Hugging Face's logo

Haitao999
/
Llama-3.1-8B-Instruct-GRPO-numia_prompt_dpo1

Text Generation
Transformers
Safetensors
llama
Generated from Trainer
open-r1
trl
grpo
conversational
text-generation-inference
Model card Files Community
Haitao999 commited on
Commit
51e1c8e
·
verified ·
1 Parent(s): 9ce4d0b

Model save

Browse files
Files changed (9) hide show
  1. README.md +67 -0
  2. all_results.json +8 -0
  3. generation_config.json +12 -0
  4. model-00001-of-00004.safetensors +1 -1
  5. model-00002-of-00004.safetensors +1 -1
  6. model-00003-of-00004.safetensors +1 -1
  7. model-00004-of-00004.safetensors +1 -1
  8. train_results.json +8 -0
  9. trainer_state.json +2000 -0
README.md ADDED
@@ -0,0 +1,67 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: transformers
3
+ model_name: Llama-3.1-8B-Instruct-GRPO-numia_prompt_dpo1
4
+ tags:
5
+ - generated_from_trainer
6
+ - trl
7
+ - grpo
8
+ licence: license
9
+ ---
10
+
11
+ # Model Card for Llama-3.1-8B-Instruct-GRPO-numia_prompt_dpo1
12
+
13
+ This model is a fine-tuned version of [None](https://huggingface.co/None).
14
+ It has been trained using [TRL](https://github.com/huggingface/trl).
15
+
16
+ ## Quick start
17
+
18
+ ```python
19
+ from transformers import pipeline
20
+
21
+ 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?"
22
+ generator = pipeline("text-generation", model="Haitao999/Llama-3.1-8B-Instruct-GRPO-numia_prompt_dpo1", device="cuda")
23
+ output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
24
+ print(output["generated_text"])
25
+ ```
26
+
27
+ ## Training procedure
28
+
29
+ [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/tjucsailab/huggingface/runs/4m3bt8vp)
30
+
31
+
32
+ This model was trained with GRPO, a method introduced in [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://huggingface.co/papers/2402.03300).
33
+
34
+ ### Framework versions
35
+
36
+ - TRL: 0.14.0
37
+ - Transformers: 4.48.3
38
+ - Pytorch: 2.5.1
39
+ - Datasets: 3.2.0
40
+ - Tokenizers: 0.21.1
41
+
42
+ ## Citations
43
+
44
+ Cite GRPO as:
45
+
46
+ ```bibtex
47
+ @article{zhihong2024deepseekmath,
48
+ title = {{DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models}},
49
+ author = {Zhihong Shao and Peiyi Wang and Qihao Zhu and Runxin Xu and Junxiao Song and Mingchuan Zhang and Y. K. Li and Y. Wu and Daya Guo},
50
+ year = 2024,
51
+ eprint = {arXiv:2402.03300},
52
+ }
53
+
54
+ ```
55
+
56
+ Cite TRL as:
57
+
58
+ ```bibtex
59
+ @misc{vonwerra2022trl,
60
+ title = {{TRL: Transformer Reinforcement Learning}},
61
+ 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 Gallouédec},
62
+ year = 2020,
63
+ journal = {GitHub repository},
64
+ publisher = {GitHub},
65
+ howpublished = {\url{https://github.com/huggingface/trl}}
66
+ }
67
+ ```
all_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "total_flos": 0.0,
3
+ "train_loss": 1.002405131310157e-08,
4
+ "train_runtime": 79864.5906,
5
+ "train_samples": 20000,
6
+ "train_samples_per_second": 0.25,
7
+ "train_steps_per_second": 0.002
8
+ }
generation_config.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 128000,
3
+ "do_sample": true,
4
+ "eos_token_id": [
5
+ 128001,
6
+ 128008,
7
+ 128009
8
+ ],
9
+ "temperature": 0.6,
10
+ "top_p": 0.9,
11
+ "transformers_version": "4.48.3"
12
+ }
model-00001-of-00004.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:5f436c0a93ab06c9ff3b3779c7f7d424f649b8fcadd480944f7360e4516b4dc2
3
  size 4976698672
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fe7e94f89ed7609bdff13f974759bff3b08f054427ad450046845bbdce5c128d
3
  size 4976698672
model-00002-of-00004.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1d0e253e4c093bcbd35b10c2f44982fab3cc678882450d76c30f645752569819
3
  size 4999802720
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:eec623267cbebbeaacf49205bb6fc1f75a3b51ad70ac1e2252fdb6c009444309
3
  size 4999802720
model-00003-of-00004.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:be23af942065b492b78114481199bf167965df6b8f9c5022632485e628a162bb
3
  size 4915916176
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c832ec7ca1b878d865cdc1c208de3cbcd7129b22c32a1f8f6100b93e1745b640
3
  size 4915916176
model-00004-of-00004.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a7916f363aeb5b36f2ad8e3f0c8603eba7da79b870bfc6374162cbb4fee11a88
3
  size 1168138808
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ea318c3130004ad59a2167dd8dc068d34e37ef1f952243835483c1993568d81
3
  size 1168138808
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "total_flos": 0.0,
3
+ "train_loss": 1.002405131310157e-08,
4
+ "train_runtime": 79864.5906,
5
+ "train_samples": 20000,
6
+ "train_samples_per_second": 0.25,
7
+ "train_steps_per_second": 0.002
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,2000 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.9965010496850945,
5
+ "eval_steps": 100,
6
+ "global_step": 178,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "completion_length": 621.5997104644775,
13
+ "epoch": 0.005598320503848845,
14
+ "grad_norm": 0.01685222238302231,
15
+ "learning_rate": 3e-07,
16
+ "loss": 0.0,
17
+ "reward": 0.21130952710518613,
18
+ "reward_std": 0.1508616954088211,
19
+ "rewards/accuracy_reward": 0.21130952710518613,
20
+ "step": 1
21
+ },
22
+ {
23
+ "completion_length": 593.0461368560791,
24
+ "epoch": 0.01119664100769769,
25
+ "grad_norm": 0.017173435539007187,
26
+ "learning_rate": 3e-07,
27
+ "loss": 0.0,
28
+ "reward": 0.21502976398915052,
29
+ "reward_std": 0.1765604354441166,
30
+ "rewards/accuracy_reward": 0.21502976398915052,
31
+ "step": 2
32
+ },
33
+ {
34
+ "completion_length": 625.5074501037598,
35
+ "epoch": 0.016794961511546535,
36
+ "grad_norm": 0.013760975562036037,
37
+ "learning_rate": 3e-07,
38
+ "loss": 0.0,
39
+ "reward": 0.15401786100119352,
40
+ "reward_std": 0.13073960039764643,
41
+ "rewards/accuracy_reward": 0.15401786100119352,
42
+ "step": 3
43
+ },
44
+ {
45
+ "completion_length": 647.694206237793,
46
+ "epoch": 0.02239328201539538,
47
+ "grad_norm": 0.015759378671646118,
48
+ "learning_rate": 3e-07,
49
+ "loss": 0.0,
50
+ "reward": 0.1808035772992298,
51
+ "reward_std": 0.1654246775433421,
52
+ "rewards/accuracy_reward": 0.1808035772992298,
53
+ "step": 4
54
+ },
55
+ {
56
+ "completion_length": 629.4486751556396,
57
+ "epoch": 0.02799160251924423,
58
+ "grad_norm": 0.014322967268526554,
59
+ "learning_rate": 3e-07,
60
+ "loss": 0.0,
61
+ "reward": 0.23883929010480642,
62
+ "reward_std": 0.15154671855270863,
63
+ "rewards/accuracy_reward": 0.23883929010480642,
64
+ "step": 5
65
+ },
66
+ {
67
+ "completion_length": 604.1823024749756,
68
+ "epoch": 0.03358992302309307,
69
+ "grad_norm": 0.014635809697210789,
70
+ "learning_rate": 3e-07,
71
+ "loss": 0.0,
72
+ "reward": 0.2194940543267876,
73
+ "reward_std": 0.1607616674154997,
74
+ "rewards/accuracy_reward": 0.2194940543267876,
75
+ "step": 6
76
+ },
77
+ {
78
+ "completion_length": 659.3504657745361,
79
+ "epoch": 0.03918824352694192,
80
+ "grad_norm": 0.015097221359610558,
81
+ "learning_rate": 3e-07,
82
+ "loss": 0.0,
83
+ "reward": 0.16666666860692203,
84
+ "reward_std": 0.13727260008454323,
85
+ "rewards/accuracy_reward": 0.16666666860692203,
86
+ "step": 7
87
+ },
88
+ {
89
+ "completion_length": 587.1049137115479,
90
+ "epoch": 0.04478656403079076,
91
+ "grad_norm": 0.01467145700007677,
92
+ "learning_rate": 3e-07,
93
+ "loss": 0.0,
94
+ "reward": 0.16815476398915052,
95
+ "reward_std": 0.15577294118702412,
96
+ "rewards/accuracy_reward": 0.16815476398915052,
97
+ "step": 8
98
+ },
99
+ {
100
+ "completion_length": 675.1614761352539,
101
+ "epoch": 0.05038488453463961,
102
+ "grad_norm": 0.015891388058662415,
103
+ "learning_rate": 3e-07,
104
+ "loss": 0.0,
105
+ "reward": 0.1755952414823696,
106
+ "reward_std": 0.16377168311737478,
107
+ "rewards/accuracy_reward": 0.1755952414823696,
108
+ "step": 9
109
+ },
110
+ {
111
+ "completion_length": 651.4523887634277,
112
+ "epoch": 0.05598320503848846,
113
+ "grad_norm": 0.014163345098495483,
114
+ "learning_rate": 3e-07,
115
+ "loss": 0.0,
116
+ "reward": 0.18080357555299997,
117
+ "reward_std": 0.13789755944162607,
118
+ "rewards/accuracy_reward": 0.18080357555299997,
119
+ "step": 10
120
+ },
121
+ {
122
+ "completion_length": 652.5982303619385,
123
+ "epoch": 0.0615815255423373,
