Hugging Face's logo

Siddeshwar1625
/
osint-checkpoints

Safetensors
Model card Files
xet
 Community
Siddeshwar1625 commited on
Commit
d68559d
·
verified ·
1 Parent(s): 6d5cc53

Intermediate checkpoint upload step=60 (answerer_train)

Browse files
Files changed (12) hide show
  1. .gitattributes +1 -0
  2. self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/chat_template.jinja +54 -0
  3. self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/config.json +57 -0
  4. self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/generation_config.json +13 -0
  5. self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/model.safetensors +3 -0
  6. self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/optimizer.pt +3 -0
  7. self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/rng_state.pth +3 -0
  8. self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/scheduler.pt +3 -0
  9. self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/tokenizer.json +3 -0
  10. self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/tokenizer_config.json +32 -0
  11. self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/trainer_state.json +1129 -0
  12. self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/training_args.bin +3 -0
.gitattributes CHANGED
@@ -88,3 +88,4 @@ self_play_hf_l40s_full/round_005/generator_train/checkpoint-90/tokenizer.json fi
88
  self_play_hf_l40s_full/round_005/generator_train/checkpoint-120/tokenizer.json filter=lfs diff=lfs merge=lfs -text
89
  self_play_hf_l40s_full/round_005/generator_train/final_model/tokenizer.json filter=lfs diff=lfs merge=lfs -text
90
  self_play_hf_l40s_full/round_005/answerer_train/checkpoint-30/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 
 
88
  self_play_hf_l40s_full/round_005/generator_train/checkpoint-120/tokenizer.json filter=lfs diff=lfs merge=lfs -text
89
  self_play_hf_l40s_full/round_005/generator_train/final_model/tokenizer.json filter=lfs diff=lfs merge=lfs -text
90
  self_play_hf_l40s_full/round_005/answerer_train/checkpoint-30/tokenizer.json filter=lfs diff=lfs merge=lfs -text
91
+ self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/tokenizer.json filter=lfs diff=lfs merge=lfs -text
self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/chat_template.jinja ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0]['role'] == 'system' %}
4
+ {{- messages[0]['content'] }}
5
+ {%- else %}
6
+ {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
7
+ {%- endif %}
8
+ {{- "\n\n# 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>" }}
9
+ {%- for tool in tools %}
10
+ {{- "\n" }}
11
+ {{- tool | tojson }}
12
+ {%- endfor %}
13
+ {{- "\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" }}
14
+ {%- else %}
15
+ {%- if messages[0]['role'] == 'system' %}
16
+ {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
17
+ {%- else %}
18
+ {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
19
+ {%- endif %}
20
+ {%- endif %}
21
+ {%- for message in messages %}
22
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
23
+ {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
24
+ {%- elif message.role == "assistant" %}
25
+ {{- '<|im_start|>' + message.role }}
26
+ {%- if message.content %}
27
+ {{- '\n' + message.content }}
28
+ {%- endif %}
29
+ {%- for tool_call in message.tool_calls %}
30
+ {%- if tool_call.function is defined %}
31
+ {%- set tool_call = tool_call.function %}
32
+ {%- endif %}
33
+ {{- '\n<tool_call>\n{"name": "' }}
34
+ {{- tool_call.name }}
35
+ {{- '", "arguments": ' }}
36
+ {{- tool_call.arguments | tojson }}
37
+ {{- '}\n</tool_call>' }}
38
+ {%- endfor %}
39
+ {{- '<|im_end|>\n' }}
40
+ {%- elif message.role == "tool" %}
41
+ {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
42
+ {{- '<|im_start|>user' }}
43
+ {%- endif %}
44
+ {{- '\n<tool_response>\n' }}
45
+ {{- message.content }}
46
+ {{- '\n</tool_response>' }}
47
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
48
+ {{- '<|im_end|>\n' }}
49
+ {%- endif %}
50
+ {%- endif %}
51
+ {%- endfor %}
52
+ {%- if add_generation_prompt %}
53
+ {{- '<|im_start|>assistant\n' }}
54
+ {%- endif %}
self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/config.json ADDED
@@ -0,0 +1,57 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen2ForCausalLM"
4
+ ],
5
+ "attention_dropout": 0.0,
6
+ "bos_token_id": null,
7
+ "dtype": "float32",
8
+ "eos_token_id": 151645,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 896,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 4864,
13
+ "layer_types": [
14
+ "full_attention",
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
+ ],
39
+ "max_position_embeddings": 32768,
40
+ "max_window_layers": 21,
41
+ "model_type": "qwen2",
42
+ "num_attention_heads": 14,
43
+ "num_hidden_layers": 24,
44
+ "num_key_value_heads": 2,
45
+ "pad_token_id": 151643,
46
+ "rms_norm_eps": 1e-06,
47
+ "rope_parameters": {
48
+ "rope_theta": 1000000.0,
49
+ "rope_type": "default"
50
+ },
51
+ "sliding_window": null,
52
+ "tie_word_embeddings": true,
53
+ "transformers_version": "5.6.2",
54
+ "use_cache": false,
55
+ "use_sliding_window": false,
56
+ "vocab_size": 151936
57
+ }
self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/generation_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_sample": true,
3
+ "eos_token_id": [
4
+ 151645,
5
+ 151643
6
+ ],
7
+ "pad_token_id": 151643,
8
+ "repetition_penalty": 1.1,
9
+ "temperature": 0.7,
10
+ "top_k": 20,
11
+ "top_p": 0.8,
12
+ "transformers_version": "5.6.2"
