diff --git a/code/orarl.egg-info/PKG-INFO b/code/orarl.egg-info/PKG-INFO
deleted file mode 100644
index 46b3ccb3ed12bfa6b4f1556877e844f0254836b1..0000000000000000000000000000000000000000
--- a/code/orarl.egg-info/PKG-INFO
+++ /dev/null
@@ -1,230 +0,0 @@
-Metadata-Version: 2.4
-Name: orarl
-Version: 0.1.0
-Summary: Annotations as rollouts for unified video MLLM reinforcement learning
-Author: Yunheng Li, Guohong Mu, Hao Li, Shengsheng Qian, Dingwen Zhang, Qibin Hou, Ming-Ming Cheng
-License-Expression: Apache-2.0
-Project-URL: Homepage, https://orarl.github.io/
-Project-URL: Documentation, https://orarl.github.io/
-Keywords: reinforcement-learning,video-understanding,multimodal,grpo
-Classifier: Development Status :: 3 - Alpha
-Classifier: Programming Language :: Python :: 3
-Classifier: Programming Language :: Python :: 3.10
-Classifier: Programming Language :: Python :: 3.11
-Classifier: Programming Language :: Python :: 3.12
-Requires-Python: >=3.10
-Description-Content-Type: text/markdown
-License-File: LICENSE
-License-File: NOTICE
-Requires-Dist: numpy
-Requires-Dist: PyYAML
-Requires-Dist: torch
-Provides-Extra: test
-Requires-Dist: pytest; extra == "test"
-Provides-Extra: lint
-Requires-Dist: ruff; extra == "lint"
-Provides-Extra: hf
-Requires-Dist: huggingface_hub; extra == "hf"
-Dynamic: license-file
-
-
简体中文
-
-
-
-## Why OraRL
-
-- **Annotation-as-rollout:** annotations become reliable positive rollouts while
- policy samples retain an on-policy baseline.
-- **Seven task families:** one update rule covers temporal and spatial grounding,
- segmentation, tracking, spatial-temporal grounding, video QA, and spatial
- intelligence.
-- **Efficient training (4B):** sign-balanced pruning delivers **1.48× faster
- updates** (**92.5 → 62.4 s/step**) while reducing peak per-GPU memory from
- **62.4 to 50.9 GB**.
-- **Efficient inference:** on one H20 with vLLM in BF16, weight loading occupies
- **8.6 GiB (4B)** and **17.6 GiB (9B)**. On ten-minute, 2-fps videos,
- answer-only decoding cuts median post-TTFT latency from **4.78 s to 130 ms**
- and total latency from **29.03 to 24.30 s**.
-- **Multimodal veRL infrastructure:** a unified video contract carries cached
- artifacts, raw paths, or inline frame tensors through vLLM rollouts and FSDP
- updates, with decode-once frame reuse, temporal metadata, task-grouped
- batching, asynchronous Ray rewards, and safe hybrid-engine cache handling.
-
-## OraRL in One Update
-
-
-
-
-
-An OraRL update separates reliable annotation guidance from on-policy
-normalization:
-
-1. **Build the group:** append one serialized annotation rollout to the policy
- samples generated for the same prompt.
-2. **Keep the baseline on-policy:** estimate the group baseline from policy
- rewards only.
-3. **Guide and select:** convert the annotation-policy reward gap into a
- correction, then retain a sign-balanced subset for the update.
-
-This design uses task-native annotations directly and requires no
-chain-of-thought supervision or decoding.
-
-## Video-ORA Results
-
-
-
-
-
-### Dataset-Level Results
-
-
-
-
-
-
-
-Video-ORA-9B leads the matched seven-family comparison without CoT decoding.
-Best and second-best values are highlighted per row; `†` denotes an
-original-report value whose frame, prompt, split, or decoding settings may
-differ. Averages require complete family coverage.
-
-
-
-### Model Scaling
-
-
-
-
-
-### Data Scaling
-
-
-
-
-
-## Models
-
-| Model | Backbone | Released recipe | Weights |
-| --- | --- | --- | --- |
-| **Video-ORA-9B** | Qwen3.5-9B | `orarl_9b.yaml` | [Hugging Face](https://huggingface.co/OraRL/Video-ORA-9B) |
-| **Video-ORA-4B** | Qwen3.5-4B | `orarl_4b.yaml` | Hugging Face (coming soon) |
-
-### vLLM Serving
-
-Both Video-ORA checkpoints load directly with **vLLM 0.19.1** for
-OpenAI-compatible serving:
-
-```bash
-MODEL=OraRL/Video-ORA-9B
-
-vllm serve "$MODEL" \
- --served-model-name Video-ORA-9B \
- --trust-remote-code \
- --dtype bfloat16 \
- --tensor-parallel-size 1 \
- --max-model-len 131072 \
- --limit-mm-per-prompt '{"image": 1, "video": 1}'
-```
-
-Set `--tensor-parallel-size` to the GPU count for multi-GPU deployment and
-lower `--max-model-len` on smaller-memory devices. Use
-`enable_thinking=false` in the chat template for answer-only inference.
-
-## Use OraRL
-
-The release is organized around three user-facing workflows:
-
-1. **[Environment](docs/environment.md):** install the pinned CUDA stack that
- covers both the bundled trainer and the evaluators.
-2. **[Training](docs/training.md):** prepare licensed local training data and
- launch GRPO or OraRL on one or multiple nodes.
-3. **[Evaluation](docs/evaluation.md):** download Video-ORA and OraRL-Data,
- then run a smoke test or the complete paper suite.
-
-Training and evaluation are dry runs by default; inspect the resolved command
-before adding `--run`. Checkpoints and evaluation media are hosted under the
-[OraRL Hugging Face organization](https://huggingface.co/OraRL).
-
-## Acknowledgements
-
-OraRL is built on [veRL](https://github.com/volcengine/verl) — a
-high-performance RL framework with HybridEngine. We thank its authors and
-contributors for open-sourcing the training infrastructure.
-
-## License
-
-OraRL source is released under [Apache-2.0](LICENSE). Datasets, models,
-benchmarks, and optional dependencies retain their original licenses; see
-[NOTICE](NOTICE).
-
-## Citation
-
-If you find OraRL useful, please consider giving this repository a ⭐ and
-citing our [paper](https://arxiv.org/abs/2608.20492).
-
-```bibtex
-@article{li2026orarl,
- title = {Annotations as Rollouts: Efficient and Scalable
- Reinforcement Learning for Video MLLMs},
- author = {Li, Yunheng and Mu, Guohong and Li, Hao and
- Qian, Shengsheng and Zhang, Dingwen and Hou, Qibin
- and Cheng, Ming-Ming},
- journal = {arXiv preprint arXiv:2608.20492},
- year = {2026},
- url = {https://arxiv.org/abs/2608.20492}
-}
-```
diff --git a/code/orarl.egg-info/SOURCES.txt b/code/orarl.egg-info/SOURCES.txt
deleted file mode 100644
index 109aa7dc0e190f0ad7152e1ed4dc70e643a673d0..0000000000000000000000000000000000000000
--- a/code/orarl.egg-info/SOURCES.txt
+++ /dev/null
@@ -1,196 +0,0 @@
-LICENSE
-MANIFEST.in
-NOTICE
-README.md
-environment.yml
-pyproject.toml
-requirements-cu129.txt
-configs/data_sources.example.yaml
-configs/grpo_4b.yaml
-configs/grpo_9b.yaml
-configs/orarl_4b.yaml
-configs/orarl_9b.yaml
-data/eval/README.md
-data/eval/datasets.jsonl
-docs/environment.md
-docs/environment_zh.md
-docs/evaluation.md
-docs/evaluation_zh.md
-docs/training.md
-docs/training_zh.md
-eval/README.md
-eval/task/canonical_data.py
-eval/task/eval.sh
-eval/task/eval_image_mc_vllm.py
-eval/task/eval_prompt.py
-eval/task/eval_revsi_multinode.sh
-eval/task/eval_vllm.py
-eval/task/qwenvl_decord_patch.py
-eval/task/mindcube/data_utils.py
-eval/task/mmsi/eval_mmsi_transformers.py
-eval/task/mmsi/run_eval_transformers.sh
-eval/task/revsi/eval_revsi_vllm.py
-eval/task/revsi/merge_multinode_shards.py
-eval/task/revsi/run_eval_vllm.sh
-eval/task/segmentation/eval_seg_vllm.py
-eval/task/segmentation/post_sam2.py
-eval/task/segmentation/run_eval_vllm.sh
-eval/task/spatial_grounding/_grounding_utils.py
-eval/task/spatial_grounding/eval_refcoco_vllm.py
-eval/task/spatial_temporal_grounding/eval_stvg_vllm.py
-eval/task/temporal_grounding/eval_timelens_hf.py
-eval/task/temporal_grounding/run_eval.sh
-eval/task/tracking/eval_tracking_vllm.py
-orarl/__init__.py
-orarl/resources.py
-orarl.egg-info/PKG-INFO
-orarl.egg-info/SOURCES.txt
-orarl.egg-info/dependency_links.txt
-orarl.egg-info/entry_points.txt
-orarl.egg-info/requires.txt
-orarl.egg-info/top_level.txt
-orarl/algorithm/__init__.py
-orarl/algorithm/_utils.py
-orarl/algorithm/advantages.py
-orarl/algorithm/config.py
-orarl/algorithm/correction.py
-orarl/algorithm/selection.py
-orarl/cli/__init__.py
-orarl/cli/eval_data.py
-orarl/cli/evaluate.py
-orarl/cli/prepare.py
-orarl/cli/train.py
-orarl/data/__init__.py
-orarl/data/build.py
-orarl/data/identity.py
-orarl/data/schema.py
-orarl/evaluation/__init__.py
-orarl/evaluation/card.py
-orarl/evaluation/converters.py
-orarl/evaluation/hub.py
-orarl/evaluation/layout.py
-orarl/evaluation/manifest.py
-orarl/evaluation/schema.py
-orarl/evaluation/sources.py
-orarl/evaluation/staging.py
-orarl/rewards/__init__.py
-orarl/rewards/router.py
-orarl/rewards/types.py
-orarl/rewards/adapters/__init__.py
-orarl/rewards/adapters/_common.py
-orarl/rewards/adapters/segmentation.py
-orarl/rewards/adapters/spatial_grounding.py
-orarl/rewards/adapters/spatial_intelligence.py
-orarl/rewards/adapters/spatial_temporal_grounding.py
-orarl/rewards/adapters/temporal_grounding.py
-orarl/rewards/adapters/tracking.py
-orarl/rewards/adapters/video_qa.py
-scripts/check_environment.py
-scripts/check_release.py
-scripts/create_conda_env.sh
-scripts/create_eval_source_manifest.py
-scripts/evaluate.sh
-scripts/finalize_eval_index.py
-scripts/install_conda_runtime_hook.sh
-scripts/launch_multinode.sh
-scripts/model_merger.py
-scripts/organize_evaluation_outputs.py
-scripts/prepare_data.sh
-scripts/smoke_evaluation.sh
-scripts/smoke_training.sh
-scripts/train_grpo.sh
-scripts/train_orarl.sh
-tests/test_algorithm_advantages.py
-tests/test_algorithm_correction.py
-tests/test_algorithm_selection.py
-tests/test_builtin_reward_adapters.py
-tests/test_cli_evaluate.py
-tests/test_cli_metadata.py
-tests/test_cli_train.py
-tests/test_data_build.py
-tests/test_data_schema.py
-tests/test_environment_spec.py
-tests/test_eval_canonical_data.py
-tests/test_eval_source_discovery.py
-tests/test_evaluation_manifest.py
-tests/test_evaluation_release.py
-tests/test_evaluation_runtime.py
-tests/test_evaluation_schema.py
-tests/test_evaluation_staging.py
-tests/test_organize_evaluation_outputs.py
-tests/test_release_hygiene.py
-tests/test_resources.py
-tests/test_rewards.py
-tests/test_runtime_reward_entrypoint.py
-tests/test_training_jsonl_dataset.py
-tests/test_training_runtime.py
-verl/__init__.py
-verl/protocol.py
-verl/models/__init__.py
-verl/models/monkey_patch.py
-verl/models/transformers/__init__.py
-verl/models/transformers/flash_attention_utils.py
-verl/models/transformers/qwen2_vl.py
-verl/models/transformers/qwen3_5.py
-verl/models/transformers/qwen3_vl.py
-verl/single_controller/__init__.py
-verl/single_controller/base/__init__.py
-verl/single_controller/base/decorator.py
-verl/single_controller/base/worker.py
-verl/single_controller/base/worker_group.py
-verl/single_controller/base/register_center/__init__.py
-verl/single_controller/base/register_center/ray.py
-verl/single_controller/ray/__init__.py
-verl/single_controller/ray/base.py
-verl/trainer/__init__.py
-verl/trainer/config.py
-verl/trainer/core_algos.py
-verl/trainer/data_loader.py
-verl/trainer/main.py
-verl/trainer/metrics.py
-verl/trainer/orarl_config.py
-verl/trainer/orarl_post_selection.py
-verl/trainer/orarl_selection.py
-verl/trainer/ray_trainer.py
-verl/utils/__init__.py
-verl/utils/dataset.py
-verl/utils/flops_counter.py
-verl/utils/fsdp_utils.py
-verl/utils/model_utils.py
-verl/utils/multimodal_contract.py
-verl/utils/prompt_template.py
-verl/utils/py_functional.py
-verl/utils/seqlen_balancing.py
-verl/utils/tokenizer.py
-verl/utils/torch_dtypes.py
-verl/utils/torch_functional.py
-verl/utils/ulysses.py
-verl/utils/checkpoint/__init__.py
-verl/utils/checkpoint/checkpoint_manager.py
-verl/utils/checkpoint/fsdp_checkpoint_manager.py
-verl/utils/logger/__init__.py
-verl/utils/logger/gen_logger.py
-verl/utils/logger/logger.py
-verl/workers/__init__.py
-verl/workers/config.py
-verl/workers/fsdp_workers.py
-verl/workers/actor/__init__.py
-verl/workers/actor/base.py
-verl/workers/actor/config.py
-verl/workers/actor/dp_actor.py
-verl/workers/critic/__init__.py
-verl/workers/critic/base.py
-verl/workers/critic/config.py
-verl/workers/critic/dp_critic.py
-verl/workers/reward/__init__.py
-verl/workers/reward/config.py
-verl/workers/reward/function.py
-verl/workers/rollout/__init__.py
-verl/workers/rollout/base.py
-verl/workers/rollout/config.py
-verl/workers/rollout/hf_rollout.py
-verl/workers/rollout/vllm_rollout_spmd.py
-verl/workers/sharding_manager/__init__.py
-verl/workers/sharding_manager/base.py
-verl/workers/sharding_manager/fsdp_ulysses.py
-verl/workers/sharding_manager/fsdp_vllm.py
\ No newline at end of file
diff --git a/code/orarl.egg-info/dependency_links.txt b/code/orarl.egg-info/dependency_links.txt
deleted file mode 100644
index 8b137891791fe96927ad78e64b0aad7bded08bdc..0000000000000000000000000000000000000000
--- a/code/orarl.egg-info/dependency_links.txt
+++ /dev/null
@@ -1 +0,0 @@
-
diff --git a/code/orarl.egg-info/entry_points.txt b/code/orarl.egg-info/entry_points.txt
deleted file mode 100644
index 709973346314612295b0da8478f38837c2896cfc..0000000000000000000000000000000000000000
--- a/code/orarl.egg-info/entry_points.txt
+++ /dev/null
@@ -1,5 +0,0 @@
-[console_scripts]
-orarl-eval = orarl.cli.evaluate:main
-orarl-eval-data = orarl.cli.eval_data:main
-orarl-prepare = orarl.cli.prepare:main
-orarl-train = orarl.cli.train:main
diff --git a/code/orarl.egg-info/requires.txt b/code/orarl.egg-info/requires.txt
deleted file mode 100644
index 3a476f18c1915282ed840b17c37848c6a66ca26d..0000000000000000000000000000000000000000
--- a/code/orarl.egg-info/requires.txt
+++ /dev/null
@@ -1,12 +0,0 @@
-numpy
-PyYAML
-torch
-
-[hf]
-huggingface_hub
-
-[lint]
-ruff
-
-[test]
-pytest
diff --git a/code/orarl.egg-info/top_level.txt b/code/orarl.egg-info/top_level.txt
deleted file mode 100644
index 515595152e429980809f3de1df69091ae10204fe..0000000000000000000000000000000000000000
--- a/code/orarl.egg-info/top_level.txt
+++ /dev/null
@@ -1,2 +0,0 @@
-orarl
-verl
diff --git a/code/runs/smoke-training-9b-20260825_174304/smoke.log b/code/runs/smoke-training-9b-20260825_174304/smoke.log
deleted file mode 100644
index 572849e23e215d0ddbbea096ad47301531b07ba3..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_174304/smoke.log
+++ /dev/null
@@ -1,467 +0,0 @@
-OraRL root: /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official
-Python: /data/miniconda3/envs/orarl/bin/python
-Model: /apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B
-Train data: /mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_joint_100k_maskaware.jsonl
-Val data: /mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_phase3_49152.canary.jsonl
-Recipe scale: 9b
-World size: 1 x 8
-Steps per run: 1
-Results: /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_174304
-Execute: 1
-0, NVIDIA H20, 0 MiB, 97871 MiB
-1, NVIDIA H20, 0 MiB, 97871 MiB
-2, NVIDIA H20, 0 MiB, 97871 MiB
-3, NVIDIA H20, 0 MiB, 97871 MiB
-4, NVIDIA H20, 0 MiB, 97871 MiB
-5, NVIDIA H20, 0 MiB, 97871 MiB
-6, NVIDIA H20, 0 MiB, 97871 MiB
-7, NVIDIA H20, 0 MiB, 97871 MiB
-/data/miniconda3/envs/orarl/lib/python3.11/site-packages/transformers/utils/import_utils.py:1006: FutureWarning: `is_flash_attn_greater_or_equal_2_10` is deprecated and will be removed in v5.8. Please use `is_flash_attn_greater_or_equal(library_version='2.1.0')` instead if needed.
- warnings.warn(
-dependency OK: numpy 2.2.6
-dependency OK: torch 2.10.0+cu129
-dependency OK: transformers 5.5.4
-dependency OK: vllm 0.19.1
-dependency OK: ray 2.54.0
-dependency OK: tensordict 0.12.2
-dependency OK: codetiming 1.4.0
-dependency OK: omegaconf 2.3.0
-dependency OK: verl.trainer.main /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/verl/trainer/main.py
-dependency OK: orarl.rewards /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/orarl/rewards/__init__.py
-
-=== grpo smoke (1 update(s)) ===
-/data/miniconda3/envs/orarl/lib/python3.11/site-packages/transformers/utils/import_utils.py:1006: FutureWarning: `is_flash_attn_greater_or_equal_2_10` is deprecated and will be removed in v5.8. Please use `is_flash_attn_greater_or_equal(library_version='2.1.0')` instead if needed.
- warnings.warn(
-2026-08-25 17:43:40,058 INFO worker.py:2004 -- Started a local Ray instance. View the dashboard at [1m[32mhttp://127.0.0.1:8265 [39m[22m
-/data/miniconda3/envs/orarl/lib/python3.11/site-packages/ray/_private/worker.py:2052: FutureWarning: Tip: In future versions of Ray, Ray will no longer override accelerator visible devices env var if num_gpus=0 or num_gpus=None (default). To enable this behavior and turn off this error message, set RAY_ACCEL_ENV_VAR_OVERRIDE_ON_ZERO=0
- warnings.warn(
-[36m(pid=1944439)[0m W0825 17:43:56.478000 1944439 site-packages/torch/utils/cpp_extension.py:118] No CUDA runtime is found, using CUDA_HOME='/data/miniconda3/envs/orarl'
-[36m(Runner pid=1944439)[0m {
-[36m(Runner pid=1944439)[0m "data": {
-[36m(Runner pid=1944439)[0m "train_files": "/mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_joint_100k_maskaware.jsonl",
-[36m(Runner pid=1944439)[0m "val_files": "/mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_phase3_49152.canary.jsonl",
-[36m(Runner pid=1944439)[0m "prompt_key": "problem",
-[36m(Runner pid=1944439)[0m "answer_key": "answer",
-[36m(Runner pid=1944439)[0m "image_key": "images",
-[36m(Runner pid=1944439)[0m "video_key": "videos",
-[36m(Runner pid=1944439)[0m "image_dir": "/",
-[36m(Runner pid=1944439)[0m "video_fps": 2.0,
-[36m(Runner pid=1944439)[0m "video_max_frames": 128,
-[36m(Runner pid=1944439)[0m "max_prompt_length": 24576,
-[36m(Runner pid=1944439)[0m "max_response_length": 4096,
-[36m(Runner pid=1944439)[0m "rollout_batch_size": 8,
-[36m(Runner pid=1944439)[0m "mini_rollout_batch_size": null,
-[36m(Runner pid=1944439)[0m "val_batch_size": 16,
-[36m(Runner pid=1944439)[0m "format_prompt": null,
-[36m(Runner pid=1944439)[0m "override_chat_template": null,
-[36m(Runner pid=1944439)[0m "enable_thinking": false,
-[36m(Runner pid=1944439)[0m "response_prefix": "",
-[36m(Runner pid=1944439)[0m "shuffle": true,
-[36m(Runner pid=1944439)[0m "seed": 42,
-[36m(Runner pid=1944439)[0m "min_pixels": 4096,
-[36m(Runner pid=1944439)[0m "max_pixels": 4194304,
-[36m(Runner pid=1944439)[0m "image_min_pixels": 4096,
-[36m(Runner pid=1944439)[0m "image_max_pixels": 1048576,
-[36m(Runner pid=1944439)[0m "video_min_pixels": 4096,
-[36m(Runner pid=1944439)[0m "video_max_pixels": 786432,
-[36m(Runner pid=1944439)[0m "video_total_pixels": 8388608,
-[36m(Runner pid=1944439)[0m "val_video_fps": 2.0,
-[36m(Runner pid=1944439)[0m "val_video_max_frames": 128,
-[36m(Runner pid=1944439)[0m "val_video_min_pixels": 4096,
-[36m(Runner pid=1944439)[0m "val_video_max_pixels": 786432,
-[36m(Runner pid=1944439)[0m "val_video_total_pixels": 8388608,
-[36m(Runner pid=1944439)[0m "filter_overlong_prompts": false,
-[36m(Runner pid=1944439)[0m "filter_overlong_prompts_workers": 16,
-[36m(Runner pid=1944439)[0m "use_preprocessed_videos": false,
-[36m(Runner pid=1944439)[0m "video_source_mode": "realtime_only",
-[36m(Runner pid=1944439)[0m "preprocessed_video_dir": null,
-[36m(Runner pid=1944439)[0m "val_preprocessed_video_dir": null,
-[36m(Runner pid=1944439)[0m "val_video_source_mode": "realtime_only",
-[36m(Runner pid=1944439)[0m "inline_video_tensors": true,
-[36m(Runner pid=1944439)[0m "group_by_task": true,
-[36m(Runner pid=1944439)[0m "group_by_task_key": "problem_type",
-[36m(Runner pid=1944439)[0m "dataloader_num_workers": 1
-[36m(Runner pid=1944439)[0m },
-[36m(Runner pid=1944439)[0m "worker": {
-[36m(Runner pid=1944439)[0m "hybrid_engine": true,
-[36m(Runner pid=1944439)[0m "actor": {
-[36m(Runner pid=1944439)[0m "strategy": "fsdp",
-[36m(Runner pid=1944439)[0m "global_batch_size": 8,
-[36m(Runner pid=1944439)[0m "micro_batch_size_per_device_for_update": 1,
-[36m(Runner pid=1944439)[0m "micro_batch_size_per_device_for_experience": 1,
-[36m(Runner pid=1944439)[0m "max_grad_norm": 1.0,
-[36m(Runner pid=1944439)[0m "clip_ratio_low": 0.2,
-[36m(Runner pid=1944439)[0m "clip_ratio_high": 0.2,
-[36m(Runner pid=1944439)[0m "clip_ratio_dual": 3.0,
-[36m(Runner pid=1944439)[0m "loss_avg_mode": "token",
-[36m(Runner pid=1944439)[0m "loss_type": "default",
-[36m(Runner pid=1944439)[0m "ppo_epochs": 1,
-[36m(Runner pid=1944439)[0m "padding_free": true,
-[36m(Runner pid=1944439)[0m "dynamic_batching": true,
-[36m(Runner pid=1944439)[0m "max_token_len_per_gpu": 32768,
-[36m(Runner pid=1944439)[0m "ulysses_size": 1,
-[36m(Runner pid=1944439)[0m "use_torch_compile": false,
-[36m(Runner pid=1944439)[0m "model": {
-[36m(Runner pid=1944439)[0m "model_path": "/apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B",
-[36m(Runner pid=1944439)[0m "tokenizer_path": "/apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B",
-[36m(Runner pid=1944439)[0m "override_config": {},
-[36m(Runner pid=1944439)[0m "enable_gradient_checkpointing": true,
-[36m(Runner pid=1944439)[0m "trust_remote_code": false,
-[36m(Runner pid=1944439)[0m "freeze_vision_tower": true,
-[36m(Runner pid=1944439)[0m "train_vision_merger": false
-[36m(Runner pid=1944439)[0m },
-[36m(Runner pid=1944439)[0m "optim": {
-[36m(Runner pid=1944439)[0m "lr": 1e-06,
-[36m(Runner pid=1944439)[0m "betas": [
-[36m(Runner pid=1944439)[0m 0.9,
-[36m(Runner pid=1944439)[0m 0.999
-[36m(Runner pid=1944439)[0m ],
-[36m(Runner pid=1944439)[0m "weight_decay": 0.0,
-[36m(Runner pid=1944439)[0m "strategy": "adamw",
-[36m(Runner pid=1944439)[0m "lr_warmup_ratio": 0.0,
-[36m(Runner pid=1944439)[0m "lr_warmup_steps": null,
-[36m(Runner pid=1944439)[0m "min_lr_ratio": null,
-[36m(Runner pid=1944439)[0m "lr_scheduler_type": "constant",
-[36m(Runner pid=1944439)[0m "training_steps": -1
-[36m(Runner pid=1944439)[0m },
-[36m(Runner pid=1944439)[0m "fsdp": {
-[36m(Runner pid=1944439)[0m "enable_full_shard": true,
-[36m(Runner pid=1944439)[0m "enable_cpu_offload": false,
-[36m(Runner pid=1944439)[0m "enable_rank0_init": true,
-[36m(Runner pid=1944439)[0m "use_orig_params": false,
-[36m(Runner pid=1944439)[0m "torch_dtype": "bf16",
-[36m(Runner pid=1944439)[0m "fsdp_size": -1,
-[36m(Runner pid=1944439)[0m "mp_param_dtype": "bf16",
-[36m(Runner pid=1944439)[0m "mp_reduce_dtype": "fp32",
-[36m(Runner pid=1944439)[0m "mp_buffer_dtype": "fp32"
-[36m(Runner pid=1944439)[0m },
-[36m(Runner pid=1944439)[0m "offload": {
-[36m(Runner pid=1944439)[0m "offload_params": false,
-[36m(Runner pid=1944439)[0m "offload_optimizer": false
-[36m(Runner pid=1944439)[0m },
-[36m(Runner pid=1944439)[0m "global_batch_size_per_device": -1,
-[36m(Runner pid=1944439)[0m "disable_kl": true,
-[36m(Runner pid=1944439)[0m "use_kl_loss": false,
-[36m(Runner pid=1944439)[0m "kl_penalty": "low_var_kl",
-[36m(Runner pid=1944439)[0m "kl_coef": 0.0
-[36m(Runner pid=1944439)[0m },
-[36m(Runner pid=1944439)[0m "critic": {
-[36m(Runner pid=1944439)[0m "strategy": "fsdp",
-[36m(Runner pid=1944439)[0m "global_batch_size": 256,
-[36m(Runner pid=1944439)[0m "micro_batch_size_per_device_for_update": 4,
-[36m(Runner pid=1944439)[0m "micro_batch_size_per_device_for_experience": 16,
-[36m(Runner pid=1944439)[0m "max_grad_norm": 1.0,
-[36m(Runner pid=1944439)[0m "cliprange_value": 0.5,
-[36m(Runner pid=1944439)[0m "loss_avg_mode": "token",
-[36m(Runner pid=1944439)[0m "ppo_epochs": 1,
-[36m(Runner pid=1944439)[0m "padding_free": false,
-[36m(Runner pid=1944439)[0m "dynamic_batching": true,
-[36m(Runner pid=1944439)[0m "max_token_len_per_gpu": null,
-[36m(Runner pid=1944439)[0m "ulysses_size": 1,
-[36m(Runner pid=1944439)[0m "model": {
-[36m(Runner pid=1944439)[0m "model_path": null,
-[36m(Runner pid=1944439)[0m "tokenizer_path": null,
-[36m(Runner pid=1944439)[0m "override_config": {},
-[36m(Runner pid=1944439)[0m "enable_gradient_checkpointing": true,
-[36m(Runner pid=1944439)[0m "trust_remote_code": true,
-[36m(Runner pid=1944439)[0m "freeze_vision_tower": false,
-[36m(Runner pid=1944439)[0m "train_vision_merger": false
-[36m(Runner pid=1944439)[0m },
-[36m(Runner pid=1944439)[0m "optim": {
-[36m(Runner pid=1944439)[0m "lr": 1e-06,
-[36m(Runner pid=1944439)[0m "betas": [
-[36m(Runner pid=1944439)[0m 0.9,[36m(Runner pid=1944439)[0m The `use_fast` parameter is deprecated and will be removed in a future version. Use `backend="torchvision"` instead of `use_fast=True`, or `backend="pil"` instead of `use_fast=False`.
-[36m(Runner pid=1944439)[0m
Generating train split: 0 examples [00:00, ? examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 2684 examples [00:00, 5741.64 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 3894 examples [00:00, 5566.39 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 6050 examples [00:01, 5916.34 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 7310 examples [00:01, 5917.78 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 8652 examples [00:01, 7062.55 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 10017 examples [00:01, 5373.29 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 11061 examples [00:01, 6122.95 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 12345 examples [00:01, 6764.91 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 13564 examples [00:02, 4332.70 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 14792 examples [00:02, 5251.29 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 15958 examples [00:02, 5219.09 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 17208 examples [00:03, 5223.77 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 18484 examples [00:03, 6319.60 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 19610 examples [00:03, 5646.33 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 20953 examples [00:03, 6782.24 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 22170 examples [00:03, 7115.81 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 23412 examples [00:03, 6220.19 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 24542 examples [00:04, 7068.41 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 25620 examples [00:04, 5887.92 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 27027 examples [00:05, 2384.25 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 29107 examples [00:06, 2709.31 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 30179 examples [00:06, 2852.90 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 32832 examples [00:06, 4368.47 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 35008 examples [00:06, 5689.84 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 38278 examples [00:07, 7779.78 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 41012 examples [00:07, 9843.69 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 43744 examples [00:07, 10491.50 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 46312 examples [00:07, 12026.91 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 49391 examples [00:08, 8630.42 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 52225 examples [00:08, 10551.99 examples/s]
-[36m(Runner pid=1944439)[0m
Generating train split: 52225 examples [00:08, 6076.51 examples/s]
-Traceback (most recent call last):
- File "", line 198, in _run_module_as_main
- File "", line 88, in _run_code
- File "/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/verl/trainer/main.py", line 137, in
- main()
- File "/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/verl/trainer/main.py", line 129, in main
- ray.get(runner.run.remote(ppo_config))
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/ray/_private/auto_init_hook.py", line 22, in auto_init_wrapper
- return fn(*args, **kwargs)
- ^^^^^^^^^^^^^^^^^^^
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/ray/_private/client_mode_hook.py", line 104, in wrapper
- return func(*args, **kwargs)
- ^^^^^^^^^^^^^^^^^^^^^
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/ray/_private/worker.py", line 2981, in get
- values, debugger_breakpoint = worker.get_objects(
- ^^^^^^^^^^^^^^^^^^^
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/ray/_private/worker.py", line 1012, in get_objects
- raise value.as_instanceof_cause()
-ray.exceptions.RayTaskError(DatasetGenerationError): [36mray::Runner.run()[39m (pid=1944439, ip=29.127.69.4, actor_id=e981ae3ef4ecbc78705cfeea01000000, repr=)
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/datasets/packaged_modules/json/json.py", line 265, in _generate_tables
- self._cast_table(pa_table, json_field_paths=json_field_paths),
- ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/datasets/packaged_modules/json/json.py", line 120, in _cast_table
- pa_table = table_cast(pa_table, self.info.features.arrow_schema)
- ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/datasets/table.py", line 2272, in table_cast
- return cast_table_to_schema(table, schema)
- ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/datasets/table.py", line 2218, in cast_table_to_schema
- raise CastError(
-datasets.table.CastError: Couldn't cast
-problem: string
-answer: string
-ground_truth: string
-problem_type: string
-videos: list
- child 0, item: string
-data_type: string
-task_source: string
-segmentation_output: string
-meta: string
-images: list
- child 0, item: string
-fps: int64
-video_second: double
-resolution: struct
- child 0, width: int64
- child 1, height: int64
-path: string
-data_source: string
-messages: list>
- child 0, item: struct
- child 0, role: string
- child 1, content: string
-solution: string
-difficulty: double
-pred_iou: double
-pred_span: list
- child 0, item: double
-pred_text: string
-checkpoint_score: double
-hardness_band: string
-selection_role: string
-selection_checkpoint: string
-identity_key: string
-media_anchors: list
- child 0, item: string
-selection_temporal_bucket: string
-question_type: string
-source_file: string
-source_id: string
-selection_weak_buckets: list
- child 0, item: string
-selection_bucket: string
-source_block: string
-selection_scores: struct)
- ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
- ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
- File "/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/verl/trainer/main.py", line 74, in run
- train_dataloader, val_dataloader = create_dataloader(
- ^^^^^^^^^^^^^^^^^^
- File "/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/verl/trainer/data_loader.py", line 41, in create_dataloader
- train_dataset = RLHFDataset(
- ^^^^^^^^^^^^
- File "/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/verl/utils/dataset.py", line 444, in __init__
- self.dataset = load_dataset(file_type, data_files=data_path, split=data_split)
- ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/datasets/load.py", line 1708, in load_dataset
- builder_instance.download_and_prepare(
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/datasets/builder.py", line 884, in download_and_prepare
- self._download_and_prepare(
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/datasets/builder.py", line 947, in _download_and_prepare
- self._prepare_split(split_generator, **prepare_split_kwargs)
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/datasets/builder.py", line 1739, in _prepare_split
- for job_id, done, content in self._prepare_split_single(
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/datasets/builder.py", line 1922, in _prepare_split_single
- raise DatasetGenerationError("An error occurred while generating the dataset") from e
-datasets.exceptions.DatasetGenerationError: An error occurred while generating the dataset
-
-[36m(Runner pid=1944439)[0m 0.999
-[36m(Runner pid=1944439)[0m ],
-[36m(Runner pid=1944439)[0m "weight_decay": 0.01,
-[36m(Runner pid=1944439)[0m "strategy": "adamw",
-[36m(Runner pid=1944439)[0m "lr_warmup_ratio": 0.0,
-[36m(Runner pid=1944439)[0m "lr_warmup_steps": null,
-[36m(Runner pid=1944439)[0m "min_lr_ratio": null,
-[36m(Runner pid=1944439)[0m "lr_scheduler_type": "constant",
-[36m(Runner pid=1944439)[0m "training_steps": -1
-[36m(Runner pid=1944439)[0m },
-[36m(Runner pid=1944439)[0m "fsdp": {
-[36m(Runner pid=1944439)[0m "enable_full_shard": true,
-[36m(Runner pid=1944439)[0m "enable_cpu_offload": false,
-[36m(Runner pid=1944439)[0m "enable_rank0_init": true,
-[36m(Runner pid=1944439)[0m "use_orig_params": false,
-[36m(Runner pid=1944439)[0m "torch_dtype": null,
-[36m(Runner pid=1944439)[0m "fsdp_size": -1,
-[36m(Runner pid=1944439)[0m "mp_param_dtype": "bf16",
-[36m(Runner pid=1944439)[0m "mp_reduce_dtype": "fp32",
-[36m(Runner pid=1944439)[0m "mp_buffer_dtype": "fp32"
-[36m(Runner pid=1944439)[0m },
-[36m(Runner pid=1944439)[0m "offload": {
-[36m(Runner pid=1944439)[0m "offload_params": false,
-[36m(Runner pid=1944439)[0m "offload_optimizer": false
-[36m(Runner pid=1944439)[0m },
-[36m(Runner pid=1944439)[0m "global_batch_size_per_device": -1
-[36m(Runner pid=1944439)[0m },
-[36m(Runner pid=1944439)[0m "ref": {
-[36m(Runner pid=1944439)[0m "strategy": "fsdp",
-[36m(Runner pid=1944439)[0m "fsdp": {
-[36m(Runner pid=1944439)[0m "enable_full_shard": true,
-[36m(Runner pid=1944439)[0m "enable_cpu_offload": false,
-[36m(Runner pid=1944439)[0m "enable_rank0_init": true,
-[36m(Runner pid=1944439)[0m "use_orig_params": false,
-[36m(Runner pid=1944439)[0m "torch_dtype": "bf16",
-[36m(Runner pid=1944439)[0m "fsdp_size": -1,
-[36m(Runner pid=1944439)[0m "mp_param_dtype": "bf16",
-[36m(Runner pid=1944439)[0m "mp_reduce_dtype": "fp32",
-[36m(Runner pid=1944439)[0m "mp_buffer_dtype": "fp32"
-[36m(Runner pid=1944439)[0m },
-[36m(Runner pid=1944439)[0m "offload": {
-[36m(Runner pid=1944439)[0m "offload_params": false,
-[36m(Runner pid=1944439)[0m "offload_optimizer": false
-[36m(Runner pid=1944439)[0m },
-[36m(Runner pid=1944439)[0m "micro_batch_size_per_device_for_experience": 1,
-[36m(Runner pid=1944439)[0m "padding_free": true,
-[36m(Runner pid=1944439)[0m "dynamic_batching": true,
-[36m(Runner pid=1944439)[0m "max_token_len_per_gpu": 32768,
-[36m(Runner pid=1944439)[0m "ulysses_size": 1,
-[36m(Runner pid=1944439)[0m "use_torch_compile": false
-[36m(Runner pid=1944439)[0m },
-[36m(Runner pid=1944439)[0m "reward": {
-[36m(Runner pid=1944439)[0m "reward_function": "orarl.rewards",
-[36m(Runner pid=1944439)[0m "reward_function_kwargs": {},
-[36m(Runner pid=1944439)[0m "skip_special_tokens": true,
-[36m(Runner pid=1944439)[0m "num_cpus": 1,
-[36m(Runner pid=1944439)[0m "reward_function_name": "compute_score",
-[36m(Runner pid=1944439)[0m "reward_function_is_module": true
-[36m(Runner pid=1944439)[0m },
-[36m(Runner pid=1944439)[0m "rollout": {
-[36m(Runner pid=1944439)[0m "name": "vllm",
-[36m(Runner pid=1944439)[0m "n": 8,
-[36m(Runner pid=1944439)[0m "temperature": 1.0,
-[36m(Runner pid=1944439)[0m "top_p": 0.85,
-[36m(Runner pid=1944439)[0m "top_k": -1,
-[36m(Runner pid=1944439)[0m "seed": 42,
-[36m(Runner pid=1944439)[0m "limit_images": 10,
-[36m(Runner pid=1944439)[0m "dtype": "bf16",
-[36m(Runner pid=1944439)[0m "gpu_memory_utilization": 0.45,
-[36m(Runner pid=1944439)[0m "ignore_eos": false,
-[36m(Runner pid=1944439)[0m "enforce_eager": false,
-[36m(Runner pid=1944439)[0m "enable_chunked_prefill": true,
-[36m(Runner pid=1944439)[0m "tensor_parallel_size": 1,
-[36m(Runner pid=1944439)[0m "max_model_len": null,
-[36m(Runner pid=1944439)[0m "max_num_batched_tokens": 32768,
-[36m(Runner pid=1944439)[0m "disable_log_stats": true,
-[36m(Runner pid=1944439)[0m "disable_tqdm": true,
-[36m(Runner pid=1944439)[0m "val_override_config": {
-[36m(Runner pid=1944439)[0m "temperature": 0.7,
-[36m(Runner pid=1944439)[0m "top_p": 0.95,
-[36m(Runner pid=1944439)[0m "n": 1
-[36m(Runner pid=1944439)[0m },
-[36m(Runner pid=1944439)[0m "kv_cache_dtype": "auto",
-[36m(Runner pid=1944439)[0m "calculate_log_probs": false,
-[36m(Runner pid=1944439)[0m "collect_seq_logprob_for_filter": false,
-[36m(Runner pid=1944439)[0m "prompt_length": 24576,
-[36m(Runner pid=1944439)[0m "response_length": 4096,
-[36m(Runner pid=1944439)[0m "trust_remote_code": false
-[36m(Runner pid=1944439)[0m }
-[36m(Runner pid=1944439)[0m },
-[36m(Runner pid=1944439)[0m "algorithm": {
-[36m(Runner pid=1944439)[0m "name": "grpo",
-[36m(Runner pid=1944439)[0m "gamma": 1.0,
-[36m(Runner pid=1944439)[0m "lam": 1.0,
-[36m(Runner pid=1944439)[0m "adv_estimator": "grpo",
-[36m(Runner pid=1944439)[0m "scale_rewards": true,
-[36m(Runner pid=1944439)[0m "disable_kl": true,
-[36m(Runner pid=1944439)[0m "use_kl_loss": false,
-[36m(Runner pid=1944439)[0m "kl_penalty": "low_var_kl",
-[36m(Runner pid=1944439)[0m "kl_coef": 0.0,
-[36m(Runner pid=1944439)[0m "kl_type": "fixed",
-[36m(Runner pid=1944439)[0m "kl_horizon": 10000.0,
-[36m(Runner pid=1944439)[0m "kl_target": 0.1
-[36m(Runner pid=1944439)[0m },
-[36m(Runner pid=1944439)[0m "trainer": {
-[36m(Runner pid=1944439)[0m "total_epochs": 1,
-[36m(Runner pid=1944439)[0m "max_steps": 1,
-[36m(Runner pid=1944439)[0m "project_name": "OraRL",
-[36m(Runner pid=1944439)[0m "experiment_name": "smoke-grpo-9b",
-[36m(Runner pid=1944439)[0m "logger": [
-[36m(Runner pid=1944439)[0m "console"
-[36m(Runner pid=1944439)[0m ],
-[36m(Runner pid=1944439)[0m "nnodes": 1,
-[36m(Runner pid=1944439)[0m "n_gpus_per_node": 8,
-[36m(Runner pid=1944439)[0m "critic_warmup": 0,
-[36m(Runner pid=1944439)[0m "val_freq": -1,
-[36m(Runner pid=1944439)[0m "val_before_train": false,
-[36m(Runner pid=1944439)[0m "val_only": false,
-[36m(Runner pid=1944439)[0m "val_generations_to_log": 0,
-[36m(Runner pid=1944439)[0m "save_freq": 1,
-[36m(Runner pid=1944439)[0m "save_limit": 3,
-[36m(Runner pid=1944439)[0m "save_model_only": false,
-[36m(Runner pid=1944439)[0m "keep_optim_only_latest": true,
-[36m(Runner pid=1944439)[0m "save_checkpoint_path": "/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_174304/grpo",
-[36m(Runner pid=1944439)[0m "load_checkpoint_path": null,
-[36m(Runner pid=1944439)[0m "ray_timeline": null,
-[36m(Runner pid=1944439)[0m "find_last_checkpoint": false,
-[36m(Runner pid=1944439)[0m "keep_best_train_ckpt": false,
-[36m(Runner pid=1944439)[0m "best_train_metric_key": "reward/overall",
-[36m(Runner pid=1944439)[0m "best_train_smooth_window": 5,
-[36m(Runner pid=1944439)[0m "best_train_min_step": 10
-[36m(Runner pid=1944439)[0m }
-[36m(Runner pid=1944439)[0m }
-[36m(AutoRewardManager pid=1946036)[0m Using reward function `compute_score` from `orarl.rewards`.
