Hugging Face's logo

pavan01729
/
sld-models

Safetensors
Model card Files
xet
 Community
pavan01729 commited on
Commit
2dbc717
·
verified ·
1 Parent(s): 34cd0c9

Upload folder using huggingface_hub

Browse files
Files changed (10) hide show
  1. sld_26.0/panel_counter_v20_27b_grpo/README.md +34 -0
  2. sld_26.0/panel_counter_v20_27b_grpo/eval/phase0_serve_stop.log +4 -0
  3. sld_26.0/panel_counter_v20_27b_grpo/eval/phase2_sft.log +0 -0
  4. sld_26.0/panel_counter_v20_27b_grpo/eval/phase3_smoke.log +0 -0
  5. sld_26.0/panel_counter_v20_27b_grpo/eval/phase4_grpo_20260705-155119.log +0 -0
  6. sld_26.0/panel_counter_v20_27b_grpo/eval/run_20260704-161913.log +108 -0
  7. sld_26.0/panel_counter_v20_27b_grpo/eval/run_20260704-162424.log +0 -0
  8. sld_26.0/panel_counter_v20_27b_grpo/eval/run_20260705-150056.log +0 -0
  9. sld_26.0/panel_counter_v20_27b_grpo/eval/v20_27b_sft_best_ckpt.txt +1 -0
  10. sld_26.0/panel_counter_v20_27b_grpo/training_logging_full.jsonl +37 -0
sld_26.0/panel_counter_v20_27b_grpo/README.md ADDED
@@ -0,0 +1,34 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # panel_counter_v20_27b_grpo (archived, superseded)
2
+
3
+ Source: `sld_26.0/training/checkpoints/panel_counter_v20_27b_grpo`
4
+ Final checkpoint archived: `checkpoint-35` (step 35) from run `v0-20260705-155130`
5
+ Training run segments (resumed sequentially): ['v0-20260705-155130']
6
+
7
+ This is an **inference-only** archive: adapter weights + configs + full training log +
8
+ eval results. Optimizer/scheduler/RNG state (needed only to resume this exact run) was
9
+ dropped to save space; this run is superseded by later versions and not intended to be
10
+ resumed.
11
+
12
+ ## Training summary
13
+ ```
14
+ {
15
+ "reward_first": 0.07907143,
16
+ "reward_last": 0.0,
17
+ "reward_peak": 0.375,
18
+ "reward_peak_step": 11,
19
+ "n_reward_points": 37,
20
+ "loss_first": -0.23855455,
21
+ "loss_last": 0.00251389
22
+ }
23
+ ```
24
+ Total logged steps across all resumed segments: 37
25
+
26
+ ## Eval files included
27
+ - eval/phase0_serve_stop.log
28
+ - eval/phase2_sft.log
29
+ - eval/phase3_smoke.log
30
+ - eval/phase4_grpo_20260705-155119.log
31
+ - eval/run_20260704-161913.log
32
+ - eval/run_20260704-162424.log
33
+ - eval/run_20260705-150056.log
34
+ - eval/v20_27b_sft_best_ckpt.txt
sld_26.0/panel_counter_v20_27b_grpo/eval/phase0_serve_stop.log ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ bash: warning: setlocale: LC_ALL: cannot change locale (en_US.UTF-8)
2
+ [serve_v19] target gpu-memory-utilization=stop extra_lora=none
3
+ [serve_v19] no existing vLLM process found
4
+ [serve_v19] stop requested — vLLM killed, not relaunching
sld_26.0/panel_counter_v20_27b_grpo/eval/phase2_sft.log ADDED
The diff for this file is too large to render. See raw diff
 
sld_26.0/panel_counter_v20_27b_grpo/eval/phase3_smoke.log ADDED
The diff for this file is too large to render. See raw diff
 
sld_26.0/panel_counter_v20_27b_grpo/eval/phase4_grpo_20260705-155119.log ADDED
The diff for this file is too large to render. See raw diff
 
sld_26.0/panel_counter_v20_27b_grpo/eval/run_20260704-161913.log ADDED
@@ -0,0 +1,108 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ bash: warning: setlocale: LC_ALL: cannot change locale (en_US.UTF-8)
2
+ [phase 0] preflight — 2026-07-04T16:19:13+00:00
3
+ 33886 MiB, 11573 MiB, 46068 MiB
4
+ /dev/nvme1n1 850G 679G 171G 80% /workspace
5
+ bash: warning: setlocale: LC_ALL: cannot change locale (en_US.UTF-8)
6
+ [serve_v19] target gpu-memory-utilization=stop extra_lora=none
7
+ [serve_v19] killing existing vLLM pids: 1098035
8
+ 1098138
9
+ [serve_v19] SIGKILL fallback for pid 1098035
10
+ [serve_v19] stop requested — vLLM killed, not relaunching
11
+ [phase 0] done at 3s
12
+ [phase 1] data check — 2026-07-04T16:19:16+00:00
13
+ reusing: sft_v19_pure30_train.jsonl (27), sft_v19_pure30_eval.jsonl (3), grpo_v19_pure30_train.jsonl (27)
14
+ [phase 1] done at 3s
15
+ [phase 2] QLoRA SFT (Qwen3.6-27B) — 2026-07-04T16:19:16+00:00
16
+ bash: warning: setlocale: LC_ALL: cannot change locale (en_US.UTF-8)
17
+ run sh: `/usr/bin/python3 /home/node/.local/lib/python3.11/site-packages/swift/cli/sft.py --model Qwen/Qwen3.6-27B --model_type qwen3_5 --template qwen3_5 --quant_method bnb --quant_bits 4 --tuner_type lora --dataset /workspace/training/sft_v19_pure30_train.jsonl --val_dataset /workspace/training/sft_v19_pure30_eval.jsonl --output_dir /workspace/training/checkpoints/panel_counter_v20_27b_sft --torch_dtype bfloat16 --lora_rank 64 --lora_alpha 128 --target_modules all-linear --num_train_epochs 15 --per_device_train_batch_size 1 --per_device_eval_batch_size 1 --gradient_accumulation_steps 8 --learning_rate 5e-5 --warmup_ratio 0.05 --weight_decay 0.1 --lr_scheduler_type cosine --max_length 8192 --eval_strategy steps --eval_steps 4 --save_strategy steps --save_steps 4 --save_total_limit 4 --load_best_model_at_end true --metric_for_best_model loss --greater_is_better false --early_stop_interval 3 --logging_steps 2 --bf16 true --gradient_checkpointing true --optim adamw_torch_fused --enable_thinking false --report_to none`
18
+ [INFO:swift] Successfully registered `/home/node/.local/lib/python3.11/site-packages/swift/dataset/data/dataset_info.json`.
