Add files using upload-large-folder tool

.gitattributes

@@ -475,3 +475,13 @@ wandb/run-20251221_201505-3d40po3i/run-3d40po3i.wandb filter=lfs diff=lfs merge=
 wandb/run-20251221_205549-vnhoirkv/run-vnhoirkv.wandb filter=lfs diff=lfs merge=lfs -text
 wandb/run-20251222_011618-lmccfedj/run-lmccfedj.wandb filter=lfs diff=lfs merge=lfs -text
 wandb/run-20251222_152015-7wc5x7qq/run-7wc5x7qq.wandb filter=lfs diff=lfs merge=lfs -text
+wandb/run-20251222_194213-wibn1p50/run-wibn1p50.wandb filter=lfs diff=lfs merge=lfs -text
+wandb/run-20251222_202905-69ev02ff/run-69ev02ff.wandb filter=lfs diff=lfs merge=lfs -text
+wandb/run-20251222_203657-pjvd64ya/run-pjvd64ya.wandb filter=lfs diff=lfs merge=lfs -text
+wandb/run-20251222_204216-q200utyn/run-q200utyn.wandb filter=lfs diff=lfs merge=lfs -text
+wandb/run-20251222_204856-wk1bayyz/run-wk1bayyz.wandb filter=lfs diff=lfs merge=lfs -text
+wandb/run-20251222_222707-8p2uoqx7/run-8p2uoqx7.wandb filter=lfs diff=lfs merge=lfs -text
+wandb/run-20251222_223912-ystnyb7f/run-ystnyb7f.wandb filter=lfs diff=lfs merge=lfs -text
+wandb/run-20251223_020703-tkcjhbb8/run-tkcjhbb8.wandb filter=lfs diff=lfs merge=lfs -text
+wandb/run-20251223_053708-okfflayq/run-okfflayq.wandb filter=lfs diff=lfs merge=lfs -text
+wandb/run-20251223_104324-n3pfi4rh/run-n3pfi4rh.wandb filter=lfs diff=lfs merge=lfs -text

wandb/run-20251222_182729-6i1mkj49/files/config.yaml

@@ -0,0 +1,963 @@
+_wandb:
+    value:
+        cli_version: 0.23.0
+        e:
+            i0axz6jgeasoze55ozdefzq2sh05xke5:
+                args:
+                    - --node-ip-address=10.119.27.78
+                    - --node-manager-port=45457
+                    - --object-store-name=/tmp/ray/session_2025-12-22_18-26-14_802758_3429006/sockets/plasma_store
+                    - --raylet-name=/tmp/ray/session_2025-12-22_18-26-14_802758_3429006/sockets/raylet
+                    - --redis-address=None
+                    - --metrics-agent-port=41966
+                    - --logging-rotate-bytes=536870912
+                    - --logging-rotate-backup-count=5
+                    - --runtime-env-agent-port=57932
+                    - --gcs-address=10.119.27.78:62444
+                    - --session-name=session_2025-12-22_18-26-14_802758_3429006
+                    - --temp-dir=/tmp/ray
+                    - --webui=127.0.0.1:8265
+                    - --cluster-id=f5d582d4084ab6f69c0ebe1c1d779c63e370565da4ba7fe85624529d
+                    - --startup-token=112
+                    - --worker-launch-time-ms=1766399184671
+                    - --node-id=d98b0f6bb40ee78c635322e643c17204ee3f035aa813492ed75c5c3e
+                    - --runtime-env-hash=-839022692
+                cpu_count: 64
+                cpu_count_logical: 128
+                cudaVersion: "12.4"
+                disk:
+                    /:
+                        total: "5153960755200"
+                        used: "471657721856"
+                email: haoyu-wa22@mails.tsinghua.edu.cn
+                executable: /root/local/miniconda3/envs/ragen/bin/python
+                git:
+                    commit: 8d73639b99b38265453f898b8d6af7d4af50d56e
+                    remote: https://github.com/mll-lab-nu/RAGEN.git
+                gpu: NVIDIA H100 80GB HBM3
+                gpu_count: 8
+                gpu_nvidia:
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-35e2d43d-4067-82ce-90d4-def9e389bf28
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-af4135e3-88f2-e9ac-518d-502c75a85429
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-d7fdeeba-fe9b-ec03-d9f7-6724fe4266b5
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-ccc4f668-3882-5a8e-2c07-c5cd08f6f666
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-7b73c0cf-d3d5-e10c-7176-a43be1e41001
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-81b58d94-5d1f-8ec2-f9d2-fd56172ed177
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-03e8bc66-3b44-6794-49fd-5392fbdda6d1
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-b7cf0ec6-7c29-1179-dceb-09565da51890
+                host: pt-4d654cf4576f4d23ad3d3919f12932fe-worker-0
+                memory:
+                    total: "2163642122240"
+                os: Linux-5.14.0-284.25.1.el9_2.x86_64-x86_64-with-glibc2.35
+                program: /root/local/miniconda3/envs/ragen/lib/python3.10/site-packages/ray/_private/workers/default_worker.py
+                python: CPython 3.10.19
+                root: /mnt/general/wanghy/RAGEN
+                startedAt: "2025-12-22T10:27:29.079503Z"
+                writerId: i0axz6jgeasoze55ozdefzq2sh05xke5
+        m: []
+        python_version: 3.10.19
+        t:
+            "1":
+                - 1
+                - 5
+                - 11
+                - 30
+                - 41
+                - 49
+                - 50
+                - 51
+                - 53
+                - 71
+                - 95
+                - 97
+                - 98
+                - 105
+            "2":
+                - 1
+                - 5
+                - 11
+                - 30
+                - 41
+                - 49
+                - 50
+                - 51
+                - 53
+                - 71
+                - 95
+                - 97
+                - 98
+                - 105
+            "3":
+                - 2
+                - 13
+                - 16
+            "4": 3.10.19
+            "5": 0.23.0
+            "6": 4.57.1
+            "12": 0.23.0
+            "13": linux-x86_64
+actor_rollout_ref:
+    value:
+        actor:
+            _target_: verl.workers.config.FSDPActorConfig
+            checkpoint:
+                _target_: verl.trainer.config.CheckpointConfig
+                async_save: false
+                load_contents:
+                    - model
+                    - optimizer
+                    - extra
+                save_contents:
+                    - model
+                    - optimizer
+                    - extra
+            clip_ratio: 0.2
+            clip_ratio_c: 3
+            clip_ratio_high: 0.28
+            clip_ratio_low: 0.2
+            entropy_checkpointing: false
+            entropy_coeff: 0.001
+            entropy_from_logits_with_chunking: false
+            freeze_vision_tower: false
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                entropy_checkpointing: false
+                entropy_from_logits_with_chunking: false
+                forward_only: false
+                forward_prefetch: false
+                fsdp_size: -1
+                model_dtype: fp32
+                offload_policy: false
+                optimizer_offload: false
+                param_offload: false
+                reshard_after_forward: true
+                strategy: fsdp
+                ulysses_sequence_parallel_size: 1
+                use_orig_params: false
+                use_torch_compile: true
+                wrap_policy:
+                    min_num_params: 0
+            grad_clip: 1
+            grpo_advantage_length_weight: false
+            kl_loss_coef: 0
+            kl_loss_type: kl
+            loss_agg_mode: token-mean
+            optim:
+                _target_: verl.workers.config.FSDPOptimizerConfig
+                betas:
+                    - 0.9
+                    - 0.999
+                clip_grad: 1
+                lr: 1e-06
+                lr_warmup_steps: -1
+                lr_warmup_steps_ratio: 0
+                min_lr_ratio: 0
+                num_cycles: 0.5
+                total_training_steps: 200
+                warmup_style: constant
+                weight_decay: 0.01
+            policy_loss:
+                _target_: verl.workers.config.PolicyLossConfig
+                clip_cov_lb: 1
+                clip_cov_ratio: 0.0002
+                clip_cov_ub: 5
+                kl_cov_ratio: 0.0002
+                loss_mode: vanilla
+                ppo_kl_coef: 0.1
+            ppo_epochs: 1
+            ppo_max_token_len_per_gpu: 16384
+            ppo_micro_batch_size: null
+            ppo_micro_batch_size_per_gpu: 1
+            ppo_mini_batch_size: 32
+            profiler:
+                _target_: verl.utils.profiler.ProfilerConfig
+                all_ranks: false
+                enable: false
+                ranks: []
+                save_path: outputs/profile
+                tool: null
+                tool_config:
+                    npu:
+                        _target_: verl.utils.profiler.config.NPUToolConfig
+                        analysis: true
+                        contents: []
+                        discrete: false
+                        level: level1
+                    nsys:
+                        _target_: verl.utils.profiler.config.NsightToolConfig
+                        discrete: false
+                    torch:
+                        _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                        step_end: null
+                        step_start: 0
+                    torch_memory:
+                        _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                        stack_depth: 32
+                        trace_alloc_max_entries: 100000
+            shuffle: false
+            strategy: fsdp
+            tis_imp_ratio_cap: -1
+            ulysses_sequence_parallel_size: 1
+            use_dynamic_bsz: false
+            use_fused_kernels: false
+            use_kl_loss: false
+            use_ref: true
+            use_remove_padding: false
+            use_torch_compile: true
+        hybrid_engine: true
+        model:
+            _target_: verl.workers.config.HFModelConfig
+            custom_chat_template: null
+            enable_activation_offload: false
+            enable_gradient_checkpointing: true
+            exclude_modules: null
+            external_lib: null
+            fused_kernel_options:
+                impl_backend: torch
+            hf_config_path: null
+            lora_alpha: 16
+            lora_rank: 0
+            path: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            target_modules: all-linear
+            tokenizer_path: null
+            torch_dtype: bfloat16
+            trust_remote_code: false
+            use_fused_kernels: false
+            use_liger: false
+            use_remove_padding: false
+            use_shm: false
+        nccl_timeout: 600
+        ref:
+            entropy_checkpointing: false
+            entropy_from_logits_with_chunking: false
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                entropy_checkpointing: false
+                entropy_from_logits_with_chunking: false
+                forward_only: false
+                forward_prefetch: false
+                fsdp_size: -1
+                model_dtype: fp32
+                offload_policy: false
+                optimizer_offload: false
+                param_offload: false
+                reshard_after_forward: true
+                strategy: fsdp
+                ulysses_sequence_parallel_size: 1
+                use_orig_params: false
+                use_torch_compile: true
+                wrap_policy:
+                    min_num_params: 0
+            log_prob_max_token_len_per_gpu: 16384
+            log_prob_micro_batch_size: null
+            log_prob_micro_batch_size_per_gpu: 1
+            log_prob_use_dynamic_bsz: false
+            model: null
+            profiler:
+                _target_: verl.utils.profiler.ProfilerConfig
+                all_ranks: false
+                enable: false
+                ranks: []
+                save_path: outputs/profile
+                tool: null
+                tool_config:
+                    npu:
+                        _target_: verl.utils.profiler.config.NPUToolConfig
+                        analysis: true
+                        contents: []
+                        discrete: false
+                        level: level1
+                    nsys:
+                        _target_: verl.utils.profiler.config.NsightToolConfig
+                        discrete: false
+                    torch:
+                        _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                        step_end: null
+                        step_start: 0
+                    torch_memory:
+                        _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                        stack_depth: 32
+                        trace_alloc_max_entries: 100000
+            strategy: fsdp
+            ulysses_sequence_parallel_size: 1
+            use_torch_compile: true
+        rollout:
+            _target_: verl.workers.config.RolloutConfig
+            agent:
+                _target_: verl.workers.config.AgentLoopConfig
+                agent_loop_config_path: null
+                custom_async_server:
+                    _target_: verl.workers.config.CustomAsyncServerConfig
+                    name: null
+                    path: null
+                num_workers: 8
+            calculate_log_probs: false
+            cudagraph_capture_sizes: null
+            data_parallel_size: 1
+            disable_log_stats: true
+            do_sample: true
+            dtype: bfloat16
+            enable_chunked_prefill: true
+            enable_prefix_caching: true
+            enforce_eager: true
+            expert_parallel_size: 1
+            free_cache_engine: true
+            gpu_memory_utilization: 0.7
+            ignore_eos: false
+            layered_summon: false
+            load_format: dummy
+            log_prob_max_token_len_per_gpu: 16384
+            log_prob_micro_batch_size: null
+            log_prob_micro_batch_size_per_gpu: 1
+            log_prob_use_dynamic_bsz: false
+            max_model_len: 7200
+            max_num_batched_tokens: 8192
+            max_num_seqs: 1024
+            mode: sync
+            multi_stage_wake_up: false
+            multi_turn:
+                _target_: verl.workers.config.MultiTurnConfig
+                enable: false
+                format: hermes
+                interaction_config_path: null
+                max_assistant_turns: null
+                max_parallel_calls: 1
+                max_tool_response_length: 256
+                max_user_turns: null
+                num_repeat_rollouts: null
+                tokenization_sanity_check_mode: strict
+                tool_config_path: null
+                tool_response_truncate_side: middle
+                use_inference_chat_template: false
+            "n": 1
+            name: vllm
+            over_sample_rate: 0
+            profiler:
+                _target_: verl.utils.profiler.ProfilerConfig
+                all_ranks: false
+                enable: false
+                ranks: []
+                save_path: outputs/profile
+                tool: null
+                tool_config:
+                    npu:
+                        _target_: verl.utils.profiler.config.NPUToolConfig
+                        analysis: true
+                        contents: []
+                        discrete: false
+                        level: level1
+                    nsys:
+                        _target_: verl.utils.profiler.config.NsightToolConfig
+                        discrete: false
+                    torch:
+                        _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                        step_end: null
+                        step_start: 0
+                    torch_memory:
+                        _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                        stack_depth: 32
+                        trace_alloc_max_entries: 100000
+            prompt_length: 1
+            response_length: 500
+            rollout_filter_metric: reward_variance
+            rollout_filter_ratio: 0.25
+            rollout_filter_type: largest
+            skip_dump_dir: /tmp/rollout_dump
+            skip_rollout: false
+            skip_tokenizer_init: true
+            temperature: 1
+            tensor_model_parallel_size: 1
+            top_k: -1
+            top_p: 1
+            trace:
+                _target_: verl.workers.config.TraceConfig
+                backend: null
+                token2text: false
+            update_weights_bucket_megabytes: 512
+            val_kwargs:
+                _target_: verl.workers.config.SamplingConfig
+                do_sample: true
+                "n": 1
+                temperature: 0.5
+                top_k: -1
+                top_p: 1
+agent_proxy:
+    value:
+        action_sep: '||'
+        enable_think: true
+        max_actions_per_turn: 1
+        max_context_window: -1
+        max_turn: 1
+        reward_normalization:
+            grouping: state
+            method: identity
+        use_turn_scores: false
+algorithm:
+    value:
+        _target_: verl.trainer.config.AlgoConfig
+        adv_estimator: gae
+        bi_level_gae: false
+        gamma: 1
+        high_level_gamma: 0.95
+        kl_ctrl:
+            _target_: verl.trainer.config.KLControlConfig
+            horizon: 10000
+            kl_coef: 0
+            target_kl: 0.1
+            type: fixed
+        kl_penalty: kl
+        lam: 1
+        norm_adv_by_std_in_grpo: true
+        pf_ppo:
+            reweight_method: pow
+            weight_pow: 2
+        use_kl_in_reward: false
+        use_pf_ppo: false
+critic:
+    value:
+        _target_: verl.workers.config.FSDPCriticConfig
+        checkpoint:
+            _target_: verl.trainer.config.CheckpointConfig
+            async_save: false
+            load_contents:
+                - model
+                - optimizer
+                - extra
+            save_contents:
+                - model
+                - optimizer
+                - extra
+        cliprange_value: 0.5
+        enable: null
+        forward_max_token_len_per_gpu: 32768
+        forward_micro_batch_size: null
+        forward_micro_batch_size_per_gpu: 1
+        grad_clip: 1
+        loss_agg_mode: token-mean
+        model:
+            _target_: verl.workers.config.FSDPCriticModelCfg
+            enable_activation_offload: false
+            enable_gradient_checkpointing: true
+            external_lib: null
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                entropy_checkpointing: false
+                entropy_from_logits_with_chunking: false
+                forward_only: false
+                forward_prefetch: false
+                fsdp_size: -1
+                model_dtype: fp32
+                offload_policy: false
+                optimizer_offload: false
+                param_offload: false
+                reshard_after_forward: true
+                strategy: fsdp
+                ulysses_sequence_parallel_size: 1
+                use_orig_params: false
+                use_torch_compile: true
+                wrap_policy:
+                    min_num_params: 0
+            lora_alpha: 16
+            lora_rank: 0
+            path: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            target_modules: all-linear
+            tokenizer_path: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            trust_remote_code: false
+            use_remove_padding: false
+            use_shm: false
+        optim:
+            _target_: verl.workers.config.FSDPOptimizerConfig
+            betas:
+                - 0.9
+                - 0.999
+            clip_grad: 1
+            lr: 1e-05
+            lr_warmup_steps: -1
+            lr_warmup_steps_ratio: 0
+            min_lr_ratio: 0
+            num_cycles: 0.5
+            total_training_steps: 200
+            warmup_style: constant
+            weight_decay: 0.01
+        ppo_epochs: 1
+        ppo_max_token_len_per_gpu: 32768
+        ppo_micro_batch_size: null
+        ppo_micro_batch_size_per_gpu: 1
+        ppo_mini_batch_size: 32
+        profiler:
+            _target_: verl.utils.profiler.ProfilerConfig
+            all_ranks: false
+            enable: false
+            ranks: []
+            save_path: outputs/profile
+            tool: null
+            tool_config:
+                npu:
+                    _target_: verl.utils.profiler.config.NPUToolConfig
+                    analysis: true
+                    contents: []
+                    discrete: false
+                    level: level1
+                nsys:
+                    _target_: verl.utils.profiler.config.NsightToolConfig
+                    discrete: false
+                torch:
+                    _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                    step_end: null
+                    step_start: 0
+                torch_memory:
+                    _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                    stack_depth: 32
+                    trace_alloc_max_entries: 100000
+        rollout_n: 1
+        shuffle: false
+        strategy: fsdp
+        ulysses_sequence_parallel_size: 1
+        use_dynamic_bsz: false
+ctx_manager:
+    value:
+        generation:
+            gen_config:
+                kwargs: null
+                response_length: 500
+                temperature: 1
+                top_k: -1
+                top_p: 1
+custom_envs:
+    value:
+        Bandit:
+            env_config:
+                split: train
+            env_instruction: ""
+            env_type: bandit
+            max_actions_per_traj: 1
+            max_tokens: 100
+        BanditTest:
+            env_config:
+                split: test
+            env_instruction: ""
+            env_type: bandit
+            max_actions_per_traj: 1
+            max_tokens: 100
+        CoordFrozenLake:
+            env_config:
+                observation_format: grid_coord
+            env_instruction: |
+                You are solving the FrozenLake puzzle. The observation includes both a symbol grid and zero-indexed coordinates for the start, goal, player, and any holes.
+                Coordinates range from the top-left corner (0, 0) to the bottom-right corner (5, 5).
+                Beware that the ice is slippery, so the agent might slide and end up in an unintended tile.
+                Respond with a sequence of actions such as <answer>Left || Up || Up</answer>.
+            env_type: frozen_lake
+            max_actions_per_traj: 25
+            max_tokens: 120
+        CoordSokoban:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                max_steps: 100
+                num_boxes: 2
+                observation_format: grid_coord
+            env_instruction: "You are solving the Sokoban puzzle. You are the player and you need to push all boxes to targets.\nYou are provided with a symbol grid and the zero-indexed coordinates of the player, each box, and each target. \nCoordinates range from the top-left corner (0, 0) to the bottom-right corner (5, 5). \nWhen you are exactly next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box.\nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>.\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 120
+        Countdown:
+            env_config: null
+            env_instruction: 'You are solving the Countdown puzzle. You should use the num list to create an equation that equals the target. Example answer format: <think> To find an equation using [3, 5, 2] to get 4. Let''s check 2 + 5 = 7, 7 - 3 = 4. So the answer is 2 + 5 - 3 = 4. </think><answer>2 + 5 - 3</answer>'
+            env_type: countdown
+            max_actions_per_traj: 1
+            max_tokens: 100
+        FrozenLake:
+            env_config: null
+            env_instruction: 'You are solving the FrozenLake puzzle. Forbid the whole and go to the target. You may move to the unintended direction due to the slippery ice. Example answer format: <think>To forbid the hole and go to the target, I should go left then go up.</think><answer>Left || Up</answer>'
+            env_type: frozen_lake
+            max_actions_per_traj: 25
+            max_tokens: 100
+        LargerSokoban:
+            env_config:
+                dim_x: 8
+                dim_y: 8
+                max_steps: 100
+                num_boxes: 2
+                search_depth: 10
+            env_instruction: "You are solving the Sokoban puzzle. \nYou are the player and you need to push all boxes to targets. \nWhen you are right next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box. \nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        Lean:
+            env_config: null
+            env_instruction: You are a Lean theorem prover. Given a Lean theorem statement, propose a sequence of tactics that completes the proof. Think step by step about which tactics to apply next. Provide tactics separated by '||', for example <answer>intro || simp || rfl</answer>.
+            env_type: lean
+            max_actions_per_traj: 30
+            max_tokens: 512
+        MediumSudoku:
+            env_config:
+                difficulty: medium
+                grid_size: 9
+                max_steps: 81
+                render_format: with_feedback
+                show_conflicts: true
+                show_valid_numbers: true
+            env_instruction: |
+                You are solving a Sudoku puzzle. Fill in the grid so that every row, column, and 3x3 box contains the numbers 1-9 without repetition.
+                Initial cells are shown in [brackets] and cannot be modified. Empty cells are shown as dots (.).
+                Place numbers one at a time using the format: <answer>place 5 at row 2 col 3</answer> or <answer>2,3,5</answer>
+                The environment will provide feedback on valid/invalid moves and show conflicts if any occur.
+            env_type: sudoku
+            max_actions_per_traj: 30
+            max_tokens: 150
+            max_workers: 32
+            parallel_friendly: false
+        MetamathQA:
+            env_config: null
+            env_instruction: 'You are solving Math problems. '
+            env_type: metamathqa
+            max_actions_per_traj: 1
+            max_tokens: 100
+        SimpleSokoban:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                max_steps: 100
+                num_boxes: 2
+            env_instruction: "You are solving the Sokoban puzzle. \nYou are the player and you need to push all boxes to targets. \nWhen you are right next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box. \nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        SimpleSudoku:
+            env_config:
+                difficulty: easy
+                grid_size: 4
+                max_steps: 20
+                render_format: with_feedback
+                show_conflicts: false
+                show_valid_numbers: false
+            env_instruction: |
+                You are solving a Sudoku puzzle. Fill in the grid so that every row, column, and 2x2 box contains the numbers 1-4 without repetition.
+                Initial cells are shown in [brackets] and cannot be modified. Empty cells are shown as dots (.).
+                Place numbers one at a time using the format, for example: <answer>place 1 at row 2 col 3</answer> or <answer>1,2,3</answer>
+                The environment will provide feedback on valid/invalid moves and show conflicts if any occur.
+            env_type: sudoku
+            max_actions_per_traj: 20
+            max_tokens: 150
+            max_workers: 32
+            parallel_friendly: false
+        SokobanDifferentGridVocab:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                grid_lookup:
+                    "0": W
+                    "1": .
+                    "2": G
+                    "3": C
+                    "4": B
+                    "5": A
+                    "6": '@'
+                grid_vocab:
+                    .: empty
+                    '@': player on target
+                    A: player
+                    B: box
+                    C: box on target
+                    G: target
+                    W: wall
+                max_steps: 100
+                num_boxes: 1
+                search_depth: 30
+            env_instruction: "You are solving the Sokoban puzzle. \nYou are the player and you need to push all boxes to targets. \nWhen you are right next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box. \nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        VisualSimpleSokoban:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                max_steps: 100
+                num_boxes: 1
+                render_mode: rgb_array
+            env_instruction: You are solving the Sokoban puzzle. You are the player and you need to push all boxes to targets. When you are right next to a box, you can push it by moving in the same direction. You cannot push a box through a wall, and you cannot pull a box. The answer should be a sequence of actions, like <answer>Right || Right || Up</answer>
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        WebShop:
+            env_config:
+                dataset: small
+            env_instruction: 'You are browsing an online shop. Based on the instruction, buy a product that close to the production description. You need to search, read the search results, pick a product, choose the size and color and buy. You should only choose action from the available actions list provided later.  Example process: I need a gingko light and 20x20 pillow cover that is hand painted. First search[gingko light 20x20 pillow cover hand painted], answer format: <answer>search[blanket with fleece throw]</answer>. Valid answer is search[<keywords>] or click[<clickable>].'
+            env_type: webshop
+            max_actions_per_traj: 9
+            max_tokens: 200
+        blackjack:
+            env_config: null
+            env_instruction: |
+                You are playing Blackjack against a dealer. The dealer must hit on 16 or less and stand on 17 or more.
+                Choose either Stick or Hit. Respond with a single action.
+                Example: <answer>Hit</answer>
+            env_type: blackjack
+            max_actions_per_traj: 10
+            max_tokens: 64
+        game_2048:
+            env_config: null
+            env_instruction: |
+                You are playing the 2048 game on a 4x4 grid. Merge equal tiles by sliding Up, Right, Down, or Left.
+                If a move is invalid (no tiles move), a small penalty is applied. Respond with a single action.
+                Example: <answer>Up</answer>
+            env_type: game_2048
+            max_actions_per_traj: 700
+            max_tokens: 8192
+        rubikscube:
+            env_config:
+                max_steps: 20
+                render_mode: text
+                scramble_depth: 1
+            env_instruction: |
+                You are solving a 2x2 Rubik's Cube (Pocket Cube). The goal is to restore the cube so that each of the faces consists of a single, unique color.
+                Available actions use standard Singmaster notation for face rotations: U, U', D, D', L, L', R, R', F, F', B, B'.
+                - Faces: U (Up), D (Down), L (Left), R (Right), F (Front), B (Back).
+                - Modifiers: A letter alone means 90° clockwise (e.g., 'R'). A letter with prime (') means 90° counter-clockwise (e.g., "R'").
+                Respond with a sequence of actions separated by "||".
+                Example: <answer>U</answer>
+            env_type: rubikscube
+            max_actions_per_traj: 20
+            max_tokens: 96
+custom_reward_function:
+    value:
+        name: compute_score
+        path: null
+data:
+    value:
+        custom_cls:
+            name: null
+            path: null
+        datagen:
+            name: null
+            path: null
+        dataloader_num_workers: 8
+        filter_overlong_prompts: false
+        filter_overlong_prompts_workers: 1
+        image_key: images
+        max_prompt_length: null
+        max_response_length: null
+        prompt_key: prompt
+        return_full_prompt: false
+        return_multi_modal_inputs: true
+        return_raw_chat: false
+        return_raw_input_ids: false
+        reward_fn_key: data_source
+        sampler:
+            class_name: null
+            class_path: null
+        shuffle: true
+        tokenizer: null
+        train_batch_size: 128
+        train_files: ~/data/rlhf/gsm8k/train.parquet
+        truncation: error
+        trust_remote_code: false
+        use_shm: false
+        val_batch_size: null
+        val_files: ~/data/rlhf/gsm8k/test.parquet
+        validation_shuffle: false
+        video_key: videos
+enable_response_mask:
+    value: true
+es_manager:
+    value:
+        format_penalty: -0.1
+        train:
+            env_configs:
+                n_groups:
+                    - 8
+                tags:
+                    - Bandit
+            env_groups: 8
+            group_size: 16
+        val:
+            env_configs:
+                n_groups:
+                    - 256
+                    - 256
+                tags:
+                    - Bandit
+                    - BanditTest
+            env_groups: 512
+            group_size: 1
+global_profiler:
+    value:
+        _target_: verl.utils.profiler.ProfilerConfig
+        global_tool_config:
+            nsys:
+                _target_: verl.utils.profiler.config.NsightToolConfig
+                controller_nsight_options:
+                    cuda-graph-trace: graph
+                    cuda-memory-usage: "true"
+                    trace: cuda,nvtx,cublas,ucx
+                discrete: false
+                worker_nsight_options:
+                    capture-range: cudaProfilerApi
+                    capture-range-end: null
+                    cuda-graph-trace: graph
+                    cuda-memory-usage: "true"
+                    kill: none
+                    trace: cuda,nvtx,cublas,ucx
+            torch_memory:
+                context: all
+                stack_depth: 32
+                stacks: all
+                trace_alloc_max_entries: 100000
+        profile_continuous_steps: false
+        save_path: outputs/profile
+        steps: null
+        tool: null
+grpo_advantage_length_weight:
+    value: false
+lora:
+    value:
+        alpha: 16
+        rank: 0
+        target_modules: all-linear
+micro_batch_size_per_gpu:
+    value: 1
+model_path:
+    value: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+ppo_mini_batch_size:
+    value: 32
+ray_kwargs:
+    value:
+        ray_init:
+            num_cpus: null
+        timeline_json_file: null
+reward_model:
+    value:
+        enable: false
+        enable_resource_pool: false
+        forward_max_token_len_per_gpu: 32768
+        launch_reward_fn_async: false
+        max_length: null
+        micro_batch_size: null
+        micro_batch_size_per_gpu: null
+        model:
+            external_lib: null
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                forward_prefetch: false
+                fsdp_size: -1
+                param_offload: false
+                reshard_after_forward: true
+                wrap_policy:
+                    min_num_params: 0
+            input_tokenizer: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            path: ~/models/FsfairX-LLaMA3-RM-v0.1
+            trust_remote_code: false
+            use_fused_kernels: false
+            use_remove_padding: false
+            use_shm: false
+        n_gpus_per_node: 0
+        nnodes: 0
+        profiler:
+            _target_: verl.utils.profiler.ProfilerConfig
+            all_ranks: false
+            enable: false
+            ranks: []
+            save_path: outputs/profile
+            tool: null
+            tool_config:
+                npu:
+                    _target_: verl.utils.profiler.config.NPUToolConfig
+                    analysis: true
+                    contents: []
+                    discrete: false
+                    level: level1
+                nsys:
+                    _target_: verl.utils.profiler.config.NsightToolConfig
+                    discrete: false
+                torch:
+                    _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                    step_end: null
+                    step_start: 0
+                torch_memory:
+                    _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                    stack_depth: 32
+                    trace_alloc_max_entries: 100000
+        reward_manager: naive
+        sandbox_fusion:
+            max_concurrent: 64
+            memory_limit_mb: 1024
+            url: null
+        strategy: fsdp
+        ulysses_sequence_parallel_size: 1
+        use_dynamic_bsz: false
+seed:
+    value:
+        train: 10000
+        val: 123
+system:
+    value:
+        CUDA_VISIBLE_DEVICES: 6,7
+trainer:
+    value:
+        balance_batch: true
+        critic_warmup: 0
+        default_hdfs_dir: null
+        default_local_dir: /mnt/general/wanghy/RAGEN/saves/
+        del_local_ckpt_after_load: false
+        device: cuda
+        esi_redundant_time: 0
+        experiment_name: bandit-starpo
+        generations_to_log_to_wandb:
+            train: 128
+            val: 20
+        local_log_dir: results/
+        log_val_generations: 0
+        logger:
+            - console
+            - wandb
+        max_actor_ckpt_to_keep: 1
+        max_critic_ckpt_to_keep: 1
+        n_gpus_per_node: 2
+        nnodes: 1
+        project_name: ragen_latest_llama_1B_it
+        ray_wait_register_center_timeout: 300
+        resume_from_path: null
+        resume_mode: auto
+        rollout_data_dir: null
+        save_freq: -1
+        test_freq: 10
+        total_epochs: 30
+        total_training_steps: 200
+        use_legacy_worker_impl: auto
+        val_before_train: true
+        val_only: false
+        validation_data_dir: null
+        validation_steps: 1