-[36m(AutoRewardManager pid=1946036)[0m Reward name: orarl, reward type: batch.
-[36m(AutoRewardManager pid=1946038)[0m Using reward function `compute_score` from `orarl.rewards`.
-[36m(AutoRewardManager pid=1946038)[0m Reward name: orarl, reward type: batch.
-method=grpo mode=run
-/data/miniconda3/envs/orarl/bin/python -m verl.trainer.main config=/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/configs/grpo_9b.yaml trainer.max_steps=1 trainer.save_freq=1 trainer.val_before_train=false trainer.experiment_name=smoke-grpo-9b data.rollout_batch_size=8 worker.actor.global_batch_size=8 worker.actor.model.model_path=/apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B worker.actor.model.tokenizer_path=/apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B data.train_files=/mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_joint_100k_maskaware.jsonl data.val_files=/mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_phase3_49152.canary.jsonl trainer.save_checkpoint_path=/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_174304/grpo trainer.nnodes=1 trainer.n_gpus_per_node=8
diff --git a/code/runs/smoke-training-9b-20260825_175213/smoke.log b/code/runs/smoke-training-9b-20260825_175213/smoke.log
deleted file mode 100644
index ec43ca43443fafe8272b4cd9c86b41fc634ccb26..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175213/smoke.log
+++ /dev/null
@@ -1,827 +0,0 @@
-OraRL root: /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official
-Python: /data/miniconda3/envs/orarl/bin/python
-Model: /apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B
-Train data: /mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_joint_100k_maskaware.jsonl
-Val data: /mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_phase3_49152.canary.jsonl
-Recipe scale: 9b
-World size: 1 x 8
-Steps per run: 1
-Results: /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175213
-Execute: 1
-0, NVIDIA H20, 0 MiB, 97871 MiB
-1, NVIDIA H20, 0 MiB, 97871 MiB
-2, NVIDIA H20, 0 MiB, 97871 MiB
-3, NVIDIA H20, 0 MiB, 97871 MiB
-4, NVIDIA H20, 0 MiB, 97871 MiB
-5, NVIDIA H20, 0 MiB, 97871 MiB
-6, NVIDIA H20, 0 MiB, 97871 MiB
-7, NVIDIA H20, 0 MiB, 97871 MiB
-/data/miniconda3/envs/orarl/lib/python3.11/site-packages/transformers/utils/import_utils.py:1006: FutureWarning: `is_flash_attn_greater_or_equal_2_10` is deprecated and will be removed in v5.8. Please use `is_flash_attn_greater_or_equal(library_version='2.1.0')` instead if needed.
- warnings.warn(
-dependency OK: numpy 2.2.6
-dependency OK: torch 2.10.0+cu129
-dependency OK: transformers 5.5.4
-dependency OK: vllm 0.19.1
-dependency OK: ray 2.54.0
-dependency OK: tensordict 0.12.2
-dependency OK: codetiming 1.4.0
-dependency OK: omegaconf 2.3.0
-dependency OK: verl.trainer.main /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/verl/trainer/main.py
-dependency OK: orarl.rewards /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/orarl/rewards/__init__.py
-
-=== grpo smoke (1 update(s)) ===
-/data/miniconda3/envs/orarl/lib/python3.11/site-packages/transformers/utils/import_utils.py:1006: FutureWarning: `is_flash_attn_greater_or_equal_2_10` is deprecated and will be removed in v5.8. Please use `is_flash_attn_greater_or_equal(library_version='2.1.0')` instead if needed.
- warnings.warn(
-2026-08-25 17:52:41,284 INFO worker.py:2004 -- Started a local Ray instance. View the dashboard at [1m[32mhttp://127.0.0.1:8265 [39m[22m
-/data/miniconda3/envs/orarl/lib/python3.11/site-packages/ray/_private/worker.py:2052: FutureWarning: Tip: In future versions of Ray, Ray will no longer override accelerator visible devices env var if num_gpus=0 or num_gpus=None (default). To enable this behavior and turn off this error message, set RAY_ACCEL_ENV_VAR_OVERRIDE_ON_ZERO=0
- warnings.warn(
-[36m(pid=1984436)[0m W0825 17:52:57.110000 1984436 site-packages/torch/utils/cpp_extension.py:118] No CUDA runtime is found, using CUDA_HOME='/data/miniconda3/envs/orarl'
-[36m(Runner pid=1984436)[0m {
-[36m(Runner pid=1984436)[0m "data": {
-[36m(Runner pid=1984436)[0m "train_files": "/mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_joint_100k_maskaware.jsonl",
-[36m(Runner pid=1984436)[0m "val_files": "/mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_phase3_49152.canary.jsonl",
-[36m(Runner pid=1984436)[0m "prompt_key": "problem",
-[36m(Runner pid=1984436)[0m "answer_key": "answer",
-[36m(Runner pid=1984436)[0m "image_key": "images",
-[36m(Runner pid=1984436)[0m "video_key": "videos",
-[36m(Runner pid=1984436)[0m "image_dir": "/",
-[36m(Runner pid=1984436)[0m "video_fps": 2.0,
-[36m(Runner pid=1984436)[0m "video_max_frames": 128,
-[36m(Runner pid=1984436)[0m "max_prompt_length": 24576,
-[36m(Runner pid=1984436)[0m "max_response_length": 4096,
-[36m(Runner pid=1984436)[0m "rollout_batch_size": 8,
-[36m(Runner pid=1984436)[0m "mini_rollout_batch_size": null,
-[36m(Runner pid=1984436)[0m "val_batch_size": 16,
-[36m(Runner pid=1984436)[0m "format_prompt": null,
-[36m(Runner pid=1984436)[0m "override_chat_template": null,
-[36m(Runner pid=1984436)[0m "enable_thinking": false,
-[36m(Runner pid=1984436)[0m "response_prefix": "",
-[36m(Runner pid=1984436)[0m "shuffle": true,
-[36m(Runner pid=1984436)[0m "seed": 42,
-[36m(Runner pid=1984436)[0m "min_pixels": 4096,
-[36m(Runner pid=1984436)[0m "max_pixels": 4194304,
-[36m(Runner pid=1984436)[0m "image_min_pixels": 4096,
-[36m(Runner pid=1984436)[0m "image_max_pixels": 1048576,
-[36m(Runner pid=1984436)[0m "video_min_pixels": 4096,
-[36m(Runner pid=1984436)[0m "video_max_pixels": 786432,
-[36m(Runner pid=1984436)[0m "video_total_pixels": 8388608,
-[36m(Runner pid=1984436)[0m "val_video_fps": 2.0,
-[36m(Runner pid=1984436)[0m "val_video_max_frames": 128,
-[36m(Runner pid=1984436)[0m "val_video_min_pixels": 4096,
-[36m(Runner pid=1984436)[0m "val_video_max_pixels": 786432,
-[36m(Runner pid=1984436)[0m "val_video_total_pixels": 8388608,
-[36m(Runner pid=1984436)[0m "filter_overlong_prompts": false,
-[36m(Runner pid=1984436)[0m "filter_overlong_prompts_workers": 16,
-[36m(Runner pid=1984436)[0m "use_preprocessed_videos": false,
-[36m(Runner pid=1984436)[0m "video_source_mode": "realtime_only",
-[36m(Runner pid=1984436)[0m "preprocessed_video_dir": null,
-[36m(Runner pid=1984436)[0m "val_preprocessed_video_dir": null,
-[36m(Runner pid=1984436)[0m "val_video_source_mode": "realtime_only",
-[36m(Runner pid=1984436)[0m "inline_video_tensors": true,
-[36m(Runner pid=1984436)[0m "group_by_task": true,
-[36m(Runner pid=1984436)[0m "group_by_task_key": "problem_type",
-[36m(Runner pid=1984436)[0m "dataloader_num_workers": 1
-[36m(Runner pid=1984436)[0m },
-[36m(Runner pid=1984436)[0m "worker": {
-[36m(Runner pid=1984436)[0m "hybrid_engine": true,
-[36m(Runner pid=1984436)[0m "actor": {
-[36m(Runner pid=1984436)[0m "strategy": "fsdp",
-[36m(Runner pid=1984436)[0m "global_batch_size": 8,
-[36m(Runner pid=1984436)[0m "micro_batch_size_per_device_for_update": 1,
-[36m(Runner pid=1984436)[0m "micro_batch_size_per_device_for_experience": 1,
-[36m(Runner pid=1984436)[0m "max_grad_norm": 1.0,
-[36m(Runner pid=1984436)[0m "clip_ratio_low": 0.2,
-[36m(Runner pid=1984436)[0m "clip_ratio_high": 0.2,
-[36m(Runner pid=1984436)[0m "clip_ratio_dual": 3.0,
-[36m(Runner pid=1984436)[0m "loss_avg_mode": "token",
-[36m(Runner pid=1984436)[0m "loss_type": "default",
-[36m(Runner pid=1984436)[0m "ppo_epochs": 1,
-[36m(Runner pid=1984436)[0m "padding_free": true,
-[36m(Runner pid=1984436)[0m "dynamic_batching": true,
-[36m(Runner pid=1984436)[0m "max_token_len_per_gpu": 32768,
-[36m(Runner pid=1984436)[0m "ulysses_size": 1,
-[36m(Runner pid=1984436)[0m "use_torch_compile": false,
-[36m(Runner pid=1984436)[0m "model": {
-[36m(Runner pid=1984436)[0m "model_path": "/apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B",
-[36m(Runner pid=1984436)[0m "tokenizer_path": "/apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B",
-[36m(Runner pid=1984436)[0m "override_config": {},
-[36m(Runner pid=1984436)[0m "enable_gradient_checkpointing": true,
-[36m(Runner pid=1984436)[0m "trust_remote_code": false,
-[36m(Runner pid=1984436)[0m "freeze_vision_tower": true,
-[36m(Runner pid=1984436)[0m "train_vision_merger": false
-[36m(Runner pid=1984436)[0m },
-[36m(Runner pid=1984436)[0m "optim": {
-[36m(Runner pid=1984436)[0m "lr": 1e-06,
-[36m(Runner pid=1984436)[0m "betas": [
-[36m(Runner pid=1984436)[0m 0.9,
-[36m(Runner pid=1984436)[0m 0.999
-[36m(Runner pid=1984436)[0m ],
-[36m(Runner pid=1984436)[0m "weight_decay": 0.0,
-[36m(Runner pid=1984436)[0m "strategy": "adamw",
-[36m(Runner pid=1984436)[0m "lr_warmup_ratio": 0.0,
-[36m(Runner pid=1984436)[0m "lr_warmup_steps": null,
-[36m(Runner pid=1984436)[0m "min_lr_ratio": null,
-[36m(Runner pid=1984436)[0m "lr_scheduler_type": "constant",
-[36m(Runner pid=1984436)[0m "training_steps": -1
-[36m(Runner pid=1984436)[0m },
-[36m(Runner pid=1984436)[0m "fsdp": {
-[36m(Runner pid=1984436)[0m "enable_full_shard": true,
-[36m(Runner pid=1984436)[0m "enable_cpu_offload": false,
-[36m(Runner pid=1984436)[0m "enable_rank0_init": true,
-[36m(Runner pid=1984436)[0m "use_orig_params": false,
-[36m(Runner pid=1984436)[0m "torch_dtype": "bf16",
-[36m(Runner pid=1984436)[0m "fsdp_size": -1,
-[36m(Runner pid=1984436)[0m "mp_param_dtype": "bf16",
-[36m(Runner pid=1984436)[0m "mp_reduce_dtype": "fp32",
-[36m(Runner pid=1984436)[0m "mp_buffer_dtype": "fp32"
-[36m(Runner pid=1984436)[0m },
-[36m(Runner pid=1984436)[0m "offload": {
-[36m(Runner pid=1984436)[0m "offload_params": false,
-[36m(Runner pid=1984436)[0m "offload_optimizer": false
-[36m(Runner pid=1984436)[0m },
-[36m(Runner pid=1984436)[0m "global_batch_size_per_device": -1,
-[36m(Runner pid=1984436)[0m "disable_kl": true,
-[36m(Runner pid=1984436)[0m "use_kl_loss": false,
-[36m(Runner pid=1984436)[0m "kl_penalty": "low_var_kl",
-[36m(Runner pid=1984436)[0m "kl_coef": 0.0
-[36m(Runner pid=1984436)[0m },
-[36m(Runner pid=1984436)[0m "critic": {
-[36m(Runner pid=1984436)[0m "strategy": "fsdp",
-[36m(Runner pid=1984436)[0m "global_batch_size": 256,
-[36m(Runner pid=1984436)[0m "micro_batch_size_per_device_for_update": 4,
-[36m(Runner pid=1984436)[0m "micro_batch_size_per_device_for_experience": 16,
-[36m(Runner pid=1984436)[0m "max_grad_norm": 1.0,
-[36m(Runner pid=1984436)[0m "cliprange_value": 0.5,
-[36m(Runner pid=1984436)[0m "loss_avg_mode": "token",
-[36m(Runner pid=1984436)[0m "ppo_epochs": 1,
-[36m(Runner pid=1984436)[0m "padding_free": false,
-[36m(Runner pid=1984436)[0m "dynamic_batching": true,
-[36m(Runner pid=1984436)[0m "max_token_len_per_gpu": null,
-[36m(Runner pid=1984436)[0m "ulysses_size": 1,
-[36m(Runner pid=1984436)[0m "model": {
-[36m(Runner pid=1984436)[0m "model_path": null,
-[36m(Runner pid=1984436)[0m "tokenizer_path": null,
-[36m(Runner pid=1984436)[0m "override_config": {},
-[36m(Runner pid=1984436)[0m "enable_gradient_checkpointing": true,
-[36m(Runner pid=1984436)[0m "trust_remote_code": true,
-[36m(Runner pid=1984436)[0m "freeze_vision_tower": false,
-[36m(Runner pid=1984436)[0m "train_vision_merger": false
-[36m(Runner pid=1984436)[0m },
-[36m(Runner pid=1984436)[0m "optim": {
-[36m(Runner pid=1984436)[0m "lr": 1e-06,
-[36m(Runner pid=1984436)[0m "betas": [
-[36m(Runner pid=1984436)[0m 0.9,[36m(Runner pid=1984436)[0m The `use_fast` parameter is deprecated and will be removed in a future version. Use `backend="torchvision"` instead of `use_fast=True`, or `backend="pil"` instead of `use_fast=False`.
-
-[36m(Runner pid=1984436)[0m 0.999
-[36m(Runner pid=1984436)[0m ],
-[36m(Runner pid=1984436)[0m "weight_decay": 0.01,
-[36m(Runner pid=1984436)[0m "strategy": "adamw",
-[36m(Runner pid=1984436)[0m "lr_warmup_ratio": 0.0,
-[36m(Runner pid=1984436)[0m "lr_warmup_steps": null,
-[36m(Runner pid=1984436)[0m "min_lr_ratio": null,
-[36m(Runner pid=1984436)[0m "lr_scheduler_type": "constant",
-[36m(Runner pid=1984436)[0m "training_steps": -1
-[36m(Runner pid=1984436)[0m },
-[36m(Runner pid=1984436)[0m "fsdp": {
-[36m(Runner pid=1984436)[0m "enable_full_shard": true,
-[36m(Runner pid=1984436)[0m "enable_cpu_offload": false,
-[36m(Runner pid=1984436)[0m "enable_rank0_init": true,
-[36m(Runner pid=1984436)[0m "use_orig_params": false,
-[36m(Runner pid=1984436)[0m "torch_dtype": null,
-[36m(Runner pid=1984436)[0m "fsdp_size": -1,
-[36m(Runner pid=1984436)[0m "mp_param_dtype": "bf16",
-[36m(Runner pid=1984436)[0m "mp_reduce_dtype": "fp32",
-[36m(Runner pid=1984436)[0m "mp_buffer_dtype": "fp32"
-[36m(Runner pid=1984436)[0m },
-[36m(Runner pid=1984436)[0m "offload": {
-[36m(Runner pid=1984436)[0m "offload_params": false,
-[36m(Runner pid=1984436)[0m "offload_optimizer": false
-[36m(Runner pid=1984436)[0m },
-[36m(Runner pid=1984436)[0m "global_batch_size_per_device": -1
-[36m(Runner pid=1984436)[0m },
-[36m(Runner pid=1984436)[0m "ref": {
-[36m(Runner pid=1984436)[0m "strategy": "fsdp",
-[36m(Runner pid=1984436)[0m "fsdp": {
-[36m(Runner pid=1984436)[0m "enable_full_shard": true,
-[36m(Runner pid=1984436)[0m "enable_cpu_offload": false,
-[36m(Runner pid=1984436)[0m "enable_rank0_init": true,
-[36m(Runner pid=1984436)[0m "use_orig_params": false,
-[36m(Runner pid=1984436)[0m "torch_dtype": "bf16",
-[36m(Runner pid=1984436)[0m "fsdp_size": -1,
-[36m(Runner pid=1984436)[0m "mp_param_dtype": "bf16",
-[36m(Runner pid=1984436)[0m "mp_reduce_dtype": "fp32",
-[36m(Runner pid=1984436)[0m "mp_buffer_dtype": "fp32"
-[36m(Runner pid=1984436)[0m },
-[36m(Runner pid=1984436)[0m "offload": {
-[36m(Runner pid=1984436)[0m "offload_params": false,
-[36m(Runner pid=1984436)[0m "offload_optimizer": false
-[36m(Runner pid=1984436)[0m },
-[36m(Runner pid=1984436)[0m "micro_batch_size_per_device_for_experience": 1,
-[36m(Runner pid=1984436)[0m "padding_free": true,
-[36m(Runner pid=1984436)[0m "dynamic_batching": true,
-[36m(Runner pid=1984436)[0m "max_token_len_per_gpu": 32768,
-[36m(Runner pid=1984436)[0m "ulysses_size": 1,
-[36m(Runner pid=1984436)[0m "use_torch_compile": false
-[36m(Runner pid=1984436)[0m },
-[36m(Runner pid=1984436)[0m "reward": {
-[36m(Runner pid=1984436)[0m "reward_function": "orarl.rewards",
-[36m(Runner pid=1984436)[0m "reward_function_kwargs": {},
-[36m(Runner pid=1984436)[0m "skip_special_tokens": true,
-[36m(Runner pid=1984436)[0m "num_cpus": 1,
-[36m(Runner pid=1984436)[0m "reward_function_name": "compute_score",
-[36m(Runner pid=1984436)[0m "reward_function_is_module": true
-[36m(Runner pid=1984436)[0m },
-[36m(Runner pid=1984436)[0m "rollout": {
-[36m(Runner pid=1984436)[0m "name": "vllm",
-[36m(Runner pid=1984436)[0m "n": 8,
-[36m(Runner pid=1984436)[0m "temperature": 1.0,
-[36m(Runner pid=1984436)[0m "top_p": 0.85,
-[36m(Runner pid=1984436)[0m "top_k": -1,
-[36m(Runner pid=1984436)[0m "seed": 42,
-[36m(Runner pid=1984436)[0m "limit_images": 10,
-[36m(Runner pid=1984436)[0m "dtype": "bf16",
-[36m(Runner pid=1984436)[0m "gpu_memory_utilization": 0.45,
-[36m(Runner pid=1984436)[0m "ignore_eos": false,
-[36m(Runner pid=1984436)[0m "enforce_eager": false,
-[36m(Runner pid=1984436)[0m "enable_chunked_prefill": true,
-[36m(Runner pid=1984436)[0m "tensor_parallel_size": 1,
-[36m(Runner pid=1984436)[0m "max_model_len": null,
-[36m(Runner pid=1984436)[0m "max_num_batched_tokens": 32768,
-[36m(Runner pid=1984436)[0m "disable_log_stats": true,
-[36m(Runner pid=1984436)[0m "disable_tqdm": true,
-[36m(Runner pid=1984436)[0m "val_override_config": {
-[36m(Runner pid=1984436)[0m "temperature": 0.7,
-[36m(Runner pid=1984436)[0m "top_p": 0.95,
-[36m(Runner pid=1984436)[0m "n": 1
-[36m(Runner pid=1984436)[0m },
-[36m(Runner pid=1984436)[0m "kv_cache_dtype": "auto",
-[36m(Runner pid=1984436)[0m "calculate_log_probs": false,
-[36m(Runner pid=1984436)[0m "collect_seq_logprob_for_filter": false,
-[36m(Runner pid=1984436)[0m "prompt_length": 24576,
-[36m(Runner pid=1984436)[0m "response_length": 4096,
-[36m(Runner pid=1984436)[0m "trust_remote_code": false
-[36m(Runner pid=1984436)[0m }
-[36m(Runner pid=1984436)[0m },
-[36m(Runner pid=1984436)[0m "algorithm": {
-[36m(Runner pid=1984436)[0m "name": "grpo",
-[36m(Runner pid=1984436)[0m "gamma": 1.0,
-[36m(Runner pid=1984436)[0m "lam": 1.0,
-[36m(Runner pid=1984436)[0m "adv_estimator": "grpo",
-[36m(Runner pid=1984436)[0m "scale_rewards": true,
-[36m(Runner pid=1984436)[0m "disable_kl": true,
-[36m(Runner pid=1984436)[0m "use_kl_loss": false,
-[36m(Runner pid=1984436)[0m "kl_penalty": "low_var_kl",
-[36m(Runner pid=1984436)[0m "kl_coef": 0.0,
-[36m(Runner pid=1984436)[0m "kl_type": "fixed",
-[36m(Runner pid=1984436)[0m "kl_horizon": 10000.0,
-[36m(Runner pid=1984436)[0m "kl_target": 0.1
-[36m(Runner pid=1984436)[0m },
-[36m(Runner pid=1984436)[0m "trainer": {
-[36m(Runner pid=1984436)[0m "total_epochs": 1,
-[36m(Runner pid=1984436)[0m "max_steps": 1,
-[36m(Runner pid=1984436)[0m "project_name": "OraRL",
-[36m(Runner pid=1984436)[0m "experiment_name": "smoke-grpo-9b",
-[36m(Runner pid=1984436)[0m "logger": [
-[36m(Runner pid=1984436)[0m "console"
-[36m(Runner pid=1984436)[0m ],
-[36m(Runner pid=1984436)[0m "nnodes": 1,
-[36m(Runner pid=1984436)[0m "n_gpus_per_node": 8,
-[36m(Runner pid=1984436)[0m "critic_warmup": 0,
-[36m(Runner pid=1984436)[0m "val_freq": -1,
-[36m(Runner pid=1984436)[0m "val_before_train": false,
-[36m(Runner pid=1984436)[0m "val_only": false,
-[36m(Runner pid=1984436)[0m "val_generations_to_log": 0,
-[36m(Runner pid=1984436)[0m "save_freq": 1,
-[36m(Runner pid=1984436)[0m "save_limit": 3,
-[36m(Runner pid=1984436)[0m "save_model_only": false,
-[36m(Runner pid=1984436)[0m "keep_optim_only_latest": true,
-[36m(Runner pid=1984436)[0m "save_checkpoint_path": "/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175213/grpo",
-[36m(Runner pid=1984436)[0m "load_checkpoint_path": null,
-[36m(Runner pid=1984436)[0m "ray_timeline": null,
-[36m(Runner pid=1984436)[0m "find_last_checkpoint": false,
-[36m(Runner pid=1984436)[0m "keep_best_train_ckpt": false,
-[36m(Runner pid=1984436)[0m "best_train_metric_key": "reward/overall",
-[36m(Runner pid=1984436)[0m "best_train_smooth_window": 5,
-[36m(Runner pid=1984436)[0m "best_train_min_step": 10
-[36m(Runner pid=1984436)[0m }
-[36m(Runner pid=1984436)[0m }
-[36m(Runner pid=1984436)[0m [TaskGroupedBatchSampler] 15 groups: image_sequence_mc_answer_only(10304), obj_appearance_order(64), object_abs_distance(1344), object_counting(960), object_rel_direction(448), object_rel_distance(640), object_size_estimation(1408), room_size_estimation(960), route_planning(960), segmentation(12032), spatial grounding(7040), spatial-temporal grounding(9536), temporal grounding(20096), tracking(13952), video_qa_mc(20288) | batch_size=8 | total_batches=12504
-[36m(Runner pid=1984436)[0m Size of train dataloader: 12504
-[36m(Runner pid=1984436)[0m Size of val dataloader: 32
-[36m(Runner pid=1984436)[0m KL is disabled, no KL metrics will be logged. Please set `kl_coef=0` to log KL metrics.[36m(pid=1986239)[0m /data/miniconda3/envs/orarl/lib/python3.11/site-packages/transformers/utils/import_utils.py:1006: FutureWarning: `is_flash_attn_greater_or_equal_2_10` is deprecated and will be removed in v5.8. Please use `is_flash_attn_greater_or_equal(library_version='2.1.0')` instead if needed.
-[36m(pid=1986239)[0m warnings.warn(
-[36m(pid=1986747)[0m /data/miniconda3/envs/orarl/lib/python3.11/site-packages/transformers/utils/import_utils.py:1006: FutureWarning: `is_flash_attn_greater_or_equal_2_10` is deprecated and will be removed in v5.8. Please use `is_flash_attn_greater_or_equal(library_version='2.1.0')` instead if needed.
-[36m(pid=1986747)[0m warnings.warn(
-[36m(pid=1986750)[0m /data/miniconda3/envs/orarl/lib/python3.11/site-packages/transformers/utils/import_utils.py:1006: FutureWarning: `is_flash_attn_greater_or_equal_2_10` is deprecated and will be removed in v5.8. Please use `is_flash_attn_greater_or_equal(library_version='2.1.0')` instead if needed.
-[36m(pid=1986750)[0m warnings.warn(
-[36m(WorkerDict pid=1986750)[0m [W825 17:53:20.598420573 ProcessGroupNCCL.cpp:929] Warning: TORCH_NCCL_AVOID_RECORD_STREAMS is the default now, this environment variable is thus deprecated. (function operator())
-[36m(pid=1986745)[0m /data/miniconda3/envs/orarl/lib/python3.11/site-packages/transformers/utils/import_utils.py:1006: FutureWarning: `is_flash_attn_greater_or_equal_2_10` is deprecated and will be removed in v5.8. Please use `is_flash_attn_greater_or_equal(library_version='2.1.0')` instead if needed.[32m [repeated 5x across cluster] (Ray deduplicates logs by default. Set RAY_DEDUP_LOGS=0 to disable log deduplication, or see https://docs.ray.io/en/master/ray-observability/user-guides/configure-logging.html#log-deduplication for more options.)[0m
-[36m(pid=1986745)[0m warnings.warn([32m [repeated 5x across cluster][0m
-[36m(WorkerDict pid=1986750)[0m The `use_fast` parameter is deprecated and will be removed in a future version. Use `backend="torchvision"` instead of `use_fast=True`, or `backend="pil"` instead of `use_fast=False`.
-[36m(WorkerDict pid=1986750)[0m `torch_dtype` is deprecated! Use `dtype` instead!
-[36m(WorkerDict pid=1986745)[0m [W825 17:53:22.611308342 ProcessGroupNCCL.cpp:929] Warning: TORCH_NCCL_AVOID_RECORD_STREAMS is the default now, this environment variable is thus deprecated. (function operator())[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=1986750)[0m The fast path is not available because one of the required library is not installed. Falling back to torch implementation. To install follow https://github.com/fla-org/flash-linear-attention#installation and https://github.com/Dao-AILab/causal-conv1d
-[36m(WorkerDict pid=1986750)[0m [rank7]:[W825 17:53:25.252368567 ProcessGroupNCCL.cpp:5138] Guessing device ID based on global rank. This can cause a hang if rank to GPU mapping is heterogeneous. You can specify device_id in init_process_group()
-[36m(WorkerDict pid=1986239)[0m The `use_fast` parameter is deprecated and will be removed in a future version. Use `backend="torchvision"` instead of `use_fast=True`, or `backend="pil"` instead of `use_fast=False`.[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=1986239)[0m
Loading weights: 0%| | 0/760 [00:00, ?it/s]
-[36m(WorkerDict pid=1986239)[0m
Loading weights: 100%|██████████| 760/760 [00:00<00:00, 9091.24it/s]
-[36m(WorkerDict pid=1986239)[0m /data/miniconda3/envs/orarl/lib/python3.11/site-packages/torch/distributed/c10d_logger.py:83: UserWarning: barrier(): using the device under current context. You can specify `device_id` in `init_process_group` to mute this warning.
-[36m(WorkerDict pid=1986239)[0m return func(*args, **kwargs)
-[36m(WorkerDict pid=1986750)[0m `Qwen2VLImageProcessorFast` is deprecated. The `Fast` suffix for image processors has been removed; use `Qwen2VLImageProcessor` instead.
-[36m(WorkerDict pid=1986748)[0m `torch_dtype` is deprecated! Use `dtype` instead![32m [repeated 6x across cluster][0m
-[36m(WorkerDict pid=1986748)[0m The fast path is not available because one of the required library is not installed. Falling back to torch implementation. To install follow https://github.com/fla-org/flash-linear-attention#installation and https://github.com/Dao-AILab/causal-conv1d[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=1986239)[0m [rank0]:[W825 17:53:28.954233606 ProcessGroupNCCL.cpp:5138] Guessing device ID based on global rank. This can cause a hang if rank to GPU mapping is heterogeneous. You can specify device_id in init_process_group()[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=1986747)[0m :1241: FutureWarning: The cuda.cudart module is deprecated and will be removed in a future release, please switch to use the cuda.bindings.runtime module instead.
-[36m(WorkerDict pid=1986747)[0m :1241: FutureWarning: The cuda.nvrtc module is deprecated and will be removed in a future release, please switch to use the cuda.bindings.nvrtc module instead.
-[36m(WorkerDict pid=1986749)[0m `Qwen2VLImageProcessorFast` is deprecated. The `Fast` suffix for image processors has been removed; use `Qwen2VLImageProcessor` instead.[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=1986239)[0m 2026-08-25 17:54:33,940 - INFO - autotuner.py:262 - flashinfer.jit: [Autotuner]: Autotuning process starts ...
