800000000000001e-07,
37
+ "loss": 14.0203,
38
+ "mean_token_accuracy": 0.7678851690143347,
39
+ "num_tokens": 801635.0,
40
+ "step": 30
41
+ },
42
+ {
43
+ "entropy": 0.8524315822869539,
44
+ "epoch": 0.07121397574273951,
45
+ "grad_norm": 37.75,
46
+ "learning_rate": 7.8e-07,
47
+ "loss": 14.3071,
48
+ "mean_token_accuracy": 0.7648834150284529,
49
+ "num_tokens": 1060812.0,
50
+ "step": 40
51
+ },
52
+ {
53
+ "entropy": 0.8403167355805635,
54
+ "epoch": 0.08901746967842439,
55
+ "grad_norm": 34.5,
56
+ "learning_rate": 9.800000000000001e-07,
57
+ "loss": 14.0049,
58
+ "mean_token_accuracy": 0.7667405389249324,
59
+ "num_tokens": 1327380.0,
60
+ "step": 50
61
+ },
62
+ {
63
+ "entropy": 0.8425804980099201,
64
+ "epoch": 0.10682096361410927,
65
+ "grad_norm": 34.25,
66
+ "learning_rate": 1.1800000000000001e-06,
67
+ "loss": 13.9802,
68
+ "mean_token_accuracy": 0.7680465068668128,
69
+ "num_tokens": 1597405.0,
70
+ "step": 60
71
+ },
72
+ {
73
+ "entropy": 0.8755916632711888,
74
+ "epoch": 0.12462445754979415,
75
+ "grad_norm": 32.25,
76
+ "learning_rate": 1.3800000000000001e-06,
77
+ "loss": 14.4557,
78
+ "mean_token_accuracy": 0.7614609662443399,
79
+ "num_tokens": 1859684.0,
80
+ "step": 70
81
+ },
82
+ {
83
+ "entropy": 0.8682003870606423,
84
+ "epoch": 0.14242795148547902,
85
+ "grad_norm": 30.375,
86
+ "learning_rate": 1.5800000000000001e-06,
87
+ "loss": 14.1174,
88
+ "mean_token_accuracy": 0.7653359699994325,
89
+ "num_tokens": 2118068.0,
90
+ "step": 80
91
+ },
92
+ {
93
+ "entropy": 0.8360268581658602,
94
+ "epoch": 0.1602314454211639,
95
+ "grad_norm": 28.375,
96
+ "learning_rate": 1.7800000000000001e-06,
97
+ "loss": 13.6773,
98
+ "mean_token_accuracy": 0.7724899325519801,
99
+ "num_tokens": 2388824.0,
100
+ "step": 90
101
+ },
102
+ {
103
+ "entropy": 0.8634540975093842,
104
+ "epoch": 0.17803493935684878,
105
+ "grad_norm": 34.0,
106
+ "learning_rate": 1.98e-06,
107
+ "loss": 14.0689,
108
+ "mean_token_accuracy": 0.7672080259770155,
109
+ "num_tokens": 2644330.0,
110
+ "step": 100
111
+ },
112
+ {
113
+ "epoch": 0.17803493935684878,
114
+ "eval_biology_entropy": 1.128741333961487,
115
+ "eval_biology_loss": 1.1601769924163818,
116
+ "eval_biology_mean_token_accuracy": 0.7042496418952942,
117
+ "eval_biology_num_tokens": 2644330.0,
118
+ "eval_biology_runtime": 54.4222,
119
+ "eval_biology_samples_per_second": 9.187,
120
+ "eval_biology_steps_per_second": 2.297,
121
+ "step": 100
122
+ },
123
+ {
124
+ "epoch": 0.17803493935684878,
125
+ "eval_chemistry_entropy": 0.8544474849700928,
126
+ "eval_chemistry_loss": 0.8545747995376587,
127
+ "eval_chemistry_mean_token_accuracy": 0.7713311352729797,
128
+ "eval_chemistry_num_tokens": 2644330.0,
129
+ "eval_chemistry_runtime": 65.3258,
130
+ "eval_chemistry_samples_per_second": 7.654,
131
+ "eval_chemistry_steps_per_second": 1.913,
132
+ "step": 100
133
+ },
134
+ {
135
+ "entropy": 0.8453936260193586,
136
+ "epoch": 0.19583843329253367,