wandb/run-20251222_182729-6i1mkj49/files/requirements.txt

@@ -0,0 +1,316 @@
+ragen==0.1
+ragen==0.1
+colorama==0.4.6
+psutil==7.1.2
+ragen==0.1
+setuptools==80.9.0
+wheel==0.45.1
+pip==25.3
+ragen==0.1
+pytz==2025.2
+nvidia-cusparselt-cu12==0.6.2
+mpmath==1.3.0
+typing_extensions==4.15.0
+sympy==1.13.1
+pylatexenc==2.10
+nvidia-cusolver-cu12==11.6.1.9
+numpy==1.26.4
+Markdown==3.10
+absl-py==2.3.1
+pyasn1_modules==0.4.2
+Werkzeug==3.1.3
+networkx==3.3
+MarkupSafe==2.1.5
+fsspec==2025.9.0
+filelock==3.19.1
+nvidia-cudnn-cu12==9.1.0.70
+Jinja2==3.1.6
+tensorboard==2.20.0
+triton==3.2.0
+pillow==11.3.0
+xxhash==3.6.0
+nvidia-nvtx-cu12==12.4.127
+nvidia-nvjitlink-cu12==12.4.127
+nvidia-nccl-cu12==2.21.5
+nvidia-curand-cu12==10.3.5.147
+wrapt==2.0.1
+nvidia-cufft-cu12==11.2.1.3
+pyjnius==1.7.0
+nvidia-cuda-runtime-cu12==12.4.127
+nvidia-cuda-nvrtc-cu12==12.4.127
+websockets==15.0.1
+nvidia-cuda-cupti-cu12==12.4.127
+zipp==3.23.0
+nvidia-cublas-cu12==12.4.5.8
+nvidia-cusparse-cu12==12.3.1.170
+flash_attn==2.8.3
+rsa==4.9.1
+einops==0.8.1
+py-cpuinfo==9.0.0
+pure_eval==0.2.3
+ptyprocess==0.7.0
+gym-notices==0.1.0
+flatbuffers==25.9.23
+fastrlock==0.8.3
+Farama-Notifications==0.0.4
+faiss-gpu==1.7.2
+antlr4-python3-runtime==4.9.3
+wcwidth==0.2.14
+wasabi==1.1.3
+uvloop==0.22.1
+urllib3==2.5.0
+tzdata==2025.2
+typing-inspection==0.4.2
+traitlets==5.14.3
+tqdm==4.67.1
+tomli==2.3.0
+threadpoolctl==3.6.0
+tenacity==9.1.2
+tabulate==0.9.0
+spacy-loggers==1.0.5
+spacy-legacy==3.0.12
+soupsieve==2.8
+sniffio==1.3.1
+smmap==5.0.2
+six==1.17.0
+shellingham==1.5.4
+sentencepiece==0.2.1
+scipy==1.15.3
+safetensors==0.6.2
+rpds-py==0.29.0
+rignore==0.7.6
+regex==2025.11.3
+RapidFuzz==3.14.3
+rank-bm25==0.2.2
+pyzmq==27.1.0
+PyYAML==6.0.3
+pytokens==0.3.0
+python-multipart==0.0.20
+python-json-logger==4.0.0
+python-dotenv==1.2.1
+PySocks==1.7.1
+pyparsing==3.2.5
+PyJWT==2.10.1
+Pygments==2.19.2
+pygame==2.6.1
+pydantic_core==2.41.5
+pycparser==2.23
+pycountry==24.6.1
+pybind11==3.0.1
+pyarrow==22.0.0
+psutil==7.1.3
+protobuf==4.25.8
+propcache==0.4.1
+prometheus_client==0.23.1
+platformdirs==4.5.0
+pexpect==4.9.0
+pathspec==0.12.1
+partial-json-parser==0.2.1.1.post6
+parso==0.8.5
+packaging==25.0
+orjson==3.11.4
+opentelemetry-semantic-conventions-ai==0.4.13
+opencv-python-headless==4.12.0.88
+ninja==1.13.0
+nest-asyncio==1.6.0
+mypy_extensions==1.1.0
+murmurhash==1.0.15
+multidict==6.7.0
+msgspec==0.19.0
+msgpack==1.1.2
+mdurl==0.1.2
+loguru==0.7.3
+llvmlite==0.44.0
+llguidance==0.7.30
+lark==1.2.2
+kiwisolver==1.4.9
+joblib==1.5.2
+jiter==0.12.0
+itsdangerous==2.2.0
+interegular==0.3.3
+ImageIO==2.37.2
+idna==3.11
+humanfriendly==10.0
+httpx-sse==0.4.3
+httptools==0.7.1
+html2text==2025.4.15
+hf-xet==1.2.0
+h11==0.16.0
+grpcio==1.76.0
+frozenlist==1.8.0
+fonttools==4.60.1
+executing==2.2.1
+exceptiongroup==1.3.0
+eval_type_backport==0.2.2
+docstring_parser==0.17.0
+dnspython==2.8.0
+distro==1.9.0
+diskcache==5.6.3
+dill==0.4.0
+decorator==5.2.1
+debugpy==1.8.17
+Cython==3.2.1
+cymem==2.0.13
+cycler==0.12.1
+cupy-cuda12x==13.6.0
+contourpy==1.3.2
+colorama==0.4.6
+codetiming==1.4.0
+cloudpickle==3.1.2
+cloudpathlib==0.23.0
+click==8.3.1
+charset-normalizer==3.4.4
+certifi==2025.11.12
+catalogue==2.0.10
+cachetools==6.2.2
+blis==1.3.2
+blinker==1.9.0
+blake3==1.0.8
+attrs==25.4.0
+async-timeout==5.0.1
+asttokens==3.0.1
+astor==0.8.1
+annotated-types==0.7.0
+annotated-doc==0.0.4
+airportsdata==20250909
+aiohappyeyeballs==2.6.1
+yarl==1.22.0
+uvicorn==0.38.0
+typer-slim==0.20.0
+thefuzz==0.22.1
+stack-data==0.6.3
+srsly==2.5.1
+smart_open==7.5.0
+sentry-sdk==2.44.0
+scikit-learn==1.7.2
+requests==2.32.5
+referencing==0.37.0
+python-dateutil==2.9.0.post0
+pydantic==2.12.4
+prompt_toolkit==3.0.52
+preshed==3.0.11
+opentelemetry-proto==1.26.0
+omegaconf==2.3.0
+numba==0.61.2
+nltk==3.9.2
+multiprocess==0.70.18
+matplotlib-inline==0.2.1
+markdown-it-py==4.0.0
+jedi==0.19.2
+importlib_metadata==8.0.0
+httpcore==1.0.9
+gymnasium==1.2.2
+gym==0.26.2
+googleapis-common-protos==1.72.0
+gitdb==4.0.12
+gguf==0.17.1
+Flask==3.1.2
+faiss-cpu==1.13.0
+email-validator==2.3.0
+depyf==0.18.0
+Deprecated==1.3.1
+coloredlogs==15.0.1
+cffi==2.0.0
+black==25.11.0
+beautifulsoup4==4.14.2
+anyio==4.11.0
+aiosignal==1.4.0
+watchfiles==1.1.1
+tiktoken==0.12.0
+starlette==0.49.3
+sse-starlette==3.0.3
+rich==14.2.0
+pydantic-settings==2.12.0
+pydantic-extra-types==2.10.6
+pandas==2.3.3
+opentelemetry-exporter-otlp-proto-common==1.26.0
+onnxruntime==1.23.2
+matplotlib==3.10.7
+lm-format-enforcer==0.10.12
+jsonschema-specifications==2025.9.1
+ipython==8.37.0
+hydra-core==1.3.2
+huggingface-hub==0.36.0
+httpx==0.28.1
+gym-sokoban==0.0.6
+GitPython==3.1.45
+cryptography==46.0.3
+confection==0.1.5
+cleantext==1.1.4
+aiohttp==3.13.2
+weasel==0.4.3
+wandb==0.23.0
+typer==0.19.2
+virtualenv==20.35.4
+pyasn1==0.6.1
+torch==2.6.0
+tokenizers==0.22.1
+thinc==8.3.9
+rich-toolkit==0.15.1
+prometheus-fastapi-instrumentator==7.1.0
+openai==2.8.0
+jsonschema==4.25.1
+gdown==5.2.0
+fastapi==0.121.2
+anthropic==0.73.0
+xformers==0.0.29.post2
+transformers==4.57.1
+proto-plus==1.26.1
+torchvision==0.21.0
+torchdata==0.11.0
+torchaudio==2.6.0
+together==1.5.30
+tensordict==0.8.3
+spacy==3.8.9
+ray==2.51.1
+outlines_core==0.1.26
+mistral_common==1.8.5
+mcp==1.21.1
+fastapi-cloud-cli==0.3.1
+fastapi-cli==0.0.16
+datasets==4.4.1
+accelerate==1.11.0
+xgrammar==0.1.18
+pyserini==1.2.0
+peft==0.18.0
+outlines==0.1.11
+opentelemetry-exporter-otlp-proto-http==1.26.0
+opentelemetry-exporter-otlp-proto-grpc==1.26.0
+kimina-client==0.2.1
+compressed-tensors==0.9.3
+opentelemetry-exporter-otlp==1.26.0
+vllm==0.8.5
+py-spy==0.4.1
+opencensus-context==0.1.3
+distlib==0.4.0
+colorful==0.5.8
+tensorboard-data-server==0.7.2
+google-auth==2.43.0
+google-api-core==2.28.1
+aiohttp-cors==0.8.1
+opentelemetry-exporter-prometheus==0.59b0
+opencensus==0.11.4
+verl==0.5.0.dev0
+opentelemetry-api==1.38.0
+opentelemetry-semantic-conventions==0.59b0
+opentelemetry-sdk==1.38.0
+typeguard==4.4.4
+shtab==1.8.0
+tyro==0.9.35
+verl==0.5.0.dev0
+autocommand==2.2.2
+backports.tarfile==1.2.0
+importlib_metadata==8.0.0
+inflect==7.3.1
+jaraco.collections==5.1.0
+jaraco.context==5.3.0
+jaraco.functools==4.0.1
+jaraco.text==3.12.1
+more-itertools==10.3.0
+packaging==24.2
+platformdirs==4.2.2
+tomli==2.0.1
+typeguard==4.3.0
+typing_extensions==4.12.2
+wheel==0.45.1
+zipp==3.19.2

wandb/run-20251222_182729-6i1mkj49/files/wandb-metadata.json

@@ -0,0 +1,107 @@
+{
+  "os": "Linux-5.14.0-284.25.1.el9_2.x86_64-x86_64-with-glibc2.35",
+  "python": "CPython 3.10.19",
+  "startedAt": "2025-12-22T10:27:29.079503Z",
+  "args": [
+    "--node-ip-address=10.119.27.78",
+    "--node-manager-port=45457",
+    "--object-store-name=/tmp/ray/session_2025-12-22_18-26-14_802758_3429006/sockets/plasma_store",
+    "--raylet-name=/tmp/ray/session_2025-12-22_18-26-14_802758_3429006/sockets/raylet",
+    "--redis-address=None",
+    "--metrics-agent-port=41966",
+    "--logging-rotate-bytes=536870912",
+    "--logging-rotate-backup-count=5",
+    "--runtime-env-agent-port=57932",
+    "--gcs-address=10.119.27.78:62444",
+    "--session-name=session_2025-12-22_18-26-14_802758_3429006",
+    "--temp-dir=/tmp/ray",
+    "--webui=127.0.0.1:8265",
+    "--cluster-id=f5d582d4084ab6f69c0ebe1c1d779c63e370565da4ba7fe85624529d",
+    "--startup-token=112",
+    "--worker-launch-time-ms=1766399184671",
+    "--node-id=d98b0f6bb40ee78c635322e643c17204ee3f035aa813492ed75c5c3e",
+    "--runtime-env-hash=-839022692"
+  ],
+  "program": "/root/local/miniconda3/envs/ragen/lib/python3.10/site-packages/ray/_private/workers/default_worker.py",
+  "git": {
+    "remote": "https://github.com/mll-lab-nu/RAGEN.git",
+    "commit": "8d73639b99b38265453f898b8d6af7d4af50d56e"
+  },
+  "email": "haoyu-wa22@mails.tsinghua.edu.cn",
+  "root": "/mnt/general/wanghy/RAGEN",
+  "host": "pt-4d654cf4576f4d23ad3d3919f12932fe-worker-0",
+  "executable": "/root/local/miniconda3/envs/ragen/bin/python",
+  "cpu_count": 64,
+  "cpu_count_logical": 128,
+  "gpu": "NVIDIA H100 80GB HBM3",
+  "gpu_count": 8,
+  "disk": {
+    "/": {
+      "total": "5153960755200",
+      "used": "471657721856"
+    }
+  },
+  "memory": {
+    "total": "2163642122240"
+  },
+  "gpu_nvidia": [
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-35e2d43d-4067-82ce-90d4-def9e389bf28"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-af4135e3-88f2-e9ac-518d-502c75a85429"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-d7fdeeba-fe9b-ec03-d9f7-6724fe4266b5"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-ccc4f668-3882-5a8e-2c07-c5cd08f6f666"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-7b73c0cf-d3d5-e10c-7176-a43be1e41001"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-81b58d94-5d1f-8ec2-f9d2-fd56172ed177"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-03e8bc66-3b44-6794-49fd-5392fbdda6d1"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-b7cf0ec6-7c29-1179-dceb-09565da51890"
+    }
+  ],
+  "cudaVersion": "12.4",
+  "writerId": "i0axz6jgeasoze55ozdefzq2sh05xke5"
+}

