Hugging Face's logo

stefanwebb
/
banking77-intent-classifier

Text Generation
Transformers
Safetensors
qwen3
Generated from Trainer
sft
trl
conversational
text-generation-inference
Model card Files
xet
 Community
stefanwebb commited on
Commit
b874299
·
1 Parent(s): 6280f7c

everything except large files

Browse files
Files changed (12) hide show
  1. README.md +58 -0
  2. added_tokens.json +28 -0
  3. chat_template.jinja +61 -0
  4. config.json +68 -0
  5. generation_config.json +12 -0
  6. merges.txt +0 -0
  7. metrics_rank0000.jsonl +107 -0
  8. special_tokens_map.json +31 -0
  9. tokenizer_config.json +239 -0
  10. trainer_state.json +1423 -0
  11. training_args.bin +0 -0
  12. vocab.json +0 -0
README.md ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: Qwen/Qwen3-4B-Instruct-2507
3
+ library_name: transformers
4
+ model_name: training_output
5
+ tags:
6
+ - generated_from_trainer
7
+ - sft
8
+ - trl
9
+ licence: license
10
+ ---
11
+
12
+ # Model Card for training_output
13
+
14
+ This model is a fine-tuned version of [Qwen/Qwen3-4B-Instruct-2507](https://huggingface.co/Qwen/Qwen3-4B-Instruct-2507).
15
+ It has been trained using [TRL](https://github.com/huggingface/trl).
16
+
17
+ ## Quick start
18
+
19
+ ```python
20
+ from transformers import pipeline
21
+
22
+ question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
23
+ generator = pipeline("text-generation", model="None", device="cuda")
24
+ output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
25
+ print(output["generated_text"])
26
+ ```
27
+
28
+ ## Training procedure
29
+
30
+
31
+
32
+
33
+ This model was trained with SFT.
34
+
35
+ ### Framework versions
36
+
37
+ - TRL: 0.26.2
38
+ - Transformers: 4.57.6
39
+ - Pytorch: 2.8.0
40
+ - Datasets: 4.5.0
41
+ - Tokenizers: 0.22.2
42
+
43
+ ## Citations
44
+
45
+
46
+
47
+ Cite TRL as:
48
+
49
+ ```bibtex
50
+ @misc{vonwerra2022trl,
51
+ title = {{TRL: Transformer Reinforcement Learning}},
52
+ author = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallou{\'e}dec},
53
+ year = 2020,
54
+ journal = {GitHub repository},
55
+ publisher = {GitHub},
56
+ howpublished = {\url{https://github.com/huggingface/trl}}
57
+ }
58
+ ```
added_tokens.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "</think>": 151668,
3
+ "</tool_call>": 151658,
4
+ "</tool_response>": 151666,
5
+ "<think>": 151667,
6
+ "<tool_call>": 151657,
7
+ "<tool_response>": 151665,
8
+ "<|box_end|>": 151649,
9
+ "<|box_start|>": 151648,
10
+ "<|endoftext|>": 151643,
11
+ "<|file_sep|>": 151664,
12
+ "<|fim_middle|>": 151660,
13
+ "<|fim_pad|>": 151662,
14
+ "<|fim_prefix|>": 151659,
15
+ "<|fim_suffix|>": 151661,
16
+ "<|im_end|>": 151645,
17
+ "<|im_start|>": 151644,
18
+ "<|image_pad|>": 151655,
19
+ "<|object_ref_end|>": 151647,
20
+ "<|object_ref_start|>": 151646,
21
+ "<|quad_end|>": 151651,
22
+ "<|quad_start|>": 151650,
23
+ "<|repo_name|>": 151663,
24
+ "<|video_pad|>": 151656,
25
+ "<|vision_end|>": 151653,
26
+ "<|vision_pad|>": 151654,
27
+ "<|vision_start|>": 151652
28
+ }
chat_template.jinja ADDED
@@ -0,0 +1,61 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0].role == 'system' %}
4
+ {{- messages[0].content + '\n\n' }}
5
+ {%- endif %}
6
+ {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
7
+ {%- for tool in tools %}
8
+ {{- "\n" }}
9
+ {{- tool | tojson }}
10
+ {%- endfor %}
11
+ {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
12
+ {%- else %}
13
+ {%- if messages[0].role == 'system' %}
14
+ {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
15
+ {%- endif %}
16
+ {%- endif %}
17
+ {%- for message in messages %}
18
+ {%- if message.content is string %}
19
+ {%- set content = message.content %}
20
+ {%- else %}
21
+ {%- set content = '' %}
22
+ {%- endif %}
23
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
24
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
25
+ {%- elif message.role == "assistant" %}
26
+ {{- '<|im_start|>' + message.role + '\n' + content }}
27
+ {%- if message.tool_calls %}
28
+ {%- for tool_call in message.tool_calls %}
29
+ {%- if (loop.first and content) or (not loop.first) %}
30
+ {{- '\n' }}
31
+ {%- endif %}
32
+ {%- if tool_call.function %}
33
+ {%- set tool_call = tool_call.function %}
34
+ {%- endif %}
35
+ {{- '<tool_call>\n{"name": "' }}
36
+ {{- tool_call.name }}
37
+ {{- '", "arguments": ' }}
38
+ {%- if tool_call.arguments is string %}
39
+ {{- tool_call.arguments }}
40
+ {%- else %}
41
+ {{- tool_call.arguments | tojson }}
42
+ {%- endif %}
43
+ {{- '}\n</tool_call>' }}
44
+ {%- endfor %}
45
+ {%- endif %}
46
+ {{- '<|im_end|>\n' }}
47
+ {%- elif message.role == "tool" %}
48
+ {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
49
+ {{- '<|im_start|>user' }}
50
+ {%- endif %}
51
+ {{- '\n<tool_response>\n' }}
52
+ {{- content }}
53
+ {{- '\n</tool_response>' }}
54
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
55
+ {{- '<|im_end|>\n' }}
56
+ {%- endif %}
57
+ {%- endif %}
58
+ {%- endfor %}
59
+ {%- if add_generation_prompt %}
60
+ {{- '<|im_start|>assistant\n' }}
61
+ {%- endif %}
config.json ADDED
@@ -0,0 +1,68 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen3ForCausalLM"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "dtype": "bfloat16",
8
+ "eos_token_id": 151645,
9
+ "head_dim": 128,
10
+ "hidden_act": "silu",
11
+ "hidden_size": 2560,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 9728,
14
+ "layer_types": [
15
+ "full_attention",
16
+ "full_attention",
17
+ "full_attention",
18
+ "full_attention",
19
+ "full_attention",
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention",
45
+ "full_attention",
46
+ "full_attention",
47
+ "full_attention",
48
+ "full_attention",
49
+ "full_attention",
50
+ "full_attention"
51
+ ],
52
+ "max_position_embeddings": 262144,
53
+ "max_window_layers": 36,
54
+ "model_type": "qwen3",
55
+ "num_attention_heads": 32,
56
+ "num_hidden_layers": 36,
57
+ "num_key_value_heads": 8,
58
+ "pad_token_id": 151643,
59
+ "rms_norm_eps": 1e-06,
60
+ "rope_scaling": null,
61
+ "rope_theta": 5000000,
62
+ "sliding_window": null,
63
+ "tie_word_embeddings": true,
64
+ "transformers_version": "4.57.6",
65
+ "use_cache": false,
66
+ "use_sliding_window": false,
67
+ "vocab_size": 151936
68
+ }
generation_config.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_sample": true,
3
+ "eos_token_id": [
4
+ 151645,
5
+ 151643
6
+ ],
7
+ "pad_token_id": 151643,
8
+ "temperature": 0.7,
9
+ "top_k": 20,
10
+ "top_p": 0.8,
11
+ "transformers_version": "4.57.6"
12
+ }
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
metrics_rank0000.jsonl ADDED
@@ -0,0 +1,107 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"loss": 2.9876, "grad_norm": 157.0, "learning_rate": 5.925925925925926e-06, "entropy": 0.825, "num_tokens": 114489.0, "mean_token_accuracy": 0.6893173575401306, "epoch": 0.01890359168241966, "num_input_tokens_seen": 115216, "train_runtime": 4.4379, "train_tokens_per_second": 25961.602}
2
+ {"loss": 0.6662, "grad_norm": 24.5, "learning_rate": 1.3333333333333333e-05, "entropy": 0.8890625, "num_tokens": 229109.0, "mean_token_accuracy": 0.8326915562152862, "epoch": 0.03780718336483932, "num_input_tokens_seen": 230592, "train_runtime": 7.7641, "train_tokens_per_second": 29699.859}
3
+ {"loss": 0.6027, "grad_norm": 18.625, "learning_rate": 2.074074074074074e-05, "entropy": 1.02890625, "num_tokens": 343545.0, "mean_token_accuracy": 0.8529165983200073, "epoch": 0.05671077504725898, "num_input_tokens_seen": 345600, "train_runtime": 13.2817, "train_tokens_per_second": 26020.836}
4
+ {"loss": 0.4447, "grad_norm": 19.125, "learning_rate": 2.814814814814815e-05, "entropy": 1.1421875, "num_tokens": 458335.0, "mean_token_accuracy": 0.881816154718399, "epoch": 0.07561436672967864, "num_input_tokens_seen": 461282, "train_runtime": 16.8766, "train_tokens_per_second": 27332.56}
5
+ {"loss": 0.3989, "grad_norm": 21.75, "learning_rate": 3.555555555555555e-05, "entropy": 1.21640625, "num_tokens": 572824.0, "mean_token_accuracy": 0.8929793298244476, "epoch": 0.0945179584120983, "num_input_tokens_seen": 576346, "train_runtime": 22.5256, "train_tokens_per_second": 25586.277}
6
+ {"loss": 0.4872, "grad_norm": 13.625, "learning_rate": 3.999843966403289e-05, "entropy": 1.26484375, "num_tokens": 687152.0, "mean_token_accuracy": 0.8781549751758575, "epoch": 0.11342155009451796, "num_input_tokens_seen": 691188, "train_runtime": 26.0077, "train_tokens_per_second": 26576.234}
7
+ {"loss": 0.28, "grad_norm": 17.375, "learning_rate": 3.99808886803243e-05, "entropy": 1.28515625, "num_tokens": 801973.0, "mean_token_accuracy": 0.9074305832386017, "epoch": 0.1323251417769376, "num_input_tokens_seen": 806740, "train_runtime": 29.7225, "train_tokens_per_second": 27142.403}
8
+ {"loss": 0.356, "grad_norm": 12.625, "learning_rate": 3.994385346473689e-05, "entropy": 1.2984375, "num_tokens": 916426.0, "mean_token_accuracy": 0.9146295249462127, "epoch": 0.15122873345935728, "num_input_tokens_seen": 921796, "train_runtime": 34.6767, "train_tokens_per_second": 26582.553}
9
+ {"loss": 0.3933, "grad_norm": 11.4375, "learning_rate": 3.9887370131917e-05, "entropy": 1.56953125, "num_tokens": 1030824.0, "mean_token_accuracy": 0.9064954161643982, "epoch": 0.17013232514177692, "num_input_tokens_seen": 1036824, "train_runtime": 38.0075, "train_tokens_per_second": 27279.472}
10
+ {"loss": 0.2873, "grad_norm": 15.1875, "learning_rate": 3.981149376121427e-05, "entropy": 1.7515625, "num_tokens": 1145500.0, "mean_token_accuracy": 0.9260397672653198, "epoch": 0.1890359168241966, "num_input_tokens_seen": 1152356, "train_runtime": 43.1792, "train_tokens_per_second": 26687.759}
11
+ {"loss": 0.3322, "grad_norm": 8.5625, "learning_rate": 3.97162983429714e-05, "entropy": 1.75859375, "num_tokens": 1260057.0, "mean_token_accuracy": 0.9256749033927918, "epoch": 0.20793950850661624, "num_input_tokens_seen": 1267634, "train_runtime": 46.6166, "train_tokens_per_second": 27192.755}
12
+ {"loss": 0.2865, "grad_norm": 11.6875, "learning_rate": 3.960187670637294e-05, "entropy": 1.734375, "num_tokens": 1374973.0, "mean_token_accuracy": 0.9282522916793823, "epoch": 0.22684310018903592, "num_input_tokens_seen": 1383494, "train_runtime": 50.5007, "train_tokens_per_second": 27395.522}
13
+ {"loss": 0.2277, "grad_norm": 9.75, "learning_rate": 3.946834042892355e-05, "entropy": 1.70859375, "num_tokens": 1489683.0, "mean_token_accuracy": 0.9320353448390961, "epoch": 0.24574669187145556, "num_input_tokens_seen": 1499052, "train_runtime": 55.2907, "train_tokens_per_second": 27112.206}
14
+ {"loss": 0.2733, "grad_norm": 10.125, "learning_rate": 3.931581972764386e-05, "entropy": 1.6578125, "num_tokens": 1604106.0, "mean_token_accuracy": 0.9363594233989716, "epoch": 0.2646502835538752, "num_input_tokens_seen": 1614146, "train_runtime": 58.5263, "train_tokens_per_second": 27579.832}
15