124
+ "grad_norm": 0.013874271884560585,
125
+ "learning_rate": 3e-07,
126
+ "loss": 0.0,
127
+ "reward": 0.17187500558793545,
128
+ "reward_std": 0.1334617892280221,
129
+ "rewards/accuracy_reward": 0.17187500558793545,
130
+ "step": 11
131
+ },
132
+ {
133
+ "completion_length": 629.9330501556396,
134
+ "epoch": 0.06717984604618614,
135
+ "grad_norm": 0.015252678655087948,
136
+ "learning_rate": 3e-07,
137
+ "loss": 0.0,
138
+ "reward": 0.15997024276293814,
139
+ "reward_std": 0.13702514953911304,
140
+ "rewards/accuracy_reward": 0.15997024276293814,
141
+ "step": 12
142
+ },
143
+ {
144
+ "completion_length": 635.7953948974609,
145
+ "epoch": 0.072778166550035,
146
+ "grad_norm": 0.014308245852589607,
147
+ "learning_rate": 3e-07,
148
+ "loss": 0.0,
149
+ "reward": 0.1398809556849301,
150
+ "reward_std": 0.13629459426738322,
151
+ "rewards/accuracy_reward": 0.1398809556849301,
152
+ "step": 13
153
+ },
154
+ {
155
+ "completion_length": 655.9814014434814,
156
+ "epoch": 0.07837648705388384,
157
+ "grad_norm": 0.014390479773283005,
158
+ "learning_rate": 3e-07,
159
+ "loss": 0.0,
160
+ "reward": 0.17485119425691664,
161
+ "reward_std": 0.14875314245000482,
162
+ "rewards/accuracy_reward": 0.17485119425691664,
163
+ "step": 14
164
+ },
165
+ {
166
+ "completion_length": 656.0022430419922,
167
+ "epoch": 0.08397480755773268,
168
+ "grad_norm": 0.017086399719119072,
169
+ "learning_rate": 3e-07,
170
+ "loss": 0.0,
171
+ "reward": 0.17633928690338507,
172
+ "reward_std": 0.13736709533259273,
173
+ "rewards/accuracy_reward": 0.17633928690338507,
174
+ "step": 15
175
+ },
176
+ {
177
+ "completion_length": 578.8586463928223,
178
+ "epoch": 0.08957312806158152,
179
+ "grad_norm": 0.015446292236447334,
180
+ "learning_rate": 3e-07,
181
+ "loss": 0.0,
182
+ "reward": 0.21205357555299997,
183
+ "reward_std": 0.15961862821131945,
184
+ "rewards/accuracy_reward": 0.21205357555299997,
185
+ "step": 16
186
+ },
187
+ {
188
+ "completion_length": 624.680814743042,
189
+ "epoch": 0.09517144856543037,
190
+ "grad_norm": 0.016500283032655716,
191
+ "learning_rate": 3e-07,
192
+ "loss": 0.0,
193
+ "reward": 0.18601190892513841,
194
+ "reward_std": 0.14123531756922603,
195
+ "rewards/accuracy_reward": 0.18601190892513841,
196
+ "step": 17
197
+ },
198
+ {
199
+ "completion_length": 661.7500152587891,
200
+ "epoch": 0.10076976906927922,
201
+ "grad_norm": 0.014794868417084217,
202
+ "learning_rate": 3e-07,
203
+ "loss": 0.0,
204
+ "reward": 0.18229167046956718,
205
+ "reward_std": 0.15161667275242507,
206
+ "rewards/accuracy_reward": 0.18229167046956718,
207
+ "step": 18
208
+ },
209
+ {
210
+ "completion_length": 632.3244209289551,
211
+ "epoch": 0.10636808957312806,
212
+ "grad_norm": 0.01493909489363432,
213
+ "learning_rate": 3e-07,
214
+ "loss": 0.0,
215
+ "reward": 0.1636904807528481,
216
+ "reward_std": 0.1566695892252028,
217
+ "rewards/accuracy_reward": 0.1636904807528481,
218
+ "step": 19
219
+ },
220
+ {
221
+ "completion_length": 627.7477836608887,
222
+ "epoch": 0.11196641007697691,
223
+ "grad_norm": 0.01570167765021324,
224
+ "learning_rate": 3e-07,
225
+ "loss": 0.0,
226
+ "reward": 0.2046130986418575,
227
+ "reward_std": 0.14942223858088255,
228
+ "rewards/accuracy_reward": 0.2046130986418575,
229
+ "step": 20
230
+ },
231
+ {
232
+ "completion_length": 625.4613227844238,
233
+ "epoch": 0.11756473058082575,
234
+ "grad_norm": 0.016287673264741898,
235
+ "learning_rate": 3e-07,
236
+ "loss": 0.0,
237
+ "reward": 0.22991071827709675,
238
+ "reward_std": 0.1482053459621966,
239
+ "rewards/accuracy_reward": 0.22991071827709675,
240
+ "step": 21
241
+ },
242
+ {
243
+ "completion_length": 695.4895973205566,
244
+ "epoch": 0.1231630510846746,
245
+ "grad_norm": 0.013338581658899784,
246
+ "learning_rate": 3e-07,
247
+ "loss": 0.0,
248
+ "reward": 0.129464287601877,
249
+ "reward_std": 0.13608421152457595,
250
+ "rewards/accuracy_reward": 0.129464287601877,
251
+ "step": 22
252
+ },
253
+ {
254
+ "completion_length": 666.1235198974609,
255
+ "epoch": 0.12876137158852344,
256
+ "grad_norm": 0.015841346234083176,
257
+ "learning_rate": 3e-07,
258
+ "loss": 0.0,
259
+ "reward": 0.20386905269697309,
260
+ "reward_std": 0.14255746873095632,
261
+ "rewards/accuracy_reward": 0.20386905269697309,
262
+ "step": 23
263
+ },
264
+ {
265
+ "completion_length": 649.9732265472412,
266
+ "epoch": 0.13435969209237228,
267
+ "grad_norm": 0.016053833067417145,
268
+ "learning_rate": 3e-07,
269
+ "loss": 0.0,
270
+ "reward": 0.17038690764456987,
271
+ "reward_std": 0.12253910815343261,
272
+ "rewards/accuracy_reward": 0.17038690764456987,
273
+ "step": 24
274
+ },
275
+ {
276
+ "completion_length": 613.6108741760254,
277
+ "epoch": 0.13995801259622112,
278
+ "grad_norm": 0.020091773942112923,
279
+ "learning_rate": 3e-07,
280
+ "loss": 0.0,
281
+ "reward": 0.22470238478854299,
282
+ "reward_std": 0.13420080952346325,
283
+ "rewards/accuracy_reward": 0.22470238478854299,
284
+ "step": 25
285
+ },
286
+ {
287
+ "completion_length": 629.3846836090088,
288
+ "epoch": 0.14555633310007,
289
+ "grad_norm": 0.016363132745027542,
290
+ "learning_rate": 3e-07,
291
+ "loss": 0.0,
292
+ "reward": 0.19642857508733869,
293
+ "reward_std": 0.14178543630987406,
294
+ "rewards/accuracy_reward": 0.19642857508733869,
295
+ "step": 26
296
+ },
297
+ {
298
+ "completion_length": 641.9903411865234,
299
+ "epoch": 0.15115465360391883,
300
+ "grad_norm": 0.016413498669862747,
301
+ "learning_rate": 3e-07,
302
+ "loss": 0.0,
303
+ "reward": 0.19642857648432255,
304
+ "reward_std": 0.16177992545999587,
305
+ "rewards/accuracy_reward": 0.19642857648432255,
306
+ "step": 27
307
+ },
308
+ {
309
+ "completion_length": 647.2522392272949,
310
+ "epoch": 0.15675297410776767,
311
+ "grad_norm": 0.015543129295110703,
312
+ "learning_rate": 3e-07,
313
+ "loss": 0.0,
314
+ "reward": 0.19345238601090387,
315
+ "reward_std": 0.13141631055623293,
316
+ "rewards/accuracy_reward": 0.19345238601090387,
317
+ "step": 28
318
+ },
319
+ {
320
+ "completion_length": 635.9523849487305,
321
+ "epoch": 0.16235129461161651,
322
+ "grad_norm": 0.02027256041765213,
323
+ "learning_rate": 3e-07,
324
+ "loss": 0.0,
325
+ "reward": 0.22023809660458937,
326
+ "reward_std": 0.12436413252726197,
327
+ "rewards/accuracy_reward": 0.22023809660458937,
328
+ "step": 29
329
+ },
330
+ {
331
+ "completion_length": 658.7291736602783,
332
+ "epoch": 0.16794961511546536,
333
+ "grad_norm": 0.019577788189053535,
334
+ "learning_rate": 3e-07,
335
+ "loss": 0.0,
336
+ "reward": 0.19642857584403828,
337
+ "reward_std": 0.161903060041368,
338
+ "rewards/accuracy_reward": 0.19642857584403828,
339
+ "step": 30
340
+ },
341
+ {
342
+ "completion_length": 666.1145973205566,
343
+ "epoch": 0.1735479356193142,
344
+ "grad_norm": 0.016409732401371002,
345
+ "learning_rate": 3e-07,
346
+ "loss": 0.0,
347
+ "reward": 0.19494048005435616,
348
+ "reward_std": 0.12808324676007032,
349
+ "rewards/accuracy_reward": 0.19494048005435616,
350
+ "step": 31
351
+ },
352
+ {
353
+ "completion_length": 662.9114723205566,
354
+ "epoch": 0.17914625612316304,
355
+ "grad_norm": 0.02575683780014515,
356
+ "learning_rate": 3e-07,
357
+ "loss": 0.0,
358
+ "reward": 0.1733630993985571,
359
+ "reward_std": 0.14043171424418688,
360
+ "rewards/accuracy_reward": 0.1733630993985571,
361
+ "step": 32
362
+ },
363
+ {
364
+ "completion_length": 672.6763496398926,
365
+ "epoch": 0.1847445766270119,
366
+ "grad_norm": 0.02134682983160019,
367
+ "learning_rate": 3e-07,
368
+ "loss": 0.0,
369
+ "reward": 0.1979166700039059,
370
+ "reward_std": 0.1591111128218472,
371
+ "rewards/accuracy_reward": 0.1979166700039059,
372
+ "step": 33
373
+ },
374
+ {
375
+ "completion_length": 640.2604217529297,
376
+ "epoch": 0.19034289713086075,
377
+ "grad_norm": 0.023335738107562065,
378
+ "learning_rate": 3e-07,
379
+ "loss": 0.0,
380
+ "reward": 0.22172619309276342,
381
+ "reward_std": 0.16011474281549454,
382
+ "rewards/accuracy_reward": 0.22172619309276342,
383
+ "step": 34
384
+ },
385
+ {
386
+ "completion_length": 679.5119152069092,
387
+ "epoch": 0.1959412176347096,
388
+ "grad_norm": 0.026544196531176567,
389
+ "learning_rate": 3e-07,
390
+ "loss": 0.0,
391
+ "reward": 0.21651785960420966,
392
+ "reward_std": 0.1384496740065515,
393
+ "rewards/accuracy_reward": 0.21651785960420966,