19
+ [INFO:swift] rank: -1, local_rank: -1, world_size: 1, local_world_size: 1
20
+ [INFO:swift] Downloading the model from HuggingFace Hub, model_id: Qwen/Qwen3.6-27B
21
+
22
+
23
+
24
+ Traceback (most recent call last):
25
+ File "/home/node/.local/lib/python3.11/site-packages/swift/cli/sft.py", line 20, in <module>
26
+ sft_main()
27
+ File "/home/node/.local/lib/python3.11/site-packages/swift/pipelines/train/sft.py", line 354, in sft_main
28
+ return SwiftSft(args).main()
29
+ ^^^^^^^^^^^^^^
30
+ File "/home/node/.local/lib/python3.11/site-packages/swift/pipelines/train/sft.py", line 27, in __init__
31
+ super().__init__(args)
32
+ File "/home/node/.local/lib/python3.11/site-packages/swift/pipelines/base.py", line 18, in __init__
33
+ self.args = self._parse_args(args)
34
+ ^^^^^^^^^^^^^^^^^^^^^^
35
+ File "/home/node/.local/lib/python3.11/site-packages/swift/pipelines/base.py", line 35, in _parse_args
36
+ args, remaining_argv = parse_args(self.args_class, args)
37
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
38
+ File "/home/node/.local/lib/python3.11/site-packages/swift/utils/utils.py", line 182, in parse_args
39
+ args, remaining_args = parser.parse_args_into_dataclasses(argv, return_remaining_strings=True)
40
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
41
+ File "/home/node/.local/lib/python3.11/site-packages/transformers/hf_argparser.py", line 345, in parse_args_into_dataclasses
42
+ obj = dtype(**inputs)
43
+ ^^^^^^^^^^^^^^^
44
+ File "<string>", line 320, in __init__
45
+ File "/home/node/.local/lib/python3.11/site-packages/swift/arguments/sft_args.py", line 204, in __post_init__
46
+ BaseArguments.__post_init__(self)
47
+ File "/home/node/.local/lib/python3.11/site-packages/swift/arguments/base_args/base_args.py", line 167, in __post_init__
48
+ ModelArguments.__post_init__(self)
49
+ File "/home/node/.local/lib/python3.11/site-packages/swift/arguments/base_args/model_args.py", line 224, in __post_init__
50
+ self._init_torch_dtype()
51
+ File "/home/node/.local/lib/python3.11/site-packages/swift/arguments/base_args/model_args.py", line 123, in _init_torch_dtype
52
+ self.torch_dtype: torch.dtype = self._init_model_info()
53
+ ^^^^^^^^^^^^^^^^^^^^^^^
54
+ File "/home/node/.local/lib/python3.11/site-packages/swift/arguments/base_args/model_args.py", line 193, in _init_model_info
55
+ self.model_info, self.model_meta = get_model_info_meta(**model_kwargs)
56
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
57
+ File "/home/node/.local/lib/python3.11/site-packages/swift/model/model_meta.py", line 265, in get_model_info_meta
58
+ model_dir = safe_snapshot_download(
59
+ ^^^^^^^^^^^^^^^^^^^^^^^
60
+ File "/home/node/.local/lib/python3.11/site-packages/swift/utils/hub_utils.py", line 100, in safe_snapshot_download
61
+ model_dir = hub.download_model(model_id_or_path, revision, ignore_patterns, token=hub_token, **kwargs)
62
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
63
+ File "/home/node/.local/lib/python3.11/site-packages/swift/hub/hub.py", line 455, in download_model
64
+ return snapshot_download(
65
+ ^^^^^^^^^^^^^^^^^^
66
+ File "/home/node/.local/lib/python3.11/site-packages/huggingface_hub/utils/_validators.py", line 88, in _inner_fn
67
+ return fn(*args, **kwargs)
68
+ ^^^^^^^^^^^^^^^^^^^
69
+ File "/home/node/.local/lib/python3.11/site-packages/huggingface_hub/_snapshot_download.py", line 455, in snapshot_download
70
+ thread_map(
71
+ File "/home/node/.local/lib/python3.11/site-packages/tqdm/contrib/concurrent.py", line 69, in thread_map
72
+ return _executor_map(ThreadPoolExecutor, fn, *iterables, **tqdm_kwargs)
73
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
74
+ File "/home/node/.local/lib/python3.11/site-packages/tqdm/contrib/concurrent.py", line 51, in _executor_map
75
+ return list(tqdm_class(ex.map(fn, *iterables, chunksize=chunksize), **kwargs))
76
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
77
+ File "/home/node/.local/lib/python3.11/site-packages/tqdm/std.py", line 1181, in __iter__
78
+ for obj in iterable:
79
+ File "/usr/lib/python3.11/concurrent/futures/_base.py", line 619, in result_iterator
80
+ yield _result_or_cancel(fs.pop())
81
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^
82
+ File "/usr/lib/python3.11/concurrent/futures/_base.py", line 317, in _result_or_cancel
83
+ return fut.result(timeout)
84
+ ^^^^^^^^^^^^^^^^^^^
85
+ File "/usr/lib/python3.11/concurrent/futures/_base.py", line 456, in result
86
+ return self.__get_result()
87
+ ^^^^^^^^^^^^^^^^^^^
88
+ File "/usr/lib/python3.11/concurrent/futures/_base.py", line 401, in __get_result
89
+ raise self._exception
90
+ File "/usr/lib/python3.11/concurrent/futures/thread.py", line 58, in run
91
+ result = self.fn(*self.args, **self.kwargs)
92
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
93