137
+ "grad_norm": 29.5,
138
+ "learning_rate": 2.1800000000000003e-06,
139
+ "loss": 13.6681,
140
+ "mean_token_accuracy": 0.7730784911662341,
141
+ "num_tokens": 2913700.0,
142
+ "step": 110
143
+ },
144
+ {
145
+ "entropy": 0.8460511896759272,
146
+ "epoch": 0.21364192722821854,
147
+ "grad_norm": 27.625,
148
+ "learning_rate": 2.38e-06,
149
+ "loss": 13.728,
150
+ "mean_token_accuracy": 0.7713178683072328,
151
+ "num_tokens": 3185255.0,
152
+ "step": 120
153
+ },
154
+ {
155
+ "entropy": 0.8403830077499151,
156
+ "epoch": 0.2314454211639034,
157
+ "grad_norm": 28.25,
158
+ "learning_rate": 2.5800000000000003e-06,
159
+ "loss": 13.5642,
160
+ "mean_token_accuracy": 0.7730850588530302,
161
+ "num_tokens": 3454750.0,
162
+ "step": 130
163
+ },
164
+ {
165
+ "entropy": 0.8515023712068797,
166
+ "epoch": 0.2492489150995883,
167
+ "grad_norm": 30.0,
168
+ "learning_rate": 2.7800000000000005e-06,
169
+ "loss": 13.8032,
170
+ "mean_token_accuracy": 0.7721988521516323,
171
+ "num_tokens": 3719113.0,
172
+ "step": 140
173
+ },
174
+ {
175
+ "entropy": 0.8474330805242062,
176
+ "epoch": 0.26705240903527316,
177
+ "grad_norm": 27.375,
178
+ "learning_rate": 2.9800000000000003e-06,
179
+ "loss": 13.728,
180
+ "mean_token_accuracy": 0.7698215767741203,
181
+ "num_tokens": 3990505.0,
182
+ "step": 150
183
+ },
184
+ {
185
+ "entropy": 0.8292176539078355,
186
+ "epoch": 0.28485590297095803,
187
+ "grad_norm": 24.625,
188
+ "learning_rate": 3.1800000000000005e-06,
189
+ "loss": 13.3391,
190
+ "mean_token_accuracy": 0.776393149420619,
191
+ "num_tokens": 4267403.0,
192
+ "step": 160
193
+ },
194
+ {
195
+ "entropy": 0.8280835278332234,
196
+ "epoch": 0.30265939690664295,
197
+ "grad_norm": 29.875,
198
+ "learning_rate": 3.3800000000000007e-06,
199
+ "loss": 13.3175,
200
+ "mean_token_accuracy": 0.7770637154579163,
201
+ "num_tokens": 4535458.0,
202
+ "step": 170
203
+ },
204
+ {
205
+ "entropy": 0.842758315615356,
206
+ "epoch": 0.3204628908423278,
207
+ "grad_norm": 27.5,
208
+ "learning_rate": 3.58e-06,
209
+ "loss": 13.6062,
210
+ "mean_token_accuracy": 0.7731371156871318,
211
+ "num_tokens": 4796815.0,
212
+ "step": 180
213
+ },
214
+ {
215
+ "entropy": 0.8247860476374627,
216
+ "epoch": 0.3382663847780127,
217
+ "grad_norm": 22.75,
218
+ "learning_rate": 3.7800000000000002e-06,
219
+ "loss": 13.2445,
220
+ "mean_token_accuracy": 0.7772165313363075,
221
+ "num_tokens": 5066948.0,
222
+ "step": 190
223
+ },
224
+ {
225
+ "entropy": 0.8495354067534209,
226
+ "epoch": 0.35606987871369755,
227
+ "grad_norm": 26.125,
228
+ "learning_rate": 3.980000000000001e-06,
229
+ "loss": 13.7137,
230
+ "mean_token_accuracy": 0.7696468211710453,
231
+ "num_tokens": 5324751.0,
232
+ "step": 200
233
+ },
234
+ {
235
+ "epoch": 0.35606987871369755,
236
+ "eval_biology_entropy": 1.1391102423667907,
237
+ "eval_biology_loss": 1.1645851135253906,
238
+ "eval_biology_mean_token_accuracy": 0.7036963219642639,