394
+ "step": 35
395
+ },
396
+ {
397
+ "completion_length": 644.0104331970215,
398
+ "epoch": 0.20153953813855843,
399
+ "grad_norm": 0.02789580263197422,
400
+ "learning_rate": 3e-07,
401
+ "loss": 0.0,
402
+ "reward": 0.23065476911142468,
403
+ "reward_std": 0.14074094174429774,
404
+ "rewards/accuracy_reward": 0.23065476911142468,
405
+ "step": 36
406
+ },
407
+ {
408
+ "completion_length": 686.8742713928223,
409
+ "epoch": 0.20713785864240727,
410
+ "grad_norm": 0.02931089699268341,
411
+ "learning_rate": 3e-07,
412
+ "loss": 0.0,
413
+ "reward": 0.2165178587893024,
414
+ "reward_std": 0.1497473274357617,
415
+ "rewards/accuracy_reward": 0.2165178587893024,
416
+ "step": 37
417
+ },
418
+ {
419
+ "completion_length": 707.8519477844238,
420
+ "epoch": 0.21273617914625612,
421
+ "grad_norm": 0.02348598837852478,
422
+ "learning_rate": 3e-07,
423
+ "loss": 0.0,
424
+ "reward": 0.17857143143191934,
425
+ "reward_std": 0.12875390285626054,
426
+ "rewards/accuracy_reward": 0.17857143143191934,
427
+ "step": 38
428
+ },
429
+ {
430
+ "completion_length": 739.4196586608887,
431
+ "epoch": 0.21833449965010496,
432
+ "grad_norm": 0.025258367881178856,
433
+ "learning_rate": 3e-07,
434
+ "loss": 0.0,
435
+ "reward": 0.16294643242144957,
436
+ "reward_std": 0.10602654609829187,
437
+ "rewards/accuracy_reward": 0.16294643242144957,
438
+ "step": 39
439
+ },
440
+ {
441
+ "completion_length": 781.2157859802246,
442
+ "epoch": 0.22393282015395383,
443
+ "grad_norm": 0.02543455921113491,
444
+ "learning_rate": 3e-07,
445
+ "loss": 0.0,
446
+ "reward": 0.17038690770277753,
447
+ "reward_std": 0.1598060461692512,
448
+ "rewards/accuracy_reward": 0.17038690770277753,
449
+ "step": 40
450
+ },
451
+ {
452
+ "completion_length": 806.7745704650879,
453
+ "epoch": 0.22953114065780267,
454
+ "grad_norm": 0.02464163675904274,
455
+ "learning_rate": 3e-07,
456
+ "loss": 0.0,
457
+ "reward": 0.1443452417734079,
458
+ "reward_std": 0.13860655017197132,
459
+ "rewards/accuracy_reward": 0.1443452417734079,
460
+ "step": 41
461
+ },
462
+ {
463
+ "completion_length": 797.2678718566895,
464
+ "epoch": 0.2351294611616515,
465
+ "grad_norm": 0.022209392860531807,
466
+ "learning_rate": 3e-07,
467
+ "loss": 0.0,
468
+ "reward": 0.1465773843228817,
469
+ "reward_std": 0.1135863265953958,
470
+ "rewards/accuracy_reward": 0.1465773843228817,
471
+ "step": 42
472
+ },
473
+ {
474
+ "completion_length": 800.4032897949219,
475
+ "epoch": 0.24072778166550035,
476
+ "grad_norm": 0.027684131637215614,
477
+ "learning_rate": 3e-07,
478
+ "loss": 0.0,
479
+ "reward": 0.1502976210322231,
480
+ "reward_std": 0.14116185018792748,
481
+ "rewards/accuracy_reward": 0.1502976210322231,
482
+ "step": 43
483
+ },
484
+ {
485
+ "completion_length": 802.7254638671875,
486
+ "epoch": 0.2463261021693492,
487
+ "grad_norm": 0.02655387856066227,
488
+ "learning_rate": 3e-07,
489
+ "loss": 0.0,
490
+ "reward": 0.16294643009314314,
491
+ "reward_std": 0.11001868033781648,
492
+ "rewards/accuracy_reward": 0.16294643009314314,
493
+ "step": 44
494
+ },
495
+ {
496
+ "completion_length": 825.2031326293945,
497
+ "epoch": 0.25192442267319803,
498
+ "grad_norm": 0.02530819922685623,
499
+ "learning_rate": 3e-07,
500
+ "loss": 0.0,
501
+ "reward": 0.16889881389215589,
502
+ "reward_std": 0.12393559236079454,
503
+ "rewards/accuracy_reward": 0.16889881389215589,
504
+ "step": 45
505
+ },
506
+ {
507
+ "completion_length": 809.6086502075195,
508
+ "epoch": 0.2575227431770469,
509
+ "grad_norm": 0.027012314647436142,
510
+ "learning_rate": 3e-07,
511
+ "loss": 0.0,
512
+ "reward": 0.1339285749127157,
513
+ "reward_std": 0.12476211739704013,
514
+ "rewards/accuracy_reward": 0.1339285749127157,
515
+ "step": 46
516
+ },
517
+ {
518
+ "completion_length": 822.8742713928223,
519
+ "epoch": 0.2631210636808957,
520
+ "grad_norm": 0.027949512004852295,
521
+ "learning_rate": 3e-07,
522
+ "loss": 0.0,
523
+ "reward": 0.1458333374466747,
524
+ "reward_std": 0.12774629099294543,
525
+ "rewards/accuracy_reward": 0.1458333374466747,
526
+ "step": 47
527
+ },
528
+ {
529
+ "completion_length": 830.1458473205566,
530
+ "epoch": 0.26871938418474456,
531
+ "grad_norm": 0.040896009653806686,
532
+ "learning_rate": 3e-07,
533
+ "loss": 0.0,
534
+ "reward": 0.09077381109818816,
535
+ "reward_std": 0.09958610264584422,
536
+ "rewards/accuracy_reward": 0.09077381109818816,
537
+ "step": 48
538
+ },
539
+ {
540
+ "completion_length": 858.4375228881836,
541
+ "epoch": 0.2743177046885934,
542
+ "grad_norm": 0.025837421417236328,
543
+ "learning_rate": 3e-07,
544
+ "loss": 0.0,
545
+ "reward": 0.12276786024449393,
546
+ "reward_std": 0.10703055281192064,
547
+ "rewards/accuracy_reward": 0.12276786024449393,
548
+ "step": 49
549
+ },
550
+ {
551
+ "completion_length": 837.8504638671875,
552
+ "epoch": 0.27991602519244224,
553
+ "grad_norm": 0.029402956366539,
554
+ "learning_rate": 3e-07,
555
+ "loss": 0.0,
556
+ "reward": 0.10639881080714986,
557
+ "reward_std": 0.09327335562556982,
558
+ "rewards/accuracy_reward": 0.10639881080714986,
559
+ "step": 50
560
+ },
561
+ {
562
+ "completion_length": 795.3244209289551,
563
+ "epoch": 0.28551434569629114,
564
+ "grad_norm": 0.035774268209934235,
565
+ "learning_rate": 3e-07,
566
+ "loss": 0.0,
567
+ "reward": 0.1502976214978844,
568
+ "reward_std": 0.15580066060647368,
569
+ "rewards/accuracy_reward": 0.1502976214978844,
570
+ "step": 51
571
+ },
572
+ {
573
+ "completion_length": 804.2485198974609,
574
+ "epoch": 0.29111266620014,
575
+ "grad_norm": 0.028347313404083252,
576
+ "learning_rate": 3e-07,
577
+ "loss": 0.0,
578
+ "reward": 0.13764881319366395,
579
+ "reward_std": 0.10927183693274856,
580
+ "rewards/accuracy_reward": 0.13764881319366395,
581
+ "step": 52
582
+ },
583
+ {
584
+ "completion_length": 806.589298248291,
585
+ "epoch": 0.2967109867039888,
586
+ "grad_norm": 0.04255390539765358,
587
+ "learning_rate": 3e-07,
588
+ "loss": 0.0,
589
+ "reward": 0.16815476428018883,
590
+ "reward_std": 0.11745391692966223,
591
+ "rewards/accuracy_reward": 0.16815476428018883,
592
+ "step": 53
593
+ },
594
+ {
595
+ "completion_length": 827.7753105163574,
596
+ "epoch": 0.30230930720783766,
597
+ "grad_norm": 0.028443966060876846,
598
+ "learning_rate": 3e-07,
599
+ "loss": 0.0,
600
+ "reward": 0.13690476393094286,
601
+ "reward_std": 0.09341412922367454,
602
+ "rewards/accuracy_reward": 0.13690476393094286,
603
+ "step": 54
604
+ },
605
+ {
606
+ "completion_length": 793.2686157226562,
607
+ "epoch": 0.3079076277116865,
608
+ "grad_norm": 0.06588064879179001,
609
+ "learning_rate": 3e-07,
610
+ "loss": 0.0,
611
+ "reward": 0.14434524159878492,
612
+ "reward_std": 0.14150456665083766,
613
+ "rewards/accuracy_reward": 0.14434524159878492,
614
+ "step": 55
615
+ },
616
+ {
617
+ "completion_length": 808.3698043823242,
618
+ "epoch": 0.31350594821553535,
619
+ "grad_norm": 0.07028539478778839,
620
+ "learning_rate": 3e-07,
621
+ "loss": 0.0,
622
+ "reward": 0.15476190706249326,
623
+ "reward_std": 0.08559148758649826,
624
+ "rewards/accuracy_reward": 0.15476190706249326,
625
+ "step": 56
626
+ },
627
+ {
628
+ "completion_length": 750.5506172180176,
629
+ "epoch": 0.3191042687193842,
630
+ "grad_norm": 0.20404821634292603,
631
+ "learning_rate": 3e-07,
632
+ "loss": 0.0,
633
+ "reward": 0.25000000395812094,
634
+ "reward_std": 0.15004885476082563,
635
+ "rewards/accuracy_reward": 0.25000000395812094,
636
+ "step": 57
637
+ },
638
+ {
639
+ "completion_length": 802.5141525268555,
640
+ "epoch": 0.32470258922323303,
641
+ "grad_norm": 0.30963581800460815,
642
+ "learning_rate": 3e-07,
643
+ "loss": 0.0,
644
+ "reward": 0.15848214778816327,
645
+ "reward_std": 0.13095595594495535,
646
+ "rewards/accuracy_reward": 0.15848214778816327,
647
+ "step": 58
648
+ },
649
+ {
650
+ "completion_length": 817.5855865478516,
651
+ "epoch": 0.33030090972708187,
652
+ "grad_norm": 0.4757382571697235,
653
+ "learning_rate": 3e-07,
654
+ "loss": 0.0,
655
+ "reward": 0.08258928690338507,
656
+ "reward_std": 0.09397146827541292,
657
+ "rewards/accuracy_reward": 0.08258928690338507,
658
+ "step": 59
659
+ },
660
+ {
661
+ "completion_length": 780.9456977844238,
662
+ "epoch": 0.3358992302309307,
663
+ "grad_norm": 0.9979881644248962,
664
+ "learning_rate": 3e-07,
665
+ "loss": 0.0,