wandb/run-20251222_182729-6i1mkj49/files/wandb-summary.json

@@ -0,0 +1 @@
+{"_wandb":{"runtime":30},"_runtime":30}

wandb/run-20251222_184600-j1l9fcd9/files/config.yaml

@@ -0,0 +1,962 @@
+_wandb:
+    value:
+        cli_version: 0.23.0
+        e:
+            2t97y3f3xn1qhzjmtyvbbboq8vk1twrh:
+                args:
+                    - --node-ip-address=10.119.27.78
+                    - --node-manager-port=42449
+                    - --object-store-name=/tmp/ray/session_2025-12-22_18-44-43_109609_3743349/sockets/plasma_store
+                    - --raylet-name=/tmp/ray/session_2025-12-22_18-44-43_109609_3743349/sockets/raylet
+                    - --redis-address=None
+                    - --metrics-agent-port=58057
+                    - --logging-rotate-bytes=536870912
+                    - --logging-rotate-backup-count=5
+                    - --runtime-env-agent-port=64142
+                    - --gcs-address=10.119.27.78:40383
+                    - --session-name=session_2025-12-22_18-44-43_109609_3743349
+                    - --temp-dir=/tmp/ray
+                    - --webui=127.0.0.1:8265
+                    - --cluster-id=bb05b0e0daa61e2e69d651e5343e9354c9782bb454bfb3b4a69c706a
+                    - --startup-token=112
+                    - --worker-launch-time-ms=1766400294644
+                    - --node-id=7db3272367b4e4a9a8c9a20b5b50b37cf6f00b4f6eae0f10b698a824
+                    - --runtime-env-hash=-839022692
+                cpu_count: 64
+                cpu_count_logical: 128
+                cudaVersion: "12.4"
+                disk:
+                    /:
+                        total: "5153960755200"
+                        used: "471669112832"
+                email: haoyu-wa22@mails.tsinghua.edu.cn
+                executable: /root/local/miniconda3/envs/ragen/bin/python
+                git:
+                    commit: 8d73639b99b38265453f898b8d6af7d4af50d56e
+                    remote: https://github.com/mll-lab-nu/RAGEN.git
+                gpu: NVIDIA H100 80GB HBM3
+                gpu_count: 8
+                gpu_nvidia:
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-35e2d43d-4067-82ce-90d4-def9e389bf28
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-af4135e3-88f2-e9ac-518d-502c75a85429
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-d7fdeeba-fe9b-ec03-d9f7-6724fe4266b5
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-ccc4f668-3882-5a8e-2c07-c5cd08f6f666
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-7b73c0cf-d3d5-e10c-7176-a43be1e41001
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-81b58d94-5d1f-8ec2-f9d2-fd56172ed177
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-03e8bc66-3b44-6794-49fd-5392fbdda6d1
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-b7cf0ec6-7c29-1179-dceb-09565da51890
+                host: pt-4d654cf4576f4d23ad3d3919f12932fe-worker-0
+                memory:
+                    total: "2163642122240"
+                os: Linux-5.14.0-284.25.1.el9_2.x86_64-x86_64-with-glibc2.35
+                program: /root/local/miniconda3/envs/ragen/lib/python3.10/site-packages/ray/_private/workers/default_worker.py
+                python: CPython 3.10.19
+                root: /mnt/general/wanghy/RAGEN
+                startedAt: "2025-12-22T10:46:00.095282Z"
+                writerId: 2t97y3f3xn1qhzjmtyvbbboq8vk1twrh
+        m: []
+        python_version: 3.10.19
+        t:
+            "1":
+                - 1
+                - 5
+                - 11
+                - 30
+                - 41
+                - 49
+                - 50
+                - 51
+                - 53
+                - 71
+                - 95
+                - 97
+                - 98
+                - 105
+            "2":
+                - 1
+                - 5
+                - 11
+                - 30
+                - 41
+                - 49
+                - 50
+                - 51
+                - 53
+                - 71
+                - 95
+                - 97
+                - 98
+                - 105
+            "3":
+                - 2
+                - 13
+                - 16
+                - 61
+            "4": 3.10.19
+            "5": 0.23.0
+            "6": 4.57.1
+            "12": 0.23.0
+            "13": linux-x86_64
+actor_rollout_ref:
+    value:
+        actor:
+            _target_: verl.workers.config.FSDPActorConfig
+            checkpoint:
+                _target_: verl.trainer.config.CheckpointConfig
+                async_save: false
+                load_contents:
+                    - model
+                    - optimizer
+                    - extra
+                save_contents:
+                    - model
+                    - optimizer
+                    - extra
+            clip_ratio: 0.2
+            clip_ratio_c: 3
+            clip_ratio_high: 0.28
+            clip_ratio_low: 0.2
+            entropy_checkpointing: false
+            entropy_coeff: 0.001
+            entropy_from_logits_with_chunking: false
+            freeze_vision_tower: false
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                entropy_checkpointing: false
+                entropy_from_logits_with_chunking: false
+                forward_only: false
+                forward_prefetch: false
+                fsdp_size: -1
+                model_dtype: fp32
+                offload_policy: false
+                optimizer_offload: false
+                param_offload: false
+                reshard_after_forward: true
+                strategy: fsdp
+                ulysses_sequence_parallel_size: 1
+                use_orig_params: false
+                use_torch_compile: true
+                wrap_policy:
+                    min_num_params: 0
+            grad_clip: 1
+            grpo_advantage_length_weight: false
+            kl_loss_coef: 0
+            kl_loss_type: kl
+            loss_agg_mode: token-mean
+            optim:
+                _target_: verl.workers.config.FSDPOptimizerConfig
+                betas:
+                    - 0.9
+                    - 0.999
+                clip_grad: 1
+                lr: 1e-06
+                lr_warmup_steps: -1
+                lr_warmup_steps_ratio: 0
+                min_lr_ratio: 0
+                num_cycles: 0.5
+                total_training_steps: 200
+                warmup_style: constant
+                weight_decay: 0.01
+            policy_loss:
+                _target_: verl.workers.config.PolicyLossConfig
+                clip_cov_lb: 1
+                clip_cov_ratio: 0.0002
+                clip_cov_ub: 5
+                kl_cov_ratio: 0.0002
+                loss_mode: vanilla
+                ppo_kl_coef: 0.1
+            ppo_epochs: 1
+            ppo_max_token_len_per_gpu: 16384
+            ppo_micro_batch_size: null
+            ppo_micro_batch_size_per_gpu: 1
+            ppo_mini_batch_size: 32
+            profiler:
+                _target_: verl.utils.profiler.ProfilerConfig
+                all_ranks: false
+                enable: false
+                ranks: []
+                save_path: outputs/profile
+                tool: null
+                tool_config:
+                    npu:
+                        _target_: verl.utils.profiler.config.NPUToolConfig
+                        analysis: true
+                        contents: []
+                        discrete: false
+                        level: level1
+                    nsys:
+                        _target_: verl.utils.profiler.config.NsightToolConfig
+                        discrete: false
+                    torch:
+                        _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                        step_end: null
+                        step_start: 0
+                    torch_memory:
+                        _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                        stack_depth: 32
+                        trace_alloc_max_entries: 100000
+            shuffle: false
+            strategy: fsdp
+            tis_imp_ratio_cap: -1
+            ulysses_sequence_parallel_size: 1
+            use_dynamic_bsz: false
+            use_fused_kernels: false
+            use_kl_loss: false
+            use_ref: true
+            use_remove_padding: false
+            use_torch_compile: true
+        hybrid_engine: true
+        model:
+            _target_: verl.workers.config.HFModelConfig
+            custom_chat_template: null
+            enable_activation_offload: false
+            enable_gradient_checkpointing: true
+            exclude_modules: null
+            external_lib: null
+            fused_kernel_options:
+                impl_backend: torch
+            hf_config_path: null
+            lora_alpha: 16
+            lora_rank: 0
+            path: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            target_modules: all-linear
+            tokenizer_path: null
+            torch_dtype: bfloat16
+            trust_remote_code: false
+            use_fused_kernels: false
+            use_liger: false
+            use_remove_padding: false
+            use_shm: false
+        nccl_timeout: 600
+        ref:
+            entropy_checkpointing: false
+            entropy_from_logits_with_chunking: false
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                entropy_checkpointing: false
+                entropy_from_logits_with_chunking: false
+                forward_only: false
+                forward_prefetch: false
+                fsdp_size: -1
+                model_dtype: fp32
+                offload_policy: false
+                optimizer_offload: false
+                param_offload: false
+                reshard_after_forward: true
+                strategy: fsdp
+                ulysses_sequence_parallel_size: 1
+                use_orig_params: false
+                use_torch_compile: true
+                wrap_policy:
+                    min_num_params: 0
+            log_prob_max_token_len_per_gpu: 16384
+            log_prob_micro_batch_size: null
+            log_prob_micro_batch_size_per_gpu: 1
+            log_prob_use_dynamic_bsz: false
+            model: null
+            profiler:
+                _target_: verl.utils.profiler.ProfilerConfig
+                all_ranks: false
+                enable: false
+                ranks: []
+                save_path: outputs/profile
+                tool: null
+                tool_config:
+                    npu:
+                        _target_: verl.utils.profiler.config.NPUToolConfig
+                        analysis: true
+                        contents: []
+                        discrete: false
+                        level: level1
+                    nsys:
+                        _target_: verl.utils.profiler.config.NsightToolConfig
+                        discrete: false
+                    torch:
+                        _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                        step_end: null
+                        step_start: 0
+                    torch_memory:
+                        _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                        stack_depth: 32
+                        trace_alloc_max_entries: 100000
+            strategy: fsdp
+            ulysses_sequence_parallel_size: 1
+            use_torch_compile: true
+        rollout:
+            _target_: verl.workers.config.RolloutConfig
+            agent:
+                _target_: verl.workers.config.AgentLoopConfig
+                agent_loop_config_path: null
+                custom_async_server:
+                    _target_: verl.workers.config.CustomAsyncServerConfig
+                    name: null
+                    path: null
+                num_workers: 8
+            calculate_log_probs: false
+            cudagraph_capture_sizes: null
+            data_parallel_size: 1
+            disable_log_stats: true
+            do_sample: true
+            dtype: bfloat16
+            enable_chunked_prefill: true
+            enable_prefix_caching: true
+            enforce_eager: true
+            expert_parallel_size: 1
+            free_cache_engine: true
+            gpu_memory_utilization: 0.7
+            ignore_eos: false
+            layered_summon: false
+            load_format: dummy
+            log_prob_max_token_len_per_gpu: 16384
+            log_prob_micro_batch_size: null
+            log_prob_micro_batch_size_per_gpu: 1
+            log_prob_use_dynamic_bsz: false
+            max_model_len: 7200
+            max_num_batched_tokens: 8192
+            max_num_seqs: 1024
+            mode: sync
+            multi_stage_wake_up: false
+            multi_turn:
+                _target_: verl.workers.config.MultiTurnConfig
+                enable: false
+                format: hermes
+                interaction_config_path: null
+                max_assistant_turns: null
+                max_parallel_calls: 1
+                max_tool_response_length: 256
+                max_user_turns: null
+                num_repeat_rollouts: null
+                tokenization_sanity_check_mode: strict
+                tool_config_path: null
+                tool_response_truncate_side: middle
+                use_inference_chat_template: false
+            "n": 1
+            name: vllm
+            over_sample_rate: 0
+            profiler:
+                _target_: verl.utils.profiler.ProfilerConfig
+                all_ranks: false
+                enable: false
+                ranks: []
+                save_path: outputs/profile
+                tool: null
+                tool_config:
+                    npu:
+                        _target_: verl.utils.profiler.config.NPUToolConfig
+                        analysis: true
+                        contents: []
+                        discrete: false
+                        level: level1
+                    nsys:
+                        _target_: verl.utils.profiler.config.NsightToolConfig
+                        discrete: false
+                    torch:
+                        _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                        step_end: null
+                        step_start: 0
+                    torch_memory:
+                        _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                        stack_depth: 32
+                        trace_alloc_max_entries: 100000
+            prompt_length: 1
+            response_length: 400
+            rollout_filter_metric: reward_variance
+            rollout_filter_ratio: 0.25
+            rollout_filter_type: largest
+            skip_dump_dir: /tmp/rollout_dump
+            skip_rollout: false
+            skip_tokenizer_init: true
+            temperature: 1
+            tensor_model_parallel_size: 1
+            top_k: -1
+            top_p: 1
+            trace:
+                _target_: verl.workers.config.TraceConfig
+                backend: null
+                token2text: false
+            update_weights_bucket_megabytes: 512
+            val_kwargs:
+                _target_: verl.workers.config.SamplingConfig
+                do_sample: true
+                "n": 1
+                temperature: 0.5
+                top_k: -1
+                top_p: 1
+agent_proxy:
+    value:
+        action_sep: '||'
+        enable_think: true
+        max_actions_per_turn: 1
+        max_context_window: -1
+        max_turn: 25
+        reward_normalization:
+            grouping: state
+            method: identity
+        use_turn_scores: false
+algorithm:
+    value:
+        _target_: verl.trainer.config.AlgoConfig
+        adv_estimator: gae
+        bi_level_gae: false
+        gamma: 1
+        high_level_gamma: 0.95
+        kl_ctrl:
+            _target_: verl.trainer.config.KLControlConfig
+            horizon: 10000
+            kl_coef: 0
+            target_kl: 0.1
+            type: fixed
+        kl_penalty: kl
+        lam: 1
+        norm_adv_by_std_in_grpo: true
+        pf_ppo:
+            reweight_method: pow
+            weight_pow: 2
+        use_kl_in_reward: false
+        use_pf_ppo: false
+critic:
+    value:
+        _target_: verl.workers.config.FSDPCriticConfig
+        checkpoint:
+            _target_: verl.trainer.config.CheckpointConfig
+            async_save: false
+            load_contents:
+                - model
+                - optimizer
+                - extra
+            save_contents:
+                - model
+                - optimizer
+                - extra
+        cliprange_value: 0.5
+        enable: null
+        forward_max_token_len_per_gpu: 32768
+        forward_micro_batch_size: null
+        forward_micro_batch_size_per_gpu: 1
+        grad_clip: 1
+        loss_agg_mode: token-mean
+        model:
+            _target_: verl.workers.config.FSDPCriticModelCfg
+            enable_activation_offload: false
+            enable_gradient_checkpointing: true
+            external_lib: null
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                entropy_checkpointing: false
+                entropy_from_logits_with_chunking: false
+                forward_only: false
+                forward_prefetch: false
+                fsdp_size: -1
+                model_dtype: fp32
+                offload_policy: false
+                optimizer_offload: false
+                param_offload: false
+                reshard_after_forward: true
+                strategy: fsdp
+                ulysses_sequence_parallel_size: 1
+                use_orig_params: false
+                use_torch_compile: true
+                wrap_policy:
+                    min_num_params: 0
+            lora_alpha: 16
+            lora_rank: 0
+            path: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            target_modules: all-linear
+            tokenizer_path: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            trust_remote_code: false
+            use_remove_padding: false
+            use_shm: false
+        optim:
+            _target_: verl.workers.config.FSDPOptimizerConfig
+            betas:
+                - 0.9
+                - 0.999
+            clip_grad: 1
+            lr: 1e-05
+            lr_warmup_steps: -1
+            lr_warmup_steps_ratio: 0
+            min_lr_ratio: 0
+            num_cycles: 0.5
+            total_training_steps: 200
+            warmup_style: constant
+            weight_decay: 0.01
+        ppo_epochs: 1
+        ppo_max_token_len_per_gpu: 32768
+        ppo_micro_batch_size: null
+        ppo_micro_batch_size_per_gpu: 1
+        ppo_mini_batch_size: 32
+        profiler:
+            _target_: verl.utils.profiler.ProfilerConfig
+            all_ranks: false
+            enable: false
+            ranks: []
+            save_path: outputs/profile
+            tool: null
+            tool_config:
+                npu:
+                    _target_: verl.utils.profiler.config.NPUToolConfig
+                    analysis: true
+                    contents: []
+                    discrete: false
+                    level: level1
+                nsys:
+                    _target_: verl.utils.profiler.config.NsightToolConfig
+                    discrete: false
+                torch:
+                    _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                    step_end: null
+                    step_start: 0
+                torch_memory:
+                    _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                    stack_depth: 32
+                    trace_alloc_max_entries: 100000
+        rollout_n: 1
+        shuffle: false
+        strategy: fsdp
+        ulysses_sequence_parallel_size: 1
+        use_dynamic_bsz: false
+ctx_manager:
+    value:
+        generation:
+            gen_config:
+                kwargs: null
+                response_length: 400
+                temperature: 1
+                top_k: -1
+                top_p: 1
+custom_envs:
+    value:
+        Bandit:
+            env_config:
+                split: train
+            env_instruction: ""
+            env_type: bandit
+            max_actions_per_traj: 1
+            max_tokens: 100
+        BanditTest:
+            env_config:
+                split: test
+            env_instruction: ""
+            env_type: bandit
+            max_actions_per_traj: 1
+            max_tokens: 100
+        CoordFrozenLake:
+            env_config:
+                observation_format: grid_coord
+            env_instruction: |
+                You are solving the FrozenLake puzzle. The observation includes both a symbol grid and zero-indexed coordinates for the start, goal, player, and any holes.
+                Coordinates range from the top-left corner (0, 0) to the bottom-right corner (5, 5).
+                Beware that the ice is slippery, so the agent might slide and end up in an unintended tile.
+                Respond with a sequence of actions such as <answer>Left || Up || Up</answer>.
+            env_type: frozen_lake
+            max_actions_per_traj: 25
+            max_tokens: 120
+        CoordSokoban:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                max_steps: 100
+                num_boxes: 2
+                observation_format: grid_coord
+            env_instruction: "You are solving the Sokoban puzzle. You are the player and you need to push all boxes to targets.\nYou are provided with a symbol grid and the zero-indexed coordinates of the player, each box, and each target. \nCoordinates range from the top-left corner (0, 0) to the bottom-right corner (5, 5). \nWhen you are exactly next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box.\nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>.\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 120
+        Countdown:
+            env_config: null
+            env_instruction: 'You are solving the Countdown puzzle. You should use the num list to create an equation that equals the target. Example answer format: <think> To find an equation using [3, 5, 2] to get 4. Let''s check 2 + 5 = 7, 7 - 3 = 4. So the answer is 2 + 5 - 3 = 4. </think><answer>2 + 5 - 3</answer>'
+            env_type: countdown
+            max_actions_per_traj: 1
+            max_tokens: 100
+        FrozenLake:
+            env_config: null
+            env_instruction: 'You are solving the FrozenLake puzzle. Forbid the whole and go to the target. You may move to the unintended direction due to the slippery ice. Example answer format: <think>To forbid the hole and go to the target, I should go left then go up.</think><answer>Left || Up</answer>'
+            env_type: frozen_lake
+            max_actions_per_traj: 25
+            max_tokens: 100
+        LargerSokoban:
+            env_config:
+                dim_x: 8
+                dim_y: 8
+                max_steps: 100
+                num_boxes: 2
+                search_depth: 10
+            env_instruction: "You are solving the Sokoban puzzle. \nYou are the player and you need to push all boxes to targets. \nWhen you are right next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box. \nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        Lean:
+            env_config: null
+            env_instruction: You are a Lean theorem prover. Given a Lean theorem statement, propose a sequence of tactics that completes the proof. Think step by step about which tactics to apply next. Provide tactics separated by '||', for example <answer>intro || simp || rfl</answer>.
+            env_type: lean
+            max_actions_per_traj: 30
+            max_tokens: 512
+        MediumSudoku:
+            env_config:
+                difficulty: medium
+                grid_size: 9
+                max_steps: 81
+                render_format: with_feedback
+                show_conflicts: true
+                show_valid_numbers: true
+            env_instruction: |
+                You are solving a Sudoku puzzle. Fill in the grid so that every row, column, and 3x3 box contains the numbers 1-9 without repetition.
+                Initial cells are shown in [brackets] and cannot be modified. Empty cells are shown as dots (.).
+                Place numbers one at a time using the format: <answer>place 5 at row 2 col 3</answer> or <answer>2,3,5</answer>
+                The environment will provide feedback on valid/invalid moves and show conflicts if any occur.
+            env_type: sudoku
+            max_actions_per_traj: 30
+            max_tokens: 150
+            max_workers: 32
+            parallel_friendly: false
+        MetamathQA:
+            env_config: null
+            env_instruction: 'You are solving Math problems. '
+            env_type: metamathqa
+            max_actions_per_traj: 1
+            max_tokens: 100
+        SimpleSokoban:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                max_steps: 100
+                num_boxes: 2
+            env_instruction: "You are solving the Sokoban puzzle. \nYou are the player and you need to push all boxes to targets. \nWhen you are right next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box. \nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        SimpleSudoku:
+            env_config:
+                difficulty: easy
+                grid_size: 4
+                max_steps: 20
+                render_format: with_feedback
+                show_conflicts: false
+                show_valid_numbers: false
+            env_instruction: |
+                You are solving a Sudoku puzzle. Fill in the grid so that every row, column, and 2x2 box contains the numbers 1-4 without repetition.
+                Initial cells are shown in [brackets] and cannot be modified. Empty cells are shown as dots (.).
+                Place numbers one at a time using the format, for example: <answer>place 1 at row 2 col 3</answer> or <answer>1,2,3</answer>
+                The environment will provide feedback on valid/invalid moves and show conflicts if any occur.
+            env_type: sudoku
+            max_actions_per_traj: 20
+            max_tokens: 150
+            max_workers: 32
+            parallel_friendly: false
+        SokobanDifferentGridVocab:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                grid_lookup:
+                    "0": W
+                    "1": .
+                    "2": G
+                    "3": C
+                    "4": B
+                    "5": A
+                    "6": '@'
+                grid_vocab:
+                    .: empty
+                    '@': player on target
+                    A: player
+                    B: box
+                    C: box on target
+                    G: target
+                    W: wall
+                max_steps: 100
+                num_boxes: 1
+                search_depth: 30
+            env_instruction: "You are solving the Sokoban puzzle. \nYou are the player and you need to push all boxes to targets. \nWhen you are right next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box. \nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        VisualSimpleSokoban:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                max_steps: 100
+                num_boxes: 1
+                render_mode: rgb_array
+            env_instruction: You are solving the Sokoban puzzle. You are the player and you need to push all boxes to targets. When you are right next to a box, you can push it by moving in the same direction. You cannot push a box through a wall, and you cannot pull a box. The answer should be a sequence of actions, like <answer>Right || Right || Up</answer>
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        WebShop:
+            env_config:
+                dataset: small
+            env_instruction: 'You are browsing an online shop. Based on the instruction, buy a product that close to the production description. You need to search, read the search results, pick a product, choose the size and color and buy. You should only choose action from the available actions list provided later.  Example process: I need a gingko light and 20x20 pillow cover that is hand painted. First search[gingko light 20x20 pillow cover hand painted], answer format: <answer>search[blanket with fleece throw]</answer>. Valid answer is search[<keywords>] or click[<clickable>].'
+            env_type: webshop
+            max_actions_per_traj: 9
+            max_tokens: 200
+        blackjack:
+            env_config: null
+            env_instruction: |
+                You are playing Blackjack against a dealer. The dealer must hit on 16 or less and stand on 17 or more.
+                Choose either Stick or Hit. Respond with a single action.
+                Example: <answer>Hit</answer>
+            env_type: blackjack
+            max_actions_per_traj: 10
+            max_tokens: 64
+        game_2048:
+            env_config: null
+            env_instruction: |
+                You are playing the 2048 game on a 4x4 grid. Merge equal tiles by sliding Up, Right, Down, or Left.
+                If a move is invalid (no tiles move), a small penalty is applied. Respond with a single action.
+                Example: <answer>Up</answer>
+            env_type: game_2048
+            max_actions_per_traj: 700
+            max_tokens: 8192
+        rubikscube:
+            env_config:
+                max_steps: 20
+                render_mode: text
+                scramble_depth: 1
+            env_instruction: |
+                You are solving a 2x2 Rubik's Cube (Pocket Cube). The goal is to restore the cube so that each of the faces consists of a single, unique color.
+                Available actions use standard Singmaster notation for face rotations: U, U', D, D', L, L', R, R', F, F', B, B'.
+                - Faces: U (Up), D (Down), L (Left), R (Right), F (Front), B (Back).
+                - Modifiers: A letter alone means 90° clockwise (e.g., 'R'). A letter with prime (') means 90° counter-clockwise (e.g., "R'").
+                Respond with a sequence of actions separated by "||".
+                Example: <answer>U</answer>
+            env_type: rubikscube
+            max_actions_per_traj: 20
+            max_tokens: 96
+custom_reward_function:
+    value:
+        name: compute_score
+        path: null
+data:
+    value:
+        custom_cls:
+            name: null
+            path: null
+        datagen:
+            name: null
+            path: null
+        dataloader_num_workers: 8
+        filter_overlong_prompts: false
+        filter_overlong_prompts_workers: 1
+        image_key: images
+        max_prompt_length: null
+        max_response_length: null
+        prompt_key: prompt
+        return_full_prompt: false
+        return_multi_modal_inputs: true
+        return_raw_chat: false
+        return_raw_input_ids: false
+        reward_fn_key: data_source
+        sampler:
+            class_name: null
+            class_path: null
+        shuffle: true
+        tokenizer: null
+        train_batch_size: 128
+        train_files: ~/data/rlhf/gsm8k/train.parquet
+        truncation: error
+        trust_remote_code: false
+        use_shm: false
+        val_batch_size: null
+        val_files: ~/data/rlhf/gsm8k/test.parquet
+        validation_shuffle: false
+        video_key: videos
+enable_response_mask:
+    value: true
+es_manager:
+    value:
+        format_penalty: -0.1
+        train:
+            env_configs:
+                n_groups:
+                    - 8
+                tags:
+                    - CoordFrozenLake
+            env_groups: 8
+            group_size: 16
+        val:
+            env_configs:
+                n_groups:
+                    - 256
+                tags:
+                    - CoordFrozenLake
+            env_groups: 256
+            group_size: 1
+global_profiler:
+    value:
+        _target_: verl.utils.profiler.ProfilerConfig
+        global_tool_config:
+            nsys:
+                _target_: verl.utils.profiler.config.NsightToolConfig
+                controller_nsight_options:
+                    cuda-graph-trace: graph
+                    cuda-memory-usage: "true"
+                    trace: cuda,nvtx,cublas,ucx
+                discrete: false
+                worker_nsight_options:
+                    capture-range: cudaProfilerApi
+                    capture-range-end: null
+                    cuda-graph-trace: graph
+                    cuda-memory-usage: "true"
+                    kill: none
+                    trace: cuda,nvtx,cublas,ucx
+            torch_memory:
+                context: all
+                stack_depth: 32
+                stacks: all
+                trace_alloc_max_entries: 100000
+        profile_continuous_steps: false
+        save_path: outputs/profile
+        steps: null
+        tool: null
+grpo_advantage_length_weight:
+    value: false
+lora:
+    value:
+        alpha: 16
+        rank: 0
+        target_modules: all-linear
+micro_batch_size_per_gpu:
+    value: 1
+model_path:
+    value: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+ppo_mini_batch_size:
+    value: 32
+ray_kwargs:
+    value:
+        ray_init:
+            num_cpus: null
+        timeline_json_file: null
+reward_model:
+    value:
+        enable: false
+        enable_resource_pool: false
+        forward_max_token_len_per_gpu: 32768
+        launch_reward_fn_async: false
+        max_length: null
+        micro_batch_size: null
+        micro_batch_size_per_gpu: null
+        model:
+            external_lib: null
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                forward_prefetch: false
+                fsdp_size: -1
+                param_offload: false
+                reshard_after_forward: true
+                wrap_policy:
+                    min_num_params: 0
+            input_tokenizer: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            path: ~/models/FsfairX-LLaMA3-RM-v0.1
+            trust_remote_code: false
+            use_fused_kernels: false
+            use_remove_padding: false
+            use_shm: false
+        n_gpus_per_node: 0
+        nnodes: 0
+        profiler:
+            _target_: verl.utils.profiler.ProfilerConfig
+            all_ranks: false
+            enable: false
+            ranks: []
+            save_path: outputs/profile
+            tool: null
+            tool_config:
+                npu:
+                    _target_: verl.utils.profiler.config.NPUToolConfig
+                    analysis: true
+                    contents: []
+                    discrete: false
+                    level: level1
+                nsys:
+                    _target_: verl.utils.profiler.config.NsightToolConfig
+                    discrete: false
+                torch:
+                    _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                    step_end: null
+                    step_start: 0
+                torch_memory:
+                    _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                    stack_depth: 32
+                    trace_alloc_max_entries: 100000
+        reward_manager: naive
+        sandbox_fusion:
+            max_concurrent: 64
+            memory_limit_mb: 1024
+            url: null
+        strategy: fsdp
+        ulysses_sequence_parallel_size: 1
+        use_dynamic_bsz: false
+seed:
+    value:
+        train: 10000
+        val: 123
+system:
+    value:
+        CUDA_VISIBLE_DEVICES: 6,7
+trainer:
+    value:
+        balance_batch: true
+        critic_warmup: 0
+        default_hdfs_dir: null
+        default_local_dir: /mnt/general/wanghy/RAGEN/saves/
+        del_local_ckpt_after_load: false
+        device: cuda
+        esi_redundant_time: 0
+        experiment_name: frozen_lake-starpo-slippery
+        generations_to_log_to_wandb:
+            train: 128
+            val: 20
+        local_log_dir: results/
+        log_val_generations: 0
+        logger:
+            - console
+            - wandb
+        max_actor_ckpt_to_keep: 1
+        max_critic_ckpt_to_keep: 1
+        n_gpus_per_node: 2
+        nnodes: 1
+        project_name: ragen_latest_llama_1B_it
+        ray_wait_register_center_timeout: 300
+        resume_from_path: null
+        resume_mode: auto
+        rollout_data_dir: null
+        save_freq: -1
+        test_freq: 10
+        total_epochs: 30
+        total_training_steps: 200
+        use_legacy_worker_impl: auto
+        val_before_train: true
+        val_only: false
+        validation_data_dir: null
+        validation_steps: 1