-[36m(WorkerDict pid=1986239)[0m 2026-08-25 17:54:33,958 - INFO - autotuner.py:268 - flashinfer.jit: [Autotuner]: Autotuning process ends
-[36m(WorkerDict pid=1986748)[0m :1241: FutureWarning: The cuda.cudart module is deprecated and will be removed in a future release, please switch to use the cuda.bindings.runtime module instead.[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=1986748)[0m :1241: FutureWarning: The cuda.nvrtc module is deprecated and will be removed in a future release, please switch to use the cuda.bindings.nvrtc module instead.[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 0%| | 0/51 [00:00, ?it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 2%|▏ | 1/51 [00:00<00:05, 9.91it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 6%|▌ | 3/51 [00:00<00:03, 12.20it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 10%|▉ | 5/51 [00:00<00:03, 13.14it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 14%|█▎ | 7/51 [00:00<00:03, 13.55it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 18%|█▊ | 9/51 [00:00<00:03, 13.17it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 22%|██▏ | 11/51 [00:00<00:02, 13.83it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 25%|██▌ | 13/51 [00:00<00:02, 14.58it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 29%|██▉ | 15/51 [00:01<00:02, 15.15it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 33%|███▎ | 17/51 [00:01<00:02, 16.11it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 39%|███▉ | 20/51 [00:01<00:01, 17.88it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 45%|████▌ | 23/51 [00:01<00:01, 18.51it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 49%|████▉ | 25/51 [00:01<00:01, 18.81it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 55%|█████▍ | 28/51 [00:01<00:01, 19.76it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 61%|██████ | 31/51 [00:01<00:00, 20.73it/s]
-[36m(WorkerDict pid=1986750)[0m 2026-08-25 17:54:33,938 - INFO - autotuner.py:262 - flashinfer.jit: [Autotuner]: Autotuning process starts ...[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=1986750)[0m 2026-08-25 17:54:33,956 - INFO - autotuner.py:268 - flashinfer.jit: [Autotuner]: Autotuning process ends[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 67%|██████▋ | 34/51 [00:01<00:00, 21.99it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 73%|███████▎ | 37/51 [00:02<00:00, 22.86it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 78%|███████▊ | 40/51 [00:02<00:00, 22.41it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 84%|████████▍ | 43/51 [00:02<00:00, 23.82it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 90%|█████████ | 46/51 [00:02<00:00, 24.27it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 96%|█████████▌| 49/51 [00:02<00:00, 25.66it/s]
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 100%|██████████| 51/51 [00:02<00:00, 19.63it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (decode, FULL): 0%| | 0/51 [00:00, ?it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (decode, FULL): 2%|▏ | 1/51 [00:00<00:10, 4.74it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (decode, FULL): 6%|▌ | 3/51 [00:00<00:05, 8.45it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (decode, FULL): 10%|▉ | 5/51 [00:00<00:04, 10.01it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (decode, FULL): 14%|█▎ | 7/51 [00:00<00:03, 11.04it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (decode, FULL): 18%|█▊ | 9/51 [00:00<00:03, 11.91it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (decode, FULL): 22%|██▏ | 11/51 [00:00<00:03, 12.70it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (decode, FULL): 25%|██▌ | 13/51 [00:01<00:03, 12.13it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (decode, FULL): 29%|██▉ | 15/51 [00:01<00:02, 12.99it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (decode, FULL): 33%|███▎ | 17/51 [00:01<00:02, 14.07it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (decode, FULL): 37%|███▋ | 19/51 [00:01<00:02, 11.67it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (decode, FULL): 41%|████ | 21/51 [00:01<00:02, 13.29it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (decode, FULL): 45%|████▌ | 23/51 [00:01<00:01, 14.62it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (decode, FULL): 49%|████▉ | 25/51 [00:01<00:01, 15.81it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (decode, FULL): 55%|█████▍ | 28/51 [00:02<00:01, 18.07it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (decode, FULL): 61%|██████ | 31/51 [00:02<00:00, 20.16it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (decode, FULL): 67%|██████▋ | 34/51 [00:02<00:00, 21.76it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (decode, FULL): 73%|███████▎ | 37/51 [00:02<00:00, 19.94it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (decode, FULL): 80%|████████ | 41/51 [00:02<00:00, 23.51it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (decode, FULL): 88%|████████▊ | 45/51 [00:02<00:00, 26.54it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (decode, FULL): 96%|█████████▌| 49/51 [00:02<00:00, 28.89it/s]
-[36m(WorkerDict pid=1986239)[0m
Capturing CUDA graphs (decode, FULL): 100%|██████████| 51/51 [00:03<00:00, 16.62it/s]
-
-[36m(Runner pid=1984436)[0m [GRPO] Scaled advantages ON: A=(r-mean)/(std+eps).
-[36m(Runner pid=1984436)[0m Total training steps: 1
-[36m(AutoRewardManager pid=1986029)[0m Using reward function `compute_score` from `orarl.rewards`.
-[36m(AutoRewardManager pid=1986029)[0m Reward name: orarl, reward type: batch.
-[36m(WorkerDict pid=1986239)[0m actor will use global batch size 64.
-[36m(AutoRewardManager pid=1986031)[0m Using reward function `compute_score` from `orarl.rewards`.
-[36m(AutoRewardManager pid=1986031)[0m Reward name: orarl, reward type: batch.
-[36m(WorkerDict pid=1986239)[0m Model config: Qwen3_5Config {
-[36m(WorkerDict pid=1986239)[0m "architectures": [
-[36m(WorkerDict pid=1986239)[0m "Qwen3_5ForConditionalGeneration"
-[36m(WorkerDict pid=1986239)[0m ],
-[36m(WorkerDict pid=1986239)[0m "dtype": "bfloat16",
-[36m(WorkerDict pid=1986239)[0m "image_token_id": 248056,
-[36m(WorkerDict pid=1986239)[0m "model_type": "qwen3_5",
-[36m(WorkerDict pid=1986239)[0m "text_config": {
-[36m(WorkerDict pid=1986239)[0m "attention_bias": false,
-[36m(WorkerDict pid=1986239)[0m "attention_dropout": 0.0,
-[36m(WorkerDict pid=1986239)[0m "attn_output_gate": true,
-[36m(WorkerDict pid=1986239)[0m "bos_token_id": null,
-[36m(WorkerDict pid=1986239)[0m "dtype": "bfloat16",
-[36m(WorkerDict pid=1986239)[0m "eos_token_id": 248044,
-[36m(WorkerDict pid=1986239)[0m "full_attention_interval": 4,
-[36m(WorkerDict pid=1986239)[0m "head_dim": 256,
-[36m(WorkerDict pid=1986239)[0m "hidden_act": "silu",
-[36m(WorkerDict pid=1986239)[0m "hidden_size": 4096,
-[36m(WorkerDict pid=1986239)[0m "initializer_range": 0.02,
-[36m(WorkerDict pid=1986239)[0m "intermediate_size": 12288,
-[36m(WorkerDict pid=1986239)[0m "layer_types": [
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "full_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "full_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "full_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "full_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "full_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "full_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "full_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "linear_attention",
-[36m(WorkerDict pid=1986239)[0m "full_attention"
-[36m(WorkerDict pid=1986239)[0m ],
-[36m(WorkerDict pid=1986239)[0m "linear_conv_kernel_dim": 4,
-[36m(WorkerDict pid=1986239)[0m "linear_key_head_dim": 128,
-[36m(WorkerDict pid=1986239)[0m "linear_num_key_heads": 16,
-[36m(WorkerDict pid=1986239)[0m "linear_num_value_heads": 32,
-[36m(WorkerDict pid=1986239)[0m "linear_value_head_dim": 128,
-[36m(WorkerDict pid=1986239)[0m "mamba_ssm_dtype": "float32",
-[36m(WorkerDict pid=1986239)[0m "max_position_embeddings": 262144,
-[36m(WorkerDict pid=1986239)[0m "mlp_only_layers": [],
-[36m(WorkerDict pid=1986239)[0m "model_type": "qwen3_5_text",
-[36m(WorkerDict pid=1986239)[0m "mtp_num_hidden_layers": 0,
-[36m(WorkerDict pid=1986239)[0m "mtp_use_dedicated_embeddings": false,
-[36m(WorkerDict pid=1986239)[0m "num_attention_heads": 16,
-[36m(WorkerDict pid=1986239)[0m "num_hidden_layers": 32,
-[36m(WorkerDict pid=1986239)[0m "num_key_value_heads": 4,
-[36m(WorkerDict pid=1986239)[0m "pad_token_id": null,
-[36m(WorkerDict pid=1986239)[0m "partial_rotary_factor": 0.25,
-[36m(WorkerDict pid=1986239)[0m "rms_norm_eps": 1e-06,
-[36m(WorkerDict pid=1986239)[0m "rope_parameters": {
-[36m(WorkerDict pid=1986239)[0m "mrope_interleaved": true,
-[36m(WorkerDict pid=1986239)[0m "mrope_section": [
-[36m(WorkerDict pid=1986239)[0m 11,
-[36m(WorkerDict pid=1986239)[0m 11,
-[36m(WorkerDict pid=1986239)[0m 10
-[36m(WorkerDict pid=1986239)[0m ],
-[36m(WorkerDict pid=1986239)[0m "partial_rotary_factor": 0.25,
-[36m(WorkerDict pid=1986239)[0m "rope_theta": 10000000,
-[36m(WorkerDict pid=1986239)[0m "rope_type": "default"
-[36m(WorkerDict pid=1986239)[0m },
-[36m(WorkerDict pid=1986239)[0m "tie_word_embeddings": false,
-[36m(WorkerDict pid=1986239)[0m "use_cache": true,
-[36m(WorkerDict pid=1986239)[0m "vocab_size": 248320
-[36m(WorkerDict pid=1986239)[0m },
-[36m(WorkerDict pid=1986239)[0m "tie_word_embeddings": false,
-[36m(WorkerDict pid=1986239)[0m "transformers_version": "5.5.4",
-[36m(WorkerDict pid=1986239)[0m "video_token_id": 248057,
-[36m(WorkerDict pid=1986239)[0m "vision_config": {
-[36m(WorkerDict pid=1986239)[0m "deepstack_visual_indexes": [],
-[36m(WorkerDict pid=1986239)[0m "depth": 27,
-[36m(WorkerDict pid=1986239)[0m "dtype": "bfloat16",
-[36m(WorkerDict pid=1986239)[0m "hidden_act": "gelu_pytorch_tanh",
-[36m(WorkerDict pid=1986239)[0m "hidden_size": 1152,
-[36m(WorkerDict pid=1986239)[0m "in_channels": 3,
-[36m(WorkerDict pid=1986239)[0m "initializer_range": 0.02,
-[36m(WorkerDict pid=1986239)[0m "intermediate_size": 4304,
-[36m(WorkerDict pid=1986239)[0m "model_type": "qwen3_5",
-[36m(WorkerDict pid=1986239)[0m "num_heads": 16,
-[36m(WorkerDict pid=1986239)[0m "num_position_embeddings": 2304,
-[36m(WorkerDict pid=1986239)[0m "out_hidden_size": 4096,
-[36m(WorkerDict pid=1986239)[0m "patch_size": 16,
-[36m(WorkerDict pid=1986239)[0m "spatial_merge_size": 2,
-[36m(WorkerDict pid=1986239)[0m "temporal_patch_size": 2
-[36m(WorkerDict pid=1986239)[0m },
-[36m(WorkerDict pid=1986239)[0m "vision_end_token_id": 248054,
-[36m(WorkerDict pid=1986239)[0m "vision_start_token_id": 248053
-[36m(WorkerDict pid=1986239)[0m }
-[36m(WorkerDict pid=1986239)[0m
-[36m(WorkerDict pid=1986239)[0m Ulysses patch applied!
-[36m(WorkerDict pid=1986239)[0m Vision backbone is set to not trainable.
-[36m(WorkerDict pid=1986239)[0m Qwen3_5ForConditionalGeneration contains 9.41B parameters.
-[36m(WorkerDict pid=1986239)[0m After huggingface model init: 1.17 GB / 95.00 GB.
-[36m(WorkerDict pid=1986239)[0m FSDP wrap policy: functools.partial(, transformer_layer_cls={, }).
-[36m(WorkerDict pid=1986239)[0m After FSDP module init: 15.42 GB / 95.00 GB.
-[36m(WorkerDict pid=1986239)[0m After optimizer init: 15.42 GB / 95.00 GB.
-[36m(WorkerDict pid=1986744)[0m Sampling params: {'max_tokens': 4096, 'detokenize': False, 'logit_bias': {248056: -100, 248057: -100}, 'n': 8, 'temperature': 1.0, 'top_p': 0.85, 'top_k': -1, 'ignore_eos': False}.
-[36m(WorkerDict pid=1986239)[0m After vllm init: 7.76 GB / 95.00 GB.
-[36m(Runner pid=1984436)[0m Config
-[36m(Runner pid=1984436)[0m algorithm:
-[36m(Runner pid=1984436)[0m adv_estimator: grpo
-[36m(Runner pid=1984436)[0m disable_kl: true
-[36m(Runner pid=1984436)[0m gamma: 1.0
-[36m(Runner pid=1984436)[0m kl_coef: 0.0
-[36m(Runner pid=1984436)[0m kl_horizon: 10000.0
-[36m(Runner pid=1984436)[0m kl_penalty: low_var_kl
-[36m(Runner pid=1984436)[0m kl_target: 0.1
-[36m(Runner pid=1984436)[0m kl_type: fixed
-[36m(Runner pid=1984436)[0m lam: 1.0
-[36m(Runner pid=1984436)[0m name: grpo
-[36m(Runner pid=1984436)[0m scale_rewards: true
-[36m(Runner pid=1984436)[0m use_kl_loss: false
-[36m(Runner pid=1984436)[0m data:
-[36m(Runner pid=1984436)[0m answer_key: answer
-[36m(Runner pid=1984436)[0m dataloader_num_workers: 1
-[36m(Runner pid=1984436)[0m enable_thinking: false
-[36m(Runner pid=1984436)[0m filter_overlong_prompts: false
-[36m(Runner pid=1984436)[0m filter_overlong_prompts_workers: 16
-[36m(Runner pid=1984436)[0m format_prompt: null
-[36m(Runner pid=1984436)[0m group_by_task: true
-[36m(Runner pid=1984436)[0m group_by_task_key: problem_type
-[36m(Runner pid=1984436)[0m image_dir: /
-[36m(Runner pid=1984436)[0m image_key: images
-[36m(Runner pid=1984436)[0m image_max_pixels: 1048576
-[36m(Runner pid=1984436)[0m image_min_pixels: 4096
-[36m(Runner pid=1984436)[0m inline_video_tensors: true
-[36m(Runner pid=1984436)[0m max_pixels: 4194304
-[36m(Runner pid=1984436)[0m max_prompt_length: 24576
-[36m(Runner pid=1984436)[0m max_response_length: 4096
-[36m(Runner pid=1984436)[0m min_pixels: 4096
-[36m(Runner pid=1984436)[0m mini_rollout_batch_size: null
-[36m(Runner pid=1984436)[0m override_chat_template: null
-[36m(Runner pid=1984436)[0m preprocessed_video_dir: null
-[36m(Runner pid=1984436)[0m prompt_key: problem
-[36m(Runner pid=1984436)[0m response_prefix: ''
-[36m(Runner pid=1984436)[0m rollout_batch_size: 8
-[36m(Runner pid=1984436)[0m seed: 42
-[36m(Runner pid=1984436)[0m shuffle: true
-[36m(Runner pid=1984436)[0m train_files: /mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_joint_100k_maskaware.jsonl
-[36m(Runner pid=1984436)[0m use_preprocessed_videos: false
-[36m(Runner pid=1984436)[0m val_batch_size: 16
-[36m(Runner pid=1984436)[0m val_files: /mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_phase3_49152.canary.jsonl
-[36m(Runner pid=1984436)[0m val_preprocessed_video_dir: null
-[36m(Runner pid=1984436)[0m val_video_fps: 2.0
-[36m(Runner pid=1984436)[0m val_video_max_frames: 128
-[36m(Runner pid=1984436)[0m val_video_max_pixels: 786432
-[36m(Runner pid=1984436)[0m val_video_min_pixels: 4096
-[36m(Runner pid=1984436)[0m val_video_source_mode: realtime_only
-[36m(Runner pid=1984436)[0m val_video_total_pixels: 8388608
-[36m(Runner pid=1984436)[0m video_fps: 2.0
-[36m(Runner pid=1984436)[0m video_key: videos
-[36m(Runner pid=1984436)[0m video_max_frames: 128
-[36m(Runner pid=1984436)[0m video_max_pixels: 786432
-[36m(Runner pid=1984436)[0m video_min_pixels: 4096
-[36m(Runner pid=1984436)[0m video_source_mode: realtime_only
-[36m(Runner pid=1984436)[0m video_total_pixels: 8388608
-[36m(Runner pid=1984436)[0m trainer:
-[36m(Runner pid=1984436)[0m best_train_metric_key: reward/overall
-[36m(Runner pid=1984436)[0m best_train_min_step: 10
-[36m(Runner pid=1984436)[0m best_train_smooth_window: 5
-[36m(Runner pid=1984436)[0m critic_warmup: 0
-[36m(Runner pid=1984436)[0m experiment_name: smoke-grpo-9b
-[36m(Runner pid=1984436)[0m find_last_checkpoint: false
-[36m(Runner pid=1984436)[0m keep_best_train_ckpt: false
-[36m(Runner pid=1984436)[0m keep_optim_only_latest: true
-[36m(Runner pid=1984436)[0m load_checkpoint_path: null
-[36m(Runner pid=1984436)[0m logger:
-[36m(Runner pid=1984436)[0m - console
-[36m(Runner pid=1984436)[0m max_steps: 1
-[36m(Runner pid=1984436)[0m n_gpus_per_node: 8
-[36m(Runner pid=1984436)[0m nnodes: 1
-[36m(Runner pid=1984436)[0m project_name: OraRL
-[36m(Runner pid=1984436)[0m ray_timeline: null
-[36m(Runner pid=1984436)[0m save_checkpoint_path: /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175213/grpo
-[36m(Runner pid=1984436)[0m save_freq: 1
-[36m(Runner pid=1984436)[0m save_limit: 3
-[36m(Runner pid=1984436)[0m save_model_only: false
-[36m(Runner pid=1984436)[0m total_epochs: 1
-[36m(Runner pid=1984436)[0m val_before_train: false
-[36m(Runner pid=1984436)[0m val_freq: -1
-[36m(Runner pid=1984436)[0m val_generations_to_log: 0
-[36m(Runner pid=1984436)[0m val_only: false
-[36m(Runner pid=1984436)[0m worker:
-[36m(Runner pid=1984436)[0m actor:
-[36m(Runner pid=1984436)[0m clip_ratio_dual: 3.0
-[36m(Runner pid=1984436)[0m clip_ratio_high: 0.2
-[36m(Runner pid=1984436)[0m clip_ratio_low: 0.2
-[36m(Runner pid=1984436)[0m disable_kl: true
-[36m(Runner pid=1984436)[0m dynamic_batching: true
-[36m(Runner pid=1984436)[0m fsdp:
-[36m(Runner pid=1984436)[0m enable_cpu_offload: false
-[36m(Runner pid=1984436)[0m enable_full_shard: true
-[36m(Runner pid=1984436)[0m enable_rank0_init: true
-[36m(Runner pid=1984436)[0m fsdp_size: -1
-[36m(Runner pid=1984436)[0m mp_buffer_dtype: fp32
-[36m(Runner pid=1984436)[0m mp_param_dtype: bf16
-[36m(Runner pid=1984436)[0m mp_reduce_dtype: fp32
-[36m(Runner pid=1984436)[0m torch_dtype: bf16
-[36m(Runner pid=1984436)[0m use_orig_params: false
-[36m(Runner pid=1984436)[0m global_batch_size: 8
-[36m(Runner pid=1984436)[0m global_batch_size_per_device: -1
-[36m(Runner pid=1984436)[0m kl_coef: 0.0
-[36m(Runner pid=1984436)[0m kl_penalty: low_var_kl
-[36m(Runner pid=1984436)[0m loss_avg_mode: token
-[36m(Runner pid=1984436)[0m loss_type: default
-[36m(Runner pid=1984436)[0m max_grad_norm: 1.0
-[36m(Runner pid=1984436)[0m max_token_len_per_gpu: 32768
-[36m(Runner pid=1984436)[0m micro_batch_size_per_device_for_experience: 1
-[36m(Runner pid=1984436)[0m micro_batch_size_per_device_for_update: 1
-[36m(Runner pid=1984436)[0m model:
-[36m(Runner pid=1984436)[0m enable_gradient_checkpointing: true
-[36m(Runner pid=1984436)[0m freeze_vision_tower: true
-[36m(Runner pid=1984436)[0m model_path: /apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B
-[36m(Runner pid=1984436)[0m override_config: {}
-[36m(Runner pid=1984436)[0m tokenizer_path: /apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B
-[36m(Runner pid=1984436)[0m train_vision_merger: false
-[36m(Runner pid=1984436)[0m trust_remote_code: false
-[36m(Runner pid=1984436)[0m offload:
-[36m(Runner pid=1984436)[0m offload_optimizer: false
-[36m(Runner pid=1984436)[0m offload_params: false
-[36m(Runner pid=1984436)[0m optim:
-[36m(Runner pid=1984436)[0m betas:
-[36m(Runner pid=1984436)[0m - 0.9
-[36m(Runner pid=1984436)[0m - 0.999
-[36m(Runner pid=1984436)[0m lr: 1.0e-06
-[36m(Runner pid=1984436)[0m lr_scheduler_type: constant
-[36m(Runner pid=1984436)[0m lr_warmup_ratio: 0.0
-[36m(Runner pid=1984436)[0m lr_warmup_steps: null
-[36m(Runner pid=1984436)[0m min_lr_ratio: null
-[36m(Runner pid=1984436)[0m strategy: adamw
-[36m(Runner pid=1984436)[0m training_steps: 1
-[36m(Runner pid=1984436)[0m weight_decay: 0.0
-[36m(Runner pid=1984436)[0m padding_free: true
-[36m(Runner pid=1984436)[0m ppo_epochs: 1
-[36m(Runner pid=1984436)[0m strategy: fsdp
-[36m(Runner pid=1984436)[0m ulysses_size: 1
-[36m(Runner pid=1984436)[0m use_kl_loss: false
-[36m(Runner pid=1984436)[0m use_torch_compile: false
-[36m(Runner pid=1984436)[0m critic:
-[36m(Runner pid=1984436)[0m cliprange_value: 0.5
-[36m(Runner pid=1984436)[0m dynamic_batching: true
-[36m(Runner pid=1984436)[0m fsdp:
-[36m(Runner pid=1984436)[0m enable_cpu_offload: false
-[36m(Runner pid=1984436)[0m enable_full_shard: true
-[36m(Runner pid=1984436)[0m enable_rank0_init: true
-[36m(Runner pid=1984436)[0m fsdp_size: -1
-[36m(Runner pid=1984436)[0m mp_buffer_dtype: fp32
-[36m(Runner pid=1984436)[0m mp_param_dtype: bf16
-[36m(Runner pid=1984436)[0m mp_reduce_dtype: fp32
-[36m(Runner pid=1984436)[0m torch_dtype: null
-[36m(Runner pid=1984436)[0m use_orig_params: false
-[36m(Runner pid=1984436)[0m global_batch_size: 256
-[36m(Runner pid=1984436)[0m global_batch_size_per_device: -1
-[36m(Runner pid=1984436)[0m loss_avg_mode: token
-[36m(Runner pid=1984436)[0m max_grad_norm: 1.0
-[36m(Runner pid=1984436)[0m max_token_len_per_gpu: null
-[36m(Runner pid=1984436)[0m micro_batch_size_per_device_for_experience: 16
-[36m(Runner pid=1984436)[0m micro_batch_size_per_device_for_update: 4
-[36m(Runner pid=1984436)[0m model:
-[36m(Runner pid=1984436)[0m enable_gradient_checkpointing: true
-[36m(Runner pid=1984436)[0m freeze_vision_tower: false
-[36m(Runner pid=1984436)[0m model_path: null
-[36m(Runner pid=1984436)[0m override_config: {}
-[36m(Runner pid=1984436)[0m tokenizer_path: null
-[36m(Runner pid=1984436)[0m train_vision_merger: false
-[36m(Runner pid=1984436)[0m trust_remote_code: true
-[36m(Runner pid=1984436)[0m offload:
-[36m(Runner pid=1984436)[0m offload_optimizer: false
-[36m(Runner pid=1984436)[0m offload_params: false
-[36m(Runner pid=1984436)[0m optim:
-[36m(Runner pid=1984436)[0m betas:
-[36m(Runner pid=1984436)[0m - 0.9
-[36m(Runner pid=1984436)[0m - 0.999
-[36m(Runner pid=1984436)[0m lr: 1.0e-06
-[36m(Runner pid=1984436)[0m lr_scheduler_type: constant
-[36m(Runner pid=1984436)[0m lr_warmup_ratio: 0.0
-[36m(Runner pid=1984436)[0m lr_warmup_steps: null
-[36m(Runner pid=1984436)[0m min_lr_ratio: null
-[36m(Runner pid=1984436)[0m strategy: adamw
-[36m(Runner pid=1984436)[0m training_steps: 1
-[36m(Runner pid=1984436)[0m weight_decay: 0.01
-[36m(Runner pid=1984436)[0m padding_free: false
-[36m(Runner pid=1984436)[0m ppo_epochs: 1
-[36m(Runner pid=1984436)[0m strategy: fsdp
-[36m(Runner pid=1984436)[0m ulysses_size: 1
-[36m(Runner pid=1984436)[0m hybrid_engine: true
-[36m(Runner pid=1984436)[0m ref:
-[36m(Runner pid=1984436)[0m dynamic_batching: true
-[36m(Runner pid=1984436)[0m fsdp:
-[36m(Runner pid=1984436)[0m enable_cpu_offload: false
-[36m(Runner pid=1984436)[0m enable_full_shard: true
-[36m(Runner pid=1984436)[0m enable_rank0_init: true
-[36m(Runner pid=1984436)[0m fsdp_size: -1
-[36m(Runner pid=1984436)[0m mp_buffer_dtype: fp32
-[36m(Runner pid=1984436)[0m mp_param_dtype: bf16
-[36m(Runner pid=1984436)[0m mp_reduce_dtype: fp32
-[36m(Runner pid=1984436)[0m torch_dtype: bf16
-[36m(Runner pid=1984436)[0m use_orig_params: false
-[36m(Runner pid=1984436)[0m max_token_len_per_gpu: 32768
-[36m(Runner pid=1984436)[0m micro_batch_size_per_device_for_experience: 1
-[36m(Runner pid=1984436)[0m offload:
-[36m(Runner pid=1984436)[0m offload_optimizer: false
-[36m(Runner pid=1984436)[0m offload_params: false
-[36m(Runner pid=1984436)[0m padding_free: true
-[36m(Runner pid=1984436)[0m strategy: fsdp
-[36m(Runner pid=1984436)[0m ulysses_size: 1
-[36m(Runner pid=1984436)[0m use_torch_compile: false
-[36m(Runner pid=1984436)[0m reward:
-[36m(Runner pid=1984436)[0m num_cpus: 1
-[36m(Runner pid=1984436)[0m reward_function: orarl.rewards
-[36m(Runner pid=1984436)[0m reward_function_is_module: true
-[36m(Runner pid=1984436)[0m reward_function_kwargs: {}
-[36m(Runner pid=1984436)[0m reward_function_name: compute_score
-[36m(Runner pid=1984436)[0m skip_special_tokens: true
-[36m(Runner pid=1984436)[0m rollout:
-[36m(Runner pid=1984436)[0m calculate_log_probs: false
-[36m(Runner pid=1984436)[0m collect_seq_logprob_for_filter: false
-[36m(Runner pid=1984436)[0m disable_log_stats: true
-[36m(Runner pid=1984436)[0m disable_tqdm: true
-[36m(Runner pid=1984436)[0m dtype: bf16
-[36m(Runner pid=1984436)[0m enable_chunked_prefill: true
-[36m(Runner pid=1984436)[0m enforce_eager: false
-[36m(Runner pid=1984436)[0m gpu_memory_utilization: 0.45
-[36m(Runner pid=1984436)[0m ignore_eos: false
-[36m(Runner pid=1984436)[0m kv_cache_dtype: auto
-[36m(Runner pid=1984436)[0m limit_images: 10
-[36m(Runner pid=1984436)[0m max_model_len: null
-[36m(Runner pid=1984436)[0m max_num_batched_tokens: 32768
-[36m(Runner pid=1984436)[0m n: 8
-[36m(Runner pid=1984436)[0m name: vllm
-[36m(Runner pid=1984436)[0m prompt_length: 24576
-[36m(Runner pid=1984436)[0m response_length: 4096
-[36m(Runner pid=1984436)[0m seed: 42
-[36m(Runner pid=1984436)[0m temperature: 1.0
-[36m(Runner pid=1984436)[0m tensor_parallel_size: 1
-[36m(Runner pid=1984436)[0m top_k: -1
-[36m(Runner pid=1984436)[0m top_p: 0.85
-[36m(Runner pid=1984436)[0m trust_remote_code: false
-[36m(Runner pid=1984436)[0m val_override_config:
-[36m(Runner pid=1984436)[0m n: 1
-[36m(Runner pid=1984436)[0m temperature: 0.7
-[36m(Runner pid=1984436)[0m top_p: 0.95
-[36m(Runner pid=1984436)[0m
-
[2m[36m(pid=1984436) [0mRunning step: 0%| | 0.00/1.00 [00:00, ?it/s]
[2m[36m(pid=1984436) [0mRunning step: 0%| | 0.00/1.00 [00:00, ?it/s]
[2m[36m(pid=1984436) [0mRunning step: 0%| | 0.00/1.00 [00:00, ?it/s]
[2m[36m(pid=1984436) [0mRunning step: 0%| | 0.00/1.00 [00:01, ?it/s]
[2m[36m(pid=1984436) [0mRunning step: 0%| | 0.00/1.00 [00:01, ?it/s]
[2m[36m(pid=1984436) [0mRunning step: 0%| | 0.00/1.00 [00:01, ?it/s]Traceback (most recent call last):
- File "", line 198, in _run_module_as_main
- File "", line 88, in _run_code
- File "/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/verl/trainer/main.py", line 137, in
- main()
- File "/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/verl/trainer/main.py", line 129, in main
- ray.get(runner.run.remote(ppo_config))
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/ray/_private/auto_init_hook.py", line 22, in auto_init_wrapper
- return fn(*args, **kwargs)
- ^^^^^^^^^^^^^^^^^^^
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/ray/_private/client_mode_hook.py", line 104, in wrapper
- return func(*args, **kwargs)
- ^^^^^^^^^^^^^^^^^^^^^
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/ray/_private/worker.py", line 2981, in get
- values, debugger_breakpoint = worker.get_objects(
- ^^^^^^^^^^^^^^^^^^^
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/ray/_private/worker.py", line 1012, in get_objects
- raise value.as_instanceof_cause()
-ray.exceptions.RayTaskError(IndexError): [36mray::Runner.run()[39m (pid=1984436, ip=29.127.69.4, actor_id=2700fe7dc5d33e51d56c429c01000000, repr=)
- ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
- ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
- File "/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/verl/trainer/main.py", line 92, in run
- trainer.fit()
- File "/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/verl/trainer/ray_trainer.py", line 1870, in fit
- batch = self._make_batch_data(metrics=metrics)
- ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
- File "/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/verl/trainer/ray_trainer.py", line 1662, in _make_batch_data
- batch_dict = next(self.data_iterator)
- ^^^^^^^^^^^^^^^^^^^^^^^^
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/torchdata/stateful_dataloader/stateful_dataloader.py", line 450, in __next__
- return super().__next__()
- ^^^^^^^^^^^^^^^^^^
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/torch/utils/data/dataloader.py", line 741, in __next__
- data = self._next_data()
- ^^^^^^^^^^^^^^^^^
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/torchdata/stateful_dataloader/stateful_dataloader.py", line 1456, in _next_data
- return self._process_data(data, worker_id, state_dict)
- ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/torchdata/stateful_dataloader/stateful_dataloader.py", line 1543, in _process_data
- data.reraise()
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/torch/_utils.py", line 775, in reraise
- raise exception
-IndexError: Caught IndexError in DataLoader worker process 0.
-Original Traceback (most recent call last):
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/torchdata/stateful_dataloader/worker.py", line 242, in _worker_loop
- data = fetcher.fetch(index) # type: ignore[union-attr]
- ^^^^^^^^^^^^^^^^^^^^
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/torch/utils/data/_utils/fetch.py", line 54, in fetch
- data = [self.dataset[idx] for idx in possibly_batched_index]
- ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/torch/utils/data/_utils/fetch.py", line 54, in
- data = [self.dataset[idx] for idx in possibly_batched_index]
- ~~~~~~~~~~~~^^^^^
- File "/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/verl/utils/dataset.py", line 826, in __getitem__
- model_inputs = self.processor(processed_images, [prompt], add_special_tokens=False, return_tensors="pt")
- ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
- File "/data/miniconda3/envs/orarl/lib/python3.11/site-packages/transformers/models/qwen3_vl/processing_qwen3_vl.py", line 132, in __call__
- num_image_tokens = image_grid_thw[index].prod() // merge_length
- ~~~~~~~~~~~~~~^^^^^^^
-IndexError: index 8 is out of bounds for dimension 0 with size 8
-
[36m(Runner pid=1984436)[0m qwen-vl-utils using decord to read video.
-
[2m[36m(pid=1984436) [0mRunning step: 0%| | 0.00/1.00 [00:02, ?it/s]
[2m[36m(pid=1984436) [0mRunning step: 0%| | 0.00/1.00 [00:02, ?it/s][36m(WorkerDict pid=1986239)[0m Before vllm wake up in sharding manager: 7.76 GB / 95.00 GB.
-[36m(WorkerDict pid=1986239)[0m After gather model weights in sharding manager: 31.26 GB / 95.00 GB.
-[36m(Runner pid=1984436)[0m Start generating batch...
-[36m(WorkerDict pid=1986239)[0m After sync model weights in sharding manager: 25.44 GB / 95.00 GB.
-[36m(WorkerDict pid=1986239)[0m After vllm wake up in sharding manager: 45.79 GB / 95.00 GB.