666
+ "reward": 0.14136905007762834,
667
+ "reward_std": 0.10418214229866862,
668
+ "rewards/accuracy_reward": 0.14136905007762834,
669
+ "step": 60
670
+ },
671
+ {
672
+ "completion_length": 806.7462959289551,
673
+ "epoch": 0.34149755073477955,
674
+ "grad_norm": 2.2850308418273926,
675
+ "learning_rate": 3e-07,
676
+ "loss": 0.0,
677
+ "reward": 0.16741071856813505,
678
+ "reward_std": 0.12904326571151614,
679
+ "rewards/accuracy_reward": 0.16741071856813505,
680
+ "step": 61
681
+ },
682
+ {
683
+ "completion_length": 796.9077606201172,
684
+ "epoch": 0.3470958712386284,
685
+ "grad_norm": 3.0867838859558105,
686
+ "learning_rate": 3e-07,
687
+ "loss": 0.0,
688
+ "reward": 0.1555059568490833,
689
+ "reward_std": 0.1193866366520524,
690
+ "rewards/accuracy_reward": 0.1555059568490833,
691
+ "step": 62
692
+ },
693
+ {
694
+ "completion_length": 815.139892578125,
695
+ "epoch": 0.35269419174247724,
696
+ "grad_norm": 9.1638822555542,
697
+ "learning_rate": 3e-07,
698
+ "loss": 0.0,
699
+ "reward": 0.12425595556851476,
700
+ "reward_std": 0.11769626522436738,
701
+ "rewards/accuracy_reward": 0.12425595556851476,
702
+ "step": 63
703
+ },
704
+ {
705
+ "completion_length": 811.334831237793,
706
+ "epoch": 0.3582925122463261,
707
+ "grad_norm": 9.835746765136719,
708
+ "learning_rate": 3e-07,
709
+ "loss": 0.0,
710
+ "reward": 0.13467262114863843,
711
+ "reward_std": 0.10124374250881374,
712
+ "rewards/accuracy_reward": 0.13467262114863843,
713
+ "step": 64
714
+ },
715
+ {
716
+ "completion_length": 816.058048248291,
717
+ "epoch": 0.363890832750175,
718
+ "grad_norm": 12.197829246520996,
719
+ "learning_rate": 3e-07,
720
+ "loss": 0.0,
721
+ "reward": 0.13467262114863843,
722
+ "reward_std": 0.11615834524855018,
723
+ "rewards/accuracy_reward": 0.13467262114863843,
724
+ "step": 65
725
+ },
726
+ {
727
+ "completion_length": 778.9769439697266,
728
+ "epoch": 0.3694891532540238,
729
+ "grad_norm": 17.09342384338379,
730
+ "learning_rate": 3e-07,
731
+ "loss": 0.0,
732
+ "reward": 0.1324404794140719,
733
+ "reward_std": 0.1296369112096727,
734
+ "rewards/accuracy_reward": 0.1324404794140719,
735
+ "step": 66
736
+ },
737
+ {
738
+ "completion_length": 791.2901954650879,
739
+ "epoch": 0.37508747375787266,
740
+ "grad_norm": 5.410295009613037,
741
+ "learning_rate": 3e-07,
742
+ "loss": 0.0,
743
+ "reward": 0.10342262109043077,
744
+ "reward_std": 0.10896145971491933,
745
+ "rewards/accuracy_reward": 0.10342262109043077,
746
+ "step": 67
747
+ },
748
+ {
749
+ "completion_length": 745.5297737121582,
750
+ "epoch": 0.3806857942617215,
751
+ "grad_norm": 13.46410846710205,
752
+ "learning_rate": 3e-07,
753
+ "loss": 0.0,
754
+ "reward": 0.21428571629803628,
755
+ "reward_std": 0.14342379802837968,
756
+ "rewards/accuracy_reward": 0.21428571629803628,
757
+ "step": 68
758
+ },
759
+ {
760
+ "completion_length": 776.0439109802246,
761
+ "epoch": 0.38628411476557034,
762
+ "grad_norm": 8.488734245300293,
763
+ "learning_rate": 3e-07,
764
+ "loss": 0.0,
765
+ "reward": 0.11830357467988506,
766
+ "reward_std": 0.11542791035026312,
767
+ "rewards/accuracy_reward": 0.11830357467988506,
768
+ "step": 69
769
+ },
770
+ {
771
+ "completion_length": 797.398078918457,
772
+ "epoch": 0.3918824352694192,
773
+ "grad_norm": 8.824621200561523,
774
+ "learning_rate": 3e-07,
775
+ "loss": 0.0,
776
+ "reward": 0.11383928917348385,
777
+ "reward_std": 0.09247074788436294,
778
+ "rewards/accuracy_reward": 0.11383928917348385,
779
+ "step": 70
780
+ },
781
+ {
782
+ "completion_length": 765.494800567627,
783
+ "epoch": 0.397480755773268,
784
+ "grad_norm": 7.611891746520996,
785
+ "learning_rate": 3e-07,
786
+ "loss": 0.0,
787
+ "reward": 0.15178571932483464,
788
+ "reward_std": 0.10585443489253521,
789
+ "rewards/accuracy_reward": 0.15178571932483464,
790
+ "step": 71
791
+ },
792
+ {
793
+ "completion_length": 778.1890029907227,
794
+ "epoch": 0.40307907627711687,
795
+ "grad_norm": 14.848072052001953,
796
+ "learning_rate": 3e-07,
797
+ "loss": 0.0,
798
+ "reward": 0.18452381214592606,
799
+ "reward_std": 0.16349461674690247,
800
+ "rewards/accuracy_reward": 0.18452381214592606,
801
+ "step": 72
802
+ },
803
+ {
804
+ "completion_length": 774.4456939697266,
805
+ "epoch": 0.4086773967809657,
806
+ "grad_norm": 5.173851490020752,
807
+ "learning_rate": 3e-07,
808
+ "loss": 0.0,
809
+ "reward": 0.20014881325187162,
810
+ "reward_std": 0.12936217105016112,
811
+ "rewards/accuracy_reward": 0.20014881325187162,
812
+ "step": 73
813
+ },
814
+ {
815
+ "completion_length": 841.1287307739258,
816
+ "epoch": 0.41427571728481455,
817
+ "grad_norm": 2.6559793949127197,
818
+ "learning_rate": 3e-07,
819
+ "loss": 0.0,
820
+ "reward": 0.13839285890571773,
821
+ "reward_std": 0.13339613983407617,
822
+ "rewards/accuracy_reward": 0.13839285890571773,
823
+ "step": 74
824
+ },
825
+ {
826
+ "completion_length": 818.7284393310547,
827
+ "epoch": 0.4198740377886634,
828
+ "grad_norm": 1.2601547241210938,
829
+ "learning_rate": 3e-07,
830
+ "loss": 0.0,
831
+ "reward": 0.15699405129998922,
832
+ "reward_std": 0.147474380210042,
833
+ "rewards/accuracy_reward": 0.15699405129998922,
834
+ "step": 75
835
+ },
836
+ {
837
+ "completion_length": 806.4762115478516,
838
+ "epoch": 0.42547235829251223,
839
+ "grad_norm": 0.6955339312553406,
840
+ "learning_rate": 3e-07,
841
+ "loss": 0.0,
842
+ "reward": 0.14434524072567,
843
+ "reward_std": 0.11006892891600728,
844
+ "rewards/accuracy_reward": 0.14434524072567,
845
+ "step": 76
846
+ },
847
+ {
848
+ "completion_length": 829.7209968566895,
849
+ "epoch": 0.4310706787963611,
850
+ "grad_norm": 0.3638227880001068,
851
+ "learning_rate": 3e-07,
852
+ "loss": 0.0,
853
+ "reward": 0.16294643277069554,
854
+ "reward_std": 0.12348859198391438,
855
+ "rewards/accuracy_reward": 0.16294643277069554,
856
+ "step": 77
857
+ },
858
+ {
859
+ "completion_length": 825.5610313415527,
860
+ "epoch": 0.4366689993002099,
861
+ "grad_norm": 0.26862311363220215,
862
+ "learning_rate": 3e-07,
863
+ "loss": 0.0,
864
+ "reward": 0.1272321459837258,
865
+ "reward_std": 0.10739197209477425,
866
+ "rewards/accuracy_reward": 0.1272321459837258,
867
+ "step": 78
868
+ },
869
+ {
870
+ "completion_length": 789.9360198974609,
871
+ "epoch": 0.44226731980405876,
872
+ "grad_norm": 0.2580260634422302,
873
+ "learning_rate": 3e-07,
874
+ "loss": 0.0,
875
+ "reward": 0.16964286100119352,
876
+ "reward_std": 0.12814528262242675,
877
+ "rewards/accuracy_reward": 0.16964286100119352,
878
+ "step": 79
879
+ },
880
+ {
881
+ "completion_length": 812.0885543823242,
882
+ "epoch": 0.44786564030790765,
883
+ "grad_norm": 0.21123050153255463,
884
+ "learning_rate": 3e-07,
885
+ "loss": 0.0,
886
+ "reward": 0.15252976340707392,
887
+ "reward_std": 0.11390745500102639,
888
+ "rewards/accuracy_reward": 0.15252976340707392,
889
+ "step": 80
890
+ },
891
+ {
892
+ "completion_length": 827.6890068054199,
893
+ "epoch": 0.4534639608117565,
894
+ "grad_norm": 0.21955974400043488,
895
+ "learning_rate": 3e-07,
896
+ "loss": 0.0,
897
+ "reward": 0.13392857264261693,
898
+ "reward_std": 0.11876308592036366,
899
+ "rewards/accuracy_reward": 0.13392857264261693,
900
+ "step": 81
901
+ },
902
+ {
903
+ "completion_length": 790.8415336608887,
904
+ "epoch": 0.45906228131560534,
905
+ "grad_norm": 0.15459412336349487,
906
+ "learning_rate": 3e-07,
907
+ "loss": 0.0,
908
+ "reward": 0.21205357508733869,
909
+ "reward_std": 0.1082809790968895,
910
+ "rewards/accuracy_reward": 0.21205357508733869,
911
+ "step": 82
912
+ },
913
+ {
914
+ "completion_length": 799.1346817016602,
915
+ "epoch": 0.4646606018194542,
916
+ "grad_norm": 0.23524288833141327,
917
+ "learning_rate": 3e-07,
918
+ "loss": 0.0,
919
+ "reward": 0.1569940506014973,
920
+ "reward_std": 0.11971638561226428,
921
+ "rewards/accuracy_reward": 0.1569940506014973,
922
+ "step": 83
923
+ },
924
+ {
925
+ "completion_length": 757.0647468566895,
926
+ "epoch": 0.470258922323303,
927
+ "grad_norm": 0.19839145243167877,
928
+ "learning_rate": 3e-07,
929
+ "loss": 0.0,
930
+ "reward": 0.2038690508925356,
931
+ "reward_std": 0.13258925126865506,
932
+ "rewards/accuracy_reward": 0.2038690508925356,
933
+ "step": 84
934
+ },
935
+ {
936
+ "completion_length": 808.5535926818848,
937
+ "epoch": 0.47585724282715186,
938