13
+ }
self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:51d9ac48ce2d5b6670e7367dc99b94d7a005d4a317a4f01da1b5fef1cf45b06f
3
+ size 1976163472
self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0061bb1e0217ca29f5bec8242217547e52ae8bb0968dd3726fde9d59807de643
3
+ size 3952509771
self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2258c8f7e6687f5fa2dcc6cd5ebbafa0a88234fdbe6ce45ab0d52ed8947a2e83
3
+ size 14645
self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aab883bcab393f0c916048be23a6c0284269524eb2950c4249e03acc7cac4f75
3
+ size 1465
self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
3
+ size 11421892
self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/tokenizer_config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": null,
5
+ "clean_up_tokenization_spaces": false,
6
+ "eos_token": "<|im_end|>",
7
+ "errors": "replace",
8
+ "extra_special_tokens": [
9
+ "<|im_start|>",
10
+ "<|im_end|>",
11
+ "<|object_ref_start|>",
12
+ "<|object_ref_end|>",
13
+ "<|box_start|>",
14
+ "<|box_end|>",
15
+ "<|quad_start|>",
16
+ "<|quad_end|>",
17
+ "<|vision_start|>",
18
+ "<|vision_end|>",
19
+ "<|vision_pad|>",
20
+ "<|image_pad|>",
21
+ "<|video_pad|>"
22
+ ],
23
+ "is_local": true,
24
+ "local_files_only": false,
25
+ "model_max_length": 131072,
26
+ "pad_token": "<|endoftext|>",
27
+ "padding_side": "left",
28
+ "split_special_tokens": false,
29
+ "tokenizer_class": "Qwen2Tokenizer",
30
+ "truncation_side": "left",
31
+ "unk_token": null
32
+ }
self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/trainer_state.json ADDED
@@ -0,0 +1,1129 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 1.5,
6
+ "eval_steps": 500,
7
+ "global_step": 60,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "clip_ratio/high_max": 0.0,
14
+ "clip_ratio/high_mean": 0.0,
15
+ "clip_ratio/low_mean": 0.0,
16
+ "clip_ratio/low_min": 0.0,
17
+ "clip_ratio/region_mean": 0.0,
18
+ "completions/clipped_ratio": 1.0,
19
+ "completions/max_length": 256.0,
20
+ "completions/max_terminated_length": 0.0,
21
+ "completions/mean_length": 256.0,
22
+ "completions/mean_terminated_length": 0.0,
23
+ "completions/min_length": 256.0,
24
+ "completions/min_terminated_length": 0.0,
25
+ "entropy": 1.7747459411621094,
26
+ "epoch": 0.025,
27
+ "frac_reward_zero_std": 0.5,
28
+ "grad_norm": 2.0766546726226807,
29
+ "kl": 0.0,
30
+ "learning_rate": 3e-06,
31
+ "loss": -0.10802466422319412,
32
+ "num_tokens": 9960.0,
33
+ "reward": 0.861758291721344,
34
+ "reward_std": 0.1182720735669136,
35
+ "rewards/AnswererRewardFunction/mean": 0.861758291721344,
36
+ "rewards/AnswererRewardFunction/std": 0.1182720735669136,
37
+ "step": 1,
38
+ "step_time": 7.9753973850001785
39
+ },
40
+ {
41
+ "clip_ratio/high_max": 0.048828125,
42
+ "clip_ratio/high_mean": 0.048828125,
43
+ "clip_ratio/low_mean": 0.0,
44
+ "clip_ratio/low_min": 0.0,
45
+ "clip_ratio/region_mean": 0.048828125,
46
+ "entropy": 1.9674677848815918,
47
+ "epoch": 0.05,
48
+ "grad_norm": 1.2367463111877441,
49
+ "kl": 0.12442716211080551,
50
+ "learning_rate": 2.9750000000000003e-06,
51
+ "loss": -0.061602603644132614,
52
+ "step": 2,
53
+ "step_time": 0.1180487430001449
54
+ },
55
+ {
56
+ "clip_ratio/high_max": 0.0,
57
+ "clip_ratio/high_mean": 0.0,
58
+ "clip_ratio/low_mean": 0.046875,
59
+ "clip_ratio/low_min": 0.046875,
60
+ "clip_ratio/region_mean": 0.046875,
61
+ "entropy": 2.1487021446228027,
62
+ "epoch": 0.075,
63
+ "grad_norm": 3.5523228645324707,
64
+ "kl": 0.12029533833265305,
65
+ "learning_rate": 2.9499999999999997e-06,
66
+ "loss": 0.15017201006412506,
67
+ "step": 3,
68
+ "step_time": 0.11217075500007923
69
+ },
70
+ {
71
+ "clip_ratio/high_max": 0.0263671875,
72
+ "clip_ratio/high_mean": 0.0263671875,
73
+ "clip_ratio/low_mean": 0.03125,
74
+ "clip_ratio/low_min": 0.03125,
75
+ "clip_ratio/region_mean": 0.0576171875,
76
+ "entropy": 2.007148265838623,
77
+ "epoch": 0.1,
78
+ "grad_norm": 1.966639518737793,
79
+ "kl": 0.10988782346248627,
80
+ "learning_rate": 2.925e-06,
81
+ "loss": 0.02496827021241188,
82
+ "step": 4,
83
+ "step_time": 0.11107316000015999
84
+ },
85
+ {
86
+ "clip_ratio/high_max": 0.0,
87
+ "clip_ratio/high_mean": 0.0,
88
+ "clip_ratio/low_mean": 0.0,
89
+ "clip_ratio/low_min": 0.0,
90
+ "clip_ratio/region_mean": 0.0,
91
+ "completions/clipped_ratio": 1.0,
92
+ "completions/max_length": 256.0,
93
+ "completions/max_terminated_length": 0.0,
94
+ "completions/mean_length": 256.0,
95
+ "completions/mean_terminated_length": 0.0,
96
+ "completions/min_length": 256.0,
97
+ "completions/min_terminated_length": 0.0,
98
+ "entropy": 2.053436756134033,
99
+ "epoch": 0.125,
100
+ "frac_reward_zero_std": 1.0,
101
+ "grad_norm": 0.023533064872026443,
102
+ "kl": 0.03712207078933716,
103
+ "learning_rate": 2.9e-06,
104
+ "loss": 9.280518133891746e-05,
105
+ "num_tokens": 19772.0,
106
+ "reward": 0.775258481502533,
107
+ "reward_std": 0.09679953753948212,
108
+ "rewards/AnswererRewardFunction/mean": 0.775258481502533,
109
+ "rewards/AnswererRewardFunction/std": 0.09679954499006271,
110
+ "step": 5,
111
+ "step_time": 7.949325107000732
112
+ },
113
+ {
114
+ "clip_ratio/high_max": 0.0,
115
+ "clip_ratio/high_mean": 0.0,
116
+ "clip_ratio/low_mean": 0.0,
117
+ "clip_ratio/low_min": 0.0,
118
+ "clip_ratio/region_mean": 0.0,
119
+ "entropy": 1.9749951362609863,
120
+ "epoch": 0.15,
121
+ "grad_norm": 0.019975125789642334,
122
+ "kl": 0.04075080528855324,
123
+ "learning_rate": 2.875e-06,
124
+ "loss": 0.00010187701263930649,