wandb/run-20251222_184600-j1l9fcd9/files/requirements.txt

@@ -0,0 +1,316 @@
+ragen==0.1
+ragen==0.1
+colorama==0.4.6
+psutil==7.1.2
+ragen==0.1
+setuptools==80.9.0
+wheel==0.45.1
+pip==25.3
+ragen==0.1
+pytz==2025.2
+nvidia-cusparselt-cu12==0.6.2
+mpmath==1.3.0
+typing_extensions==4.15.0
+sympy==1.13.1
+pylatexenc==2.10
+nvidia-cusolver-cu12==11.6.1.9
+numpy==1.26.4
+Markdown==3.10
+absl-py==2.3.1
+pyasn1_modules==0.4.2
+Werkzeug==3.1.3
+networkx==3.3
+MarkupSafe==2.1.5
+fsspec==2025.9.0
+filelock==3.19.1
+nvidia-cudnn-cu12==9.1.0.70
+Jinja2==3.1.6
+tensorboard==2.20.0
+triton==3.2.0
+pillow==11.3.0
+xxhash==3.6.0
+nvidia-nvtx-cu12==12.4.127
+nvidia-nvjitlink-cu12==12.4.127
+nvidia-nccl-cu12==2.21.5
+nvidia-curand-cu12==10.3.5.147
+wrapt==2.0.1
+nvidia-cufft-cu12==11.2.1.3
+pyjnius==1.7.0
+nvidia-cuda-runtime-cu12==12.4.127
+nvidia-cuda-nvrtc-cu12==12.4.127
+websockets==15.0.1
+nvidia-cuda-cupti-cu12==12.4.127
+zipp==3.23.0
+nvidia-cublas-cu12==12.4.5.8
+nvidia-cusparse-cu12==12.3.1.170
+flash_attn==2.8.3
+rsa==4.9.1
+einops==0.8.1
+py-cpuinfo==9.0.0
+pure_eval==0.2.3
+ptyprocess==0.7.0
+gym-notices==0.1.0
+flatbuffers==25.9.23
+fastrlock==0.8.3
+Farama-Notifications==0.0.4
+faiss-gpu==1.7.2
+antlr4-python3-runtime==4.9.3
+wcwidth==0.2.14
+wasabi==1.1.3
+uvloop==0.22.1
+urllib3==2.5.0
+tzdata==2025.2
+typing-inspection==0.4.2
+traitlets==5.14.3
+tqdm==4.67.1
+tomli==2.3.0
+threadpoolctl==3.6.0
+tenacity==9.1.2
+tabulate==0.9.0
+spacy-loggers==1.0.5
+spacy-legacy==3.0.12
+soupsieve==2.8
+sniffio==1.3.1
+smmap==5.0.2
+six==1.17.0
+shellingham==1.5.4
+sentencepiece==0.2.1
+scipy==1.15.3
+safetensors==0.6.2
+rpds-py==0.29.0
+rignore==0.7.6
+regex==2025.11.3
+RapidFuzz==3.14.3
+rank-bm25==0.2.2
+pyzmq==27.1.0
+PyYAML==6.0.3
+pytokens==0.3.0
+python-multipart==0.0.20
+python-json-logger==4.0.0
+python-dotenv==1.2.1
+PySocks==1.7.1
+pyparsing==3.2.5
+PyJWT==2.10.1
+Pygments==2.19.2
+pygame==2.6.1
+pydantic_core==2.41.5
+pycparser==2.23
+pycountry==24.6.1
+pybind11==3.0.1
+pyarrow==22.0.0
+psutil==7.1.3
+protobuf==4.25.8
+propcache==0.4.1
+prometheus_client==0.23.1
+platformdirs==4.5.0
+pexpect==4.9.0
+pathspec==0.12.1
+partial-json-parser==0.2.1.1.post6
+parso==0.8.5
+packaging==25.0
+orjson==3.11.4
+opentelemetry-semantic-conventions-ai==0.4.13
+opencv-python-headless==4.12.0.88
+ninja==1.13.0
+nest-asyncio==1.6.0
+mypy_extensions==1.1.0
+murmurhash==1.0.15
+multidict==6.7.0
+msgspec==0.19.0
+msgpack==1.1.2
+mdurl==0.1.2
+loguru==0.7.3
+llvmlite==0.44.0
+llguidance==0.7.30
+lark==1.2.2
+kiwisolver==1.4.9
+joblib==1.5.2
+jiter==0.12.0
+itsdangerous==2.2.0
+interegular==0.3.3
+ImageIO==2.37.2
+idna==3.11
+humanfriendly==10.0
+httpx-sse==0.4.3
+httptools==0.7.1
+html2text==2025.4.15
+hf-xet==1.2.0
+h11==0.16.0
+grpcio==1.76.0
+frozenlist==1.8.0
+fonttools==4.60.1
+executing==2.2.1
+exceptiongroup==1.3.0
+eval_type_backport==0.2.2
+docstring_parser==0.17.0
+dnspython==2.8.0
+distro==1.9.0
+diskcache==5.6.3
+dill==0.4.0
+decorator==5.2.1
+debugpy==1.8.17
+Cython==3.2.1
+cymem==2.0.13
+cycler==0.12.1
+cupy-cuda12x==13.6.0
+contourpy==1.3.2
+colorama==0.4.6
+codetiming==1.4.0
+cloudpickle==3.1.2
+cloudpathlib==0.23.0
+click==8.3.1
+charset-normalizer==3.4.4
+certifi==2025.11.12
+catalogue==2.0.10
+cachetools==6.2.2
+blis==1.3.2
+blinker==1.9.0
+blake3==1.0.8
+attrs==25.4.0
+async-timeout==5.0.1
+asttokens==3.0.1
+astor==0.8.1
+annotated-types==0.7.0
+annotated-doc==0.0.4
+airportsdata==20250909
+aiohappyeyeballs==2.6.1
+yarl==1.22.0
+uvicorn==0.38.0
+typer-slim==0.20.0
+thefuzz==0.22.1
+stack-data==0.6.3
+srsly==2.5.1
+smart_open==7.5.0
+sentry-sdk==2.44.0
+scikit-learn==1.7.2
+requests==2.32.5
+referencing==0.37.0
+python-dateutil==2.9.0.post0
+pydantic==2.12.4
+prompt_toolkit==3.0.52
+preshed==3.0.11
+opentelemetry-proto==1.26.0
+omegaconf==2.3.0
+numba==0.61.2
+nltk==3.9.2
+multiprocess==0.70.18
+matplotlib-inline==0.2.1
+markdown-it-py==4.0.0
+jedi==0.19.2
+importlib_metadata==8.0.0
+httpcore==1.0.9
+gymnasium==1.2.2
+gym==0.26.2
+googleapis-common-protos==1.72.0
+gitdb==4.0.12
+gguf==0.17.1
+Flask==3.1.2
+faiss-cpu==1.13.0
+email-validator==2.3.0
+depyf==0.18.0
+Deprecated==1.3.1
+coloredlogs==15.0.1
+cffi==2.0.0
+black==25.11.0
+beautifulsoup4==4.14.2
+anyio==4.11.0
+aiosignal==1.4.0
+watchfiles==1.1.1
+tiktoken==0.12.0
+starlette==0.49.3
+sse-starlette==3.0.3
+rich==14.2.0
+pydantic-settings==2.12.0
+pydantic-extra-types==2.10.6
+pandas==2.3.3
+opentelemetry-exporter-otlp-proto-common==1.26.0
+onnxruntime==1.23.2
+matplotlib==3.10.7
+lm-format-enforcer==0.10.12
+jsonschema-specifications==2025.9.1
+ipython==8.37.0
+hydra-core==1.3.2
+huggingface-hub==0.36.0
+httpx==0.28.1
+gym-sokoban==0.0.6
+GitPython==3.1.45
+cryptography==46.0.3
+confection==0.1.5
+cleantext==1.1.4
+aiohttp==3.13.2
+weasel==0.4.3
+wandb==0.23.0
+typer==0.19.2
+virtualenv==20.35.4
+pyasn1==0.6.1
+torch==2.6.0
+tokenizers==0.22.1
+thinc==8.3.9
+rich-toolkit==0.15.1
+prometheus-fastapi-instrumentator==7.1.0
+openai==2.8.0
+jsonschema==4.25.1
+gdown==5.2.0
+fastapi==0.121.2
+anthropic==0.73.0
+xformers==0.0.29.post2
+transformers==4.57.1
+proto-plus==1.26.1
+torchvision==0.21.0
+torchdata==0.11.0
+torchaudio==2.6.0
+together==1.5.30
+tensordict==0.8.3
+spacy==3.8.9
+ray==2.51.1
+outlines_core==0.1.26
+mistral_common==1.8.5
+mcp==1.21.1
+fastapi-cloud-cli==0.3.1
+fastapi-cli==0.0.16
+datasets==4.4.1
+accelerate==1.11.0
+xgrammar==0.1.18
+pyserini==1.2.0
+peft==0.18.0
+outlines==0.1.11
+opentelemetry-exporter-otlp-proto-http==1.26.0
+opentelemetry-exporter-otlp-proto-grpc==1.26.0
+kimina-client==0.2.1
+compressed-tensors==0.9.3
+opentelemetry-exporter-otlp==1.26.0
+vllm==0.8.5
+py-spy==0.4.1
+opencensus-context==0.1.3
+distlib==0.4.0
+colorful==0.5.8
+tensorboard-data-server==0.7.2
+google-auth==2.43.0
+google-api-core==2.28.1
+aiohttp-cors==0.8.1
+opentelemetry-exporter-prometheus==0.59b0
+opencensus==0.11.4
+verl==0.5.0.dev0
+opentelemetry-api==1.38.0
+opentelemetry-semantic-conventions==0.59b0
+opentelemetry-sdk==1.38.0
+typeguard==4.4.4
+shtab==1.8.0
+tyro==0.9.35
+verl==0.5.0.dev0
+autocommand==2.2.2
+backports.tarfile==1.2.0
+importlib_metadata==8.0.0
+inflect==7.3.1
+jaraco.collections==5.1.0
+jaraco.context==5.3.0
+jaraco.functools==4.0.1
+jaraco.text==3.12.1
+more-itertools==10.3.0
+packaging==24.2
+platformdirs==4.2.2
+tomli==2.0.1
+typeguard==4.3.0
+typing_extensions==4.12.2
+wheel==0.45.1
+zipp==3.19.2

wandb/run-20251222_184600-j1l9fcd9/files/wandb-metadata.json

@@ -0,0 +1,107 @@
+{
+  "os": "Linux-5.14.0-284.25.1.el9_2.x86_64-x86_64-with-glibc2.35",
+  "python": "CPython 3.10.19",
+  "startedAt": "2025-12-22T10:46:00.095282Z",
+  "args": [
+    "--node-ip-address=10.119.27.78",
+    "--node-manager-port=42449",
+    "--object-store-name=/tmp/ray/session_2025-12-22_18-44-43_109609_3743349/sockets/plasma_store",
+    "--raylet-name=/tmp/ray/session_2025-12-22_18-44-43_109609_3743349/sockets/raylet",
+    "--redis-address=None",
+    "--metrics-agent-port=58057",
+    "--logging-rotate-bytes=536870912",
+    "--logging-rotate-backup-count=5",
+    "--runtime-env-agent-port=64142",
+    "--gcs-address=10.119.27.78:40383",
+    "--session-name=session_2025-12-22_18-44-43_109609_3743349",
+    "--temp-dir=/tmp/ray",
+    "--webui=127.0.0.1:8265",
+    "--cluster-id=bb05b0e0daa61e2e69d651e5343e9354c9782bb454bfb3b4a69c706a",
+    "--startup-token=112",
+    "--worker-launch-time-ms=1766400294644",
+    "--node-id=7db3272367b4e4a9a8c9a20b5b50b37cf6f00b4f6eae0f10b698a824",
+    "--runtime-env-hash=-839022692"
+  ],
+  "program": "/root/local/miniconda3/envs/ragen/lib/python3.10/site-packages/ray/_private/workers/default_worker.py",
+  "git": {
+    "remote": "https://github.com/mll-lab-nu/RAGEN.git",
+    "commit": "8d73639b99b38265453f898b8d6af7d4af50d56e"
+  },
+  "email": "haoyu-wa22@mails.tsinghua.edu.cn",
+  "root": "/mnt/general/wanghy/RAGEN",
+  "host": "pt-4d654cf4576f4d23ad3d3919f12932fe-worker-0",
+  "executable": "/root/local/miniconda3/envs/ragen/bin/python",
+  "cpu_count": 64,
+  "cpu_count_logical": 128,
+  "gpu": "NVIDIA H100 80GB HBM3",
+  "gpu_count": 8,
+  "disk": {
+    "/": {
+      "total": "5153960755200",
+      "used": "471669112832"
+    }
+  },
+  "memory": {
+    "total": "2163642122240"
+  },
+  "gpu_nvidia": [
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-35e2d43d-4067-82ce-90d4-def9e389bf28"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-af4135e3-88f2-e9ac-518d-502c75a85429"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-d7fdeeba-fe9b-ec03-d9f7-6724fe4266b5"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-ccc4f668-3882-5a8e-2c07-c5cd08f6f666"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-7b73c0cf-d3d5-e10c-7176-a43be1e41001"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-81b58d94-5d1f-8ec2-f9d2-fd56172ed177"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-03e8bc66-3b44-6794-49fd-5392fbdda6d1"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-b7cf0ec6-7c29-1179-dceb-09565da51890"
+    }
+  ],
+  "cudaVersion": "12.4",
+  "writerId": "2t97y3f3xn1qhzjmtyvbbboq8vk1twrh"
+}

wandb/run-20251222_184600-j1l9fcd9/files/wandb-summary.json

@@ -0,0 +1 @@
+{"val-env/CoordFrozenLake/action_is_valid":0.5771020026288547,"val-aux/unknown/reward/worst@256/std":0,"val-aux/unknown/reward/worst@2/std":1.2319831135449713,"val-aux/unknown/reward/best@16/std":0.18152682993805366,"val-aux/unknown/reward/best@16/mean":0.9651000000238419,"val-aux/unknown/reward/worst@128/std":0,"val-aux/unknown/reward/worst@128/mean":-2.5,"val-env/CoordFrozenLake/non-zero/pass@1":1,"val-env/CoordFrozenLake/episodic_return":0,"val-aux/unknown/reward/std@256":1.4174913120536703,"_wandb":{"runtime":201},"val-aux/unknown/reward/worst@256/mean":-2.5,"val-aux/unknown/reward/best@128/mean":1,"val-env/charts/avg_episode_return":0,"val-env/CoordFrozenLake/pass@1":0.18359375,"val-env/CoordFrozenLake/action_is_effective":0.4441965304500113,"val-aux/unknown/reward/best@64/mean":1,"_timestamp":1.766400469857334e+09,"val-aux/unknown/reward/worst@64/std":0,"val-env/CoordFrozenLake/raw_reward":0,"val-env/CoordFrozenLake/non-zero/num_actions":11.21656050955414,"val-aux/unknown/reward/best@32/std":0.044783814092462425,"val-core/unknown/reward/mean@256":-0.8539062514901161,"val-aux/unknown/reward/worst@2/mean":-1.6181000021398066,"val-aux/unknown/reward/worst@16/std":0,"val-aux/unknown/reward/best@2/std":1.1431289716308746,"val-env/response_length":1497.265625,"val-aux/unknown/reward/worst@4/std":0.7894730835754459,"val/generations":{"artifact_path":"wandb-client-artifact://uhw10o0iccwc6ouhgeexfvzwqf83dk5rprcg4ec846drsfr0sgb517rhd1ah3unpjdgp3pnc23jfkxy8qs9yhurhfhtdj41d2mld39stnnu5ym9th0ha8ptwas1jyo6i/val/generations.table.json","_latest_artifact_path":"wandb-client-artifact://czcnoviyqj1x68zfn1cdriz0kpnobj118f4ncs7873jfukpj5pk57bzqnrc3jh3hwnlx3sa6wpwu15du009faftul22najnmaljoyjs27ezqkdthgw1b3yu4qdfdbkn1:latest/val/generations.table.json","size":272160,"log_mode":"IMMUTABLE","_type":"table-file","ncols":61,"path":"media/table/val/generations_0_3ee3d4aebd0def73e996.table.json","nrows":1,"sha256":"3ee3d4aebd0def73e996beedd6e809e890171518abc39efb032d73925217b45a"},"val-env/CoordFrozenLake/num_actions":6.87890625,"val-aux/unknown/reward/best@4/mean":0.5091999997347594,"val-aux/unknown/reward/worst@64/mean":-2.5,"val-env/CoordFrozenLake/non-zero/success":1,"val-aux/unknown/reward/worst@32/std":0,"val-aux/unknown/reward/best@2/mean":-0.029600001491606236,"val-env/CoordFrozenLake/success":0.18359375,"val-aux/unknown/reward/best@32/mean":0.9978999999761582,"val-env/CoordFrozenLake/non-zero/action_is_effective":0.738404622046772,"_step":0,"val-aux/unknown/reward/best@4/std":0.6776100360002214,"_runtime":201.855420511,"val-aux/unknown/reward/worst@4/mean":-2.2065000013932585,"val-aux/unknown/reward/worst@16/mean":-2.5,"val-aux/unknown/reward/worst@8/mean":-2.471100000195205,"val-env/CoordFrozenLake/non-zero/action_is_valid":0.9531491140192697,"val-aux/unknown/reward/worst@8/std":0.25497997913746756,"val-aux/unknown/reward/best@128/std":0,"val-aux/unknown/reward/worst@32/mean":-2.5,"val-aux/unknown/reward/best@64/std":0,"val-aux/unknown/reward/best@8/std":0.39004244638698615,"val-aux/unknown/reward/best@8/mean":0.8116999999135732,"val-core/unknown/reward/best@256/mean":1,"val-core/unknown/reward/best@256/std":0}