+ {"loss": 0.2164, "grad_norm": 6.34375, "learning_rate": 3.91444633320903e-05, "entropy": 1.57109375, "num_tokens": 1718632.0, "mean_token_accuracy": 0.9349239528179168, "epoch": 0.2835538752362949, "num_input_tokens_seen": 1729362, "train_runtime": 63.5177, "train_tokens_per_second": 27226.454}
16
+ {"loss": 0.2173, "grad_norm": 10.875, "learning_rate": 3.8954438339322366e-05, "entropy": 1.57734375, "num_tokens": 1833068.0, "mean_token_accuracy": 0.9350460767745972, "epoch": 0.30245746691871456, "num_input_tokens_seen": 1844444, "train_runtime": 66.8194, "train_tokens_per_second": 27603.402}
17
+ {"loss": 0.2337, "grad_norm": 11.0, "learning_rate": 3.874593005095909e-05, "entropy": 1.62421875, "num_tokens": 1947640.0, "mean_token_accuracy": 0.929820317029953, "epoch": 0.32136105860113423, "num_input_tokens_seen": 1959682, "train_runtime": 70.4744, "train_tokens_per_second": 27806.99}
18
+ {"loss": 0.2156, "grad_norm": 8.625, "learning_rate": 3.851914179248333e-05, "entropy": 1.71171875, "num_tokens": 2062310.0, "mean_token_accuracy": 0.9308744966983795, "epoch": 0.34026465028355385, "num_input_tokens_seen": 2075138, "train_runtime": 75.9345, "train_tokens_per_second": 27327.991}
19
+ {"loss": 0.2718, "grad_norm": 13.5625, "learning_rate": 3.82742947149703e-05, "entropy": 1.86875, "num_tokens": 2176716.0, "mean_token_accuracy": 0.9264281988143921, "epoch": 0.3591682419659735, "num_input_tokens_seen": 2190160, "train_runtime": 79.4416, "train_tokens_per_second": 27569.42}
20
+ {"loss": 0.3385, "grad_norm": 6.125, "learning_rate": 3.801162757943359e-05, "entropy": 1.94765625, "num_tokens": 2291230.0, "mean_token_accuracy": 0.9164456725120544, "epoch": 0.3780718336483932, "num_input_tokens_seen": 2305250, "train_runtime": 84.7105, "train_tokens_per_second": 27213.265}
21
+ {"loss": 0.1811, "grad_norm": 15.1875, "learning_rate": 3.773139652399884e-05, "entropy": 1.846875, "num_tokens": 2405904.0, "mean_token_accuracy": 0.944804173707962, "epoch": 0.39697542533081287, "num_input_tokens_seen": 2420666, "train_runtime": 88.7231, "train_tokens_per_second": 27283.383}
22
+ {"loss": 0.1974, "grad_norm": 6.3125, "learning_rate": 3.743387481413243e-05, "entropy": 1.84765625, "num_tokens": 2520235.0, "mean_token_accuracy": 0.9379207909107208, "epoch": 0.4158790170132325, "num_input_tokens_seen": 2535606, "train_runtime": 93.0343, "train_tokens_per_second": 27254.523}
23
+ {"loss": 0.1266, "grad_norm": 2.875, "learning_rate": 3.711935257616842e-05, "entropy": 1.83984375, "num_tokens": 2634592.0, "mean_token_accuracy": 0.9594786465167999, "epoch": 0.43478260869565216, "num_input_tokens_seen": 2650514, "train_runtime": 98.3865, "train_tokens_per_second": 26939.815}
24
+ {"loss": 0.1993, "grad_norm": 3.6875, "learning_rate": 3.678813651439376e-05, "entropy": 1.78984375, "num_tokens": 2749299.0, "mean_token_accuracy": 0.9459972441196441, "epoch": 0.45368620037807184, "num_input_tokens_seen": 2766004, "train_runtime": 102.0946, "train_tokens_per_second": 27092.567}
25
+ {"loss": 0.2075, "grad_norm": 9.1875, "learning_rate": 3.6440549611967656e-05, "entropy": 1.78828125, "num_tokens": 2863713.0, "mean_token_accuracy": 0.940614128112793, "epoch": 0.4725897920604915, "num_input_tokens_seen": 2880990, "train_runtime": 107.8735, "train_tokens_per_second": 26707.121}
26
+ {"loss": 0.236, "grad_norm": 7.15625, "learning_rate": 3.6076930815966654e-05, "entropy": 1.85234375, "num_tokens": 2978032.0, "mean_token_accuracy": 0.9343804061412812, "epoch": 0.4914933837429111, "num_input_tokens_seen": 2995844, "train_runtime": 111.3362, "train_tokens_per_second": 26908.095}
27
+ {"loss": 0.162, "grad_norm": 7.375, "learning_rate": 3.569763470686262e-05, "entropy": 1.95859375, "num_tokens": 3092605.0, "mean_token_accuracy": 0.9484993875026703, "epoch": 0.5103969754253308, "num_input_tokens_seen": 3111092, "train_runtime": 115.8418, "train_tokens_per_second": 26856.393}
28
+ {"loss": 0.1892, "grad_norm": 7.125, "learning_rate": 3.530303115275597e-05, "entropy": 1.99921875, "num_tokens": 3207190.0, "mean_token_accuracy": 0.9394895970821381, "epoch": 0.5293005671077504, "num_input_tokens_seen": 3226396, "train_runtime": 120.7172, "train_tokens_per_second": 26726.892}
29
+ {"loss": 0.1614, "grad_norm": 4.5625, "learning_rate": 3.4893504948701185e-05, "entropy": 1.96484375, "num_tokens": 3321840.0, "mean_token_accuracy": 0.9600624740123749, "epoch": 0.5482041587901701, "num_input_tokens_seen": 3341802, "train_runtime": 124.4268, "train_tokens_per_second": 26857.576}
30
+ {"loss": 0.1334, "grad_norm": 6.96875, "learning_rate": 3.4469455441476475e-05, "entropy": 1.90859375, "num_tokens": 3436339.0, "mean_token_accuracy": 0.9625543296337128, "epoch": 0.5671077504725898, "num_input_tokens_seen": 3456964, "train_runtime": 130.3081, "train_tokens_per_second": 26529.148}
31
+ {"loss": 0.1427, "grad_norm": 12.4375, "learning_rate": 3.403129614016339e-05, "entropy": 1.92421875, "num_tokens": 3550813.0, "mean_token_accuracy": 0.9588114261627197, "epoch": 0.5860113421550095, "num_input_tokens_seen": 3572084, "train_runtime": 133.8989, "train_tokens_per_second": 26677.47}
32
+ {"loss": 0.2129, "grad_norm": 7.3125, "learning_rate": 3.357945431291618e-05, "entropy": 1.98671875, "num_tokens": 3665300.0, "mean_token_accuracy": 0.9367718935012818, "epoch": 0.6049149338374291, "num_input_tokens_seen": 3687248, "train_runtime": 138.2948, "train_tokens_per_second": 26662.235}
33
+ {"loss": 0.2219, "grad_norm": 3.09375, "learning_rate": 3.311437057031406e-05, "entropy": 2.1359375, "num_tokens": 3779809.0, "mean_token_accuracy": 0.9387097895145416, "epoch": 0.6238185255198487, "num_input_tokens_seen": 3802458, "train_runtime": 142.569, "train_tokens_per_second": 26671.004}
34
+ {"loss": 0.1355, "grad_norm": 4.53125, "learning_rate": 3.263649843570271e-05, "entropy": 2.0859375, "num_tokens": 3894322.0, "mean_token_accuracy": 0.9585716307163239, "epoch": 0.6427221172022685, "num_input_tokens_seen": 3917580, "train_runtime": 145.9767, "train_tokens_per_second": 26837.021}
35
+ {"loss": 0.2962, "grad_norm": 6.53125, "learning_rate": 3.214630390294396e-05, "entropy": 1.946875, "num_tokens": 4008844.0, "mean_token_accuracy": 0.9372412860393524, "epoch": 0.6616257088846881, "num_input_tokens_seen": 4032748, "train_runtime": 151.6027, "train_tokens_per_second": 26600.765}
36
+ {"loss": 0.1841, "grad_norm": 6.96875, "learning_rate": 3.1644264982005e-05, "entropy": 1.98671875, "num_tokens": 4123487.0, "mean_token_accuracy": 0.9490657150745392, "epoch": 0.6805293005671077, "num_input_tokens_seen": 4148142, "train_runtime": 154.9764, "train_tokens_per_second": 26766.274}
37
+ {"loss": 0.124, "grad_norm": 2.953125, "learning_rate": 3.113087123283002e-05, "entropy": 2.021875, "num_tokens": 4238014.0, "mean_token_accuracy": 0.964401924610138, "epoch": 0.6994328922495274, "num_input_tokens_seen": 4263312, "train_runtime": 159.4694, "train_tokens_per_second": 26734.354}
38
+ {"loss": 0.1498, "grad_norm": 3.4375, "learning_rate": 3.060662328794916e-05, "entropy": 1.96171875, "num_tokens": 4352627.0, "mean_token_accuracy": 0.9481843888759613, "epoch": 0.718336483931947, "num_input_tokens_seen": 4378630, "train_runtime": 163.6223, "train_tokens_per_second": 26760.595}
39
+ {"loss": 0.1076, "grad_norm": 4.1875, "learning_rate": 3.0072032364289914e-05, "entropy": 1.9640625, "num_tokens": 4467053.0, "mean_token_accuracy": 0.9691859900951385, "epoch": 0.7372400756143668, "num_input_tokens_seen": 4493600, "train_runtime": 166.9247, "train_tokens_per_second": 26919.915}
40
+ {"loss": 0.2501, "grad_norm": 3.875, "learning_rate": 2.9527619764667376e-05, "entropy": 2.02734375, "num_tokens": 4581812.0, "mean_token_accuracy": 0.9455641567707062, "epoch": 0.7561436672967864, "num_input_tokens_seen": 4609216, "train_runtime": 172.0695, "train_tokens_per_second": 26786.938}
41
+ {"loss": 0.2157, "grad_norm": 5.4375, "learning_rate": 2.8973916369439194e-05, "entropy": 2.14375, "num_tokens": 4696178.0, "mean_token_accuracy": 0.9492439985275268, "epoch": 0.775047258979206, "num_input_tokens_seen": 4724086, "train_runtime": 175.6473, "train_tokens_per_second": 26895.294}
42
+ {"loss": 0.1762, "grad_norm": 3.75, "learning_rate": 2.84114621188211e-05, "entropy": 2.2625, "num_tokens": 4810939.0, "mean_token_accuracy": 0.9574925601482391, "epoch": 0.7939508506616257, "num_input_tokens_seen": 4839702, "train_runtime": 180.4712, "train_tokens_per_second": 26817.036}
43
+ {"loss": 0.1703, "grad_norm": 3.9375, "learning_rate": 2.7840805486367792e-05, "entropy": 2.2953125, "num_tokens": 4925591.0, "mean_token_accuracy": 0.9540181159973145, "epoch": 0.8128544423440454, "num_input_tokens_seen": 4955098, "train_runtime": 184.4177, "train_tokens_per_second": 26868.891}
44
+ {"loss": 0.0938, "grad_norm": 4.625, "learning_rate": 2.7262502944132526e-05, "entropy": 2.2828125, "num_tokens": 5040089.0, "mean_token_accuracy": 0.9725252389907837, "epoch": 0.831758034026465, "num_input_tokens_seen": 5070258, "train_runtime": 188.065, "train_tokens_per_second": 26960.132}
45
+ {"loss": 0.1704, "grad_norm": 5.71875, "learning_rate": 2.667711842002707e-05, "entropy": 2.1265625, "num_tokens": 5154604.0, "mean_token_accuracy": 0.9579161703586578, "epoch": 0.8506616257088847, "num_input_tokens_seen": 5185478, "train_runtime": 192.8301, "train_tokens_per_second": 26891.43}
46
+ {"loss": 0.4284, "grad_norm": 4.3125, "learning_rate": 2.6085222747911155e-05, "entropy": 2.0484375, "num_tokens": 5269357.0, "mean_token_accuracy": 0.9190201222896576, "epoch": 0.8695652173913043, "num_input_tokens_seen": 5301020, "train_runtime": 196.1744, "train_tokens_per_second": 27021.971}
47
+ {"loss": 0.1346, "grad_norm": 13.3125, "learning_rate": 2.5487393110947557e-05, "entropy": 1.98671875, "num_tokens": 5384069.0, "mean_token_accuracy": 0.9579481542110443, "epoch": 0.888468809073724, "num_input_tokens_seen": 5416464, "train_runtime": 201.21, "train_tokens_per_second": 26919.463}
48
+ {"loss": 0.097, "grad_norm": 3.84375, "learning_rate": 2.4884212478765747e-05, "entropy": 1.9875, "num_tokens": 5498568.0, "mean_token_accuracy": 0.9672803819179535, "epoch": 0.9073724007561437, "num_input_tokens_seen": 5531644, "train_runtime": 205.075, "train_tokens_per_second": 26973.766}
49
+ {"loss": 0.2298, "grad_norm": 4.9375, "learning_rate": 2.427626903898292e-05, "entropy": 2.00234375, "num_tokens": 5613157.0, "mean_token_accuracy": 0.9443018674850464, "epoch": 0.9262759924385633, "num_input_tokens_seen": 5646952, "train_runtime": 208.4891, "train_tokens_per_second": 27085.115}
50
+ {"loss": 0.1732, "grad_norm": 6.03125, "learning_rate": 2.3664155623636715e-05, "entropy": 2.0140625, "num_tokens": 5727795.0, "mean_token_accuracy": 0.9442705571651459, "epoch": 0.945179584120983, "num_input_tokens_seen": 5762366, "train_runtime": 214.059, "train_tokens_per_second": 26919.525}
51
+ {"loss": 0.1083, "grad_norm": 3.546875, "learning_rate": 2.304846913108891e-05, "entropy": 2.0125, "num_tokens": 5842437.0, "mean_token_accuracy": 0.9664817750453949, "epoch": 0.9640831758034026, "num_input_tokens_seen": 5877646, "train_runtime": 275.7098, "train_tokens_per_second": 21318.232}