125
+ "step": 6,
126
+ "step_time": 0.11253355799999554
127
+ },
128
+ {
129
+ "clip_ratio/high_max": 0.0,
130
+ "clip_ratio/high_mean": 0.0,
131
+ "clip_ratio/low_mean": 0.0,
132
+ "clip_ratio/low_min": 0.0,
133
+ "clip_ratio/region_mean": 0.0,
134
+ "entropy": 1.901890754699707,
135
+ "epoch": 0.175,
136
+ "grad_norm": 0.040124714374542236,
137
+ "kl": 0.06329751014709473,
138
+ "learning_rate": 2.85e-06,
139
+ "loss": 0.00015824378351680934,
140
+ "step": 7,
141
+ "step_time": 0.1121210709998195
142
+ },
143
+ {
144
+ "clip_ratio/high_max": 0.0,
145
+ "clip_ratio/high_mean": 0.0,
146
+ "clip_ratio/low_mean": 0.0,
147
+ "clip_ratio/low_min": 0.0,
148
+ "clip_ratio/region_mean": 0.0,
149
+ "entropy": 1.9866925477981567,
150
+ "epoch": 0.2,
151
+ "grad_norm": 0.28295770287513733,
152
+ "kl": 0.07546718418598175,
153
+ "learning_rate": 2.825e-06,
154
+ "loss": 0.00018866793834604323,
155
+ "step": 8,
156
+ "step_time": 0.11143768600049953
157
+ },
158
+ {
159
+ "clip_ratio/high_max": 0.0,
160
+ "clip_ratio/high_mean": 0.0,
161
+ "clip_ratio/low_mean": 0.0,
162
+ "clip_ratio/low_min": 0.0,
163
+ "clip_ratio/region_mean": 0.0,
164
+ "completions/clipped_ratio": 1.0,
165
+ "completions/max_length": 256.0,
166
+ "completions/max_terminated_length": 0.0,
167
+ "completions/mean_length": 256.0,
168
+ "completions/mean_terminated_length": 0.0,
169
+ "completions/min_length": 256.0,
170
+ "completions/min_terminated_length": 0.0,
171
+ "entropy": 1.887934923171997,
172
+ "epoch": 0.225,
173
+ "frac_reward_zero_std": 1.0,
174
+ "grad_norm": 0.02717243880033493,
175
+ "kl": 0.05967765301465988,
176
+ "learning_rate": 2.8000000000000003e-06,
177
+ "loss": 0.00014919412205927074,
178
+ "num_tokens": 29608.0,
179
+ "reward": 0.8826282620429993,
180
+ "reward_std": 0.10600326210260391,
181
+ "rewards/AnswererRewardFunction/mean": 0.8826282620429993,
182
+ "rewards/AnswererRewardFunction/std": 0.10600324720144272,
183
+ "step": 9,
184
+ "step_time": 7.95751762799955
185
+ },
186
+ {
187
+ "clip_ratio/high_max": 0.0,
188
+ "clip_ratio/high_mean": 0.0,
189
+ "clip_ratio/low_mean": 0.0,
190
+ "clip_ratio/low_min": 0.0,
191
+ "clip_ratio/region_mean": 0.0,
192
+ "entropy": 1.672785758972168,
193
+ "epoch": 0.25,
194
+ "grad_norm": 0.03259706869721413,
195
+ "kl": 0.07357163727283478,
196
+ "learning_rate": 2.775e-06,
197
+ "loss": 0.00018392910715192556,
198
+ "step": 10,
199
+ "step_time": 0.1110475139994378
200
+ },
201
+ {
202
+ "clip_ratio/high_max": 0.0,
203
+ "clip_ratio/high_mean": 0.0,
204
+ "clip_ratio/low_mean": 0.0,
205
+ "clip_ratio/low_min": 0.0,
206
+ "clip_ratio/region_mean": 0.0,
207
+ "entropy": 2.054617404937744,
208
+ "epoch": 0.275,
209
+ "grad_norm": 0.034495558589696884,
210
+ "kl": 0.07685650885105133,
211
+ "learning_rate": 2.75e-06,
212
+ "loss": 0.00019214126223232597,
213
+ "step": 11,
214
+ "step_time": 0.1110302529996261
215
+ },
216
+ {
217
+ "clip_ratio/high_max": 0.0,
218
+ "clip_ratio/high_mean": 0.0,
219
+ "clip_ratio/low_mean": 0.0,
220
+ "clip_ratio/low_min": 0.0,
221
+ "clip_ratio/region_mean": 0.0,
222
+ "entropy": 2.0550127029418945,
223
+ "epoch": 0.3,
224
+ "grad_norm": 0.029428452253341675,
225
+ "kl": 0.07294180989265442,
226
+ "learning_rate": 2.725e-06,
227
+ "loss": 0.0001823545026127249,
228
+ "step": 12,
229
+ "step_time": 0.11090638099994976
230
+ },
231
+ {
232
+ "clip_ratio/high_max": 0.0,
233
+ "clip_ratio/high_mean": 0.0,
234
+ "clip_ratio/low_mean": 0.0,
235
+ "clip_ratio/low_min": 0.0,
236
+ "clip_ratio/region_mean": 0.0,
237
+ "completions/clipped_ratio": 1.0,
238
+ "completions/max_length": 256.0,
239
+ "completions/max_terminated_length": 0.0,
240
+ "completions/mean_length": 256.0,
241
+ "completions/mean_terminated_length": 0.0,
242
+ "completions/min_length": 256.0,
243
+ "completions/min_terminated_length": 0.0,
244
+ "entropy": 1.8360331058502197,
245
+ "epoch": 0.325,
246
+ "frac_reward_zero_std": 0.75,
247
+ "grad_norm": 2.9922423362731934,
248
+ "kl": 0.07551632076501846,
249
+ "learning_rate": 2.7e-06,
250
+ "loss": -0.06226314604282379,
251
+ "num_tokens": 39580.0,
252
+ "reward": 0.9046050310134888,
253
+ "reward_std": 0.08236623555421829,
254
+ "rewards/AnswererRewardFunction/mean": 0.9046050310134888,
255
+ "rewards/AnswererRewardFunction/std": 0.0823662281036377,
256
+ "step": 13,
257
+ "step_time": 7.9857917509998515
258
+ },
259
+ {
260
+ "clip_ratio/high_max": 0.0,
261
+ "clip_ratio/high_mean": 0.0,
262
+ "clip_ratio/low_mean": 0.0,
263
+ "clip_ratio/low_min": 0.0,
264
+ "clip_ratio/region_mean": 0.0,
265
+ "entropy": 2.2757010459899902,
266
+ "epoch": 0.35,
267
+ "grad_norm": 0.032990459352731705,
268
+ "kl": 0.07481437176465988,
269
+ "learning_rate": 2.6750000000000002e-06,
270
+ "loss": 0.0001870359410531819,
271
+ "step": 14,
272
+ "step_time": 0.11140432199954375
273
+ },
274
+ {
275
+ "clip_ratio/high_max": 0.0,
276
+ "clip_ratio/high_mean": 0.0,
277
+ "clip_ratio/low_mean": 0.0009765625,
278
+ "clip_ratio/low_min": 0.0009765625,
279
+ "clip_ratio/region_mean": 0.0009765625,
280
+ "entropy": 1.8182785511016846,
281
+ "epoch": 0.375,
282
+ "grad_norm": 0.8941988348960876,
283
+ "kl": 0.10425875335931778,
284
+ "learning_rate": 2.65e-06,
285
+ "loss": 0.0317252017557621,
286
+ "step": 15,
287
+ "step_time": 0.1321398499994757
288
+ },
289
+ {
290
+ "clip_ratio/high_max": 0.0,
291
+ "clip_ratio/high_mean": 0.0,
292
+ "clip_ratio/low_mean": 0.0068359375,
293
+ "clip_ratio/low_min": 0.0068359375,
294
+ "clip_ratio/region_mean": 0.0068359375,