239
+ "eval_biology_num_tokens": 5324751.0,
240
+ "eval_biology_runtime": 47.5522,
241
+ "eval_biology_samples_per_second": 10.515,
242
+ "eval_biology_steps_per_second": 2.629,
243
+ "step": 200
244
+ },
245
+ {
246
+ "epoch": 0.35606987871369755,
247
+ "eval_chemistry_entropy": 0.8357059621810913,
248
+ "eval_chemistry_loss": 0.8271888494491577,
249
+ "eval_chemistry_mean_token_accuracy": 0.7765080814361572,
250
+ "eval_chemistry_num_tokens": 5324751.0,
251
+ "eval_chemistry_runtime": 58.2133,
252
+ "eval_chemistry_samples_per_second": 8.589,
253
+ "eval_chemistry_steps_per_second": 2.147,
254
+ "step": 200
255
+ },
256
+ {
257
+ "entropy": 0.8138596788048744,
258
+ "epoch": 0.3738733726493824,
259
+ "grad_norm": 27.875,
260
+ "learning_rate": 4.18e-06,
261
+ "loss": 13.0337,
262
+ "mean_token_accuracy": 0.7797718059271574,
263
+ "num_tokens": 5585508.0,
264
+ "step": 210
265
+ },
266
+ {
267
+ "entropy": 0.8215312957763672,
268
+ "epoch": 0.39167686658506734,
269
+ "grad_norm": 25.75,
270
+ "learning_rate": 4.38e-06,
271
+ "loss": 13.2224,
272
+ "mean_token_accuracy": 0.7784637857228518,
273
+ "num_tokens": 5848889.0,
274
+ "step": 220
275
+ },
276
+ {
277
+ "entropy": 0.8100055737420917,
278
+ "epoch": 0.4094803605207522,
279
+ "grad_norm": 27.625,
280
+ "learning_rate": 4.58e-06,
281
+ "loss": 13.0714,
282
+ "mean_token_accuracy": 0.7809950839728117,
283
+ "num_tokens": 6114855.0,
284
+ "step": 230
285
+ },
286
+ {
287
+ "entropy": 0.8117449183017016,
288
+ "epoch": 0.4272838544564371,
289
+ "grad_norm": 25.875,
290
+ "learning_rate": 4.78e-06,
291
+ "loss": 13.0914,
292
+ "mean_token_accuracy": 0.7797230206429958,
293
+ "num_tokens": 6378152.0,
294
+ "step": 240
295
+ },
296
+ {
297
+ "entropy": 0.8300687098875642,
298
+ "epoch": 0.44508734839212194,
299
+ "grad_norm": 24.5,
300
+ "learning_rate": 4.980000000000001e-06,
301
+ "loss": 13.3199,
302
+ "mean_token_accuracy": 0.7753712415695191,
303
+ "num_tokens": 6637273.0,
304
+ "step": 250
305
+ },
306
+ {
307
+ "entropy": 0.8000802919268608,
308
+ "epoch": 0.4628908423278068,
309
+ "grad_norm": 23.75,
310
+ "learning_rate": 5.18e-06,
311
+ "loss": 12.8911,
312
+ "mean_token_accuracy": 0.7822641927748919,
313
+ "num_tokens": 6896684.0,
314
+ "step": 260
315
+ },
316
+ {
317
+ "entropy": 0.8173866732046008,
318
+ "epoch": 0.48069433626349173,
319
+ "grad_norm": 25.75,
320
+ "learning_rate": 5.380000000000001e-06,
321
+ "loss": 13.1839,
322
+ "mean_token_accuracy": 0.7791078709065914,
323
+ "num_tokens": 7166608.0,
324
+ "step": 270
325
+ },
326
+ {
327
+ "entropy": 0.7851757485419512,
328
+ "epoch": 0.4984978301991766,
329
+ "grad_norm": 23.125,
330
+ "learning_rate": 5.580000000000001e-06,
331
+ "loss": 12.5647,
332
+ "mean_token_accuracy": 0.7872234936803579,
333
+ "num_tokens": 7444923.0,
334
+ "step": 280
335
+ },
336
+ {
337
+ "entropy": 0.8196133345365524,
338
+ "epoch": 0.5163013241348615,
339
+ "grad_norm": 23.75,
340