52
+ {"loss": 0.0875, "grad_norm": 2.671875, "learning_rate": 2.242980994396401e-05, "entropy": 2.0, "num_tokens": 5956870.0, "mean_token_accuracy": 0.9795427262783051, "epoch": 0.9829867674858223, "num_input_tokens_seen": 5992710, "train_runtime": 280.9684, "train_tokens_per_second": 21328.766}
53
+ {"loss": 0.1654, "grad_norm": 7.4375, "learning_rate": 2.1808781343690027e-05, "entropy": 1.9513888888888888, "num_tokens": 6059927.0, "mean_token_accuracy": 0.9603289763132731, "epoch": 1.0, "num_input_tokens_seen": 6096342, "train_runtime": 284.3725, "train_tokens_per_second": 21437.877}
54
+ {"loss": 0.0783, "grad_norm": 3.453125, "learning_rate": 2.118598892221257e-05, "entropy": 1.903125, "num_tokens": 6174483.0, "mean_token_accuracy": 0.9817151129245758, "epoch": 1.0189035916824196, "num_input_tokens_seen": 6211574, "train_runtime": 288.2049, "train_tokens_per_second": 21552.63}
55
+ {"loss": 0.1214, "grad_norm": 1.734375, "learning_rate": 2.0562039991455877e-05, "entropy": 1.84375, "num_tokens": 6289163.0, "mean_token_accuracy": 0.9741188943386078, "epoch": 1.0378071833648392, "num_input_tokens_seen": 6327000, "train_runtime": 293.7126, "train_tokens_per_second": 21541.469}
56
+ {"loss": 0.1393, "grad_norm": 3.78125, "learning_rate": 1.99375429911066e-05, "entropy": 1.8421875, "num_tokens": 6403766.0, "mean_token_accuracy": 0.9579156279563904, "epoch": 1.056710775047259, "num_input_tokens_seen": 6442290, "train_runtime": 297.1668, "train_tokens_per_second": 21679.038}
57
+ {"loss": 0.0872, "grad_norm": 3.953125, "learning_rate": 1.931310689529781e-05, "entropy": 1.85078125, "num_tokens": 6518469.0, "mean_token_accuracy": 0.9788394093513488, "epoch": 1.0756143667296787, "num_input_tokens_seen": 6557852, "train_runtime": 301.7702, "train_tokens_per_second": 21731.276}
58
+ {"loss": 0.0637, "grad_norm": 7.1875, "learning_rate": 1.8689340618771937e-05, "entropy": 1.8234375, "num_tokens": 6632963.0, "mean_token_accuracy": 0.972537738084793, "epoch": 1.0945179584120983, "num_input_tokens_seen": 6673032, "train_runtime": 306.4769, "train_tokens_per_second": 21773.362}
59
+ {"loss": 0.0565, "grad_norm": 5.78125, "learning_rate": 1.806685242310156e-05, "entropy": 1.78359375, "num_tokens": 6747403.0, "mean_token_accuracy": 0.9854797184467315, "epoch": 1.113421550094518, "num_input_tokens_seen": 6788174, "train_runtime": 310.3851, "train_tokens_per_second": 21870.17}
60
+ {"loss": 0.0973, "grad_norm": 8.8125, "learning_rate": 1.7446249323547117e-05, "entropy": 1.76015625, "num_tokens": 6861788.0, "mean_token_accuracy": 0.9734237968921662, "epoch": 1.1323251417769375, "num_input_tokens_seen": 6903146, "train_runtime": 315.4655, "train_tokens_per_second": 21882.41}
61
+ {"loss": 0.0681, "grad_norm": 1.4453125, "learning_rate": 1.6828136497130014e-05, "entropy": 1.75078125, "num_tokens": 6976277.0, "mean_token_accuracy": 0.9820096373558045, "epoch": 1.1512287334593574, "num_input_tokens_seen": 7018350, "train_runtime": 319.0527, "train_tokens_per_second": 21997.465}
62
+ {"loss": 0.0625, "grad_norm": 4.90625, "learning_rate": 1.6213116692498206e-05, "entropy": 1.740625, "num_tokens": 7090874.0, "mean_token_accuracy": 0.9826828062534332, "epoch": 1.170132325141777, "num_input_tokens_seen": 7133636, "train_runtime": 323.8986, "train_tokens_per_second": 22024.29}
63
+ {"loss": 0.077, "grad_norm": 0.66015625, "learning_rate": 1.560178964215987e-05, "entropy": 1.7328125, "num_tokens": 7205391.0, "mean_token_accuracy": 0.978941410779953, "epoch": 1.1890359168241966, "num_input_tokens_seen": 7248866, "train_runtime": 327.5895, "train_tokens_per_second": 22127.897}
64
+ {"loss": 0.067, "grad_norm": 4.0625, "learning_rate": 1.4994751477658139e-05, "entropy": 1.73203125, "num_tokens": 7319827.0, "mean_token_accuracy": 0.9818780541419982, "epoch": 1.2079395085066162, "num_input_tokens_seen": 7363900, "train_runtime": 331.4598, "train_tokens_per_second": 22216.571}
65
+ {"loss": 0.1153, "grad_norm": 2.734375, "learning_rate": 1.4392594148257426e-05, "entropy": 1.73515625, "num_tokens": 7434543.0, "mean_token_accuracy": 0.9638942897319793, "epoch": 1.2268431001890359, "num_input_tokens_seen": 7479394, "train_runtime": 336.2629, "train_tokens_per_second": 22242.696}
66
+ {"loss": 0.0359, "grad_norm": 2.046875, "learning_rate": 1.3795904843707959e-05, "entropy": 1.74609375, "num_tokens": 7549134.0, "mean_token_accuracy": 0.9886789560317993, "epoch": 1.2457466918714555, "num_input_tokens_seen": 7594632, "train_runtime": 339.6052, "train_tokens_per_second": 22363.12}
67
+ {"loss": 0.0808, "grad_norm": 2.25, "learning_rate": 1.3205265421651588e-05, "entropy": 1.740625, "num_tokens": 7663583.0, "mean_token_accuracy": 0.9852688193321228, "epoch": 1.264650283553875, "num_input_tokens_seen": 7709704, "train_runtime": 344.9458, "train_tokens_per_second": 22350.48}
68
+ {"loss": 0.0663, "grad_norm": 1.8125, "learning_rate": 1.2621251840227112e-05, "entropy": 1.75078125, "num_tokens": 7778064.0, "mean_token_accuracy": 0.9817369997501373, "epoch": 1.283553875236295, "num_input_tokens_seen": 7824834, "train_runtime": 348.223, "train_tokens_per_second": 22470.756}
69
+ {"loss": 0.0678, "grad_norm": 4.28125, "learning_rate": 1.2044433596428537e-05, "entropy": 1.75234375, "num_tokens": 7892415.0, "mean_token_accuracy": 0.9812626421451569, "epoch": 1.3024574669187146, "num_input_tokens_seen": 7939832, "train_runtime": 352.0847, "train_tokens_per_second": 22550.916}
70
+ {"loss": 0.0465, "grad_norm": 2.703125, "learning_rate": 1.1475373170763819e-05, "entropy": 1.746875, "num_tokens": 8006926.0, "mean_token_accuracy": 0.9823280215263367, "epoch": 1.3213610586011342, "num_input_tokens_seen": 8054988, "train_runtime": 357.1271, "train_tokens_per_second": 22554.962}
71
+ {"loss": 0.1174, "grad_norm": 1.4921875, "learning_rate": 1.0914625478755672e-05, "entropy": 1.74765625, "num_tokens": 8121373.0, "mean_token_accuracy": 0.9695515096187591, "epoch": 1.3402646502835538, "num_input_tokens_seen": 8170098, "train_runtime": 360.7524, "train_tokens_per_second": 22647.381}
72
+ {"loss": 0.045, "grad_norm": 1.1015625, "learning_rate": 1.0362737329819413e-05, "entropy": 1.74453125, "num_tokens": 8235981.0, "mean_token_accuracy": 0.9885900497436524, "epoch": 1.3591682419659734, "num_input_tokens_seen": 8285346, "train_runtime": 366.0216, "train_tokens_per_second": 22636.221}
73
+ {"loss": 0.0428, "grad_norm": 5.15625, "learning_rate": 9.820246894045316e-06, "entropy": 1.74296875, "num_tokens": 8350356.0, "mean_token_accuracy": 0.9822307825088501, "epoch": 1.3780718336483933, "num_input_tokens_seen": 8400240, "train_runtime": 369.6364, "train_tokens_per_second": 22725.685}
74
+ {"loss": 0.099, "grad_norm": 4.09375, "learning_rate": 9.28768317740564e-06, "entropy": 1.73515625, "num_tokens": 8465025.0, "mean_token_accuracy": 0.9710565328598022, "epoch": 1.3969754253308129, "num_input_tokens_seen": 8515740, "train_runtime": 373.5701, "train_tokens_per_second": 22795.56}
75
+ {"loss": 0.0736, "grad_norm": 4.96875, "learning_rate": 8.765565505897902e-06, "entropy": 1.7328125, "num_tokens": 8579648.0, "mean_token_accuracy": 0.9741575241088867, "epoch": 1.4158790170132325, "num_input_tokens_seen": 8631054, "train_runtime": 378.7394, "train_tokens_per_second": 22788.901}
76
+ {"loss": 0.0806, "grad_norm": 3.265625, "learning_rate": 8.254403019127566e-06, "entropy": 1.73359375, "num_tokens": 8694249.0, "mean_token_accuracy": 0.9791056990623475, "epoch": 1.434782608695652, "num_input_tokens_seen": 8746364, "train_runtime": 382.0615, "train_tokens_per_second": 22892.552}
77
+ {"loss": 0.0404, "grad_norm": 3.75, "learning_rate": 7.754694173823947e-06, "entropy": 1.73515625, "num_tokens": 8808789.0, "mean_token_accuracy": 0.9839386224746705, "epoch": 1.4536862003780717, "num_input_tokens_seen": 8861574, "train_runtime": 387.2205, "train_tokens_per_second": 22885.084}
78
+ {"loss": 0.0926, "grad_norm": 5.09375, "learning_rate": 7.266926257773346e-06, "entropy": 1.73359375, "num_tokens": 8923407.0, "mean_token_accuracy": 0.9714232623577118, "epoch": 1.4725897920604916, "num_input_tokens_seen": 8976944, "train_runtime": 390.891, "train_tokens_per_second": 22965.336}
79
+ {"loss": 0.0519, "grad_norm": 5.0, "learning_rate": 6.7915749146436415e-06, "entropy": 1.72265625, "num_tokens": 9037924.0, "mean_token_accuracy": 0.9837916433811188, "epoch": 1.4914933837429112, "num_input_tokens_seen": 9092050, "train_runtime": 395.3397, "train_tokens_per_second": 22998.071}
80
+ {"loss": 0.2115, "grad_norm": 3.875, "learning_rate": 6.329103680163495e-06, "entropy": 1.71796875, "num_tokens": 9152659.0, "mean_token_accuracy": 0.9516554296016693, "epoch": 1.5103969754253308, "num_input_tokens_seen": 9207594, "train_runtime": 399.5499, "train_tokens_per_second": 23044.916}
81
+ {"loss": 0.0348, "grad_norm": 0.51953125, "learning_rate": 5.879963530108506e-06, "entropy": 1.71640625, "num_tokens": 9267059.0, "mean_token_accuracy": 0.9919346511363983, "epoch": 1.5293005671077504, "num_input_tokens_seen": 9322572, "train_runtime": 403.4031, "train_tokens_per_second": 23109.815}
82
+ {"loss": 0.0374, "grad_norm": 2.234375, "learning_rate": 5.444592440535177e-06, "entropy": 1.7125, "num_tokens": 9381725.0, "mean_token_accuracy": 0.9837370038032531, "epoch": 1.54820415879017, "num_input_tokens_seen": 9438004, "train_runtime": 407.9692, "train_tokens_per_second": 23134.111}
83
+ {"loss": 0.0325, "grad_norm": 3.09375, "learning_rate": 5.023414960691469e-06, "entropy": 1.7078125, "num_tokens": 9496255.0, "mean_token_accuracy": 0.9918534696102143, "epoch": 1.5671077504725899, "num_input_tokens_seen": 9553156, "train_runtime": 412.1408, "train_tokens_per_second": 23179.35}
84
+ {"loss": 0.0618, "grad_norm": 4.78125, "learning_rate": 4.616841799020364e-06, "entropy": 1.703125, "num_tokens": 9610808.0, "mean_token_accuracy": 0.9808044970035553, "epoch": 1.5860113421550095, "num_input_tokens_seen": 9668364, "train_runtime": 416.3235, "train_tokens_per_second": 23223.203}
85
+ {"loss": 0.0493, "grad_norm": 6.1875, "learning_rate": 4.225269422660258e-06, "entropy": 1.703125, "num_tokens": 9725283.0, "mean_token_accuracy": 0.9843941271305084, "epoch": 1.6049149338374291, "num_input_tokens_seen": 9783552, "train_runtime": 421.1316, "train_tokens_per_second": 23231.576}
86
+ {"loss": 0.0595, "grad_norm": 1.5078125, "learning_rate": 3.8490796708326404e-06, "entropy": 1.70390625, "num_tokens": 9839878.0, "mean_token_accuracy": 0.9822299420833588, "epoch": 1.6238185255198487, "num_input_tokens_seen": 9898934, "train_runtime": 424.7606, "train_tokens_per_second": 23304.735}
87
+ {"loss": 0.059, "grad_norm": 1.328125, "learning_rate": 3.4886393824940924e-06, "entropy": 1.7015625, "num_tokens": 9954403.0, "mean_token_accuracy": 0.9807979345321656, "epoch": 1.6427221172022684, "num_input_tokens_seen": 10014142, "train_runtime": 429.8927, "train_tokens_per_second": 23294.514}
88