295
+ "entropy": 1.4916911125183105,
296
+ "epoch": 0.4,
297
+ "grad_norm": 0.8881436586380005,
298
+ "kl": 0.08638335019350052,
299
+ "learning_rate": 2.6250000000000003e-06,
300
+ "loss": 0.031212178990244865,
301
+ "step": 16,
302
+ "step_time": 0.11174384800051484
303
+ },
304
+ {
305
+ "clip_ratio/high_max": 0.0,
306
+ "clip_ratio/high_mean": 0.0,
307
+ "clip_ratio/low_mean": 0.0,
308
+ "clip_ratio/low_min": 0.0,
309
+ "clip_ratio/region_mean": 0.0,
310
+ "completions/clipped_ratio": 1.0,
311
+ "completions/max_length": 256.0,
312
+ "completions/max_terminated_length": 0.0,
313
+ "completions/mean_length": 256.0,
314
+ "completions/mean_terminated_length": 0.0,
315
+ "completions/min_length": 256.0,
316
+ "completions/min_terminated_length": 0.0,
317
+ "entropy": 2.031963348388672,
318
+ "epoch": 0.425,
319
+ "frac_reward_zero_std": 1.0,
320
+ "grad_norm": 0.045982908457517624,
321
+ "kl": 0.10716703534126282,
322
+ "learning_rate": 2.6e-06,
323
+ "loss": 0.00026791758136823773,
324
+ "num_tokens": 49960.0,
325
+ "reward": 0.9323133230209351,
326
+ "reward_std": 0.0,
327
+ "rewards/AnswererRewardFunction/mean": 0.9323133230209351,
328
+ "rewards/AnswererRewardFunction/std": 0.0,
329
+ "step": 17,
330
+ "step_time": 7.907421807999526
331
+ },
332
+ {
333
+ "clip_ratio/high_max": 0.0,
334
+ "clip_ratio/high_mean": 0.0,
335
+ "clip_ratio/low_mean": 0.0,
336
+ "clip_ratio/low_min": 0.0,
337
+ "clip_ratio/region_mean": 0.0,
338
+ "entropy": 1.8369736671447754,
339
+ "epoch": 0.45,
340
+ "grad_norm": 0.045823127031326294,
341
+ "kl": 0.10512565076351166,
342
+ "learning_rate": 2.575e-06,
343
+ "loss": 0.00026281410828232765,
344
+ "step": 18,
345
+ "step_time": 0.11191840300034528
346
+ },
347
+ {
348
+ "clip_ratio/high_max": 0.0,
349
+ "clip_ratio/high_mean": 0.0,
350
+ "clip_ratio/low_mean": 0.0,
351
+ "clip_ratio/low_min": 0.0,
352
+ "clip_ratio/region_mean": 0.0,
353
+ "entropy": 1.9750962257385254,
354
+ "epoch": 0.475,
355
+ "grad_norm": 0.03864246979355812,
356
+ "kl": 0.0918373316526413,
357
+ "learning_rate": 2.55e-06,
358
+ "loss": 0.00022959333728067577,
359
+ "step": 19,
360
+ "step_time": 0.11069956699975592
361
+ },
362
+ {
363
+ "clip_ratio/high_max": 0.0,
364
+ "clip_ratio/high_mean": 0.0,
365
+ "clip_ratio/low_mean": 0.0,
366
+ "clip_ratio/low_min": 0.0,
367
+ "clip_ratio/region_mean": 0.0,
368
+ "entropy": 2.031843662261963,
369
+ "epoch": 0.5,
370
+ "grad_norm": 0.05079154297709465,
371
+ "kl": 0.12332479655742645,
372
+ "learning_rate": 2.525e-06,
373
+ "loss": 0.0003083120100200176,
374
+ "step": 20,
375
+ "step_time": 0.11133921100008592
376
+ },
377
+ {
378
+ "clip_ratio/high_max": 0.0,
379
+ "clip_ratio/high_mean": 0.0,
380
+ "clip_ratio/low_mean": 0.0,
381
+ "clip_ratio/low_min": 0.0,
382
+ "clip_ratio/region_mean": 0.0,
383
+ "completions/clipped_ratio": 1.0,
384
+ "completions/max_length": 256.0,
385
+ "completions/max_terminated_length": 0.0,
386
+ "completions/mean_length": 256.0,
387
+ "completions/mean_terminated_length": 0.0,
388
+ "completions/min_length": 256.0,
389
+ "completions/min_terminated_length": 0.0,
390
+ "entropy": 1.8194365501403809,
391
+ "epoch": 0.525,
392
+ "frac_reward_zero_std": 1.0,
393
+ "grad_norm": 0.23991960287094116,
394
+ "kl": 0.13239404559135437,
395
+ "learning_rate": 2.5e-06,
396
+ "loss": 0.00033098511630669236,
397
+ "num_tokens": 60148.0,
398
+ "reward": 0.819347620010376,
399
+ "reward_std": 0.11957819759845734,
400
+ "rewards/AnswererRewardFunction/mean": 0.819347620010376,
401
+ "rewards/AnswererRewardFunction/std": 0.11957821249961853,
402
+ "step": 21,
403
+ "step_time": 7.9142402379993655
404
+ },
405
+ {
406
+ "clip_ratio/high_max": 0.0,
407
+ "clip_ratio/high_mean": 0.0,
408
+ "clip_ratio/low_mean": 0.0,
409
+ "clip_ratio/low_min": 0.0,
410
+ "clip_ratio/region_mean": 0.0,
411
+ "entropy": 1.333341360092163,
412
+ "epoch": 0.55,
413
+ "grad_norm": 0.04438238963484764,
414
+ "kl": 0.07663387060165405,
415
+ "learning_rate": 2.475e-06,
416
+ "loss": 0.00019158469513058662,
417
+ "step": 22,
418
+ "step_time": 0.11160182699950383
419
+ },
420
+ {
421
+ "clip_ratio/high_max": 0.0,
422
+ "clip_ratio/high_mean": 0.0,
423
+ "clip_ratio/low_mean": 0.0,
424
+ "clip_ratio/low_min": 0.0,
425
+ "clip_ratio/region_mean": 0.0,
426
+ "entropy": 1.8490242958068848,
427
+ "epoch": 0.575,
428
+ "grad_norm": 0.04169222339987755,
429
+ "kl": 0.10077071189880371,
430
+ "learning_rate": 2.45e-06,
431
+ "loss": 0.00025192677276208997,
432
+ "step": 23,
433
+ "step_time": 0.11161970700050006
434
+ },
435
+ {
436
+ "clip_ratio/high_max": 0.0,
437
+ "clip_ratio/high_mean": 0.0,
438
+ "clip_ratio/low_mean": 0.0,
439
+ "clip_ratio/low_min": 0.0,
440
+ "clip_ratio/region_mean": 0.0,
441
+ "entropy": 1.6197292804718018,
442
+ "epoch": 0.6,
443
+ "grad_norm": 0.05068081617355347,
444
+ "kl": 0.12023425102233887,
445
+ "learning_rate": 2.425e-06,
446
+ "loss": 0.0003005856415256858,
447
+ "step": 24,
448
+ "step_time": 0.11153809700044803
449
+ },
450
+ {
451
+ "clip_ratio/high_max": 0.0,
452
+ "clip_ratio/high_mean": 0.0,
453
+ "clip_ratio/low_mean": 0.0,
454
+ "clip_ratio/low_min": 0.0,
455
+ "clip_ratio/region_mean": 0.0,
456
+ "completions/clipped_ratio": 1.0,
457
+ "completions/max_length": 256.0,
458
+ "completions/max_terminated_length": 0.0,
459
+ "completions/mean_length": 256.0,
460
+ "completions/mean_terminated_length": 0.0,
461
+ "completions/min_length": 256.0,
462
+ "completions/min_terminated_length": 0.0,
463