-[36m(WorkerDict pid=1986746)[0m Sampling params: {'max_tokens': 4096, 'detokenize': False, 'logit_bias': {248056: -100, 248057: -100}, 'n': 8, 'temperature': 1.0, 'top_p': 0.85, 'top_k': -1, 'ignore_eos': False}.[32m [repeated 7x across cluster][0m
-
[2m[36m(pid=1984436) [0mRunning step: 0%| | 0.00/1.00 [00:06, ?it/s]
-method=grpo mode=run
-/data/miniconda3/envs/orarl/bin/python -m verl.trainer.main config=/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/configs/grpo_9b.yaml trainer.max_steps=1 trainer.save_freq=1 trainer.val_before_train=false trainer.experiment_name=smoke-grpo-9b data.rollout_batch_size=8 worker.actor.global_batch_size=8 worker.actor.model.model_path=/apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B worker.actor.model.tokenizer_path=/apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B data.train_files=/mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_joint_100k_maskaware.jsonl data.val_files=/mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_phase3_49152.canary.jsonl trainer.save_checkpoint_path=/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175213/grpo trainer.nnodes=1 trainer.n_gpus_per_node=8
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/checkpoint_tracker.json b/code/runs/smoke-training-9b-20260825_175949/grpo/checkpoint_tracker.json
deleted file mode 100644
index 0fcf0afbb9e157734d78d06eda5f8cf105fc7736..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/checkpoint_tracker.json
+++ /dev/null
@@ -1,6 +0,0 @@
-{
- "best_global_step": null,
- "best_val_reward_score": -1.0,
- "last_global_step": 1,
- "last_actor_path": "/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor"
-}
\ No newline at end of file
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_0.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_0.pt
deleted file mode 100644
index 9ddf9fa959f9f7e04210a74d1c92d33701d6b33e..0000000000000000000000000000000000000000
Binary files a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_0.pt and /dev/null differ
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_1.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_1.pt
deleted file mode 100644
index ff1c6d6f862719840b74d28237a295340dabc090..0000000000000000000000000000000000000000
Binary files a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_1.pt and /dev/null differ
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_2.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_2.pt
deleted file mode 100644
index 36e5d1767618d404ae9ef7a17bb7ab62bce1e0ac..0000000000000000000000000000000000000000
Binary files a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_2.pt and /dev/null differ
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_3.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_3.pt
deleted file mode 100644
index e8e74bd7a72db76984c7c1d48e12ef2c44d43042..0000000000000000000000000000000000000000
Binary files a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_3.pt and /dev/null differ
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_4.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_4.pt
deleted file mode 100644
index f02841fbe96b46e5590e8a124c6636fcc2848b80..0000000000000000000000000000000000000000
Binary files a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_4.pt and /dev/null differ
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_5.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_5.pt
deleted file mode 100644
index 09e2a3f4856e2e4e56c7538dfc8eb632e89f0987..0000000000000000000000000000000000000000
Binary files a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_5.pt and /dev/null differ
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_6.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_6.pt
deleted file mode 100644
index 63eb7ddfc6d3921af129980eaa14baed168b2dc4..0000000000000000000000000000000000000000
Binary files a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_6.pt and /dev/null differ
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_7.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_7.pt
deleted file mode 100644
index 7f00686a15d8d21702a4c9063f36fcc958eb18ba..0000000000000000000000000000000000000000
Binary files a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_7.pt and /dev/null differ
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/huggingface/chat_template.jinja b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/huggingface/chat_template.jinja
deleted file mode 100644
index a585dec894e63da457d9440ec6aa7caa16d20860..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/huggingface/chat_template.jinja
+++ /dev/null
@@ -1,154 +0,0 @@
-{%- set image_count = namespace(value=0) %}
-{%- set video_count = namespace(value=0) %}
-{%- macro render_content(content, do_vision_count, is_system_content=false) %}
- {%- if content is string %}
- {{- content }}
- {%- elif content is iterable and content is not mapping %}
- {%- for item in content %}
- {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
- {%- if is_system_content %}
- {{- raise_exception('System message cannot contain images.') }}
- {%- endif %}
- {%- if do_vision_count %}
- {%- set image_count.value = image_count.value + 1 %}
- {%- endif %}
- {%- if add_vision_id %}
- {{- 'Picture ' ~ image_count.value ~ ': ' }}
- {%- endif %}
- {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
- {%- elif 'video' in item or item.type == 'video' %}
- {%- if is_system_content %}
- {{- raise_exception('System message cannot contain videos.') }}
- {%- endif %}
- {%- if do_vision_count %}
- {%- set video_count.value = video_count.value + 1 %}
- {%- endif %}
- {%- if add_vision_id %}
- {{- 'Video ' ~ video_count.value ~ ': ' }}
- {%- endif %}
- {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
- {%- elif 'text' in item %}
- {{- item.text }}
- {%- else %}
- {{- raise_exception('Unexpected item type in content.') }}
- {%- endif %}
- {%- endfor %}
- {%- elif content is none or content is undefined %}
- {{- '' }}
- {%- else %}
- {{- raise_exception('Unexpected content type.') }}
- {%- endif %}
-{%- endmacro %}
-{%- if not messages %}
- {{- raise_exception('No messages provided.') }}
-{%- endif %}
-{%- if tools and tools is iterable and tools is not mapping %}
- {{- '<|im_start|>system\n' }}
- {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
- {%- for tool in tools %}
- {{- "\n" }}
- {{- tool | tojson }}
- {%- endfor %}
- {{- "\n " }}
- {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n \n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n \n \n \n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n ' }}
- {%- if messages[0].role == 'system' %}
- {%- set content = render_content(messages[0].content, false, true)|trim %}
- {%- if content %}
- {{- '\n\n' + content }}
- {%- endif %}
- {%- endif %}
- {{- '<|im_end|>\n' }}
-{%- else %}
- {%- if messages[0].role == 'system' %}
- {%- set content = render_content(messages[0].content, false, true)|trim %}
- {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
- {%- endif %}
-{%- endif %}
-{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
-{%- for message in messages[::-1] %}
- {%- set index = (messages|length - 1) - loop.index0 %}
- {%- if ns.multi_step_tool and message.role == "user" %}
- {%- set content = render_content(message.content, false)|trim %}
- {%- if not(content.startswith('') and content.endswith(' ')) %}
- {%- set ns.multi_step_tool = false %}
- {%- set ns.last_query_index = index %}
- {%- endif %}
- {%- endif %}
-{%- endfor %}
-{%- if ns.multi_step_tool %}
- {{- raise_exception('No user query found in messages.') }}
-{%- endif %}
-{%- for message in messages %}
- {%- set content = render_content(message.content, true)|trim %}
- {%- if message.role == "system" %}
- {%- if not loop.first %}
- {{- raise_exception('System message must be at the beginning.') }}
- {%- endif %}
- {%- elif message.role == "user" %}
- {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
- {%- elif message.role == "assistant" %}
- {%- set reasoning_content = '' %}
- {%- if message.reasoning_content is string %}
- {%- set reasoning_content = message.reasoning_content %}
- {%- else %}
- {%- if '' in content %}
- {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
- {%- set content = content.split(' ')[-1].lstrip('\n') %}
- {%- endif %}
- {%- endif %}
- {%- set reasoning_content = reasoning_content|trim %}
- {%- if loop.index0 > ns.last_query_index %}
- {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n \n\n' + content }}
- {%- else %}
- {{- '<|im_start|>' + message.role + '\n' + content }}
- {%- endif %}
- {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
- {%- for tool_call in message.tool_calls %}
- {%- if tool_call.function is defined %}
- {%- set tool_call = tool_call.function %}
- {%- endif %}
- {%- if loop.first %}
- {%- if content|trim %}
- {{- '\n\n\n\n' }}
- {%- else %}
- {{- '\n\n' }}
- {%- endif %}
- {%- else %}
- {{- '\n\n\n' }}
- {%- endif %}
- {%- if tool_call.arguments is defined %}
- {%- for args_name, args_value in tool_call.arguments|items %}
- {{- '\n' }}
- {%- set args_value = args_value | tojson | safe if args_value is mapping or (args_value is sequence and args_value is not string) else args_value | string %}
- {{- args_value }}
- {{- '\n \n' }}
- {%- endfor %}
- {%- endif %}
- {{- ' \n ' }}
- {%- endfor %}
- {%- endif %}
- {{- '<|im_end|>\n' }}
- {%- elif message.role == "tool" %}
- {%- if loop.previtem and loop.previtem.role != "tool" %}
- {{- '<|im_start|>user' }}
- {%- endif %}
- {{- '\n\n' }}
- {{- content }}
- {{- '\n ' }}
- {%- if not loop.last and loop.nextitem.role != "tool" %}
- {{- '<|im_end|>\n' }}
- {%- elif loop.last %}
- {{- '<|im_end|>\n' }}
- {%- endif %}
- {%- else %}
- {{- raise_exception('Unexpected message role.') }}
- {%- endif %}
-{%- endfor %}
-{%- if add_generation_prompt %}
- {{- '<|im_start|>assistant\n' }}
- {%- if enable_thinking is defined and enable_thinking is false %}
- {{- '\n\n \n\n' }}
- {%- else %}
- {{- '\n' }}
- {%- endif %}
-{%- endif %}
\ No newline at end of file
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/huggingface/config.json b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/huggingface/config.json
deleted file mode 100644
index d13e73613c775be9cd78fc28ff9c7aa65090debc..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/huggingface/config.json
+++ /dev/null
@@ -1,109 +0,0 @@
-{
- "architectures": [
- "Qwen3_5ForConditionalGeneration"
- ],
- "dtype": "bfloat16",
- "image_token_id": 248056,
- "model_type": "qwen3_5",
- "text_config": {
- "attention_bias": false,
- "attention_dropout": 0.0,
- "attn_output_gate": true,
- "bos_token_id": null,
- "dtype": "bfloat16",
- "eos_token_id": 248044,
- "full_attention_interval": 4,
- "head_dim": 256,
- "hidden_act": "silu",
- "hidden_size": 4096,
- "initializer_range": 0.02,
- "intermediate_size": 12288,
- "layer_types": [
- "linear_attention",
- "linear_attention",
- "linear_attention",
- "full_attention",
- "linear_attention",
- "linear_attention",
- "linear_attention",
- "full_attention",
- "linear_attention",
- "linear_attention",
- "linear_attention",
- "full_attention",
- "linear_attention",
- "linear_attention",
- "linear_attention",
- "full_attention",
- "linear_attention",
- "linear_attention",
- "linear_attention",
- "full_attention",
- "linear_attention",
- "linear_attention",
- "linear_attention",
- "full_attention",
- "linear_attention",
- "linear_attention",
- "linear_attention",
- "full_attention",
- "linear_attention",
- "linear_attention",
- "linear_attention",
- "full_attention"
- ],
- "linear_conv_kernel_dim": 4,
- "linear_key_head_dim": 128,
- "linear_num_key_heads": 16,
- "linear_num_value_heads": 32,
- "linear_value_head_dim": 128,
- "mamba_ssm_dtype": "float32",
- "max_position_embeddings": 262144,
- "mlp_only_layers": [],
- "model_type": "qwen3_5_text",
- "mtp_num_hidden_layers": 0,
- "mtp_use_dedicated_embeddings": false,
- "num_attention_heads": 16,
- "num_hidden_layers": 32,
- "num_key_value_heads": 4,
- "pad_token_id": null,
- "partial_rotary_factor": 0.25,
- "rms_norm_eps": 1e-06,
- "rope_parameters": {
- "mrope_interleaved": true,
- "mrope_section": [
- 11,
- 11,
- 10
- ],
- "partial_rotary_factor": 0.25,
- "rope_theta": 10000000,
- "rope_type": "default"
- },
- "tie_word_embeddings": false,
- "use_cache": true,
- "vocab_size": 248320
- },
- "tie_word_embeddings": false,
- "transformers_version": "5.5.4",
- "video_token_id": 248057,
- "vision_config": {
- "deepstack_visual_indexes": [],
- "depth": 27,
- "dtype": "bfloat16",
- "hidden_act": "gelu_pytorch_tanh",
- "hidden_size": 1152,
- "in_channels": 3,
- "initializer_range": 0.02,
- "intermediate_size": 4304,
- "model_type": "qwen3_5",
- "num_heads": 16,
- "num_position_embeddings": 2304,
- "out_hidden_size": 4096,
- "patch_size": 16,
- "spatial_merge_size": 2,
- "temporal_patch_size": 2
- },
- "vision_end_token_id": 248054,
- "vision_start_token_id": 248053
-}
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/huggingface/generation_config.json b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/huggingface/generation_config.json
deleted file mode 100644
index 23a0a961f467960631ac5faedc9547a7523a99c3..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/huggingface/generation_config.json
+++ /dev/null
@@ -1,13 +0,0 @@
-{
- "bos_token_id": 248044,
- "do_sample": true,
- "eos_token_id": [
- 248046,
- 248044
- ],
- "pad_token_id": 248044,
- "temperature": 1.0,
- "top_k": 20,
- "top_p": 0.95,
- "transformers_version": "5.5.4"
-}
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/huggingface/processor_config.json b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/huggingface/processor_config.json
deleted file mode 100644
index 33818c7f9e991ad735fd240209f4fa73e6c28c50..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/huggingface/processor_config.json
+++ /dev/null
@@ -1,60 +0,0 @@
-{
- "image_processor": {
- "do_convert_rgb": true,
- "do_normalize": true,
- "do_rescale": true,
- "do_resize": true,
- "image_mean": [
- 0.5,
- 0.5,
- 0.5
- ],
- "image_processor_type": "Qwen2VLImageProcessor",
- "image_std": [
- 0.5,
- 0.5,
- 0.5
- ],
- "merge_size": 2,
- "patch_size": 16,
- "resample": 3,
- "rescale_factor": 0.00392156862745098,
- "size": {
- "longest_edge": 16777216,
- "shortest_edge": 65536
- },
- "temporal_patch_size": 2
- },
- "processor_class": "Qwen3VLProcessor",
- "video_processor": {
- "do_convert_rgb": true,
- "do_normalize": true,
- "do_rescale": true,
- "do_resize": true,
- "do_sample_frames": true,
- "fps": 2,
- "image_mean": [
- 0.5,
- 0.5,
- 0.5
- ],
- "image_std": [
- 0.5,
- 0.5,
- 0.5
- ],
- "max_frames": 768,
- "merge_size": 2,
- "min_frames": 4,
- "patch_size": 16,
- "resample": 3,
- "rescale_factor": 0.00392156862745098,
- "return_metadata": false,
- "size": {
- "longest_edge": 25165824,
- "shortest_edge": 4096
- },
- "temporal_patch_size": 2,
- "video_processor_type": "Qwen3VLVideoProcessor"
- }
-}
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/huggingface/tokenizer.json b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/huggingface/tokenizer.json
deleted file mode 100644
index 67741b04f23bfdb46501f748ce27865ec82eccfb..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/huggingface/tokenizer.json
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:87a7830d63fcf43bf241c3c5242e96e62dd3fdc29224ca26fed8ea333db72de4
-size 19989343
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/huggingface/tokenizer_config.json b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/huggingface/tokenizer_config.json
deleted file mode 100644
index a068e2468cff426a9b105006e74e044030a6faf4..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/huggingface/tokenizer_config.json
+++ /dev/null
@@ -1,32 +0,0 @@
-{
- "add_prefix_space": false,
- "audio_bos_token": "<|audio_start|>",
- "audio_eos_token": "<|audio_end|>",
- "audio_token": "<|audio_pad|>",
- "backend": "tokenizers",
- "bos_token": null,
- "clean_up_tokenization_spaces": false,
- "eos_token": "<|im_end|>",
- "errors": "replace",
- "image_token": "<|image_pad|>",
- "is_local": true,
- "model_max_length": 262144,
- "model_specific_special_tokens": {
- "audio_bos_token": "<|audio_start|>",
- "audio_eos_token": "<|audio_end|>",
- "audio_token": "<|audio_pad|>",
- "image_token": "<|image_pad|>",
- "video_token": "<|video_pad|>",
- "vision_bos_token": "<|vision_start|>",
- "vision_eos_token": "<|vision_end|>"
- },
- "pad_token": "<|endoftext|>",
- "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
- "processor_class": "Qwen3VLProcessor",
- "split_special_tokens": false,
- "tokenizer_class": "TokenizersBackend",
- "unk_token": null,
- "video_token": "<|video_pad|>",
- "vision_bos_token": "<|vision_start|>",
- "vision_eos_token": "<|vision_end|>"
-}
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_0.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_0.pt
deleted file mode 100644
index 880f25f27588d18338bad74731696b06ad1d5069..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_0.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:e7a112d3ef8e9efd0c7404343835e8d1cacb68f1bf47ec9e70a6035d628bfeae
-size 2352797236
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_1.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_1.pt
deleted file mode 100644
index 8ef782c0875bd58d1b344b7b9210181f0a73c911..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_1.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:929bde074b4648a5e40ea9d59f1b37f587f605044b3013c2d9663496ff270997
-size 2352797236
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_2.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_2.pt
deleted file mode 100644
index f058aa02326eebbfdd6ada84a9fca495c4e8d8c2..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_2.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:5f191a9b777c259be21b12b8ed02585b790915f74c547f0ac2600a45ce3a598a
-size 2352797236
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_3.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_3.pt
deleted file mode 100644
index 4eaa7a0a3388cef91ac20367d03a2844da7264ef..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_3.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ff8e73e975bbe1ea882acbebe7533b050a9c127beb1c39b8a5d3cfd72ffbc17c
-size 2352797236
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_4.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_4.pt
deleted file mode 100644
index 43b28746068bd1bc4f689d867d183ee0a6b2d845..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_4.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:49a0e15c74448d5debb04be1f773463910f08df82e32e33fc164e1e674a1d3e2
-size 2352797236
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_5.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_5.pt
deleted file mode 100644
index 99a1dcf49b6a2a434da998b7c75ed95ca8cf2f84..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_5.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:84a14b9ff89d4c3769d06de5e98af7f71b0b2c7d82226c3e9761d37af480a7db
-size 2352797236
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_6.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_6.pt
deleted file mode 100644
index 93876eceb78280fcfdeebc8a43cb000e143864b7..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_6.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:53206b46e825c183bee092c93d115c7863e368745f0b4a9340f477845fff6521
-size 2352797236
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_7.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_7.pt
deleted file mode 100644
index 6dad326e8199db4e1d35257f90ea26820f3d8378..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_7.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:5f69907e184a8e128ebd0087e9e47ebd62f2cfa47dd93b356b2b03df20317cda
-size 2352797236
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_0.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_0.pt
deleted file mode 100644
index 96582c2dad336962a44f62e4cac0203859483a87..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_0.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:dc1018349737fd19af9fe2ae532ee8aed53a1cd2a6289bd51efc0e893092b5e7
-size 4477429805
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_1.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_1.pt
deleted file mode 100644
index f62cef8911fdf35efc7061d2221bc4606731850c..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_1.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:72b7df9ff763ae286ab239e33b2d6711bff9f19a16f9043418b8c68accc08cb2
-size 4477429805
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_2.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_2.pt
deleted file mode 100644
index 09b9bd2a1e73a35547da89bba6ad3648bb0920ad..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_2.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:fad4ce51bcfbc75a0bd746f41af31c8c44cd57974acc1a36adcc4153bf303a6b
-size 4477429805
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_3.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_3.pt
deleted file mode 100644
index aa6b5487fb06482430ec76e98a0654e0605b4cab..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_3.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:11ed1bcf8aa7199792a3f094d8844b6840cdd53efb3fcdfd7f1ea6931b07ea75
-size 4477429805
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_4.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_4.pt
deleted file mode 100644
index 2b5480769f41a56f2df48815aec56d3e40ce5384..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_4.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ccb3231b0406316170091c991dd49652d43e89bcc3f8544cbe436742baf2e960
-size 4477429805
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_5.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_5.pt
deleted file mode 100644
index 2b57e8782e887d1442ec5b5c931eb9ad2249a8e6..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_5.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:8b1e07b9349724c6f93fd95d24146c217483921a201c48db3cd1742e22527b70
-size 4477429805
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_6.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_6.pt
deleted file mode 100644
index 5e407504f5f07f27deaa2e30082b11fd8111c5f1..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_6.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:92c3605f87040e3ee654db866052d94620101854bc2ee697c10a0536bc5b3dbb
-size 4477429805
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_7.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_7.pt
deleted file mode 100644
index 4829ff68eb7dff55f11034b8df1e65bbfc6769c0..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_7.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:c2f8b77f8faef58fce850c407f800d2cd0e7c9e4388e7c7a08f6465357fb9891
-size 4477429805
diff --git a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/dataloader.pt b/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/dataloader.pt
deleted file mode 100644
index 237b2a50543b7937e88f70166e880f4e7a02342f..0000000000000000000000000000000000000000
Binary files a/code/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/dataloader.pt and /dev/null differ
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/checkpoint_tracker.json b/code/runs/smoke-training-9b-20260825_175949/orarl/checkpoint_tracker.json
deleted file mode 100644
index 89e57145ec755ed0b11d88f01c9cda722eb0237d..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/checkpoint_tracker.json
+++ /dev/null
@@ -1,6 +0,0 @@
-{
- "best_global_step": null,
- "best_val_reward_score": -1.0,
- "last_global_step": 1,
- "last_actor_path": "/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor"
-}
\ No newline at end of file
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_0.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_0.pt
deleted file mode 100644
index 9ddf9fa959f9f7e04210a74d1c92d33701d6b33e..0000000000000000000000000000000000000000
Binary files a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_0.pt and /dev/null differ
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_1.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_1.pt
deleted file mode 100644
index ff1c6d6f862719840b74d28237a295340dabc090..0000000000000000000000000000000000000000
Binary files a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_1.pt and /dev/null differ
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_2.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_2.pt
deleted file mode 100644
index 36e5d1767618d404ae9ef7a17bb7ab62bce1e0ac..0000000000000000000000000000000000000000
Binary files a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_2.pt and /dev/null differ
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_3.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_3.pt
deleted file mode 100644
index e8e74bd7a72db76984c7c1d48e12ef2c44d43042..0000000000000000000000000000000000000000
Binary files a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_3.pt and /dev/null differ
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_4.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_4.pt
deleted file mode 100644
index f02841fbe96b46e5590e8a124c6636fcc2848b80..0000000000000000000000000000000000000000
Binary files a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_4.pt and /dev/null differ
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_5.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_5.pt
deleted file mode 100644
index 09e2a3f4856e2e4e56c7538dfc8eb632e89f0987..0000000000000000000000000000000000000000
Binary files a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_5.pt and /dev/null differ
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_6.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_6.pt
deleted file mode 100644
index 63eb7ddfc6d3921af129980eaa14baed168b2dc4..0000000000000000000000000000000000000000
Binary files a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_6.pt and /dev/null differ
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_7.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_7.pt
deleted file mode 100644
index 7f00686a15d8d21702a4c9063f36fcc958eb18ba..0000000000000000000000000000000000000000
Binary files a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_7.pt and /dev/null differ
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/huggingface/chat_template.jinja b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/huggingface/chat_template.jinja
deleted file mode 100644
index a585dec894e63da457d9440ec6aa7caa16d20860..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/huggingface/chat_template.jinja
+++ /dev/null
@@ -1,154 +0,0 @@
-{%- set image_count = namespace(value=0) %}
-{%- set video_count = namespace(value=0) %}
-{%- macro render_content(content, do_vision_count, is_system_content=false) %}
- {%- if content is string %}
- {{- content }}
- {%- elif content is iterable and content is not mapping %}
- {%- for item in content %}
- {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
- {%- if is_system_content %}
- {{- raise_exception('System message cannot contain images.') }}
- {%- endif %}
- {%- if do_vision_count %}
- {%- set image_count.value = image_count.value + 1 %}
- {%- endif %}
- {%- if add_vision_id %}
- {{- 'Picture ' ~ image_count.value ~ ': ' }}
- {%- endif %}
- {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
- {%- elif 'video' in item or item.type == 'video' %}
- {%- if is_system_content %}
- {{- raise_exception('System message cannot contain videos.') }}
- {%- endif %}
- {%- if do_vision_count %}
- {%- set video_count.value = video_count.value + 1 %}
- {%- endif %}
- {%- if add_vision_id %}
- {{- 'Video ' ~ video_count.value ~ ': ' }}
- {%- endif %}
- {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
- {%- elif 'text' in item %}
- {{- item.text }}
- {%- else %}
- {{- raise_exception('Unexpected item type in content.') }}
- {%- endif %}
- {%- endfor %}
- {%- elif content is none or content is undefined %}
- {{- '' }}
- {%- else %}
- {{- raise_exception('Unexpected content type.') }}
- {%- endif %}
-{%- endmacro %}
-{%- if not messages %}
- {{- raise_exception('No messages provided.') }}
-{%- endif %}
-{%- if tools and tools is iterable and tools is not mapping %}
- {{- '<|im_start|>system\n' }}
- {{- "# Tools\n\nYou have access to the following functions:\n\n" }}
- {%- for tool in tools %}
- {{- "\n" }}
- {{- tool | tojson }}
- {%- endfor %}
- {{- "\n " }}
- {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n\n\n\nvalue_1\n \n\nThis is the value for the second parameter\nthat can span\nmultiple lines\n \n \n \n\n\nReminder:\n- Function calls MUST follow the specified format: an inner block must be nested within XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n ' }}
- {%- if messages[0].role == 'system' %}
- {%- set content = render_content(messages[0].content, false, true)|trim %}
- {%- if content %}
- {{- '\n\n' + content }}
- {%- endif %}
- {%- endif %}
- {{- '<|im_end|>\n' }}
-{%- else %}
- {%- if messages[0].role == 'system' %}
- {%- set content = render_content(messages[0].content, false, true)|trim %}
- {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
- {%- endif %}
-{%- endif %}
-{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
-{%- for message in messages[::-1] %}
- {%- set index = (messages|length - 1) - loop.index0 %}
- {%- if ns.multi_step_tool and message.role == "user" %}
- {%- set content = render_content(message.content, false)|trim %}
- {%- if not(content.startswith('') and content.endswith(' ')) %}
- {%- set ns.multi_step_tool = false %}
- {%- set ns.last_query_index = index %}
- {%- endif %}
- {%- endif %}
-{%- endfor %}
-{%- if ns.multi_step_tool %}
- {{- raise_exception('No user query found in messages.') }}
-{%- endif %}
-{%- for message in messages %}
- {%- set content = render_content(message.content, true)|trim %}
- {%- if message.role == "system" %}
- {%- if not loop.first %}
- {{- raise_exception('System message must be at the beginning.') }}
- {%- endif %}
- {%- elif message.role == "user" %}
- {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
- {%- elif message.role == "assistant" %}
- {%- set reasoning_content = '' %}
- {%- if message.reasoning_content is string %}
- {%- set reasoning_content = message.reasoning_content %}
- {%- else %}
- {%- if ' ' in content %}
- {%- set reasoning_content = content.split('')[0].rstrip('\n').split('')[-1].lstrip('\n') %}
- {%- set content = content.split(' ')[-1].lstrip('\n') %}
- {%- endif %}
- {%- endif %}
- {%- set reasoning_content = reasoning_content|trim %}
- {%- if loop.index0 > ns.last_query_index %}
- {{- '<|im_start|>' + message.role + '\n\n' + reasoning_content + '\n \n\n' + content }}
- {%- else %}
- {{- '<|im_start|>' + message.role + '\n' + content }}
- {%- endif %}
- {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
- {%- for tool_call in message.tool_calls %}
- {%- if tool_call.function is defined %}
- {%- set tool_call = tool_call.function %}
- {%- endif %}
- {%- if loop.first %}
- {%- if content|trim %}
- {{- '\n\n\n\n' }}
- {%- else %}
- {{- '\n\n' }}
- {%- endif %}
- {%- else %}
- {{- '\n\n\n' }}
- {%- endif %}
- {%- if tool_call.arguments is defined %}
- {%- for args_name, args_value in tool_call.arguments|items %}
- {{- '\n' }}
- {%- set args_value = args_value | tojson | safe if args_value is mapping or (args_value is sequence and args_value is not string) else args_value | string %}
- {{- args_value }}
- {{- '\n \n' }}
- {%- endfor %}
- {%- endif %}
- {{- ' \n ' }}
- {%- endfor %}
- {%- endif %}
- {{- '<|im_end|>\n' }}
- {%- elif message.role == "tool" %}
- {%- if loop.previtem and loop.previtem.role != "tool" %}
- {{- '<|im_start|>user' }}
- {%- endif %}
- {{- '\n\n' }}
- {{- content }}
- {{- '\n ' }}
- {%- if not loop.last and loop.nextitem.role != "tool" %}
- {{- '<|im_end|>\n' }}
- {%- elif loop.last %}
- {{- '<|im_end|>\n' }}
- {%- endif %}
- {%- else %}
- {{- raise_exception('Unexpected message role.') }}
- {%- endif %}
-{%- endfor %}
-{%- if add_generation_prompt %}
- {{- '<|im_start|>assistant\n' }}
- {%- if enable_thinking is defined and enable_thinking is false %}
- {{- '\n\n \n\n' }}
- {%- else %}
- {{- '\n' }}
- {%- endif %}
-{%- endif %}
\ No newline at end of file
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/huggingface/config.json b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/huggingface/config.json
deleted file mode 100644
index d13e73613c775be9cd78fc28ff9c7aa65090debc..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/huggingface/config.json
+++ /dev/null
@@ -1,109 +0,0 @@
-{
- "architectures": [
- "Qwen3_5ForConditionalGeneration"
- ],
- "dtype": "bfloat16",
- "image_token_id": 248056,
- "model_type": "qwen3_5",
- "text_config": {
- "attention_bias": false,
- "attention_dropout": 0.0,
- "attn_output_gate": true,
- "bos_token_id": null,
- "dtype": "bfloat16",
- "eos_token_id": 248044,
- "full_attention_interval": 4,
- "head_dim": 256,
- "hidden_act": "silu",
- "hidden_size": 4096,
- "initializer_range": 0.02,
- "intermediate_size": 12288,
- "layer_types": [
- "linear_attention",
- "linear_attention",
- "linear_attention",
- "full_attention",
- "linear_attention",
- "linear_attention",
- "linear_attention",
- "full_attention",
- "linear_attention",
- "linear_attention",
- "linear_attention",
- "full_attention",
- "linear_attention",
- "linear_attention",
- "linear_attention",
- "full_attention",
- "linear_attention",
- "linear_attention",
- "linear_attention",
- "full_attention",
- "linear_attention",
- "linear_attention",
- "linear_attention",
- "full_attention",
- "linear_attention",
- "linear_attention",
- "linear_attention",
- "full_attention",
- "linear_attention",
- "linear_attention",
- "linear_attention",
- "full_attention"
- ],
- "linear_conv_kernel_dim": 4,
- "linear_key_head_dim": 128,
- "linear_num_key_heads": 16,
- "linear_num_value_heads": 32,
- "linear_value_head_dim": 128,
- "mamba_ssm_dtype": "float32",
- "max_position_embeddings": 262144,
- "mlp_only_layers": [],
- "model_type": "qwen3_5_text",
- "mtp_num_hidden_layers": 0,
- "mtp_use_dedicated_embeddings": false,
- "num_attention_heads": 16,
- "num_hidden_layers": 32,
- "num_key_value_heads": 4,
- "pad_token_id": null,
- "partial_rotary_factor": 0.25,
- "rms_norm_eps": 1e-06,
- "rope_parameters": {
- "mrope_interleaved": true,
- "mrope_section": [
- 11,
- 11,
- 10
- ],
- "partial_rotary_factor": 0.25,
- "rope_theta": 10000000,
- "rope_type": "default"
- },
- "tie_word_embeddings": false,
- "use_cache": true,
- "vocab_size": 248320
- },
- "tie_word_embeddings": false,
- "transformers_version": "5.5.4",
- "video_token_id": 248057,
- "vision_config": {
- "deepstack_visual_indexes": [],
- "depth": 27,
- "dtype": "bfloat16",
- "hidden_act": "gelu_pytorch_tanh",
- "hidden_size": 1152,
- "in_channels": 3,
- "initializer_range": 0.02,
- "intermediate_size": 4304,
- "model_type": "qwen3_5",
- "num_heads": 16,
- "num_position_embeddings": 2304,
- "out_hidden_size": 4096,
- "patch_size": 16,
- "spatial_merge_size": 2,
- "temporal_patch_size": 2
- },
- "vision_end_token_id": 248054,
- "vision_start_token_id": 248053
-}
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/huggingface/generation_config.json b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/huggingface/generation_config.json
deleted file mode 100644
index 23a0a961f467960631ac5faedc9547a7523a99c3..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/huggingface/generation_config.json
+++ /dev/null
@@ -1,13 +0,0 @@
-{
- "bos_token_id": 248044,
- "do_sample": true,
- "eos_token_id": [
- 248046,
- 248044
- ],
- "pad_token_id": 248044,
- "temperature": 1.0,
- "top_k": 20,
- "top_p": 0.95,
- "transformers_version": "5.5.4"
-}
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/huggingface/processor_config.json b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/huggingface/processor_config.json
deleted file mode 100644
index 33818c7f9e991ad735fd240209f4fa73e6c28c50..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/huggingface/processor_config.json
+++ /dev/null
@@ -1,60 +0,0 @@
-{
- "image_processor": {
- "do_convert_rgb": true,
- "do_normalize": true,
- "do_rescale": true,
- "do_resize": true,
- "image_mean": [
- 0.5,
- 0.5,
- 0.5
- ],
- "image_processor_type": "Qwen2VLImageProcessor",
- "image_std": [
- 0.5,
- 0.5,
- 0.5
- ],
- "merge_size": 2,
- "patch_size": 16,
- "resample": 3,
- "rescale_factor": 0.00392156862745098,
- "size": {
- "longest_edge": 16777216,
- "shortest_edge": 65536
- },
- "temporal_patch_size": 2
- },
- "processor_class": "Qwen3VLProcessor",
- "video_processor": {
- "do_convert_rgb": true,
- "do_normalize": true,
- "do_rescale": true,
- "do_resize": true,
- "do_sample_frames": true,
- "fps": 2,
- "image_mean": [
- 0.5,
- 0.5,
- 0.5
- ],
- "image_std": [
- 0.5,
- 0.5,
- 0.5
- ],
- "max_frames": 768,
- "merge_size": 2,
- "min_frames": 4,
- "patch_size": 16,
- "resample": 3,
- "rescale_factor": 0.00392156862745098,
- "return_metadata": false,
- "size": {
- "longest_edge": 25165824,
- "shortest_edge": 4096
- },
- "temporal_patch_size": 2,
- "video_processor_type": "Qwen3VLVideoProcessor"
- }
-}
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/huggingface/tokenizer.json b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/huggingface/tokenizer.json
deleted file mode 100644
index 67741b04f23bfdb46501f748ce27865ec82eccfb..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/huggingface/tokenizer.json
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:87a7830d63fcf43bf241c3c5242e96e62dd3fdc29224ca26fed8ea333db72de4
-size 19989343
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/huggingface/tokenizer_config.json b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/huggingface/tokenizer_config.json
deleted file mode 100644
index a068e2468cff426a9b105006e74e044030a6faf4..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/huggingface/tokenizer_config.json
+++ /dev/null
@@ -1,32 +0,0 @@
-{
- "add_prefix_space": false,
- "audio_bos_token": "<|audio_start|>",
- "audio_eos_token": "<|audio_end|>",
- "audio_token": "<|audio_pad|>",
- "backend": "tokenizers",
- "bos_token": null,
- "clean_up_tokenization_spaces": false,
- "eos_token": "<|im_end|>",
- "errors": "replace",
- "image_token": "<|image_pad|>",
- "is_local": true,
- "model_max_length": 262144,
- "model_specific_special_tokens": {
- "audio_bos_token": "<|audio_start|>",
- "audio_eos_token": "<|audio_end|>",
- "audio_token": "<|audio_pad|>",
- "image_token": "<|image_pad|>",
- "video_token": "<|video_pad|>",
- "vision_bos_token": "<|vision_start|>",
- "vision_eos_token": "<|vision_end|>"
- },
- "pad_token": "<|endoftext|>",
- "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
- "processor_class": "Qwen3VLProcessor",
- "split_special_tokens": false,
- "tokenizer_class": "TokenizersBackend",
- "unk_token": null,
- "video_token": "<|video_pad|>",
- "vision_bos_token": "<|vision_start|>",
- "vision_eos_token": "<|vision_end|>"
-}
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_0.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_0.pt
deleted file mode 100644
index 965bcc65404bdb192443f340dcb3d2ae70da0d82..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_0.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:68d916aca907a48f093fcb05501c28b5e0a46a127dae4e7077e1bf7da273a578
-size 2352797236
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_1.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_1.pt
deleted file mode 100644
index 59aac22dc11af461393131f923317783bbf79a9a..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_1.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:59a86054eef4b1dc5ad060c150e8435918f4221890f037bef4800944e065ac9c
-size 2352797236
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_2.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_2.pt
deleted file mode 100644
index d61124c62cf6d85809c98cde069ab62cc93e5fbc..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_2.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:dbdea55e015c41e8b330efd62a209990d6396cc7e9bbefb3db3a01615e130094
-size 2352797236
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_3.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_3.pt
deleted file mode 100644
index b943ad5dee63423f2944d482cafc8fb16d7b28af..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_3.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:2a4b11162a92caf7e46a3df91b572fc32bd6e899d751e80b5c29e24f44bb3d3d
-size 2352797236
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_4.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_4.pt
deleted file mode 100644
index 533a04be9c2909452ad93c4908a226ac8f2b0f8c..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_4.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:b967583161f9f4c2e32ad4e4e1e6143aef915a0eb86d8cc648bbba99b1110fd4
-size 2352797236
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_5.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_5.pt
deleted file mode 100644
index 94deff6a447bd4c4a12ef7e467bd95b190ebd5ed..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_5.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:681ba029790747845f2e435e9515175d2f62475f8c5df86d8cd73641c8805790
-size 2352797236
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_6.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_6.pt
deleted file mode 100644
index 8896d9b1ec7c92dd781bb0f3c16a2305db3b4313..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_6.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:3a8d4c40826eb2dd2f293b9d24132b7099c718f232badccc7595930d9567b8e8
-size 2352797236
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_7.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_7.pt
deleted file mode 100644
index 00bacbdd1dbdeaae2ff6794802f06bb5d2dd5f14..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_7.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:629e7633a242b848283f4f61d2a8e63a7b3e1bc7ed1fa96ee8820c7145700c4d
-size 2352797236
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_0.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_0.pt
deleted file mode 100644
index f14b6d6ad7ea58e61effea441cb67d17545f3561..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_0.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:9f55ce17accdc1232fa2501877d6ed35f409bd8d6dc4cc9374f193bc5623ed20
-size 4477429805
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_1.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_1.pt
deleted file mode 100644
index 4908c9a622c8f0bbe76f4102754bd7f2e9b5c71c..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_1.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:4c8607d4247f9852139f11f1ac2fea7630177c0bccdb25383a75b549ebb3fe0c
-size 4477429805
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_2.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_2.pt
deleted file mode 100644
index 3082d4018cd73ad883996034c5bde012cb49576f..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_2.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:783a7bdb5c7289b31bb284c1b5f4e1fbcdc14b0386e61c12d33a29a624f619e5
-size 4477429805
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_3.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_3.pt
deleted file mode 100644
index 6bf10bb33bf914ba7a4692396058035a1d445e5d..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_3.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:1e47f4381f34e93df73d5005da02800a003e906bab5ec093ec433a4b9fc0234f
-size 4477429805
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_4.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_4.pt
deleted file mode 100644
index cbe647e71dd1f6074b22248d1a279e2c387fb130..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_4.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:a275a020afff2f926632a3c18dfc66d881e44a8a39922107128102d7afdfa076
-size 4477429805
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_5.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_5.pt
deleted file mode 100644
index 1e46e5c6f599fd7888466d152a8665715293600c..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_5.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:f1fb064bf27a05dd600fcac0bf25685dae85254178e71fc05e7413e547a415cb
-size 4477429805
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_6.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_6.pt
deleted file mode 100644
index c8e0a2d15efe3f25b6cbc6ec894443c9489707c7..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_6.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:719541230bc7a694e4f98a0162f19b5bf70adabd627e43df1a7cb3832661f45d
-size 4477429805
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_7.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_7.pt
deleted file mode 100644
index 16bc4663efd6064401fa86e7ae3bb19bc6488665..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_7.pt
+++ /dev/null
@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:53aabad2855f21fb69be25b681885b9c45d2cbbe9ce0ab8bd077f9f81be6b6b5
-size 4477429805
diff --git a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/dataloader.pt b/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/dataloader.pt
deleted file mode 100644
index 237b2a50543b7937e88f70166e880f4e7a02342f..0000000000000000000000000000000000000000
Binary files a/code/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/dataloader.pt and /dev/null differ
diff --git a/code/runs/smoke-training-9b-20260825_175949/smoke.log b/code/runs/smoke-training-9b-20260825_175949/smoke.log
deleted file mode 100644
index 6c47f32837c70e758a87251ad7564cfeba5494e0..0000000000000000000000000000000000000000
--- a/code/runs/smoke-training-9b-20260825_175949/smoke.log
+++ /dev/null
@@ -1,5748 +0,0 @@
-OraRL root: /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official
-Python: /data/miniconda3/envs/orarl/bin/python
-Model: /apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B
-Train data: /mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_joint_100k_maskaware.jsonl
-Val data: /mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_phase3_49152.canary.jsonl
-Recipe scale: 9b
-World size: 1 x 8
-Steps per run: 1
-Results: /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949
-Execute: 1
-0, NVIDIA H20, 0 MiB, 97871 MiB
-1, NVIDIA H20, 0 MiB, 97871 MiB
-2, NVIDIA H20, 0 MiB, 97871 MiB
-3, NVIDIA H20, 0 MiB, 97871 MiB
-4, NVIDIA H20, 0 MiB, 97871 MiB
-5, NVIDIA H20, 0 MiB, 97871 MiB
-6, NVIDIA H20, 0 MiB, 97871 MiB
-7, NVIDIA H20, 0 MiB, 97871 MiB
-/data/miniconda3/envs/orarl/lib/python3.11/site-packages/transformers/utils/import_utils.py:1006: FutureWarning: `is_flash_attn_greater_or_equal_2_10` is deprecated and will be removed in v5.8. Please use `is_flash_attn_greater_or_equal(library_version='2.1.0')` instead if needed.