+ {"loss": 0.0574, "grad_norm": 2.09375, "learning_rate": 3.144300038615691e-06, "entropy": 1.70390625, "num_tokens": 10068933.0, "mean_token_accuracy": 0.9839386105537414, "epoch": 1.6616257088846882, "num_input_tokens_seen": 10129264, "train_runtime": 433.4828, "train_tokens_per_second": 23367.164}
89
+ {"loss": 0.0669, "grad_norm": 4.25, "learning_rate": 2.8163974194386766e-06, "entropy": 1.6984375, "num_tokens": 10183591.0, "mean_token_accuracy": 0.9792383193969727, "epoch": 1.6805293005671076, "num_input_tokens_seen": 10244732, "train_runtime": 437.9792, "train_tokens_per_second": 23390.909}
90
+ {"loss": 0.0801, "grad_norm": 3.46875, "learning_rate": 2.5052512770405434e-06, "entropy": 1.7015625, "num_tokens": 10298251.0, "mean_token_accuracy": 0.9761136710643769, "epoch": 1.6994328922495274, "num_input_tokens_seen": 10360212, "train_runtime": 442.481, "train_tokens_per_second": 23413.915}
91
+ {"loss": 0.0297, "grad_norm": 0.59765625, "learning_rate": 2.2111650235309147e-06, "entropy": 1.70234375, "num_tokens": 10412810.0, "mean_token_accuracy": 0.9904489517211914, "epoch": 1.718336483931947, "num_input_tokens_seen": 10475400, "train_runtime": 446.3738, "train_tokens_per_second": 23467.773}
92
+ {"loss": 0.0989, "grad_norm": 4.0625, "learning_rate": 1.9344254351812287e-06, "entropy": 1.69921875, "num_tokens": 10527389.0, "mean_token_accuracy": 0.9743396818637848, "epoch": 1.7372400756143667, "num_input_tokens_seen": 10590710, "train_runtime": 451.1755, "train_tokens_per_second": 23473.591}
93
+ {"loss": 0.0476, "grad_norm": 0.890625, "learning_rate": 1.6753023727767436e-06, "entropy": 1.703125, "num_tokens": 10641918.0, "mean_token_accuracy": 0.9838890075683594, "epoch": 1.7561436672967865, "num_input_tokens_seen": 10705900, "train_runtime": 454.754, "train_tokens_per_second": 23542.179}
94
+ {"loss": 0.0556, "grad_norm": 1.359375, "learning_rate": 1.4340485184635712e-06, "entropy": 1.7, "num_tokens": 10756496.0, "mean_token_accuracy": 0.9777659058570862, "epoch": 1.775047258979206, "num_input_tokens_seen": 10821144, "train_runtime": 459.2027, "train_tokens_per_second": 23565.072}
95
+ {"loss": 0.0595, "grad_norm": 1.171875, "learning_rate": 1.2108991293473627e-06, "entropy": 1.69921875, "num_tokens": 10871124.0, "mean_token_accuracy": 0.9741835057735443, "epoch": 1.7939508506616257, "num_input_tokens_seen": 10936460, "train_runtime": 463.6099, "train_tokens_per_second": 23589.79}
96
+ {"loss": 0.0541, "grad_norm": 3.265625, "learning_rate": 1.0060718080838683e-06, "entropy": 1.69765625, "num_tokens": 10985594.0, "mean_token_accuracy": 0.9831156551837921, "epoch": 1.8128544423440454, "num_input_tokens_seen": 11051508, "train_runtime": 467.1593, "train_tokens_per_second": 23656.828}
97
+ {"loss": 0.0835, "grad_norm": 2.4375, "learning_rate": 8.197662906851534e-07, "entropy": 1.70078125, "num_tokens": 11100230.0, "mean_token_accuracy": 0.9726030707359314, "epoch": 1.831758034026465, "num_input_tokens_seen": 11166904, "train_runtime": 472.195, "train_tokens_per_second": 23648.922}
98
+ {"loss": 0.0532, "grad_norm": 2.765625, "learning_rate": 6.521642517483573e-07, "entropy": 1.69921875, "num_tokens": 11214624.0, "mean_token_accuracy": 0.9853454470634461, "epoch": 1.8506616257088848, "num_input_tokens_seen": 11281802, "train_runtime": 475.7718, "train_tokens_per_second": 23712.635}
99
+ {"loss": 0.027, "grad_norm": 2.171875, "learning_rate": 5.034291272968772e-07, "entropy": 1.70078125, "num_tokens": 11329098.0, "mean_token_accuracy": 0.9934648215770722, "epoch": 1.8695652173913042, "num_input_tokens_seen": 11396946, "train_runtime": 480.2436, "train_tokens_per_second": 23731.596}
100
+ {"loss": 0.0742, "grad_norm": 4.0625, "learning_rate": 3.737059554068334e-07, "entropy": 1.6984375, "num_tokens": 11443715.0, "mean_token_accuracy": 0.9744843065738678, "epoch": 1.888468809073724, "num_input_tokens_seen": 11512282, "train_runtime": 484.6792, "train_tokens_per_second": 23752.376}
101
+ {"loss": 0.1322, "grad_norm": 6.84375, "learning_rate": 2.631212347741352e-07, "entropy": 1.69921875, "num_tokens": 11558513.0, "mean_token_accuracy": 0.9680740118026734, "epoch": 1.9073724007561437, "num_input_tokens_seen": 11627828, "train_runtime": 544.5283, "train_tokens_per_second": 21353.945}
102
+ {"loss": 0.0864, "grad_norm": 1.0078125, "learning_rate": 1.7178280136011417e-07, "entropy": 1.69921875, "num_tokens": 11673010.0, "mean_token_accuracy": 0.9749818980693817, "epoch": 1.9262759924385633, "num_input_tokens_seen": 11743010, "train_runtime": 549.7569, "train_tokens_per_second": 21360.369}
103
+ {"loss": 0.1175, "grad_norm": 2.5625, "learning_rate": 9.977972323599095e-08, "entropy": 1.69921875, "num_tokens": 11787637.0, "mean_token_accuracy": 0.9680160820484162, "epoch": 1.9451795841209831, "num_input_tokens_seen": 11858430, "train_runtime": 553.6509, "train_tokens_per_second": 21418.605}
104
+ {"loss": 0.0695, "grad_norm": 2.921875, "learning_rate": 4.718221372874254e-08, "entropy": 1.69765625, "num_tokens": 11902111.0, "mean_token_accuracy": 0.9804269134998321, "epoch": 1.9640831758034025, "num_input_tokens_seen": 11973576, "train_runtime": 557.8609, "train_tokens_per_second": 21463.371}
105
+ {"loss": 0.1152, "grad_norm": 5.8125, "learning_rate": 1.4041562953031051e-08, "entropy": 1.69609375, "num_tokens": 12016759.0, "mean_token_accuracy": 0.9696780204772949, "epoch": 1.9829867674858224, "num_input_tokens_seen": 12088990, "train_runtime": 561.9991, "train_tokens_per_second": 21510.694}
106
+ {"loss": 0.0989, "grad_norm": 3.75, "learning_rate": 3.900877959917004e-10, "entropy": 1.6961805555555556, "num_tokens": 12119827.0, "mean_token_accuracy": 0.9715293182267083, "epoch": 2.0, "num_input_tokens_seen": 12192662, "train_runtime": 565.5622, "train_tokens_per_second": 21558.482}
107
+ {"train_runtime": 612.9949, "train_samples_per_second": 27.592, "train_steps_per_second": 0.865, "train_tokens_per_second": 2486.879, "total_flos": 3.3226637176733696e+16, "train_loss": 0.1822078584218925, "epoch": 2.0, "num_input_tokens_seen": 12192662}
special_tokens_map.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<|im_start|>",
4
+ "<|im_end|>",
5
+ "<|object_ref_start|>",
6
+ "<|object_ref_end|>",
7
+ "<|box_start|>",
8
+ "<|box_end|>",
9
+ "<|quad_start|>",
10
+ "<|quad_end|>",
11
+ "<|vision_start|>",
12
+ "<|vision_end|>",
13
+ "<|vision_pad|>",
14
+ "<|image_pad|>",
15
+ "<|video_pad|>"
16
+ ],
17
+ "eos_token": {
18
+ "content": "<|im_end|>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ },
24
+ "pad_token": {
25
+ "content": "<|endoftext|>",
26
+ "lstrip": false,
27
+ "normalized": false,
28
+ "rstrip": false,
29
+ "single_word": false
30
+ }
31
+ }
tokenizer_config.json ADDED
@@ -0,0 +1,239 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": false,
3
+ "add_prefix_space": false,
4
+ "added_tokens_decoder": {
5
+ "151643": {
6
+ "content": "<|endoftext|>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "151644": {
14
+ "content": "<|im_start|>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "151645": {
22
+ "content": "<|im_end|>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ },
29
+ "151646": {
30
+ "content": "<|object_ref_start|>",
31
+ "lstrip": false,
32
+ "normalized": false,
33
+ "rstrip": false,
34
+ "single_word": false,
35
+ "special": true
36
+ },
37
+ "151647": {
38
+ "content": "<|object_ref_end|>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false,
43
+ "special": true
44
+ },
45
+ "151648": {
46
+ "content": "<|box_start|>",
47
+ "lstrip": false,
48
+ "normalized": false,
49
+ "rstrip": false,
50
+ "single_word": false,
51
+ "special": true
52
+ },
53
+ "151649": {
54
+ "content": "<|box_end|>",
55
+ "lstrip": false,
56
+ "normalized": false,
57
+ "rstrip": false,
58
+ "single_word": false,
59
+ "special": true
60
+ },
61
+ "151650": {
62
+ "content": "<|quad_start|>",
63
+ "lstrip": false,
64
+ "normalized": false,
65
+ "rstrip": false,
66
+ "single_word": false,
67
+ "special": true
68
+ },
69
+ "151651": {
70
+ "content": "<|quad_end|>",
71
+ "lstrip": false,
72
+ "normalized": false,
73
+ "rstrip": false,
74
+ "single_word": false,
75
+ "special": true
76
+ },
77
+ "151652": {
78
+ "content": "<|vision_start|>",
79
+ "lstrip": false,
80
+ "normalized": false,
81
+ "rstrip": false,
82
+ "single_word": false,
83
+ "special": true
84
+ },
85
+ "151653": {
86
+ "content": "<|vision_end|>",
87
+ "lstrip": false,
88
+ "normalized": false,
89
+ "rstrip": false,
90
+ "single_word": false,
91
+ "special": true
92
+ },
93
+ "151654": {
94
+ "content": "<|vision_pad|>",
95
+ "lstrip": false,
96
+ "normalized": false,
97
+ "rstrip": false,
98
+ "single_word": false,
99
+ "special": true
100
+ },
101
+ "151655": {
102
+ "content": "<|image_pad|>",
103
+ "lstrip": false,
104
+ "normalized": false,
105
+ "rstrip": false,
106
+ "single_word": false,
107
+ "special": true
108
+ },
109
+ "151656": {
110
+ "content": "<|video_pad|>",
111
+ "lstrip": false,
112
+ "normalized": false,
113
+ "rstrip": false,
114
+ "single_word": false,
115
+ "special": true
116
+ },
117
+ "151657": {
118
+ "content": "<tool_call>",
119
+ "lstrip": false,
120
+ "normalized": false,
121
+ "rstrip": false,
122
+ "single_word": false,
123
+ "special": false
124
+ },
125
+ "151658": {
126
+ "content": "</tool_call>",
127
+ "lstrip": false,
128
+ "normalized": false,
129
+ "rstrip": false,
130
+ "single_word": false,
131
+ "special": false
132
+ },
133
+ "151659": {
134
+ "content": "<|fim_prefix|>",
135
+ "lstrip": false,
136
+ "normalized": false,
137
+ "rstrip": false,
138
+ "single_word": false,
139
+ "special": false
140
+ },
141
+ "151660": {
142
+ "content": "<|fim_middle|>",
143
+ "lstrip": false,
144
+ "normalized": false,
145
+ "rstrip": false,
146
+ "single_word": false,
147
+ "special": false
148
+ },
149
+ "151661": {
150
+ "content": "<|fim_suffix|>",
151
+ "lstrip": false,
152
+ "normalized": false,
153
+ "rstrip": false,
154
+ "single_word": false,
155
+ "special": false
156
+ },
157
+ "151662": {
158
+ "content": "<|fim_pad|>",
159
+ "lstrip": false,
160
+ "normalized": false,
161
+ "rstrip": false,
162
+ "single_word": false,
163
+ "special": false
164
+ },
165
+ "151663": {
166
+ "content": "<|repo_name|>",
167
+ "lstrip": false,
168
+ "normalized": false,
169
+ "rstrip": false,
170
+ "single_word": false,
171
+ "special": false
172
+ },
173
+ "151664": {
174
+ "content": "<|file_sep|>",
175
+ "lstrip": false,
176
+ "normalized": false,
177
+ "rstrip": false,
178
+ "single_word": false,
179
+ "special": false
180
+ },
181
+ "151665": {
182
+ "content": "<tool_response>",
183
+ "lstrip": false,
184
+ "normalized": false,
185
+ "rstrip": false,
186
+ "single_word": false,
187
+ "special": false
188
+ },
189
+ "151666": {
190
+ "content": "</tool_response>",
191
+ "lstrip": false,
192
+ "normalized": false,
193
+ "rstrip": false,
194
+ "single_word": false,
195
+ "special": false
196
+ },
197
+ "151667": {
198
+ "content": "<think>",
199
+ "lstrip": false,
200
+ "normalized": false,
201
+ "rstrip": false,
202