+ "grad_norm": 0.32531610131263733,
939
+ "learning_rate": 3e-07,
940
+ "loss": 0.0,
941
+ "reward": 0.1666666684905067,
942
+ "reward_std": 0.13159588165581226,
943
+ "rewards/accuracy_reward": 0.1666666684905067,
944
+ "step": 85
945
+ },
946
+ {
947
+ "completion_length": 813.6451110839844,
948
+ "epoch": 0.4814555633310007,
949
+ "grad_norm": 0.18594929575920105,
950
+ "learning_rate": 3e-07,
951
+ "loss": 0.0,
952
+ "reward": 0.144345240900293,
953
+ "reward_std": 0.1254395372234285,
954
+ "rewards/accuracy_reward": 0.144345240900293,
955
+ "step": 86
956
+ },
957
+ {
958
+ "completion_length": 774.9620704650879,
959
+ "epoch": 0.48705388383484954,
960
+ "grad_norm": 0.18389521539211273,
961
+ "learning_rate": 3e-07,
962
+ "loss": 0.0,
963
+ "reward": 0.19270833674818277,
964
+ "reward_std": 0.15590714011341333,
965
+ "rewards/accuracy_reward": 0.19270833674818277,
966
+ "step": 87
967
+ },
968
+ {
969
+ "completion_length": 769.0573043823242,
970
+ "epoch": 0.4926522043386984,
971
+ "grad_norm": 0.2218749225139618,
972
+ "learning_rate": 3e-07,
973
+ "loss": 0.0,
974
+ "reward": 0.2105654807528481,
975
+ "reward_std": 0.12664200831204653,
976
+ "rewards/accuracy_reward": 0.2105654807528481,
977
+ "step": 88
978
+ },
979
+ {
980
+ "completion_length": 792.6644401550293,
981
+ "epoch": 0.4982505248425472,
982
+ "grad_norm": 1.4022585153579712,
983
+ "learning_rate": 3e-07,
984
+ "loss": 0.0,
985
+ "reward": 0.15476190886693075,
986
+ "reward_std": 0.11095877178013325,
987
+ "rewards/accuracy_reward": 0.15476190886693075,
988
+ "step": 89
989
+ },
990
+ {
991
+ "completion_length": 782.799861907959,
992
+ "epoch": 0.5038488453463961,
993
+ "grad_norm": 28.66908073425293,
994
+ "learning_rate": 3e-07,
995
+ "loss": 0.0,
996
+ "reward": 0.15104167041135952,
997
+ "reward_std": 0.15992113715037704,
998
+ "rewards/accuracy_reward": 0.15104167041135952,
999
+ "step": 90
1000
+ },
1001
+ {
1002
+ "completion_length": 795.1547698974609,
1003
+ "epoch": 0.509447165850245,
1004
+ "grad_norm": 8.09515380859375,
1005
+ "learning_rate": 3e-07,
1006
+ "loss": 0.0,
1007
+ "reward": 0.1569940495537594,
1008
+ "reward_std": 0.12339623901061714,
1009
+ "rewards/accuracy_reward": 0.1569940495537594,
1010
+ "step": 91
1011
+ },
1012
+ {
1013
+ "completion_length": 823.6614723205566,
1014
+ "epoch": 0.5150454863540938,
1015
+ "grad_norm": 4.843145847320557,
1016
+ "learning_rate": 3e-07,
1017
+ "loss": 0.0,
1018
+ "reward": 0.1197916705859825,
1019
+ "reward_std": 0.11639799829572439,
1020
+ "rewards/accuracy_reward": 0.1197916705859825,
1021
+ "step": 92
1022
+ },
1023
+ {
1024
+ "completion_length": 814.1956977844238,
1025
+ "epoch": 0.5206438068579426,
1026
+ "grad_norm": 3.3891513347625732,
1027
+ "learning_rate": 3e-07,
1028
+ "loss": 0.0,
1029
+ "reward": 0.12276785861467943,
1030
+ "reward_std": 0.11077928310260177,
1031
+ "rewards/accuracy_reward": 0.12276785861467943,
1032
+ "step": 93
1033
+ },
1034
+ {
1035
+ "completion_length": 832.5439147949219,
1036
+ "epoch": 0.5262421273617914,
1037
+ "grad_norm": 3.9651668071746826,
1038
+ "learning_rate": 3e-07,
1039
+ "loss": 0.0,
1040
+ "reward": 0.12797619338380173,
1041
+ "reward_std": 0.13058934407308698,
1042
+ "rewards/accuracy_reward": 0.12797619338380173,
1043
+ "step": 94
1044
+ },
1045
+ {
1046
+ "completion_length": 845.0171318054199,
1047
+ "epoch": 0.5318404478656403,
1048
+ "grad_norm": 2.0770602226257324,
1049
+ "learning_rate": 3e-07,
1050
+ "loss": 0.0,
1051
+ "reward": 0.11235119315097108,
1052
+ "reward_std": 0.08664647908881307,
1053
+ "rewards/accuracy_reward": 0.11235119315097108,
1054
+ "step": 95
1055
+ },
1056
+ {
1057
+ "completion_length": 790.6480865478516,
1058
+ "epoch": 0.5374387683694891,
1059
+ "grad_norm": 1.959823727607727,
1060
+ "learning_rate": 3e-07,
1061
+ "loss": 0.0,
1062
+ "reward": 0.12351190787740052,
1063
+ "reward_std": 0.10759956203401089,
1064
+ "rewards/accuracy_reward": 0.12351190787740052,
1065
+ "step": 96
1066
+ },
1067
+ {
1068
+ "completion_length": 816.3586463928223,
1069
+ "epoch": 0.543037088873338,
1070
+ "grad_norm": 1.8126938343048096,
1071
+ "learning_rate": 3e-07,
1072
+ "loss": 0.0,
1073
+ "reward": 0.1688988117966801,
1074
+ "reward_std": 0.13670154195278883,
1075
+ "rewards/accuracy_reward": 0.1688988117966801,
1076
+ "step": 97
1077
+ },
1078
+ {
1079
+ "completion_length": 787.2901878356934,
1080
+ "epoch": 0.5486354093771868,
1081
+ "grad_norm": 1.0017977952957153,
1082
+ "learning_rate": 3e-07,
1083
+ "loss": 0.0,
1084
+ "reward": 0.15104166907258332,
1085
+ "reward_std": 0.12999275140464306,
1086
+ "rewards/accuracy_reward": 0.15104166907258332,
1087
+ "step": 98
1088
+ },
1089
+ {
1090
+ "completion_length": 755.874267578125,
1091
+ "epoch": 0.5542337298810357,
1092
+ "grad_norm": 1.0542266368865967,
1093
+ "learning_rate": 3e-07,
1094
+ "loss": 0.0,
1095
+ "reward": 0.24479167215758935,
1096
+ "reward_std": 0.15221792878583074,
1097
+ "rewards/accuracy_reward": 0.24479167215758935,
1098
+ "step": 99
1099
+ },
1100
+ {
1101
+ "completion_length": 791.4122161865234,
1102
+ "epoch": 0.5598320503848845,
1103
+ "grad_norm": 0.7178149223327637,
1104
+ "learning_rate": 3e-07,
1105
+ "loss": 0.0,
1106
+ "reward": 0.17038690840126947,
1107
+ "reward_std": 0.13499723048880696,
1108
+ "rewards/accuracy_reward": 0.17038690840126947,
1109
+ "step": 100
1110
+ },
1111
+ {
1112
+ "completion_length": 744.4151840209961,
1113
+ "epoch": 0.5654303708887334,
1114
+ "grad_norm": 0.7150797843933105,
1115
+ "learning_rate": 3e-07,
1116
+ "loss": 0.0,
1117
+ "reward": 0.20610119280172512,
1118
+ "reward_std": 0.1412361622788012,
1119
+ "rewards/accuracy_reward": 0.20610119280172512,
1120
+ "step": 101
1121
+ },
1122
+ {
1123
+ "completion_length": 789.238109588623,
1124
+ "epoch": 0.5710286913925823,
1125
+ "grad_norm": 0.9325894117355347,
1126
+ "learning_rate": 3e-07,
1127
+ "loss": 0.0,
1128
+ "reward": 0.16443452634848654,
1129
+ "reward_std": 0.11632257583551109,
1130
+ "rewards/accuracy_reward": 0.16443452634848654,
1131
+ "step": 102
1132
+ },
1133
+ {
1134
+ "completion_length": 768.8869132995605,
1135
+ "epoch": 0.5766270118964311,
1136
+ "grad_norm": 7.234724998474121,
1137
+ "learning_rate": 3e-07,
1138
+ "loss": 0.0,
1139
+ "reward": 0.125744050310459,
1140
+ "reward_std": 0.10205298336222768,
1141
+ "rewards/accuracy_reward": 0.125744050310459,
1142
+ "step": 103
1143
+ },
1144
+ {
1145
+ "completion_length": 751.9308242797852,
1146
+ "epoch": 0.58222533240028,
1147
+ "grad_norm": 3.879939079284668,
1148
+ "learning_rate": 3e-07,
1149
+ "loss": 0.0,
1150
+ "reward": 0.1755952414823696,
1151
+ "reward_std": 0.12294519320130348,
1152
+ "rewards/accuracy_reward": 0.1755952414823696,
1153
+ "step": 104
1154
+ },
1155
+ {
1156
+ "completion_length": 627.394359588623,
1157
+ "epoch": 0.5878236529041287,
1158
+ "grad_norm": 3.7597320079803467,
1159
+ "learning_rate": 3e-07,
1160
+ "loss": 0.0,
1161
+ "reward": 0.2224702457897365,
1162
+ "reward_std": 0.15529246046207845,
1163
+ "rewards/accuracy_reward": 0.2224702457897365,
1164
+ "step": 105
1165
+ },
1166
+ {
1167
+ "completion_length": 617.4479293823242,
1168
+ "epoch": 0.5934219734079776,
1169
+ "grad_norm": 1.7537007331848145,
1170
+ "learning_rate": 3e-07,
1171
+ "loss": 0.0,
1172
+ "reward": 0.2157738134264946,
1173
+ "reward_std": 0.14103552885353565,
1174
+ "rewards/accuracy_reward": 0.2157738134264946,
1175
+ "step": 106
1176
+ },
1177
+ {
1178
+ "completion_length": 654.3489608764648,
1179
+ "epoch": 0.5990202939118264,
1180
+ "grad_norm": 1.367697834968567,
1181
+ "learning_rate": 3e-07,
1182
+ "loss": 0.0,
1183
+ "reward": 0.17485119379125535,
1184
+ "reward_std": 0.14419334661215544,
1185
+ "rewards/accuracy_reward": 0.17485119379125535,
1186
+ "step": 107
1187
+ },
1188
+ {
1189
+ "completion_length": 674.4724807739258,
1190
+ "epoch": 0.6046186144156753,
1191
+ "grad_norm": 0.4473717212677002,
1192
+ "learning_rate": 3e-07,
1193
+ "loss": 0.0,
1194
+ "reward": 0.17857143364381045,
1195
+ "reward_std": 0.11048215441405773,
1196
+ "rewards/accuracy_reward": 0.17857143364381045,
1197
+ "step": 108
1198
+ },
1199
+ {
1200
+ "completion_length": 725.7738227844238,
1201
+ "epoch": 0.6102169349195241,
1202
+ "grad_norm": 3.6793627738952637,
1203
+ "learning_rate": 3e-07,
1204
+ "loss": 0.0,
1205