+ "learning_rate": 5.78e-06,
341
+ "loss": 13.195,
342
+ "mean_token_accuracy": 0.7780409008264542,
343
+ "num_tokens": 7706502.0,
344
+ "step": 290
345
+ },
346
+ {
347
+ "entropy": 0.8108824253082275,
348
+ "epoch": 0.5341048180705463,
349
+ "grad_norm": 25.75,
350
+ "learning_rate": 5.98e-06,
351
+ "loss": 13.1243,
352
+ "mean_token_accuracy": 0.7791608296334743,
353
+ "num_tokens": 7969704.0,
354
+ "step": 300
355
+ },
356
+ {
357
+ "epoch": 0.5341048180705463,
358
+ "eval_biology_entropy": 1.1140506463050843,
359
+ "eval_biology_loss": 1.1709669828414917,
360
+ "eval_biology_mean_token_accuracy": 0.7033902740478516,
361
+ "eval_biology_num_tokens": 7969704.0,
362
+ "eval_biology_runtime": 47.6418,
363
+ "eval_biology_samples_per_second": 10.495,
364
+ "eval_biology_steps_per_second": 2.624,
365
+ "step": 300
366
+ },
367
+ {
368
+ "epoch": 0.5341048180705463,
369
+ "eval_chemistry_entropy": 0.7940924577713012,
370
+ "eval_chemistry_loss": 0.8012509942054749,
371
+ "eval_chemistry_mean_token_accuracy": 0.7820386853218079,
372
+ "eval_chemistry_num_tokens": 7969704.0,
373
+ "eval_chemistry_runtime": 59.4475,
374
+ "eval_chemistry_samples_per_second": 8.411,
375
+ "eval_chemistry_steps_per_second": 2.103,
376
+ "step": 300
377
+ },
378
+ {
379
+ "entropy": 0.7919084688648581,
380
+ "epoch": 0.5519083120062312,
381
+ "grad_norm": 24.375,
382
+ "learning_rate": 6.18e-06,
383
+ "loss": 12.7303,
384
+ "mean_token_accuracy": 0.7834540419280529,
385
+ "num_tokens": 8242162.0,
386
+ "step": 310
387
+ },
388
+ {
389
+ "entropy": 0.8003328915685415,
390
+ "epoch": 0.5697118059419161,
391
+ "grad_norm": 22.875,
392
+ "learning_rate": 6.380000000000001e-06,
393
+ "loss": 12.8333,
394
+ "mean_token_accuracy": 0.7824427511543035,
395
+ "num_tokens": 8497852.0,
396
+ "step": 320
397
+ },
398
+ {
399
+ "entropy": 0.7982260027900339,
400
+ "epoch": 0.587515299877601,
401
+ "grad_norm": 28.0,
402
+ "learning_rate": 6.5800000000000005e-06,
403
+ "loss": 12.8827,
404
+ "mean_token_accuracy": 0.7826927099376917,
405
+ "num_tokens": 8757753.0,
406
+ "step": 330
407
+ },
408
+ {
409
+ "entropy": 0.7888958260416985,
410
+ "epoch": 0.6053187938132859,
411
+ "grad_norm": 25.125,
412
+ "learning_rate": 6.780000000000001e-06,
413
+ "loss": 12.7464,
414
+ "mean_token_accuracy": 0.7863177515566349,
415
+ "num_tokens": 9024677.0,
416
+ "step": 340
417
+ },
418
+ {
419
+ "entropy": 0.7913781819865108,
420
+ "epoch": 0.6231222877489707,
421
+ "grad_norm": 26.0,
422
+ "learning_rate": 6.98e-06,
423
+ "loss": 12.7471,
424
+ "mean_token_accuracy": 0.7842035111039877,
425
+ "num_tokens": 9291760.0,
426
+ "step": 350
427
+ },
428
+ {
429
+ "entropy": 0.7765169985592365,
430
+ "epoch": 0.6409257816846556,
431
+ "grad_norm": 22.0,
432
+ "learning_rate": 7.180000000000001e-06,
433
+ "loss": 12.5378,
434
+ "mean_token_accuracy": 0.7859202962368727,
435
+ "num_tokens": 9561091.0,
436
+ "step": 360
437
+ },
438
+ {
439