+ "single_word": false,
203
+ "special": false
204
+ },
205
+ "151668": {
206
+ "content": "</think>",
207
+ "lstrip": false,
208
+ "normalized": false,
209
+ "rstrip": false,
210
+ "single_word": false,
211
+ "special": false
212
+ }
213
+ },
214
+ "additional_special_tokens": [
215
+ "<|im_start|>",
216
+ "<|im_end|>",
217
+ "<|object_ref_start|>",
218
+ "<|object_ref_end|>",
219
+ "<|box_start|>",
220
+ "<|box_end|>",
221
+ "<|quad_start|>",
222
+ "<|quad_end|>",
223
+ "<|vision_start|>",
224
+ "<|vision_end|>",
225
+ "<|vision_pad|>",
226
+ "<|image_pad|>",
227
+ "<|video_pad|>"
228
+ ],
229
+ "bos_token": null,
230
+ "clean_up_tokenization_spaces": false,
231
+ "eos_token": "<|im_end|>",
232
+ "errors": "replace",
233
+ "extra_special_tokens": {},
234
+ "model_max_length": 32768,
235
+ "pad_token": "<|endoftext|>",
236
+ "split_special_tokens": false,
237
+ "tokenizer_class": "Qwen2Tokenizer",
238
+ "unk_token": null
239
+ }
trainer_state.json ADDED
@@ -0,0 +1,1423 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 2.0,
6
+ "eval_steps": 250,
7
+ "global_step": 530,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "entropy": 0.825,
14
+ "epoch": 0.01890359168241966,
15
+ "grad_norm": 157.0,
16
+ "learning_rate": 5.925925925925926e-06,
17
+ "loss": 2.9876,
18
+ "mean_token_accuracy": 0.6893173575401306,
19
+ "num_input_tokens_seen": 115216,
20
+ "num_tokens": 114489.0,
21
+ "step": 5,
22
+ "train_runtime": 4.4379,
23
+ "train_tokens_per_second": 25961.602
24
+ },
25
+ {
26
+ "entropy": 0.8890625,
27
+ "epoch": 0.03780718336483932,
28
+ "grad_norm": 24.5,
29
+ "learning_rate": 1.3333333333333333e-05,
30
+ "loss": 0.6662,
31
+ "mean_token_accuracy": 0.8326915562152862,
32
+ "num_input_tokens_seen": 230592,
33
+ "num_tokens": 229109.0,
34
+ "step": 10,
35
+ "train_runtime": 7.7641,
36
+ "train_tokens_per_second": 29699.859
37
+ },
38
+ {
39
+ "entropy": 1.02890625,
40
+ "epoch": 0.05671077504725898,
41
+ "grad_norm": 18.625,
42
+ "learning_rate": 2.074074074074074e-05,
43
+ "loss": 0.6027,
44
+ "mean_token_accuracy": 0.8529165983200073,
45
+ "num_input_tokens_seen": 345600,
46
+ "num_tokens": 343545.0,
47
+ "step": 15,
48
+ "train_runtime": 13.2817,
49
+ "train_tokens_per_second": 26020.836
50
+ },
51
+ {
52
+ "entropy": 1.1421875,
53
+ "epoch": 0.07561436672967864,
54
+ "grad_norm": 19.125,
55
+ "learning_rate": 2.814814814814815e-05,
56
+ "loss": 0.4447,
57
+ "mean_token_accuracy": 0.881816154718399,
58
+ "num_input_tokens_seen": 461282,
59
+ "num_tokens": 458335.0,
60
+ "step": 20,
61
+ "train_runtime": 16.8766,
62
+ "train_tokens_per_second": 27332.56
63
+ },
64
+ {
65
+ "entropy": 1.21640625,
66
+ "epoch": 0.0945179584120983,
67
+ "grad_norm": 21.75,
68
+ "learning_rate": 3.555555555555555e-05,
69
+ "loss": 0.3989,
70
+ "mean_token_accuracy": 0.8929793298244476,
71
+ "num_input_tokens_seen": 576346,
72
+ "num_tokens": 572824.0,
73
+ "step": 25,
74
+ "train_runtime": 22.5256,
75
+ "train_tokens_per_second": 25586.277
76
+ },
77
+ {
78
+ "entropy": 1.26484375,
79
+ "epoch": 0.11342155009451796,
80
+ "grad_norm": 13.625,
81
+ "learning_rate": 3.999843966403289e-05,
82
+ "loss": 0.4872,
83
+ "mean_token_accuracy": 0.8781549751758575,
84
+ "num_input_tokens_seen": 691188,
85
+ "num_tokens": 687152.0,
86
+ "step": 30,
87
+ "train_runtime": 26.0077,
88
+ "train_tokens_per_second": 26576.234
89
+ },
90
+ {
91
+ "entropy": 1.28515625,
92
+ "epoch": 0.1323251417769376,
93
+ "grad_norm": 17.375,
94
+ "learning_rate": 3.99808886803243e-05,
95
+ "loss": 0.28,
96
+ "mean_token_accuracy": 0.9074305832386017,
97
+ "num_input_tokens_seen": 806740,
98
+ "num_tokens": 801973.0,
99
+ "step": 35,
100
+ "train_runtime": 29.7225,
101
+ "train_tokens_per_second": 27142.403
102
+ },
103
+ {
104
+ "entropy": 1.2984375,
105
+ "epoch": 0.15122873345935728,
106
+ "grad_norm": 12.625,
107
+ "learning_rate": 3.994385346473689e-05,
108
+ "loss": 0.356,
109
+ "mean_token_accuracy": 0.9146295249462127,
110
+ "num_input_tokens_seen": 921796,
111
+ "num_tokens": 916426.0,
112
+ "step": 40,
113
+ "train_runtime": 34.6767,
114
+ "train_tokens_per_second": 26582.553
115
+ },
116
+ {
117
+ "entropy": 1.56953125,
118
+ "epoch": 0.17013232514177692,
119
+ "grad_norm": 11.4375,
120
+ "learning_rate": 3.9887370131917e-05,
121
+ "loss": 0.3933,
122
+ "mean_token_accuracy": 0.9064954161643982,
123
+ "num_input_tokens_seen": 1036824,
124
+ "num_tokens": 1030824.0,
125
+ "step": 45,
126
+ "train_runtime": 38.0075,
127
+ "train_tokens_per_second": 27279.472
128
+ },
129
+ {
130
+ "entropy": 1.7515625,
131
+ "epoch": 0.1890359168241966,
132
+ "grad_norm": 15.1875,
133
+ "learning_rate": 3.981149376121427e-05,
134
+ "loss": 0.2873,
135
+ "mean_token_accuracy": 0.9260397672653198,
136
+ "num_input_tokens_seen": 1152356,
137
+ "num_tokens": 1145500.0,
138
+ "step": 50,
139
+ "train_runtime": 43.1792,
140
+ "train_tokens_per_second": 26687.759
141
+ },
142
+ {
143
+ "entropy": 1.75859375,
144
+ "epoch": 0.20793950850661624,
145
+ "grad_norm": 8.5625,
146
+ "learning_rate": 3.97162983429714e-05,
147
+ "loss": 0.3322,
148
+ "mean_token_accuracy": 0.9256749033927918,
149
+ "num_input_tokens_seen": 1267634,
150
+ "num_tokens": 1260057.0,
151
+ "step": 55,
152
+ "train_runtime": 46.6166,
153
+ "train_tokens_per_second": 27192.755
154
+ },
155
+ {
156
+ "entropy": 1.734375,
157
+ "epoch": 0.22684310018903592,
158
+ "grad_norm": 11.6875,
159
+ "learning_rate": 3.960187670637294e-05,
160
+ "loss": 0.2865,
161
+ "mean_token_accuracy": 0.9282522916793823,
162
+ "num_input_tokens_seen": 1383494,
163
+ "num_tokens": 1374973.0,
164
+ "step": 60,
165
+ "train_runtime": 50.5007,
166
+ "train_tokens_per_second": 27395.522
167
+ },
168
+ {
169
+ "entropy": 1.70859375,
170
+ "epoch": 0.24574669187145556,
171
+ "grad_norm": 9.75,
172
+ "learning_rate": 3.946834042892355e-05,
173
+ "loss": 0.2277,
174
+ "mean_token_accuracy": 0.9320353448390961,
175
+ "num_input_tokens_seen": 1499052,
176
+ "num_tokens": 1489683.0,
177
+ "step": 65,
178
+ "train_runtime": 55.2907,
179
+ "train_tokens_per_second": 27112.206
180
+ },
181
+ {
182
+ "entropy": 1.6578125,
183
+ "epoch": 0.2646502835538752,
184
+ "grad_norm": 10.125,
185
+ "learning_rate": 3.931581972764386e-05,
186
+ "loss": 0.2733,
187
+ "mean_token_accuracy": 0.9363594233989716,
188
+ "num_input_tokens_seen": 1614146,
189
+ "num_tokens": 1604106.0,
190
+ "step": 70,
191
+ "train_runtime": 58.5263,
192
+ "train_tokens_per_second": 27579.832
193
+ },
194
+ {
195
+ "entropy": 1.57109375,
196
+ "epoch": 0.2835538752362949,
197
+ "grad_norm": 6.34375,
198
+ "learning_rate": 3.91444633320903e-05,
199
+ "loss": 0.2164,
200
+ "mean_token_accuracy": 0.9349239528179168,
201
+ "num_input_tokens_seen": 1729362,
202
+ "num_tokens": 1718632.0,
203
+ "step": 75,
204
+ "train_runtime": 63.5177,
205
+ "train_tokens_per_second": 27226.454
206
+ },
207
+ {
208
+ "entropy": 1.57734375,
209
+ "epoch": 0.30245746691871456,
210
+ "grad_norm": 10.875,
211
+ "learning_rate": 3.8954438339322366e-05,
212
+ "loss": 0.2173,
213
+ "mean_token_accuracy": 0.9350460767745972,
214
+ "num_input_tokens_seen": 1844444,
215
+ "num_tokens": 1833068.0,
216
+ "step": 80,
217
+ "train_runtime": 66.8194,
218
+ "train_tokens_per_second": 27603.402
219
+ },
220
+ {
221
+ "entropy": 1.62421875,
222
+ "epoch": 0.32136105860113423,
223
+ "grad_norm": 11.0,
224
+ "learning_rate": 3.874593005095909e-05,
225
+ "loss": 0.2337,
226
+ "mean_token_accuracy": 0.929820317029953,
227
+ "num_input_tokens_seen": 1959682,
228
+ "num_tokens": 1947640.0,
229
+ "step": 85,
230
+ "train_runtime": 70.4744,
231
+ "train_tokens_per_second": 27806.99
232
+ },
233
+ {
234
+ "entropy": 1.71171875,
235
+ "epoch": 0.34026465028355385,
236
+ "grad_norm": 8.625,
237
+ "learning_rate": 3.851914179248333e-05,
238
+ "loss": 0.2156,
239
+ "mean_token_accuracy": 0.9308744966983795,
240
+ "num_input_tokens_seen": 2075138,
241
+ "num_tokens": 2062310.0,
242
+ "step": 90,
243
+ "train_runtime": 75.9345,
244
+ "train_tokens_per_second": 27327.991
245
+ },
246
+ {
247
+ "entropy": 1.86875,
248
+ "epoch": 0.3591682419659735,
249
+ "grad_norm": 13.5625,
250
+ "learning_rate": 3.82742947149703e-05,
251
+ "loss": 0.2718,
252
+ "mean_token_accuracy": 0.9264281988143921,
253
+ "num_input_tokens_seen": 2190160,
254
+ "num_tokens": 2176716.0,
255
+ "step": 95,
256
+ "train_runtime": 79.4416,
257
+ "train_tokens_per_second": 27569.42
258
+ },
259
+ {
260
+ "entropy": 1.94765625,
261
+ "epoch": 0.3780718336483932,
262
+ "grad_norm": 6.125,
263
+ "learning_rate": 3.801162757943359e-05,
264
+ "loss": 0.3385,
265
+ "mean_token_accuracy": 0.9164456725120544,
266
+ "num_input_tokens_seen": 2305250,
267
+ "num_tokens": 2291230.0,
268
+ "step": 100,
269
+ "train_runtime": 84.7105,
270
+ "train_tokens_per_second": 27213.265
271
+ },
272
+ {
273
+ "entropy": 1.846875,
274
+ "epoch": 0.39697542533081287,
275
+ "grad_norm": 15.1875,
276
+ "learning_rate": 3.773139652399884e-05,
277
+ "loss": 0.1811,
278
+ "mean_token_accuracy": 0.944804173707962,
279
+ "num_input_tokens_seen": 2420666,
280
+ "num_tokens": 2405904.0,
281
+ "step": 105,
282
+ "train_runtime": 88.7231,
283
+ "train_tokens_per_second": 27283.383
284
+ },
285
+ {
286
+ "entropy": 1.84765625,
287
+ "epoch": 0.4158790170132325,
288
+ "grad_norm": 6.3125,
289
+ "learning_rate": 3.743387481413243e-05,
290
+ "loss": 0.1974,
291
+ "mean_token_accuracy": 0.9379207909107208,
292
+ "num_input_tokens_seen": 2535606,
293
+ "num_tokens": 2520235.0,
294
+ "step": 110,
295
+ "train_runtime": 93.0343,
296
+ "train_tokens_per_second": 27254.523
297
+ },
298
+ {
299
+ "entropy": 1.83984375,
300
+ "epoch": 0.43478260869565216,
301
+ "grad_norm": 2.875,
302
+ "learning_rate": 3.711935257616842e-05,
303
+ "loss": 0.1266,
304
+ "mean_token_accuracy": 0.9594786465167999,
305
+ "num_input_tokens_seen": 2650514,
306
+ "num_tokens": 2634592.0,
307
+ "step": 115,
308
+ "train_runtime": 98.3865,
309
+ "train_tokens_per_second": 26939.815
310
+ },
311
+ {
312
+ "entropy": 1.78984375,
313
+ "epoch": 0.45368620037807184,
314
+ "grad_norm": 3.6875,
315
+ "learning_rate": 3.678813651439376e-05,
316
+ "loss": 0.1993,
317
+ "mean_token_accuracy": 0.9459972441196441,
318
+ "num_input_tokens_seen": 2766004,
319
+ "num_tokens": 2749299.0,
320
+ "step": 120,
321
+ "train_runtime": 102.0946,
322
+ "train_tokens_per_second": 27092.567
323
+ },
324
+ {
325
+ "entropy": 1.78828125,
326
+ "epoch": 0.4725897920604915,
327
+ "grad_norm": 9.1875,
328
+ "learning_rate": 3.6440549611967656e-05,
329
+ "loss": 0.2075,
330
+ "mean_token_accuracy": 0.940614128112793,
331
+ "num_input_tokens_seen": 2880990,
332
+ "num_tokens": 2863713.0,
333
+ "step": 125,
334
+ "train_runtime": 107.8735,
335
+ "train_tokens_per_second": 26707.121
336
+ },
337
+ {
338
+ "entropy": 1.85234375,
339
+ "epoch": 0.4914933837429111,
340
+ "grad_norm": 7.15625,
341
+ "learning_rate": 3.6076930815966654e-05,
342
+ "loss": 0.236,
343
+ "mean_token_accuracy": 0.9343804061412812,
344