+ File "/home/node/.local/lib/python3.11/site-packages/huggingface_hub/_snapshot_download.py", line 435, in _inner_hf_hub_download
94
+ hf_hub_download( # type: ignore
95
+ File "/home/node/.local/lib/python3.11/site-packages/huggingface_hub/utils/_validators.py", line 88, in _inner_fn
96
+ return fn(*args, **kwargs)
97
+ ^^^^^^^^^^^^^^^^^^^
98
+ File "/home/node/.local/lib/python3.11/site-packages/huggingface_hub/file_download.py", line 997, in hf_hub_download
99
+ return _hf_hub_download_to_cache_dir(
100
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
101
+ File "/home/node/.local/lib/python3.11/site-packages/huggingface_hub/file_download.py", line 1216, in _hf_hub_download_to_cache_dir
102
+ _download_to_tmp_and_move(
103
+ File "/home/node/.local/lib/python3.11/site-packages/huggingface_hub/file_download.py", line 1842, in _download_to_tmp_and_move
104
+ xet_get(
105
+ File "/home/node/.local/lib/python3.11/site-packages/huggingface_hub/file_download.py", line 558, in xet_get
106
+ download_files(
107
+ RuntimeError: Data processing error: I/O error: Permission denied (os error 13)
108
+ FATAL: train_v20_27b_sft.sh failed (exit 1) — check phase2_sft.log (likely model download, template mismatch, or OOM)
sld_26.0/panel_counter_v20_27b_grpo/eval/run_20260704-162424.log ADDED
The diff for this file is too large to render. See raw diff
 
sld_26.0/panel_counter_v20_27b_grpo/eval/run_20260705-150056.log ADDED
The diff for this file is too large to render. See raw diff
 
sld_26.0/panel_counter_v20_27b_grpo/eval/v20_27b_sft_best_ckpt.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ /workspace/training/checkpoints/panel_counter_v20_27b_sft/v0-20260704-162435/checkpoint-20
sld_26.0/panel_counter_v20_27b_grpo/training_logging_full.jsonl ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"loss": -0.23855455, "grad_norm": 0.55953109, "learning_rate": 5e-06, "completions/mean_length": 1804.4375, "completions/min_length": 1102.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.5625, "reward": 0.07907143, "reward_std": 0.06675064, "frac_reward_zero_std": 0.25, "rewards/MultiComponentReward/mean": 0.07907143, "rewards/MultiComponentReward/std": 0.11098644, "kl": 0.10359544, "clip_ratio/low_mean": 0.00012207, "clip_ratio/low_min": 0.00012207, "clip_ratio/high_mean": 0.00276154, "clip_ratio/high_max": 0.00276154, "clip_ratio/region_mean": 0.00288361, "step_time": 677.21864874, "epoch": 0.07692308, "global_step/max_steps": "1/150", "elapsed_time": "22m 35s", "remaining_time": "2d 8h 4m 1s", "memory(GiB)": 33.89, "train_speed(s/it)": 1354.636805}
2
+ {"loss": -0.15833452, "grad_norm": 0.21745902, "learning_rate": 5e-06, "completions/mean_length": 1807.3125, "completions/min_length": 952.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.75, "reward": 0.14347115, "reward_std": 0.19707843, "frac_reward_zero_std": 0.5, "rewards/MultiComponentReward/mean": 0.14347115, "rewards/MultiComponentReward/std": 0.30085115, "kl": 0.10078631, "clip_ratio/low_mean": 0.00036585, "clip_ratio/low_min": 0.00036585, "clip_ratio/high_mean": 0.00050599, "clip_ratio/high_max": 0.00050599, "clip_ratio/region_mean": 0.00087184, "step_time": 671.07863164, "epoch": 0.15384615, "global_step/max_steps": "2/150", "elapsed_time": "44m 57s", "remaining_time": "2d 7h 26m 8s", "memory(GiB)": 38.2, "train_speed(s/it)": 1348.43571}
3
+ {"loss": -0.12163342, "grad_norm": 0.20743255, "learning_rate": 5e-06, "completions/mean_length": 1976.4375, "completions/min_length": 1475.5, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.9375, "reward": 0.01194853, "reward_std": 0.02389706, "frac_reward_zero_std": 0.75, "rewards/MultiComponentReward/mean": 0.01194853, "rewards/MultiComponentReward/std": 0.03379555, "kl": 0.0837781, "clip_ratio/low_mean": 0.00024414, "clip_ratio/low_min": 0.00024414, "clip_ratio/high_mean": 0.00027685, "clip_ratio/high_max": 0.00027685, "clip_ratio/region_mean": 0.000521, "step_time": 681.86782524, "epoch": 0.23076923, "global_step/max_steps": "3/150", "elapsed_time": "1h 7m 41s", "remaining_time": "2d 7h 16m 14s", "memory(GiB)": 38.2, "train_speed(s/it)": 1353.562994}
4
+ {"loss": 0.01899067, "grad_norm": 0.42097899, "learning_rate": 4.99e-06, "completions/mean_length": 1859.625, "completions/min_length": 723.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.8125, "reward": 0.13050595, "reward_std": 0.24677275, "frac_reward_zero_std": 0.5, "rewards/MultiComponentReward/mean": 0.13050595, "rewards/MultiComponentReward/std": 0.35200787, "kl": 0.10089052, "clip_ratio/low_mean": 0.00030518, "clip_ratio/low_min": 0.00030518, "clip_ratio/high_mean": 0.00025853, "clip_ratio/high_max": 0.00025853, "clip_ratio/region_mean": 0.00056371, "step_time": 682.23811183, "epoch": 0.30769231, "global_step/max_steps": "4/150", "elapsed_time": "1h 30m 25s", "remaining_time": "2d 7h 0m 22s", "memory(GiB)": 38.2, "train_speed(s/it)": 1356.311753}