- warnings.warn(
-dependency OK: numpy 2.2.6
-dependency OK: torch 2.10.0+cu129
-dependency OK: transformers 5.5.4
-dependency OK: vllm 0.19.1
-dependency OK: ray 2.54.0
-dependency OK: tensordict 0.12.2
-dependency OK: codetiming 1.4.0
-dependency OK: omegaconf 2.3.0
-dependency OK: verl.trainer.main /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/verl/trainer/main.py
-dependency OK: orarl.rewards /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/orarl/rewards/__init__.py
-
-=== grpo smoke (1 update(s)) ===
-/data/miniconda3/envs/orarl/lib/python3.11/site-packages/transformers/utils/import_utils.py:1006: FutureWarning: `is_flash_attn_greater_or_equal_2_10` is deprecated and will be removed in v5.8. Please use `is_flash_attn_greater_or_equal(library_version='2.1.0')` instead if needed.
- warnings.warn(
-2026-08-25 18:00:17,309 INFO worker.py:2004 -- Started a local Ray instance. View the dashboard at [1m[32mhttp://127.0.0.1:8265 [39m[22m
-/data/miniconda3/envs/orarl/lib/python3.11/site-packages/ray/_private/worker.py:2052: FutureWarning: Tip: In future versions of Ray, Ray will no longer override accelerator visible devices env var if num_gpus=0 or num_gpus=None (default). To enable this behavior and turn off this error message, set RAY_ACCEL_ENV_VAR_OVERRIDE_ON_ZERO=0
- warnings.warn(
-[36m(pid=2031595)[0m W0825 18:00:33.227000 2031595 site-packages/torch/utils/cpp_extension.py:118] No CUDA runtime is found, using CUDA_HOME='/data/miniconda3/envs/orarl'
-[36m(Runner pid=2031595)[0m {
-[36m(Runner pid=2031595)[0m "data": {
-[36m(Runner pid=2031595)[0m "train_files": "/mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_joint_100k_maskaware.jsonl",
-[36m(Runner pid=2031595)[0m "val_files": "/mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_phase3_49152.canary.jsonl",
-[36m(Runner pid=2031595)[0m "prompt_key": "problem",
-[36m(Runner pid=2031595)[0m "answer_key": "answer",
-[36m(Runner pid=2031595)[0m "image_key": "images",
-[36m(Runner pid=2031595)[0m "video_key": "videos",
-[36m(Runner pid=2031595)[0m "image_dir": "/",
-[36m(Runner pid=2031595)[0m "video_fps": 2.0,
-[36m(Runner pid=2031595)[0m "video_max_frames": 128,
-[36m(Runner pid=2031595)[0m "max_prompt_length": 24576,
-[36m(Runner pid=2031595)[0m "max_response_length": 4096,
-[36m(Runner pid=2031595)[0m "rollout_batch_size": 8,
-[36m(Runner pid=2031595)[0m "mini_rollout_batch_size": null,
-[36m(Runner pid=2031595)[0m "val_batch_size": 16,
-[36m(Runner pid=2031595)[0m "format_prompt": null,
-[36m(Runner pid=2031595)[0m "override_chat_template": null,
-[36m(Runner pid=2031595)[0m "enable_thinking": false,
-[36m(Runner pid=2031595)[0m "response_prefix": "",
-[36m(Runner pid=2031595)[0m "shuffle": true,
-[36m(Runner pid=2031595)[0m "seed": 42,
-[36m(Runner pid=2031595)[0m "min_pixels": 4096,
-[36m(Runner pid=2031595)[0m "max_pixels": 4194304,
-[36m(Runner pid=2031595)[0m "image_min_pixels": 4096,
-[36m(Runner pid=2031595)[0m "image_max_pixels": 1048576,
-[36m(Runner pid=2031595)[0m "video_min_pixels": 4096,
-[36m(Runner pid=2031595)[0m "video_max_pixels": 786432,
-[36m(Runner pid=2031595)[0m "video_total_pixels": 8388608,
-[36m(Runner pid=2031595)[0m "val_video_fps": 2.0,
-[36m(Runner pid=2031595)[0m "val_video_max_frames": 128,
-[36m(Runner pid=2031595)[0m "val_video_min_pixels": 4096,
-[36m(Runner pid=2031595)[0m "val_video_max_pixels": 786432,
-[36m(Runner pid=2031595)[0m "val_video_total_pixels": 8388608,
-[36m(Runner pid=2031595)[0m "filter_overlong_prompts": false,
-[36m(Runner pid=2031595)[0m "filter_overlong_prompts_workers": 16,
-[36m(Runner pid=2031595)[0m "use_preprocessed_videos": false,
-[36m(Runner pid=2031595)[0m "video_source_mode": "realtime_only",
-[36m(Runner pid=2031595)[0m "preprocessed_video_dir": null,
-[36m(Runner pid=2031595)[0m "val_preprocessed_video_dir": null,
-[36m(Runner pid=2031595)[0m "val_video_source_mode": "realtime_only",
-[36m(Runner pid=2031595)[0m "inline_video_tensors": true,
-[36m(Runner pid=2031595)[0m "group_by_task": true,
-[36m(Runner pid=2031595)[0m "group_by_task_key": "problem_type",
-[36m(Runner pid=2031595)[0m "dataloader_num_workers": 0
-[36m(Runner pid=2031595)[0m },
-[36m(Runner pid=2031595)[0m "worker": {
-[36m(Runner pid=2031595)[0m "hybrid_engine": true,
-[36m(Runner pid=2031595)[0m "actor": {
-[36m(Runner pid=2031595)[0m "strategy": "fsdp",
-[36m(Runner pid=2031595)[0m "global_batch_size": 8,
-[36m(Runner pid=2031595)[0m "micro_batch_size_per_device_for_update": 1,
-[36m(Runner pid=2031595)[0m "micro_batch_size_per_device_for_experience": 1,
-[36m(Runner pid=2031595)[0m "max_grad_norm": 1.0,
-[36m(Runner pid=2031595)[0m "clip_ratio_low": 0.2,
-[36m(Runner pid=2031595)[0m "clip_ratio_high": 0.2,
-[36m(Runner pid=2031595)[0m "clip_ratio_dual": 3.0,
-[36m(Runner pid=2031595)[0m "loss_avg_mode": "token",
-[36m(Runner pid=2031595)[0m "loss_type": "default",
-[36m(Runner pid=2031595)[0m "ppo_epochs": 1,
-[36m(Runner pid=2031595)[0m "padding_free": true,
-[36m(Runner pid=2031595)[0m "dynamic_batching": true,
-[36m(Runner pid=2031595)[0m "max_token_len_per_gpu": 32768,
-[36m(Runner pid=2031595)[0m "ulysses_size": 1,
-[36m(Runner pid=2031595)[0m "use_torch_compile": false,
-[36m(Runner pid=2031595)[0m "model": {
-[36m(Runner pid=2031595)[0m "model_path": "/apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B",
-[36m(Runner pid=2031595)[0m "tokenizer_path": "/apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B",
-[36m(Runner pid=2031595)[0m "override_config": {},
-[36m(Runner pid=2031595)[0m "enable_gradient_checkpointing": true,
-[36m(Runner pid=2031595)[0m "trust_remote_code": false,
-[36m(Runner pid=2031595)[0m "freeze_vision_tower": true,
-[36m(Runner pid=2031595)[0m "train_vision_merger": false
-[36m(Runner pid=2031595)[0m },
-[36m(Runner pid=2031595)[0m "optim": {
-[36m(Runner pid=2031595)[0m "lr": 1e-06,
-[36m(Runner pid=2031595)[0m "betas": [
-[36m(Runner pid=2031595)[0m 0.9,
-[36m(Runner pid=2031595)[0m 0.999
-[36m(Runner pid=2031595)[0m ],
-[36m(Runner pid=2031595)[0m "weight_decay": 0.0,
-[36m(Runner pid=2031595)[0m "strategy": "adamw",
-[36m(Runner pid=2031595)[0m "lr_warmup_ratio": 0.0,
-[36m(Runner pid=2031595)[0m "lr_warmup_steps": null,
-[36m(Runner pid=2031595)[0m "min_lr_ratio": null,
-[36m(Runner pid=2031595)[0m "lr_scheduler_type": "constant",
-[36m(Runner pid=2031595)[0m "training_steps": -1
-[36m(Runner pid=2031595)[0m },
-[36m(Runner pid=2031595)[0m "fsdp": {
-[36m(Runner pid=2031595)[0m "enable_full_shard": true,
-[36m(Runner pid=2031595)[0m "enable_cpu_offload": false,
-[36m(Runner pid=2031595)[0m "enable_rank0_init": true,
-[36m(Runner pid=2031595)[0m "use_orig_params": false,
-[36m(Runner pid=2031595)[0m "torch_dtype": "bf16",
-[36m(Runner pid=2031595)[0m "fsdp_size": -1,
-[36m(Runner pid=2031595)[0m "mp_param_dtype": "bf16",
-[36m(Runner pid=2031595)[0m "mp_reduce_dtype": "fp32",
-[36m(Runner pid=2031595)[0m "mp_buffer_dtype": "fp32"
-[36m(Runner pid=2031595)[0m },
-[36m(Runner pid=2031595)[0m "offload": {
-[36m(Runner pid=2031595)[0m "offload_params": false,
-[36m(Runner pid=2031595)[0m "offload_optimizer": false
-[36m(Runner pid=2031595)[0m },
-[36m(Runner pid=2031595)[0m "global_batch_size_per_device": -1,
-[36m(Runner pid=2031595)[0m "disable_kl": true,
-[36m(Runner pid=2031595)[0m "use_kl_loss": false,
-[36m(Runner pid=2031595)[0m "kl_penalty": "low_var_kl",
-[36m(Runner pid=2031595)[0m "kl_coef": 0.0
-[36m(Runner pid=2031595)[0m },
-[36m(Runner pid=2031595)[0m "critic": {
-[36m(Runner pid=2031595)[0m "strategy": "fsdp",
-[36m(Runner pid=2031595)[0m "global_batch_size": 256,
-[36m(Runner pid=2031595)[0m "micro_batch_size_per_device_for_update": 4,
-[36m(Runner pid=2031595)[0m "micro_batch_size_per_device_for_experience": 16,
-[36m(Runner pid=2031595)[0m "max_grad_norm": 1.0,
-[36m(Runner pid=2031595)[0m "cliprange_value": 0.5,
-[36m(Runner pid=2031595)[0m "loss_avg_mode": "token",
-[36m(Runner pid=2031595)[0m "ppo_epochs": 1,
-[36m(Runner pid=2031595)[0m "padding_free": false,
-[36m(Runner pid=2031595)[0m "dynamic_batching": true,
-[36m(Runner pid=2031595)[0m "max_token_len_per_gpu": null,
-[36m(Runner pid=2031595)[0m "ulysses_size": 1,
-[36m(Runner pid=2031595)[0m "model": {
-[36m(Runner pid=2031595)[0m "model_path": null,
-[36m(Runner pid=2031595)[0m "tokenizer_path": null,
-[36m(Runner pid=2031595)[0m "override_config": {},
-[36m(Runner pid=2031595)[0m "enable_gradient_checkpointing": true,
-[36m(Runner pid=2031595)[0m "trust_remote_code": true,
-[36m(Runner pid=2031595)[0m "freeze_vision_tower": false,
-[36m(Runner pid=2031595)[0m "train_vision_merger": false
-[36m(Runner pid=2031595)[0m },
-[36m(Runner pid=2031595)[0m "optim": {
-[36m(Runner pid=2031595)[0m "lr": 1e-06,
-[36m(Runner pid=2031595)[0m "betas": [
-[36m(Runner pid=2031595)[0m 0.9,[36m(Runner pid=2031595)[0m The `use_fast` parameter is deprecated and will be removed in a future version. Use `backend="torchvision"` instead of `use_fast=True`, or `backend="pil"` instead of `use_fast=False`.
-
-[36m(Runner pid=2031595)[0m 0.999
-[36m(Runner pid=2031595)[0m ],
-[36m(Runner pid=2031595)[0m "weight_decay": 0.01,
-[36m(Runner pid=2031595)[0m "strategy": "adamw",
-[36m(Runner pid=2031595)[0m "lr_warmup_ratio": 0.0,
-[36m(Runner pid=2031595)[0m "lr_warmup_steps": null,
-[36m(Runner pid=2031595)[0m "min_lr_ratio": null,
-[36m(Runner pid=2031595)[0m "lr_scheduler_type": "constant",
-[36m(Runner pid=2031595)[0m "training_steps": -1
-[36m(Runner pid=2031595)[0m },
-[36m(Runner pid=2031595)[0m "fsdp": {
-[36m(Runner pid=2031595)[0m "enable_full_shard": true,
-[36m(Runner pid=2031595)[0m "enable_cpu_offload": false,
-[36m(Runner pid=2031595)[0m "enable_rank0_init": true,
-[36m(Runner pid=2031595)[0m "use_orig_params": false,
-[36m(Runner pid=2031595)[0m "torch_dtype": null,
-[36m(Runner pid=2031595)[0m "fsdp_size": -1,
-[36m(Runner pid=2031595)[0m "mp_param_dtype": "bf16",
-[36m(Runner pid=2031595)[0m "mp_reduce_dtype": "fp32",
-[36m(Runner pid=2031595)[0m "mp_buffer_dtype": "fp32"
-[36m(Runner pid=2031595)[0m },
-[36m(Runner pid=2031595)[0m "offload": {
-[36m(Runner pid=2031595)[0m "offload_params": false,
-[36m(Runner pid=2031595)[0m "offload_optimizer": false
-[36m(Runner pid=2031595)[0m },
-[36m(Runner pid=2031595)[0m "global_batch_size_per_device": -1
-[36m(Runner pid=2031595)[0m },
-[36m(Runner pid=2031595)[0m "ref": {
-[36m(Runner pid=2031595)[0m "strategy": "fsdp",
-[36m(Runner pid=2031595)[0m "fsdp": {
-[36m(Runner pid=2031595)[0m "enable_full_shard": true,
-[36m(Runner pid=2031595)[0m "enable_cpu_offload": false,
-[36m(Runner pid=2031595)[0m "enable_rank0_init": true,
-[36m(Runner pid=2031595)[0m "use_orig_params": false,
-[36m(Runner pid=2031595)[0m "torch_dtype": "bf16",
-[36m(Runner pid=2031595)[0m "fsdp_size": -1,
-[36m(Runner pid=2031595)[0m "mp_param_dtype": "bf16",
-[36m(Runner pid=2031595)[0m "mp_reduce_dtype": "fp32",
-[36m(Runner pid=2031595)[0m "mp_buffer_dtype": "fp32"
-[36m(Runner pid=2031595)[0m },
-[36m(Runner pid=2031595)[0m "offload": {
-[36m(Runner pid=2031595)[0m "offload_params": false,
-[36m(Runner pid=2031595)[0m "offload_optimizer": false
-[36m(Runner pid=2031595)[0m },
-[36m(Runner pid=2031595)[0m "micro_batch_size_per_device_for_experience": 1,
-[36m(Runner pid=2031595)[0m "padding_free": true,
-[36m(Runner pid=2031595)[0m "dynamic_batching": true,
-[36m(Runner pid=2031595)[0m "max_token_len_per_gpu": 32768,
-[36m(Runner pid=2031595)[0m "ulysses_size": 1,
-[36m(Runner pid=2031595)[0m "use_torch_compile": false
-[36m(Runner pid=2031595)[0m },
-[36m(Runner pid=2031595)[0m "reward": {
-[36m(Runner pid=2031595)[0m "reward_function": "orarl.rewards",
-[36m(Runner pid=2031595)[0m "reward_function_kwargs": {},
-[36m(Runner pid=2031595)[0m "skip_special_tokens": true,
-[36m(Runner pid=2031595)[0m "num_cpus": 1,
-[36m(Runner pid=2031595)[0m "reward_function_name": "compute_score",
-[36m(Runner pid=2031595)[0m "reward_function_is_module": true
-[36m(Runner pid=2031595)[0m },
-[36m(Runner pid=2031595)[0m "rollout": {
-[36m(Runner pid=2031595)[0m "name": "vllm",
-[36m(Runner pid=2031595)[0m "n": 8,
-[36m(Runner pid=2031595)[0m "temperature": 1.0,
-[36m(Runner pid=2031595)[0m "top_p": 0.85,
-[36m(Runner pid=2031595)[0m "top_k": -1,
-[36m(Runner pid=2031595)[0m "seed": 42,
-[36m(Runner pid=2031595)[0m "limit_images": 10,
-[36m(Runner pid=2031595)[0m "dtype": "bf16",
-[36m(Runner pid=2031595)[0m "gpu_memory_utilization": 0.45,
-[36m(Runner pid=2031595)[0m "ignore_eos": false,
-[36m(Runner pid=2031595)[0m "enforce_eager": false,
-[36m(Runner pid=2031595)[0m "enable_chunked_prefill": true,
-[36m(Runner pid=2031595)[0m "tensor_parallel_size": 1,
-[36m(Runner pid=2031595)[0m "max_model_len": null,
-[36m(Runner pid=2031595)[0m "max_num_batched_tokens": 32768,
-[36m(Runner pid=2031595)[0m "disable_log_stats": true,
-[36m(Runner pid=2031595)[0m "disable_tqdm": true,
-[36m(Runner pid=2031595)[0m "val_override_config": {
-[36m(Runner pid=2031595)[0m "temperature": 0.7,
-[36m(Runner pid=2031595)[0m "top_p": 0.95,
-[36m(Runner pid=2031595)[0m "n": 1
-[36m(Runner pid=2031595)[0m },
-[36m(Runner pid=2031595)[0m "kv_cache_dtype": "auto",
-[36m(Runner pid=2031595)[0m "calculate_log_probs": false,
-[36m(Runner pid=2031595)[0m "collect_seq_logprob_for_filter": false,
-[36m(Runner pid=2031595)[0m "prompt_length": 24576,
-[36m(Runner pid=2031595)[0m "response_length": 4096,
-[36m(Runner pid=2031595)[0m "trust_remote_code": false
-[36m(Runner pid=2031595)[0m }
-[36m(Runner pid=2031595)[0m },
-[36m(Runner pid=2031595)[0m "algorithm": {
-[36m(Runner pid=2031595)[0m "name": "grpo",
-[36m(Runner pid=2031595)[0m "gamma": 1.0,
-[36m(Runner pid=2031595)[0m "lam": 1.0,
-[36m(Runner pid=2031595)[0m "adv_estimator": "grpo",
-[36m(Runner pid=2031595)[0m "scale_rewards": true,
-[36m(Runner pid=2031595)[0m "disable_kl": true,
-[36m(Runner pid=2031595)[0m "use_kl_loss": false,
-[36m(Runner pid=2031595)[0m "kl_penalty": "low_var_kl",
-[36m(Runner pid=2031595)[0m "kl_coef": 0.0,
-[36m(Runner pid=2031595)[0m "kl_type": "fixed",
-[36m(Runner pid=2031595)[0m "kl_horizon": 10000.0,
-[36m(Runner pid=2031595)[0m "kl_target": 0.1
-[36m(Runner pid=2031595)[0m },
-[36m(Runner pid=2031595)[0m "trainer": {
-[36m(Runner pid=2031595)[0m "total_epochs": 1,
-[36m(Runner pid=2031595)[0m "max_steps": 1,
-[36m(Runner pid=2031595)[0m "project_name": "OraRL",
-[36m(Runner pid=2031595)[0m "experiment_name": "smoke-grpo-9b",
-[36m(Runner pid=2031595)[0m "logger": [
-[36m(Runner pid=2031595)[0m "console"
-[36m(Runner pid=2031595)[0m ],
-[36m(Runner pid=2031595)[0m "nnodes": 1,
-[36m(Runner pid=2031595)[0m "n_gpus_per_node": 8,
-[36m(Runner pid=2031595)[0m "critic_warmup": 0,
-[36m(Runner pid=2031595)[0m "val_freq": -1,
-[36m(Runner pid=2031595)[0m "val_before_train": false,
-[36m(Runner pid=2031595)[0m "val_only": false,
-[36m(Runner pid=2031595)[0m "val_generations_to_log": 0,
-[36m(Runner pid=2031595)[0m "save_freq": 1,
-[36m(Runner pid=2031595)[0m "save_limit": 3,
-[36m(Runner pid=2031595)[0m "save_model_only": false,
-[36m(Runner pid=2031595)[0m "keep_optim_only_latest": true,
-[36m(Runner pid=2031595)[0m "save_checkpoint_path": "/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/grpo",
-[36m(Runner pid=2031595)[0m "load_checkpoint_path": null,
-[36m(Runner pid=2031595)[0m "ray_timeline": null,
-[36m(Runner pid=2031595)[0m "find_last_checkpoint": false,
-[36m(Runner pid=2031595)[0m "keep_best_train_ckpt": false,
-[36m(Runner pid=2031595)[0m "best_train_metric_key": "reward/overall",
-[36m(Runner pid=2031595)[0m "best_train_smooth_window": 5,
-[36m(Runner pid=2031595)[0m "best_train_min_step": 10
-[36m(Runner pid=2031595)[0m }
-[36m(Runner pid=2031595)[0m }
-[36m(Runner pid=2031595)[0m [TaskGroupedBatchSampler] 15 groups: image_sequence_mc_answer_only(10304), obj_appearance_order(64), object_abs_distance(1344), object_counting(960), object_rel_direction(448), object_rel_distance(640), object_size_estimation(1408), room_size_estimation(960), route_planning(960), segmentation(12032), spatial grounding(7040), spatial-temporal grounding(9536), temporal grounding(20096), tracking(13952), video_qa_mc(20288) | batch_size=8 | total_batches=12504
-[36m(Runner pid=2031595)[0m Size of train dataloader: 12504
-[36m(Runner pid=2031595)[0m Size of val dataloader: 32
-[36m(Runner pid=2031595)[0m KL is disabled, no KL metrics will be logged. Please set `kl_coef=0` to log KL metrics.[36m(pid=2033394)[0m /data/miniconda3/envs/orarl/lib/python3.11/site-packages/transformers/utils/import_utils.py:1006: FutureWarning: `is_flash_attn_greater_or_equal_2_10` is deprecated and will be removed in v5.8. Please use `is_flash_attn_greater_or_equal(library_version='2.1.0')` instead if needed.
-[36m(pid=2033394)[0m warnings.warn(
-[36m(pid=2033896)[0m /data/miniconda3/envs/orarl/lib/python3.11/site-packages/transformers/utils/import_utils.py:1006: FutureWarning: `is_flash_attn_greater_or_equal_2_10` is deprecated and will be removed in v5.8. Please use `is_flash_attn_greater_or_equal(library_version='2.1.0')` instead if needed.
-[36m(pid=2033896)[0m warnings.warn(
-[36m(pid=2033897)[0m /data/miniconda3/envs/orarl/lib/python3.11/site-packages/transformers/utils/import_utils.py:1006: FutureWarning: `is_flash_attn_greater_or_equal_2_10` is deprecated and will be removed in v5.8. Please use `is_flash_attn_greater_or_equal(library_version='2.1.0')` instead if needed.
-[36m(pid=2033897)[0m warnings.warn(
-[36m(WorkerDict pid=2033898)[0m [W825 18:00:56.039186514 ProcessGroupNCCL.cpp:929] Warning: TORCH_NCCL_AVOID_RECORD_STREAMS is the default now, this environment variable is thus deprecated. (function operator())
-[36m(pid=2033901)[0m /data/miniconda3/envs/orarl/lib/python3.11/site-packages/transformers/utils/import_utils.py:1006: FutureWarning: `is_flash_attn_greater_or_equal_2_10` is deprecated and will be removed in v5.8. Please use `is_flash_attn_greater_or_equal(library_version='2.1.0')` instead if needed.[32m [repeated 5x across cluster] (Ray deduplicates logs by default. Set RAY_DEDUP_LOGS=0 to disable log deduplication, or see https://docs.ray.io/en/master/ray-observability/user-guides/configure-logging.html#log-deduplication for more options.)[0m
-[36m(pid=2033901)[0m warnings.warn([32m [repeated 5x across cluster][0m
-[36m(WorkerDict pid=2033898)[0m The `use_fast` parameter is deprecated and will be removed in a future version. Use `backend="torchvision"` instead of `use_fast=True`, or `backend="pil"` instead of `use_fast=False`.
-[36m(WorkerDict pid=2033898)[0m `torch_dtype` is deprecated! Use `dtype` instead!
-[36m(WorkerDict pid=2033899)[0m [W825 18:00:58.504154548 ProcessGroupNCCL.cpp:929] Warning: TORCH_NCCL_AVOID_RECORD_STREAMS is the default now, this environment variable is thus deprecated. (function operator())[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2033898)[0m The fast path is not available because one of the required library is not installed. Falling back to torch implementation. To install follow https://github.com/fla-org/flash-linear-attention#installation and https://github.com/Dao-AILab/causal-conv1d
-[36m(WorkerDict pid=2033898)[0m [rank3]:[W825 18:01:02.812976524 ProcessGroupNCCL.cpp:5138] Guessing device ID based on global rank. This can cause a hang if rank to GPU mapping is heterogeneous. You can specify device_id in init_process_group()
-[36m(WorkerDict pid=2033394)[0m The `use_fast` parameter is deprecated and will be removed in a future version. Use `backend="torchvision"` instead of `use_fast=True`, or `backend="pil"` instead of `use_fast=False`.[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2033394)[0m
Loading weights: 0%| | 0/760 [00:00, ?it/s]
Loading weights: 100%|██████████| 760/760 [00:00<00:00, 10448.47it/s]
-[36m(WorkerDict pid=2033394)[0m /data/miniconda3/envs/orarl/lib/python3.11/site-packages/torch/distributed/c10d_logger.py:83: UserWarning: barrier(): using the device under current context. You can specify `device_id` in `init_process_group` to mute this warning.
-[36m(WorkerDict pid=2033394)[0m return func(*args, **kwargs)
-[36m(WorkerDict pid=2033896)[0m `Qwen2VLImageProcessorFast` is deprecated. The `Fast` suffix for image processors has been removed; use `Qwen2VLImageProcessor` instead.
-[36m(WorkerDict pid=2033899)[0m `torch_dtype` is deprecated! Use `dtype` instead![32m [repeated 6x across cluster][0m
-[36m(WorkerDict pid=2033899)[0m The fast path is not available because one of the required library is not installed. Falling back to torch implementation. To install follow https://github.com/fla-org/flash-linear-attention#installation and https://github.com/Dao-AILab/causal-conv1d[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2033899)[0m [rank4]:[W825 18:01:05.292398250 ProcessGroupNCCL.cpp:5138] Guessing device ID based on global rank. This can cause a hang if rank to GPU mapping is heterogeneous. You can specify device_id in init_process_group()[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2033896)[0m :1241: FutureWarning: The cuda.cudart module is deprecated and will be removed in a future release, please switch to use the cuda.bindings.runtime module instead.
-[36m(WorkerDict pid=2033896)[0m :1241: FutureWarning: The cuda.nvrtc module is deprecated and will be removed in a future release, please switch to use the cuda.bindings.nvrtc module instead.
-[36m(WorkerDict pid=2033394)[0m `Qwen2VLImageProcessorFast` is deprecated. The `Fast` suffix for image processors has been removed; use `Qwen2VLImageProcessor` instead.[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2033897)[0m 2026-08-25 18:02:00,379 - INFO - autotuner.py:262 - flashinfer.jit: [Autotuner]: Autotuning process starts ...
-[36m(WorkerDict pid=2033902)[0m :1241: FutureWarning: The cuda.cudart module is deprecated and will be removed in a future release, please switch to use the cuda.bindings.runtime module instead.[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2033902)[0m :1241: FutureWarning: The cuda.nvrtc module is deprecated and will be removed in a future release, please switch to use the cuda.bindings.nvrtc module instead.[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2033897)[0m 2026-08-25 18:02:00,394 - INFO - autotuner.py:268 - flashinfer.jit: [Autotuner]: Autotuning process ends
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 0%| | 0/51 [00:00, ?it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 2%|▏ | 1/51 [00:00<00:05, 9.37it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 4%|▍ | 2/51 [00:00<00:05, 9.22it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 8%|▊ | 4/51 [00:00<00:04, 10.28it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 12%|█▏ | 6/51 [00:00<00:04, 11.03it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 16%|█▌ | 8/51 [00:00<00:03, 11.47it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 20%|█▉ | 10/51 [00:00<00:03, 11.38it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 24%|██▎ | 12/51 [00:01<00:03, 11.34it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 27%|██▋ | 14/51 [00:01<00:03, 11.36it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 31%|███▏ | 16/51 [00:01<00:02, 11.99it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 35%|███▌ | 18/51 [00:01<00:02, 11.84it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 39%|███▉ | 20/51 [00:01<00:02, 12.37it/s]
-[36m(WorkerDict pid=2033898)[0m 2026-08-25 18:02:00,394 - INFO - autotuner.py:262 - flashinfer.jit: [Autotuner]: Autotuning process starts ...[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2033898)[0m 2026-08-25 18:02:00,406 - INFO - autotuner.py:268 - flashinfer.jit: [Autotuner]: Autotuning process ends[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 43%|████▎ | 22/51 [00:01<00:02, 12.30it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 47%|████▋ | 24/51 [00:02<00:02, 12.79it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 51%|█████ | 26/51 [00:02<00:02, 11.79it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 55%|█████▍ | 28/51 [00:02<00:01, 11.50it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 59%|█████▉ | 30/51 [00:02<00:01, 11.95it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 63%|██████▎ | 32/51 [00:02<00:01, 12.85it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 67%|██████▋ | 34/51 [00:02<00:01, 12.62it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 71%|███████ | 36/51 [00:03<00:01, 12.60it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 75%|███████▍ | 38/51 [00:03<00:00, 13.01it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 78%|███████▊ | 40/51 [00:03<00:00, 13.53it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 82%|████████▏ | 42/51 [00:03<00:00, 13.58it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 86%|████████▋ | 44/51 [00:03<00:00, 11.08it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 90%|█████████ | 46/51 [00:03<00:00, 12.19it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 94%|█████████▍| 48/51 [00:03<00:00, 13.01it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 98%|█████████▊| 50/51 [00:04<00:00, 13.66it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 100%|██████████| 51/51 [00:04<00:00, 12.33it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 0%| | 0/51 [00:00, ?it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 2%|▏ | 1/51 [00:00<00:18, 2.72it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 4%|▍ | 2/51 [00:00<00:12, 3.95it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 6%|▌ | 3/51 [00:00<00:09, 5.28it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 8%|▊ | 4/51 [00:00<00:07, 6.05it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 10%|▉ | 5/51 [00:00<00:07, 6.53it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 12%|█▏ | 6/51 [00:01<00:06, 7.14it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 14%|█▎ | 7/51 [00:01<00:05, 7.44it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 18%|█▊ | 9/51 [00:01<00:04, 8.49it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 22%|██▏ | 11/51 [00:01<00:04, 9.59it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 25%|██▌ | 13/51 [00:01<00:03, 10.73it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 29%|██▉ | 15/51 [00:01<00:03, 11.18it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 33%|███▎ | 17/51 [00:01<00:02, 11.96it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 37%|███▋ | 19/51 [00:02<00:02, 12.55it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 41%|████ | 21/51 [00:02<00:02, 13.68it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 45%|████▌ | 23/51 [00:02<00:02, 13.59it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 49%|████▉ | 25/51 [00:02<00:01, 14.38it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 53%|█████▎ | 27/51 [00:02<00:01, 15.27it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 59%|█████▉ | 30/51 [00:02<00:01, 17.49it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 65%|██████▍ | 33/51 [00:02<00:00, 18.60it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 71%|███████ | 36/51 [00:03<00:00, 20.94it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 78%|███████▊ | 40/51 [00:03<00:00, 23.69it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 86%|████████▋ | 44/51 [00:03<00:00, 25.25it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 94%|█████████▍| 48/51 [00:03<00:00, 27.71it/s]
-[36m(WorkerDict pid=2033394)[0m
Capturing CUDA graphs (decode, FULL): 100%|██████████| 51/51 [00:03<00:00, 14.65it/s]
-
-[36m(Runner pid=2031595)[0m [GRPO] Scaled advantages ON: A=(r-mean)/(std+eps).
-[36m(Runner pid=2031595)[0m Total training steps: 1
-[36m(AutoRewardManager pid=2033189)[0m Using reward function `compute_score` from `orarl.rewards`.
-[36m(AutoRewardManager pid=2033189)[0m Reward name: orarl, reward type: batch.
-[36m(WorkerDict pid=2033394)[0m actor will use global batch size 64.
-[36m(AutoRewardManager pid=2033186)[0m Using reward function `compute_score` from `orarl.rewards`.
-[36m(AutoRewardManager pid=2033186)[0m Reward name: orarl, reward type: batch.
-[36m(WorkerDict pid=2033394)[0m Model config: Qwen3_5Config {
-[36m(WorkerDict pid=2033394)[0m "architectures": [
-[36m(WorkerDict pid=2033394)[0m "Qwen3_5ForConditionalGeneration"
-[36m(WorkerDict pid=2033394)[0m ],
-[36m(WorkerDict pid=2033394)[0m "dtype": "bfloat16",
-[36m(WorkerDict pid=2033394)[0m "image_token_id": 248056,
-[36m(WorkerDict pid=2033394)[0m "model_type": "qwen3_5",
-[36m(WorkerDict pid=2033394)[0m "text_config": {
-[36m(WorkerDict pid=2033394)[0m "attention_bias": false,
-[36m(WorkerDict pid=2033394)[0m "attention_dropout": 0.0,
-[36m(WorkerDict pid=2033394)[0m "attn_output_gate": true,
-[36m(WorkerDict pid=2033394)[0m "bos_token_id": null,
-[36m(WorkerDict pid=2033394)[0m "dtype": "bfloat16",
-[36m(WorkerDict pid=2033394)[0m "eos_token_id": 248044,
-[36m(WorkerDict pid=2033394)[0m "full_attention_interval": 4,
-[36m(WorkerDict pid=2033394)[0m "head_dim": 256,
-[36m(WorkerDict pid=2033394)[0m "hidden_act": "silu",
-[36m(WorkerDict pid=2033394)[0m "hidden_size": 4096,
-[36m(WorkerDict pid=2033394)[0m "initializer_range": 0.02,
-[36m(WorkerDict pid=2033394)[0m "intermediate_size": 12288,
-[36m(WorkerDict pid=2033394)[0m "layer_types": [
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "full_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "full_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "full_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "full_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "full_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "full_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "full_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "linear_attention",
-[36m(WorkerDict pid=2033394)[0m "full_attention"
-[36m(WorkerDict pid=2033394)[0m ],
-[36m(WorkerDict pid=2033394)[0m "linear_conv_kernel_dim": 4,
-[36m(WorkerDict pid=2033394)[0m "linear_key_head_dim": 128,
-[36m(WorkerDict pid=2033394)[0m "linear_num_key_heads": 16,
-[36m(WorkerDict pid=2033394)[0m "linear_num_value_heads": 32,
-[36m(WorkerDict pid=2033394)[0m "linear_value_head_dim": 128,
-[36m(WorkerDict pid=2033394)[0m "mamba_ssm_dtype": "float32",
-[36m(WorkerDict pid=2033394)[0m "max_position_embeddings": 262144,
-[36m(WorkerDict pid=2033394)[0m "mlp_only_layers": [],
-[36m(WorkerDict pid=2033394)[0m "model_type": "qwen3_5_text",
-[36m(WorkerDict pid=2033394)[0m "mtp_num_hidden_layers": 0,
-[36m(WorkerDict pid=2033394)[0m "mtp_use_dedicated_embeddings": false,
-[36m(WorkerDict pid=2033394)[0m "num_attention_heads": 16,
-[36m(WorkerDict pid=2033394)[0m "num_hidden_layers": 32,
-[36m(WorkerDict pid=2033394)[0m "num_key_value_heads": 4,
-[36m(WorkerDict pid=2033394)[0m "pad_token_id": null,
-[36m(WorkerDict pid=2033394)[0m "partial_rotary_factor": 0.25,
-[36m(WorkerDict pid=2033394)[0m "rms_norm_eps": 1e-06,
-[36m(WorkerDict pid=2033394)[0m "rope_parameters": {
-[36m(WorkerDict pid=2033394)[0m "mrope_interleaved": true,
-[36m(WorkerDict pid=2033394)[0m "mrope_section": [
-[36m(WorkerDict pid=2033394)[0m 11,
-[36m(WorkerDict pid=2033394)[0m 11,
-[36m(WorkerDict pid=2033394)[0m 10
-[36m(WorkerDict pid=2033394)[0m ],
-[36m(WorkerDict pid=2033394)[0m "partial_rotary_factor": 0.25,
-[36m(WorkerDict pid=2033394)[0m "rope_theta": 10000000,
-[36m(WorkerDict pid=2033394)[0m "rope_type": "default"
-[36m(WorkerDict pid=2033394)[0m },
-[36m(WorkerDict pid=2033394)[0m "tie_word_embeddings": false,
-[36m(WorkerDict pid=2033394)[0m "use_cache": true,
-[36m(WorkerDict pid=2033394)[0m "vocab_size": 248320
-[36m(WorkerDict pid=2033394)[0m },
-[36m(WorkerDict pid=2033394)[0m "tie_word_embeddings": false,
-[36m(WorkerDict pid=2033394)[0m "transformers_version": "5.5.4",
-[36m(WorkerDict pid=2033394)[0m "video_token_id": 248057,
-[36m(WorkerDict pid=2033394)[0m "vision_config": {
-[36m(WorkerDict pid=2033394)[0m "deepstack_visual_indexes": [],
-[36m(WorkerDict pid=2033394)[0m "depth": 27,
-[36m(WorkerDict pid=2033394)[0m "dtype": "bfloat16",
-[36m(WorkerDict pid=2033394)[0m "hidden_act": "gelu_pytorch_tanh",
-[36m(WorkerDict pid=2033394)[0m "hidden_size": 1152,
-[36m(WorkerDict pid=2033394)[0m "in_channels": 3,
-[36m(WorkerDict pid=2033394)[0m "initializer_range": 0.02,
-[36m(WorkerDict pid=2033394)[0m "intermediate_size": 4304,
-[36m(WorkerDict pid=2033394)[0m "model_type": "qwen3_5",
-[36m(WorkerDict pid=2033394)[0m "num_heads": 16,
-[36m(WorkerDict pid=2033394)[0m "num_position_embeddings": 2304,
-[36m(WorkerDict pid=2033394)[0m "out_hidden_size": 4096,
-[36m(WorkerDict pid=2033394)[0m "patch_size": 16,
-[36m(WorkerDict pid=2033394)[0m "spatial_merge_size": 2,
-[36m(WorkerDict pid=2033394)[0m "temporal_patch_size": 2
-[36m(WorkerDict pid=2033394)[0m },
-[36m(WorkerDict pid=2033394)[0m "vision_end_token_id": 248054,
-[36m(WorkerDict pid=2033394)[0m "vision_start_token_id": 248053
-[36m(WorkerDict pid=2033394)[0m }
-[36m(WorkerDict pid=2033394)[0m
-[36m(WorkerDict pid=2033394)[0m Ulysses patch applied!