+ "entropy": 1.7150523662567139,
464
+ "epoch": 0.625,
465
+ "frac_reward_zero_std": 1.0,
466
+ "grad_norm": 0.03134142979979515,
467
+ "kl": 0.09788648784160614,
468
+ "learning_rate": 2.4000000000000003e-06,
469
+ "loss": 0.0002447162405587733,
470
+ "num_tokens": 70092.0,
471
+ "reward": 0.8280093669891357,
472
+ "reward_std": 0.10857455432415009,
473
+ "rewards/AnswererRewardFunction/mean": 0.8280093669891357,
474
+ "rewards/AnswererRewardFunction/std": 0.10857456177473068,
475
+ "step": 25,
476
+ "step_time": 7.957390386000043
477
+ },
478
+ {
479
+ "clip_ratio/high_max": 0.0,
480
+ "clip_ratio/high_mean": 0.0,
481
+ "clip_ratio/low_mean": 0.0,
482
+ "clip_ratio/low_min": 0.0,
483
+ "clip_ratio/region_mean": 0.0,
484
+ "entropy": 1.5895636081695557,
485
+ "epoch": 0.65,
486
+ "grad_norm": 0.04325919598340988,
487
+ "kl": 0.1107931137084961,
488
+ "learning_rate": 2.375e-06,
489
+ "loss": 0.00027698278427124023,
490
+ "step": 26,
491
+ "step_time": 0.1117739569999685
492
+ },
493
+ {
494
+ "clip_ratio/high_max": 0.0,
495
+ "clip_ratio/high_mean": 0.0,
496
+ "clip_ratio/low_mean": 0.0,
497
+ "clip_ratio/low_min": 0.0,
498
+ "clip_ratio/region_mean": 0.0,
499
+ "entropy": 1.73920476436615,
500
+ "epoch": 0.675,
501
+ "grad_norm": 0.03872699663043022,
502
+ "kl": 0.10983294248580933,
503
+ "learning_rate": 2.35e-06,
504
+ "loss": 0.00027458235854282975,
505
+ "step": 27,
506
+ "step_time": 0.11149289300010423
507
+ },
508
+ {
509
+ "clip_ratio/high_max": 0.0,
510
+ "clip_ratio/high_mean": 0.0,
511
+ "clip_ratio/low_mean": 0.0,
512
+ "clip_ratio/low_min": 0.0,
513
+ "clip_ratio/region_mean": 0.0,
514
+ "entropy": 1.898261308670044,
515
+ "epoch": 0.7,
516
+ "grad_norm": 0.04079338535666466,
517
+ "kl": 0.11259151250123978,
518
+ "learning_rate": 2.325e-06,
519
+ "loss": 0.00028147880220785737,
520
+ "step": 28,
521
+ "step_time": 0.11117797499991866
522
+ },
523
+ {
524
+ "clip_ratio/high_max": 0.0,
525
+ "clip_ratio/high_mean": 0.0,
526
+ "clip_ratio/low_mean": 0.0,
527
+ "clip_ratio/low_min": 0.0,
528
+ "clip_ratio/region_mean": 0.0,
529
+ "completions/clipped_ratio": 1.0,
530
+ "completions/max_length": 256.0,
531
+ "completions/max_terminated_length": 0.0,
532
+ "completions/mean_length": 256.0,
533
+ "completions/mean_terminated_length": 0.0,
534
+ "completions/min_length": 256.0,
535
+ "completions/min_terminated_length": 0.0,
536
+ "entropy": 1.9487004280090332,
537
+ "epoch": 0.725,
538
+ "frac_reward_zero_std": 1.0,
539
+ "grad_norm": 0.047494642436504364,
540
+ "kl": 0.11588452756404877,
541
+ "learning_rate": 2.3000000000000004e-06,
542
+ "loss": 0.0002897112863138318,
543
+ "num_tokens": 80364.0,
544
+ "reward": 0.8668582439422607,
545
+ "reward_std": 0.1170896589756012,
546
+ "rewards/AnswererRewardFunction/mean": 0.8668582439422607,
547
+ "rewards/AnswererRewardFunction/std": 0.1170896664261818,
548
+ "step": 29,
549
+ "step_time": 7.972899151000092
550
+ },
551
+ {
552
+ "clip_ratio/high_max": 0.0,
553
+ "clip_ratio/high_mean": 0.0,
554
+ "clip_ratio/low_mean": 0.0,
555
+ "clip_ratio/low_min": 0.0,
556
+ "clip_ratio/region_mean": 0.0,
557
+ "entropy": 1.7704535722732544,
558
+ "epoch": 0.75,
559
+ "grad_norm": 0.02996332012116909,
560
+ "kl": 0.09660449624061584,
561
+ "learning_rate": 2.275e-06,
562
+ "loss": 0.00024151125398930162,
563
+ "step": 30,
564
+ "step_time": 0.11398699500023213
565
+ },
566
+ {
567
+ "clip_ratio/high_max": 0.0,
568
+ "clip_ratio/high_mean": 0.0,
569
+ "clip_ratio/low_mean": 0.0,
570
+ "clip_ratio/low_min": 0.0,
571
+ "clip_ratio/region_mean": 0.0,
572
+ "entropy": 1.766472339630127,
573
+ "epoch": 0.775,
574
+ "grad_norm": 0.03504747897386551,
575
+ "kl": 0.08477599918842316,
576
+ "learning_rate": 2.25e-06,
577
+ "loss": 0.00021194000146351755,
578
+ "step": 31,
579
+ "step_time": 0.11211207500036835
580
+ },
581
+ {
582
+ "clip_ratio/high_max": 0.0,
583
+ "clip_ratio/high_mean": 0.0,
584
+ "clip_ratio/low_mean": 0.0,
585
+ "clip_ratio/low_min": 0.0,
586
+ "clip_ratio/region_mean": 0.0,
587
+ "entropy": 1.4578940868377686,
588
+ "epoch": 0.8,
589
+ "grad_norm": 0.02072232775390148,
590
+ "kl": 0.06009065359830856,
591
+ "learning_rate": 2.2250000000000003e-06,
592
+ "loss": 0.00015022660954855382,
593
+ "step": 32,
594
+ "step_time": 0.1105231419996926
595
+ },
596
+ {
597
+ "clip_ratio/high_max": 0.0,
598
+ "clip_ratio/high_mean": 0.0,
599
+ "clip_ratio/low_mean": 0.0,
600
+ "clip_ratio/low_min": 0.0,
601
+ "clip_ratio/region_mean": 0.0,
602
+ "completions/clipped_ratio": 1.0,
603
+ "completions/max_length": 256.0,
604
+ "completions/max_terminated_length": 0.0,
605
+ "completions/mean_length": 256.0,
606
+ "completions/mean_terminated_length": 0.0,
607
+ "completions/min_length": 256.0,
608
+ "completions/min_terminated_length": 0.0,
609
+ "entropy": 1.7355934381484985,
610
+ "epoch": 0.825,
611
+ "frac_reward_zero_std": 0.75,
612
+ "grad_norm": 1.4318349361419678,
613
+ "kl": 0.10250318050384521,
614
+ "learning_rate": 2.1999999999999997e-06,
615
+ "loss": -0.06233750656247139,
616
+ "num_tokens": 90796.0,
617
+ "reward": 0.9159495830535889,
618
+ "reward_std": 0.06545510143041611,
619
+ "rewards/AnswererRewardFunction/mean": 0.9159495830535889,
620
+ "rewards/AnswererRewardFunction/std": 0.0654551088809967,
621
+ "step": 33,
622
+ "step_time": 7.989859212000738
623
+ },
624
+ {
625
+ "clip_ratio/high_max": 0.0,
626
+ "clip_ratio/high_mean": 0.0,
627
+ "clip_ratio/low_mean": 0.0009765625,
628
+ "clip_ratio/low_min": 0.0009765625,
629
+ "clip_ratio/region_mean": 0.0009765625,