+ "reward": 0.14806547824991867,
1206
+ "reward_std": 0.08863410213962197,
1207
+ "rewards/accuracy_reward": 0.14806547824991867,
1208
+ "step": 109
1209
+ },
1210
+ {
1211
+ "completion_length": 677.5327529907227,
1212
+ "epoch": 0.615815255423373,
1213
+ "grad_norm": 1.7502440214157104,
1214
+ "learning_rate": 3e-07,
1215
+ "loss": 0.0,
1216
+ "reward": 0.24627976631745696,
1217
+ "reward_std": 0.16675466718152165,
1218
+ "rewards/accuracy_reward": 0.24627976631745696,
1219
+ "step": 110
1220
+ },
1221
+ {
1222
+ "completion_length": 738.4843883514404,
1223
+ "epoch": 0.6214135759272218,
1224
+ "grad_norm": 5.396337509155273,
1225
+ "learning_rate": 3e-07,
1226
+ "loss": 0.0,
1227
+ "reward": 0.17485119419870898,
1228
+ "reward_std": 0.1276370887644589,
1229
+ "rewards/accuracy_reward": 0.17485119419870898,
1230
+ "step": 111
1231
+ },
1232
+ {
1233
+ "completion_length": 669.8638534545898,
1234
+ "epoch": 0.6270118964310707,
1235
+ "grad_norm": 13.67143726348877,
1236
+ "learning_rate": 3e-07,
1237
+ "loss": 0.0,
1238
+ "reward": 0.21726190828485414,
1239
+ "reward_std": 0.13963406695984304,
1240
+ "rewards/accuracy_reward": 0.21726190828485414,
1241
+ "step": 112
1242
+ },
1243
+ {
1244
+ "completion_length": 665.2529907226562,
1245
+ "epoch": 0.6326102169349195,
1246
+ "grad_norm": 44.409034729003906,
1247
+ "learning_rate": 3e-07,
1248
+ "loss": 0.0,
1249
+ "reward": 0.19940476591000333,
1250
+ "reward_std": 0.14031004393473268,
1251
+ "rewards/accuracy_reward": 0.19940476591000333,
1252
+ "step": 113
1253
+ },
1254
+ {
1255
+ "completion_length": 632.8326091766357,
1256
+ "epoch": 0.6382085374387684,
1257
+ "grad_norm": 10.954506874084473,
1258
+ "learning_rate": 3e-07,
1259
+ "loss": 0.0,
1260
+ "reward": 0.2090773865347728,
1261
+ "reward_std": 0.13506855070590973,
1262
+ "rewards/accuracy_reward": 0.2090773865347728,
1263
+ "step": 114
1264
+ },
1265
+ {
1266
+ "completion_length": 684.7834930419922,
1267
+ "epoch": 0.6438068579426172,
1268
+ "grad_norm": 3.232715606689453,
1269
+ "learning_rate": 3e-07,
1270
+ "loss": 0.0,
1271
+ "reward": 0.18601190974004567,
1272
+ "reward_std": 0.14560958743095398,
1273
+ "rewards/accuracy_reward": 0.18601190974004567,
1274
+ "step": 115
1275
+ },
1276
+ {
1277
+ "completion_length": 707.04465675354,
1278
+ "epoch": 0.6494051784464661,
1279
+ "grad_norm": 0.662769079208374,
1280
+ "learning_rate": 3e-07,
1281
+ "loss": 0.0,
1282
+ "reward": 0.21875000512227416,
1283
+ "reward_std": 0.1376757239922881,
1284
+ "rewards/accuracy_reward": 0.21875000512227416,
1285
+ "step": 116
1286
+ },
1287
+ {
1288
+ "completion_length": 669.0446586608887,
1289
+ "epoch": 0.655003498950315,
1290
+ "grad_norm": 0.33869776129722595,
1291
+ "learning_rate": 3e-07,
1292
+ "loss": 0.0,
1293
+ "reward": 0.21056548063643277,
1294
+ "reward_std": 0.14122196286916733,
1295
+ "rewards/accuracy_reward": 0.21056548063643277,
1296
+ "step": 117
1297
+ },
1298
+ {
1299
+ "completion_length": 684.3393020629883,
1300
+ "epoch": 0.6606018194541637,
1301
+ "grad_norm": 0.7248935103416443,
1302
+ "learning_rate": 3e-07,
1303
+ "loss": 0.0,
1304
+ "reward": 0.1614583363989368,
1305
+ "reward_std": 0.11945095984265208,
1306
+ "rewards/accuracy_reward": 0.1614583363989368,
1307
+ "step": 118
1308
+ },
1309
+ {
1310
+ "completion_length": 666.3139991760254,
1311
+ "epoch": 0.6662001399580126,
1312
+ "grad_norm": 43.89714431762695,
1313
+ "learning_rate": 3e-07,
1314
+ "loss": 0.0,
1315
+ "reward": 0.22470238665118814,
1316
+ "reward_std": 0.10662061581388116,
1317
+ "rewards/accuracy_reward": 0.22470238665118814,
1318
+ "step": 119
1319
+ },
1320
+ {
1321
+ "completion_length": 679.056568145752,
1322
+ "epoch": 0.6717984604618614,
1323
+ "grad_norm": 8.356165885925293,
1324
+ "learning_rate": 3e-07,
1325
+ "loss": 0.0,
1326
+ "reward": 0.21279762336052954,
1327
+ "reward_std": 0.124493746785447,
1328
+ "rewards/accuracy_reward": 0.21279762336052954,
1329
+ "step": 120
1330
+ },
1331
+ {
1332
+ "completion_length": 687.572925567627,
1333
+ "epoch": 0.6773967809657103,
1334
+ "grad_norm": 7.832132816314697,
1335
+ "learning_rate": 3e-07,
1336
+ "loss": 0.0,
1337
+ "reward": 0.18750000436557457,
1338
+ "reward_std": 0.14207694586366415,
1339
+ "rewards/accuracy_reward": 0.18750000436557457,
1340
+ "step": 121
1341
+ },
1342
+ {
1343
+ "completion_length": 689.9590873718262,
1344
+ "epoch": 0.6829951014695591,
1345
+ "grad_norm": 2.498194932937622,
1346
+ "learning_rate": 3e-07,
1347
+ "loss": 0.0,
1348
+ "reward": 0.15773809858364984,
1349
+ "reward_std": 0.10648334585130215,
1350
+ "rewards/accuracy_reward": 0.15773809858364984,
1351
+ "step": 122
1352
+ },
1353
+ {
1354
+ "completion_length": 704.2939109802246,
1355
+ "epoch": 0.688593421973408,
1356
+ "grad_norm": 6.899082183837891,
1357
+ "learning_rate": 3e-07,
1358
+ "loss": 0.0,
1359
+ "reward": 0.18377976468764246,
1360
+ "reward_std": 0.1484173396602273,
1361
+ "rewards/accuracy_reward": 0.18377976468764246,
1362
+ "step": 123
1363
+ },
1364
+ {
1365
+ "completion_length": 651.5878105163574,
1366
+ "epoch": 0.6941917424772568,
1367
+ "grad_norm": 7.5064544677734375,
1368
+ "learning_rate": 3e-07,
1369
+ "loss": 0.0,
1370
+ "reward": 0.2500000016298145,
1371
+ "reward_std": 0.16986942291259766,
1372
+ "rewards/accuracy_reward": 0.2500000016298145,
1373
+ "step": 124
1374
+ },
1375
+ {
1376
+ "completion_length": 695.3638496398926,
1377
+ "epoch": 0.6997900629811057,
1378
+ "grad_norm": 4.8156867027282715,
1379
+ "learning_rate": 3e-07,
1380
+ "loss": 0.0,
1381
+ "reward": 0.20238095673266798,
1382
+ "reward_std": 0.13348530931398273,
1383
+ "rewards/accuracy_reward": 0.20238095673266798,
1384
+ "step": 125
1385
+ },
1386
+ {
1387
+ "completion_length": 702.0327548980713,
1388
+ "epoch": 0.7053883834849545,
1389
+ "grad_norm": 7.651693344116211,
1390
+ "learning_rate": 3e-07,
1391
+ "loss": 0.0,
1392
+ "reward": 0.21875000500585884,
1393
+ "reward_std": 0.16874144971370697,
1394
+ "rewards/accuracy_reward": 0.21875000500585884,
1395
+ "step": 126
1396
+ },
1397
+ {
1398
+ "completion_length": 671.8259048461914,
1399
+ "epoch": 0.7109867039888034,
1400
+ "grad_norm": 6.942368507385254,
1401
+ "learning_rate": 3e-07,
1402
+ "loss": 0.0,
1403
+ "reward": 0.24330357555299997,
1404
+ "reward_std": 0.1387807228602469,
1405
+ "rewards/accuracy_reward": 0.24330357555299997,
1406
+ "step": 127
1407
+ },
1408
+ {
1409
+ "completion_length": 694.0915260314941,
1410
+ "epoch": 0.7165850244926522,
1411
+ "grad_norm": 4.375088691711426,
1412
+ "learning_rate": 3e-07,
1413
+ "loss": 0.0,
1414
+ "reward": 0.20833333634072915,
1415
+ "reward_std": 0.12121187802404165,
1416
+ "rewards/accuracy_reward": 0.20833333634072915,
1417
+ "step": 128
1418
+ },
1419
+ {
1420
+ "completion_length": 703.304328918457,
1421
+ "epoch": 0.722183344996501,
1422
+ "grad_norm": 3.8942065238952637,
1423
+ "learning_rate": 3e-07,
1424
+ "loss": 0.0,
1425
+ "reward": 0.1785714317811653,
1426
+ "reward_std": 0.1311384318396449,
1427
+ "rewards/accuracy_reward": 0.1785714317811653,
1428
+ "step": 129
1429
+ },
1430
+ {
1431
+ "completion_length": 681.0907897949219,
1432
+ "epoch": 0.72778166550035,
1433
+ "grad_norm": 2.7340543270111084,
1434
+ "learning_rate": 3e-07,
1435
+ "loss": 0.0,
1436
+ "reward": 0.1979166711680591,
1437
+ "reward_std": 0.12608638824895024,
1438
+ "rewards/accuracy_reward": 0.1979166711680591,
1439
+ "step": 130
1440
+ },
1441
+ {
1442
+ "completion_length": 695.2366180419922,
1443
+ "epoch": 0.7333799860041987,
1444
+ "grad_norm": 1.5587713718414307,
1445
+ "learning_rate": 3e-07,
1446
+ "loss": 0.0,
1447
+ "reward": 0.15922619513003156,
1448
+ "reward_std": 0.11638409737497568,
1449
+ "rewards/accuracy_reward": 0.15922619513003156,
1450
+ "step": 131
1451
+ },
1452
+ {
1453
+ "completion_length": 709.5788822174072,
1454
+ "epoch": 0.7389783065080476,
1455
+ "grad_norm": 3.3193564414978027,
1456
+ "learning_rate": 3e-07,
1457
+ "loss": 0.0,
1458
+ "reward": 0.159226194024086,
1459
+ "reward_std": 0.13807840831577778,
1460
+ "rewards/accuracy_reward": 0.159226194024086,
1461
+ "step": 132
1462
+ },
1463
+ {
1464
+ "completion_length": 698.4427185058594,
1465
+ "epoch": 0.7445766270118964,
1466
+ "grad_norm": 1.9349188804626465,
1467
+ "learning_rate": 3e-07,
1468
+ "loss": 0.0,
1469