-[36m(WorkerDict pid=2033394)[0m Vision backbone is set to not trainable.
-[36m(WorkerDict pid=2033394)[0m Qwen3_5ForConditionalGeneration contains 9.41B parameters.
-[36m(WorkerDict pid=2033394)[0m After huggingface model init: 1.17 GB / 95.00 GB.
-[36m(WorkerDict pid=2033394)[0m FSDP wrap policy: functools.partial(, transformer_layer_cls={, }).
-[36m(WorkerDict pid=2033394)[0m After FSDP module init: 15.42 GB / 95.00 GB.
-[36m(WorkerDict pid=2033394)[0m After optimizer init: 15.42 GB / 95.00 GB.
-[36m(WorkerDict pid=2033899)[0m Sampling params: {'max_tokens': 4096, 'detokenize': False, 'logit_bias': {248056: -100, 248057: -100}, 'n': 8, 'temperature': 1.0, 'top_p': 0.85, 'top_k': -1, 'ignore_eos': False}.
-[36m(Runner pid=2031595)[0m Config
-[36m(Runner pid=2031595)[0m algorithm:
-[36m(Runner pid=2031595)[0m adv_estimator: grpo
-[36m(Runner pid=2031595)[0m disable_kl: true
-[36m(Runner pid=2031595)[0m gamma: 1.0
-[36m(Runner pid=2031595)[0m kl_coef: 0.0
-[36m(Runner pid=2031595)[0m kl_horizon: 10000.0
-[36m(Runner pid=2031595)[0m kl_penalty: low_var_kl
-[36m(Runner pid=2031595)[0m kl_target: 0.1
-[36m(Runner pid=2031595)[0m kl_type: fixed
-[36m(Runner pid=2031595)[0m lam: 1.0
-[36m(Runner pid=2031595)[0m name: grpo
-[36m(Runner pid=2031595)[0m scale_rewards: true
-[36m(Runner pid=2031595)[0m use_kl_loss: false
-[36m(Runner pid=2031595)[0m data:
-[36m(Runner pid=2031595)[0m answer_key: answer
-[36m(Runner pid=2031595)[0m dataloader_num_workers: 0
-[36m(Runner pid=2031595)[0m enable_thinking: false
-[36m(Runner pid=2031595)[0m filter_overlong_prompts: false
-[36m(Runner pid=2031595)[0m filter_overlong_prompts_workers: 16
-[36m(Runner pid=2031595)[0m format_prompt: null
-[36m(Runner pid=2031595)[0m group_by_task: true
-[36m(Runner pid=2031595)[0m group_by_task_key: problem_type
-[36m(Runner pid=2031595)[0m image_dir: /
-[36m(Runner pid=2031595)[0m image_key: images
-[36m(Runner pid=2031595)[0m image_max_pixels: 1048576
-[36m(Runner pid=2031595)[0m image_min_pixels: 4096
-[36m(Runner pid=2031595)[0m inline_video_tensors: true
-[36m(Runner pid=2031595)[0m max_pixels: 4194304
-[36m(Runner pid=2031595)[0m max_prompt_length: 24576
-[36m(Runner pid=2031595)[0m max_response_length: 4096
-[36m(Runner pid=2031595)[0m min_pixels: 4096
-[36m(Runner pid=2031595)[0m mini_rollout_batch_size: null
-[36m(Runner pid=2031595)[0m override_chat_template: null
-[36m(Runner pid=2031595)[0m preprocessed_video_dir: null
-[36m(Runner pid=2031595)[0m prompt_key: problem
-[36m(Runner pid=2031595)[0m response_prefix: ''
-[36m(Runner pid=2031595)[0m rollout_batch_size: 8
-[36m(Runner pid=2031595)[0m seed: 42
-[36m(Runner pid=2031595)[0m shuffle: true
-[36m(Runner pid=2031595)[0m train_files: /mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_joint_100k_maskaware.jsonl
-[36m(Runner pid=2031595)[0m use_preprocessed_videos: false
-[36m(Runner pid=2031595)[0m val_batch_size: 16
-[36m(Runner pid=2031595)[0m val_files: /mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_phase3_49152.canary.jsonl
-[36m(Runner pid=2031595)[0m val_preprocessed_video_dir: null
-[36m(Runner pid=2031595)[0m val_video_fps: 2.0
-[36m(Runner pid=2031595)[0m val_video_max_frames: 128
-[36m(Runner pid=2031595)[0m val_video_max_pixels: 786432
-[36m(Runner pid=2031595)[0m val_video_min_pixels: 4096
-[36m(Runner pid=2031595)[0m val_video_source_mode: realtime_only
-[36m(Runner pid=2031595)[0m val_video_total_pixels: 8388608
-[36m(Runner pid=2031595)[0m video_fps: 2.0
-[36m(Runner pid=2031595)[0m video_key: videos
-[36m(Runner pid=2031595)[0m video_max_frames: 128
-[36m(Runner pid=2031595)[0m video_max_pixels: 786432
-[36m(Runner pid=2031595)[0m video_min_pixels: 4096
-[36m(Runner pid=2031595)[0m video_source_mode: realtime_only
-[36m(Runner pid=2031595)[0m video_total_pixels: 8388608
-[36m(Runner pid=2031595)[0m trainer:
-[36m(Runner pid=2031595)[0m best_train_metric_key: reward/overall
-[36m(Runner pid=2031595)[0m best_train_min_step: 10
-[36m(Runner pid=2031595)[0m best_train_smooth_window: 5
-[36m(Runner pid=2031595)[0m critic_warmup: 0
-[36m(Runner pid=2031595)[0m experiment_name: smoke-grpo-9b
-[36m(Runner pid=2031595)[0m find_last_checkpoint: false
-[36m(Runner pid=2031595)[0m keep_best_train_ckpt: false
-[36m(Runner pid=2031595)[0m keep_optim_only_latest: true
-[36m(Runner pid=2031595)[0m load_checkpoint_path: null
-[36m(Runner pid=2031595)[0m logger:
-[36m(Runner pid=2031595)[0m - console
-[36m(Runner pid=2031595)[0m max_steps: 1
-[36m(Runner pid=2031595)[0m n_gpus_per_node: 8
-[36m(Runner pid=2031595)[0m nnodes: 1
-[36m(Runner pid=2031595)[0m project_name: OraRL
-[36m(Runner pid=2031595)[0m ray_timeline: null
-[36m(Runner pid=2031595)[0m save_checkpoint_path: /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/grpo
-[36m(Runner pid=2031595)[0m save_freq: 1
-[36m(Runner pid=2031595)[0m save_limit: 3
-[36m(Runner pid=2031595)[0m save_model_only: false
-[36m(Runner pid=2031595)[0m total_epochs: 1
-[36m(Runner pid=2031595)[0m val_before_train: false
-[36m(Runner pid=2031595)[0m val_freq: -1
-[36m(Runner pid=2031595)[0m val_generations_to_log: 0
-[36m(Runner pid=2031595)[0m val_only: false
-[36m(Runner pid=2031595)[0m worker:
-[36m(Runner pid=2031595)[0m actor:
-[36m(Runner pid=2031595)[0m clip_ratio_dual: 3.0
-[36m(Runner pid=2031595)[0m clip_ratio_high: 0.2
-[36m(Runner pid=2031595)[0m clip_ratio_low: 0.2
-[36m(Runner pid=2031595)[0m disable_kl: true
-[36m(Runner pid=2031595)[0m dynamic_batching: true
-[36m(Runner pid=2031595)[0m fsdp:
-[36m(Runner pid=2031595)[0m enable_cpu_offload: false
-[36m(Runner pid=2031595)[0m enable_full_shard: true
-[36m(Runner pid=2031595)[0m enable_rank0_init: true
-[36m(Runner pid=2031595)[0m fsdp_size: -1
-[36m(Runner pid=2031595)[0m mp_buffer_dtype: fp32
-[36m(Runner pid=2031595)[0m mp_param_dtype: bf16
-[36m(Runner pid=2031595)[0m mp_reduce_dtype: fp32
-[36m(Runner pid=2031595)[0m torch_dtype: bf16
-[36m(Runner pid=2031595)[0m use_orig_params: false
-[36m(Runner pid=2031595)[0m global_batch_size: 8
-[36m(Runner pid=2031595)[0m global_batch_size_per_device: -1
-[36m(Runner pid=2031595)[0m kl_coef: 0.0
-[36m(Runner pid=2031595)[0m kl_penalty: low_var_kl
-[36m(Runner pid=2031595)[0m loss_avg_mode: token
-[36m(Runner pid=2031595)[0m loss_type: default
-[36m(Runner pid=2031595)[0m max_grad_norm: 1.0
-[36m(Runner pid=2031595)[0m max_token_len_per_gpu: 32768
-[36m(Runner pid=2031595)[0m micro_batch_size_per_device_for_experience: 1
-[36m(Runner pid=2031595)[0m micro_batch_size_per_device_for_update: 1
-[36m(Runner pid=2031595)[0m model:
-[36m(Runner pid=2031595)[0m enable_gradient_checkpointing: true
-[36m(Runner pid=2031595)[0m freeze_vision_tower: true
-[36m(Runner pid=2031595)[0m model_path: /apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B
-[36m(Runner pid=2031595)[0m override_config: {}
-[36m(Runner pid=2031595)[0m tokenizer_path: /apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B
-[36m(Runner pid=2031595)[0m train_vision_merger: false
-[36m(Runner pid=2031595)[0m trust_remote_code: false
-[36m(Runner pid=2031595)[0m offload:
-[36m(Runner pid=2031595)[0m offload_optimizer: false
-[36m(Runner pid=2031595)[0m offload_params: false
-[36m(Runner pid=2031595)[0m optim:
-[36m(Runner pid=2031595)[0m betas:
-[36m(Runner pid=2031595)[0m - 0.9
-[36m(Runner pid=2031595)[0m - 0.999
-[36m(Runner pid=2031595)[0m lr: 1.0e-06
-[36m(Runner pid=2031595)[0m lr_scheduler_type: constant
-[36m(Runner pid=2031595)[0m lr_warmup_ratio: 0.0
-[36m(Runner pid=2031595)[0m lr_warmup_steps: null
-[36m(Runner pid=2031595)[0m min_lr_ratio: null
-[36m(Runner pid=2031595)[0m strategy: adamw
-[36m(Runner pid=2031595)[0m training_steps: 1
-[36m(Runner pid=2031595)[0m weight_decay: 0.0
-[36m(Runner pid=2031595)[0m padding_free: true
-[36m(Runner pid=2031595)[0m ppo_epochs: 1
-[36m(Runner pid=2031595)[0m strategy: fsdp
-[36m(Runner pid=2031595)[0m ulysses_size: 1
-[36m(Runner pid=2031595)[0m use_kl_loss: false
-[36m(Runner pid=2031595)[0m use_torch_compile: false
-[36m(Runner pid=2031595)[0m critic:
-[36m(Runner pid=2031595)[0m cliprange_value: 0.5
-[36m(Runner pid=2031595)[0m dynamic_batching: true
-[36m(Runner pid=2031595)[0m fsdp:
-[36m(Runner pid=2031595)[0m enable_cpu_offload: false
-[36m(Runner pid=2031595)[0m enable_full_shard: true
-[36m(Runner pid=2031595)[0m enable_rank0_init: true
-[36m(Runner pid=2031595)[0m fsdp_size: -1
-[36m(Runner pid=2031595)[0m mp_buffer_dtype: fp32
-[36m(Runner pid=2031595)[0m mp_param_dtype: bf16
-[36m(Runner pid=2031595)[0m mp_reduce_dtype: fp32
-[36m(Runner pid=2031595)[0m torch_dtype: null
-[36m(Runner pid=2031595)[0m use_orig_params: false
-[36m(Runner pid=2031595)[0m global_batch_size: 256
-[36m(Runner pid=2031595)[0m global_batch_size_per_device: -1
-[36m(Runner pid=2031595)[0m loss_avg_mode: token
-[36m(Runner pid=2031595)[0m max_grad_norm: 1.0
-[36m(Runner pid=2031595)[0m max_token_len_per_gpu: null
-[36m(Runner pid=2031595)[0m micro_batch_size_per_device_for_experience: 16
-[36m(Runner pid=2031595)[0m micro_batch_size_per_device_for_update: 4
-[36m(Runner pid=2031595)[0m model:
-[36m(Runner pid=2031595)[0m enable_gradient_checkpointing: true
-[36m(Runner pid=2031595)[0m freeze_vision_tower: false
-[36m(Runner pid=2031595)[0m model_path: null
-[36m(Runner pid=2031595)[0m override_config: {}
-[36m(Runner pid=2031595)[0m tokenizer_path: null
-[36m(Runner pid=2031595)[0m train_vision_merger: false
-[36m(Runner pid=2031595)[0m trust_remote_code: true
-[36m(Runner pid=2031595)[0m offload:
-[36m(Runner pid=2031595)[0m offload_optimizer: false
-[36m(Runner pid=2031595)[0m offload_params: false
-[36m(Runner pid=2031595)[0m optim:
-[36m(Runner pid=2031595)[0m betas:
-[36m(Runner pid=2031595)[0m - 0.9
-[36m(Runner pid=2031595)[0m - 0.999
-[36m(Runner pid=2031595)[0m lr: 1.0e-06
-[36m(Runner pid=2031595)[0m lr_scheduler_type: constant
-[36m(Runner pid=2031595)[0m lr_warmup_ratio: 0.0
-[36m(Runner pid=2031595)[0m lr_warmup_steps: null
-[36m(Runner pid=2031595)[0m min_lr_ratio: null
-[36m(Runner pid=2031595)[0m strategy: adamw
-[36m(Runner pid=2031595)[0m training_steps: 1
-[36m(Runner pid=2031595)[0m weight_decay: 0.01
-[36m(Runner pid=2031595)[0m padding_free: false
-[36m(Runner pid=2031595)[0m ppo_epochs: 1
-[36m(Runner pid=2031595)[0m strategy: fsdp
-[36m(Runner pid=2031595)[0m ulysses_size: 1
-[36m(Runner pid=2031595)[0m hybrid_engine: true
-[36m(Runner pid=2031595)[0m ref:
-[36m(Runner pid=2031595)[0m dynamic_batching: true
-[36m(Runner pid=2031595)[0m fsdp:
-[36m(Runner pid=2031595)[0m enable_cpu_offload: false
-[36m(Runner pid=2031595)[0m enable_full_shard: true
-[36m(Runner pid=2031595)[0m enable_rank0_init: true
-[36m(Runner pid=2031595)[0m fsdp_size: -1
-[36m(Runner pid=2031595)[0m mp_buffer_dtype: fp32
-[36m(Runner pid=2031595)[0m mp_param_dtype: bf16
-[36m(Runner pid=2031595)[0m mp_reduce_dtype: fp32
-[36m(Runner pid=2031595)[0m torch_dtype: bf16
-[36m(Runner pid=2031595)[0m use_orig_params: false
-[36m(Runner pid=2031595)[0m max_token_len_per_gpu: 32768
-[36m(Runner pid=2031595)[0m micro_batch_size_per_device_for_experience: 1
-[36m(Runner pid=2031595)[0m offload:
-[36m(Runner pid=2031595)[0m offload_optimizer: false
-[36m(Runner pid=2031595)[0m offload_params: false
-[36m(Runner pid=2031595)[0m padding_free: true
-[36m(Runner pid=2031595)[0m strategy: fsdp
-[36m(Runner pid=2031595)[0m ulysses_size: 1
-[36m(Runner pid=2031595)[0m use_torch_compile: false
-[36m(Runner pid=2031595)[0m reward:
-[36m(Runner pid=2031595)[0m num_cpus: 1
-[36m(Runner pid=2031595)[0m reward_function: orarl.rewards
-[36m(Runner pid=2031595)[0m reward_function_is_module: true
-[36m(Runner pid=2031595)[0m reward_function_kwargs: {}
-[36m(Runner pid=2031595)[0m reward_function_name: compute_score
-[36m(Runner pid=2031595)[0m skip_special_tokens: true
-[36m(Runner pid=2031595)[0m rollout:
-[36m(Runner pid=2031595)[0m calculate_log_probs: false
-[36m(Runner pid=2031595)[0m collect_seq_logprob_for_filter: false
-[36m(Runner pid=2031595)[0m disable_log_stats: true
-[36m(Runner pid=2031595)[0m disable_tqdm: true
-[36m(Runner pid=2031595)[0m dtype: bf16
-[36m(Runner pid=2031595)[0m enable_chunked_prefill: true
-[36m(Runner pid=2031595)[0m enforce_eager: false
-[36m(Runner pid=2031595)[0m gpu_memory_utilization: 0.45
-[36m(Runner pid=2031595)[0m ignore_eos: false
-[36m(Runner pid=2031595)[0m kv_cache_dtype: auto
-[36m(Runner pid=2031595)[0m limit_images: 10
-[36m(Runner pid=2031595)[0m max_model_len: null
-[36m(Runner pid=2031595)[0m max_num_batched_tokens: 32768
-[36m(Runner pid=2031595)[0m n: 8
-[36m(Runner pid=2031595)[0m name: vllm
-[36m(Runner pid=2031595)[0m prompt_length: 24576
-[36m(Runner pid=2031595)[0m response_length: 4096
-[36m(Runner pid=2031595)[0m seed: 42
-[36m(Runner pid=2031595)[0m temperature: 1.0
-[36m(Runner pid=2031595)[0m tensor_parallel_size: 1
-[36m(Runner pid=2031595)[0m top_k: -1
-[36m(Runner pid=2031595)[0m top_p: 0.85
-[36m(Runner pid=2031595)[0m trust_remote_code: false
-[36m(Runner pid=2031595)[0m val_override_config:
-[36m(Runner pid=2031595)[0m n: 1
-[36m(Runner pid=2031595)[0m temperature: 0.7
-[36m(Runner pid=2031595)[0m top_p: 0.95
-[36m(Runner pid=2031595)[0m
-
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [00:00, ?it/s]
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [00:00, ?it/s]
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [00:00, ?it/s]
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [00:01, ?it/s]
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [00:02, ?it/s]
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [00:02, ?it/s]
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [00:02, ?it/s]
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [00:22, ?it/s]
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [00:22, ?it/s]
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [00:22, ?it/s][36m(WorkerDict pid=2033394)[0m After vllm init: 7.01 GB / 95.00 GB.
-[36m(WorkerDict pid=2033394)[0m Before vllm wake up in sharding manager: 7.01 GB / 95.00 GB.
-[36m(WorkerDict pid=2033394)[0m After gather model weights in sharding manager: 30.92 GB / 95.00 GB.
-[36m(WorkerDict pid=2033394)[0m After sync model weights in sharding manager: 24.69 GB / 95.00 GB.
-[36m(WorkerDict pid=2033394)[0m After vllm wake up in sharding manager: 45.04 GB / 95.00 GB.
-[36m(Runner pid=2031595)[0m Start generating batch...
-[36m(WorkerDict pid=2033394)[0m Before vllm offload in sharding manager: 50.43 GB / 95.00 GB.
-[36m(WorkerDict pid=2033902)[0m Sampling params: {'max_tokens': 4096, 'detokenize': False, 'logit_bias': {248056: -100, 248057: -100}, 'n': 8, 'temperature': 1.0, 'top_p': 0.85, 'top_k': -1, 'ignore_eos': False}.[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2033394)[0m After vllm offload in sharding manager: 7.40 GB / 95.00 GB.
-
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 0%| | 0.00/2.00 [00:00, ?it/s][A[A
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [00:14<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [00:14<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 0%| | 0.00/1.00 [00:00, ?it/s][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 0%| | 0.00/2.00 [00:00, ?it/s][A[A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [01:22<00:00, 41.2s/it][A[A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [01:22<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A[36m(WorkerDict pid=2033394)[0m /data/miniconda3/envs/orarl/lib/python3.11/site-packages/torch/distributed/c10d_logger.py:83: UserWarning: barrier(): using the device under current context. You can specify `device_id` in `init_process_group` to mute this warning.
-
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [02:35, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:05<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [01:44<00:00, 105s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [01:44<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A[36m(WorkerDict pid=2033394)[0m return func(*args, **kwargs)
-
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [02:35, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:05<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [01:44<00:00, 105s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [01:44<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [02:43, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:12<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [01:52<00:00, 112s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [01:52<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [02:43, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:12<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [01:52<00:00, 112s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [01:52<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [02:43, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:12<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [01:52<00:00, 112s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [01:52<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A[36m(WorkerDict pid=2033394)[0m /data/miniconda3/envs/orarl/lib/python3.11/site-packages/torch/distributed/c10d_logger.py:83: UserWarning: barrier(): using the device under current context. You can specify `device_id` in `init_process_group` to mute this warning.
-
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:32, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:02<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:41<00:00, 222s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:41<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A[36m(WorkerDict pid=2033394)[0m return func(*args, **kwargs)
-
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:32, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:02<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:41<00:00, 222s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:41<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 0%| | 0.00/1.00 [04:34, ?it/s]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
[2m[36m(pid=2031595) [0mRunning step: 100%|██████████| 1.00/1.00 [04:34<00:00, 275s/it]
[2m[36m(pid=2031595) [0mRunning step: 100%|██████████| 1.00/1.00 [04:34<00:00, 275s/it]
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 100%|██████████| 1.00/1.00 [04:34<00:00, 275s/it]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 100%|██████████| 1.00/1.00 [04:34<00:00, 275s/it]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2031595) [0mRunning step: 100%|██████████| 1.00/1.00 [04:34<00:00, 275s/it]
-
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:04<00:00, 7.28s/it][A[A
-
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:43<00:00, 224s/it][A[A
-
-
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:43<00:00, 41.2s/it][A[A[A[36m(WorkerDict pid=2033394)[0m [rank-0]: Saving model to /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_0.pt.
-[36m(WorkerDict pid=2033394)[0m [rank-0]: Saving optimizer to /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_0.pt.
-[36m(WorkerDict pid=2033394)[0m [rank-0]: Saving extra_state to /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_0.pt.
-[36m(Runner pid=2031595)[0m [TRAIN STEP] 1/1 step_s=274.7 elapsed=4m34s eta=0m00s
-[36m(Runner pid=2031595)[0m Step 1
-[36m(Runner pid=2031595)[0m actor:
-[36m(Runner pid=2031595)[0m backward_rows: 64.0
-[36m(Runner pid=2031595)[0m backward_rows_per_prompt: 8.0
-[36m(Runner pid=2031595)[0m entropy_loss: 0.112
-[36m(Runner pid=2031595)[0m grad_norm: 7.625
-[36m(Runner pid=2031595)[0m lr: 1.0e-06
-[36m(Runner pid=2031595)[0m pg_clipfrac_higher: 0.0
-[36m(Runner pid=2031595)[0m pg_clipfrac_lower: 0.0
-[36m(Runner pid=2031595)[0m pg_loss: 3.725290298461914e-09
-[36m(Runner pid=2031595)[0m ppo_kl: 0.0
-[36m(Runner pid=2031595)[0m critic:
-[36m(Runner pid=2031595)[0m advantages:
-[36m(Runner pid=2031595)[0m max: 1.62
-[36m(Runner pid=2031595)[0m mean: -4.789658891724002e-09
-[36m(Runner pid=2031595)[0m min: -1.62
-[36m(Runner pid=2031595)[0m std: 0.74
-[36m(Runner pid=2031595)[0m returns:
-[36m(Runner pid=2031595)[0m max: 1.62
-[36m(Runner pid=2031595)[0m mean: -4.789658891724002e-09
-[36m(Runner pid=2031595)[0m min: -1.62
-[36m(Runner pid=2031595)[0m rewards:
-[36m(Runner pid=2031595)[0m max: 1.0
-[36m(Runner pid=2031595)[0m mean: 0.578
-[36m(Runner pid=2031595)[0m min: 0.0
-[36m(Runner pid=2031595)[0m std: 0.494
-[36m(Runner pid=2031595)[0m score:
-[36m(Runner pid=2031595)[0m max: 1.0
-[36m(Runner pid=2031595)[0m mean: 0.578
-[36m(Runner pid=2031595)[0m min: 0.0
-[36m(Runner pid=2031595)[0m std: 0.494
-[36m(Runner pid=2031595)[0m var: 0.244
-[36m(Runner pid=2031595)[0m global_seqlen:
-[36m(Runner pid=2031595)[0m balanced_max: 54644
-[36m(Runner pid=2031595)[0m balanced_min: 54644
-[36m(Runner pid=2031595)[0m max: 68488
-[36m(Runner pid=2031595)[0m mean: 54644.0
-[36m(Runner pid=2031595)[0m min: 18296
-[36m(Runner pid=2031595)[0m minmax_diff: 50192
-[36m(Runner pid=2031595)[0m perf:
-[36m(Runner pid=2031595)[0m cpu_memory_used_gb: 1091.043
-[36m(Runner pid=2031595)[0m max_memory_allocated_gb: 69.333
-[36m(Runner pid=2031595)[0m max_memory_reserved_gb: 83.51
-[36m(Runner pid=2031595)[0m mfu_actor: 0.217
-[36m(Runner pid=2031595)[0m throughput: 198.904
-[36m(Runner pid=2031595)[0m time_per_step: 274.726
-[36m(Runner pid=2031595)[0m total_num_tokens: 437152
-[36m(Runner pid=2031595)[0m prompt_length:
-[36m(Runner pid=2031595)[0m clip_ratio: 0.0
-[36m(Runner pid=2031595)[0m max: 8554.0
-[36m(Runner pid=2031595)[0m mean: 6823.5
-[36m(Runner pid=2031595)[0m min: 2280.0
-[36m(Runner pid=2031595)[0m response_length:
-[36m(Runner pid=2031595)[0m clip_ratio: 0.0
-[36m(Runner pid=2031595)[0m max: 7.0
-[36m(Runner pid=2031595)[0m mean: 7.0
-[36m(Runner pid=2031595)[0m min: 7.0
-[36m(Runner pid=2031595)[0m reward:
-[36m(Runner pid=2031595)[0m accuracy: 0.578
-[36m(Runner pid=2031595)[0m format: 1.0
-[36m(Runner pid=2031595)[0m overall: 0.578
-[36m(Runner pid=2031595)[0m rollout:
-[36m(Runner pid=2031595)[0m all_identical_group_fraction: 0.375
-[36m(Runner pid=2031595)[0m exact_unique_fraction: 0.281
-[36m(Runner pid=2031595)[0m pairwise_token_agreement_mean: 0.943
-[36m(Runner pid=2031595)[0m response_length_std_mean: 0.0
-[36m(Runner pid=2031595)[0m reward_group_std_mean: 0.29
-[36m(Runner pid=2031595)[0m reward_unique_fraction: 0.203
-[36m(Runner pid=2031595)[0m zero_reward_std_group_fraction: 0.375
-[36m(Runner pid=2031595)[0m timing_per_token_ms:
-[36m(Runner pid=2031595)[0m adv: 7.939586091993053e-06
-[36m(Runner pid=2031595)[0m gen: 51.144
-[36m(Runner pid=2031595)[0m old: 0.064
-[36m(Runner pid=2031595)[0m reward: 0.008
-[36m(Runner pid=2031595)[0m update_actor: 0.239
-[36m(Runner pid=2031595)[0m timing_s:
-[36m(Runner pid=2031595)[0m adv: 0.003
-[36m(Runner pid=2031595)[0m gen: 22.913
-[36m(Runner pid=2031595)[0m old: 28.067
-[36m(Runner pid=2031595)[0m reward: 0.004
-[36m(Runner pid=2031595)[0m save_checkpoint: 119.155
-[36m(Runner pid=2031595)[0m step: 274.726
-[36m(Runner pid=2031595)[0m update_actor: 104.535
-[36m(Runner pid=2031595)[0m
-[36m(WorkerDict pid=2033899)[0m [rank-4]: Saving model to /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/model_world_size_8_rank_4.pt.[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2033899)[0m [rank-4]: Saving optimizer to /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/optim_world_size_8_rank_4.pt.[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2033899)[0m [rank-4]: Saving extra_state to /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/grpo/global_step_1/actor/extra_state_world_size_8_rank_4.pt.[32m [repeated 7x across cluster][0m
-
[2m[36m(pid=2031595) [0mRunning step: 100%|██████████| 1.00/1.00 [04:39<00:00, 280s/it]
-
[2m[36m(pid=2033394) [0mCompute log probs: 100%|██████████| 2.00/2.00 [04:09<00:00, 125s/it]
-
[2m[36m(pid=2033394) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [03:48<00:00, 229s/it]
-
[2m[36m(pid=2033394) [0mUpdate policy: 100%|██████████| 2.00/2.00 [03:48<00:00, 114s/it]
-method=grpo mode=run
-/data/miniconda3/envs/orarl/bin/python -m verl.trainer.main config=/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/configs/grpo_9b.yaml trainer.max_steps=1 trainer.save_freq=1 trainer.val_before_train=false trainer.experiment_name=smoke-grpo-9b data.rollout_batch_size=8 worker.actor.global_batch_size=8 worker.actor.model.model_path=/apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B worker.actor.model.tokenizer_path=/apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B data.train_files=/mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_joint_100k_maskaware.jsonl data.val_files=/mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_phase3_49152.canary.jsonl trainer.save_checkpoint_path=/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/grpo trainer.nnodes=1 trainer.n_gpus_per_node=8
-checkpoint OK: grpo -> /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/grpo/global_step_1
-
-=== orarl smoke (1 update(s)) ===
-/data/miniconda3/envs/orarl/lib/python3.11/site-packages/transformers/utils/import_utils.py:1006: FutureWarning: `is_flash_attn_greater_or_equal_2_10` is deprecated and will be removed in v5.8. Please use `is_flash_attn_greater_or_equal(library_version='2.1.0')` instead if needed.