5
+ {"loss": 0.01631328, "grad_norm": 0.12264023, "learning_rate": 4.99e-06, "completions/mean_length": 1975.3125, "completions/min_length": 1516.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.875, "reward": 0.07076389, "reward_std": 0.08200339, "frac_reward_zero_std": 0.75, "rewards/MultiComponentReward/mean": 0.07076389, "rewards/MultiComponentReward/std": 0.13134193, "kl": 0.08819868, "clip_ratio/low_mean": 6.104e-05, "clip_ratio/low_min": 6.104e-05, "clip_ratio/high_mean": 0.0003811, "clip_ratio/high_max": 0.0003811, "clip_ratio/region_mean": 0.00044213, "step_time": 681.80834189, "epoch": 0.38461538, "global_step/max_steps": "5/150", "elapsed_time": "1h 53m 9s", "remaining_time": "2d 6h 41m 20s", "memory(GiB)": 38.2, "train_speed(s/it)": 1357.792591}
6
+ {"loss": 0.12725414, "grad_norm": 0.09394885, "learning_rate": 4.98e-06, "completions/mean_length": 1995.125, "completions/min_length": 1625.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.9375, "reward": 0.05912698, "reward_std": 0.11825397, "frac_reward_zero_std": 0.75, "rewards/MultiComponentReward/mean": 0.05912698, "rewards/MultiComponentReward/std": 0.16723637, "kl": 0.05552185, "clip_ratio/low_mean": 0.00012207, "clip_ratio/low_min": 0.00012207, "clip_ratio/high_mean": 0.0, "clip_ratio/high_max": 0.0, "clip_ratio/region_mean": 0.00012207, "step_time": 664.67729748, "epoch": 0.46153846, "global_step/max_steps": "6/150", "elapsed_time": "2h 15m 23s", "remaining_time": "2d 6h 9m 4s", "memory(GiB)": 38.2, "train_speed(s/it)": 1353.774453}
7
+ {"loss": -0.12138511, "grad_norm": 0.48947242, "learning_rate": 4.97e-06, "completions/mean_length": 1811.3125, "completions/min_length": 884.5, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.75, "reward": 0.07884615, "reward_std": 0.15769231, "frac_reward_zero_std": 0.25, "rewards/MultiComponentReward/mean": 0.07884615, "rewards/MultiComponentReward/std": 0.19085121, "kl": 0.08899206, "clip_ratio/low_mean": 0.00061035, "clip_ratio/low_min": 0.00061035, "clip_ratio/high_mean": 0.00011792, "clip_ratio/high_max": 0.00011792, "clip_ratio/region_mean": 0.00072828, "step_time": 672.10490665, "epoch": 0.53846154, "global_step/max_steps": "7/150", "elapsed_time": "2h 37m 47s", "remaining_time": "2d 5h 43m 16s", "memory(GiB)": 38.2, "train_speed(s/it)": 1352.419602}
8
+ {"loss": -0.1224158, "grad_norm": 0.10015026, "learning_rate": 4.96e-06, "completions/mean_length": 2015.0, "completions/min_length": 1784.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.9375, "reward": 0.03392857, "reward_std": 0.06785715, "frac_reward_zero_std": 0.75, "rewards/MultiComponentReward/mean": 0.03392857, "rewards/MultiComponentReward/std": 0.0959645, "kl": 0.06419932, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/high_mean": 0.00032895, "clip_ratio/high_max": 0.00032895, "clip_ratio/region_mean": 0.00032895, "step_time": 679.55552981, "epoch": 0.61538462, "global_step/max_steps": "8/150", "elapsed_time": "3h 0m 26s", "remaining_time": "2d 5h 22m 44s", "memory(GiB)": 38.2, "train_speed(s/it)": 1353.266182}
9
+ {"loss": 0.09515999, "grad_norm": 0.20102166, "learning_rate": 4.96e-06, "completions/mean_length": 1896.75, "completions/min_length": 896.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.8125, "reward": 0.1875, "reward_std": 0.26933756, "frac_reward_zero_std": 0.5, "rewards/MultiComponentReward/mean": 0.1875, "rewards/MultiComponentReward/std": 0.40823172, "kl": 0.08924169, "clip_ratio/low_mean": 0.00036621, "clip_ratio/low_min": 0.00036621, "clip_ratio/high_mean": 0.00010647, "clip_ratio/high_max": 0.00010647, "clip_ratio/region_mean": 0.00047268, "step_time": 670.89900102, "epoch": 0.69230769, "global_step/max_steps": "9/150", "elapsed_time": "3h 22m 48s", "remaining_time": "2d 4h 57m 12s", "memory(GiB)": 38.2, "train_speed(s/it)": 1352.000025}
10
+ {"loss": 0.00238286, "grad_norm": 0.12615055, "learning_rate": 4.95e-06, "completions/mean_length": 1960.4375, "completions/min_length": 1347.5, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.875, "reward": 0.09242295, "reward_std": 0.18484589, "frac_reward_zero_std": 0.5, "rewards/MultiComponentReward/mean": 0.09242295, "rewards/MultiComponentReward/std": 0.26141156, "kl": 0.06393088, "clip_ratio/low_mean": 0.00036621, "clip_ratio/low_min": 0.00036621, "clip_ratio/high_mean": 0.00021866, "clip_ratio/high_max": 0.00021866, "clip_ratio/region_mean": 0.00058487, "step_time": 684.90496191, "epoch": 0.76923077, "global_step/max_steps": "10/150", "elapsed_time": "3h 45m 38s", "remaining_time": "2d 4h 38m 51s", "memory(GiB)": 38.2, "train_speed(s/it)": 1353.789302}
11