+ "num_input_tokens_seen": 2995844,
345
+ "num_tokens": 2978032.0,
346
+ "step": 130,
347
+ "train_runtime": 111.3362,
348
+ "train_tokens_per_second": 26908.095
349
+ },
350
+ {
351
+ "entropy": 1.95859375,
352
+ "epoch": 0.5103969754253308,
353
+ "grad_norm": 7.375,
354
+ "learning_rate": 3.569763470686262e-05,
355
+ "loss": 0.162,
356
+ "mean_token_accuracy": 0.9484993875026703,
357
+ "num_input_tokens_seen": 3111092,
358
+ "num_tokens": 3092605.0,
359
+ "step": 135,
360
+ "train_runtime": 115.8418,
361
+ "train_tokens_per_second": 26856.393
362
+ },
363
+ {
364
+ "entropy": 1.99921875,
365
+ "epoch": 0.5293005671077504,
366
+ "grad_norm": 7.125,
367
+ "learning_rate": 3.530303115275597e-05,
368
+ "loss": 0.1892,
369
+ "mean_token_accuracy": 0.9394895970821381,
370
+ "num_input_tokens_seen": 3226396,
371
+ "num_tokens": 3207190.0,
372
+ "step": 140,
373
+ "train_runtime": 120.7172,
374
+ "train_tokens_per_second": 26726.892
375
+ },
376
+ {
377
+ "entropy": 1.96484375,
378
+ "epoch": 0.5482041587901701,
379
+ "grad_norm": 4.5625,
380
+ "learning_rate": 3.4893504948701185e-05,
381
+ "loss": 0.1614,
382
+ "mean_token_accuracy": 0.9600624740123749,
383
+ "num_input_tokens_seen": 3341802,
384
+ "num_tokens": 3321840.0,
385
+ "step": 145,
386
+ "train_runtime": 124.4268,
387
+ "train_tokens_per_second": 26857.576
388
+ },
389
+ {
390
+ "entropy": 1.90859375,
391
+ "epoch": 0.5671077504725898,
392
+ "grad_norm": 6.96875,
393
+ "learning_rate": 3.4469455441476475e-05,
394
+ "loss": 0.1334,
395
+ "mean_token_accuracy": 0.9625543296337128,
396
+ "num_input_tokens_seen": 3456964,
397
+ "num_tokens": 3436339.0,
398
+ "step": 150,
399
+ "train_runtime": 130.3081,
400
+ "train_tokens_per_second": 26529.148
401
+ },
402
+ {
403
+ "entropy": 1.92421875,
404
+ "epoch": 0.5860113421550095,
405
+ "grad_norm": 12.4375,
406
+ "learning_rate": 3.403129614016339e-05,
407
+ "loss": 0.1427,
408
+ "mean_token_accuracy": 0.9588114261627197,
409
+ "num_input_tokens_seen": 3572084,
410
+ "num_tokens": 3550813.0,
411
+ "step": 155,
412
+ "train_runtime": 133.8989,
413
+ "train_tokens_per_second": 26677.47
414
+ },
415
+ {
416
+ "entropy": 1.98671875,
417
+ "epoch": 0.6049149338374291,
418
+ "grad_norm": 7.3125,
419
+ "learning_rate": 3.357945431291618e-05,
420
+ "loss": 0.2129,
421
+ "mean_token_accuracy": 0.9367718935012818,
422
+ "num_input_tokens_seen": 3687248,
423
+ "num_tokens": 3665300.0,
424
+ "step": 160,
425
+ "train_runtime": 138.2948,
426
+ "train_tokens_per_second": 26662.235
427
+ },
428
+ {
429
+ "entropy": 2.1359375,
430
+ "epoch": 0.6238185255198487,
431
+ "grad_norm": 3.09375,
432
+ "learning_rate": 3.311437057031406e-05,
433
+ "loss": 0.2219,
434
+ "mean_token_accuracy": 0.9387097895145416,
435
+ "num_input_tokens_seen": 3802458,
436
+ "num_tokens": 3779809.0,
437
+ "step": 165,
438
+ "train_runtime": 142.569,
439
+ "train_tokens_per_second": 26671.004
440
+ },
441
+ {
442
+ "entropy": 2.0859375,
443
+ "epoch": 0.6427221172022685,
444
+ "grad_norm": 4.53125,
445
+ "learning_rate": 3.263649843570271e-05,
446
+ "loss": 0.1355,
447
+ "mean_token_accuracy": 0.9585716307163239,
448
+ "num_input_tokens_seen": 3917580,
449
+ "num_tokens": 3894322.0,
450
+ "step": 170,
451
+ "train_runtime": 145.9767,
452
+ "train_tokens_per_second": 26837.021
453
+ },
454
+ {
455
+ "entropy": 1.946875,
456
+ "epoch": 0.6616257088846881,
457
+ "grad_norm": 6.53125,
458
+ "learning_rate": 3.214630390294396e-05,
459
+ "loss": 0.2962,
460
+ "mean_token_accuracy": 0.9372412860393524,
461
+ "num_input_tokens_seen": 4032748,
462
+ "num_tokens": 4008844.0,
463
+ "step": 175,
464
+ "train_runtime": 151.6027,
465
+ "train_tokens_per_second": 26600.765
466
+ },
467
+ {
468
+ "entropy": 1.98671875,
469
+ "epoch": 0.6805293005671077,
470
+ "grad_norm": 6.96875,
471
+ "learning_rate": 3.1644264982005e-05,
472
+ "loss": 0.1841,
473
+ "mean_token_accuracy": 0.9490657150745392,
474
+ "num_input_tokens_seen": 4148142,
475
+ "num_tokens": 4123487.0,
476
+ "step": 180,
477
+ "train_runtime": 154.9764,
478
+ "train_tokens_per_second": 26766.274
479
+ },
480
+ {
481
+ "entropy": 2.021875,
482
+ "epoch": 0.6994328922495274,
483
+ "grad_norm": 2.953125,
484
+ "learning_rate": 3.113087123283002e-05,
485
+ "loss": 0.124,
486
+ "mean_token_accuracy": 0.964401924610138,
487
+ "num_input_tokens_seen": 4263312,
488
+ "num_tokens": 4238014.0,
489
+ "step": 185,
490
+ "train_runtime": 159.4694,
491
+ "train_tokens_per_second": 26734.354
492
+ },
493
+ {
494
+ "entropy": 1.96171875,
495
+ "epoch": 0.718336483931947,
496
+ "grad_norm": 3.4375,
497
+ "learning_rate": 3.060662328794916e-05,
498
+ "loss": 0.1498,
499
+ "mean_token_accuracy": 0.9481843888759613,
500
+ "num_input_tokens_seen": 4378630,
501
+ "num_tokens": 4352627.0,
502
+ "step": 190,
503
+ "train_runtime": 163.6223,
504
+ "train_tokens_per_second": 26760.595
505
+ },
506
+ {
507
+ "entropy": 1.9640625,
508
+ "epoch": 0.7372400756143668,
509
+ "grad_norm": 4.1875,
510
+ "learning_rate": 3.0072032364289914e-05,
511
+ "loss": 0.1076,
512
+ "mean_token_accuracy": 0.9691859900951385,
513
+ "num_input_tokens_seen": 4493600,
514
+ "num_tokens": 4467053.0,
515
+ "step": 195,
516
+ "train_runtime": 166.9247,
517
+ "train_tokens_per_second": 26919.915
518
+ },
519
+ {
520
+ "entropy": 2.02734375,
521
+ "epoch": 0.7561436672967864,
522
+ "grad_norm": 3.875,
523
+ "learning_rate": 2.9527619764667376e-05,
524
+ "loss": 0.2501,
525
+ "mean_token_accuracy": 0.9455641567707062,
526
+ "num_input_tokens_seen": 4609216,
527
+ "num_tokens": 4581812.0,
528
+ "step": 200,
529
+ "train_runtime": 172.0695,
530
+ "train_tokens_per_second": 26786.938
531
+ },
532
+ {
533
+ "entropy": 2.14375,
534
+ "epoch": 0.775047258979206,
535
+ "grad_norm": 5.4375,
536
+ "learning_rate": 2.8973916369439194e-05,
537
+ "loss": 0.2157,
538
+ "mean_token_accuracy": 0.9492439985275268,
539
+ "num_input_tokens_seen": 4724086,
540
+ "num_tokens": 4696178.0,
541
+ "step": 205,
542
+ "train_runtime": 175.6473,
543
+ "train_tokens_per_second": 26895.294
544
+ },
545
+ {
546
+ "entropy": 2.2625,
547
+ "epoch": 0.7939508506616257,
548
+ "grad_norm": 3.75,
549
+ "learning_rate": 2.84114621188211e-05,
550
+ "loss": 0.1762,
551
+ "mean_token_accuracy": 0.9574925601482391,
552
+ "num_input_tokens_seen": 4839702,
553
+ "num_tokens": 4810939.0,
554
+ "step": 210,
555
+ "train_runtime": 180.4712,
556
+ "train_tokens_per_second": 26817.036
557
+ },
558
+ {
559
+ "entropy": 2.2953125,
560
+ "epoch": 0.8128544423440454,
561
+ "grad_norm": 3.9375,
562
+ "learning_rate": 2.7840805486367792e-05,
563
+ "loss": 0.1703,
564
+ "mean_token_accuracy": 0.9540181159973145,
565
+ "num_input_tokens_seen": 4955098,
566
+ "num_tokens": 4925591.0,
567
+ "step": 215,
568
+ "train_runtime": 184.4177,
569
+ "train_tokens_per_second": 26868.891
570
+ },
571
+ {
572
+ "entropy": 2.2828125,
573
+ "epoch": 0.831758034026465,
574
+ "grad_norm": 4.625,
575
+ "learning_rate": 2.7262502944132526e-05,
576
+ "loss": 0.0938,
577
+ "mean_token_accuracy": 0.9725252389907837,
578
+ "num_input_tokens_seen": 5070258,
579
+ "num_tokens": 5040089.0,
580
+ "step": 220,
581
+ "train_runtime": 188.065,
582
+ "train_tokens_per_second": 26960.132
583
+ },
584
+ {
585
+ "entropy": 2.1265625,
586
+ "epoch": 0.8506616257088847,
587
+ "grad_norm": 5.71875,
588
+ "learning_rate": 2.667711842002707e-05,
589
+ "loss": 0.1704,
590
+ "mean_token_accuracy": 0.9579161703586578,
591
+ "num_input_tokens_seen": 5185478,
592
+ "num_tokens": 5154604.0,
593
+ "step": 225,
594
+ "train_runtime": 192.8301,
595
+ "train_tokens_per_second": 26891.43
596
+ },
597
+ {
598
+ "entropy": 2.0484375,
599
+ "epoch": 0.8695652173913043,
600
+ "grad_norm": 4.3125,
601
+ "learning_rate": 2.6085222747911155e-05,
602
+ "loss": 0.4284,
603
+ "mean_token_accuracy": 0.9190201222896576,
604
+ "num_input_tokens_seen": 5301020,
605
+ "num_tokens": 5269357.0,
606
+ "step": 230,
607
+ "train_runtime": 196.1744,
608
+ "train_tokens_per_second": 27021.971
609
+ },
610
+ {
611
+ "entropy": 1.98671875,
612
+ "epoch": 0.888468809073724,
613
+ "grad_norm": 13.3125,
614
+ "learning_rate": 2.5487393110947557e-05,
615
+ "loss": 0.1346,
616
+ "mean_token_accuracy": 0.9579481542110443,
617
+ "num_input_tokens_seen": 5416464,
618
+ "num_tokens": 5384069.0,
619
+ "step": 235,
620
+ "train_runtime": 201.21,
621
+ "train_tokens_per_second": 26919.463
622
+ },
623
+ {
624
+ "entropy": 1.9875,
625
+ "epoch": 0.9073724007561437,
626
+ "grad_norm": 3.84375,
627
+ "learning_rate": 2.4884212478765747e-05,
628
+ "loss": 0.097,
629
+ "mean_token_accuracy": 0.9672803819179535,
630
+ "num_input_tokens_seen": 5531644,
631
+ "num_tokens": 5498568.0,
632
+ "step": 240,
633
+ "train_runtime": 205.075,
634
+ "train_tokens_per_second": 26973.766
635
+ },
636
+ {
637
+ "entropy": 2.00234375,
638
+ "epoch": 0.9262759924385633,
639
+ "grad_norm": 4.9375,
640
+ "learning_rate": 2.427626903898292e-05,
641
+ "loss": 0.2298,
642
+ "mean_token_accuracy": 0.9443018674850464,
643
+ "num_input_tokens_seen": 5646952,
644
+ "num_tokens": 5613157.0,
645
+ "step": 245,
646
+ "train_runtime": 208.4891,
647
+ "train_tokens_per_second": 27085.115
648
+ },
649
+ {
650
+ "entropy": 2.0140625,
651
+ "epoch": 0.945179584120983,
652
+ "grad_norm": 6.03125,
653
+ "learning_rate": 2.3664155623636715e-05,
654
+ "loss": 0.1732,
655
+ "mean_token_accuracy": 0.9442705571651459,
656
+ "num_input_tokens_seen": 5762366,
657
+ "num_tokens": 5727795.0,
658
+ "step": 250,
659
+ "train_runtime": 214.059,
660
+ "train_tokens_per_second": 26919.525
661
+ },
662
+ {
663
+ "entropy": 2.0125,
664
+ "epoch": 0.9640831758034026,
665
+ "grad_norm": 3.546875,
666
+ "learning_rate": 2.304846913108891e-05,
667
+ "loss": 0.1083,
668
+ "mean_token_accuracy": 0.9664817750453949,
669
+ "num_input_tokens_seen": 5877646,
670
+ "num_tokens": 5842437.0,
671
+ "step": 255,
672
+ "train_runtime": 275.7098,
673
+ "train_tokens_per_second": 21318.232
674
+ },
675
+ {
676
+ "entropy": 2.0,
677
+ "epoch": 0.9829867674858223,
678
+ "grad_norm": 2.671875,
679
+ "learning_rate": 2.242980994396401e-05,
680
+ "loss": 0.0875,
681
+ "mean_token_accuracy": 0.9795427262783051,
682
+ "num_input_tokens_seen": 5992710,
683
+ "num_tokens": 5956870.0,
684
+ "step": 260,
685
+ "train_runtime": 280.9684,
686
+ "train_tokens_per_second": 21328.766
687
+ },
688
+ {
689
+ "entropy": 1.9513888888888888,
690
+ "epoch": 1.0,
691
+ "grad_norm": 7.4375,
692
+ "learning_rate": 2.1808781343690027e-05,
693
+ "loss": 0.1654,
694
+ "mean_token_accuracy": 0.9603289763132731,
695
+ "num_input_tokens_seen": 6096342,
696
+ "num_tokens": 6059927.0,
697
+ "step": 265,
698
+ "train_runtime": 284.3725,
699
+ "train_tokens_per_second": 21437.877
700
+ },
701
+ {
702
+ "entropy": 1.903125,
703
+ "epoch": 1.0189035916824196,
704
+ "grad_norm": 3.453125,
705
+ "learning_rate": 2.118598892221257e-05,
706
+ "loss": 0.0783,
707
+ "mean_token_accuracy": 0.9817151129245758,
708
+ "num_input_tokens_seen": 6211574,
709
+ "num_tokens": 6174483.0,
710