+ "reward": 0.19717262242920697,
1470
+ "reward_std": 0.12813474284484982,
1471
+ "rewards/accuracy_reward": 0.19717262242920697,
1472
+ "step": 133
1473
+ },
1474
+ {
1475
+ "completion_length": 688.2210006713867,
1476
+ "epoch": 0.7501749475157453,
1477
+ "grad_norm": 2.0050582885742188,
1478
+ "learning_rate": 3e-07,
1479
+ "loss": 0.0,
1480
+ "reward": 0.20758929062867537,
1481
+ "reward_std": 0.1788321421481669,
1482
+ "rewards/accuracy_reward": 0.20758929062867537,
1483
+ "step": 134
1484
+ },
1485
+ {
1486
+ "completion_length": 696.2515087127686,
1487
+ "epoch": 0.7557732680195941,
1488
+ "grad_norm": 0.6444657444953918,
1489
+ "learning_rate": 3e-07,
1490
+ "loss": 0.0,
1491
+ "reward": 0.19642857677536085,
1492
+ "reward_std": 0.12122054304927588,
1493
+ "rewards/accuracy_reward": 0.19642857677536085,
1494
+ "step": 135
1495
+ },
1496
+ {
1497
+ "completion_length": 694.3489761352539,
1498
+ "epoch": 0.761371588523443,
1499
+ "grad_norm": 0.5755271315574646,
1500
+ "learning_rate": 3e-07,
1501
+ "loss": 0.0,
1502
+ "reward": 0.2113095311797224,
1503
+ "reward_std": 0.15177863789722323,
1504
+ "rewards/accuracy_reward": 0.2113095311797224,
1505
+ "step": 136
1506
+ },
1507
+ {
1508
+ "completion_length": 703.0320091247559,
1509
+ "epoch": 0.7669699090272918,
1510
+ "grad_norm": 0.382269024848938,
1511
+ "learning_rate": 3e-07,
1512
+ "loss": 0.0,
1513
+ "reward": 0.19419643247965723,
1514
+ "reward_std": 0.11693665431812406,
1515
+ "rewards/accuracy_reward": 0.19419643247965723,
1516
+ "step": 137
1517
+ },
1518
+ {
1519
+ "completion_length": 705.675615310669,
1520
+ "epoch": 0.7725682295311407,
1521
+ "grad_norm": 0.23526860773563385,
1522
+ "learning_rate": 3e-07,
1523
+ "loss": 0.0,
1524
+ "reward": 0.17187500471482053,
1525
+ "reward_std": 0.12707782164216042,
1526
+ "rewards/accuracy_reward": 0.17187500471482053,
1527
+ "step": 138
1528
+ },
1529
+ {
1530
+ "completion_length": 673.7269477844238,
1531
+ "epoch": 0.7781665500349895,
1532
+ "grad_norm": 0.22726382315158844,
1533
+ "learning_rate": 3e-07,
1534
+ "loss": 0.0,
1535
+ "reward": 0.2581845277454704,
1536
+ "reward_std": 0.1626861086115241,
1537
+ "rewards/accuracy_reward": 0.2581845277454704,
1538
+ "step": 139
1539
+ },
1540
+ {
1541
+ "completion_length": 701.4620704650879,
1542
+ "epoch": 0.7837648705388384,
1543
+ "grad_norm": 0.19594760239124298,
1544
+ "learning_rate": 3e-07,
1545
+ "loss": 0.0,
1546
+ "reward": 0.21726190811023116,
1547
+ "reward_std": 0.15826332941651344,
1548
+ "rewards/accuracy_reward": 0.21726190811023116,
1549
+ "step": 140
1550
+ },
1551
+ {
1552
+ "completion_length": 726.7105808258057,
1553
+ "epoch": 0.7893631910426872,
1554
+ "grad_norm": 0.21776525676250458,
1555
+ "learning_rate": 3e-07,
1556
+ "loss": 0.0,
1557
+ "reward": 0.19122024084208533,
1558
+ "reward_std": 0.13169277971610427,
1559
+ "rewards/accuracy_reward": 0.19122024084208533,
1560
+ "step": 141
1561
+ },
1562
+ {
1563
+ "completion_length": 709.7366180419922,
1564
+ "epoch": 0.794961511546536,
1565
+ "grad_norm": 0.17518886923789978,
1566
+ "learning_rate": 3e-07,
1567
+ "loss": 0.0,
1568
+ "reward": 0.19642857578583062,
1569
+ "reward_std": 0.13794927392154932,
1570
+ "rewards/accuracy_reward": 0.19642857578583062,
1571
+ "step": 142
1572
+ },
1573
+ {
1574
+ "completion_length": 700.7433128356934,
1575
+ "epoch": 0.8005598320503848,
1576
+ "grad_norm": 0.1962345391511917,
1577
+ "learning_rate": 3e-07,
1578
+ "loss": 0.0,
1579
+ "reward": 0.2261904808692634,
1580
+ "reward_std": 0.17080192361027002,
1581
+ "rewards/accuracy_reward": 0.2261904808692634,
1582
+ "step": 143
1583
+ },
1584
+ {
1585
+ "completion_length": 749.479175567627,
1586
+ "epoch": 0.8061581525542337,
1587
+ "grad_norm": 0.09322026371955872,
1588
+ "learning_rate": 3e-07,
1589
+ "loss": 0.0,
1590
+ "reward": 0.14434524165699258,
1591
+ "reward_std": 0.07725285878404975,
1592
+ "rewards/accuracy_reward": 0.14434524165699258,
1593
+ "step": 144
1594
+ },
1595
+ {
1596
+ "completion_length": 726.3467464447021,
1597
+ "epoch": 0.8117564730580826,
1598
+ "grad_norm": 0.1386028379201889,
1599
+ "learning_rate": 3e-07,
1600
+ "loss": 0.0,
1601
+ "reward": 0.2031250053551048,
1602
+ "reward_std": 0.14093447849154472,
1603
+ "rewards/accuracy_reward": 0.2031250053551048,
1604
+ "step": 145
1605
+ },
1606
+ {
1607
+ "completion_length": 738.7120704650879,
1608
+ "epoch": 0.8173547935619314,
1609
+ "grad_norm": 0.12930598855018616,
1610
+ "learning_rate": 3e-07,
1611
+ "loss": 0.0,
1612
+ "reward": 0.17857143195578828,
1613
+ "reward_std": 0.12077544536441565,
1614
+ "rewards/accuracy_reward": 0.17857143195578828,
1615
+ "step": 146
1616
+ },
1617
+ {
1618
+ "completion_length": 765.6473388671875,
1619
+ "epoch": 0.8229531140657803,
1620
+ "grad_norm": 0.13917317986488342,
1621
+ "learning_rate": 3e-07,
1622
+ "loss": 0.0,
1623
+ "reward": 0.12425595498643816,
1624
+ "reward_std": 0.12115136627107859,
1625
+ "rewards/accuracy_reward": 0.12425595498643816,
1626
+ "step": 147
1627
+ },
1628
+ {
1629
+ "completion_length": 741.0476341247559,
1630
+ "epoch": 0.8285514345696291,
1631
+ "grad_norm": 0.5330355763435364,
1632
+ "learning_rate": 3e-07,
1633
+ "loss": 0.0,
1634
+ "reward": 0.18080357613507658,
1635
+ "reward_std": 0.12820188701152802,
1636
+ "rewards/accuracy_reward": 0.18080357613507658,
1637
+ "step": 148
1638
+ },
1639
+ {
1640
+ "completion_length": 715.349723815918,
1641
+ "epoch": 0.834149755073478,
1642
+ "grad_norm": 0.1551419347524643,
1643
+ "learning_rate": 3e-07,
1644
+ "loss": 0.0,
1645
+ "reward": 0.18452381377574056,
1646
+ "reward_std": 0.13737005554139614,
1647
+ "rewards/accuracy_reward": 0.18452381377574056,
1648
+ "step": 149
1649
+ },
1650
+ {
1651
+ "completion_length": 748.0156364440918,
1652
+ "epoch": 0.8397480755773268,
1653
+ "grad_norm": 0.17706456780433655,
1654
+ "learning_rate": 3e-07,
1655
+ "loss": 0.0,
1656
+ "reward": 0.19047619315097108,
1657
+ "reward_std": 0.10885073570534587,
1658
+ "rewards/accuracy_reward": 0.19047619315097108,
1659
+ "step": 150
1660
+ },
1661
+ {
1662
+ "completion_length": 772.0535888671875,
1663
+ "epoch": 0.8453463960811757,
1664
+ "grad_norm": 0.3694287836551666,
1665
+ "learning_rate": 3e-07,
1666
+ "loss": 0.0,
1667
+ "reward": 0.17336309852544218,
1668
+ "reward_std": 0.11179409665055573,
1669
+ "rewards/accuracy_reward": 0.17336309852544218,
1670
+ "step": 151
1671
+ },
1672
+ {
1673
+ "completion_length": 774.1912307739258,
1674
+ "epoch": 0.8509447165850245,
1675
+ "grad_norm": 0.2840059697628021,
1676
+ "learning_rate": 3e-07,
1677
+ "loss": 0.0,
1678
+ "reward": 0.15252976468764246,
1679
+ "reward_std": 0.11831123428419232,
1680
+ "rewards/accuracy_reward": 0.15252976468764246,
1681
+ "step": 152
1682
+ },
1683
+ {
1684
+ "completion_length": 760.9918327331543,
1685
+ "epoch": 0.8565430370888734,
1686
+ "grad_norm": 0.2995321452617645,
1687
+ "learning_rate": 3e-07,
1688
+ "loss": 0.0,
1689
+ "reward": 0.1778273859526962,
1690
+ "reward_std": 0.15920937061309814,
1691
+ "rewards/accuracy_reward": 0.1778273859526962,
1692
+ "step": 153
1693
+ },
1694
+ {
1695
+ "completion_length": 763.8556671142578,
1696
+ "epoch": 0.8621413575927221,
1697
+ "grad_norm": 1.4188278913497925,
1698
+ "learning_rate": 3e-07,
1699
+ "loss": 0.0,
1700
+ "reward": 0.1696428614668548,
1701
+ "reward_std": 0.12167660798877478,
1702
+ "rewards/accuracy_reward": 0.1696428614668548,
1703
+ "step": 154
1704
+ },
1705
+ {
1706
+ "completion_length": 771.9032936096191,
1707
+ "epoch": 0.867739678096571,
1708
+ "grad_norm": 0.15566886961460114,
1709
+ "learning_rate": 3e-07,
1710
+ "loss": 0.0,
1711
+ "reward": 0.20684524130774662,
1712
+ "reward_std": 0.13866402814164758,
1713
+ "rewards/accuracy_reward": 0.20684524130774662,
1714
+ "step": 155
1715
+ },
1716
+ {
1717
+ "completion_length": 797.9323081970215,
1718
+ "epoch": 0.8733379986004198,
1719
+ "grad_norm": 0.21956497430801392,
1720
+ "learning_rate": 3e-07,
1721
+ "loss": 0.0,
1722
+ "reward": 0.11979166796663776,
1723
+ "reward_std": 0.08592891087755561,
1724
+ "rewards/accuracy_reward": 0.11979166796663776,
1725
+ "step": 156
1726
+ },
1727
+ {
1728
+ "completion_length": 803.9062576293945,
1729
+ "epoch": 0.8789363191042687,
1730
+ "grad_norm": 0.4193490743637085,
1731
+ "learning_rate": 3e-07,