+ {"loss": -0.33839542, "grad_norm": 0.23330766, "learning_rate": 4.93e-06, "completions/mean_length": 1962.75, "completions/min_length": 1697.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.8125, "reward": 0.1875, "reward_std": 0.26933756, "frac_reward_zero_std": 0.5, "rewards/MultiComponentReward/mean": 0.1875, "rewards/MultiComponentReward/std": 0.40823172, "kl": 0.07551282, "clip_ratio/low_mean": 0.00012207, "clip_ratio/low_min": 0.00012207, "clip_ratio/high_mean": 0.00069854, "clip_ratio/high_max": 0.00069854, "clip_ratio/region_mean": 0.00082061, "step_time": 678.64507908, "epoch": 0.84615385, "global_step/max_steps": "11/150", "elapsed_time": "4h 8m 19s", "remaining_time": "2d 4h 17m 51s", "memory(GiB)": 38.2, "train_speed(s/it)": 1354.465982}
12
+ {"loss": -0.21339521, "grad_norm": 0.18358865, "learning_rate": 4.92e-06, "completions/mean_length": 1872.4375, "completions/min_length": 1071.5, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.625, "reward": 0.375, "reward_std": 0.53867513, "frac_reward_zero_std": 0.0, "rewards/MultiComponentReward/mean": 0.375, "rewards/MultiComponentReward/std": 0.49871626, "kl": 0.08055659, "clip_ratio/low_mean": 0.00042725, "clip_ratio/low_min": 0.00042725, "clip_ratio/high_mean": 0.00058642, "clip_ratio/high_max": 0.00058642, "clip_ratio/region_mean": 0.00101366, "step_time": 674.55871501, "epoch": 0.92307692, "global_step/max_steps": "12/150", "elapsed_time": "4h 30m 48s", "remaining_time": "2d 3h 54m 16s", "memory(GiB)": 38.2, "train_speed(s/it)": 1354.027088}
13
+ {"loss": -0.21283938, "grad_norm": 0.17627691, "learning_rate": 4.91e-06, "completions/mean_length": 1851.875, "completions/min_length": 1254.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.5625, "reward": 0.32764423, "reward_std": 0.29962602, "frac_reward_zero_std": 0.25, "rewards/MultiComponentReward/mean": 0.32764423, "rewards/MultiComponentReward/std": 0.48478188, "kl": 0.09238788, "clip_ratio/low_mean": 0.00019676, "clip_ratio/low_min": 0.00019676, "clip_ratio/high_mean": 0.00101498, "clip_ratio/high_max": 0.00101498, "clip_ratio/region_mean": 0.00121174, "step_time": 676.13337755, "epoch": 1.0, "global_step/max_steps": "13/150", "elapsed_time": "4h 53m 21s", "remaining_time": "2d 3h 31m 25s", "memory(GiB)": 38.2, "train_speed(s/it)": 1353.907024}
14
+ {"loss": -0.12241498, "grad_norm": 0.23631455, "learning_rate": 4.89e-06, "completions/mean_length": 1938.3125, "completions/min_length": 1518.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.8125, "reward": 0.15625, "reward_std": 0.20683756, "frac_reward_zero_std": 0.5, "rewards/MultiComponentReward/mean": 0.15625, "rewards/MultiComponentReward/std": 0.22903135, "kl": 0.06989579, "clip_ratio/low_mean": 0.00018311, "clip_ratio/low_min": 0.00018311, "clip_ratio/high_mean": 0.00038892, "clip_ratio/high_max": 0.00038892, "clip_ratio/region_mean": 0.00057202, "step_time": 660.82616335, "epoch": 1.07692308, "global_step/max_steps": "14/150", "elapsed_time": "5h 15m 23s", "remaining_time": "2d 3h 3m 40s", "memory(GiB)": 38.2, "train_speed(s/it)": 1351.618717}
15
+ {"loss": -0.06816661, "grad_norm": 0.26185188, "learning_rate": 4.88e-06, "completions/mean_length": 1814.6875, "completions/min_length": 955.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.75, "reward": 0.18125001, "reward_std": 0.3, "frac_reward_zero_std": 0.25, "rewards/MultiComponentReward/mean": 0.18125001, "rewards/MultiComponentReward/std": 0.36638924, "kl": 0.09874486, "clip_ratio/low_mean": 0.00061035, "clip_ratio/low_min": 0.00061035, "clip_ratio/high_mean": 0.00015263, "clip_ratio/high_max": 0.00015263, "clip_ratio/region_mean": 0.00076298, "step_time": 677.7168158, "epoch": 1.15384615, "global_step/max_steps": "15/150", "elapsed_time": "5h 37m 58s", "remaining_time": "2d 2h 41m 44s", "memory(GiB)": 38.2, "train_speed(s/it)": 1351.878406}
16
+ {"loss": -0.24655516, "grad_norm": 0.19601341, "learning_rate": 4.86e-06, "completions/mean_length": 1961.5625, "completions/min_length": 1356.5, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.875, "reward": 0.07048611, "reward_std": 0.14097223, "frac_reward_zero_std": 0.5, "rewards/MultiComponentReward/mean": 0.07048611, "rewards/MultiComponentReward/std": 0.19936484, "kl": 0.08533393, "clip_ratio/low_mean": 0.00012207, "clip_ratio/low_min": 0.00012207, "clip_ratio/high_mean": 0.00014156, "clip_ratio/high_max": 0.00014156, "clip_ratio/region_mean": 0.00026363, "step_time": 659.2746419, "epoch": 1.23076923, "global_step/max_steps": "16/150", "elapsed_time": "6h 0m 1s", "remaining_time": "2d 2h 15m 7s", "memory(GiB)": 38.2, "train_speed(s/it)": 1350.051763}
17
+ {"loss": 0.12819873, "grad_norm": 0.07010439, "learning_rate": 4.84e-06, "completions/mean_length": 2045.0625, "completions/min_length": 2024.5, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.9375, "reward": 0.03383152, "reward_std": 0.06766304, "frac_reward_zero_std": 0.75, "rewards/MultiComponentReward/mean": 0.03383152, "rewards/MultiComponentReward/std": 0.09569, "kl": 0.08098024, "clip_ratio/low_mean": 6.104e-05, "clip_ratio/low_min": 6.104e-05, "clip_ratio/high_mean": 0.0, "clip_ratio/high_max": 0.0, "clip_ratio/region_mean": 6.104e-05, "step_time": 685.79876713, "epoch": 1.30769231, "global_step/max_steps": "17/150", "elapsed_time": "6h 22m 53s", "remaining_time": "2d 1h 55m 26s", "memory(GiB)": 38.2, "train_speed(s/it)": 1351.324041}