+ "step": 270,
711
+ "train_runtime": 288.2049,
712
+ "train_tokens_per_second": 21552.63
713
+ },
714
+ {
715
+ "entropy": 1.84375,
716
+ "epoch": 1.0378071833648392,
717
+ "grad_norm": 1.734375,
718
+ "learning_rate": 2.0562039991455877e-05,
719
+ "loss": 0.1214,
720
+ "mean_token_accuracy": 0.9741188943386078,
721
+ "num_input_tokens_seen": 6327000,
722
+ "num_tokens": 6289163.0,
723
+ "step": 275,
724
+ "train_runtime": 293.7126,
725
+ "train_tokens_per_second": 21541.469
726
+ },
727
+ {
728
+ "entropy": 1.8421875,
729
+ "epoch": 1.056710775047259,
730
+ "grad_norm": 3.78125,
731
+ "learning_rate": 1.99375429911066e-05,
732
+ "loss": 0.1393,
733
+ "mean_token_accuracy": 0.9579156279563904,
734
+ "num_input_tokens_seen": 6442290,
735
+ "num_tokens": 6403766.0,
736
+ "step": 280,
737
+ "train_runtime": 297.1668,
738
+ "train_tokens_per_second": 21679.038
739
+ },
740
+ {
741
+ "entropy": 1.85078125,
742
+ "epoch": 1.0756143667296787,
743
+ "grad_norm": 3.953125,
744
+ "learning_rate": 1.931310689529781e-05,
745
+ "loss": 0.0872,
746
+ "mean_token_accuracy": 0.9788394093513488,
747
+ "num_input_tokens_seen": 6557852,
748
+ "num_tokens": 6518469.0,
749
+ "step": 285,
750
+ "train_runtime": 301.7702,
751
+ "train_tokens_per_second": 21731.276
752
+ },
753
+ {
754
+ "entropy": 1.8234375,
755
+ "epoch": 1.0945179584120983,
756
+ "grad_norm": 7.1875,
757
+ "learning_rate": 1.8689340618771937e-05,
758
+ "loss": 0.0637,
759
+ "mean_token_accuracy": 0.972537738084793,
760
+ "num_input_tokens_seen": 6673032,
761
+ "num_tokens": 6632963.0,
762
+ "step": 290,
763
+ "train_runtime": 306.4769,
764
+ "train_tokens_per_second": 21773.362
765
+ },
766
+ {
767
+ "entropy": 1.78359375,
768
+ "epoch": 1.113421550094518,
769
+ "grad_norm": 5.78125,
770
+ "learning_rate": 1.806685242310156e-05,
771
+ "loss": 0.0565,
772
+ "mean_token_accuracy": 0.9854797184467315,
773
+ "num_input_tokens_seen": 6788174,
774
+ "num_tokens": 6747403.0,
775
+ "step": 295,
776
+ "train_runtime": 310.3851,
777
+ "train_tokens_per_second": 21870.17
778
+ },
779
+ {
780
+ "entropy": 1.76015625,
781
+ "epoch": 1.1323251417769375,
782
+ "grad_norm": 8.8125,
783
+ "learning_rate": 1.7446249323547117e-05,
784
+ "loss": 0.0973,
785
+ "mean_token_accuracy": 0.9734237968921662,
786
+ "num_input_tokens_seen": 6903146,
787
+ "num_tokens": 6861788.0,
788
+ "step": 300,
789
+ "train_runtime": 315.4655,
790
+ "train_tokens_per_second": 21882.41
791
+ },
792
+ {
793
+ "entropy": 1.75078125,
794
+ "epoch": 1.1512287334593574,
795
+ "grad_norm": 1.4453125,
796
+ "learning_rate": 1.6828136497130014e-05,
797
+ "loss": 0.0681,
798
+ "mean_token_accuracy": 0.9820096373558045,
799
+ "num_input_tokens_seen": 7018350,
800
+ "num_tokens": 6976277.0,
801
+ "step": 305,
802
+ "train_runtime": 319.0527,
803
+ "train_tokens_per_second": 21997.465
804
+ },
805
+ {
806
+ "entropy": 1.740625,
807
+ "epoch": 1.170132325141777,
808
+ "grad_norm": 4.90625,
809
+ "learning_rate": 1.6213116692498206e-05,
810
+ "loss": 0.0625,
811
+ "mean_token_accuracy": 0.9826828062534332,
812
+ "num_input_tokens_seen": 7133636,
813
+ "num_tokens": 7090874.0,
814
+ "step": 310,
815
+ "train_runtime": 323.8986,
816
+ "train_tokens_per_second": 22024.29
817
+ },
818
+ {
819
+ "entropy": 1.7328125,
820
+ "epoch": 1.1890359168241966,
821
+ "grad_norm": 0.66015625,
822
+ "learning_rate": 1.560178964215987e-05,
823
+ "loss": 0.077,
824
+ "mean_token_accuracy": 0.978941410779953,
825
+ "num_input_tokens_seen": 7248866,
826
+ "num_tokens": 7205391.0,
827
+ "step": 315,
828
+ "train_runtime": 327.5895,
829
+ "train_tokens_per_second": 22127.897
830
+ },
831
+ {
832
+ "entropy": 1.73203125,
833
+ "epoch": 1.2079395085066162,
834
+ "grad_norm": 4.0625,
835
+ "learning_rate": 1.4994751477658139e-05,
836
+ "loss": 0.067,
837
+ "mean_token_accuracy": 0.9818780541419982,
838
+ "num_input_tokens_seen": 7363900,
839
+ "num_tokens": 7319827.0,
840
+ "step": 320,
841
+ "train_runtime": 331.4598,
842
+ "train_tokens_per_second": 22216.571
843
+ },
844
+ {
845
+ "entropy": 1.73515625,
846
+ "epoch": 1.2268431001890359,
847
+ "grad_norm": 2.734375,
848
+ "learning_rate": 1.4392594148257426e-05,
849
+ "loss": 0.1153,
850
+ "mean_token_accuracy": 0.9638942897319793,
851
+ "num_input_tokens_seen": 7479394,
852
+ "num_tokens": 7434543.0,
853
+ "step": 325,
854
+ "train_runtime": 336.2629,
855
+ "train_tokens_per_second": 22242.696
856
+ },
857
+ {
858
+ "entropy": 1.74609375,
859
+ "epoch": 1.2457466918714555,
860
+ "grad_norm": 2.046875,
861
+ "learning_rate": 1.3795904843707959e-05,
862
+ "loss": 0.0359,
863
+ "mean_token_accuracy": 0.9886789560317993,
864
+ "num_input_tokens_seen": 7594632,
865
+ "num_tokens": 7549134.0,
866
+ "step": 330,
867
+ "train_runtime": 339.6052,
868
+ "train_tokens_per_second": 22363.12
869
+ },
870
+ {
871
+ "entropy": 1.740625,
872
+ "epoch": 1.264650283553875,
873
+ "grad_norm": 2.25,
874
+ "learning_rate": 1.3205265421651588e-05,
875
+ "loss": 0.0808,
876
+ "mean_token_accuracy": 0.9852688193321228,
877
+ "num_input_tokens_seen": 7709704,
878
+ "num_tokens": 7663583.0,
879
+ "step": 335,
880
+ "train_runtime": 344.9458,
881
+ "train_tokens_per_second": 22350.48
882
+ },
883
+ {
884
+ "entropy": 1.75078125,
885
+ "epoch": 1.283553875236295,
886
+ "grad_norm": 1.8125,
887
+ "learning_rate": 1.2621251840227112e-05,
888
+ "loss": 0.0663,
889
+ "mean_token_accuracy": 0.9817369997501373,
890
+ "num_input_tokens_seen": 7824834,
891
+ "num_tokens": 7778064.0,
892
+ "step": 340,
893
+ "train_runtime": 348.223,
894
+ "train_tokens_per_second": 22470.756
895
+ },
896
+ {
897
+ "entropy": 1.75234375,
898
+ "epoch": 1.3024574669187146,
899
+ "grad_norm": 4.28125,
900
+ "learning_rate": 1.2044433596428537e-05,
901
+ "loss": 0.0678,
902
+ "mean_token_accuracy": 0.9812626421451569,
903
+ "num_input_tokens_seen": 7939832,
904
+ "num_tokens": 7892415.0,
905
+ "step": 345,
906
+ "train_runtime": 352.0847,
907
+ "train_tokens_per_second": 22550.916
908
+ },
909
+ {
910
+ "entropy": 1.746875,
911
+ "epoch": 1.3213610586011342,
912
+ "grad_norm": 2.703125,
913
+ "learning_rate": 1.1475373170763819e-05,
914
+ "loss": 0.0465,
915
+ "mean_token_accuracy": 0.9823280215263367,
916
+ "num_input_tokens_seen": 8054988,
917
+ "num_tokens": 8006926.0,
918
+ "step": 350,
919
+ "train_runtime": 357.1271,
920
+ "train_tokens_per_second": 22554.962
921
+ },
922
+ {
923
+ "entropy": 1.74765625,
924
+ "epoch": 1.3402646502835538,
925
+ "grad_norm": 1.4921875,
926
+ "learning_rate": 1.0914625478755672e-05,
927
+ "loss": 0.1174,
928
+ "mean_token_accuracy": 0.9695515096187591,
929
+ "num_input_tokens_seen": 8170098,
930
+ "num_tokens": 8121373.0,
931
+ "step": 355,
932
+ "train_runtime": 360.7524,
933
+ "train_tokens_per_second": 22647.381
934
+ },
935
+ {
936
+ "entropy": 1.74453125,
937
+ "epoch": 1.3591682419659734,
938
+ "grad_norm": 1.1015625,
939
+ "learning_rate": 1.0362737329819413e-05,
940
+ "loss": 0.045,
941
+ "mean_token_accuracy": 0.9885900497436524,
942
+ "num_input_tokens_seen": 8285346,
943
+ "num_tokens": 8235981.0,
944
+ "step": 360,
945
+ "train_runtime": 366.0216,
946
+ "train_tokens_per_second": 22636.221
947
+ },
948
+ {
949
+ "entropy": 1.74296875,
950
+ "epoch": 1.3780718336483933,
951
+ "grad_norm": 5.15625,
952
+ "learning_rate": 9.820246894045316e-06,
953
+ "loss": 0.0428,
954
+ "mean_token_accuracy": 0.9822307825088501,
955
+ "num_input_tokens_seen": 8400240,
956
+ "num_tokens": 8350356.0,
957
+ "step": 365,
958
+ "train_runtime": 369.6364,
959
+ "train_tokens_per_second": 22725.685
960
+ },
961
+ {
962
+ "entropy": 1.73515625,
963
+ "epoch": 1.3969754253308129,
964
+ "grad_norm": 4.09375,
965
+ "learning_rate": 9.28768317740564e-06,
966
+ "loss": 0.099,
967
+ "mean_token_accuracy": 0.9710565328598022,
968
+ "num_input_tokens_seen": 8515740,
969
+ "num_tokens": 8465025.0,
970
+ "step": 370,
971
+ "train_runtime": 373.5701,
972
+ "train_tokens_per_second": 22795.56
973
+ },
974
+ {
975
+ "entropy": 1.7328125,
976
+ "epoch": 1.4158790170132325,
977
+ "grad_norm": 4.96875,
978
+ "learning_rate": 8.765565505897902e-06,
979
+ "loss": 0.0736,
980
+ "mean_token_accuracy": 0.9741575241088867,
981
+ "num_input_tokens_seen": 8631054,
982
+ "num_tokens": 8579648.0,
983
+ "step": 375,
984
+ "train_runtime": 378.7394,
985
+ "train_tokens_per_second": 22788.901
986
+ },
987
+ {
988
+ "entropy": 1.73359375,
989
+ "epoch": 1.434782608695652,
990
+ "grad_norm": 3.265625,
991
+ "learning_rate": 8.254403019127566e-06,
992
+ "loss": 0.0806,
993
+ "mean_token_accuracy": 0.9791056990623475,
994
+ "num_input_tokens_seen": 8746364,
995
+ "num_tokens": 8694249.0,
996
+ "step": 380,
997
+ "train_runtime": 382.0615,
998
+ "train_tokens_per_second": 22892.552
999
+ },
1000
+ {
1001
+ "entropy": 1.73515625,
1002
+ "epoch": 1.4536862003780717,
1003
+ "grad_norm": 3.75,
1004
+ "learning_rate": 7.754694173823947e-06,
1005
+ "loss": 0.0404,
1006
+ "mean_token_accuracy": 0.9839386224746705,
1007
+ "num_input_tokens_seen": 8861574,
1008
+ "num_tokens": 8808789.0,
1009
+ "step": 385,
1010
+ "train_runtime": 387.2205,
1011
+ "train_tokens_per_second": 22885.084
1012
+ },
1013
+ {
1014
+ "entropy": 1.73359375,
1015
+ "epoch": 1.4725897920604916,
1016
+ "grad_norm": 5.09375,
1017
+ "learning_rate": 7.266926257773346e-06,
1018
+ "loss": 0.0926,
1019
+ "mean_token_accuracy": 0.9714232623577118,
1020
+ "num_input_tokens_seen": 8976944,
1021
+ "num_tokens": 8923407.0,
1022
+ "step": 390,
1023
+ "train_runtime": 390.891,
1024
+ "train_tokens_per_second": 22965.336
1025
+ },
1026
+ {
1027
+ "entropy": 1.72265625,
1028
+ "epoch": 1.4914933837429112,
1029
+ "grad_norm": 5.0,
1030
+ "learning_rate": 6.7915749146436415e-06,
1031
+ "loss": 0.0519,
1032
+ "mean_token_accuracy": 0.9837916433811188,
1033
+ "num_input_tokens_seen": 9092050,
1034
+ "num_tokens": 9037924.0,
1035
+ "step": 395,
1036
+ "train_runtime": 395.3397,
1037
+ "train_tokens_per_second": 22998.071
1038
+ },
1039
+ {
1040
+ "entropy": 1.71796875,
1041
+ "epoch": 1.5103969754253308,
1042
+ "grad_norm": 3.875,
1043
+ "learning_rate": 6.329103680163495e-06,
1044
+ "loss": 0.2115,
1045
+ "mean_token_accuracy": 0.9516554296016693,
1046
+ "num_input_tokens_seen": 9207594,
1047
+ "num_tokens": 9152659.0,
1048
+ "step": 400,
1049
+ "train_runtime": 399.5499,
1050
+ "train_tokens_per_second": 23044.916
1051
+ },
1052
+ {
1053
+ "entropy": 1.71640625,
1054
+ "epoch": 1.5293005671077504,
1055
+ "grad_norm": 0.51953125,
1056
+ "learning_rate": 5.879963530108506e-06,
1057
+ "loss": 0.0348,
1058
+ "mean_token_accuracy": 0.9919346511363983,
1059
+ "num_input_tokens_seen": 9322572,
1060
+ "num_tokens": 9267059.0,
1061
+ "step": 405,
1062
+ "train_runtime": 403.4031,
1063
+ "train_tokens_per_second": 23109.815
1064
+ },
1065
+ {
1066
+ "entropy": 1.7125,
1067
+ "epoch": 1.54820415879017,
1068
+ "grad_norm": 2.234375,
1069
+ "learning_rate": 5.444592440535177e-06,
1070
+ "loss": 0.0374,
1071
+ "mean_token_accuracy": 0.9837370038032531,
1072