630
+ "entropy": 1.732600450515747,
631
+ "epoch": 0.85,
632
+ "grad_norm": 2.944581985473633,
633
+ "kl": 0.08287997543811798,
634
+ "learning_rate": 2.175e-06,
635
+ "loss": 0.06262950599193573,
636
+ "step": 34,
637
+ "step_time": 0.11203606500021124
638
+ },
639
+ {
640
+ "clip_ratio/high_max": 0.0,
641
+ "clip_ratio/high_mean": 0.0,
642
+ "clip_ratio/low_mean": 0.0,
643
+ "clip_ratio/low_min": 0.0,
644
+ "clip_ratio/region_mean": 0.0,
645
+ "entropy": 1.9220362901687622,
646
+ "epoch": 0.875,
647
+ "grad_norm": 0.03256967291235924,
648
+ "kl": 0.08927878737449646,
649
+ "learning_rate": 2.15e-06,
650
+ "loss": 0.00022319695563055575,
651
+ "step": 35,
652
+ "step_time": 0.1124586719997751
653
+ },
654
+ {
655
+ "clip_ratio/high_max": 0.0,
656
+ "clip_ratio/high_mean": 0.0,
657
+ "clip_ratio/low_mean": 0.0,
658
+ "clip_ratio/low_min": 0.0,
659
+ "clip_ratio/region_mean": 0.0,
660
+ "entropy": 1.7540044784545898,
661
+ "epoch": 0.9,
662
+ "grad_norm": 0.03057776764035225,
663
+ "kl": 0.08942355215549469,
664
+ "learning_rate": 2.125e-06,
665
+ "loss": 0.0002235588908661157,
666
+ "step": 36,
667
+ "step_time": 0.11161630500009778
668
+ },
669
+ {
670
+ "clip_ratio/high_max": 0.0,
671
+ "clip_ratio/high_mean": 0.0,
672
+ "clip_ratio/low_mean": 0.0,
673
+ "clip_ratio/low_min": 0.0,
674
+ "clip_ratio/region_mean": 0.0,
675
+ "completions/clipped_ratio": 1.0,
676
+ "completions/max_length": 256.0,
677
+ "completions/max_terminated_length": 0.0,
678
+ "completions/mean_length": 256.0,
679
+ "completions/mean_terminated_length": 0.0,
680
+ "completions/min_length": 256.0,
681
+ "completions/min_terminated_length": 0.0,
682
+ "entropy": 1.8761208057403564,
683
+ "epoch": 0.925,
684
+ "frac_reward_zero_std": 1.0,
685
+ "grad_norm": 0.053888410329818726,
686
+ "kl": 0.10505490750074387,
687
+ "learning_rate": 2.1e-06,
688
+ "loss": 0.00026263727340847254,
689
+ "num_tokens": 100696.0,
690
+ "reward": 0.8227691054344177,
691
+ "reward_std": 0.11510813981294632,
692
+ "rewards/AnswererRewardFunction/mean": 0.8227691054344177,
693
+ "rewards/AnswererRewardFunction/std": 0.11510813981294632,
694
+ "step": 37,
695
+ "step_time": 7.976926989000276
696
+ },
697
+ {
698
+ "clip_ratio/high_max": 0.0,
699
+ "clip_ratio/high_mean": 0.0,
700
+ "clip_ratio/low_mean": 0.0,
701
+ "clip_ratio/low_min": 0.0,
702
+ "clip_ratio/region_mean": 0.0,
703
+ "entropy": 1.6150931119918823,
704
+ "epoch": 0.95,
705
+ "grad_norm": 0.028759539127349854,
706
+ "kl": 0.0988161563873291,
707
+ "learning_rate": 2.075e-06,
708
+ "loss": 0.0002470404142513871,
709
+ "step": 38,
710
+ "step_time": 0.1113082760002726
711
+ },
712
+ {
713
+ "clip_ratio/high_max": 0.0,
714
+ "clip_ratio/high_mean": 0.0,
715
+ "clip_ratio/low_mean": 0.0,
716
+ "clip_ratio/low_min": 0.0,
717
+ "clip_ratio/region_mean": 0.0,
718
+ "entropy": 2.035503387451172,
719
+ "epoch": 0.975,
720
+ "grad_norm": 0.04202085733413696,
721
+ "kl": 0.11065234243869781,
722
+ "learning_rate": 2.0500000000000003e-06,
723
+ "loss": 0.00027663083164952695,
724
+ "step": 39,
725
+ "step_time": 0.11134837800000241
726
+ },
727
+ {
728
+ "clip_ratio/high_max": 0.0,
729
+ "clip_ratio/high_mean": 0.0,
730
+ "clip_ratio/low_mean": 0.0,
731
+ "clip_ratio/low_min": 0.0,
732
+ "clip_ratio/region_mean": 0.0,
733
+ "entropy": 1.9588853120803833,
734
+ "epoch": 1.0,
735
+ "grad_norm": 0.029340602457523346,
736
+ "kl": 0.09918107837438583,
737
+ "learning_rate": 2.025e-06,
738
+ "loss": 0.0002479527029208839,
739
+ "step": 40,
740
+ "step_time": 0.11250619200018264
741
+ },
742
+ {
743
+ "clip_ratio/high_max": 0.0,
744
+ "clip_ratio/high_mean": 0.0,
745
+ "clip_ratio/low_mean": 0.0,
746
+ "clip_ratio/low_min": 0.0,
747
+ "clip_ratio/region_mean": 0.0,
748
+ "completions/clipped_ratio": 1.0,
749
+ "completions/max_length": 256.0,
750
+ "completions/max_terminated_length": 0.0,
751
+ "completions/mean_length": 256.0,
752
+ "completions/mean_terminated_length": 0.0,
753
+ "completions/min_length": 256.0,
754
+ "completions/min_terminated_length": 0.0,
755
+ "entropy": 1.903963327407837,
756
+ "epoch": 1.025,
757
+ "frac_reward_zero_std": 0.25,
758
+ "grad_norm": 2.2654471397399902,
759
+ "kl": 0.10239098221063614,
760
+ "learning_rate": 2e-06,
761
+ "loss": 0.035676077008247375,
762
+ "num_tokens": 110284.0,
763
+ "reward": 0.7121027708053589,
764
+ "reward_std": 0.12839390337467194,
765
+ "rewards/AnswererRewardFunction/mean": 0.7121027708053589,
766
+ "rewards/AnswererRewardFunction/std": 0.12839393317699432,
767
+ "step": 41,
768
+ "step_time": 7.889583379000214
769
+ },
770
+ {
771
+ "clip_ratio/high_max": 0.0,
772
+ "clip_ratio/high_mean": 0.0,
773
+ "clip_ratio/low_mean": 0.0,
774
+ "clip_ratio/low_min": 0.0,
775
+ "clip_ratio/region_mean": 0.0,
776
+ "entropy": 2.0613226890563965,
777
+ "epoch": 1.05,
778
+ "grad_norm": 3.5304839611053467,
779
+ "kl": 0.10499992966651917,
780
+ "learning_rate": 1.975e-06,
781
+ "loss": 0.13377703726291656,
782
+ "step": 42,
783
+ "step_time": 0.11177234499973565
784
+ },
785
+ {
786
+ "clip_ratio/high_max": 0.0146484375,
787
+ "clip_ratio/high_mean": 0.0146484375,
788
+ "clip_ratio/low_mean": 0.0,
789
+ "clip_ratio/low_min": 0.0,
790
+ "clip_ratio/region_mean": 0.0146484375,
791
+ "entropy": 1.7200109958648682,
792
+ "epoch": 1.075,
793
+ "grad_norm": 2.346640110015869,
794
+ "kl": 0.11238373816013336,
795
+ "learning_rate": 1.95e-06,
796
+ "loss": -0.1459183692932129,
797
+ "step": 43,
798
+ "step_time": 0.11070189299971389
799