+ "entropy": 0.7792716162279248,
440
+ "epoch": 0.6587292756203404,
441
+ "grad_norm": 25.0,
442
+ "learning_rate": 7.3800000000000005e-06,
443
+ "loss": 12.4956,
444
+ "mean_token_accuracy": 0.7880564954131841,
445
+ "num_tokens": 9827272.0,
446
+ "step": 370
447
+ },
448
+ {
449
+ "entropy": 0.758336128294468,
450
+ "epoch": 0.6765327695560254,
451
+ "grad_norm": 25.0,
452
+ "learning_rate": 7.58e-06,
453
+ "loss": 12.1732,
454
+ "mean_token_accuracy": 0.7918921418488025,
455
+ "num_tokens": 10096065.0,
456
+ "step": 380
457
+ },
458
+ {
459
+ "entropy": 0.7533971995115281,
460
+ "epoch": 0.6943362634917103,
461
+ "grad_norm": 24.375,
462
+ "learning_rate": 7.78e-06,
463
+ "loss": 12.1132,
464
+ "mean_token_accuracy": 0.7922064792364836,
465
+ "num_tokens": 10364601.0,
466
+ "step": 390
467
+ },
468
+ {
469
+ "entropy": 0.7620638139545918,
470
+ "epoch": 0.7121397574273951,
471
+ "grad_norm": 23.625,
472
+ "learning_rate": 7.980000000000002e-06,
473
+ "loss": 12.3379,
474
+ "mean_token_accuracy": 0.7897147350013256,
475
+ "num_tokens": 10633325.0,
476
+ "step": 400
477
+ },
478
+ {
479
+ "epoch": 0.7121397574273951,
480
+ "eval_biology_entropy": 1.1209784712791442,
481
+ "eval_biology_loss": 1.175487995147705,
482
+ "eval_biology_mean_token_accuracy": 0.7023502192497253,
483
+ "eval_biology_num_tokens": 10633325.0,
484
+ "eval_biology_runtime": 45.7349,
485
+ "eval_biology_samples_per_second": 10.933,
486
+ "eval_biology_steps_per_second": 2.733,
487
+ "step": 400
488
+ },
489
+ {
490
+ "epoch": 0.7121397574273951,
491
+ "eval_chemistry_entropy": 0.7846209690570831,
492
+ "eval_chemistry_loss": 0.7767007946968079,
493
+ "eval_chemistry_mean_token_accuracy": 0.7876848134994506,
494
+ "eval_chemistry_num_tokens": 10633325.0,
495
+ "eval_chemistry_runtime": 56.279,
496
+ "eval_chemistry_samples_per_second": 8.884,
497
+ "eval_chemistry_steps_per_second": 2.221,
498
+ "step": 400
499
+ },
500
+ {
501
+ "entropy": 0.7557655736804009,
502
+ "epoch": 0.72994325136308,
503
+ "grad_norm": 25.0,
504
+ "learning_rate": 8.18e-06,
505
+ "loss": 12.1754,
506
+ "mean_token_accuracy": 0.7914514016360045,
507
+ "num_tokens": 10897916.0,
508
+ "step": 410
509
+ },
510
+ {
511
+ "entropy": 0.7626162808388471,
512
+ "epoch": 0.7477467452987648,
513
+ "grad_norm": 22.375,
514
+ "learning_rate": 8.380000000000001e-06,
515
+ "loss": 12.1991,
516
+ "mean_token_accuracy": 0.7916438620537519,
517
+ "num_tokens": 11165356.0,
518
+ "step": 420
519
+ },
520
+ {
521
+ "entropy": 0.7673018729314208,
522
+ "epoch": 0.7655502392344498,
523
+ "grad_norm": 24.375,
524
+ "learning_rate": 8.580000000000001e-06,
525
+ "loss": 12.3987,
526
+ "mean_token_accuracy": 0.7888187035918236,
527
+ "num_tokens": 11436799.0,
528
+ "step": 430
529
+ },
530
+ {
531
+ "entropy": 0.7744929634034634,
532
+ "epoch": 0.7833537331701347,
533
+ "grad_norm": 24.125,
534
+ "learning_rate": 8.78e-06,
535
+ "loss": 12.4708,
536