wandb/run-20251222_194213-wibn1p50/files/config.yaml

@@ -0,0 +1,962 @@
+_wandb:
+    value:
+        cli_version: 0.23.0
+        e:
+            zwjnig5cm80n0kt195r3qdekxa46t9np:
+                args:
+                    - --node-ip-address=10.119.27.78
+                    - --node-manager-port=37753
+                    - --object-store-name=/tmp/ray/session_2025-12-22_19-40-56_553503_470386/sockets/plasma_store
+                    - --raylet-name=/tmp/ray/session_2025-12-22_19-40-56_553503_470386/sockets/raylet
+                    - --redis-address=None
+                    - --metrics-agent-port=51403
+                    - --logging-rotate-bytes=536870912
+                    - --logging-rotate-backup-count=5
+                    - --runtime-env-agent-port=43669
+                    - --gcs-address=10.119.27.78:65001
+                    - --session-name=session_2025-12-22_19-40-56_553503_470386
+                    - --temp-dir=/tmp/ray
+                    - --webui=127.0.0.1:8265
+                    - --cluster-id=964748dea50d5bc75e51be8ae26c5636d236c0d7cce789a34e7246cb
+                    - --startup-token=112
+                    - --worker-launch-time-ms=1766403665793
+                    - --node-id=5213fd2b78ee903994429a6ae8f18c3b184d8bacf0a0a232caab64a8
+                    - --runtime-env-hash=-839022692
+                cpu_count: 64
+                cpu_count_logical: 128
+                cudaVersion: "12.4"
+                disk:
+                    /:
+                        total: "5153960755200"
+                        used: "471721852928"
+                email: haoyu-wa22@mails.tsinghua.edu.cn
+                executable: /root/local/miniconda3/envs/ragen/bin/python
+                git:
+                    commit: 8d73639b99b38265453f898b8d6af7d4af50d56e
+                    remote: https://github.com/mll-lab-nu/RAGEN.git
+                gpu: NVIDIA H100 80GB HBM3
+                gpu_count: 8
+                gpu_nvidia:
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-35e2d43d-4067-82ce-90d4-def9e389bf28
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-af4135e3-88f2-e9ac-518d-502c75a85429
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-d7fdeeba-fe9b-ec03-d9f7-6724fe4266b5
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-ccc4f668-3882-5a8e-2c07-c5cd08f6f666
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-7b73c0cf-d3d5-e10c-7176-a43be1e41001
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-81b58d94-5d1f-8ec2-f9d2-fd56172ed177
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-03e8bc66-3b44-6794-49fd-5392fbdda6d1
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-b7cf0ec6-7c29-1179-dceb-09565da51890
+                host: pt-4d654cf4576f4d23ad3d3919f12932fe-worker-0
+                memory:
+                    total: "2163642122240"
+                os: Linux-5.14.0-284.25.1.el9_2.x86_64-x86_64-with-glibc2.35
+                program: /root/local/miniconda3/envs/ragen/lib/python3.10/site-packages/ray/_private/workers/default_worker.py
+                python: CPython 3.10.19
+                root: /mnt/general/wanghy/RAGEN
+                startedAt: "2025-12-22T11:42:13.576643Z"
+                writerId: zwjnig5cm80n0kt195r3qdekxa46t9np
+        m: []
+        python_version: 3.10.19
+        t:
+            "1":
+                - 1
+                - 5
+                - 11
+                - 30
+                - 41
+                - 49
+                - 50
+                - 51
+                - 53
+                - 71
+                - 95
+                - 97
+                - 98
+                - 105
+            "2":
+                - 1
+                - 5
+                - 11
+                - 30
+                - 41
+                - 49
+                - 50
+                - 51
+                - 53
+                - 71
+                - 95
+                - 97
+                - 98
+                - 105
+            "3":
+                - 2
+                - 13
+                - 16
+                - 61
+            "4": 3.10.19
+            "5": 0.23.0
+            "6": 4.57.1
+            "12": 0.23.0
+            "13": linux-x86_64
+actor_rollout_ref:
+    value:
+        actor:
+            _target_: verl.workers.config.FSDPActorConfig
+            checkpoint:
+                _target_: verl.trainer.config.CheckpointConfig
+                async_save: false
+                load_contents:
+                    - model
+                    - optimizer
+                    - extra
+                save_contents:
+                    - model
+                    - optimizer
+                    - extra
+            clip_ratio: 0.2
+            clip_ratio_c: 3
+            clip_ratio_high: 0.28
+            clip_ratio_low: 0.2
+            entropy_checkpointing: false
+            entropy_coeff: 0.001
+            entropy_from_logits_with_chunking: false
+            freeze_vision_tower: false
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                entropy_checkpointing: false
+                entropy_from_logits_with_chunking: false
+                forward_only: false
+                forward_prefetch: false
+                fsdp_size: -1
+                model_dtype: fp32
+                offload_policy: false
+                optimizer_offload: false
+                param_offload: false
+                reshard_after_forward: true
+                strategy: fsdp
+                ulysses_sequence_parallel_size: 1
+                use_orig_params: false
+                use_torch_compile: true
+                wrap_policy:
+                    min_num_params: 0
+            grad_clip: 1
+            grpo_advantage_length_weight: false
+            kl_loss_coef: 0
+            kl_loss_type: kl
+            loss_agg_mode: token-mean
+            optim:
+                _target_: verl.workers.config.FSDPOptimizerConfig
+                betas:
+                    - 0.9
+                    - 0.999
+                clip_grad: 1
+                lr: 1e-06
+                lr_warmup_steps: -1
+                lr_warmup_steps_ratio: 0
+                min_lr_ratio: 0
+                num_cycles: 0.5
+                total_training_steps: 200
+                warmup_style: constant
+                weight_decay: 0.01
+            policy_loss:
+                _target_: verl.workers.config.PolicyLossConfig
+                clip_cov_lb: 1
+                clip_cov_ratio: 0.0002
+                clip_cov_ub: 5
+                kl_cov_ratio: 0.0002
+                loss_mode: vanilla
+                ppo_kl_coef: 0.1
+            ppo_epochs: 1
+            ppo_max_token_len_per_gpu: 16384
+            ppo_micro_batch_size: null
+            ppo_micro_batch_size_per_gpu: 1
+            ppo_mini_batch_size: 32
+            profiler:
+                _target_: verl.utils.profiler.ProfilerConfig
+                all_ranks: false
+                enable: false
+                ranks: []
+                save_path: outputs/profile
+                tool: null
+                tool_config:
+                    npu:
+                        _target_: verl.utils.profiler.config.NPUToolConfig
+                        analysis: true
+                        contents: []
+                        discrete: false
+                        level: level1
+                    nsys:
+                        _target_: verl.utils.profiler.config.NsightToolConfig
+                        discrete: false
+                    torch:
+                        _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                        step_end: null
+                        step_start: 0
+                    torch_memory:
+                        _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                        stack_depth: 32
+                        trace_alloc_max_entries: 100000
+            shuffle: false
+            strategy: fsdp
+            tis_imp_ratio_cap: -1
+            ulysses_sequence_parallel_size: 1
+            use_dynamic_bsz: false
+            use_fused_kernels: false
+            use_kl_loss: false
+            use_ref: true
+            use_remove_padding: false
+            use_torch_compile: true
+        hybrid_engine: true
+        model:
+            _target_: verl.workers.config.HFModelConfig
+            custom_chat_template: null
+            enable_activation_offload: false
+            enable_gradient_checkpointing: true
+            exclude_modules: null
+            external_lib: null
+            fused_kernel_options:
+                impl_backend: torch
+            hf_config_path: null
+            lora_alpha: 16
+            lora_rank: 0
+            path: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            target_modules: all-linear
+            tokenizer_path: null
+            torch_dtype: bfloat16
+            trust_remote_code: false
+            use_fused_kernels: false
+            use_liger: false
+            use_remove_padding: false
+            use_shm: false
+        nccl_timeout: 600
+        ref:
+            entropy_checkpointing: false
+            entropy_from_logits_with_chunking: false
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                entropy_checkpointing: false
+                entropy_from_logits_with_chunking: false
+                forward_only: false
+                forward_prefetch: false
+                fsdp_size: -1
+                model_dtype: fp32
+                offload_policy: false
+                optimizer_offload: false
+                param_offload: false
+                reshard_after_forward: true
+                strategy: fsdp
+                ulysses_sequence_parallel_size: 1
+                use_orig_params: false
+                use_torch_compile: true
+                wrap_policy:
+                    min_num_params: 0
+            log_prob_max_token_len_per_gpu: 16384
+            log_prob_micro_batch_size: null
+            log_prob_micro_batch_size_per_gpu: 1
+            log_prob_use_dynamic_bsz: false
+            model: null
+            profiler:
+                _target_: verl.utils.profiler.ProfilerConfig
+                all_ranks: false
+                enable: false
+                ranks: []
+                save_path: outputs/profile
+                tool: null
+                tool_config:
+                    npu:
+                        _target_: verl.utils.profiler.config.NPUToolConfig
+                        analysis: true
+                        contents: []
+                        discrete: false
+                        level: level1
+                    nsys:
+                        _target_: verl.utils.profiler.config.NsightToolConfig
+                        discrete: false
+                    torch:
+                        _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                        step_end: null
+                        step_start: 0
+                    torch_memory:
+                        _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                        stack_depth: 32
+                        trace_alloc_max_entries: 100000
+            strategy: fsdp
+            ulysses_sequence_parallel_size: 1
+            use_torch_compile: true
+        rollout:
+            _target_: verl.workers.config.RolloutConfig
+            agent:
+                _target_: verl.workers.config.AgentLoopConfig
+                agent_loop_config_path: null
+                custom_async_server:
+                    _target_: verl.workers.config.CustomAsyncServerConfig
+                    name: null
+                    path: null
+                num_workers: 8
+            calculate_log_probs: false
+            cudagraph_capture_sizes: null
+            data_parallel_size: 1
+            disable_log_stats: true
+            do_sample: true
+            dtype: bfloat16
+            enable_chunked_prefill: true
+            enable_prefix_caching: true
+            enforce_eager: true
+            expert_parallel_size: 1
+            free_cache_engine: true
+            gpu_memory_utilization: 0.7
+            ignore_eos: false
+            layered_summon: false
+            load_format: dummy
+            log_prob_max_token_len_per_gpu: 16384
+            log_prob_micro_batch_size: null
+            log_prob_micro_batch_size_per_gpu: 1
+            log_prob_use_dynamic_bsz: false
+            max_model_len: 7200
+            max_num_batched_tokens: 8192
+            max_num_seqs: 1024
+            mode: sync
+            multi_stage_wake_up: false
+            multi_turn:
+                _target_: verl.workers.config.MultiTurnConfig
+                enable: false
+                format: hermes
+                interaction_config_path: null
+                max_assistant_turns: null
+                max_parallel_calls: 1
+                max_tool_response_length: 256
+                max_user_turns: null
+                num_repeat_rollouts: null
+                tokenization_sanity_check_mode: strict
+                tool_config_path: null
+                tool_response_truncate_side: middle
+                use_inference_chat_template: false
+            "n": 1
+            name: vllm
+            over_sample_rate: 0
+            profiler:
+                _target_: verl.utils.profiler.ProfilerConfig
+                all_ranks: false
+                enable: false
+                ranks: []
+                save_path: outputs/profile
+                tool: null
+                tool_config:
+                    npu:
+                        _target_: verl.utils.profiler.config.NPUToolConfig
+                        analysis: true
+                        contents: []
+                        discrete: false
+                        level: level1
+                    nsys:
+                        _target_: verl.utils.profiler.config.NsightToolConfig
+                        discrete: false
+                    torch:
+                        _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                        step_end: null
+                        step_start: 0
+                    torch_memory:
+                        _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                        stack_depth: 32
+                        trace_alloc_max_entries: 100000
+            prompt_length: 1
+            response_length: 400
+            rollout_filter_metric: reward_variance
+            rollout_filter_ratio: 0.25
+            rollout_filter_type: largest
+            skip_dump_dir: /tmp/rollout_dump
+            skip_rollout: false
+            skip_tokenizer_init: true
+            temperature: 1
+            tensor_model_parallel_size: 1
+            top_k: -1
+            top_p: 1
+            trace:
+                _target_: verl.workers.config.TraceConfig
+                backend: null
+                token2text: false
+            update_weights_bucket_megabytes: 512
+            val_kwargs:
+                _target_: verl.workers.config.SamplingConfig
+                do_sample: true
+                "n": 1
+                temperature: 0.5
+                top_k: -1
+                top_p: 1
+agent_proxy:
+    value:
+        action_sep: '||'
+        enable_think: true
+        max_actions_per_turn: 1
+        max_context_window: -1
+        max_turn: 25
+        reward_normalization:
+            grouping: state
+            method: identity
+        use_turn_scores: false
+algorithm:
+    value:
+        _target_: verl.trainer.config.AlgoConfig
+        adv_estimator: gae
+        bi_level_gae: false
+        gamma: 1
+        high_level_gamma: 0.95
+        kl_ctrl:
+            _target_: verl.trainer.config.KLControlConfig
+            horizon: 10000
+            kl_coef: 0
+            target_kl: 0.1
+            type: fixed
+        kl_penalty: kl
+        lam: 1
+        norm_adv_by_std_in_grpo: true
+        pf_ppo:
+            reweight_method: pow
+            weight_pow: 2
+        use_kl_in_reward: false
+        use_pf_ppo: false
+critic:
+    value:
+        _target_: verl.workers.config.FSDPCriticConfig
+        checkpoint:
+            _target_: verl.trainer.config.CheckpointConfig
+            async_save: false
+            load_contents:
+                - model
+                - optimizer
+                - extra
+            save_contents:
+                - model
+                - optimizer
+                - extra
+        cliprange_value: 0.5
+        enable: null
+        forward_max_token_len_per_gpu: 32768
+        forward_micro_batch_size: null
+        forward_micro_batch_size_per_gpu: 1
+        grad_clip: 1
+        loss_agg_mode: token-mean
+        model:
+            _target_: verl.workers.config.FSDPCriticModelCfg
+            enable_activation_offload: false
+            enable_gradient_checkpointing: true
+            external_lib: null
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                entropy_checkpointing: false
+                entropy_from_logits_with_chunking: false
+                forward_only: false
+                forward_prefetch: false
+                fsdp_size: -1
+                model_dtype: fp32
+                offload_policy: false
+                optimizer_offload: false
+                param_offload: false
+                reshard_after_forward: true
+                strategy: fsdp
+                ulysses_sequence_parallel_size: 1
+                use_orig_params: false
+                use_torch_compile: true
+                wrap_policy:
+                    min_num_params: 0
+            lora_alpha: 16
+            lora_rank: 0
+            path: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            target_modules: all-linear
+            tokenizer_path: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            trust_remote_code: false
+            use_remove_padding: false
+            use_shm: false
+        optim:
+            _target_: verl.workers.config.FSDPOptimizerConfig
+            betas:
+                - 0.9
+                - 0.999
+            clip_grad: 1
+            lr: 1e-05
+            lr_warmup_steps: -1
+            lr_warmup_steps_ratio: 0
+            min_lr_ratio: 0
+            num_cycles: 0.5
+            total_training_steps: 200
+            warmup_style: constant
+            weight_decay: 0.01
+        ppo_epochs: 1
+        ppo_max_token_len_per_gpu: 32768
+        ppo_micro_batch_size: null
+        ppo_micro_batch_size_per_gpu: 1
+        ppo_mini_batch_size: 32
+        profiler:
+            _target_: verl.utils.profiler.ProfilerConfig
+            all_ranks: false
+            enable: false
+            ranks: []
+            save_path: outputs/profile
+            tool: null
+            tool_config:
+                npu:
+                    _target_: verl.utils.profiler.config.NPUToolConfig
+                    analysis: true
+                    contents: []
+                    discrete: false
+                    level: level1
+                nsys:
+                    _target_: verl.utils.profiler.config.NsightToolConfig
+                    discrete: false
+                torch:
+                    _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                    step_end: null
+                    step_start: 0
+                torch_memory:
+                    _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                    stack_depth: 32
+                    trace_alloc_max_entries: 100000
+        rollout_n: 1
+        shuffle: false
+        strategy: fsdp
+        ulysses_sequence_parallel_size: 1
+        use_dynamic_bsz: false
+ctx_manager:
+    value:
+        generation:
+            gen_config:
+                kwargs: null
+                response_length: 400
+                temperature: 1
+                top_k: -1
+                top_p: 1
+custom_envs:
+    value:
+        Bandit:
+            env_config:
+                split: train
+            env_instruction: ""
+            env_type: bandit
+            max_actions_per_traj: 1
+            max_tokens: 100
+        BanditTest:
+            env_config:
+                split: test
+            env_instruction: ""
+            env_type: bandit
+            max_actions_per_traj: 1
+            max_tokens: 100
+        CoordFrozenLake:
+            env_config:
+                observation_format: grid_coord
+            env_instruction: |
+                You are solving the FrozenLake puzzle. The observation includes both a symbol grid and zero-indexed coordinates for the start, goal, player, and any holes.
+                Coordinates range from the top-left corner (0, 0) to the bottom-right corner (5, 5).
+                Beware that the ice is slippery, so the agent might slide and end up in an unintended tile.
+                Respond with a sequence of actions such as <answer>Left || Up || Up</answer>.
+            env_type: frozen_lake
+            max_actions_per_traj: 25
+            max_tokens: 120
+        CoordSokoban:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                max_steps: 100
+                num_boxes: 2
+                observation_format: grid_coord
+            env_instruction: "You are solving the Sokoban puzzle. You are the player and you need to push all boxes to targets.\nYou are provided with a symbol grid and the zero-indexed coordinates of the player, each box, and each target. \nCoordinates range from the top-left corner (0, 0) to the bottom-right corner (5, 5). \nWhen you are exactly next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box.\nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>.\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 120
+        Countdown:
+            env_config: null
+            env_instruction: 'You are solving the Countdown puzzle. You should use the num list to create an equation that equals the target. Example answer format: <think> To find an equation using [3, 5, 2] to get 4. Let''s check 2 + 5 = 7, 7 - 3 = 4. So the answer is 2 + 5 - 3 = 4. </think><answer>2 + 5 - 3</answer>'
+            env_type: countdown
+            max_actions_per_traj: 1
+            max_tokens: 100
+        FrozenLake:
+            env_config: null
+            env_instruction: 'You are solving the FrozenLake puzzle. Forbid the whole and go to the target. You may move to the unintended direction due to the slippery ice. Example answer format: <think>To forbid the hole and go to the target, I should go left then go up.</think><answer>Left || Up</answer>'
+            env_type: frozen_lake
+            max_actions_per_traj: 25
+            max_tokens: 100
+        LargerSokoban:
+            env_config:
+                dim_x: 8
+                dim_y: 8
+                max_steps: 100
+                num_boxes: 2
+                search_depth: 10
+            env_instruction: "You are solving the Sokoban puzzle. \nYou are the player and you need to push all boxes to targets. \nWhen you are right next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box. \nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        Lean:
+            env_config: null
+            env_instruction: You are a Lean theorem prover. Given a Lean theorem statement, propose a sequence of tactics that completes the proof. Think step by step about which tactics to apply next. Provide tactics separated by '||', for example <answer>intro || simp || rfl</answer>.
+            env_type: lean
+            max_actions_per_traj: 30
+            max_tokens: 512
+        MediumSudoku:
+            env_config:
+                difficulty: medium
+                grid_size: 9
+                max_steps: 81
+                render_format: with_feedback
+                show_conflicts: true
+                show_valid_numbers: true
+            env_instruction: |
+                You are solving a Sudoku puzzle. Fill in the grid so that every row, column, and 3x3 box contains the numbers 1-9 without repetition.
+                Initial cells are shown in [brackets] and cannot be modified. Empty cells are shown as dots (.).
+                Place numbers one at a time using the format: <answer>place 5 at row 2 col 3</answer> or <answer>2,3,5</answer>
+                The environment will provide feedback on valid/invalid moves and show conflicts if any occur.
+            env_type: sudoku
+            max_actions_per_traj: 30
+            max_tokens: 150
+            max_workers: 32
+            parallel_friendly: false
+        MetamathQA:
+            env_config: null
+            env_instruction: 'You are solving Math problems. '
+            env_type: metamathqa
+            max_actions_per_traj: 1
+            max_tokens: 100
+        SimpleSokoban:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                max_steps: 100
+                num_boxes: 2
+            env_instruction: "You are solving the Sokoban puzzle. \nYou are the player and you need to push all boxes to targets. \nWhen you are right next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box. \nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        SimpleSudoku:
+            env_config:
+                difficulty: easy
+                grid_size: 4
+                max_steps: 20
+                render_format: with_feedback
+                show_conflicts: false
+                show_valid_numbers: false
+            env_instruction: |
+                You are solving a Sudoku puzzle. Fill in the grid so that every row, column, and 2x2 box contains the numbers 1-4 without repetition.
+                Initial cells are shown in [brackets] and cannot be modified. Empty cells are shown as dots (.).
+                Place numbers one at a time using the format, for example: <answer>place 1 at row 2 col 3</answer> or <answer>1,2,3</answer>
+                The environment will provide feedback on valid/invalid moves and show conflicts if any occur.
+            env_type: sudoku
+            max_actions_per_traj: 20
+            max_tokens: 150
+            max_workers: 32
+            parallel_friendly: false
+        SokobanDifferentGridVocab:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                grid_lookup:
+                    "0": W
+                    "1": .
+                    "2": G
+                    "3": C
+                    "4": B
+                    "5": A
+                    "6": '@'
+                grid_vocab:
+                    .: empty
+                    '@': player on target
+                    A: player
+                    B: box
+                    C: box on target
+                    G: target
+                    W: wall
+                max_steps: 100
+                num_boxes: 1
+                search_depth: 30
+            env_instruction: "You are solving the Sokoban puzzle. \nYou are the player and you need to push all boxes to targets. \nWhen you are right next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box. \nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        VisualSimpleSokoban:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                max_steps: 100
+                num_boxes: 1
+                render_mode: rgb_array
+            env_instruction: You are solving the Sokoban puzzle. You are the player and you need to push all boxes to targets. When you are right next to a box, you can push it by moving in the same direction. You cannot push a box through a wall, and you cannot pull a box. The answer should be a sequence of actions, like <answer>Right || Right || Up</answer>
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        WebShop:
+            env_config:
+                dataset: small
+            env_instruction: 'You are browsing an online shop. Based on the instruction, buy a product that close to the production description. You need to search, read the search results, pick a product, choose the size and color and buy. You should only choose action from the available actions list provided later.  Example process: I need a gingko light and 20x20 pillow cover that is hand painted. First search[gingko light 20x20 pillow cover hand painted], answer format: <answer>search[blanket with fleece throw]</answer>. Valid answer is search[<keywords>] or click[<clickable>].'
+            env_type: webshop
+            max_actions_per_traj: 9
+            max_tokens: 200
+        blackjack:
+            env_config: null
+            env_instruction: |
+                You are playing Blackjack against a dealer. The dealer must hit on 16 or less and stand on 17 or more.
+                Choose either Stick or Hit. Respond with a single action.
+                Example: <answer>Hit</answer>
+            env_type: blackjack
+            max_actions_per_traj: 10
+            max_tokens: 64
+        game_2048:
+            env_config: null
+            env_instruction: |
+                You are playing the 2048 game on a 4x4 grid. Merge equal tiles by sliding Up, Right, Down, or Left.
+                If a move is invalid (no tiles move), a small penalty is applied. Respond with a single action.
+                Example: <answer>Up</answer>
+            env_type: game_2048
+            max_actions_per_traj: 700
+            max_tokens: 8192
+        rubikscube:
+            env_config:
+                max_steps: 20
+                render_mode: text
+                scramble_depth: 1
+            env_instruction: |
+                You are solving a 2x2 Rubik's Cube (Pocket Cube). The goal is to restore the cube so that each of the faces consists of a single, unique color.
+                Available actions use standard Singmaster notation for face rotations: U, U', D, D', L, L', R, R', F, F', B, B'.
+                - Faces: U (Up), D (Down), L (Left), R (Right), F (Front), B (Back).
+                - Modifiers: A letter alone means 90° clockwise (e.g., 'R'). A letter with prime (') means 90° counter-clockwise (e.g., "R'").
+                Respond with a sequence of actions separated by "||".
+                Example: <answer>U</answer>
+            env_type: rubikscube
+            max_actions_per_traj: 20
+            max_tokens: 96
+custom_reward_function:
+    value:
+        name: compute_score
+        path: null
+data:
+    value:
+        custom_cls:
+            name: null
+            path: null
+        datagen:
+            name: null
+            path: null
+        dataloader_num_workers: 8
+        filter_overlong_prompts: false
+        filter_overlong_prompts_workers: 1
+        image_key: images
+        max_prompt_length: null
+        max_response_length: null
+        prompt_key: prompt
+        return_full_prompt: false
+        return_multi_modal_inputs: true
+        return_raw_chat: false
+        return_raw_input_ids: false
+        reward_fn_key: data_source
+        sampler:
+            class_name: null
+            class_path: null
+        shuffle: true
+        tokenizer: null
+        train_batch_size: 128
+        train_files: ~/data/rlhf/gsm8k/train.parquet
+        truncation: error
+        trust_remote_code: false
+        use_shm: false
+        val_batch_size: null
+        val_files: ~/data/rlhf/gsm8k/test.parquet
+        validation_shuffle: false
+        video_key: videos
+enable_response_mask:
+    value: true
+es_manager:
+    value:
+        format_penalty: -0.1
+        train:
+            env_configs:
+                n_groups:
+                    - 8
+                tags:
+                    - CoordFrozenLake
+            env_groups: 8
+            group_size: 16
+        val:
+            env_configs:
+                n_groups:
+                    - 256
+                tags:
+                    - CoordFrozenLake
+            env_groups: 256
+            group_size: 1
+global_profiler:
+    value:
+        _target_: verl.utils.profiler.ProfilerConfig
+        global_tool_config:
+            nsys:
+                _target_: verl.utils.profiler.config.NsightToolConfig
+                controller_nsight_options:
+                    cuda-graph-trace: graph
+                    cuda-memory-usage: "true"
+                    trace: cuda,nvtx,cublas,ucx
+                discrete: false
+                worker_nsight_options:
+                    capture-range: cudaProfilerApi
+                    capture-range-end: null
+                    cuda-graph-trace: graph
+                    cuda-memory-usage: "true"
+                    kill: none
+                    trace: cuda,nvtx,cublas,ucx
+            torch_memory:
+                context: all
+                stack_depth: 32
+                stacks: all
+                trace_alloc_max_entries: 100000
+        profile_continuous_steps: false
+        save_path: outputs/profile
+        steps: null
+        tool: null
+grpo_advantage_length_weight:
+    value: false
+lora:
+    value:
+        alpha: 16
+        rank: 0
+        target_modules: all-linear
+micro_batch_size_per_gpu:
+    value: 1
+model_path:
+    value: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+ppo_mini_batch_size:
+    value: 32
+ray_kwargs:
+    value:
+        ray_init:
+            num_cpus: null
+        timeline_json_file: null
+reward_model:
+    value:
+        enable: false
+        enable_resource_pool: false
+        forward_max_token_len_per_gpu: 32768
+        launch_reward_fn_async: false
+        max_length: null
+        micro_batch_size: null
+        micro_batch_size_per_gpu: null
+        model:
+            external_lib: null
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                forward_prefetch: false
+                fsdp_size: -1
+                param_offload: false
+                reshard_after_forward: true
+                wrap_policy:
+                    min_num_params: 0
+            input_tokenizer: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            path: ~/models/FsfairX-LLaMA3-RM-v0.1
+            trust_remote_code: false
+            use_fused_kernels: false
+            use_remove_padding: false
+            use_shm: false
+        n_gpus_per_node: 0
+        nnodes: 0
+        profiler:
+            _target_: verl.utils.profiler.ProfilerConfig
+            all_ranks: false
+            enable: false
+            ranks: []
+            save_path: outputs/profile
+            tool: null
+            tool_config:
+                npu:
+                    _target_: verl.utils.profiler.config.NPUToolConfig
+                    analysis: true
+                    contents: []
+                    discrete: false
+                    level: level1
+                nsys:
+                    _target_: verl.utils.profiler.config.NsightToolConfig
+                    discrete: false
+                torch:
+                    _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                    step_end: null
+                    step_start: 0
+                torch_memory:
+                    _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                    stack_depth: 32
+                    trace_alloc_max_entries: 100000
+        reward_manager: naive
+        sandbox_fusion:
+            max_concurrent: 64
+            memory_limit_mb: 1024
+            url: null
+        strategy: fsdp
+        ulysses_sequence_parallel_size: 1
+        use_dynamic_bsz: false
+seed:
+    value:
+        train: 10000
+        val: 123
+system:
+    value:
+        CUDA_VISIBLE_DEVICES: 0,1,2,3,4,5,6,7
+trainer:
+    value:
+        balance_batch: true
+        critic_warmup: 0
+        default_hdfs_dir: null
+        default_local_dir: /mnt/general/wanghy/RAGEN/saves/
+        del_local_ckpt_after_load: false
+        device: cuda
+        esi_redundant_time: 0
+        experiment_name: frozen_lake-starpo-slippery
+        generations_to_log_to_wandb:
+            train: 128
+            val: 20
+        local_log_dir: results/
+        log_val_generations: 0
+        logger:
+            - console
+            - wandb
+        max_actor_ckpt_to_keep: 1
+        max_critic_ckpt_to_keep: 1
+        n_gpus_per_node: 8
+        nnodes: 1
+        project_name: ragen_latest_llama_1B_it
+        ray_wait_register_center_timeout: 300
+        resume_from_path: null
+        resume_mode: auto
+        rollout_data_dir: null
+        save_freq: -1
+        test_freq: 10
+        total_epochs: 30
+        total_training_steps: 200
+        use_legacy_worker_impl: auto
+        val_before_train: true
+        val_only: false
+        validation_data_dir: null
+        validation_steps: 1

wandb/run-20251222_194213-wibn1p50/files/requirements.txt

@@ -0,0 +1,316 @@
+ragen==0.1
+ragen==0.1
+colorama==0.4.6
+psutil==7.1.2
+ragen==0.1
+setuptools==80.9.0
+wheel==0.45.1
+pip==25.3
+ragen==0.1
+pytz==2025.2
+nvidia-cusparselt-cu12==0.6.2
+mpmath==1.3.0
+typing_extensions==4.15.0
+sympy==1.13.1
+pylatexenc==2.10
+nvidia-cusolver-cu12==11.6.1.9
+numpy==1.26.4
+Markdown==3.10
+absl-py==2.3.1
+pyasn1_modules==0.4.2
+Werkzeug==3.1.3
+networkx==3.3
+MarkupSafe==2.1.5
+fsspec==2025.9.0
+filelock==3.19.1
+nvidia-cudnn-cu12==9.1.0.70
+Jinja2==3.1.6
+tensorboard==2.20.0
+triton==3.2.0
+pillow==11.3.0
+xxhash==3.6.0
+nvidia-nvtx-cu12==12.4.127
+nvidia-nvjitlink-cu12==12.4.127
+nvidia-nccl-cu12==2.21.5
+nvidia-curand-cu12==10.3.5.147
+wrapt==2.0.1
+nvidia-cufft-cu12==11.2.1.3
+pyjnius==1.7.0
+nvidia-cuda-runtime-cu12==12.4.127
+nvidia-cuda-nvrtc-cu12==12.4.127
+websockets==15.0.1
+nvidia-cuda-cupti-cu12==12.4.127
+zipp==3.23.0
+nvidia-cublas-cu12==12.4.5.8
+nvidia-cusparse-cu12==12.3.1.170
+flash_attn==2.8.3
+rsa==4.9.1
+einops==0.8.1
+py-cpuinfo==9.0.0
+pure_eval==0.2.3
+ptyprocess==0.7.0
+gym-notices==0.1.0
+flatbuffers==25.9.23
+fastrlock==0.8.3
+Farama-Notifications==0.0.4
+faiss-gpu==1.7.2
+antlr4-python3-runtime==4.9.3
+wcwidth==0.2.14
+wasabi==1.1.3
+uvloop==0.22.1
+urllib3==2.5.0
+tzdata==2025.2
+typing-inspection==0.4.2
+traitlets==5.14.3
+tqdm==4.67.1
+tomli==2.3.0
+threadpoolctl==3.6.0
+tenacity==9.1.2
+tabulate==0.9.0
+spacy-loggers==1.0.5
+spacy-legacy==3.0.12
+soupsieve==2.8
+sniffio==1.3.1
+smmap==5.0.2
+six==1.17.0
+shellingham==1.5.4
+sentencepiece==0.2.1
+scipy==1.15.3
+safetensors==0.6.2
+rpds-py==0.29.0
+rignore==0.7.6
+regex==2025.11.3
+RapidFuzz==3.14.3
+rank-bm25==0.2.2
+pyzmq==27.1.0
+PyYAML==6.0.3
+pytokens==0.3.0
+python-multipart==0.0.20
+python-json-logger==4.0.0
+python-dotenv==1.2.1
+PySocks==1.7.1
+pyparsing==3.2.5
+PyJWT==2.10.1
+Pygments==2.19.2
+pygame==2.6.1
+pydantic_core==2.41.5
+pycparser==2.23
+pycountry==24.6.1
+pybind11==3.0.1
+pyarrow==22.0.0
+psutil==7.1.3
+protobuf==4.25.8
+propcache==0.4.1
+prometheus_client==0.23.1
+platformdirs==4.5.0
+pexpect==4.9.0
+pathspec==0.12.1
+partial-json-parser==0.2.1.1.post6
+parso==0.8.5
+packaging==25.0
+orjson==3.11.4
+opentelemetry-semantic-conventions-ai==0.4.13
+opencv-python-headless==4.12.0.88
+ninja==1.13.0
+nest-asyncio==1.6.0
+mypy_extensions==1.1.0
+murmurhash==1.0.15
+multidict==6.7.0
+msgspec==0.19.0
+msgpack==1.1.2
+mdurl==0.1.2
+loguru==0.7.3
+llvmlite==0.44.0
+llguidance==0.7.30
+lark==1.2.2
+kiwisolver==1.4.9
+joblib==1.5.2
+jiter==0.12.0
+itsdangerous==2.2.0
+interegular==0.3.3
+ImageIO==2.37.2
+idna==3.11
+humanfriendly==10.0
+httpx-sse==0.4.3
+httptools==0.7.1
+html2text==2025.4.15
+hf-xet==1.2.0
+h11==0.16.0
+grpcio==1.76.0
+frozenlist==1.8.0
+fonttools==4.60.1
+executing==2.2.1
+exceptiongroup==1.3.0
+eval_type_backport==0.2.2
+docstring_parser==0.17.0
+dnspython==2.8.0
+distro==1.9.0
+diskcache==5.6.3
+dill==0.4.0
+decorator==5.2.1
+debugpy==1.8.17
+Cython==3.2.1
+cymem==2.0.13
+cycler==0.12.1
+cupy-cuda12x==13.6.0
+contourpy==1.3.2
+colorama==0.4.6
+codetiming==1.4.0
+cloudpickle==3.1.2
+cloudpathlib==0.23.0
+click==8.3.1
+charset-normalizer==3.4.4
+certifi==2025.11.12
+catalogue==2.0.10
+cachetools==6.2.2
+blis==1.3.2
+blinker==1.9.0
+blake3==1.0.8
+attrs==25.4.0
+async-timeout==5.0.1
+asttokens==3.0.1
+astor==0.8.1
+annotated-types==0.7.0
+annotated-doc==0.0.4
+airportsdata==20250909
+aiohappyeyeballs==2.6.1
+yarl==1.22.0
+uvicorn==0.38.0
+typer-slim==0.20.0
+thefuzz==0.22.1
+stack-data==0.6.3
+srsly==2.5.1
+smart_open==7.5.0
+sentry-sdk==2.44.0
+scikit-learn==1.7.2
+requests==2.32.5
+referencing==0.37.0
+python-dateutil==2.9.0.post0
+pydantic==2.12.4
+prompt_toolkit==3.0.52
+preshed==3.0.11
+opentelemetry-proto==1.26.0
+omegaconf==2.3.0
+numba==0.61.2
+nltk==3.9.2
+multiprocess==0.70.18
+matplotlib-inline==0.2.1
+markdown-it-py==4.0.0
+jedi==0.19.2
+importlib_metadata==8.0.0
+httpcore==1.0.9
+gymnasium==1.2.2
+gym==0.26.2
+googleapis-common-protos==1.72.0
+gitdb==4.0.12
+gguf==0.17.1
+Flask==3.1.2
+faiss-cpu==1.13.0
+email-validator==2.3.0
+depyf==0.18.0
+Deprecated==1.3.1
+coloredlogs==15.0.1
+cffi==2.0.0
+black==25.11.0
+beautifulsoup4==4.14.2
+anyio==4.11.0
+aiosignal==1.4.0
+watchfiles==1.1.1
+tiktoken==0.12.0
+starlette==0.49.3
+sse-starlette==3.0.3
+rich==14.2.0
+pydantic-settings==2.12.0
+pydantic-extra-types==2.10.6
+pandas==2.3.3
+opentelemetry-exporter-otlp-proto-common==1.26.0
+onnxruntime==1.23.2
+matplotlib==3.10.7
+lm-format-enforcer==0.10.12
+jsonschema-specifications==2025.9.1
+ipython==8.37.0
+hydra-core==1.3.2
+huggingface-hub==0.36.0
+httpx==0.28.1
+gym-sokoban==0.0.6
+GitPython==3.1.45
+cryptography==46.0.3
+confection==0.1.5
+cleantext==1.1.4
+aiohttp==3.13.2
+weasel==0.4.3
+wandb==0.23.0
+typer==0.19.2
+virtualenv==20.35.4
+pyasn1==0.6.1
+torch==2.6.0
+tokenizers==0.22.1
+thinc==8.3.9
+rich-toolkit==0.15.1
+prometheus-fastapi-instrumentator==7.1.0
+openai==2.8.0
+jsonschema==4.25.1
+gdown==5.2.0
+fastapi==0.121.2
+anthropic==0.73.0
+xformers==0.0.29.post2
+transformers==4.57.1
+proto-plus==1.26.1
+torchvision==0.21.0
+torchdata==0.11.0
+torchaudio==2.6.0
+together==1.5.30
+tensordict==0.8.3
+spacy==3.8.9
+ray==2.51.1
+outlines_core==0.1.26
+mistral_common==1.8.5
+mcp==1.21.1
+fastapi-cloud-cli==0.3.1
+fastapi-cli==0.0.16
+datasets==4.4.1
+accelerate==1.11.0
+xgrammar==0.1.18
+pyserini==1.2.0
+peft==0.18.0
+outlines==0.1.11
+opentelemetry-exporter-otlp-proto-http==1.26.0
+opentelemetry-exporter-otlp-proto-grpc==1.26.0
+kimina-client==0.2.1
+compressed-tensors==0.9.3
+opentelemetry-exporter-otlp==1.26.0
+vllm==0.8.5
+py-spy==0.4.1
+opencensus-context==0.1.3
+distlib==0.4.0
+colorful==0.5.8
+tensorboard-data-server==0.7.2
+google-auth==2.43.0
+google-api-core==2.28.1
+aiohttp-cors==0.8.1
+opentelemetry-exporter-prometheus==0.59b0
+opencensus==0.11.4
+verl==0.5.0.dev0
+opentelemetry-api==1.38.0
+opentelemetry-semantic-conventions==0.59b0
+opentelemetry-sdk==1.38.0
+typeguard==4.4.4
+shtab==1.8.0
+tyro==0.9.35
+verl==0.5.0.dev0
+autocommand==2.2.2
+backports.tarfile==1.2.0
+importlib_metadata==8.0.0
+inflect==7.3.1
+jaraco.collections==5.1.0
+jaraco.context==5.3.0
+jaraco.functools==4.0.1
+jaraco.text==3.12.1
+more-itertools==10.3.0
+packaging==24.2
+platformdirs==4.2.2
+tomli==2.0.1
+typeguard==4.3.0
+typing_extensions==4.12.2
+wheel==0.45.1
+zipp==3.19.2