- warnings.warn(
-2026-08-25 18:07:46,220 INFO worker.py:2004 -- Started a local Ray instance. View the dashboard at [1m[32mhttp://127.0.0.1:8265 [39m[22m
-/data/miniconda3/envs/orarl/lib/python3.11/site-packages/ray/_private/worker.py:2052: FutureWarning: Tip: In future versions of Ray, Ray will no longer override accelerator visible devices env var if num_gpus=0 or num_gpus=None (default). To enable this behavior and turn off this error message, set RAY_ACCEL_ENV_VAR_OVERRIDE_ON_ZERO=0
- warnings.warn(
-[36m(pid=2088024)[0m W0825 18:08:01.584000 2088024 site-packages/torch/utils/cpp_extension.py:118] No CUDA runtime is found, using CUDA_HOME='/data/miniconda3/envs/orarl'
-[36m(Runner pid=2088024)[0m {
-[36m(Runner pid=2088024)[0m "data": {
-[36m(Runner pid=2088024)[0m "train_files": "/mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_joint_100k_maskaware.jsonl",
-[36m(Runner pid=2088024)[0m "val_files": "/mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_phase3_49152.canary.jsonl",
-[36m(Runner pid=2088024)[0m "prompt_key": "problem",
-[36m(Runner pid=2088024)[0m "answer_key": "answer",
-[36m(Runner pid=2088024)[0m "image_key": "images",
-[36m(Runner pid=2088024)[0m "video_key": "videos",
-[36m(Runner pid=2088024)[0m "image_dir": "/",
-[36m(Runner pid=2088024)[0m "video_fps": 2.0,
-[36m(Runner pid=2088024)[0m "video_max_frames": 128,
-[36m(Runner pid=2088024)[0m "max_prompt_length": 24576,
-[36m(Runner pid=2088024)[0m "max_response_length": 4096,
-[36m(Runner pid=2088024)[0m "rollout_batch_size": 8,
-[36m(Runner pid=2088024)[0m "mini_rollout_batch_size": null,
-[36m(Runner pid=2088024)[0m "val_batch_size": 16,
-[36m(Runner pid=2088024)[0m "format_prompt": null,
-[36m(Runner pid=2088024)[0m "override_chat_template": null,
-[36m(Runner pid=2088024)[0m "enable_thinking": false,
-[36m(Runner pid=2088024)[0m "response_prefix": "",
-[36m(Runner pid=2088024)[0m "shuffle": true,
-[36m(Runner pid=2088024)[0m "seed": 42,
-[36m(Runner pid=2088024)[0m "min_pixels": 4096,
-[36m(Runner pid=2088024)[0m "max_pixels": 4194304,
-[36m(Runner pid=2088024)[0m "image_min_pixels": 4096,
-[36m(Runner pid=2088024)[0m "image_max_pixels": 1048576,
-[36m(Runner pid=2088024)[0m "video_min_pixels": 4096,
-[36m(Runner pid=2088024)[0m "video_max_pixels": 786432,
-[36m(Runner pid=2088024)[0m "video_total_pixels": 8388608,
-[36m(Runner pid=2088024)[0m "val_video_fps": 2.0,
-[36m(Runner pid=2088024)[0m "val_video_max_frames": 128,
-[36m(Runner pid=2088024)[0m "val_video_min_pixels": 4096,
-[36m(Runner pid=2088024)[0m "val_video_max_pixels": 786432,
-[36m(Runner pid=2088024)[0m "val_video_total_pixels": 8388608,
-[36m(Runner pid=2088024)[0m "filter_overlong_prompts": false,
-[36m(Runner pid=2088024)[0m "filter_overlong_prompts_workers": 16,
-[36m(Runner pid=2088024)[0m "use_preprocessed_videos": false,
-[36m(Runner pid=2088024)[0m "video_source_mode": "realtime_only",
-[36m(Runner pid=2088024)[0m "preprocessed_video_dir": null,
-[36m(Runner pid=2088024)[0m "val_preprocessed_video_dir": null,
-[36m(Runner pid=2088024)[0m "val_video_source_mode": "realtime_only",
-[36m(Runner pid=2088024)[0m "inline_video_tensors": true,
-[36m(Runner pid=2088024)[0m "group_by_task": true,
-[36m(Runner pid=2088024)[0m "group_by_task_key": "problem_type",
-[36m(Runner pid=2088024)[0m "dataloader_num_workers": 0
-[36m(Runner pid=2088024)[0m },
-[36m(Runner pid=2088024)[0m "worker": {
-[36m(Runner pid=2088024)[0m "hybrid_engine": true,
-[36m(Runner pid=2088024)[0m "actor": {
-[36m(Runner pid=2088024)[0m "strategy": "fsdp",
-[36m(Runner pid=2088024)[0m "global_batch_size": 8,
-[36m(Runner pid=2088024)[0m "micro_batch_size_per_device_for_update": 1,
-[36m(Runner pid=2088024)[0m "micro_batch_size_per_device_for_experience": 1,
-[36m(Runner pid=2088024)[0m "max_grad_norm": 1.0,
-[36m(Runner pid=2088024)[0m "clip_ratio_low": 0.2,
-[36m(Runner pid=2088024)[0m "clip_ratio_high": 0.2,
-[36m(Runner pid=2088024)[0m "clip_ratio_dual": 3.0,
-[36m(Runner pid=2088024)[0m "loss_avg_mode": "token",
-[36m(Runner pid=2088024)[0m "loss_type": "default",
-[36m(Runner pid=2088024)[0m "ppo_epochs": 1,
-[36m(Runner pid=2088024)[0m "padding_free": true,
-[36m(Runner pid=2088024)[0m "dynamic_batching": true,
-[36m(Runner pid=2088024)[0m "max_token_len_per_gpu": 32768,
-[36m(Runner pid=2088024)[0m "ulysses_size": 1,
-[36m(Runner pid=2088024)[0m "use_torch_compile": false,
-[36m(Runner pid=2088024)[0m "model": {
-[36m(Runner pid=2088024)[0m "model_path": "/apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B",
-[36m(Runner pid=2088024)[0m "tokenizer_path": "/apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B",
-[36m(Runner pid=2088024)[0m "override_config": {},
-[36m(Runner pid=2088024)[0m "enable_gradient_checkpointing": true,
-[36m(Runner pid=2088024)[0m "trust_remote_code": false,
-[36m(Runner pid=2088024)[0m "freeze_vision_tower": true,
-[36m(Runner pid=2088024)[0m "train_vision_merger": false
-[36m(Runner pid=2088024)[0m },
-[36m(Runner pid=2088024)[0m "optim": {
-[36m(Runner pid=2088024)[0m "lr": 1e-06,
-[36m(Runner pid=2088024)[0m "betas": [
-[36m(Runner pid=2088024)[0m 0.9,
-[36m(Runner pid=2088024)[0m 0.999
-[36m(Runner pid=2088024)[0m ],
-[36m(Runner pid=2088024)[0m "weight_decay": 0.0,
-[36m(Runner pid=2088024)[0m "strategy": "adamw",
-[36m(Runner pid=2088024)[0m "lr_warmup_ratio": 0.0,
-[36m(Runner pid=2088024)[0m "lr_warmup_steps": null,
-[36m(Runner pid=2088024)[0m "min_lr_ratio": null,
-[36m(Runner pid=2088024)[0m "lr_scheduler_type": "constant",
-[36m(Runner pid=2088024)[0m "training_steps": -1
-[36m(Runner pid=2088024)[0m },
-[36m(Runner pid=2088024)[0m "fsdp": {
-[36m(Runner pid=2088024)[0m "enable_full_shard": true,
-[36m(Runner pid=2088024)[0m "enable_cpu_offload": false,
-[36m(Runner pid=2088024)[0m "enable_rank0_init": true,
-[36m(Runner pid=2088024)[0m "use_orig_params": false,
-[36m(Runner pid=2088024)[0m "torch_dtype": "bf16",
-[36m(Runner pid=2088024)[0m "fsdp_size": -1,
-[36m(Runner pid=2088024)[0m "mp_param_dtype": "bf16",
-[36m(Runner pid=2088024)[0m "mp_reduce_dtype": "fp32",
-[36m(Runner pid=2088024)[0m "mp_buffer_dtype": "fp32"
-[36m(Runner pid=2088024)[0m },
-[36m(Runner pid=2088024)[0m "offload": {
-[36m(Runner pid=2088024)[0m "offload_params": false,
-[36m(Runner pid=2088024)[0m "offload_optimizer": false
-[36m(Runner pid=2088024)[0m },
-[36m(Runner pid=2088024)[0m "global_batch_size_per_device": -1,
-[36m(Runner pid=2088024)[0m "disable_kl": true,
-[36m(Runner pid=2088024)[0m "use_kl_loss": false,
-[36m(Runner pid=2088024)[0m "kl_penalty": "low_var_kl",
-[36m(Runner pid=2088024)[0m "kl_coef": 0.0,
-[36m(Runner pid=2088024)[0m "selection_prune_ratio": 0.5
-[36m(Runner pid=2088024)[0m },
-[36m(Runner pid=2088024)[0m "critic": {
-[36m(Runner pid=2088024)[0m "strategy": "fsdp",
-[36m(Runner pid=2088024)[0m "global_batch_size": 256,
-[36m(Runner pid=2088024)[0m "micro_batch_size_per_device_for_update": 4,
-[36m(Runner pid=2088024)[0m "micro_batch_size_per_device_for_experience": 16,
-[36m(Runner pid=2088024)[0m "max_grad_norm": 1.0,
-[36m(Runner pid=2088024)[0m "cliprange_value": 0.5,
-[36m(Runner pid=2088024)[0m "loss_avg_mode": "token",
-[36m(Runner pid=2088024)[0m "ppo_epochs": 1,
-[36m(Runner pid=2088024)[0m "padding_free": false,
-[36m(Runner pid=2088024)[0m "dynamic_batching": true,
-[36m(Runner pid=2088024)[0m "max_token_len_per_gpu": null,
-[36m(Runner pid=2088024)[0m "ulysses_size": 1,
-[36m(Runner pid=2088024)[0m "model": {
-[36m(Runner pid=2088024)[0m "model_path": null,
-[36m(Runner pid=2088024)[0m "tokenizer_path": null,
-[36m(Runner pid=2088024)[0m "override_config": {},
-[36m(Runner pid=2088024)[0m "enable_gradient_checkpointing": true,
-[36m(Runner pid=2088024)[0m "trust_remote_code": true,
-[36m(Runner pid=2088024)[0m "freeze_vision_tower": false,
-[36m(Runner pid=2088024)[0m "train_vision_merger": false
-[36m(Runner pid=2088024)[0m },
-[36m(Runner pid=2088024)[0m "optim": {
-[36m(Runner pid=2088024)[0m "lr": 1e-06,
-[36m(Runner pid=2088024)[0m "betas": [
-[36m(Runner pid=2088024)[0m 0.9,
-[36m(Runner pid=2088024)[0m 0.999
-[36m(Runner pid=2088024)[0m ],
-[36m(Runner pid=2088024)[0m "weight_decay": 0.01,
-[36m(Runner pid=2088024)[0m "strategy": "adamw",
-[36m(Runner pid=2088024)[0m "lr_warmup_ratio": 0.0,
-[36m(Runner pid=2088024)[0m "lr_warmup_steps": null,
-[36m(Runner pid=2088024)[0m "min_lr_ratio": null,
-[36m(Runner pid=2088024)[0m "lr_scheduler_type": "constant",
-[36m(Runner pid=2088024)[0m "training_steps": -1
-[36m(Runner pid=2088024)[0m },
-[36m(Runner pid=2088024)[0m "fsdp": {
-[36m(Runner pid=2088024)[0m "enable_full_shard": true,
-[36m(Runner pid=2088024)[0m "enable_cpu_offload": false,
-[36m(Runner pid=2088024)[0m "enable_rank0_init": true,
-[36m(Runner pid=2088024)[0m "use_orig_params": false,
-[36m(Runner pid=2088024)[0m "torch_dtype": null,
-[36m(Runner pid=2088024)[0m "fsdp_size": -1,
-[36m(Runner pid=2088024)[0m "mp_param_dtype": "bf16",
-[36m(Runner pid=2088024)[0m "mp_reduce_dtype": "fp32",
-[36m(Runner pid=2088024)[0m "mp_buffer_dtype": "fp32"
-[36m(Runner pid=2088024)[0m },
-[36m(Runner pid=2088024)[0m "offload": {
-[36m(Runner pid=2088024)[0m "offload_params": false,
-[36m(Runner pid=2088024)[0m "offload_optimizer": false
-[36m(Runner pid=2088024)[0m },
-[36m(Runner pid=2088024)[0m "global_batch_size_per_device": -1
-[36m(Runner pid=2088024)[0m },
-[36m(Runner pid=2088024)[0m "ref": {
-[36m(Runner pid=2088024)[0m "strategy": "fsdp",
-[36m(Runner pid=2088024)[0m "fsdp": {
-[36m(Runner pid=2088024)[0m "enable_full_shard": true,
-[36m(Runner pid=2088024)[0m "enable_cpu_offload": false,
-[36m(Runner pid=2088024)[0m "enable_rank0_init": true,
-[36m(Runner pid=2088024)[0m "use_orig_params": false,
-[36m(Runner pid=2088024)[0m "torch_dtype": "bf16",
-[36m(Runner pid=2088024)[0m "fsdp_size": -1,
-[36m(Runner pid=2088024)[0m "mp_param_dtype": "bf16",
-[36m(Runner pid=2088024)[0m "mp_reduce_dtype": "fp32",
-[36m(Runner pid=2088024)[0m "mp_buffer_dtype": "fp32"
-[36m(Runner pid=2088024)[0m },
-[36m(Runner pid=2088024)[0m "offload": {
-[36m(Runner pid=2088024)[0m "offload_params": false,
-[36m(Runner pid=2088024)[0m "offload_optimizer": false
-[36m(Runner pid=2088024)[0m },
-[36m(Runner pid=2088024)[0m "micro_batch_size_per_device_for_experience": 1,
-[36m(Runner pid=2088024)[0m "padding_free": true,
-[36m(Runner pid=2088024)[0m "dynamic_batching": true,
-[36m(Runner pid=2088024)[0m "max_token_len_per_gpu": 32768,
-[36m(Runner pid=2088024)[0m "ulysses_size": 1,
-[36m(Runner pid=2088024)[0m "use_torch_compile": false
-[36m(Runner pid=2088024)[0m },
-[36m(Runner pid=2088024)[0m "reward": {
-[36m(Runner pid=2088024)[0m "reward_function": "orarl.rewards",
-[36m(Runner pid=2088024)[0m "reward_function_kwargs": {},
-[36m(Runner pid=2088024)[0m "skip_special_tokens": true,
-[36m(Runner pid=2088024)[0m "num_cpus": 1,
-[36m(Runner pid=2088024)[0m "reward_function_name": "compute_score",
-[36m(Runner pid=2088024)[0m "reward_function_is_module": true
-[36m(Runner pid=2088024)[0m },
-[36m(Runner pid=2088024)[0m "rollout": {
-[36m(Runner pid=2088024)[0m "name": "vllm",
-[36m(Runner pid=2088024)[0m "n": 8,
-[36m(Runner pid=2088024)[0m "temperature": 1.0,
-[36m(Runner pid=2088024)[0m "top_p": 0.85,
-[36m(Runner pid=2088024)[0m "top_k": -1,
-[36m(Runner pid=2088024)[0m "seed": 42,
-[36m(Runner pid=2088024)[0m "limit_images": 10,
-[36m(Runner pid=2088024)[0m "dtype": "bf16",
-[36m(Runner pid=2088024)[0m "gpu_memory_utilization": 0.45,
-[36m(Runner pid=2088024)[0m "ignore_eos": false,
-[36m(Runner pid=2088024)[0m "enforce_eager": false,
-[36m(Runner pid=2088024)[0m "enable_chunked_prefill": true,
-[36m(Runner pid=2088024)[0m "tensor_parallel_size": 1,
-[36m(Runner pid=2088024)[0m "max_model_len": null,
-[36m(Runner pid=2088024)[0m "max_num_batched_tokens": 32768,
-[36m(Runner pid=2088024)[0m "disable_log_stats": true,
-[36m(Runner pid=2088024)[0m "disable_tqdm": true,
-[36m(Runner pid=2088024)[0m "val_override_config": {
-[36m(Runner pid=2088024)[0m "temperature": 0.7,
-[36m(Runner pid=2088024)[0m "top_p": 0.95,
-[36m(Runner pid=2088024)[0m "n": 1
-[36m(Runner pid=2088024)[0m },
-[36m(Runner pid=2088024)[0m "kv_cache_dtype": "auto",
-[36m(Runner pid=2088024)[0m "calculate_log_probs": false,
-[36m(Runner pid=2088024)[0m "collect_seq_logprob_for_filter": false,
-[36m(Runner pid=2088024)[0m "prompt_length": 24576,
-[36m(Runner pid=2088024)[0m "response_length": 4096,
-[36m(Runner pid=2088024)[0m "trust_remote_code": false
-[36m(Runner pid=2088024)[0m }
-[36m(Runner pid=2088024)[0m },
-[36m(Runner pid=2088024)[0m "algorithm": {
-[36m(Runner pid=2088024)[0m "name": "orarl",
-[36m(Runner pid=2088024)[0m "gamma": 1.0,
-[36m(Runner pid=2088024)[0m "lam": 1.0,
-[36m(Runner pid=2088024)[0m "adv_estimator": "grpo",
-[36m(Runner pid=2088024)[0m "scale_rewards": false,
-[36m(Runner pid=2088024)[0m "disable_kl": true,
-[36m(Runner pid=2088024)[0m "use_kl_loss": false,
-[36m(Runner pid=2088024)[0m "kl_penalty": "low_var_kl",
-[36m(Runner pid=2088024)[0m "kl_coef": 0.0,
-[36m(Runner pid=2088024)[0m "kl_type": "fixed",
-[36m(Runner pid=2088024)[0m "kl_horizon": 10000.0,
-[36m(Runner pid=2088024)[0m "kl_target": 0.1,
-[36m(Runner pid=2088024)[0m "oracle_injection": true,
-[36m(Runner pid=2088024)[0m "oracle_injection_mode": "append",
-[36m(Runner pid=2088024)[0m "oracle_builder": "orarl.rewards:build_oracle_response_from_ground_truth",
-[36m(Runner pid=2088024)[0m "oracle_replace_index": -1,
-[36m(Runner pid=2088024)[0m "oracle_log_exclude": true,
-[36m(Runner pid=2088024)[0m "directional_gain": true,
-[36m(Runner pid=2088024)[0m "directional_gain_gamma": 0.25,
-[36m(Runner pid=2088024)[0m "directional_gain_positive_only": true,
-[36m(Runner pid=2088024)[0m "directional_gain_recenter": true,
-[36m(Runner pid=2088024)[0m "detached_oracle_advantage": true,
-[36m(Runner pid=2088024)[0m "detached_oracle_advantage_scale": 2.0,
-[36m(Runner pid=2088024)[0m "detached_oracle_use_directional_gain": false,
-[36m(Runner pid=2088024)[0m "detached_oracle_match_best_ratio": 1.2,
-[36m(Runner pid=2088024)[0m "detached_oracle_match_best_min": 0.05,
-[36m(Runner pid=2088024)[0m "detached_oracle_match_best_max": 1.0,
-[36m(Runner pid=2088024)[0m "oracle_reward_gate_beta": 2.0,
-[36m(Runner pid=2088024)[0m "selection_prune_ratio": 0.5,
-[36m(Runner pid=2088024)[0m "selection_keep_oracle": true,
-[36m(Runner pid=2088024)[0m "selection_positive_quota": 1,
-[36m(Runner pid=2088024)[0m "selection_negative_quota": 2,
-[36m(Runner pid=2088024)[0m "selection_strict_sign_balance": true,
-[36m(Runner pid=2088024)[0m "post_selection_recenter": true,
-[36m(Runner pid=2088024)[0m "post_selection_rms_match": true,
-[36m(Runner pid=2088024)[0m "post_selection_rms_min_scale": 0.25
-[36m(Runner pid=2088024)[0m },
-[36m(Runner pid=2088024)[0m "trainer": {
-[36m(Runner pid=2088024)[0m "total_epochs": 1,
-[36m(Runner pid=2088024)[0m "max_steps": 1,
-[36m(Runner pid=2088024)[0m "project_name": "OraRL",
-[36m(Runner pid=2088024)[0m "experiment_name": "smoke-orarl-9b",
-[36m(Runner pid=2088024)[0m "logger": [
-[36m(Runner pid=2088024)[0m "console"
-[36m(Runner pid=2088024)[0m ],
-[36m(Runner pid=2088024)[0m "nnodes": 1,
-[36m(Runner pid=2088024)[0m "n_gpus_per_node": 8,
-[36m(Runner pid=2088024)[0m "critic_warmup": 0,[36m(Runner pid=2088024)[0m The `use_fast` parameter is deprecated and will be removed in a future version. Use `backend="torchvision"` instead of `use_fast=True`, or `backend="pil"` instead of `use_fast=False`.
-[36m(pid=2089852)[0m /data/miniconda3/envs/orarl/lib/python3.11/site-packages/transformers/utils/import_utils.py:1006: FutureWarning: `is_flash_attn_greater_or_equal_2_10` is deprecated and will be removed in v5.8. Please use `is_flash_attn_greater_or_equal(library_version='2.1.0')` instead if needed.
-[36m(pid=2089852)[0m warnings.warn(
-[36m(pid=2090322)[0m /data/miniconda3/envs/orarl/lib/python3.11/site-packages/transformers/utils/import_utils.py:1006: FutureWarning: `is_flash_attn_greater_or_equal_2_10` is deprecated and will be removed in v5.8. Please use `is_flash_attn_greater_or_equal(library_version='2.1.0')` instead if needed.
-[36m(pid=2090322)[0m warnings.warn(
-[36m(pid=2090324)[0m /data/miniconda3/envs/orarl/lib/python3.11/site-packages/transformers/utils/import_utils.py:1006: FutureWarning: `is_flash_attn_greater_or_equal_2_10` is deprecated and will be removed in v5.8. Please use `is_flash_attn_greater_or_equal(library_version='2.1.0')` instead if needed.
-[36m(pid=2090324)[0m warnings.warn(
-[36m(WorkerDict pid=2090328)[0m [W825 18:08:26.116194738 ProcessGroupNCCL.cpp:929] Warning: TORCH_NCCL_AVOID_RECORD_STREAMS is the default now, this environment variable is thus deprecated. (function operator())
-[36m(pid=2090327)[0m /data/miniconda3/envs/orarl/lib/python3.11/site-packages/transformers/utils/import_utils.py:1006: FutureWarning: `is_flash_attn_greater_or_equal_2_10` is deprecated and will be removed in v5.8. Please use `is_flash_attn_greater_or_equal(library_version='2.1.0')` instead if needed.[32m [repeated 5x across cluster] (Ray deduplicates logs by default. Set RAY_DEDUP_LOGS=0 to disable log deduplication, or see https://docs.ray.io/en/master/ray-observability/user-guides/configure-logging.html#log-deduplication for more options.)[0m
-[36m(pid=2090327)[0m warnings.warn([32m [repeated 5x across cluster][0m
-[36m(WorkerDict pid=2090328)[0m The `use_fast` parameter is deprecated and will be removed in a future version. Use `backend="torchvision"` instead of `use_fast=True`, or `backend="pil"` instead of `use_fast=False`.
-
-[36m(Runner pid=2088024)[0m "val_freq": -1,
-[36m(Runner pid=2088024)[0m "val_before_train": false,
-[36m(Runner pid=2088024)[0m "val_only": false,
-[36m(Runner pid=2088024)[0m "val_generations_to_log": 0,
-[36m(Runner pid=2088024)[0m "save_freq": 1,
-[36m(Runner pid=2088024)[0m "save_limit": 3,
-[36m(Runner pid=2088024)[0m "save_model_only": false,
-[36m(Runner pid=2088024)[0m "keep_optim_only_latest": true,
-[36m(Runner pid=2088024)[0m "save_checkpoint_path": "/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/orarl",
-[36m(Runner pid=2088024)[0m "load_checkpoint_path": null,
-[36m(Runner pid=2088024)[0m "ray_timeline": null,
-[36m(Runner pid=2088024)[0m "find_last_checkpoint": false,
-[36m(Runner pid=2088024)[0m "keep_best_train_ckpt": false,
-[36m(Runner pid=2088024)[0m "best_train_metric_key": "reward/overall",
-[36m(Runner pid=2088024)[0m "best_train_smooth_window": 5,
-[36m(Runner pid=2088024)[0m "best_train_min_step": 10
-[36m(Runner pid=2088024)[0m }
-[36m(Runner pid=2088024)[0m }
-[36m(Runner pid=2088024)[0m [TaskGroupedBatchSampler] 15 groups: image_sequence_mc_answer_only(10304), obj_appearance_order(64), object_abs_distance(1344), object_counting(960), object_rel_direction(448), object_rel_distance(640), object_size_estimation(1408), room_size_estimation(960), route_planning(960), segmentation(12032), spatial grounding(7040), spatial-temporal grounding(9536), temporal grounding(20096), tracking(13952), video_qa_mc(20288) | batch_size=8 | total_batches=12504
-[36m(Runner pid=2088024)[0m Size of train dataloader: 12504
-[36m(Runner pid=2088024)[0m Size of val dataloader: 32
-[36m(Runner pid=2088024)[0m KL is disabled, no KL metrics will be logged. Please set `kl_coef=0` to log KL metrics.
-[36m(Runner pid=2088024)[0m [GRPO] Raw-centered advantages ON: A=r-mean (no std whitening).
-[36m(Runner pid=2088024)[0m [OraRL] enabled | append annotation-as-rollout | directional_gain_gamma=0.25 | detached oracle | strict sign-balanced selection | post-selection moment correction
-[36m(Runner pid=2088024)[0m Total training steps: 1
-[36m(AutoRewardManager pid=2089612)[0m Using reward function `compute_score` from `orarl.rewards`.
-[36m(AutoRewardManager pid=2089612)[0m Reward name: orarl, reward type: batch.
-[36m(WorkerDict pid=2089852)[0m actor OraRL selection active (P=0.5): scaling global_batch_size with k=4 instead of n=8.
-[36m(WorkerDict pid=2089852)[0m actor will use global batch size 32.
-[36m(AutoRewardManager pid=2089614)[0m Using reward function `compute_score` from `orarl.rewards`.
-[36m(AutoRewardManager pid=2089614)[0m Reward name: orarl, reward type: batch.
-[36m(WorkerDict pid=2089852)[0m Model config: Qwen3_5Config {
-[36m(WorkerDict pid=2089852)[0m "architectures": [
-[36m(WorkerDict pid=2089852)[0m "Qwen3_5ForConditionalGeneration"
-[36m(WorkerDict pid=2089852)[0m ],
-[36m(WorkerDict pid=2089852)[0m "dtype": "bfloat16",
-[36m(WorkerDict pid=2089852)[0m "image_token_id": 248056,
-[36m(WorkerDict pid=2089852)[0m "model_type": "qwen3_5",
-[36m(WorkerDict pid=2089852)[0m "text_config": {
-[36m(WorkerDict pid=2089852)[0m "attention_bias": false,
-[36m(WorkerDict pid=2089852)[0m "attention_dropout": 0.0,
-[36m(WorkerDict pid=2089852)[0m "attn_output_gate": true,
-[36m(WorkerDict pid=2089852)[0m "bos_token_id": null,
-[36m(WorkerDict pid=2089852)[0m "dtype": "bfloat16",
-[36m(WorkerDict pid=2089852)[0m "eos_token_id": 248044,
-[36m(WorkerDict pid=2089852)[0m "full_attention_interval": 4,
-[36m(WorkerDict pid=2089852)[0m "head_dim": 256,
-[36m(WorkerDict pid=2089852)[0m "hidden_act": "silu",
-[36m(WorkerDict pid=2089852)[0m "hidden_size": 4096,
-[36m(WorkerDict pid=2089852)[0m "initializer_range": 0.02,
-[36m(WorkerDict pid=2089852)[0m "intermediate_size": 12288,
-[36m(WorkerDict pid=2089852)[0m "layer_types": [
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "full_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "full_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "full_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "full_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "full_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "full_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "full_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "linear_attention",
-[36m(WorkerDict pid=2089852)[0m "full_attention"
-[36m(WorkerDict pid=2089852)[0m ],
-[36m(WorkerDict pid=2089852)[0m "linear_conv_kernel_dim": 4,
-[36m(WorkerDict pid=2089852)[0m "linear_key_head_dim": 128,
-[36m(WorkerDict pid=2089852)[0m "linear_num_key_heads": 16,
-[36m(WorkerDict pid=2089852)[0m "linear_num_value_heads": 32,
-[36m(WorkerDict pid=2089852)[0m "linear_value_head_dim": 128,
-[36m(WorkerDict pid=2089852)[0m "mamba_ssm_dtype": "float32",
-[36m(WorkerDict pid=2089852)[0m "max_position_embeddings": 262144,
-[36m(WorkerDict pid=2089852)[0m "mlp_only_layers": [],
-[36m(WorkerDict pid=2089852)[0m "model_type": "qwen3_5_text",
-[36m(WorkerDict pid=2089852)[0m "mtp_num_hidden_layers": 0,
-[36m(WorkerDict pid=2089852)[0m "mtp_use_dedicated_embeddings": false,
-[36m(WorkerDict pid=2089852)[0m "num_attention_heads": 16,
-[36m(WorkerDict pid=2089852)[0m "num_hidden_layers": 32,
-[36m(WorkerDict pid=2089852)[0m "num_key_value_heads": 4,
-[36m(WorkerDict pid=2089852)[0m "pad_token_id": null,
-[36m(WorkerDict pid=2089852)[0m "partial_rotary_factor": 0.25,
-[36m(WorkerDict pid=2089852)[0m "rms_norm_eps": 1e-06,
-[36m(WorkerDict pid=2089852)[0m "rope_parameters": {
-[36m(WorkerDict pid=2089852)[0m "mrope_interleaved": true,
-[36m(WorkerDict pid=2089852)[0m "mrope_section": [
-[36m(WorkerDict pid=2089852)[0m 11,
-[36m(WorkerDict pid=2089852)[0m 11,
-[36m(WorkerDict pid=2089852)[0m 10
-[36m(WorkerDict pid=2089852)[0m ],
-[36m(WorkerDict pid=2089852)[0m "partial_rotary_factor": 0.25,
-[36m(WorkerDict pid=2089852)[0m "rope_theta": 10000000,
-[36m(WorkerDict pid=2089852)[0m "rope_type": "default"
-[36m(WorkerDict pid=2089852)[0m },
-[36m(WorkerDict pid=2089852)[0m "tie_word_embeddings": false,
-[36m(WorkerDict pid=2089852)[0m "use_cache": true,
-[36m(WorkerDict pid=2089852)[0m "vocab_size": 248320
-[36m(WorkerDict pid=2089852)[0m },
-[36m(WorkerDict pid=2089852)[0m "tie_word_embeddings": false,
-[36m(WorkerDict pid=2089852)[0m "transformers_version": "5.5.4",
-[36m(WorkerDict pid=2089852)[0m "video_token_id": 248057,
-[36m(WorkerDict pid=2089852)[0m "vision_config": {
-[36m(WorkerDict pid=2089852)[0m "deepstack_visual_indexes": [],[36m(WorkerDict pid=2090328)[0m `torch_dtype` is deprecated! Use `dtype` instead!
-[36m(WorkerDict pid=2090327)[0m [W825 18:08:27.154508901 ProcessGroupNCCL.cpp:929] Warning: TORCH_NCCL_AVOID_RECORD_STREAMS is the default now, this environment variable is thus deprecated. (function operator())[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2090328)[0m The fast path is not available because one of the required library is not installed. Falling back to torch implementation. To install follow https://github.com/fla-org/flash-linear-attention#installation and https://github.com/Dao-AILab/causal-conv1d
-[36m(WorkerDict pid=2090328)[0m [rank7]:[W825 18:08:32.884323464 ProcessGroupNCCL.cpp:5138] Guessing device ID based on global rank. This can cause a hang if rank to GPU mapping is heterogeneous. You can specify device_id in init_process_group()
-[36m(WorkerDict pid=2089852)[0m
Loading weights: 0%| | 0/760 [00:00, ?it/s]
Loading weights: 100%|██████████| 760/760 [00:00<00:00, 10572.92it/s]
-[36m(WorkerDict pid=2089852)[0m /data/miniconda3/envs/orarl/lib/python3.11/site-packages/torch/distributed/c10d_logger.py:83: UserWarning: barrier(): using the device under current context. You can specify `device_id` in `init_process_group` to mute this warning.
-[36m(WorkerDict pid=2089852)[0m return func(*args, **kwargs)
-[36m(WorkerDict pid=2090322)[0m `Qwen2VLImageProcessorFast` is deprecated. The `Fast` suffix for image processors has been removed; use `Qwen2VLImageProcessor` instead.
-[36m(WorkerDict pid=2090326)[0m The `use_fast` parameter is deprecated and will be removed in a future version. Use `backend="torchvision"` instead of `use_fast=True`, or `backend="pil"` instead of `use_fast=False`.[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2090323)[0m `torch_dtype` is deprecated! Use `dtype` instead![32m [repeated 6x across cluster][0m
-[36m(WorkerDict pid=2090326)[0m The fast path is not available because one of the required library is not installed. Falling back to torch implementation. To install follow https://github.com/fla-org/flash-linear-attention#installation and https://github.com/Dao-AILab/causal-conv1d[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2089852)[0m [rank0]:[W825 18:08:33.067054600 ProcessGroupNCCL.cpp:5138] Guessing device ID based on global rank. This can cause a hang if rank to GPU mapping is heterogeneous. You can specify device_id in init_process_group()[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2090324)[0m :1241: FutureWarning: The cuda.cudart module is deprecated and will be removed in a future release, please switch to use the cuda.bindings.runtime module instead.
-[36m(WorkerDict pid=2090324)[0m :1241: FutureWarning: The cuda.nvrtc module is deprecated and will be removed in a future release, please switch to use the cuda.bindings.nvrtc module instead.
-[36m(WorkerDict pid=2089852)[0m `Qwen2VLImageProcessorFast` is deprecated. The `Fast` suffix for image processors has been removed; use `Qwen2VLImageProcessor` instead.[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2089852)[0m 2026-08-25 18:09:26,834 - INFO - autotuner.py:262 - flashinfer.jit: [Autotuner]: Autotuning process starts ...
-[36m(WorkerDict pid=2089852)[0m 2026-08-25 18:09:26,851 - INFO - autotuner.py:268 - flashinfer.jit: [Autotuner]: Autotuning process ends
-[36m(WorkerDict pid=2090325)[0m :1241: FutureWarning: The cuda.cudart module is deprecated and will be removed in a future release, please switch to use the cuda.bindings.runtime module instead.[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2090325)[0m :1241: FutureWarning: The cuda.nvrtc module is deprecated and will be removed in a future release, please switch to use the cuda.bindings.nvrtc module instead.[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 0%| | 0/51 [00:00, ?it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 4%|▍ | 2/51 [00:00<00:03, 12.77it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 8%|▊ | 4/51 [00:00<00:03, 12.95it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 12%|█▏ | 6/51 [00:00<00:03, 13.19it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 16%|█▌ | 8/51 [00:00<00:03, 13.43it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 20%|█▉ | 10/51 [00:00<00:02, 13.97it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 24%|██▎ | 12/51 [00:00<00:02, 14.03it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 27%|██▋ | 14/51 [00:01<00:02, 14.62it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 31%|███▏ | 16/51 [00:01<00:02, 15.10it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 37%|███▋ | 19/51 [00:01<00:01, 16.91it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 41%|████ | 21/51 [00:01<00:01, 17.49it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 45%|████▌ | 23/51 [00:01<00:01, 17.56it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 49%|████▉ | 25/51 [00:01<00:01, 17.89it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 55%|█████▍ | 28/51 [00:01<00:01, 18.95it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 61%|██████ | 31/51 [00:01<00:01, 19.67it/s]
-[36m(WorkerDict pid=2090326)[0m 2026-08-25 18:09:26,835 - INFO - autotuner.py:262 - flashinfer.jit: [Autotuner]: Autotuning process starts ...[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2090326)[0m 2026-08-25 18:09:26,849 - INFO - autotuner.py:268 - flashinfer.jit: [Autotuner]: Autotuning process ends[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 65%|██████▍ | 33/51 [00:01<00:00, 19.28it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 71%|███████ | 36/51 [00:02<00:00, 20.71it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 76%|███████▋ | 39/51 [00:02<00:00, 21.55it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 82%|████████▏ | 42/51 [00:02<00:00, 22.59it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 88%|████████▊ | 45/51 [00:02<00:00, 22.90it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 94%|█████████▍| 48/51 [00:02<00:00, 23.73it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (mixed prefill-decode, PIECEWISE): 100%|██████████| 51/51 [00:02<00:00, 18.92it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (decode, FULL): 0%| | 0/51 [00:00, ?it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (decode, FULL): 4%|▍ | 2/51 [00:00<00:04, 10.91it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (decode, FULL): 8%|▊ | 4/51 [00:00<00:04, 11.31it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (decode, FULL): 12%|█▏ | 6/51 [00:00<00:03, 11.61it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (decode, FULL): 16%|█▌ | 8/51 [00:00<00:03, 12.02it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (decode, FULL): 20%|█▉ | 10/51 [00:00<00:03, 12.59it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (decode, FULL): 24%|██▎ | 12/51 [00:00<00:02, 13.20it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (decode, FULL): 27%|██▋ | 14/51 [00:01<00:02, 13.90it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (decode, FULL): 31%|███▏ | 16/51 [00:01<00:02, 14.48it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (decode, FULL): 35%|███▌ | 18/51 [00:01<00:02, 15.16it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (decode, FULL): 39%|███▉ | 20/51 [00:01<00:02, 15.43it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (decode, FULL): 43%|████▎ | 22/51 [00:01<00:01, 15.36it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (decode, FULL): 47%|████▋ | 24/51 [00:01<00:01, 15.86it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (decode, FULL): 53%|█████▎ | 27/51 [00:01<00:01, 17.42it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (decode, FULL): 59%|█████▉ | 30/51 [00:01<00:01, 18.81it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (decode, FULL): 65%|██████▍ | 33/51 [00:02<00:00, 20.36it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (decode, FULL): 71%|███████ | 36/51 [00:02<00:00, 21.63it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (decode, FULL): 76%|███████▋ | 39/51 [00:02<00:00, 23.55it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (decode, FULL): 82%|████████▏ | 42/51 [00:02<00:00, 24.61it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (decode, FULL): 88%|████████▊ | 45/51 [00:02<00:00, 21.28it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (decode, FULL): 96%|█████████▌| 49/51 [00:02<00:00, 23.73it/s]
-[36m(WorkerDict pid=2089852)[0m
Capturing CUDA graphs (decode, FULL): 100%|██████████| 51/51 [00:02<00:00, 18.02it/s]
-
-[36m(WorkerDict pid=2089852)[0m "depth": 27,
-[36m(WorkerDict pid=2089852)[0m "dtype": "bfloat16",
-[36m(WorkerDict pid=2089852)[0m "hidden_act": "gelu_pytorch_tanh",
-[36m(WorkerDict pid=2089852)[0m "hidden_size": 1152,
-[36m(WorkerDict pid=2089852)[0m "in_channels": 3,
-[36m(WorkerDict pid=2089852)[0m "initializer_range": 0.02,
-[36m(WorkerDict pid=2089852)[0m "intermediate_size": 4304,
-[36m(WorkerDict pid=2089852)[0m "model_type": "qwen3_5",
-[36m(WorkerDict pid=2089852)[0m "num_heads": 16,
-[36m(WorkerDict pid=2089852)[0m "num_position_embeddings": 2304,
-[36m(WorkerDict pid=2089852)[0m "out_hidden_size": 4096,
-[36m(WorkerDict pid=2089852)[0m "patch_size": 16,
-[36m(WorkerDict pid=2089852)[0m "spatial_merge_size": 2,
-[36m(WorkerDict pid=2089852)[0m "temporal_patch_size": 2
-[36m(WorkerDict pid=2089852)[0m },
-[36m(WorkerDict pid=2089852)[0m "vision_end_token_id": 248054,
-[36m(WorkerDict pid=2089852)[0m "vision_start_token_id": 248053
-[36m(WorkerDict pid=2089852)[0m }
-[36m(WorkerDict pid=2089852)[0m
-[36m(WorkerDict pid=2089852)[0m Ulysses patch applied!
-[36m(WorkerDict pid=2089852)[0m Vision backbone is set to not trainable.
-[36m(WorkerDict pid=2089852)[0m Qwen3_5ForConditionalGeneration contains 9.41B parameters.
-[36m(WorkerDict pid=2089852)[0m After huggingface model init: 1.17 GB / 95.00 GB.
-[36m(WorkerDict pid=2089852)[0m FSDP wrap policy: functools.partial(, transformer_layer_cls={, }).
-[36m(WorkerDict pid=2089852)[0m After FSDP module init: 15.42 GB / 95.00 GB.
-[36m(WorkerDict pid=2089852)[0m After optimizer init: 15.42 GB / 95.00 GB.
-[36m(WorkerDict pid=2090327)[0m Sampling params: {'max_tokens': 4096, 'detokenize': False, 'logit_bias': {248056: -100, 248057: -100}, 'n': 8, 'temperature': 1.0, 'top_p': 0.85, 'top_k': -1, 'ignore_eos': False}.