+ "mean_token_accuracy": 0.787474300712347,
537
+ "num_tokens": 11703496.0,
538
+ "step": 440
539
+ },
540
+ {
541
+ "entropy": 0.7655632747337222,
542
+ "epoch": 0.8011572271058195,
543
+ "grad_norm": 24.875,
544
+ "learning_rate": 8.98e-06,
545
+ "loss": 12.299,
546
+ "mean_token_accuracy": 0.7900604665279388,
547
+ "num_tokens": 11965530.0,
548
+ "step": 450
549
+ },
550
+ {
551
+ "entropy": 0.7661685338243842,
552
+ "epoch": 0.8189607210415044,
553
+ "grad_norm": 23.0,
554
+ "learning_rate": 9.180000000000002e-06,
555
+ "loss": 12.3101,
556
+ "mean_token_accuracy": 0.7891276117414237,
557
+ "num_tokens": 12224427.0,
558
+ "step": 460
559
+ },
560
+ {
561
+ "entropy": 0.7463042287155985,
562
+ "epoch": 0.8367642149771892,
563
+ "grad_norm": 24.125,
564
+ "learning_rate": 9.38e-06,
565
+ "loss": 12.0346,
566
+ "mean_token_accuracy": 0.7945169288665056,
567
+ "num_tokens": 12509124.0,
568
+ "step": 470
569
+ },
570
+ {
571
+ "entropy": 0.753023486584425,
572
+ "epoch": 0.8545677089128741,
573
+ "grad_norm": 21.25,
574
+ "learning_rate": 9.58e-06,
575
+ "loss": 12.0236,
576
+ "mean_token_accuracy": 0.7931863989681005,
577
+ "num_tokens": 12778408.0,
578
+ "step": 480
579
+ },
580
+ {
581
+ "entropy": 0.7310503415763379,
582
+ "epoch": 0.8723712028485591,
583
+ "grad_norm": 22.25,
584
+ "learning_rate": 9.780000000000001e-06,
585
+ "loss": 11.8094,
586
+ "mean_token_accuracy": 0.7964685469865799,
587
+ "num_tokens": 13046473.0,
588
+ "step": 490
589
+ },
590
+ {
591
+ "entropy": 0.7645759535953403,
592
+ "epoch": 0.8901746967842439,
593
+ "grad_norm": 22.375,
594
+ "learning_rate": 9.980000000000001e-06,
595
+ "loss": 12.2645,
596
+ "mean_token_accuracy": 0.7906601417809724,
597
+ "num_tokens": 13301659.0,
598
+ "step": 500
599
+ },
600
+ {
601
+ "epoch": 0.8901746967842439,
602
+ "eval_biology_entropy": 1.092752426624298,
603
+ "eval_biology_loss": 1.1775906085968018,
604
+ "eval_biology_mean_token_accuracy": 0.7025688862800599,
605
+ "eval_biology_num_tokens": 13301659.0,
606
+ "eval_biology_runtime": 268.8833,
607
+ "eval_biology_samples_per_second": 1.86,
608
+ "eval_biology_steps_per_second": 0.465,
609
+ "step": 500
610
+ },
611
+ {
612
+ "epoch": 0.8901746967842439,
613
+ "eval_chemistry_entropy": 0.7384080934524536,
614
+ "eval_chemistry_loss": 0.7555699944496155,
615
+ "eval_chemistry_mean_token_accuracy": 0.7920897974967956,
616
+ "eval_chemistry_num_tokens": 13301659.0,
617
+ "eval_chemistry_runtime": 450.4029,
618
+ "eval_chemistry_samples_per_second": 1.11,
619
+ "eval_chemistry_steps_per_second": 0.278,
620
+ "step": 500
621
+ }
622
+ ],
623
+ "logging_steps": 10,
624
+ "max_steps": 10000,
625
+ "num_input_tokens_seen": 0,
626
+ "num_train_epochs": 18,
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": 1.3874307456966482e+18,
641
+ "train_batch_size": 4,
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:b9131b82800c4c22eb7beea6d65762112da81a779f69c0cfd077fb270bfc4764
3
+ size 6481