wandb/run-20251222_194213-wibn1p50/files/wandb-metadata.json

@@ -0,0 +1,107 @@
+{
+  "os": "Linux-5.14.0-284.25.1.el9_2.x86_64-x86_64-with-glibc2.35",
+  "python": "CPython 3.10.19",
+  "startedAt": "2025-12-22T11:42:13.576643Z",
+  "args": [
+    "--node-ip-address=10.119.27.78",
+    "--node-manager-port=37753",
+    "--object-store-name=/tmp/ray/session_2025-12-22_19-40-56_553503_470386/sockets/plasma_store",
+    "--raylet-name=/tmp/ray/session_2025-12-22_19-40-56_553503_470386/sockets/raylet",
+    "--redis-address=None",
+    "--metrics-agent-port=51403",
+    "--logging-rotate-bytes=536870912",
+    "--logging-rotate-backup-count=5",
+    "--runtime-env-agent-port=43669",
+    "--gcs-address=10.119.27.78:65001",
+    "--session-name=session_2025-12-22_19-40-56_553503_470386",
+    "--temp-dir=/tmp/ray",
+    "--webui=127.0.0.1:8265",
+    "--cluster-id=964748dea50d5bc75e51be8ae26c5636d236c0d7cce789a34e7246cb",
+    "--startup-token=112",
+    "--worker-launch-time-ms=1766403665793",
+    "--node-id=5213fd2b78ee903994429a6ae8f18c3b184d8bacf0a0a232caab64a8",
+    "--runtime-env-hash=-839022692"
+  ],
+  "program": "/root/local/miniconda3/envs/ragen/lib/python3.10/site-packages/ray/_private/workers/default_worker.py",
+  "git": {
+    "remote": "https://github.com/mll-lab-nu/RAGEN.git",
+    "commit": "8d73639b99b38265453f898b8d6af7d4af50d56e"
+  },
+  "email": "haoyu-wa22@mails.tsinghua.edu.cn",
+  "root": "/mnt/general/wanghy/RAGEN",
+  "host": "pt-4d654cf4576f4d23ad3d3919f12932fe-worker-0",
+  "executable": "/root/local/miniconda3/envs/ragen/bin/python",
+  "cpu_count": 64,
+  "cpu_count_logical": 128,
+  "gpu": "NVIDIA H100 80GB HBM3",
+  "gpu_count": 8,
+  "disk": {
+    "/": {
+      "total": "5153960755200",
+      "used": "471721852928"
+    }
+  },
+  "memory": {
+    "total": "2163642122240"
+  },
+  "gpu_nvidia": [
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-35e2d43d-4067-82ce-90d4-def9e389bf28"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-af4135e3-88f2-e9ac-518d-502c75a85429"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-d7fdeeba-fe9b-ec03-d9f7-6724fe4266b5"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-ccc4f668-3882-5a8e-2c07-c5cd08f6f666"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-7b73c0cf-d3d5-e10c-7176-a43be1e41001"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-81b58d94-5d1f-8ec2-f9d2-fd56172ed177"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-03e8bc66-3b44-6794-49fd-5392fbdda6d1"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-b7cf0ec6-7c29-1179-dceb-09565da51890"
+    }
+  ],
+  "cudaVersion": "12.4",
+  "writerId": "zwjnig5cm80n0kt195r3qdekxa46t9np"
+}

wandb/run-20251222_194213-wibn1p50/files/wandb-summary.json

@@ -0,0 +1 @@
+{"val-aux/unknown/reward/best@128/mean":1,"rollout/in_group_reward_mean":-1.150781273841858,"_step":1,"timing_per_token_ms/update_critic":0.00868963071298127,"val-env/CoordFrozenLake/raw_reward":0,"critic/values/max":6.1875,"response_length/mean":4105.09375,"val-env/CoordFrozenLake/non-zero/pass@1":1,"prompt_length/mean":0.1875,"val-aux/unknown/reward/best@16/std":0.17352521449027572,"val-core/unknown/reward/mean@256":-0.8128906251804437,"val-aux/unknown/reward/best@2/std":1.1340807676744216,"timing_s/old_log_prob":1.258065678179264,"timing_s/update_actor":1.6892315223813057,"val-env/CoordFrozenLake/success":0.18359375,"critic/rewards/min":-2.5,"critic/score/max":1,"train/CoordFrozenLake/non-zero/action_is_valid":0.7212219967052853,"response_length/clip_ratio":0.1875,"rollout/ref_log_prob":-8.952258110046387,"train/response_length":2451.859375,"critic/vpred_mean":0.3396296985447407,"critic/vf_explained_var":-1.5507891178131104,"prompt_length/min":0,"val-aux/unknown/reward/best@32/std":0.031914886956126036,"val-aux/unknown/reward/worst@128/mean":-2.5,"val-aux/unknown/reward/worst@64/std":0,"response/aborted_ratio":0,"val/generations":{"sha256":"48ef2484dcca7a8afd4a1bbfb43d9975db4cae66783fa0443288adfca6fffb92","log_mode":"IMMUTABLE","artifact_path":"wandb-client-artifact://766ohtlw8i9y56c77x66uod67l3qqoj6tej4rnkz774ste0p25lh5uaq1y8d269259kgj0z9vajp43d48yvfczt0s66nzysqqm4g19yzrt06kfvd8p1q525686w5pre0/val/generations.table.json","nrows":1,"path":"media/table/val/generations_0_48ef2484dcca7a8afd4a.table.json","_latest_artifact_path":"wandb-client-artifact://hdrpe676aqegq88o0o9bxsknqq1frob87fs8vlnhmb4go3o7giwmqv9yfr0j7vxqd3mc0hqlhtxsgvxdji53re63ehqmz59woml8kr71qty9xtfoymedanvia2dwrz6o:latest/val/generations.table.json","ncols":61,"size":221910,"_type":"table-file"},"rollout/chosen_in_group_reward_max":1,"actor/lr":1e-06,"timing_s/total":78.10700941085815,"train/CoordFrozenLake/pass@16":1,"response_length/max":7200,"response_length_non_aborted/max":7200,"val-aux/unknown/reward/worst@256/mean":-2.5,"rollout/chosen_in_group_reward_mean":-1.0343749523162842,"val-aux/unknown/reward/best@4/mean":0.4936999987959862,"rollout/chosen_in_group_std":1.6747982501983643,"val-aux/unknown/reward/best@16/mean":0.9669999995231628,"train/CoordFrozenLake/episodic_return":0,"val-aux/unknown/reward/best@64/std":0,"train/CoordFrozenLake/action_is_valid":0.4000528262974629,"train/CoordFrozenLake/raw_reward":0,"train/CoordFrozenLake/non-zero/num_actions":6.858974358974359,"global_seqlen/minmax_diff":14836,"val-env/charts/avg_episode_return":0,"perf/max_memory_allocated_gb":73.36607027053833,"val-env/CoordFrozenLake/non-zero/action_is_valid":0.9749324455790079,"critic/lr":1e-05,"val-aux/unknown/reward/worst@4/mean":-2.1617999996989967,"val-aux/unknown/reward/worst@4/std":0.8463809778113507,"critic/advantages/max":6.445878028869629,"rollout/chosen_in_group_mean":-1.0343749523162842,"critic/values/mean":-0.08795215934515,"val-aux/unknown/reward/worst@8/std":0.3541489938383426,"timing_s/values":0.2973985821008682,"critic/rewards/max":1,"val-aux/unknown/reward/best@8/mean":0.8145999987125396,"timing_s/update_critic":1.1415480971336365,"critic/vf_clipfrac":0,"val-aux/unknown/reward/best@2/mean":-0.07410000060498714,"perf/time_per_step":78.10407362133265,"critic/score/mean":-1.0343749523162842,"train/CoordFrozenLake/action_is_effective":0.3324561600461736,"timing_s/gen":72.23815268278122,"val-aux/unknown/reward/worst@128/std":0,"timing_per_token_ms/adv":0.002395370991513186,"response_length_non_aborted/min":340,"timing_s/ref":1.1617145091295242,"val-core/unknown/reward/best@256/std":0,"timing_s/step":78.10407362133265,"val-aux/unknown/reward/worst@64/mean":-2.5,"val-aux/unknown/reward/best@4/std":0.6590829303247507,"actor/ppo_kl":0,"perf/mfu/actor":0.11348438595994421,"val-aux/unknown/reward/worst@16/std":0.07585670697344167,"perf/max_memory_reserved_gb":81.806640625,"train/CoordFrozenLake/non-zero/pass@16":1,"rollout/in_group_max":1,"perf/cpu_memory_used_gb":103.3122787475586,"critic/advantages/min":-6.103915214538574,"train/charts/avg_episode_return":0,"actor/entropy":8.921223640441895,"critic/grad_norm":173.9873046875,"_timestamp":1.7664038917806118e+09,"global_seqlen/mean":16421.125,"val-env/CoordFrozenLake/episodic_return":0,"timing_per_token_ms/values":0.0022638414093193084,"response_length/min":340,"timing_per_token_ms/ref":0.008843140384181384,"global_seqlen/balanced_min":15936,"val-aux/unknown/reward/best@32/mean":0.9987999999523163,"_wandb":{"runtime":193},"val-aux/unknown/reward/worst@2/std":1.254008037631705,"perf/total_num_tokens":131369,"critic/rewards/mean":-1.0343749523162842,"perf/mfu/critic":0.1691245609411814,"val-aux/unknown/reward/best@128/std":0,"train/CoordFrozenLake/success":0.28125,"actor/pg_clipfrac_lower":0,"train/CoordFrozenLake/num_actions":4.1796875,"rollout/in_group_reward_std":1.464717984199524,"actor/pg_clipfrac":0,"val-env/CoordFrozenLake/num_actions":6.78515625,"prompt_length/clip_ratio":0.1875,"perf/throughput":210.24671619067615,"val-env/CoordFrozenLake/action_is_valid":0.597907788890251,"timing_per_token_ms/update_actor":0.012858676874919544,"val-env/CoordFrozenLake/action_is_effective":0.4822152752149087,"critic/returns/min":-2.5,"critic/returns/mean":-2.2920210361480713,"critic/vf_loss":0.7555297240614891,"val-core/unknown/reward/best@256/mean":1,"_runtime":193.246442713,"critic/returns/max":1,"prompt_length/max":1,"train/CoordFrozenLake/non-zero/action_is_effective":0.6167302679117422,"val-env/CoordFrozenLake/non-zero/success":1,"global_seqlen/max":22258,"critic/values/min":-5.1875,"rollout/chosen_in_group_reward_std":1.6747982501983643,"val-env/response_length":1341.08984375,"critic/score/min":-2.5,"response_length_non_aborted/clip_ratio":0.1875,"val-aux/unknown/reward/worst@32/std":0,"val-aux/unknown/reward/worst@2/mean":-1.5127999997437,"val-env/CoordFrozenLake/pass@1":0.18359375,"global_seqlen/balanced_max":17225,"response_length_non_aborted/mean":4105.09375,"timing_per_token_ms/gen":0.5499124767459728,"val-aux/unknown/reward/best@64/mean":1,"rollout/chosen_in_group_max":1,"critic/advantages/mean":-1.1889634521367043e-07,"rollout/in_group_mean":-1.150781273841858,"rollout/in_group_std":1.464717984199524,"actor/pg_loss":-0.09265413228422403,"global_seqlen/min":7422,"train/CoordFrozenLake/non-zero/success":1,"val-env/CoordFrozenLake/non-zero/action_is_effective":0.7913276311219014,"rollout/in_group_reward_max":1,"actor/grad_norm":36.663455963134766,"val-env/CoordFrozenLake/non-zero/num_actions":10.99367088607595,"val-aux/unknown/reward/best@8/std":0.3823177209862261,"val-aux/unknown/reward/worst@8/mean":-2.445700000055134,"val-aux/unknown/reward/worst@32/mean":-2.5,"val-aux/unknown/reward/worst@256/std":0,"timing_s/adv":0.31467749178409576,"val-aux/unknown/reward/worst@16/mean":-2.49760000000149,"val-aux/unknown/reward/std@256":1.4064409039854844}

wandb/run-20251222_194213-wibn1p50/run-wibn1p50.wandb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a44ddc818459776ed1d43ddb58135c20f6361b801d0ca3c2c64f630164cd7678
+size 101274