18
+ {"loss": -0.1399432, "grad_norm": 0.21691871, "learning_rate": 4.82e-06, "completions/mean_length": 1918.3125, "completions/min_length": 1104.5, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.75, "reward": 0.19895834, "reward_std": 0.36933336, "frac_reward_zero_std": 0.25, "rewards/MultiComponentReward/mean": 0.19895834, "rewards/MultiComponentReward/std": 0.40656483, "kl": 0.08030141, "clip_ratio/low_mean": 0.00024414, "clip_ratio/low_min": 0.00024414, "clip_ratio/high_mean": 0.00030457, "clip_ratio/high_max": 0.00030457, "clip_ratio/region_mean": 0.00054871, "step_time": 677.04516985, "epoch": 1.38461538, "global_step/max_steps": "18/150", "elapsed_time": "6h 45m 27s", "remaining_time": "2d 1h 33m 16s", "memory(GiB)": 38.2, "train_speed(s/it)": 1351.48234}
19
+ {"loss": 0.00242862, "grad_norm": 0.02024536, "learning_rate": 4.8e-06, "completions/mean_length": 2048.0, "completions/min_length": 2048.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 1.0, "reward": 0.0, "reward_std": 0.0, "frac_reward_zero_std": 1.0, "rewards/MultiComponentReward/mean": 0.0, "rewards/MultiComponentReward/std": 0.0, "kl": 0.06071548, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/high_max": 0.0, "clip_ratio/region_mean": 0.0, "step_time": 665.19226843, "epoch": 1.46153846, "global_step/max_steps": "19/150", "elapsed_time": "7h 7m 37s", "remaining_time": "2d 1h 8m 19s", "memory(GiB)": 38.2, "train_speed(s/it)": 1350.376287}
20
+ {"loss": 0.00289571, "grad_norm": 0.03989724, "learning_rate": 4.78e-06, "completions/mean_length": 2048.0, "completions/min_length": 2048.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 1.0, "reward": 0.0, "reward_std": 0.0, "frac_reward_zero_std": 1.0, "rewards/MultiComponentReward/mean": 0.0, "rewards/MultiComponentReward/std": 0.0, "kl": 0.07239284, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/high_max": 0.0, "clip_ratio/region_mean": 0.0, "step_time": 688.06284196, "epoch": 1.53846154, "global_step/max_steps": "20/150", "elapsed_time": "7h 30m 33s", "remaining_time": "2d 0h 48m 37s", "memory(GiB)": 38.2, "train_speed(s/it)": 1351.667863}
21
+ {"loss": 0.12816253, "grad_norm": 0.06102687, "learning_rate": 4.76e-06, "completions/mean_length": 2027.25, "completions/min_length": 1882.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.9375, "reward": 0.0375, "reward_std": 0.075, "frac_reward_zero_std": 0.75, "rewards/MultiComponentReward/mean": 0.0375, "rewards/MultiComponentReward/std": 0.10606603, "kl": 0.07852038, "clip_ratio/low_mean": 0.00012207, "clip_ratio/low_min": 0.00012207, "clip_ratio/high_mean": 0.0, "clip_ratio/high_max": 0.0, "clip_ratio/region_mean": 0.00012207, "step_time": 685.13871602, "epoch": 1.61538462, "global_step/max_steps": "21/150", "elapsed_time": "7h 53m 27s", "remaining_time": "2d 0h 28m 20s", "memory(GiB)": 38.2, "train_speed(s/it)": 1352.712056}
22
+ {"loss": -0.1219337, "grad_norm": 0.12813629, "learning_rate": 4.74e-06, "completions/mean_length": 2009.75, "completions/min_length": 1742.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.9375, "reward": 0.011875, "reward_std": 0.02375, "frac_reward_zero_std": 0.75, "rewards/MultiComponentReward/mean": 0.011875, "rewards/MultiComponentReward/std": 0.03358757, "kl": 0.07422316, "clip_ratio/low_mean": 0.00012207, "clip_ratio/low_min": 0.00012207, "clip_ratio/high_mean": 0.0, "clip_ratio/high_max": 0.0, "clip_ratio/region_mean": 0.00012207, "step_time": 675.02420676, "epoch": 1.69230769, "global_step/max_steps": "22/150", "elapsed_time": "8h 15m 57s", "remaining_time": "2d 0h 5m 32s", "memory(GiB)": 38.2, "train_speed(s/it)": 1352.594248}
23
+ {"loss": 0.00253558, "grad_norm": 0.02347771, "learning_rate": 4.72e-06, "completions/mean_length": 2048.0, "completions/min_length": 2048.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 1.0, "reward": 0.0, "reward_std": 0.0, "frac_reward_zero_std": 1.0, "rewards/MultiComponentReward/mean": 0.0, "rewards/MultiComponentReward/std": 0.0, "kl": 0.06338951, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/high_max": 0.0, "clip_ratio/region_mean": 0.0, "step_time": 683.53620696, "epoch": 1.76923077, "global_step/max_steps": "23/150", "elapsed_time": "8h 38m 44s", "remaining_time": "1d 23h 44m 20s", "memory(GiB)": 38.2, "train_speed(s/it)": 1353.227087}
24