+ },
800
+ {
801
+ "clip_ratio/high_max": 0.0146484375,
802
+ "clip_ratio/high_mean": 0.0146484375,
803
+ "clip_ratio/low_mean": 0.00390625,
804
+ "clip_ratio/low_min": 0.00390625,
805
+ "clip_ratio/region_mean": 0.0185546875,
806
+ "entropy": 1.9567177295684814,
807
+ "epoch": 1.1,
808
+ "grad_norm": 3.576796531677246,
809
+ "kl": 0.12578752636909485,
810
+ "learning_rate": 1.925e-06,
811
+ "loss": -0.021488450467586517,
812
+ "step": 44,
813
+ "step_time": 0.11009783999998035
814
+ },
815
+ {
816
+ "clip_ratio/high_max": 0.0,
817
+ "clip_ratio/high_mean": 0.0,
818
+ "clip_ratio/low_mean": 0.0,
819
+ "clip_ratio/low_min": 0.0,
820
+ "clip_ratio/region_mean": 0.0,
821
+ "completions/clipped_ratio": 1.0,
822
+ "completions/max_length": 256.0,
823
+ "completions/max_terminated_length": 0.0,
824
+ "completions/mean_length": 256.0,
825
+ "completions/mean_terminated_length": 0.0,
826
+ "completions/min_length": 256.0,
827
+ "completions/min_terminated_length": 0.0,
828
+ "entropy": 2.0115418434143066,
829
+ "epoch": 1.125,
830
+ "frac_reward_zero_std": 0.75,
831
+ "grad_norm": 0.046743664890527725,
832
+ "kl": 0.11043193191289902,
833
+ "learning_rate": 1.9e-06,
834
+ "loss": 0.00027607983793132007,
835
+ "num_tokens": 120284.0,
836
+ "reward": 0.8719930648803711,
837
+ "reward_std": 0.10185166448354721,
838
+ "rewards/AnswererRewardFunction/mean": 0.8719930648803711,
839
+ "rewards/AnswererRewardFunction/std": 0.10185166448354721,
840
+ "step": 45,
841
+ "step_time": 8.05356791899976
842
+ },
843
+ {
844
+ "clip_ratio/high_max": 0.001953125,
845
+ "clip_ratio/high_mean": 0.001953125,
846
+ "clip_ratio/low_mean": 0.0,
847
+ "clip_ratio/low_min": 0.0,
848
+ "clip_ratio/region_mean": 0.001953125,
849
+ "entropy": 1.8525512218475342,
850
+ "epoch": 1.15,
851
+ "grad_norm": 0.9486414194107056,
852
+ "kl": 0.12013739347457886,
853
+ "learning_rate": 1.875e-06,
854
+ "loss": -0.03101479448378086,
855
+ "step": 46,
856
+ "step_time": 0.11169548499947268
857
+ },
858
+ {
859
+ "clip_ratio/high_max": 0.0048828125,
860
+ "clip_ratio/high_mean": 0.0048828125,
861
+ "clip_ratio/low_mean": 0.0,
862
+ "clip_ratio/low_min": 0.0,
863
+ "clip_ratio/region_mean": 0.0048828125,
864
+ "entropy": 2.05437970161438,
865
+ "epoch": 1.175,
866
+ "grad_norm": 0.9893139600753784,
867
+ "kl": 0.0978676825761795,
868
+ "learning_rate": 1.85e-06,
869
+ "loss": -0.03111003153026104,
870
+ "step": 47,
871
+ "step_time": 0.1112122829999862
872
+ },
873
+ {
874
+ "clip_ratio/high_max": 0.0048828125,
875
+ "clip_ratio/high_mean": 0.0048828125,
876
+ "clip_ratio/low_mean": 0.0,
877
+ "clip_ratio/low_min": 0.0,
878
+ "clip_ratio/region_mean": 0.0048828125,
879
+ "entropy": 1.9310542345046997,
880
+ "epoch": 1.2,
881
+ "grad_norm": 2.8787803649902344,
882
+ "kl": 0.10207061469554901,
883
+ "learning_rate": 1.8249999999999999e-06,
884
+ "loss": 0.06266620755195618,
885
+ "step": 48,
886
+ "step_time": 0.12508205599988287
887
+ },
888
+ {
889
+ "clip_ratio/high_max": 0.0,
890
+ "clip_ratio/high_mean": 0.0,
891
+ "clip_ratio/low_mean": 0.0,
892
+ "clip_ratio/low_min": 0.0,
893
+ "clip_ratio/region_mean": 0.0,
894
+ "completions/clipped_ratio": 1.0,
895
+ "completions/max_length": 256.0,
896
+ "completions/max_terminated_length": 0.0,
897
+ "completions/mean_length": 256.0,
898
+ "completions/mean_terminated_length": 0.0,
899
+ "completions/min_length": 256.0,
900
+ "completions/min_terminated_length": 0.0,
901
+ "entropy": 1.9703929424285889,
902
+ "epoch": 1.225,
903
+ "frac_reward_zero_std": 0.5,
904
+ "grad_norm": 3.0901620388031006,
905
+ "kl": 0.0971868485212326,
906
+ "learning_rate": 1.8e-06,
907
+ "loss": 0.06205549091100693,
908
+ "num_tokens": 130188.0,
909
+ "reward": 0.7701587677001953,
910
+ "reward_std": 0.17226041853427887,
911
+ "rewards/AnswererRewardFunction/mean": 0.7701587677001953,
912
+ "rewards/AnswererRewardFunction/std": 0.17226043343544006,
913
+ "step": 49,
914
+ "step_time": 7.936510957000792
915
+ },
916
+ {
917
+ "clip_ratio/high_max": 0.0048828125,
918
+ "clip_ratio/high_mean": 0.0048828125,
919
+ "clip_ratio/low_mean": 0.0,
920
+ "clip_ratio/low_min": 0.0,
921
+ "clip_ratio/region_mean": 0.0048828125,
922
+ "entropy": 1.9568400382995605,
923
+ "epoch": 1.25,
924
+ "grad_norm": 1.540932297706604,
925
+ "kl": 0.11889415979385376,
926
+ "learning_rate": 1.7750000000000002e-06,
927
+ "loss": -0.09337170422077179,
928
+ "step": 50,
929
+ "step_time": 0.10965072100043471
930
+ },
931
+ {
932
+ "clip_ratio/high_max": 0.0068359375,
933
+ "clip_ratio/high_mean": 0.0068359375,
934
+ "clip_ratio/low_mean": 0.0009765625,
935
+ "clip_ratio/low_min": 0.0009765625,
936
+ "clip_ratio/region_mean": 0.0078125,
937
+ "entropy": 1.9066113233566284,
938
+ "epoch": 1.275,
939
+ "grad_norm": 3.2399439811706543,
940
+ "kl": 0.10549439489841461,
941
+ "learning_rate": 1.7500000000000002e-06,
942
+ "loss": 0.03147980943322182,
943
+ "step": 51,
944
+ "step_time": 0.10916036200069357
945
+ },
946
+ {
947
+ "clip_ratio/high_max": 0.0,
948
+ "clip_ratio/high_mean": 0.0,
949
+ "clip_ratio/low_mean": 0.0,
950
+ "clip_ratio/low_min": 0.0,
951
+ "clip_ratio/region_mean": 0.0,
952
+ "entropy": 1.7676732540130615,
953
+ "epoch": 1.3,
954
+ "grad_norm": 0.038625847548246384,
955
+ "kl": 0.1173200011253357,
956
+ "learning_rate": 1.725e-06,
957
+ "loss": 0.00029329999233596027,
958
+ "step": 52,
959
+ "step_time": 0.10919338299936499
960
+ },
961
+ {
962
+ "clip_ratio/high_max": 0.0,
963
+ "clip_ratio/high_mean": 0.0,
964
+ "clip_ratio/low_mean": 0.0,