wandb/run-20251222_202905-69ev02ff/files/config.yaml

@@ -0,0 +1,962 @@
+_wandb:
+    value:
+        cli_version: 0.23.0
+        e:
+            z84gczlr8c1c496ea0tnn6zrntod8u7k:
+                args:
+                    - --node-ip-address=10.119.27.78
+                    - --node-manager-port=38739
+                    - --object-store-name=/tmp/ray/session_2025-12-22_20-27-49_028907_487744/sockets/plasma_store
+                    - --raylet-name=/tmp/ray/session_2025-12-22_20-27-49_028907_487744/sockets/raylet
+                    - --redis-address=None
+                    - --metrics-agent-port=45094
+                    - --logging-rotate-bytes=536870912
+                    - --logging-rotate-backup-count=5
+                    - --runtime-env-agent-port=54281
+                    - --gcs-address=10.119.27.78:36231
+                    - --session-name=session_2025-12-22_20-27-49_028907_487744
+                    - --temp-dir=/tmp/ray
+                    - --webui=127.0.0.1:8265
+                    - --cluster-id=cf93e6bfeed4aa4ef758a9bb28b14e8f7aeb59402129513d9cc31269
+                    - --startup-token=112
+                    - --worker-launch-time-ms=1766406478308
+                    - --node-id=f950bfc2cfb0e6e294c1e34ef1cf9e9eac2db5dc349184b20fe78f90
+                    - --runtime-env-hash=-839022692
+                cpu_count: 64
+                cpu_count_logical: 128
+                cudaVersion: "12.4"
+                disk:
+                    /:
+                        total: "5153960755200"
+                        used: "471729799168"
+                email: haoyu-wa22@mails.tsinghua.edu.cn
+                executable: /root/local/miniconda3/envs/ragen/bin/python
+                git:
+                    commit: 8d73639b99b38265453f898b8d6af7d4af50d56e
+                    remote: https://github.com/mll-lab-nu/RAGEN.git
+                gpu: NVIDIA H100 80GB HBM3
+                gpu_count: 8
+                gpu_nvidia:
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-35e2d43d-4067-82ce-90d4-def9e389bf28
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-af4135e3-88f2-e9ac-518d-502c75a85429
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-d7fdeeba-fe9b-ec03-d9f7-6724fe4266b5
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-ccc4f668-3882-5a8e-2c07-c5cd08f6f666
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-7b73c0cf-d3d5-e10c-7176-a43be1e41001
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-81b58d94-5d1f-8ec2-f9d2-fd56172ed177
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-03e8bc66-3b44-6794-49fd-5392fbdda6d1
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-b7cf0ec6-7c29-1179-dceb-09565da51890
+                host: pt-4d654cf4576f4d23ad3d3919f12932fe-worker-0
+                memory:
+                    total: "2163642122240"
+                os: Linux-5.14.0-284.25.1.el9_2.x86_64-x86_64-with-glibc2.35
+                program: /root/local/miniconda3/envs/ragen/lib/python3.10/site-packages/ray/_private/workers/default_worker.py
+                python: CPython 3.10.19
+                root: /mnt/general/wanghy/RAGEN
+                startedAt: "2025-12-22T12:29:05.056794Z"
+                writerId: z84gczlr8c1c496ea0tnn6zrntod8u7k
+        m: []
+        python_version: 3.10.19
+        t:
+            "1":
+                - 1
+                - 5
+                - 11
+                - 30
+                - 41
+                - 49
+                - 50
+                - 51
+                - 53
+                - 71
+                - 95
+                - 97
+                - 98
+                - 105
+            "2":
+                - 1
+                - 5
+                - 11
+                - 30
+                - 41
+                - 49
+                - 50
+                - 51
+                - 53
+                - 71
+                - 95
+                - 97
+                - 98
+                - 105
+            "3":
+                - 2
+                - 13
+                - 16
+                - 61
+            "4": 3.10.19
+            "5": 0.23.0
+            "6": 4.57.1
+            "12": 0.23.0
+            "13": linux-x86_64
+actor_rollout_ref:
+    value:
+        actor:
+            _target_: verl.workers.config.FSDPActorConfig
+            checkpoint:
+                _target_: verl.trainer.config.CheckpointConfig
+                async_save: false
+                load_contents:
+                    - model
+                    - optimizer
+                    - extra
+                save_contents:
+                    - model
+                    - optimizer
+                    - extra
+            clip_ratio: 0.2
+            clip_ratio_c: 3
+            clip_ratio_high: 0.28
+            clip_ratio_low: 0.2
+            entropy_checkpointing: false
+            entropy_coeff: 0.001
+            entropy_from_logits_with_chunking: false
+            freeze_vision_tower: false
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                entropy_checkpointing: false
+                entropy_from_logits_with_chunking: false
+                forward_only: false
+                forward_prefetch: false
+                fsdp_size: -1
+                model_dtype: fp32
+                offload_policy: false
+                optimizer_offload: false
+                param_offload: false
+                reshard_after_forward: true
+                strategy: fsdp
+                ulysses_sequence_parallel_size: 1
+                use_orig_params: false
+                use_torch_compile: true
+                wrap_policy:
+                    min_num_params: 0
+            grad_clip: 1
+            grpo_advantage_length_weight: false
+            kl_loss_coef: 0
+            kl_loss_type: kl
+            loss_agg_mode: token-mean
+            optim:
+                _target_: verl.workers.config.FSDPOptimizerConfig
+                betas:
+                    - 0.9
+                    - 0.999
+                clip_grad: 1
+                lr: 1e-06
+                lr_warmup_steps: -1
+                lr_warmup_steps_ratio: 0
+                min_lr_ratio: 0
+                num_cycles: 0.5
+                total_training_steps: 200
+                warmup_style: constant
+                weight_decay: 0.01
+            policy_loss:
+                _target_: verl.workers.config.PolicyLossConfig
+                clip_cov_lb: 1
+                clip_cov_ratio: 0.0002
+                clip_cov_ub: 5
+                kl_cov_ratio: 0.0002
+                loss_mode: vanilla
+                ppo_kl_coef: 0.1
+            ppo_epochs: 1
+            ppo_max_token_len_per_gpu: 16384
+            ppo_micro_batch_size: null
+            ppo_micro_batch_size_per_gpu: 1
+            ppo_mini_batch_size: 32
+            profiler:
+                _target_: verl.utils.profiler.ProfilerConfig
+                all_ranks: false
+                enable: false
+                ranks: []
+                save_path: outputs/profile
+                tool: null
+                tool_config:
+                    npu:
+                        _target_: verl.utils.profiler.config.NPUToolConfig
+                        analysis: true
+                        contents: []
+                        discrete: false
+                        level: level1
+                    nsys:
+                        _target_: verl.utils.profiler.config.NsightToolConfig
+                        discrete: false
+                    torch:
+                        _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                        step_end: null
+                        step_start: 0
+                    torch_memory:
+                        _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                        stack_depth: 32
+                        trace_alloc_max_entries: 100000
+            shuffle: false
+            strategy: fsdp
+            tis_imp_ratio_cap: -1
+            ulysses_sequence_parallel_size: 1
+            use_dynamic_bsz: false
+            use_fused_kernels: false
+            use_kl_loss: false
+            use_ref: true
+            use_remove_padding: false
+            use_torch_compile: true
+        hybrid_engine: true
+        model:
+            _target_: verl.workers.config.HFModelConfig
+            custom_chat_template: null
+            enable_activation_offload: false
+            enable_gradient_checkpointing: true
+            exclude_modules: null
+            external_lib: null
+            fused_kernel_options:
+                impl_backend: torch
+            hf_config_path: null
+            lora_alpha: 16
+            lora_rank: 0
+            path: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            target_modules: all-linear
+            tokenizer_path: null
+            torch_dtype: bfloat16
+            trust_remote_code: false
+            use_fused_kernels: false
+            use_liger: false
+            use_remove_padding: false
+            use_shm: false
+        nccl_timeout: 600
+        ref:
+            entropy_checkpointing: false
+            entropy_from_logits_with_chunking: false
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                entropy_checkpointing: false
+                entropy_from_logits_with_chunking: false
+                forward_only: false
+                forward_prefetch: false
+                fsdp_size: -1
+                model_dtype: fp32
+                offload_policy: false
+                optimizer_offload: false
+                param_offload: false
+                reshard_after_forward: true
+                strategy: fsdp
+                ulysses_sequence_parallel_size: 1
+                use_orig_params: false
+                use_torch_compile: true
+                wrap_policy:
+                    min_num_params: 0
+            log_prob_max_token_len_per_gpu: 16384
+            log_prob_micro_batch_size: null
+            log_prob_micro_batch_size_per_gpu: 1
+            log_prob_use_dynamic_bsz: false
+            model: null
+            profiler:
+                _target_: verl.utils.profiler.ProfilerConfig
+                all_ranks: false
+                enable: false
+                ranks: []
+                save_path: outputs/profile
+                tool: null
+                tool_config:
+                    npu:
+                        _target_: verl.utils.profiler.config.NPUToolConfig
+                        analysis: true
+                        contents: []
+                        discrete: false
+                        level: level1
+                    nsys:
+                        _target_: verl.utils.profiler.config.NsightToolConfig
+                        discrete: false
+                    torch:
+                        _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                        step_end: null
+                        step_start: 0
+                    torch_memory:
+                        _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                        stack_depth: 32
+                        trace_alloc_max_entries: 100000
+            strategy: fsdp
+            ulysses_sequence_parallel_size: 1
+            use_torch_compile: true
+        rollout:
+            _target_: verl.workers.config.RolloutConfig
+            agent:
+                _target_: verl.workers.config.AgentLoopConfig
+                agent_loop_config_path: null
+                custom_async_server:
+                    _target_: verl.workers.config.CustomAsyncServerConfig
+                    name: null
+                    path: null
+                num_workers: 8
+            calculate_log_probs: false
+            cudagraph_capture_sizes: null
+            data_parallel_size: 1
+            disable_log_stats: true
+            do_sample: true
+            dtype: bfloat16
+            enable_chunked_prefill: true
+            enable_prefix_caching: true
+            enforce_eager: true
+            expert_parallel_size: 1
+            free_cache_engine: true
+            gpu_memory_utilization: 0.7
+            ignore_eos: false
+            layered_summon: false
+            load_format: dummy
+            log_prob_max_token_len_per_gpu: 16384
+            log_prob_micro_batch_size: null
+            log_prob_micro_batch_size_per_gpu: 1
+            log_prob_use_dynamic_bsz: false
+            max_model_len: 7200
+            max_num_batched_tokens: 8192
+            max_num_seqs: 1024
+            mode: sync
+            multi_stage_wake_up: false
+            multi_turn:
+                _target_: verl.workers.config.MultiTurnConfig
+                enable: false
+                format: hermes
+                interaction_config_path: null
+                max_assistant_turns: null
+                max_parallel_calls: 1
+                max_tool_response_length: 256
+                max_user_turns: null
+                num_repeat_rollouts: null
+                tokenization_sanity_check_mode: strict
+                tool_config_path: null
+                tool_response_truncate_side: middle
+                use_inference_chat_template: false
+            "n": 1
+            name: vllm
+            over_sample_rate: 0
+            profiler:
+                _target_: verl.utils.profiler.ProfilerConfig
+                all_ranks: false
+                enable: false
+                ranks: []
+                save_path: outputs/profile
+                tool: null
+                tool_config:
+                    npu:
+                        _target_: verl.utils.profiler.config.NPUToolConfig
+                        analysis: true
+                        contents: []
+                        discrete: false
+                        level: level1
+                    nsys:
+                        _target_: verl.utils.profiler.config.NsightToolConfig
+                        discrete: false
+                    torch:
+                        _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                        step_end: null
+                        step_start: 0
+                    torch_memory:
+                        _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                        stack_depth: 32
+                        trace_alloc_max_entries: 100000
+            prompt_length: 1
+            response_length: 400
+            rollout_filter_metric: reward_variance
+            rollout_filter_ratio: 0.25
+            rollout_filter_type: largest
+            skip_dump_dir: /tmp/rollout_dump
+            skip_rollout: false
+            skip_tokenizer_init: true
+            temperature: 1
+            tensor_model_parallel_size: 1
+            top_k: -1
+            top_p: 1
+            trace:
+                _target_: verl.workers.config.TraceConfig
+                backend: null
+                token2text: false
+            update_weights_bucket_megabytes: 512
+            val_kwargs:
+                _target_: verl.workers.config.SamplingConfig
+                do_sample: true
+                "n": 1
+                temperature: 0.5
+                top_k: -1
+                top_p: 1
+agent_proxy:
+    value:
+        action_sep: '||'
+        enable_think: true
+        max_actions_per_turn: 1
+        max_context_window: -1
+        max_turn: 25
+        reward_normalization:
+            grouping: state
+            method: identity
+        use_turn_scores: false
+algorithm:
+    value:
+        _target_: verl.trainer.config.AlgoConfig
+        adv_estimator: gae
+        bi_level_gae: false
+        gamma: 1
+        high_level_gamma: 0.95
+        kl_ctrl:
+            _target_: verl.trainer.config.KLControlConfig
+            horizon: 10000
+            kl_coef: 0
+            target_kl: 0.1
+            type: fixed
+        kl_penalty: kl
+        lam: 1
+        norm_adv_by_std_in_grpo: true
+        pf_ppo:
+            reweight_method: pow
+            weight_pow: 2
+        use_kl_in_reward: false
+        use_pf_ppo: false
+critic:
+    value:
+        _target_: verl.workers.config.FSDPCriticConfig
+        checkpoint:
+            _target_: verl.trainer.config.CheckpointConfig
+            async_save: false
+            load_contents:
+                - model
+                - optimizer
+                - extra
+            save_contents:
+                - model
+                - optimizer
+                - extra
+        cliprange_value: 0.5
+        enable: null
+        forward_max_token_len_per_gpu: 32768
+        forward_micro_batch_size: null
+        forward_micro_batch_size_per_gpu: 1
+        grad_clip: 1
+        loss_agg_mode: token-mean
+        model:
+            _target_: verl.workers.config.FSDPCriticModelCfg
+            enable_activation_offload: false
+            enable_gradient_checkpointing: true
+            external_lib: null
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                entropy_checkpointing: false
+                entropy_from_logits_with_chunking: false
+                forward_only: false
+                forward_prefetch: false
+                fsdp_size: -1
+                model_dtype: fp32
+                offload_policy: false
+                optimizer_offload: false
+                param_offload: false
+                reshard_after_forward: true
+                strategy: fsdp
+                ulysses_sequence_parallel_size: 1
+                use_orig_params: false
+                use_torch_compile: true
+                wrap_policy:
+                    min_num_params: 0
+            lora_alpha: 16
+            lora_rank: 0
+            path: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            target_modules: all-linear
+            tokenizer_path: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            trust_remote_code: false
+            use_remove_padding: false
+            use_shm: false
+        optim:
+            _target_: verl.workers.config.FSDPOptimizerConfig
+            betas:
+                - 0.9
+                - 0.999
+            clip_grad: 1
+            lr: 1e-05
+            lr_warmup_steps: -1
+            lr_warmup_steps_ratio: 0
+            min_lr_ratio: 0
+            num_cycles: 0.5
+            total_training_steps: 200
+            warmup_style: constant
+            weight_decay: 0.01
+        ppo_epochs: 1
+        ppo_max_token_len_per_gpu: 32768
+        ppo_micro_batch_size: null
+        ppo_micro_batch_size_per_gpu: 1
+        ppo_mini_batch_size: 32
+        profiler:
+            _target_: verl.utils.profiler.ProfilerConfig
+            all_ranks: false
+            enable: false
+            ranks: []
+            save_path: outputs/profile
+            tool: null
+            tool_config:
+                npu:
+                    _target_: verl.utils.profiler.config.NPUToolConfig
+                    analysis: true
+                    contents: []
+                    discrete: false
+                    level: level1
+                nsys:
+                    _target_: verl.utils.profiler.config.NsightToolConfig
+                    discrete: false
+                torch:
+                    _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                    step_end: null
+                    step_start: 0
+                torch_memory:
+                    _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                    stack_depth: 32
+                    trace_alloc_max_entries: 100000
+        rollout_n: 1
+        shuffle: false
+        strategy: fsdp
+        ulysses_sequence_parallel_size: 1
+        use_dynamic_bsz: false
+ctx_manager:
+    value:
+        generation:
+            gen_config:
+                kwargs: null
+                response_length: 400
+                temperature: 1
+                top_k: -1
+                top_p: 1
+custom_envs:
+    value:
+        Bandit:
+            env_config:
+                split: train
+            env_instruction: ""
+            env_type: bandit
+            max_actions_per_traj: 1
+            max_tokens: 100
+        BanditTest:
+            env_config:
+                split: test
+            env_instruction: ""
+            env_type: bandit
+            max_actions_per_traj: 1
+            max_tokens: 100
+        CoordFrozenLake:
+            env_config:
+                observation_format: grid_coord
+            env_instruction: |
+                You are solving the FrozenLake puzzle. The observation includes both a symbol grid and zero-indexed coordinates for the start, goal, player, and any holes.
+                Coordinates range from the top-left corner (0, 0) to the bottom-right corner (5, 5).
+                Beware that the ice is slippery, so the agent might slide and end up in an unintended tile.
+                Respond with a sequence of actions such as <answer>Left || Up || Up</answer>.
+            env_type: frozen_lake
+            max_actions_per_traj: 25
+            max_tokens: 120
+        CoordSokoban:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                max_steps: 100
+                num_boxes: 2
+                observation_format: grid_coord
+            env_instruction: "You are solving the Sokoban puzzle. You are the player and you need to push all boxes to targets.\nYou are provided with a symbol grid and the zero-indexed coordinates of the player, each box, and each target. \nCoordinates range from the top-left corner (0, 0) to the bottom-right corner (5, 5). \nWhen you are exactly next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box.\nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>.\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 120
+        Countdown:
+            env_config: null
+            env_instruction: 'You are solving the Countdown puzzle. You should use the num list to create an equation that equals the target. Example answer format: <think> To find an equation using [3, 5, 2] to get 4. Let''s check 2 + 5 = 7, 7 - 3 = 4. So the answer is 2 + 5 - 3 = 4. </think><answer>2 + 5 - 3</answer>'
+            env_type: countdown
+            max_actions_per_traj: 1
+            max_tokens: 100
+        FrozenLake:
+            env_config: null
+            env_instruction: 'You are solving the FrozenLake puzzle. Forbid the whole and go to the target. You may move to the unintended direction due to the slippery ice. Example answer format: <think>To forbid the hole and go to the target, I should go left then go up.</think><answer>Left || Up</answer>'
+            env_type: frozen_lake
+            max_actions_per_traj: 25
+            max_tokens: 100
+        LargerSokoban:
+            env_config:
+                dim_x: 8
+                dim_y: 8
+                max_steps: 100
+                num_boxes: 2
+                search_depth: 10
+            env_instruction: "You are solving the Sokoban puzzle. \nYou are the player and you need to push all boxes to targets. \nWhen you are right next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box. \nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        Lean:
+            env_config: null
+            env_instruction: You are a Lean theorem prover. Given a Lean theorem statement, propose a sequence of tactics that completes the proof. Think step by step about which tactics to apply next. Provide tactics separated by '||', for example <answer>intro || simp || rfl</answer>.
+            env_type: lean
+            max_actions_per_traj: 30
+            max_tokens: 512
+        MediumSudoku:
+            env_config:
+                difficulty: medium
+                grid_size: 9
+                max_steps: 81
+                render_format: with_feedback
+                show_conflicts: true
+                show_valid_numbers: true
+            env_instruction: |
+                You are solving a Sudoku puzzle. Fill in the grid so that every row, column, and 3x3 box contains the numbers 1-9 without repetition.
+                Initial cells are shown in [brackets] and cannot be modified. Empty cells are shown as dots (.).
+                Place numbers one at a time using the format: <answer>place 5 at row 2 col 3</answer> or <answer>2,3,5</answer>
+                The environment will provide feedback on valid/invalid moves and show conflicts if any occur.
+            env_type: sudoku
+            max_actions_per_traj: 30
+            max_tokens: 150
+            max_workers: 32
+            parallel_friendly: false
+        MetamathQA:
+            env_config: null
+            env_instruction: 'You are solving Math problems. '
+            env_type: metamathqa
+            max_actions_per_traj: 1
+            max_tokens: 100
+        SimpleSokoban:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                max_steps: 100
+                num_boxes: 2
+            env_instruction: "You are solving the Sokoban puzzle. \nYou are the player and you need to push all boxes to targets. \nWhen you are right next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box. \nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        SimpleSudoku:
+            env_config:
+                difficulty: easy
+                grid_size: 4
+                max_steps: 20
+                render_format: with_feedback
+                show_conflicts: false
+                show_valid_numbers: false
+            env_instruction: |
+                You are solving a Sudoku puzzle. Fill in the grid so that every row, column, and 2x2 box contains the numbers 1-4 without repetition.
+                Initial cells are shown in [brackets] and cannot be modified. Empty cells are shown as dots (.).
+                Place numbers one at a time using the format, for example: <answer>place 1 at row 2 col 3</answer> or <answer>1,2,3</answer>
+                The environment will provide feedback on valid/invalid moves and show conflicts if any occur.
+            env_type: sudoku
+            max_actions_per_traj: 20
+            max_tokens: 150
+            max_workers: 32
+            parallel_friendly: false
+        SokobanDifferentGridVocab:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                grid_lookup:
+                    "0": W
+                    "1": .
+                    "2": G
+                    "3": C
+                    "4": B
+                    "5": A
+                    "6": '@'
+                grid_vocab:
+                    .: empty
+                    '@': player on target
+                    A: player
+                    B: box
+                    C: box on target
+                    G: target
+                    W: wall
+                max_steps: 100
+                num_boxes: 1
+                search_depth: 30
+            env_instruction: "You are solving the Sokoban puzzle. \nYou are the player and you need to push all boxes to targets. \nWhen you are right next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box. \nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        VisualSimpleSokoban:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                max_steps: 100
+                num_boxes: 1
+                render_mode: rgb_array
+            env_instruction: You are solving the Sokoban puzzle. You are the player and you need to push all boxes to targets. When you are right next to a box, you can push it by moving in the same direction. You cannot push a box through a wall, and you cannot pull a box. The answer should be a sequence of actions, like <answer>Right || Right || Up</answer>
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        WebShop:
+            env_config:
+                dataset: small
+            env_instruction: 'You are browsing an online shop. Based on the instruction, buy a product that close to the production description. You need to search, read the search results, pick a product, choose the size and color and buy. You should only choose action from the available actions list provided later.  Example process: I need a gingko light and 20x20 pillow cover that is hand painted. First search[gingko light 20x20 pillow cover hand painted], answer format: <answer>search[blanket with fleece throw]</answer>. Valid answer is search[<keywords>] or click[<clickable>].'
+            env_type: webshop
+            max_actions_per_traj: 9
+            max_tokens: 200
+        blackjack:
+            env_config: null
+            env_instruction: |
+                You are playing Blackjack against a dealer. The dealer must hit on 16 or less and stand on 17 or more.
+                Choose either Stick or Hit. Respond with a single action.
+                Example: <answer>Hit</answer>
+            env_type: blackjack
+            max_actions_per_traj: 10
+            max_tokens: 64
+        game_2048:
+            env_config: null
+            env_instruction: |
+                You are playing the 2048 game on a 4x4 grid. Merge equal tiles by sliding Up, Right, Down, or Left.
+                If a move is invalid (no tiles move), a small penalty is applied. Respond with a single action.
+                Example: <answer>Up</answer>
+            env_type: game_2048
+            max_actions_per_traj: 700
+            max_tokens: 8192
+        rubikscube:
+            env_config:
+                max_steps: 20
+                render_mode: text
+                scramble_depth: 1
+            env_instruction: |
+                You are solving a 2x2 Rubik's Cube (Pocket Cube). The goal is to restore the cube so that each of the faces consists of a single, unique color.
+                Available actions use standard Singmaster notation for face rotations: U, U', D, D', L, L', R, R', F, F', B, B'.
+                - Faces: U (Up), D (Down), L (Left), R (Right), F (Front), B (Back).
+                - Modifiers: A letter alone means 90° clockwise (e.g., 'R'). A letter with prime (') means 90° counter-clockwise (e.g., "R'").
+                Respond with a sequence of actions separated by "||".
+                Example: <answer>U</answer>
+            env_type: rubikscube
+            max_actions_per_traj: 20
+            max_tokens: 96
+custom_reward_function:
+    value:
+        name: compute_score
+        path: null
+data:
+    value:
+        custom_cls:
+            name: null
+            path: null
+        datagen:
+            name: null
+            path: null
+        dataloader_num_workers: 8
+        filter_overlong_prompts: false
+        filter_overlong_prompts_workers: 1
+        image_key: images
+        max_prompt_length: null
+        max_response_length: null
+        prompt_key: prompt
+        return_full_prompt: false
+        return_multi_modal_inputs: true
+        return_raw_chat: false
+        return_raw_input_ids: false
+        reward_fn_key: data_source
+        sampler:
+            class_name: null
+            class_path: null
+        shuffle: true
+        tokenizer: null
+        train_batch_size: 128
+        train_files: ~/data/rlhf/gsm8k/train.parquet
+        truncation: error
+        trust_remote_code: false
+        use_shm: false
+        val_batch_size: null
+        val_files: ~/data/rlhf/gsm8k/test.parquet
+        validation_shuffle: false
+        video_key: videos
+enable_response_mask:
+    value: true
+es_manager:
+    value:
+        format_penalty: -0.1
+        train:
+            env_configs:
+                n_groups:
+                    - 8
+                tags:
+                    - CoordFrozenLake
+            env_groups: 8
+            group_size: 16
+        val:
+            env_configs:
+                n_groups:
+                    - 256
+                tags:
+                    - CoordFrozenLake
+            env_groups: 256
+            group_size: 1
+global_profiler:
+    value:
+        _target_: verl.utils.profiler.ProfilerConfig
+        global_tool_config:
+            nsys:
+                _target_: verl.utils.profiler.config.NsightToolConfig
+                controller_nsight_options:
+                    cuda-graph-trace: graph
+                    cuda-memory-usage: "true"
+                    trace: cuda,nvtx,cublas,ucx
+                discrete: false
+                worker_nsight_options:
+                    capture-range: cudaProfilerApi
+                    capture-range-end: null
+                    cuda-graph-trace: graph
+                    cuda-memory-usage: "true"
+                    kill: none
+                    trace: cuda,nvtx,cublas,ucx
+            torch_memory:
+                context: all
+                stack_depth: 32
+                stacks: all
+                trace_alloc_max_entries: 100000
+        profile_continuous_steps: false
+        save_path: outputs/profile
+        steps: null
+        tool: null
+grpo_advantage_length_weight:
+    value: false
+lora:
+    value:
+        alpha: 16
+        rank: 0
+        target_modules: all-linear
+micro_batch_size_per_gpu:
+    value: 1
+model_path:
+    value: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+ppo_mini_batch_size:
+    value: 32
+ray_kwargs:
+    value:
+        ray_init:
+            num_cpus: null
+        timeline_json_file: null
+reward_model:
+    value:
+        enable: false
+        enable_resource_pool: false
+        forward_max_token_len_per_gpu: 32768
+        launch_reward_fn_async: false
+        max_length: null
+        micro_batch_size: null
+        micro_batch_size_per_gpu: null
+        model:
+            external_lib: null
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                forward_prefetch: false
+                fsdp_size: -1
+                param_offload: false
+                reshard_after_forward: true
+                wrap_policy:
+                    min_num_params: 0
+            input_tokenizer: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            path: ~/models/FsfairX-LLaMA3-RM-v0.1
+            trust_remote_code: false
+            use_fused_kernels: false
+            use_remove_padding: false
+            use_shm: false
+        n_gpus_per_node: 0
+        nnodes: 0
+        profiler:
+            _target_: verl.utils.profiler.ProfilerConfig
+            all_ranks: false
+            enable: false
+            ranks: []
+            save_path: outputs/profile
+            tool: null
+            tool_config:
+                npu:
+                    _target_: verl.utils.profiler.config.NPUToolConfig
+                    analysis: true
+                    contents: []
+                    discrete: false
+                    level: level1
+                nsys:
+                    _target_: verl.utils.profiler.config.NsightToolConfig
+                    discrete: false
+                torch:
+                    _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                    step_end: null
+                    step_start: 0
+                torch_memory:
+                    _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                    stack_depth: 32
+                    trace_alloc_max_entries: 100000
+        reward_manager: naive
+        sandbox_fusion:
+            max_concurrent: 64
+            memory_limit_mb: 1024
+            url: null
+        strategy: fsdp
+        ulysses_sequence_parallel_size: 1
+        use_dynamic_bsz: false
+seed:
+    value:
+        train: 10000
+        val: 123
+system:
+    value:
+        CUDA_VISIBLE_DEVICES: 0,1,2,3,4,5,6,7
+trainer:
+    value:
+        balance_batch: true
+        critic_warmup: 0
+        default_hdfs_dir: null
+        default_local_dir: /mnt/general/wanghy/RAGEN/saves/
+        del_local_ckpt_after_load: false
+        device: cuda
+        esi_redundant_time: 0
+        experiment_name: frozen_lake-starpo-slippery
+        generations_to_log_to_wandb:
+            train: 128
+            val: 20
+        local_log_dir: results/
+        log_val_generations: 0
+        logger:
+            - console
+            - wandb
+        max_actor_ckpt_to_keep: 1
+        max_critic_ckpt_to_keep: 1
+        n_gpus_per_node: 8
+        nnodes: 1
+        project_name: ragen_latest_llama_1B_it
+        ray_wait_register_center_timeout: 300
+        resume_from_path: null
+        resume_mode: auto
+        rollout_data_dir: null
+        save_freq: -1
+        test_freq: 10
+        total_epochs: 30
+        total_training_steps: 200
+        use_legacy_worker_impl: auto
+        val_before_train: true
+        val_only: false
+        validation_data_dir: null
+        validation_steps: 1

wandb/run-20251222_202905-69ev02ff/files/requirements.txt

@@ -0,0 +1,316 @@
+ragen==0.1
+ragen==0.1
+colorama==0.4.6
+psutil==7.1.2
+ragen==0.1
+setuptools==80.9.0
+wheel==0.45.1
+pip==25.3
+ragen==0.1
+pytz==2025.2
+nvidia-cusparselt-cu12==0.6.2
+mpmath==1.3.0
+typing_extensions==4.15.0
+sympy==1.13.1
+pylatexenc==2.10
+nvidia-cusolver-cu12==11.6.1.9
+numpy==1.26.4
+Markdown==3.10
+absl-py==2.3.1
+pyasn1_modules==0.4.2
+Werkzeug==3.1.3
+networkx==3.3
+MarkupSafe==2.1.5
+fsspec==2025.9.0
+filelock==3.19.1
+nvidia-cudnn-cu12==9.1.0.70
+Jinja2==3.1.6
+tensorboard==2.20.0
+triton==3.2.0
+pillow==11.3.0
+xxhash==3.6.0
+nvidia-nvtx-cu12==12.4.127
+nvidia-nvjitlink-cu12==12.4.127
+nvidia-nccl-cu12==2.21.5
+nvidia-curand-cu12==10.3.5.147
+wrapt==2.0.1
+nvidia-cufft-cu12==11.2.1.3
+pyjnius==1.7.0
+nvidia-cuda-runtime-cu12==12.4.127
+nvidia-cuda-nvrtc-cu12==12.4.127
+websockets==15.0.1
+nvidia-cuda-cupti-cu12==12.4.127
+zipp==3.23.0
+nvidia-cublas-cu12==12.4.5.8
+nvidia-cusparse-cu12==12.3.1.170
+flash_attn==2.8.3
+rsa==4.9.1
+einops==0.8.1
+py-cpuinfo==9.0.0
+pure_eval==0.2.3
+ptyprocess==0.7.0
+gym-notices==0.1.0
+flatbuffers==25.9.23
+fastrlock==0.8.3
+Farama-Notifications==0.0.4
+faiss-gpu==1.7.2
+antlr4-python3-runtime==4.9.3
+wcwidth==0.2.14
+wasabi==1.1.3
+uvloop==0.22.1
+urllib3==2.5.0
+tzdata==2025.2
+typing-inspection==0.4.2
+traitlets==5.14.3
+tqdm==4.67.1
+tomli==2.3.0
+threadpoolctl==3.6.0
+tenacity==9.1.2
+tabulate==0.9.0
+spacy-loggers==1.0.5
+spacy-legacy==3.0.12
+soupsieve==2.8
+sniffio==1.3.1
+smmap==5.0.2
+six==1.17.0
+shellingham==1.5.4
+sentencepiece==0.2.1
+scipy==1.15.3
+safetensors==0.6.2
+rpds-py==0.29.0
+rignore==0.7.6
+regex==2025.11.3
+RapidFuzz==3.14.3
+rank-bm25==0.2.2
+pyzmq==27.1.0
+PyYAML==6.0.3
+pytokens==0.3.0
+python-multipart==0.0.20
+python-json-logger==4.0.0
+python-dotenv==1.2.1
+PySocks==1.7.1
+pyparsing==3.2.5
+PyJWT==2.10.1
+Pygments==2.19.2
+pygame==2.6.1
+pydantic_core==2.41.5
+pycparser==2.23
+pycountry==24.6.1
+pybind11==3.0.1
+pyarrow==22.0.0
+psutil==7.1.3
+protobuf==4.25.8
+propcache==0.4.1
+prometheus_client==0.23.1
+platformdirs==4.5.0
+pexpect==4.9.0
+pathspec==0.12.1
+partial-json-parser==0.2.1.1.post6
+parso==0.8.5
+packaging==25.0
+orjson==3.11.4
+opentelemetry-semantic-conventions-ai==0.4.13
+opencv-python-headless==4.12.0.88
+ninja==1.13.0
+nest-asyncio==1.6.0
+mypy_extensions==1.1.0
+murmurhash==1.0.15
+multidict==6.7.0
+msgspec==0.19.0
+msgpack==1.1.2
+mdurl==0.1.2
+loguru==0.7.3
+llvmlite==0.44.0
+llguidance==0.7.30
+lark==1.2.2
+kiwisolver==1.4.9
+joblib==1.5.2
+jiter==0.12.0
+itsdangerous==2.2.0
+interegular==0.3.3
+ImageIO==2.37.2
+idna==3.11
+humanfriendly==10.0
+httpx-sse==0.4.3
+httptools==0.7.1
+html2text==2025.4.15
+hf-xet==1.2.0
+h11==0.16.0
+grpcio==1.76.0
+frozenlist==1.8.0
+fonttools==4.60.1
+executing==2.2.1
+exceptiongroup==1.3.0
+eval_type_backport==0.2.2
+docstring_parser==0.17.0
+dnspython==2.8.0
+distro==1.9.0
+diskcache==5.6.3
+dill==0.4.0
+decorator==5.2.1
+debugpy==1.8.17
+Cython==3.2.1
+cymem==2.0.13
+cycler==0.12.1
+cupy-cuda12x==13.6.0
+contourpy==1.3.2
+colorama==0.4.6
+codetiming==1.4.0
+cloudpickle==3.1.2
+cloudpathlib==0.23.0
+click==8.3.1
+charset-normalizer==3.4.4
+certifi==2025.11.12
+catalogue==2.0.10
+cachetools==6.2.2
+blis==1.3.2
+blinker==1.9.0
+blake3==1.0.8
+attrs==25.4.0
+async-timeout==5.0.1
+asttokens==3.0.1
+astor==0.8.1
+annotated-types==0.7.0
+annotated-doc==0.0.4
+airportsdata==20250909
+aiohappyeyeballs==2.6.1
+yarl==1.22.0
+uvicorn==0.38.0
+typer-slim==0.20.0
+thefuzz==0.22.1
+stack-data==0.6.3
+srsly==2.5.1
+smart_open==7.5.0
+sentry-sdk==2.44.0
+scikit-learn==1.7.2
+requests==2.32.5
+referencing==0.37.0
+python-dateutil==2.9.0.post0
+pydantic==2.12.4
+prompt_toolkit==3.0.52
+preshed==3.0.11
+opentelemetry-proto==1.26.0
+omegaconf==2.3.0
+numba==0.61.2
+nltk==3.9.2
+multiprocess==0.70.18
+matplotlib-inline==0.2.1
+markdown-it-py==4.0.0
+jedi==0.19.2
+importlib_metadata==8.0.0
+httpcore==1.0.9
+gymnasium==1.2.2
+gym==0.26.2
+googleapis-common-protos==1.72.0
+gitdb==4.0.12
+gguf==0.17.1
+Flask==3.1.2
+faiss-cpu==1.13.0
+email-validator==2.3.0
+depyf==0.18.0
+Deprecated==1.3.1
+coloredlogs==15.0.1
+cffi==2.0.0
+black==25.11.0
+beautifulsoup4==4.14.2
+anyio==4.11.0
+aiosignal==1.4.0
+watchfiles==1.1.1
+tiktoken==0.12.0
+starlette==0.49.3
+sse-starlette==3.0.3
+rich==14.2.0
+pydantic-settings==2.12.0
+pydantic-extra-types==2.10.6
+pandas==2.3.3
+opentelemetry-exporter-otlp-proto-common==1.26.0
+onnxruntime==1.23.2
+matplotlib==3.10.7
+lm-format-enforcer==0.10.12
+jsonschema-specifications==2025.9.1
+ipython==8.37.0
+hydra-core==1.3.2
+huggingface-hub==0.36.0
+httpx==0.28.1
+gym-sokoban==0.0.6
+GitPython==3.1.45
+cryptography==46.0.3
+confection==0.1.5
+cleantext==1.1.4
+aiohttp==3.13.2
+weasel==0.4.3
+wandb==0.23.0
+typer==0.19.2
+virtualenv==20.35.4
+pyasn1==0.6.1
+torch==2.6.0
+tokenizers==0.22.1
+thinc==8.3.9
+rich-toolkit==0.15.1
+prometheus-fastapi-instrumentator==7.1.0
+openai==2.8.0
+jsonschema==4.25.1
+gdown==5.2.0
+fastapi==0.121.2
+anthropic==0.73.0
+xformers==0.0.29.post2
+transformers==4.57.1
+proto-plus==1.26.1
+torchvision==0.21.0
+torchdata==0.11.0
+torchaudio==2.6.0
+together==1.5.30
+tensordict==0.8.3
+spacy==3.8.9
+ray==2.51.1
+outlines_core==0.1.26
+mistral_common==1.8.5
+mcp==1.21.1
+fastapi-cloud-cli==0.3.1
+fastapi-cli==0.0.16
+datasets==4.4.1
+accelerate==1.11.0
+xgrammar==0.1.18
+pyserini==1.2.0
+peft==0.18.0
+outlines==0.1.11
+opentelemetry-exporter-otlp-proto-http==1.26.0
+opentelemetry-exporter-otlp-proto-grpc==1.26.0
+kimina-client==0.2.1
+compressed-tensors==0.9.3
+opentelemetry-exporter-otlp==1.26.0
+vllm==0.8.5
+py-spy==0.4.1
+opencensus-context==0.1.3
+distlib==0.4.0
+colorful==0.5.8
+tensorboard-data-server==0.7.2
+google-auth==2.43.0
+google-api-core==2.28.1
+aiohttp-cors==0.8.1
+opentelemetry-exporter-prometheus==0.59b0
+opencensus==0.11.4
+verl==0.5.0.dev0
+opentelemetry-api==1.38.0
+opentelemetry-semantic-conventions==0.59b0
+opentelemetry-sdk==1.38.0
+typeguard==4.4.4
+shtab==1.8.0
+tyro==0.9.35
+verl==0.5.0.dev0
+autocommand==2.2.2
+backports.tarfile==1.2.0
+importlib_metadata==8.0.0
+inflect==7.3.1
+jaraco.collections==5.1.0
+jaraco.context==5.3.0
+jaraco.functools==4.0.1
+jaraco.text==3.12.1
+more-itertools==10.3.0
+packaging==24.2
+platformdirs==4.2.2
+tomli==2.0.1
+typeguard==4.3.0
+typing_extensions==4.12.2
+wheel==0.45.1
+zipp==3.19.2