+ "num_input_tokens_seen": 9438004,
1073
+ "num_tokens": 9381725.0,
1074
+ "step": 410,
1075
+ "train_runtime": 407.9692,
1076
+ "train_tokens_per_second": 23134.111
1077
+ },
1078
+ {
1079
+ "entropy": 1.7078125,
1080
+ "epoch": 1.5671077504725899,
1081
+ "grad_norm": 3.09375,
1082
+ "learning_rate": 5.023414960691469e-06,
1083
+ "loss": 0.0325,
1084
+ "mean_token_accuracy": 0.9918534696102143,
1085
+ "num_input_tokens_seen": 9553156,
1086
+ "num_tokens": 9496255.0,
1087
+ "step": 415,
1088
+ "train_runtime": 412.1408,
1089
+ "train_tokens_per_second": 23179.35
1090
+ },
1091
+ {
1092
+ "entropy": 1.703125,
1093
+ "epoch": 1.5860113421550095,
1094
+ "grad_norm": 4.78125,
1095
+ "learning_rate": 4.616841799020364e-06,
1096
+ "loss": 0.0618,
1097
+ "mean_token_accuracy": 0.9808044970035553,
1098
+ "num_input_tokens_seen": 9668364,
1099
+ "num_tokens": 9610808.0,
1100
+ "step": 420,
1101
+ "train_runtime": 416.3235,
1102
+ "train_tokens_per_second": 23223.203
1103
+ },
1104
+ {
1105
+ "entropy": 1.703125,
1106
+ "epoch": 1.6049149338374291,
1107
+ "grad_norm": 6.1875,
1108
+ "learning_rate": 4.225269422660258e-06,
1109
+ "loss": 0.0493,
1110
+ "mean_token_accuracy": 0.9843941271305084,
1111
+ "num_input_tokens_seen": 9783552,
1112
+ "num_tokens": 9725283.0,
1113
+ "step": 425,
1114
+ "train_runtime": 421.1316,
1115
+ "train_tokens_per_second": 23231.576
1116
+ },
1117
+ {
1118
+ "entropy": 1.70390625,
1119
+ "epoch": 1.6238185255198487,
1120
+ "grad_norm": 1.5078125,
1121
+ "learning_rate": 3.8490796708326404e-06,
1122
+ "loss": 0.0595,
1123
+ "mean_token_accuracy": 0.9822299420833588,
1124
+ "num_input_tokens_seen": 9898934,
1125
+ "num_tokens": 9839878.0,
1126
+ "step": 430,
1127
+ "train_runtime": 424.7606,
1128
+ "train_tokens_per_second": 23304.735
1129
+ },
1130
+ {
1131
+ "entropy": 1.7015625,
1132
+ "epoch": 1.6427221172022684,
1133
+ "grad_norm": 1.328125,
1134
+ "learning_rate": 3.4886393824940924e-06,
1135
+ "loss": 0.059,
1136
+ "mean_token_accuracy": 0.9807979345321656,
1137
+ "num_input_tokens_seen": 10014142,
1138
+ "num_tokens": 9954403.0,
1139
+ "step": 435,
1140
+ "train_runtime": 429.8927,
1141
+ "train_tokens_per_second": 23294.514
1142
+ },
1143
+ {
1144
+ "entropy": 1.70390625,
1145
+ "epoch": 1.6616257088846882,
1146
+ "grad_norm": 2.09375,
1147
+ "learning_rate": 3.144300038615691e-06,
1148
+ "loss": 0.0574,
1149
+ "mean_token_accuracy": 0.9839386105537414,
1150
+ "num_input_tokens_seen": 10129264,
1151
+ "num_tokens": 10068933.0,
1152
+ "step": 440,
1153
+ "train_runtime": 433.4828,
1154
+ "train_tokens_per_second": 23367.164
1155
+ },
1156
+ {
1157
+ "entropy": 1.6984375,
1158
+ "epoch": 1.6805293005671076,
1159
+ "grad_norm": 4.25,
1160
+ "learning_rate": 2.8163974194386766e-06,
1161
+ "loss": 0.0669,
1162
+ "mean_token_accuracy": 0.9792383193969727,
1163
+ "num_input_tokens_seen": 10244732,
1164
+ "num_tokens": 10183591.0,
1165
+ "step": 445,
1166
+ "train_runtime": 437.9792,
1167
+ "train_tokens_per_second": 23390.909
1168
+ },
1169
+ {
1170
+ "entropy": 1.7015625,
1171
+ "epoch": 1.6994328922495274,
1172
+ "grad_norm": 3.46875,
1173
+ "learning_rate": 2.5052512770405434e-06,
1174
+ "loss": 0.0801,
1175
+ "mean_token_accuracy": 0.9761136710643769,
1176
+ "num_input_tokens_seen": 10360212,
1177
+ "num_tokens": 10298251.0,
1178
+ "step": 450,
1179
+ "train_runtime": 442.481,
1180
+ "train_tokens_per_second": 23413.915
1181
+ },
1182
+ {
1183
+ "entropy": 1.70234375,
1184
+ "epoch": 1.718336483931947,
1185
+ "grad_norm": 0.59765625,
1186
+ "learning_rate": 2.2111650235309147e-06,
1187
+ "loss": 0.0297,
1188
+ "mean_token_accuracy": 0.9904489517211914,
1189
+ "num_input_tokens_seen": 10475400,
1190
+ "num_tokens": 10412810.0,
1191
+ "step": 455,
1192
+ "train_runtime": 446.3738,
1193
+ "train_tokens_per_second": 23467.773
1194
+ },
1195
+ {
1196
+ "entropy": 1.69921875,
1197
+ "epoch": 1.7372400756143667,
1198
+ "grad_norm": 4.0625,
1199
+ "learning_rate": 1.9344254351812287e-06,
1200
+ "loss": 0.0989,
1201
+ "mean_token_accuracy": 0.9743396818637848,
1202
+ "num_input_tokens_seen": 10590710,
1203
+ "num_tokens": 10527389.0,
1204
+ "step": 460,
1205
+ "train_runtime": 451.1755,
1206
+ "train_tokens_per_second": 23473.591
1207
+ },
1208
+ {
1209
+ "entropy": 1.703125,
1210
+ "epoch": 1.7561436672967865,
1211
+ "grad_norm": 0.890625,
1212
+ "learning_rate": 1.6753023727767436e-06,
1213
+ "loss": 0.0476,
1214
+ "mean_token_accuracy": 0.9838890075683594,
1215
+ "num_input_tokens_seen": 10705900,
1216
+ "num_tokens": 10641918.0,
1217
+ "step": 465,
1218
+ "train_runtime": 454.754,
1219
+ "train_tokens_per_second": 23542.179
1220
+ },
1221
+ {
1222
+ "entropy": 1.7,
1223
+ "epoch": 1.775047258979206,
1224
+ "grad_norm": 1.359375,
1225
+ "learning_rate": 1.4340485184635712e-06,
1226
+ "loss": 0.0556,
1227
+ "mean_token_accuracy": 0.9777659058570862,
1228
+ "num_input_tokens_seen": 10821144,
1229
+ "num_tokens": 10756496.0,
1230
+ "step": 470,
1231
+ "train_runtime": 459.2027,
1232
+ "train_tokens_per_second": 23565.072
1233
+ },
1234
+ {
1235
+ "entropy": 1.69921875,
1236
+ "epoch": 1.7939508506616257,
1237
+ "grad_norm": 1.171875,
1238
+ "learning_rate": 1.2108991293473627e-06,
1239
+ "loss": 0.0595,
1240
+ "mean_token_accuracy": 0.9741835057735443,
1241
+ "num_input_tokens_seen": 10936460,
1242
+ "num_tokens": 10871124.0,
1243
+ "step": 475,
1244
+ "train_runtime": 463.6099,
1245
+ "train_tokens_per_second": 23589.79
1246
+ },
1247
+ {
1248
+ "entropy": 1.69765625,
1249
+ "epoch": 1.8128544423440454,
1250
+ "grad_norm": 3.265625,
1251
+ "learning_rate": 1.0060718080838683e-06,
1252
+ "loss": 0.0541,
1253
+ "mean_token_accuracy": 0.9831156551837921,
1254
+ "num_input_tokens_seen": 11051508,
1255
+ "num_tokens": 10985594.0,
1256
+ "step": 480,
1257
+ "train_runtime": 467.1593,
1258
+ "train_tokens_per_second": 23656.828
1259
+ },
1260
+ {
1261
+ "entropy": 1.70078125,
1262
+ "epoch": 1.831758034026465,
1263
+ "grad_norm": 2.4375,
1264
+ "learning_rate": 8.197662906851534e-07,
1265
+ "loss": 0.0835,
1266
+ "mean_token_accuracy": 0.9726030707359314,
1267
+ "num_input_tokens_seen": 11166904,
1268
+ "num_tokens": 11100230.0,
1269
+ "step": 485,
1270
+ "train_runtime": 472.195,
1271
+ "train_tokens_per_second": 23648.922
1272
+ },
1273
+ {
1274
+ "entropy": 1.69921875,
1275
+ "epoch": 1.8506616257088848,
1276
+ "grad_norm": 2.765625,
1277
+ "learning_rate": 6.521642517483573e-07,
1278
+ "loss": 0.0532,
1279
+ "mean_token_accuracy": 0.9853454470634461,
1280
+ "num_input_tokens_seen": 11281802,
1281
+ "num_tokens": 11214624.0,
1282
+ "step": 490,
1283
+ "train_runtime": 475.7718,
1284
+ "train_tokens_per_second": 23712.635
1285
+ },
1286
+ {
1287
+ "entropy": 1.70078125,
1288
+ "epoch": 1.8695652173913042,
1289
+ "grad_norm": 2.171875,
1290
+ "learning_rate": 5.034291272968772e-07,
1291
+ "loss": 0.027,
1292
+ "mean_token_accuracy": 0.9934648215770722,
1293
+ "num_input_tokens_seen": 11396946,
1294
+ "num_tokens": 11329098.0,
1295
+ "step": 495,
1296
+ "train_runtime": 480.2436,
1297
+ "train_tokens_per_second": 23731.596
1298
+ },
1299
+ {
1300
+ "entropy": 1.6984375,
1301
+ "epoch": 1.888468809073724,
1302
+ "grad_norm": 4.0625,
1303
+ "learning_rate": 3.737059554068334e-07,
1304
+ "loss": 0.0742,
1305
+ "mean_token_accuracy": 0.9744843065738678,
1306
+ "num_input_tokens_seen": 11512282,
1307
+ "num_tokens": 11443715.0,
1308
+ "step": 500,
1309
+ "train_runtime": 484.6792,
1310
+ "train_tokens_per_second": 23752.376
1311
+ },
1312
+ {
1313
+ "entropy": 1.69921875,
1314
+ "epoch": 1.9073724007561437,
1315
+ "grad_norm": 6.84375,
1316
+ "learning_rate": 2.631212347741352e-07,
1317
+ "loss": 0.1322,
1318
+ "mean_token_accuracy": 0.9680740118026734,
1319
+ "num_input_tokens_seen": 11627828,
1320
+ "num_tokens": 11558513.0,
1321
+ "step": 505,
1322
+ "train_runtime": 544.5283,
1323
+ "train_tokens_per_second": 21353.945
1324
+ },
1325
+ {
1326
+ "entropy": 1.69921875,
1327
+ "epoch": 1.9262759924385633,
1328
+ "grad_norm": 1.0078125,
1329
+ "learning_rate": 1.7178280136011417e-07,
1330
+ "loss": 0.0864,
1331
+ "mean_token_accuracy": 0.9749818980693817,
1332
+ "num_input_tokens_seen": 11743010,
1333
+ "num_tokens": 11673010.0,
1334
+ "step": 510,
1335
+ "train_runtime": 549.7569,
1336
+ "train_tokens_per_second": 21360.369
1337
+ },
1338
+ {
1339
+ "entropy": 1.69921875,
1340
+ "epoch": 1.9451795841209831,
1341
+ "grad_norm": 2.5625,
1342
+ "learning_rate": 9.977972323599095e-08,
1343
+ "loss": 0.1175,
1344
+ "mean_token_accuracy": 0.9680160820484162,
1345
+ "num_input_tokens_seen": 11858430,
1346
+ "num_tokens": 11787637.0,
1347
+ "step": 515,
1348
+ "train_runtime": 553.6509,
1349
+ "train_tokens_per_second": 21418.605
1350
+ },
1351
+ {
1352
+ "entropy": 1.69765625,
1353
+ "epoch": 1.9640831758034025,
1354
+ "grad_norm": 2.921875,
1355
+ "learning_rate": 4.718221372874254e-08,
1356
+ "loss": 0.0695,
1357
+ "mean_token_accuracy": 0.9804269134998321,
1358
+ "num_input_tokens_seen": 11973576,
1359
+ "num_tokens": 11902111.0,
1360
+ "step": 520,
1361
+ "train_runtime": 557.8609,
1362
+ "train_tokens_per_second": 21463.371
1363
+ },
1364
+ {
1365
+ "entropy": 1.69609375,
1366
+ "epoch": 1.9829867674858224,
1367
+ "grad_norm": 5.8125,
1368
+ "learning_rate": 1.4041562953031051e-08,
1369
+ "loss": 0.1152,
1370
+ "mean_token_accuracy": 0.9696780204772949,
1371
+ "num_input_tokens_seen": 12088990,
1372
+ "num_tokens": 12016759.0,
1373
+ "step": 525,
1374
+ "train_runtime": 561.9991,
1375
+ "train_tokens_per_second": 21510.694
1376
+ },
1377
+ {
1378
+ "entropy": 1.6961805555555556,
1379
+ "epoch": 2.0,
1380
+ "grad_norm": 3.75,
1381
+ "learning_rate": 3.900877959917004e-10,
1382
+ "loss": 0.0989,
1383
+ "mean_token_accuracy": 0.9715293182267083,
1384
+ "num_input_tokens_seen": 12192662,
1385
+ "num_tokens": 12119827.0,
1386
+ "step": 530,
1387
+ "train_runtime": 565.5622,
1388
+ "train_tokens_per_second": 21558.482
1389
+ },
1390
+ {
1391
+ "epoch": 2.0,
1392
+ "num_input_tokens_seen": 12192662,
1393
+ "step": 530,
1394
+ "total_flos": 3.3226637176733696e+16,
1395
+ "train_loss": 0.1822078584218925,
1396
+ "train_runtime": 612.9949,
1397
+ "train_samples_per_second": 27.592,
1398
+ "train_steps_per_second": 0.865,
1399
+ "train_tokens_per_second": 2486.879
1400
+ }
1401
+ ],
1402
+ "logging_steps": 5,
1403
+ "max_steps": 530,
1404
+ "num_input_tokens_seen": 12192662,
1405
+ "num_train_epochs": 2,
1406
+ "save_steps": 250,
1407
+ "stateful_callbacks": {
1408
+ "TrainerControl": {
1409
+ "args": {
1410
+ "should_epoch_stop": false,
1411
+ "should_evaluate": false,
1412
+ "should_log": false,
1413
+ "should_save": true,
1414
+ "should_training_stop": true
1415
+ },
1416
+ "attributes": {}
1417
+ }
1418
+ },
1419
+ "total_flos": 3.3226637176733696e+16,
1420
+ "train_batch_size": 2,
1421
+ "trial_name": null,
1422
+ "trial_params": null
1423
+ }
training_args.bin ADDED
Binary file (6.35 kB). View file
 
vocab.json ADDED
The diff for this file is too large to render. See raw diff