+ {"loss": -0.19086739, "grad_norm": 0.17088385, "learning_rate": 4.69e-06, "completions/mean_length": 1904.0625, "completions/min_length": 1165.5, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.6875, "reward": 0.2825, "reward_std": 0.2569498, "frac_reward_zero_std": 0.5, "rewards/MultiComponentReward/mean": 0.2825, "rewards/MultiComponentReward/std": 0.4550633, "kl": 0.07825906, "clip_ratio/low_mean": 0.00018311, "clip_ratio/low_min": 0.00018311, "clip_ratio/high_mean": 0.00074374, "clip_ratio/high_max": 0.00074374, "clip_ratio/region_mean": 0.00092684, "step_time": 681.81483265, "epoch": 1.84615385, "global_step/max_steps": "24/150", "elapsed_time": "9h 1m 28s", "remaining_time": "1d 23h 22m 42s", "memory(GiB)": 38.2, "train_speed(s/it)": 1353.663924}
25
+ {"loss": 0.12792847, "grad_norm": 0.13080813, "learning_rate": 4.67e-06, "completions/mean_length": 1929.8125, "completions/min_length": 1163.5, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.8125, "reward": 0.13269231, "reward_std": 0.26538461, "frac_reward_zero_std": 0.25, "rewards/MultiComponentReward/mean": 0.13269231, "rewards/MultiComponentReward/std": 0.35177518, "kl": 0.07278564, "clip_ratio/low_mean": 0.00048828, "clip_ratio/low_min": 0.00048828, "clip_ratio/high_mean": 6.358e-05, "clip_ratio/high_max": 6.358e-05, "clip_ratio/region_mean": 0.00055186, "step_time": 683.64309176, "epoch": 1.92307692, "global_step/max_steps": "25/150", "elapsed_time": "9h 24m 15s", "remaining_time": "1d 23h 1m 17s", "memory(GiB)": 38.2, "train_speed(s/it)": 1354.212054}
26
+ {"loss": 0.00232365, "grad_norm": 0.03190837, "learning_rate": 4.64e-06, "completions/mean_length": 2048.0, "completions/min_length": 2048.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 1.0, "reward": 0.0, "reward_std": 0.0, "frac_reward_zero_std": 1.0, "rewards/MultiComponentReward/mean": 0.0, "rewards/MultiComponentReward/std": 0.0, "kl": 0.05809125, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/high_max": 0.0, "clip_ratio/region_mean": 0.0, "step_time": 686.39553702, "epoch": 2.0, "global_step/max_steps": "26/150", "elapsed_time": "9h 47m 12s", "remaining_time": "1d 22h 40m 28s", "memory(GiB)": 38.2, "train_speed(s/it)": 1355.06808}
27
+ {"loss": 0.00231494, "grad_norm": 0.02198596, "learning_rate": 4.61e-06, "completions/mean_length": 2048.0, "completions/min_length": 2048.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 1.0, "reward": 0.0, "reward_std": 0.0, "frac_reward_zero_std": 1.0, "rewards/MultiComponentReward/mean": 0.0, "rewards/MultiComponentReward/std": 0.0, "kl": 0.0578735, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/high_max": 0.0, "clip_ratio/region_mean": 0.0, "step_time": 683.90345745, "epoch": 2.07692308, "global_step/max_steps": "27/150", "elapsed_time": "10h 9m 60s", "remaining_time": "1d 22h 18m 52s", "memory(GiB)": 38.2, "train_speed(s/it)": 1355.548065}
28
+ {"loss": 0.00208247, "grad_norm": 0.020062, "learning_rate": 4.58e-06, "completions/mean_length": 2048.0, "completions/min_length": 2048.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 1.0, "reward": 0.0, "reward_std": 0.0, "frac_reward_zero_std": 1.0, "rewards/MultiComponentReward/mean": 0.0, "rewards/MultiComponentReward/std": 0.0, "kl": 0.05206163, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/high_max": 0.0, "clip_ratio/region_mean": 0.0, "step_time": 684.97339285, "epoch": 2.15384615, "global_step/max_steps": "28/150", "elapsed_time": "10h 32m 50s", "remaining_time": "1d 21h 57m 20s", "memory(GiB)": 38.2, "train_speed(s/it)": 1356.06527}
29
+ {"loss": 0.12791054, "grad_norm": 0.07112586, "learning_rate": 4.55e-06, "completions/mean_length": 2036.5625, "completions/min_length": 1956.5, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.9375, "reward": 0.03135776, "reward_std": 0.06271552, "frac_reward_zero_std": 0.75, "rewards/MultiComponentReward/mean": 0.03135776, "rewards/MultiComponentReward/std": 0.08869313, "kl": 0.07098061, "clip_ratio/low_mean": 0.00030518, "clip_ratio/low_min": 0.00030518, "clip_ratio/high_mean": 0.0, "clip_ratio/high_max": 0.0, "clip_ratio/region_mean": 0.00030518, "step_time": 683.29152919, "epoch": 2.23076923, "global_step/max_steps": "29/150", "elapsed_time": "10h 55m 36s", "remaining_time": "1d 21h 35m 28s", "memory(GiB)": 38.2, "train_speed(s/it)": 1356.43056}
30
+ {"loss": -0.12208285, "grad_norm": 0.13502258, "learning_rate": 4.52e-06, "completions/mean_length": 1994.625, "completions/min_length": 1621.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.9375, "reward": 0.0625, "reward_std": 0.125, "frac_reward_zero_std": 0.75, "rewards/MultiComponentReward/mean": 0.0625, "rewards/MultiComponentReward/std": 0.17677669, "kl": 0.06843431, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/high_max": 0.0, "clip_ratio/region_mean": 0.0, "step_time": 663.14332553, "epoch": 2.30769231, "global_step/max_steps": "30/150", "elapsed_time": "11h 17m 43s", "remaining_time": "1d 21h 10m 51s", "memory(GiB)": 38.2, "train_speed(s/it)": 1355.428513}
31