-[36m(Runner pid=2088024)[0m Config
-[36m(Runner pid=2088024)[0m algorithm:
-[36m(Runner pid=2088024)[0m adv_estimator: grpo
-[36m(Runner pid=2088024)[0m detached_oracle_advantage: true
-[36m(Runner pid=2088024)[0m detached_oracle_advantage_scale: 2.0
-[36m(Runner pid=2088024)[0m detached_oracle_match_best_max: 1.0
-[36m(Runner pid=2088024)[0m detached_oracle_match_best_min: 0.05
-[36m(Runner pid=2088024)[0m detached_oracle_match_best_ratio: 1.2
-[36m(Runner pid=2088024)[0m detached_oracle_use_directional_gain: false
-[36m(Runner pid=2088024)[0m directional_gain: true
-[36m(Runner pid=2088024)[0m directional_gain_gamma: 0.25
-[36m(Runner pid=2088024)[0m directional_gain_positive_only: true
-[36m(Runner pid=2088024)[0m directional_gain_recenter: true
-[36m(Runner pid=2088024)[0m disable_kl: true
-[36m(Runner pid=2088024)[0m gamma: 1.0
-[36m(Runner pid=2088024)[0m kl_coef: 0.0
-[36m(Runner pid=2088024)[0m kl_horizon: 10000.0
-[36m(Runner pid=2088024)[0m kl_penalty: low_var_kl
-[36m(Runner pid=2088024)[0m kl_target: 0.1
-[36m(Runner pid=2088024)[0m kl_type: fixed
-[36m(Runner pid=2088024)[0m lam: 1.0
-[36m(Runner pid=2088024)[0m name: orarl
-[36m(Runner pid=2088024)[0m oracle_builder: orarl.rewards:build_oracle_response_from_ground_truth
-[36m(Runner pid=2088024)[0m oracle_injection: true
-[36m(Runner pid=2088024)[0m oracle_injection_mode: append
-[36m(Runner pid=2088024)[0m oracle_log_exclude: true
-[36m(Runner pid=2088024)[0m oracle_replace_index: -1
-[36m(Runner pid=2088024)[0m oracle_reward_gate_beta: 2.0
-[36m(Runner pid=2088024)[0m post_selection_recenter: true
-[36m(Runner pid=2088024)[0m post_selection_rms_match: true
-[36m(Runner pid=2088024)[0m post_selection_rms_min_scale: 0.25
-[36m(Runner pid=2088024)[0m scale_rewards: false
-[36m(Runner pid=2088024)[0m selection_keep_oracle: true
-[36m(Runner pid=2088024)[0m selection_negative_quota: 2
-[36m(Runner pid=2088024)[0m selection_positive_quota: 1
-[36m(Runner pid=2088024)[0m selection_prune_ratio: 0.5
-[36m(Runner pid=2088024)[0m selection_strict_sign_balance: true
-[36m(Runner pid=2088024)[0m use_kl_loss: false
-[36m(Runner pid=2088024)[0m data:
-[36m(Runner pid=2088024)[0m answer_key: answer
-[36m(Runner pid=2088024)[0m dataloader_num_workers: 0
-[36m(Runner pid=2088024)[0m enable_thinking: false
-[36m(Runner pid=2088024)[0m filter_overlong_prompts: false
-[36m(Runner pid=2088024)[0m filter_overlong_prompts_workers: 16
-[36m(Runner pid=2088024)[0m format_prompt: null
-[36m(Runner pid=2088024)[0m group_by_task: true
-[36m(Runner pid=2088024)[0m group_by_task_key: problem_type
-[36m(Runner pid=2088024)[0m image_dir: /
-[36m(Runner pid=2088024)[0m image_key: images
-[36m(Runner pid=2088024)[0m image_max_pixels: 1048576
-[36m(Runner pid=2088024)[0m image_min_pixels: 4096
-[36m(Runner pid=2088024)[0m inline_video_tensors: true
-[36m(Runner pid=2088024)[0m max_pixels: 4194304
-[36m(Runner pid=2088024)[0m max_prompt_length: 24576
-[36m(Runner pid=2088024)[0m max_response_length: 4096
-[36m(Runner pid=2088024)[0m min_pixels: 4096
-[36m(Runner pid=2088024)[0m mini_rollout_batch_size: null
-[36m(Runner pid=2088024)[0m override_chat_template: null
-[36m(Runner pid=2088024)[0m preprocessed_video_dir: null
-[36m(Runner pid=2088024)[0m prompt_key: problem
-[36m(Runner pid=2088024)[0m response_prefix: ''
-[36m(Runner pid=2088024)[0m rollout_batch_size: 8
-[36m(Runner pid=2088024)[0m seed: 42
-[36m(Runner pid=2088024)[0m shuffle: true
-[36m(Runner pid=2088024)[0m train_files: /mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_joint_100k_maskaware.jsonl
-[36m(Runner pid=2088024)[0m use_preprocessed_videos: false
-[36m(Runner pid=2088024)[0m val_batch_size: 16
-[36m(Runner pid=2088024)[0m val_files: /mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_phase3_49152.canary.jsonl
-[36m(Runner pid=2088024)[0m val_preprocessed_video_dir: null
-[36m(Runner pid=2088024)[0m val_video_fps: 2.0
-[36m(Runner pid=2088024)[0m val_video_max_frames: 128
-[36m(Runner pid=2088024)[0m val_video_max_pixels: 786432
-[36m(Runner pid=2088024)[0m val_video_min_pixels: 4096
-[36m(Runner pid=2088024)[0m val_video_source_mode: realtime_only
-[36m(Runner pid=2088024)[0m val_video_total_pixels: 8388608
-[36m(Runner pid=2088024)[0m video_fps: 2.0
-[36m(Runner pid=2088024)[0m video_key: videos
-[36m(Runner pid=2088024)[0m video_max_frames: 128
-[36m(Runner pid=2088024)[0m video_max_pixels: 786432
-[36m(Runner pid=2088024)[0m video_min_pixels: 4096
-[36m(Runner pid=2088024)[0m video_source_mode: realtime_only
-[36m(Runner pid=2088024)[0m video_total_pixels: 8388608
-[36m(Runner pid=2088024)[0m trainer:
-[36m(Runner pid=2088024)[0m best_train_metric_key: reward/overall
-[36m(Runner pid=2088024)[0m best_train_min_step: 10
-[36m(Runner pid=2088024)[0m best_train_smooth_window: 5
-[36m(Runner pid=2088024)[0m critic_warmup: 0
-[36m(Runner pid=2088024)[0m experiment_name: smoke-orarl-9b
-[36m(Runner pid=2088024)[0m find_last_checkpoint: false
-[36m(Runner pid=2088024)[0m keep_best_train_ckpt: false
-[36m(Runner pid=2088024)[0m keep_optim_only_latest: true
-[36m(Runner pid=2088024)[0m load_checkpoint_path: null
-[36m(Runner pid=2088024)[0m logger:
-[36m(Runner pid=2088024)[0m - console
-[36m(Runner pid=2088024)[0m max_steps: 1
-[36m(Runner pid=2088024)[0m n_gpus_per_node: 8
-[36m(Runner pid=2088024)[0m nnodes: 1
-[36m(Runner pid=2088024)[0m project_name: OraRL
-[36m(Runner pid=2088024)[0m ray_timeline: null
-[36m(Runner pid=2088024)[0m save_checkpoint_path: /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/orarl
-[36m(Runner pid=2088024)[0m save_freq: 1
-[36m(Runner pid=2088024)[0m save_limit: 3
-[36m(Runner pid=2088024)[0m save_model_only: false
-[36m(Runner pid=2088024)[0m total_epochs: 1
-[36m(Runner pid=2088024)[0m val_before_train: false
-[36m(Runner pid=2088024)[0m val_freq: -1
-[36m(Runner pid=2088024)[0m val_generations_to_log: 0
-[36m(Runner pid=2088024)[0m val_only: false
-[36m(Runner pid=2088024)[0m worker:
-[36m(Runner pid=2088024)[0m actor:
-[36m(Runner pid=2088024)[0m clip_ratio_dual: 3.0
-[36m(Runner pid=2088024)[0m clip_ratio_high: 0.2
-[36m(Runner pid=2088024)[0m clip_ratio_low: 0.2
-[36m(Runner pid=2088024)[0m disable_kl: true
-[36m(Runner pid=2088024)[0m dynamic_batching: true
-[36m(Runner pid=2088024)[0m fsdp:
-[36m(Runner pid=2088024)[0m enable_cpu_offload: false
-[36m(Runner pid=2088024)[0m enable_full_shard: true
-[36m(Runner pid=2088024)[0m enable_rank0_init: true
-[36m(Runner pid=2088024)[0m fsdp_size: -1
-[36m(Runner pid=2088024)[0m mp_buffer_dtype: fp32
-[36m(Runner pid=2088024)[0m mp_param_dtype: bf16
-[36m(Runner pid=2088024)[0m mp_reduce_dtype: fp32
-[36m(Runner pid=2088024)[0m torch_dtype: bf16
-[36m(Runner pid=2088024)[0m use_orig_params: false
-[36m(Runner pid=2088024)[0m global_batch_size: 8
-[36m(Runner pid=2088024)[0m global_batch_size_per_device: -1
-[36m(Runner pid=2088024)[0m kl_coef: 0.0
-[36m(Runner pid=2088024)[0m kl_penalty: low_var_kl
-[36m(Runner pid=2088024)[0m loss_avg_mode: token
-[36m(Runner pid=2088024)[0m loss_type: default
-[36m(Runner pid=2088024)[0m max_grad_norm: 1.0
-[36m(Runner pid=2088024)[0m max_token_len_per_gpu: 32768
-[36m(Runner pid=2088024)[0m micro_batch_size_per_device_for_experience: 1
-[36m(Runner pid=2088024)[0m micro_batch_size_per_device_for_update: 1
-[36m(Runner pid=2088024)[0m model:
-[36m(Runner pid=2088024)[0m enable_gradient_checkpointing: true
-[36m(Runner pid=2088024)[0m freeze_vision_tower: true
-[36m(Runner pid=2088024)[0m model_path: /apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B
-[36m(Runner pid=2088024)[0m override_config: {}
-[36m(Runner pid=2088024)[0m tokenizer_path: /apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B
-[36m(Runner pid=2088024)[0m train_vision_merger: false
-[36m(Runner pid=2088024)[0m trust_remote_code: false
-[36m(Runner pid=2088024)[0m offload:
-[36m(Runner pid=2088024)[0m offload_optimizer: false
-[36m(Runner pid=2088024)[0m offload_params: false
-[36m(Runner pid=2088024)[0m optim:
-[36m(Runner pid=2088024)[0m betas:
-[36m(Runner pid=2088024)[0m - 0.9
-[36m(Runner pid=2088024)[0m - 0.999
-[36m(Runner pid=2088024)[0m lr: 1.0e-06
-[36m(Runner pid=2088024)[0m lr_scheduler_type: constant
-[36m(Runner pid=2088024)[0m lr_warmup_ratio: 0.0
-[36m(Runner pid=2088024)[0m lr_warmup_steps: null
-[36m(Runner pid=2088024)[0m min_lr_ratio: null
-[36m(Runner pid=2088024)[0m strategy: adamw
-[36m(Runner pid=2088024)[0m training_steps: 1
-[36m(Runner pid=2088024)[0m weight_decay: 0.0
-[36m(Runner pid=2088024)[0m padding_free: true
-[36m(Runner pid=2088024)[0m ppo_epochs: 1
-[36m(Runner pid=2088024)[0m selection_prune_ratio: 0.5
-[36m(Runner pid=2088024)[0m strategy: fsdp
-[36m(Runner pid=2088024)[0m ulysses_size: 1
-[36m(Runner pid=2088024)[0m use_kl_loss: false
-[36m(Runner pid=2088024)[0m use_torch_compile: false
-[36m(Runner pid=2088024)[0m critic:
-[36m(Runner pid=2088024)[0m cliprange_value: 0.5
-[36m(Runner pid=2088024)[0m dynamic_batching: true
-[36m(Runner pid=2088024)[0m fsdp:
-[36m(Runner pid=2088024)[0m enable_cpu_offload: false
-[36m(Runner pid=2088024)[0m enable_full_shard: true
-[36m(Runner pid=2088024)[0m enable_rank0_init: true
-[36m(Runner pid=2088024)[0m fsdp_size: -1
-[36m(Runner pid=2088024)[0m mp_buffer_dtype: fp32
-[36m(Runner pid=2088024)[0m mp_param_dtype: bf16
-[36m(Runner pid=2088024)[0m mp_reduce_dtype: fp32
-[36m(Runner pid=2088024)[0m torch_dtype: null
-[36m(Runner pid=2088024)[0m use_orig_params: false
-[36m(Runner pid=2088024)[0m global_batch_size: 256
-[36m(Runner pid=2088024)[0m global_batch_size_per_device: -1
-[36m(Runner pid=2088024)[0m loss_avg_mode: token
-[36m(Runner pid=2088024)[0m max_grad_norm: 1.0
-[36m(Runner pid=2088024)[0m max_token_len_per_gpu: null
-[36m(Runner pid=2088024)[0m micro_batch_size_per_device_for_experience: 16
-[36m(Runner pid=2088024)[0m micro_batch_size_per_device_for_update: 4
-[36m(Runner pid=2088024)[0m model:
-[36m(Runner pid=2088024)[0m enable_gradient_checkpointing: true
-[36m(Runner pid=2088024)[0m freeze_vision_tower: false
-[36m(Runner pid=2088024)[0m model_path: null
-[36m(Runner pid=2088024)[0m override_config: {}
-[36m(Runner pid=2088024)[0m tokenizer_path: null
-[36m(Runner pid=2088024)[0m train_vision_merger: false
-[36m(Runner pid=2088024)[0m trust_remote_code: true
-[36m(Runner pid=2088024)[0m offload:
-[36m(Runner pid=2088024)[0m offload_optimizer: false
-[36m(Runner pid=2088024)[0m offload_params: false
-[36m(Runner pid=2088024)[0m optim:
-[36m(Runner pid=2088024)[0m betas:
-[36m(Runner pid=2088024)[0m - 0.9
-[36m(Runner pid=2088024)[0m - 0.999
-[36m(Runner pid=2088024)[0m lr: 1.0e-06
-[36m(Runner pid=2088024)[0m lr_scheduler_type: constant
-[36m(Runner pid=2088024)[0m lr_warmup_ratio: 0.0
-[36m(Runner pid=2088024)[0m lr_warmup_steps: null
-[36m(Runner pid=2088024)[0m min_lr_ratio: null
-[36m(Runner pid=2088024)[0m strategy: adamw
-[36m(Runner pid=2088024)[0m training_steps: 1
-[36m(Runner pid=2088024)[0m weight_decay: 0.01
-[36m(Runner pid=2088024)[0m padding_free: false
-[36m(Runner pid=2088024)[0m ppo_epochs: 1
-[36m(Runner pid=2088024)[0m strategy: fsdp
-[36m(Runner pid=2088024)[0m ulysses_size: 1
-[36m(Runner pid=2088024)[0m hybrid_engine: true
-[36m(Runner pid=2088024)[0m ref:
-[36m(Runner pid=2088024)[0m dynamic_batching: true
-[36m(Runner pid=2088024)[0m fsdp:
-[36m(Runner pid=2088024)[0m enable_cpu_offload: false
-[36m(Runner pid=2088024)[0m enable_full_shard: true
-[36m(Runner pid=2088024)[0m enable_rank0_init: true
-[36m(Runner pid=2088024)[0m fsdp_size: -1
-[36m(Runner pid=2088024)[0m mp_buffer_dtype: fp32
-[36m(Runner pid=2088024)[0m mp_param_dtype: bf16
-[36m(Runner pid=2088024)[0m mp_reduce_dtype: fp32
-[36m(Runner pid=2088024)[0m torch_dtype: bf16
-[36m(Runner pid=2088024)[0m use_orig_params: false
-[36m(Runner pid=2088024)[0m max_token_len_per_gpu: 32768
-[36m(Runner pid=2088024)[0m micro_batch_size_per_device_for_experience: 1
-[36m(Runner pid=2088024)[0m offload:
-[36m(Runner pid=2088024)[0m offload_optimizer: false
-[36m(Runner pid=2088024)[0m offload_params: false
-[36m(Runner pid=2088024)[0m padding_free: true
-[36m(Runner pid=2088024)[0m strategy: fsdp
-[36m(Runner pid=2088024)[0m ulysses_size: 1
-[36m(Runner pid=2088024)[0m use_torch_compile: false
-[36m(Runner pid=2088024)[0m reward:
-[36m(Runner pid=2088024)[0m num_cpus: 1
-[36m(Runner pid=2088024)[0m reward_function: orarl.rewards
-[36m(Runner pid=2088024)[0m reward_function_is_module: true
-[36m(Runner pid=2088024)[0m reward_function_kwargs: {}
-[36m(Runner pid=2088024)[0m reward_function_name: compute_score
-[36m(Runner pid=2088024)[0m skip_special_tokens: true
-[36m(Runner pid=2088024)[0m rollout:
-[36m(Runner pid=2088024)[0m calculate_log_probs: false
-[36m(Runner pid=2088024)[0m collect_seq_logprob_for_filter: false
-[36m(Runner pid=2088024)[0m disable_log_stats: true
-[36m(Runner pid=2088024)[0m disable_tqdm: true
-[36m(Runner pid=2088024)[0m dtype: bf16
-[36m(Runner pid=2088024)[0m enable_chunked_prefill: true
-[36m(Runner pid=2088024)[0m enforce_eager: false
-[36m(Runner pid=2088024)[0m gpu_memory_utilization: 0.45
-[36m(Runner pid=2088024)[0m ignore_eos: false
-[36m(Runner pid=2088024)[0m kv_cache_dtype: auto
-[36m(Runner pid=2088024)[0m limit_images: 10
-[36m(Runner pid=2088024)[0m max_model_len: null
-[36m(Runner pid=2088024)[0m max_num_batched_tokens: 32768
-[36m(Runner pid=2088024)[0m n: 8
-[36m(Runner pid=2088024)[0m name: vllm
-[36m(Runner pid=2088024)[0m prompt_length: 24576
-[36m(Runner pid=2088024)[0m response_length: 4096
-[36m(Runner pid=2088024)[0m seed: 42
-[36m(Runner pid=2088024)[0m temperature: 1.0
-[36m(Runner pid=2088024)[0m tensor_parallel_size: 1
-[36m(Runner pid=2088024)[0m top_k: -1
-[36m(Runner pid=2088024)[0m top_p: 0.85
-[36m(Runner pid=2088024)[0m trust_remote_code: false
-[36m(Runner pid=2088024)[0m val_override_config:
-[36m(Runner pid=2088024)[0m n: 1
-[36m(Runner pid=2088024)[0m temperature: 0.7
-[36m(Runner pid=2088024)[0m top_p: 0.95
-[36m(Runner pid=2088024)[0m
-
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [00:00, ?it/s]
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [00:00, ?it/s]
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [00:00, ?it/s]
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [00:01, ?it/s]
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [00:02, ?it/s]
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [00:02, ?it/s]
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [00:02, ?it/s]
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [00:21, ?it/s]
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [00:21, ?it/s]
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [00:22, ?it/s][36m(WorkerDict pid=2089852)[0m After vllm init: 7.01 GB / 95.00 GB.
-[36m(WorkerDict pid=2089852)[0m Before vllm wake up in sharding manager: 7.01 GB / 95.00 GB.
-[36m(WorkerDict pid=2089852)[0m After gather model weights in sharding manager: 30.92 GB / 95.00 GB.
-[36m(WorkerDict pid=2089852)[0m After sync model weights in sharding manager: 24.69 GB / 95.00 GB.
-[36m(WorkerDict pid=2089852)[0m After vllm wake up in sharding manager: 45.04 GB / 95.00 GB.
-[36m(Runner pid=2088024)[0m Start generating batch...
-[36m(WorkerDict pid=2089852)[0m Before vllm offload in sharding manager: 50.43 GB / 95.00 GB.
-[36m(WorkerDict pid=2090325)[0m Sampling params: {'max_tokens': 4096, 'detokenize': False, 'logit_bias': {248056: -100, 248057: -100}, 'n': 8, 'temperature': 1.0, 'top_p': 0.85, 'top_k': -1, 'ignore_eos': False}.[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2089852)[0m After vllm offload in sharding manager: 7.40 GB / 95.00 GB.
-
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 0%| | 0.00/2.00 [00:00, ?it/s][A[A
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [00:08<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [00:08<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 0%| | 0.00/1.00 [00:00, ?it/s][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 0%| | 0.00/1.00 [00:00, ?it/s][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A[36m(WorkerDict pid=2089852)[0m /data/miniconda3/envs/orarl/lib/python3.11/site-packages/torch/distributed/c10d_logger.py:83: UserWarning: barrier(): using the device under current context. You can specify `device_id` in `init_process_group` to mute this warning.
-
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [01:10, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [00:43<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [00:26<00:00, 26.3s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [00:25<00:00, 25.7s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A[36m(WorkerDict pid=2089852)[0m return func(*args, **kwargs)
-
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [01:10, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [00:43<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [00:26<00:00, 26.3s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [00:25<00:00, 25.7s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [01:17, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [00:50<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [00:33<00:00, 33.2s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [00:32<00:00, 32.6s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [01:17, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [00:50<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [00:33<00:00, 33.2s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [00:32<00:00, 32.6s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [01:17, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [00:50<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [00:33<00:00, 33.2s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [00:32<00:00, 32.6s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A[36m(WorkerDict pid=2089852)[0m /data/miniconda3/envs/orarl/lib/python3.11/site-packages/torch/distributed/c10d_logger.py:83: UserWarning: barrier(): using the device under current context. You can specify `device_id` in `init_process_group` to mute this warning.
-
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:57, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:30<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:13<00:00, 133s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:12<00:00, 133s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A[36m(WorkerDict pid=2089852)[0m return func(*args, **kwargs)
-
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:57, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:30<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:13<00:00, 133s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:12<00:00, 133s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A[36m(WorkerDict pid=2089852)[0m [rank-0]: Saving model to /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_0.pt.
-[36m(WorkerDict pid=2089852)[0m [rank-0]: Saving optimizer to /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_0.pt.
-[36m(WorkerDict pid=2089852)[0m [rank-0]: Saving extra_state to /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_0.pt.
-[36m(Runner pid=2088024)[0m [TRAIN STEP] 1/1 step_s=179.2 elapsed=2m59s eta=0m00s
-[36m(Runner pid=2088024)[0m Step 1
-[36m(Runner pid=2088024)[0m actor:
-[36m(Runner pid=2088024)[0m backward_rows: 32.0
-[36m(Runner pid=2088024)[0m backward_rows_per_prompt: 4.0
-[36m(Runner pid=2088024)[0m entropy_loss: 0.355
-[36m(Runner pid=2088024)[0m grad_norm: 5.594
-[36m(Runner pid=2088024)[0m lr: 1.0e-06
-[36m(Runner pid=2088024)[0m oracle_injection_enabled: 1.0
-[36m(Runner pid=2088024)[0m orarl_selection_enabled: 1.0
-[36m(Runner pid=2088024)[0m pg_clipfrac_higher: 0.027
-[36m(Runner pid=2088024)[0m pg_clipfrac_lower: 0.0
-[36m(Runner pid=2088024)[0m pg_loss: 0.002
-[36m(Runner pid=2088024)[0m ppo_kl: 0.005
-[36m(Runner pid=2088024)[0m critic:
-[36m(Runner pid=2088024)[0m advantages:
-[36m(Runner pid=2088024)[0m max: 0.541
-[36m(Runner pid=2088024)[0m mean: 1.0643687042843908e-09
-[36m(Runner pid=2088024)[0m min: -0.484
-[36m(Runner pid=2088024)[0m std: 0.366
-[36m(Runner pid=2088024)[0m returns:
-[36m(Runner pid=2088024)[0m max: 0.913
-[36m(Runner pid=2088024)[0m mean: -0.015
-[36m(Runner pid=2088024)[0m min: -0.75
-[36m(Runner pid=2088024)[0m rewards:
-[36m(Runner pid=2088024)[0m max: 1.0
-[36m(Runner pid=2088024)[0m mean: 0.594
-[36m(Runner pid=2088024)[0m min: 0.0
-[36m(Runner pid=2088024)[0m std: 0.491
-[36m(Runner pid=2088024)[0m score:
-[36m(Runner pid=2088024)[0m max: 1.0
-[36m(Runner pid=2088024)[0m mean: 0.594
-[36m(Runner pid=2088024)[0m min: 0.0
-[36m(Runner pid=2088024)[0m std: 0.491
-[36m(Runner pid=2088024)[0m var: 0.241
-[36m(Runner pid=2088024)[0m global_seqlen:
-[36m(Runner pid=2088024)[0m balanced_max: 63162
-[36m(Runner pid=2088024)[0m balanced_min: 60075
-[36m(Runner pid=2088024)[0m max: 76879
-[36m(Runner pid=2088024)[0m mean: 61474.5
-[36m(Runner pid=2088024)[0m min: 42686
-[36m(Runner pid=2088024)[0m minmax_diff: 34193
-[36m(Runner pid=2088024)[0m orarl:
-[36m(Runner pid=2088024)[0m degenerate_op_groups: 3.0
-[36m(Runner pid=2088024)[0m detached_a_oracle_mean: 0.3
-[36m(Runner pid=2088024)[0m detached_a_oracle_raw_mean: 0.574
-[36m(Runner pid=2088024)[0m detached_best_op_adv_mean: 0.298
-[36m(Runner pid=2088024)[0m detached_gain_mean: 1.0
-[36m(Runner pid=2088024)[0m detached_groups: 8.0
-[36m(Runner pid=2088024)[0m detached_match_cap_applied_frac: 0.375
-[36m(Runner pid=2088024)[0m detached_match_cap_mean: 0.377
-[36m(Runner pid=2088024)[0m detached_w_mean: 0.287
-[36m(Runner pid=2088024)[0m detached_w_mean_by_task:
-[36m(Runner pid=2088024)[0m image_sequence_mc_answer_only: 0.287
-[36m(Runner pid=2088024)[0m directional_gain_amplified_frac: 0.328
-[36m(Runner pid=2088024)[0m directional_gain_max: 4.0
-[36m(Runner pid=2088024)[0m directional_gain_mean: 1.378
-[36m(Runner pid=2088024)[0m directional_gain_positive_only: 1.0
-[36m(Runner pid=2088024)[0m directional_gain_recenter: 1.0
-[36m(Runner pid=2088024)[0m directional_gain_recenter_abs_shift_max: 0.004
-[36m(Runner pid=2088024)[0m directional_gain_recenter_abs_shift_mean: 0.001
-[36m(Runner pid=2088024)[0m groups_with_no_op_pos_baseline: 3.0
-[36m(Runner pid=2088024)[0m mu_g_mean: 0.625
-[36m(Runner pid=2088024)[0m mu_op_mean: 0.578
-[36m(Runner pid=2088024)[0m n_groups: 8.0
-[36m(Runner pid=2088024)[0m op_adv_neg_count: 19.0
-[36m(Runner pid=2088024)[0m op_adv_neg_mean: 0.455
-[36m(Runner pid=2088024)[0m op_adv_pos_count: 21.0
-[36m(Runner pid=2088024)[0m op_adv_pos_mean: 0.412
-[36m(Runner pid=2088024)[0m oracle:
-[36m(Runner pid=2088024)[0m groups_injected: 8
-[36m(Runner pid=2088024)[0m oracle_adv_mag_max: 1.0
-[36m(Runner pid=2088024)[0m oracle_adv_mag_mean: 0.422
-[36m(Runner pid=2088024)[0m oracle_mean_contamination_of_means: 0.162
-[36m(Runner pid=2088024)[0m oracle_mean_contamination_per_group: 0.114
-[36m(Runner pid=2088024)[0m post_selection_active_mean_after: 9.313225746154785e-10
-[36m(Runner pid=2088024)[0m post_selection_active_mean_before: -0.015
-[36m(Runner pid=2088024)[0m post_selection_active_policy_rows: 3.0
-[36m(Runner pid=2088024)[0m post_selection_active_rms_after: 0.292
-[36m(Runner pid=2088024)[0m post_selection_active_rms_before: 0.344
-[36m(Runner pid=2088024)[0m post_selection_active_rows: 4.0
-[36m(Runner pid=2088024)[0m post_selection_enabled: 1.0
-[36m(Runner pid=2088024)[0m post_selection_groups: 8.0
-[36m(Runner pid=2088024)[0m post_selection_groups_skipped: 0.0
-[36m(Runner pid=2088024)[0m post_selection_on_policy_rms: 0.291
-[36m(Runner pid=2088024)[0m post_selection_oracle_advantage_after: 0.289
-[36m(Runner pid=2088024)[0m post_selection_oracle_rows: 1.0
-[36m(Runner pid=2088024)[0m post_selection_oracle_sign_projection: 0.0
-[36m(Runner pid=2088024)[0m post_selection_recenter: 1.0
-[36m(Runner pid=2088024)[0m post_selection_rms_match: 1.0
-[36m(Runner pid=2088024)[0m post_selection_rms_scale: 0.954
-[36m(Runner pid=2088024)[0m post_selection_sigma_op_fallback: 0.375
-[36m(Runner pid=2088024)[0m scale_rewards: 0.0
-[36m(Runner pid=2088024)[0m selection:
-[36m(Runner pid=2088024)[0m abs_advantage_dropped_mean: 0.241
-[36m(Runner pid=2088024)[0m abs_advantage_kept_mean: 0.314
-[36m(Runner pid=2088024)[0m cross_sign_fallback_rows: 0.0
-[36m(Runner pid=2088024)[0m dropped_rows: 40.0
-[36m(Runner pid=2088024)[0m effective_keep_ratio: 0.444
-[36m(Runner pid=2088024)[0m full_batch_score_max: 1.0
-[36m(Runner pid=2088024)[0m full_batch_score_mean: 0.625
-[36m(Runner pid=2088024)[0m full_batch_score_min: 0.0
-[36m(Runner pid=2088024)[0m groups: 8.0
-[36m(Runner pid=2088024)[0m keep_per_group: 4.0
-[36m(Runner pid=2088024)[0m kept_rows: 32.0
-[36m(Runner pid=2088024)[0m negative_policy_rows_kept: 10.0
-[36m(Runner pid=2088024)[0m oracle_rows_forced: 8.0
-[36m(Runner pid=2088024)[0m positive_policy_rows_kept: 5.0
-[36m(Runner pid=2088024)[0m zero_fallback_rows: 9.0
-[36m(Runner pid=2088024)[0m zero_policy_rows_kept: 9.0
-[36m(Runner pid=2088024)[0m sigma_g_mean: 0.33
-[36m(Runner pid=2088024)[0m sigma_op_fallback_count: 3.0
-[36m(Runner pid=2088024)[0m sigma_op_mean: 0.29
-[36m(Runner pid=2088024)[0m sigma_ratio_of_means: 1.139
-[36m(Runner pid=2088024)[0m sigma_ratio_per_group: 1.005
-[36m(Runner pid=2088024)[0m sign_flip_count: 0.0
-[36m(Runner pid=2088024)[0m sign_flip_rate: 0.0
-[36m(Runner pid=2088024)[0m orarl_selection_seqlen:
-[36m(Runner pid=2088024)[0m balanced_max: 27878
-[36m(Runner pid=2088024)[0m balanced_min: 26766
-[36m(Runner pid=2088024)[0m max: 31041
-[36m(Runner pid=2088024)[0m mean: 27322.0
-[36m(Runner pid=2088024)[0m min: 23594
-[36m(Runner pid=2088024)[0m minmax_diff: 7447
-[36m(Runner pid=2088024)[0m perf:
-[36m(Runner pid=2088024)[0m cpu_memory_used_gb: 1082.023
-[36m(Runner pid=2088024)[0m max_memory_allocated_gb: 68.308
-[36m(Runner pid=2088024)[0m max_memory_reserved_gb: 84.883
-[36m(Runner pid=2088024)[0m mfu_actor: 0.444
-[36m(Runner pid=2088024)[0m throughput: 152.448
-[36m(Runner pid=2088024)[0m time_per_step: 179.222
-[36m(Runner pid=2088024)[0m total_num_tokens: 218576
-[36m(Runner pid=2088024)[0m prompt_length:
-[36m(Runner pid=2088024)[0m clip_ratio: 0.0
-[36m(Runner pid=2088024)[0m max: 8554.0
-[36m(Runner pid=2088024)[0m mean: 6823.5
-[36m(Runner pid=2088024)[0m min: 2280.0
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:14<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 0%| | 0.00/1.00 [02:59, ?it/s]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
[2m[36m(pid=2088024) [0mRunning step: 100%|██████████| 1.00/1.00 [02:59<00:00, 179s/it]
[2m[36m(pid=2088024) [0mRunning step: 100%|██████████| 1.00/1.00 [02:59<00:00, 179s/it]
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 100%|██████████| 1.00/1.00 [02:59<00:00, 179s/it]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 100%|██████████| 1.00/1.00 [02:59<00:00, 179s/it]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-
-
[A[A
-
-
[A[A
-
-
-
[A[A[A
[2m[36m(pid=2088024) [0mRunning step: 100%|██████████| 1.00/1.00 [02:59<00:00, 179s/it]
-
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:32<00:00, 4.02s/it][A[A
-
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:15<00:00, 135s/it][A[A
-
-
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:14<00:00, 134s/it][A[A[A
-[36m(Runner pid=2088024)[0m response_length:
-[36m(Runner pid=2088024)[0m clip_ratio: 0.0
-[36m(Runner pid=2088024)[0m max: 7.0
-[36m(Runner pid=2088024)[0m mean: 7.0
-[36m(Runner pid=2088024)[0m min: 7.0
-[36m(Runner pid=2088024)[0m reward:
-[36m(Runner pid=2088024)[0m accuracy: 0.625
-[36m(Runner pid=2088024)[0m format: 1.0
-[36m(Runner pid=2088024)[0m overall: 0.625
-[36m(Runner pid=2088024)[0m rollout:
-[36m(Runner pid=2088024)[0m all_identical_group_fraction: 0.375
-[36m(Runner pid=2088024)[0m exact_unique_fraction: 0.281
-[36m(Runner pid=2088024)[0m pairwise_token_agreement_mean: 0.943
-[36m(Runner pid=2088024)[0m response_length_std_mean: 0.0
-[36m(Runner pid=2088024)[0m reward_group_std_mean: 0.29
-[36m(Runner pid=2088024)[0m reward_unique_fraction: 0.203
-[36m(Runner pid=2088024)[0m zero_reward_std_group_fraction: 0.375
-[36m(Runner pid=2088024)[0m timing_per_token_ms:
-[36m(Runner pid=2088024)[0m adv: 4.752148470787695e-05
-[36m(Runner pid=2088024)[0m gen: 100.661
-[36m(Runner pid=2088024)[0m old: 0.089
-[36m(Runner pid=2088024)[0m reward: 0.016
-[36m(Runner pid=2088024)[0m update_actor: 0.128
-[36m(Runner pid=2088024)[0m timing_s:
-[36m(Runner pid=2088024)[0m adv: 0.01
-[36m(Runner pid=2088024)[0m gen: 22.548
-[36m(Runner pid=2088024)[0m old: 19.508
-[36m(Runner pid=2088024)[0m orarl_selection: 0.013
-[36m(Runner pid=2088024)[0m post_selection_balance: 0.002
-[36m(Runner pid=2088024)[0m reward: 0.004
-[36m(Runner pid=2088024)[0m save_checkpoint: 109.13
-[36m(Runner pid=2088024)[0m step: 179.222
-[36m(Runner pid=2088024)[0m update_actor: 27.961
-[36m(Runner pid=2088024)[0m
-[36m(WorkerDict pid=2090326)[0m [rank-5]: Saving model to /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/model_world_size_8_rank_5.pt.[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2090326)[0m [rank-5]: Saving optimizer to /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/optim_world_size_8_rank_5.pt.[32m [repeated 7x across cluster][0m
-[36m(WorkerDict pid=2090326)[0m [rank-5]: Saving extra_state to /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/orarl/global_step_1/actor/extra_state_world_size_8_rank_5.pt.[32m [repeated 7x across cluster][0m
-
[2m[36m(pid=2088024) [0mRunning step: 100%|██████████| 1.00/1.00 [03:03<00:00, 184s/it]
-
[2m[36m(pid=2089852) [0mCompute log probs: 100%|██████████| 2.00/2.00 [02:36<00:00, 78.4s/it]
-
[2m[36m(pid=2089852) [0mTrain mini-batches: 100%|██████████| 1.00/1.00 [02:19<00:00, 139s/it]
-
[2m[36m(pid=2089852) [0mUpdate policy: 100%|██████████| 1.00/1.00 [02:18<00:00, 139s/it]
-method=orarl mode=run
-/data/miniconda3/envs/orarl/bin/python -m verl.trainer.main config=/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/configs/orarl_9b.yaml trainer.max_steps=1 trainer.save_freq=1 trainer.val_before_train=false trainer.experiment_name=smoke-orarl-9b data.rollout_batch_size=8 worker.actor.global_batch_size=8 worker.actor.model.model_path=/apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B worker.actor.model.tokenizer_path=/apdcephfs_gy7/share_305340263/hunyuan/yunhengli/checkpoint/Video-ORA-9B data.train_files=/mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_joint_100k_maskaware.jsonl data.val_files=/mnt/lyh/TempSamp/EasyVideoR1/data/joint_final/rl_phase3_49152.canary.jsonl trainer.save_checkpoint_path=/mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/orarl trainer.nnodes=1 trainer.n_gpus_per_node=8
-checkpoint OK: orarl -> /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/orarl/global_step_1
-
-Smoke training finished.
-Log: /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/smoke.log
-GRPO output: /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/grpo
-OraRL output: /mnt/lyh/TempSamp/EasyVideoR1/OraRL-official/runs/smoke-training-9b-20260825_175949/orarl