1732
+ "loss": 0.0,
1733
+ "reward": 0.14955357427243143,
1734
+ "reward_std": 0.13559656590223312,
1735
+ "rewards/accuracy_reward": 0.14955357427243143,
1736
+ "step": 157
1737
+ },
1738
+ {
1739
+ "completion_length": 734.5416793823242,
1740
+ "epoch": 0.8845346396081175,
1741
+ "grad_norm": 0.32442227005958557,
1742
+ "learning_rate": 3e-07,
1743
+ "loss": 0.0,
1744
+ "reward": 0.20312500384170562,
1745
+ "reward_std": 0.13238779548555613,
1746
+ "rewards/accuracy_reward": 0.20312500384170562,
1747
+ "step": 158
1748
+ },
1749
+ {
1750
+ "completion_length": 780.0461502075195,
1751
+ "epoch": 0.8901329601119664,
1752
+ "grad_norm": 0.2237851321697235,
1753
+ "learning_rate": 3e-07,
1754
+ "loss": 0.0,
1755
+ "reward": 0.15773809980601072,
1756
+ "reward_std": 0.11771008232608438,
1757
+ "rewards/accuracy_reward": 0.15773809980601072,
1758
+ "step": 159
1759
+ },
1760
+ {
1761
+ "completion_length": 725.6488285064697,
1762
+ "epoch": 0.8957312806158153,
1763
+ "grad_norm": 0.20033074915409088,
1764
+ "learning_rate": 3e-07,
1765
+ "loss": 0.0,
1766
+ "reward": 0.19122024055104703,
1767
+ "reward_std": 0.13583081704564393,
1768
+ "rewards/accuracy_reward": 0.19122024055104703,
1769
+ "step": 160
1770
+ },
1771
+ {
1772
+ "completion_length": 790.0461349487305,
1773
+ "epoch": 0.9013296011196641,
1774
+ "grad_norm": 0.2317737340927124,
1775
+ "learning_rate": 3e-07,
1776
+ "loss": 0.0,
1777
+ "reward": 0.19345238577807322,
1778
+ "reward_std": 0.17005419172346592,
1779
+ "rewards/accuracy_reward": 0.19345238577807322,
1780
+ "step": 161
1781
+ },
1782
+ {
1783
+ "completion_length": 770.9166851043701,
1784
+ "epoch": 0.906927921623513,
1785
+ "grad_norm": 0.19784215092658997,
1786
+ "learning_rate": 3e-07,
1787
+ "loss": 0.0,
1788
+ "reward": 0.16592262306949124,
1789
+ "reward_std": 0.12148957094177604,
1790
+ "rewards/accuracy_reward": 0.16592262306949124,
1791
+ "step": 162
1792
+ },
1793
+ {
1794
+ "completion_length": 801.5565567016602,
1795
+ "epoch": 0.9125262421273618,
1796
+ "grad_norm": 0.16940854489803314,
1797
+ "learning_rate": 3e-07,
1798
+ "loss": 0.0,
1799
+ "reward": 0.14806547743501142,
1800
+ "reward_std": 0.14629250299185514,
1801
+ "rewards/accuracy_reward": 0.14806547743501142,
1802
+ "step": 163
1803
+ },
1804
+ {
1805
+ "completion_length": 758.389892578125,
1806
+ "epoch": 0.9181245626312107,
1807
+ "grad_norm": 0.251857727766037,
1808
+ "learning_rate": 3e-07,
1809
+ "loss": 0.0,
1810
+ "reward": 0.1979166735545732,
1811
+ "reward_std": 0.13391391886398196,
1812
+ "rewards/accuracy_reward": 0.1979166735545732,
1813
+ "step": 164
1814
+ },
1815
+ {
1816
+ "completion_length": 778.1622200012207,
1817
+ "epoch": 0.9237228831350595,
1818
+ "grad_norm": 0.2375626116991043,
1819
+ "learning_rate": 3e-07,
1820
+ "loss": 0.0,
1821
+ "reward": 0.17038691049674526,
1822
+ "reward_std": 0.12663642317056656,
1823
+ "rewards/accuracy_reward": 0.17038691049674526,
1824
+ "step": 165
1825
+ },
1826
+ {
1827
+ "completion_length": 788.4985237121582,
1828
+ "epoch": 0.9293212036389084,
1829
+ "grad_norm": 0.2811802327632904,
1830
+ "learning_rate": 3e-07,
1831
+ "loss": 0.0,
1832
+ "reward": 0.1428571462747641,
1833
+ "reward_std": 0.09295102627947927,
1834
+ "rewards/accuracy_reward": 0.1428571462747641,
1835
+ "step": 166
1836
+ },
1837
+ {
1838
+ "completion_length": 706.411470413208,
1839
+ "epoch": 0.9349195241427571,
1840
+ "grad_norm": 0.75251305103302,
1841
+ "learning_rate": 3e-07,
1842
+ "loss": 0.0,
1843
+ "reward": 0.2306547665502876,
1844
+ "reward_std": 0.14719042740762234,
1845
+ "rewards/accuracy_reward": 0.2306547665502876,
1846
+ "step": 167
1847
+ },
1848
+ {
1849
+ "completion_length": 744.7879638671875,
1850
+ "epoch": 0.940517844646606,
1851
+ "grad_norm": 0.21151037514209747,
1852
+ "learning_rate": 3e-07,
1853
+ "loss": 0.0,
1854
+ "reward": 0.16964286006987095,
1855
+ "reward_std": 0.13533117901533842,
1856
+ "rewards/accuracy_reward": 0.16964286006987095,
1857
+ "step": 168
1858
+ },
1859
+ {
1860
+ "completion_length": 734.4300689697266,
1861
+ "epoch": 0.9461161651504548,
1862
+ "grad_norm": 0.24678246676921844,
1863
+ "learning_rate": 3e-07,
1864
+ "loss": 0.0,
1865
+ "reward": 0.17336309817619622,
1866
+ "reward_std": 0.14309520740061998,
1867
+ "rewards/accuracy_reward": 0.17336309817619622,
1868
+ "step": 169
1869
+ },
1870
+ {
1871
+ "completion_length": 724.0357284545898,
1872
+ "epoch": 0.9517144856543037,
1873
+ "grad_norm": 0.36960089206695557,
1874
+ "learning_rate": 3e-07,
1875
+ "loss": 0.0,
1876
+ "reward": 0.22023809846723452,
1877
+ "reward_std": 0.16868228651583195,
1878
+ "rewards/accuracy_reward": 0.22023809846723452,
1879
+ "step": 170
1880
+ },
1881
+ {
1882
+ "completion_length": 681.9323043823242,
1883
+ "epoch": 0.9573128061581525,
1884
+ "grad_norm": 0.3980717062950134,
1885
+ "learning_rate": 3e-07,
1886
+ "loss": 0.0,
1887
+ "reward": 0.25297619425691664,
1888
+ "reward_std": 0.1777943717315793,
1889
+ "rewards/accuracy_reward": 0.25297619425691664,
1890
+ "step": 171
1891
+ },
1892
+ {
1893
+ "completion_length": 706.1183166503906,
1894
+ "epoch": 0.9629111266620014,
1895
+ "grad_norm": 0.1279979944229126,
1896
+ "learning_rate": 3e-07,
1897
+ "loss": 0.0,
1898
+ "reward": 0.18601190811023116,
1899
+ "reward_std": 0.12339989701285958,
1900
+ "rewards/accuracy_reward": 0.18601190811023116,
1901
+ "step": 172
1902
+ },
1903
+ {
1904
+ "completion_length": 696.6547718048096,
1905
+ "epoch": 0.9685094471658502,
1906
+ "grad_norm": 0.10279275476932526,
1907
+ "learning_rate": 3e-07,
1908
+ "loss": 0.0,
1909
+ "reward": 0.17187500558793545,
1910
+ "reward_std": 0.1345887309871614,
1911
+ "rewards/accuracy_reward": 0.17187500558793545,
1912
+ "step": 173
1913
+ },
1914
+ {
1915
+ "completion_length": 659.0766487121582,
1916
+ "epoch": 0.9741077676696991,
1917
+ "grad_norm": 0.14932937920093536,
1918
+ "learning_rate": 3e-07,
1919
+ "loss": 0.0,
1920
+ "reward": 0.23511904943734407,
1921
+ "reward_std": 0.12750320974737406,
1922
+ "rewards/accuracy_reward": 0.23511904943734407,
1923
+ "step": 174
1924
+ },
1925
+ {
1926
+ "completion_length": 710.5974884033203,
1927
+ "epoch": 0.979706088173548,
1928
+ "grad_norm": 0.15650224685668945,
1929
+ "learning_rate": 3e-07,
1930
+ "loss": 0.0,
1931
+ "reward": 0.20089286181610078,
1932
+ "reward_std": 0.1395609532482922,
1933
+ "rewards/accuracy_reward": 0.20089286181610078,
1934
+ "step": 175
1935
+ },
1936
+ {
1937
+ "completion_length": 752.3147430419922,
1938
+ "epoch": 0.9853044086773968,
1939
+ "grad_norm": 0.14169637858867645,
1940
+ "learning_rate": 3e-07,
1941
+ "loss": 0.0,
1942
+ "reward": 0.1659226217889227,
1943
+ "reward_std": 0.14061464229598641,
1944
+ "rewards/accuracy_reward": 0.1659226217889227,
1945
+ "step": 176
1946
+ },
1947
+ {
1948
+ "completion_length": 718.4464435577393,
1949
+ "epoch": 0.9909027291812457,
1950
+ "grad_norm": 0.1935679018497467,
1951
+ "learning_rate": 3e-07,
1952
+ "loss": 0.0,
1953
+ "reward": 0.20907738618552685,
1954
+ "reward_std": 0.13396611297503114,
1955
+ "rewards/accuracy_reward": 0.20907738618552685,
1956
+ "step": 177
1957
+ },
1958
+ {
1959
+ "completion_length": 753.9546356201172,
1960
+ "epoch": 0.9965010496850945,
1961
+ "grad_norm": 0.23016896843910217,
1962
+ "learning_rate": 3e-07,
1963
+ "loss": 0.0,
1964
+ "reward": 0.1510416681994684,
1965
+ "reward_std": 0.11266338312998414,
1966
+ "rewards/accuracy_reward": 0.1510416681994684,
1967
+ "step": 178
1968
+ },
1969
+ {
1970
+ "epoch": 0.9965010496850945,
1971
+ "step": 178,
1972
+ "total_flos": 0.0,
1973
+ "train_loss": 1.002405131310157e-08,
1974
+ "train_runtime": 79864.5906,
1975
+ "train_samples_per_second": 0.25,
1976
+ "train_steps_per_second": 0.002
1977
+ }
1978
+ ],
1979
+ "logging_steps": 1,
1980
+ "max_steps": 178,
1981
+ "num_input_tokens_seen": 0,
1982
+ "num_train_epochs": 1,
1983
+ "save_steps": 10,
1984
+ "stateful_callbacks": {
1985
+ "TrainerControl": {
1986
+ "args": {
1987
+ "should_epoch_stop": false,
1988
+ "should_evaluate": false,
1989
+ "should_log": false,
1990
+ "should_save": true,
1991
+ "should_training_stop": true
1992
+ },
1993
+ "attributes": {}
1994
+ }
1995
+ },
1996
+ "total_flos": 0.0,
1997
+ "train_batch_size": 1,
1998
+ "trial_name": null,
1999
+ "trial_params": null
2000
+ }