965
+ "clip_ratio/low_min": 0.0,
966
+ "clip_ratio/region_mean": 0.0,
967
+ "completions/clipped_ratio": 1.0,
968
+ "completions/max_length": 256.0,
969
+ "completions/max_terminated_length": 0.0,
970
+ "completions/mean_length": 256.0,
971
+ "completions/mean_terminated_length": 0.0,
972
+ "completions/min_length": 256.0,
973
+ "completions/min_terminated_length": 0.0,
974
+ "entropy": 1.6084643602371216,
975
+ "epoch": 1.325,
976
+ "frac_reward_zero_std": 1.0,
977
+ "grad_norm": 0.04859758913516998,
978
+ "kl": 0.11696670204401016,
979
+ "learning_rate": 1.7e-06,
980
+ "loss": 0.00029241672018542886,
981
+ "num_tokens": 140400.0,
982
+ "reward": 0.8848026990890503,
983
+ "reward_std": 0.08498957008123398,
984
+ "rewards/AnswererRewardFunction/mean": 0.8848026990890503,
985
+ "rewards/AnswererRewardFunction/std": 0.08498956263065338,
986
+ "step": 53,
987
+ "step_time": 7.985873733000517
988
+ },
989
+ {
990
+ "clip_ratio/high_max": 0.0,
991
+ "clip_ratio/high_mean": 0.0,
992
+ "clip_ratio/low_mean": 0.0,
993
+ "clip_ratio/low_min": 0.0,
994
+ "clip_ratio/region_mean": 0.0,
995
+ "entropy": 1.9126839637756348,
996
+ "epoch": 1.35,
997
+ "grad_norm": 0.05618106201291084,
998
+ "kl": 0.12067677080631256,
999
+ "learning_rate": 1.675e-06,
1000
+ "loss": 0.0003016919654328376,
1001
+ "step": 54,
1002
+ "step_time": 0.11533255099948292
1003
+ },
1004
+ {
1005
+ "clip_ratio/high_max": 0.0,
1006
+ "clip_ratio/high_mean": 0.0,
1007
+ "clip_ratio/low_mean": 0.0,
1008
+ "clip_ratio/low_min": 0.0,
1009
+ "clip_ratio/region_mean": 0.0,
1010
+ "entropy": 1.9060513973236084,
1011
+ "epoch": 1.375,
1012
+ "grad_norm": 0.03235335648059845,
1013
+ "kl": 0.10564382374286652,
1014
+ "learning_rate": 1.65e-06,
1015
+ "loss": 0.0002641095488797873,
1016
+ "step": 55,
1017
+ "step_time": 0.11275608800042392
1018
+ },
1019
+ {
1020
+ "clip_ratio/high_max": 0.0,
1021
+ "clip_ratio/high_mean": 0.0,
1022
+ "clip_ratio/low_mean": 0.0,
1023
+ "clip_ratio/low_min": 0.0,
1024
+ "clip_ratio/region_mean": 0.0,
1025
+ "entropy": 1.9660732746124268,
1026
+ "epoch": 1.4,
1027
+ "grad_norm": 0.03671698272228241,
1028
+ "kl": 0.10332442820072174,
1029
+ "learning_rate": 1.625e-06,
1030
+ "loss": 0.00025831107632257044,
1031
+ "step": 56,
1032
+ "step_time": 0.11206929399941146
1033
+ },
1034
+ {
1035
+ "clip_ratio/high_max": 0.0,
1036
+ "clip_ratio/high_mean": 0.0,
1037
+ "clip_ratio/low_mean": 0.0,
1038
+ "clip_ratio/low_min": 0.0,
1039
+ "clip_ratio/region_mean": 0.0,
1040
+ "completions/clipped_ratio": 1.0,
1041
+ "completions/max_length": 256.0,
1042
+ "completions/max_terminated_length": 0.0,
1043
+ "completions/mean_length": 256.0,
1044
+ "completions/mean_terminated_length": 0.0,
1045
+ "completions/min_length": 256.0,
1046
+ "completions/min_terminated_length": 0.0,
1047
+ "entropy": 2.006600856781006,
1048
+ "epoch": 1.425,
1049
+ "frac_reward_zero_std": 1.0,
1050
+ "grad_norm": 0.04220745339989662,
1051
+ "kl": 0.14282318949699402,
1052
+ "learning_rate": 1.6e-06,
1053
+ "loss": 0.00035705792834050953,
1054
+ "num_tokens": 150720.0,
1055
+ "reward": 0.9323133230209351,
1056
+ "reward_std": 0.0,
1057
+ "rewards/AnswererRewardFunction/mean": 0.9323133230209351,
1058
+ "rewards/AnswererRewardFunction/std": 0.0,
1059
+ "step": 57,
1060
+ "step_time": 7.929392387000007
1061
+ },
1062
+ {
1063
+ "clip_ratio/high_max": 0.0,
1064
+ "clip_ratio/high_mean": 0.0,
1065
+ "clip_ratio/low_mean": 0.0,
1066
+ "clip_ratio/low_min": 0.0,
1067
+ "clip_ratio/region_mean": 0.0,
1068
+ "entropy": 1.845897912979126,
1069
+ "epoch": 1.45,
1070
+ "grad_norm": 0.03133726865053177,
1071
+ "kl": 0.09734172374010086,
1072
+ "learning_rate": 1.5750000000000002e-06,
1073
+ "loss": 0.00024335432681255043,
1074
+ "step": 58,
1075
+ "step_time": 0.11144289299954835
1076
+ },
1077
+ {
1078
+ "clip_ratio/high_max": 0.0,
1079
+ "clip_ratio/high_mean": 0.0,
1080
+ "clip_ratio/low_mean": 0.0,
1081
+ "clip_ratio/low_min": 0.0,
1082
+ "clip_ratio/region_mean": 0.0,
1083
+ "entropy": 1.9323326349258423,
1084
+ "epoch": 1.475,
1085
+ "grad_norm": 0.031240344047546387,
1086
+ "kl": 0.1022108644247055,
1087
+ "learning_rate": 1.5500000000000002e-06,
1088
+ "loss": 0.0002555271494202316,
1089
+ "step": 59,
1090
+ "step_time": 0.11132388100031676
1091
+ },
1092
+ {
1093
+ "clip_ratio/high_max": 0.0,
1094
+ "clip_ratio/high_mean": 0.0,
1095
+ "clip_ratio/low_mean": 0.0,
1096
+ "clip_ratio/low_min": 0.0,
1097
+ "clip_ratio/region_mean": 0.0,
1098
+ "entropy": 1.9500852823257446,
1099
+ "epoch": 1.5,
1100
+ "grad_norm": 0.04029398411512375,
1101
+ "kl": 0.12107711285352707,
1102
+ "learning_rate": 1.525e-06,
1103
+ "loss": 0.0003026928170584142,
1104
+ "step": 60,
1105
+ "step_time": 0.11092958100016403
1106
+ }
1107
+ ],
1108
+ "logging_steps": 1,
1109
+ "max_steps": 120,
1110
+ "num_input_tokens_seen": 150720,
1111
+ "num_train_epochs": 3,
1112
+ "save_steps": 30,
1113
+ "stateful_callbacks": {
1114
+ "TrainerControl": {
1115
+ "args": {
1116
+ "should_epoch_stop": false,
1117
+ "should_evaluate": false,
1118
+ "should_log": false,
1119
+ "should_save": true,
1120
+ "should_training_stop": false
1121
+ },
1122
+ "attributes": {}
1123
+ }
1124
+ },
1125
+ "total_flos": 0.0,
1126
+ "train_batch_size": 4,
1127
+ "trial_name": null,
1128
+ "trial_params": null
1129
+ }
self_play_hf_l40s_full/round_005/answerer_train/checkpoint-60/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:176482d0784a3473c5f8dbadcd601422b0dadd5afadeeee68d055f1eaaf7d577
3
+ size 7249