+ {"loss": 0.00233378, "grad_norm": 0.02243361, "learning_rate": 4.49e-06, "completions/mean_length": 2048.0, "completions/min_length": 2048.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 1.0, "reward": 0.0, "reward_std": 0.0, "frac_reward_zero_std": 1.0, "rewards/MultiComponentReward/mean": 0.0, "rewards/MultiComponentReward/std": 0.0, "kl": 0.05834462, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/high_max": 0.0, "clip_ratio/region_mean": 0.0, "step_time": 687.5377189, "epoch": 2.38461538, "global_step/max_steps": "31/150", "elapsed_time": "11h 40m 42s", "remaining_time": "1d 20h 49m 46s", "memory(GiB)": 38.2, "train_speed(s/it)": 1356.182654}
32
+ {"loss": 0.00209491, "grad_norm": 0.01674586, "learning_rate": 4.46e-06, "completions/mean_length": 2048.0, "completions/min_length": 2048.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 1.0, "reward": 0.0, "reward_std": 0.0, "frac_reward_zero_std": 1.0, "rewards/MultiComponentReward/mean": 0.0, "rewards/MultiComponentReward/std": 0.0, "kl": 0.0523727, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/high_max": 0.0, "clip_ratio/region_mean": 0.0, "step_time": 690.64276414, "epoch": 2.46153846, "global_step/max_steps": "32/150", "elapsed_time": "12h 3m 43s", "remaining_time": "1d 20h 28m 42s", "memory(GiB)": 38.2, "train_speed(s/it)": 1356.969685}
33
+ {"loss": 0.00227442, "grad_norm": 0.02443803, "learning_rate": 4.43e-06, "completions/mean_length": 2048.0, "completions/min_length": 2048.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 1.0, "reward": 0.0, "reward_std": 0.0, "frac_reward_zero_std": 1.0, "rewards/MultiComponentReward/mean": 0.0, "rewards/MultiComponentReward/std": 0.0, "kl": 0.05686055, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/high_max": 0.0, "clip_ratio/region_mean": 0.0, "step_time": 686.68584064, "epoch": 2.53846154, "global_step/max_steps": "33/150", "elapsed_time": "12h 26m 36s", "remaining_time": "1d 20h 7m 4s", "memory(GiB)": 38.2, "train_speed(s/it)": 1357.469178}
34
+ {"loss": -0.122279, "grad_norm": 0.10082797, "learning_rate": 4.39e-06, "completions/mean_length": 2040.5625, "completions/min_length": 1988.5, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.9375, "reward": 0.06, "reward_std": 0.12, "frac_reward_zero_std": 0.75, "rewards/MultiComponentReward/mean": 0.06, "rewards/MultiComponentReward/std": 0.16970563, "kl": 0.06928595, "clip_ratio/low_mean": 6.104e-05, "clip_ratio/low_min": 6.104e-05, "clip_ratio/high_mean": 0.0001944, "clip_ratio/high_max": 0.0001944, "clip_ratio/region_mean": 0.00025544, "step_time": 665.45649702, "epoch": 2.61538462, "global_step/max_steps": "34/150", "elapsed_time": "12h 48m 47s", "remaining_time": "1d 19h 42m 56s", "memory(GiB)": 38.2, "train_speed(s/it)": 1356.69051}
35
+ {"loss": 0.12798718, "grad_norm": 0.10790416, "learning_rate": 4.36e-06, "completions/mean_length": 2019.5625, "completions/min_length": 1870.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.8125, "reward": 0.1875, "reward_std": 0.26933756, "frac_reward_zero_std": 0.5, "rewards/MultiComponentReward/mean": 0.1875, "rewards/MultiComponentReward/std": 0.40823172, "kl": 0.07556018, "clip_ratio/low_mean": 0.00024414, "clip_ratio/low_min": 0.00024414, "clip_ratio/high_mean": 0.00038481, "clip_ratio/high_max": 0.00038481, "clip_ratio/region_mean": 0.00062895, "step_time": 677.80443673, "epoch": 2.69230769, "global_step/max_steps": "35/150", "elapsed_time": "13h 11m 23s", "remaining_time": "1d 19h 20m 16s", "memory(GiB)": 38.2, "train_speed(s/it)": 1356.661674}
36
+ {"loss": 0.00250079, "grad_norm": 0.07767824, "learning_rate": 4.32e-06, "completions/mean_length": 2039.1875, "completions/min_length": 2003.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 0.875, "reward": 0.125, "reward_std": 0.14433756, "frac_reward_zero_std": 0.75, "rewards/MultiComponentReward/mean": 0.125, "rewards/MultiComponentReward/std": 0.23145503, "kl": 0.06197426, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/high_mean": 0.00012768, "clip_ratio/high_max": 0.00012768, "clip_ratio/region_mean": 0.00012768, "step_time": 686.03054158, "epoch": 2.76923077, "global_step/max_steps": "36/150", "elapsed_time": "13h 34m 19s", "remaining_time": "1d 18h 58m 39s", "memory(GiB)": 38.2, "train_speed(s/it)": 1357.187754}
37
+ {"loss": 0.00251389, "grad_norm": 0.01933305, "learning_rate": 4.29e-06, "completions/mean_length": 2048.0, "completions/min_length": 2048.0, "completions/max_length": 2048.0, "completions/clipped_ratio": 1.0, "reward": 0.0, "reward_std": 0.0, "frac_reward_zero_std": 1.0, "rewards/MultiComponentReward/mean": 0.0, "rewards/MultiComponentReward/std": 0.0, "kl": 0.06284722, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/high_max": 0.0, "clip_ratio/region_mean": 0.0, "step_time": 665.24035633, "epoch": 2.84615385, "global_step/max_steps": "37/150", "elapsed_time": "13h 56m 29s", "remaining_time": "1d 18h 34m 41s", "memory(GiB)": 38.2, "train_speed(s/it)": 1356.468197}