wandb/run-20251222_202905-69ev02ff/files/wandb-metadata.json

@@ -0,0 +1,107 @@
+{
+  "os": "Linux-5.14.0-284.25.1.el9_2.x86_64-x86_64-with-glibc2.35",
+  "python": "CPython 3.10.19",
+  "startedAt": "2025-12-22T12:29:05.056794Z",
+  "args": [
+    "--node-ip-address=10.119.27.78",
+    "--node-manager-port=38739",
+    "--object-store-name=/tmp/ray/session_2025-12-22_20-27-49_028907_487744/sockets/plasma_store",
+    "--raylet-name=/tmp/ray/session_2025-12-22_20-27-49_028907_487744/sockets/raylet",
+    "--redis-address=None",
+    "--metrics-agent-port=45094",
+    "--logging-rotate-bytes=536870912",
+    "--logging-rotate-backup-count=5",
+    "--runtime-env-agent-port=54281",
+    "--gcs-address=10.119.27.78:36231",
+    "--session-name=session_2025-12-22_20-27-49_028907_487744",
+    "--temp-dir=/tmp/ray",
+    "--webui=127.0.0.1:8265",
+    "--cluster-id=cf93e6bfeed4aa4ef758a9bb28b14e8f7aeb59402129513d9cc31269",
+    "--startup-token=112",
+    "--worker-launch-time-ms=1766406478308",
+    "--node-id=f950bfc2cfb0e6e294c1e34ef1cf9e9eac2db5dc349184b20fe78f90",
+    "--runtime-env-hash=-839022692"
+  ],
+  "program": "/root/local/miniconda3/envs/ragen/lib/python3.10/site-packages/ray/_private/workers/default_worker.py",
+  "git": {
+    "remote": "https://github.com/mll-lab-nu/RAGEN.git",
+    "commit": "8d73639b99b38265453f898b8d6af7d4af50d56e"
+  },
+  "email": "haoyu-wa22@mails.tsinghua.edu.cn",
+  "root": "/mnt/general/wanghy/RAGEN",
+  "host": "pt-4d654cf4576f4d23ad3d3919f12932fe-worker-0",
+  "executable": "/root/local/miniconda3/envs/ragen/bin/python",
+  "cpu_count": 64,
+  "cpu_count_logical": 128,
+  "gpu": "NVIDIA H100 80GB HBM3",
+  "gpu_count": 8,
+  "disk": {
+    "/": {
+      "total": "5153960755200",
+      "used": "471729799168"
+    }
+  },
+  "memory": {
+    "total": "2163642122240"
+  },
+  "gpu_nvidia": [
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-35e2d43d-4067-82ce-90d4-def9e389bf28"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-af4135e3-88f2-e9ac-518d-502c75a85429"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-d7fdeeba-fe9b-ec03-d9f7-6724fe4266b5"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-ccc4f668-3882-5a8e-2c07-c5cd08f6f666"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-7b73c0cf-d3d5-e10c-7176-a43be1e41001"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-81b58d94-5d1f-8ec2-f9d2-fd56172ed177"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-03e8bc66-3b44-6794-49fd-5392fbdda6d1"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-b7cf0ec6-7c29-1179-dceb-09565da51890"
+    }
+  ],
+  "cudaVersion": "12.4",
+  "writerId": "z84gczlr8c1c496ea0tnn6zrntod8u7k"
+}

wandb/run-20251222_202905-69ev02ff/files/wandb-summary.json

@@ -0,0 +1 @@
+{"timing_per_token_ms/update_critic":0.008671362954917283,"train/CoordFrozenLake/non-zero/action_is_effective":0.6167302679117422,"timing_s/gen":71.00658818334341,"response_length_non_aborted/clip_ratio":0.1875,"response_length/mean":4105.09375,"global_seqlen/balanced_min":15936,"rollout/in_group_reward_mean":-1.150781273841858,"response_length_non_aborted/max":7200,"val-env/charts/avg_episode_return":0,"rollout/chosen_in_group_reward_mean":-1.0343749523162842,"critic/values/min":-3.703125,"val-aux/unknown/reward/best@128/std":0,"prompt_length/min":0,"val-aux/unknown/reward/best@2/mean":-0.07410000060498714,"val-aux/unknown/reward/best@4/mean":0.4936999987959862,"_runtime":192.877234237,"actor/grad_norm":14.57581615447998,"val-aux/unknown/reward/worst@256/std":0,"val-env/CoordFrozenLake/non-zero/success":1,"timing_s/old_log_prob":1.1578870043158531,"critic/score/max":1,"timing_per_token_ms/gen":0.5405371998457968,"val-env/CoordFrozenLake/num_actions":6.78515625,"timing_per_token_ms/ref":0.008699123257084133,"train/CoordFrozenLake/non-zero/pass@16":1,"global_seqlen/mean":16421.125,"val-aux/unknown/reward/worst@32/std":0,"train/CoordFrozenLake/non-zero/num_actions":6.858974358974359,"val-aux/unknown/reward/best@16/mean":0.9669999995231628,"val-aux/unknown/reward/worst@256/mean":-2.5,"critic/values/max":8.6875,"response_length_non_aborted/mean":4105.09375,"val-aux/unknown/reward/best@16/std":0.17352521449027572,"val-aux/unknown/reward/best@2/std":1.1340807676744216,"val-aux/unknown/reward/worst@64/mean":-2.5,"timing_per_token_ms/values":0.0022152539187725064,"rollout/chosen_in_group_reward_max":1,"global_seqlen/max":22258,"_timestamp":1.7664067005738702e+09,"val-aux/unknown/reward/std@256":1.4064409039854844,"rollout/chosen_in_group_max":1,"val-aux/unknown/reward/best@64/mean":1,"perf/mfu/actor":0.10476092967073983,"perf/max_memory_allocated_gb":73.42222261428833,"val-aux/unknown/reward/worst@2/mean":-1.5127999997437,"critic/vf_clipfrac":0,"val-core/unknown/reward/mean@256":-0.8128906251804437,"val-aux/unknown/reward/worst@16/mean":-2.49760000000149,"val/generations":{"sha256":"48ef2484dcca7a8afd4a1bbfb43d9975db4cae66783fa0443288adfca6fffb92","log_mode":"IMMUTABLE","nrows":1,"size":221910,"artifact_path":"wandb-client-artifact://ms0qkuinkj2jncxq9x3qalfr77qj7q9x5eq868gk3iga9v2htiuh4qajs52v3v4y9gad1s0bzjj2ccfrs2n555co4ltss5n71iwnim7q0ijy2afkz13qzzg8yiewdppu/val/generations.table.json","ncols":61,"_latest_artifact_path":"wandb-client-artifact://k19zq8afs66ju7v8athh8ngnmdj6nh4jbx9z8xr8ujldkfuzsbbxg3h9j1zysz39l7zkngux9zft5q42nwk2sjbvrx76o0lkpf0lgrv2vb8gu2sfuqfd1duma0pqw5ni:latest/val/generations.table.json","path":"media/table/val/generations_0_48ef2484dcca7a8afd4a.table.json","_type":"table-file"},"perf/cpu_memory_used_gb":102.75568389892578,"train/charts/avg_episode_return":0,"critic/rewards/max":1,"val-env/CoordFrozenLake/action_is_valid":0.597907788890251,"perf/total_num_tokens":131369,"rollout/in_group_max":1,"critic/vf_explained_var":-7.614336967468262,"rollout/in_group_std":1.464717984199524,"val-aux/unknown/reward/worst@128/std":0,"critic/score/mean":-1.0343749523162842,"val-aux/unknown/reward/worst@4/mean":-2.1617999996989967,"train/CoordFrozenLake/non-zero/action_is_valid":0.7212219967052853,"timing_per_token_ms/update_actor":0.013922594988402997,"perf/throughput":213.57489198900478,"val-aux/unknown/reward/worst@64/std":0,"_step":1,"critic/returns/mean":-2.2920210361480713,"perf/max_memory_reserved_gb":81.94921875,"val-aux/unknown/reward/worst@8/mean":-2.445700000055134,"prompt_length/mean":0.1875,"rollout/chosen_in_group_reward_std":1.6747982501983643,"response_length/max":7200,"val-env/CoordFrozenLake/episodic_return":0,"train/response_length":2451.859375,"rollout/ref_log_prob":-8.952258110046387,"critic/grad_norm":392.3230285644531,"timing_s/adv":0.31686798483133316,"val-env/response_length":1341.08984375,"val-env/CoordFrozenLake/pass@1":0.18359375,"response_length/min":340,"val-env/CoordFrozenLake/action_is_effective":0.4822152752149087,"val-aux/unknown/reward/best@8/mean":0.8145999987125396,"timing_s/values":0.2910156920552254,"critic/score/min":-2.5,"critic/values/mean":-0.9089804291725159,"actor/ppo_kl":0,"critic/advantages/max":2.9235451221466064,"rollout/in_group_reward_max":1,"actor/entropy":8.921223640441895,"actor/lr":1e-06,"global_seqlen/balanced_max":17225,"train/CoordFrozenLake/action_is_valid":0.4000528262974629,"train/CoordFrozenLake/raw_reward":0,"val-env/CoordFrozenLake/non-zero/pass@1":1,"_wandb":{"runtime":192},"prompt_length/clip_ratio":0.1875,"timing_s/ref":1.1427951231598854,"response/aborted_ratio":0,"val-core/unknown/reward/best@256/std":0,"response_length_non_aborted/min":340,"rollout/in_group_mean":-1.150781273841858,"val-aux/unknown/reward/worst@16/std":0.07585670697344167,"global_seqlen/minmax_diff":14836,"train/CoordFrozenLake/action_is_effective":0.3324561600461736,"val-aux/unknown/reward/worst@4/std":0.8463809778113507,"val-aux/unknown/reward/best@32/mean":0.9987999999523163,"global_seqlen/min":7422,"response_length/clip_ratio":0.1875,"val-core/unknown/reward/best@256/mean":1,"train/CoordFrozenLake/num_actions":4.1796875,"val-aux/unknown/reward/best@32/std":0.031914886956126036,"val-aux/unknown/reward/worst@2/std":1.254008037631705,"val-aux/unknown/reward/best@64/std":0,"train/CoordFrozenLake/success":0.28125,"perf/mfu/critic":0.16965798966678863,"val-env/CoordFrozenLake/non-zero/action_is_valid":0.9749324455790079,"val-env/CoordFrozenLake/non-zero/num_actions":10.99367088607595,"timing_s/update_actor":1.8289973810315132,"rollout/chosen_in_group_std":1.6747982501983643,"val-aux/unknown/reward/best@8/std":0.3823177209862261,"train/CoordFrozenLake/episodic_return":0,"actor/pg_clipfrac":0,"val-aux/unknown/reward/best@128/mean":1,"critic/advantages/mean":-5.72340610460742e-08,"val-aux/unknown/reward/worst@128/mean":-2.5,"perf/time_per_step":76.8869638517499,"val-aux/unknown/reward/worst@32/mean":-2.5,"train/CoordFrozenLake/non-zero/success":1,"actor/pg_clipfrac_lower":0,"val-env/CoordFrozenLake/raw_reward":0,"actor/pg_loss":0.1883917967788875,"critic/rewards/min":-2.5,"timing_s/total":76.89122772216797,"prompt_length/max":1,"critic/lr":1e-05,"val-env/CoordFrozenLake/success":0.18359375,"timing_s/update_critic":1.1391482800245285,"critic/vf_loss":1.7385750859975815,"critic/returns/min":-2.5,"val-env/CoordFrozenLake/non-zero/action_is_effective":0.7913276311219014,"rollout/chosen_in_group_mean":-1.0343749523162842,"critic/rewards/mean":-1.0343749523162842,"timing_s/step":76.8869638517499,"critic/vpred_mean":1.2447870895266533,"val-aux/unknown/reward/best@4/std":0.6590829303247507,"val-aux/unknown/reward/worst@8/std":0.3541489938383426,"rollout/in_group_reward_std":1.464717984199524,"critic/advantages/min":-5.4256205558776855,"train/CoordFrozenLake/pass@16":1,"timing_per_token_ms/adv":0.0024120453442694483,"critic/returns/max":1}

wandb/run-20251222_202905-69ev02ff/run-69ev02ff.wandb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f1b96d87cd653a831d7ad60078d7722ca758df870d7947ae72feafe32da6565
+size 101309

wandb/run-20251222_203657-pjvd64ya/files/config.yaml

@@ -0,0 +1,962 @@
+_wandb:
+    value:
+        cli_version: 0.23.0
+        e:
+            vece8z0wtyrd6guwhtkw2pg8e4uzzajh:
+                args:
+                    - --node-ip-address=10.119.27.78
+                    - --node-manager-port=33093
+                    - --object-store-name=/tmp/ray/session_2025-12-22_20-35-38_490044_519970/sockets/plasma_store
+                    - --raylet-name=/tmp/ray/session_2025-12-22_20-35-38_490044_519970/sockets/raylet
+                    - --redis-address=None
+                    - --metrics-agent-port=64448
+                    - --logging-rotate-bytes=536870912
+                    - --logging-rotate-backup-count=5
+                    - --runtime-env-agent-port=61324
+                    - --gcs-address=10.119.27.78:59659
+                    - --session-name=session_2025-12-22_20-35-38_490044_519970
+                    - --temp-dir=/tmp/ray
+                    - --webui=127.0.0.1:8265
+                    - --cluster-id=0dc698b87906cc59debdc15a5634e568861d5234a9cc80ce98248a23
+                    - --startup-token=112
+                    - --worker-launch-time-ms=1766406950439
+                    - --node-id=6fb8b54e88b3c54a306633a87e28c055cbbe44a1a5b5852b0d26078f
+                    - --runtime-env-hash=-839022692
+                cpu_count: 64
+                cpu_count_logical: 128
+                cudaVersion: "12.4"
+                disk:
+                    /:
+                        total: "5153960755200"
+                        used: "471741583360"
+                email: haoyu-wa22@mails.tsinghua.edu.cn
+                executable: /root/local/miniconda3/envs/ragen/bin/python
+                git:
+                    commit: 8d73639b99b38265453f898b8d6af7d4af50d56e
+                    remote: https://github.com/mll-lab-nu/RAGEN.git
+                gpu: NVIDIA H100 80GB HBM3
+                gpu_count: 8
+                gpu_nvidia:
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-35e2d43d-4067-82ce-90d4-def9e389bf28
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-af4135e3-88f2-e9ac-518d-502c75a85429
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-d7fdeeba-fe9b-ec03-d9f7-6724fe4266b5
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-ccc4f668-3882-5a8e-2c07-c5cd08f6f666
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-7b73c0cf-d3d5-e10c-7176-a43be1e41001
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-81b58d94-5d1f-8ec2-f9d2-fd56172ed177
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-03e8bc66-3b44-6794-49fd-5392fbdda6d1
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-b7cf0ec6-7c29-1179-dceb-09565da51890
+                host: pt-4d654cf4576f4d23ad3d3919f12932fe-worker-0
+                memory:
+                    total: "2163642122240"
+                os: Linux-5.14.0-284.25.1.el9_2.x86_64-x86_64-with-glibc2.35
+                program: /root/local/miniconda3/envs/ragen/lib/python3.10/site-packages/ray/_private/workers/default_worker.py
+                python: CPython 3.10.19
+                root: /mnt/general/wanghy/RAGEN
+                startedAt: "2025-12-22T12:36:57.314412Z"
+                writerId: vece8z0wtyrd6guwhtkw2pg8e4uzzajh
+        m: []
+        python_version: 3.10.19
+        t:
+            "1":
+                - 1
+                - 5
+                - 11
+                - 30
+                - 41
+                - 49
+                - 50
+                - 51
+                - 53
+                - 71
+                - 95
+                - 97
+                - 98
+                - 105
+            "2":
+                - 1
+                - 5
+                - 11
+                - 30
+                - 41
+                - 49
+                - 50
+                - 51
+                - 53
+                - 71
+                - 95
+                - 97
+                - 98
+                - 105
+            "3":
+                - 2
+                - 13
+                - 16
+                - 61
+            "4": 3.10.19
+            "5": 0.23.0
+            "6": 4.57.1
+            "12": 0.23.0
+            "13": linux-x86_64
+actor_rollout_ref:
+    value:
+        actor:
+            _target_: verl.workers.config.FSDPActorConfig
+            checkpoint:
+                _target_: verl.trainer.config.CheckpointConfig
+                async_save: false
+                load_contents:
+                    - model
+                    - optimizer
+                    - extra
+                save_contents:
+                    - model
+                    - optimizer
+                    - extra
+            clip_ratio: 0.2
+            clip_ratio_c: 3
+            clip_ratio_high: 0.28
+            clip_ratio_low: 0.2
+            entropy_checkpointing: false
+            entropy_coeff: 0.001
+            entropy_from_logits_with_chunking: false
+            freeze_vision_tower: false
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                entropy_checkpointing: false
+                entropy_from_logits_with_chunking: false
+                forward_only: false
+                forward_prefetch: false
+                fsdp_size: -1
+                model_dtype: fp32
+                offload_policy: false
+                optimizer_offload: false
+                param_offload: false
+                reshard_after_forward: true
+                strategy: fsdp
+                ulysses_sequence_parallel_size: 1
+                use_orig_params: false
+                use_torch_compile: true
+                wrap_policy:
+                    min_num_params: 0
+            grad_clip: 1
+            grpo_advantage_length_weight: false
+            kl_loss_coef: 0
+            kl_loss_type: kl
+            loss_agg_mode: token-mean
+            optim:
+                _target_: verl.workers.config.FSDPOptimizerConfig
+                betas:
+                    - 0.9
+                    - 0.999
+                clip_grad: 1
+                lr: 1e-06
+                lr_warmup_steps: -1
+                lr_warmup_steps_ratio: 0
+                min_lr_ratio: 0
+                num_cycles: 0.5
+                total_training_steps: 200
+                warmup_style: constant
+                weight_decay: 0.01
+            policy_loss:
+                _target_: verl.workers.config.PolicyLossConfig
+                clip_cov_lb: 1
+                clip_cov_ratio: 0.0002
+                clip_cov_ub: 5
+                kl_cov_ratio: 0.0002
+                loss_mode: vanilla
+                ppo_kl_coef: 0.1
+            ppo_epochs: 1
+            ppo_max_token_len_per_gpu: 16384
+            ppo_micro_batch_size: null
+            ppo_micro_batch_size_per_gpu: 1
+            ppo_mini_batch_size: 32
+            profiler:
+                _target_: verl.utils.profiler.ProfilerConfig
+                all_ranks: false
+                enable: false
+                ranks: []
+                save_path: outputs/profile
+                tool: null
+                tool_config:
+                    npu:
+                        _target_: verl.utils.profiler.config.NPUToolConfig
+                        analysis: true
+                        contents: []
+                        discrete: false
+                        level: level1
+                    nsys:
+                        _target_: verl.utils.profiler.config.NsightToolConfig
+                        discrete: false
+                    torch:
+                        _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                        step_end: null
+                        step_start: 0
+                    torch_memory:
+                        _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                        stack_depth: 32
+                        trace_alloc_max_entries: 100000
+            shuffle: false
+            strategy: fsdp
+            tis_imp_ratio_cap: -1
+            ulysses_sequence_parallel_size: 1
+            use_dynamic_bsz: false
+            use_fused_kernels: false
+            use_kl_loss: false
+            use_ref: true
+            use_remove_padding: false
+            use_torch_compile: true
+        hybrid_engine: true
+        model:
+            _target_: verl.workers.config.HFModelConfig
+            custom_chat_template: null
+            enable_activation_offload: false
+            enable_gradient_checkpointing: true
+            exclude_modules: null
+            external_lib: null
+            fused_kernel_options:
+                impl_backend: torch
+            hf_config_path: null
+            lora_alpha: 16
+            lora_rank: 0
+            path: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            target_modules: all-linear
+            tokenizer_path: null
+            torch_dtype: bfloat16
+            trust_remote_code: false
+            use_fused_kernels: false
+            use_liger: false
+            use_remove_padding: false
+            use_shm: false
+        nccl_timeout: 600
+        ref:
+            entropy_checkpointing: false
+            entropy_from_logits_with_chunking: false
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                entropy_checkpointing: false
+                entropy_from_logits_with_chunking: false
+                forward_only: false
+                forward_prefetch: false
+                fsdp_size: -1
+                model_dtype: fp32
+                offload_policy: false
+                optimizer_offload: false
+                param_offload: false
+                reshard_after_forward: true
+                strategy: fsdp
+                ulysses_sequence_parallel_size: 1
+                use_orig_params: false
+                use_torch_compile: true
+                wrap_policy:
+                    min_num_params: 0
+            log_prob_max_token_len_per_gpu: 16384
+            log_prob_micro_batch_size: null
+            log_prob_micro_batch_size_per_gpu: 1
+            log_prob_use_dynamic_bsz: false
+            model: null
+            profiler:
+                _target_: verl.utils.profiler.ProfilerConfig
+                all_ranks: false
+                enable: false
+                ranks: []
+                save_path: outputs/profile
+                tool: null
+                tool_config:
+                    npu:
+                        _target_: verl.utils.profiler.config.NPUToolConfig
+                        analysis: true
+                        contents: []
+                        discrete: false
+                        level: level1
+                    nsys:
+                        _target_: verl.utils.profiler.config.NsightToolConfig
+                        discrete: false
+                    torch:
+                        _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                        step_end: null
+                        step_start: 0
+                    torch_memory:
+                        _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                        stack_depth: 32
+                        trace_alloc_max_entries: 100000
+            strategy: fsdp
+            ulysses_sequence_parallel_size: 1
+            use_torch_compile: true
+        rollout:
+            _target_: verl.workers.config.RolloutConfig
+            agent:
+                _target_: verl.workers.config.AgentLoopConfig
+                agent_loop_config_path: null
+                custom_async_server:
+                    _target_: verl.workers.config.CustomAsyncServerConfig
+                    name: null
+                    path: null
+                num_workers: 8
+            calculate_log_probs: false
+            cudagraph_capture_sizes: null
+            data_parallel_size: 1
+            disable_log_stats: true
+            do_sample: true
+            dtype: bfloat16
+            enable_chunked_prefill: true
+            enable_prefix_caching: true
+            enforce_eager: true
+            expert_parallel_size: 1
+            free_cache_engine: true
+            gpu_memory_utilization: 0.7
+            ignore_eos: false
+            layered_summon: false
+            load_format: dummy
+            log_prob_max_token_len_per_gpu: 16384
+            log_prob_micro_batch_size: null
+            log_prob_micro_batch_size_per_gpu: 1
+            log_prob_use_dynamic_bsz: false
+            max_model_len: 8000
+            max_num_batched_tokens: 8192
+            max_num_seqs: 1024
+            mode: sync
+            multi_stage_wake_up: false
+            multi_turn:
+                _target_: verl.workers.config.MultiTurnConfig
+                enable: false
+                format: hermes
+                interaction_config_path: null
+                max_assistant_turns: null
+                max_parallel_calls: 1
+                max_tool_response_length: 256
+                max_user_turns: null
+                num_repeat_rollouts: null
+                tokenization_sanity_check_mode: strict
+                tool_config_path: null
+                tool_response_truncate_side: middle
+                use_inference_chat_template: false
+            "n": 1
+            name: vllm
+            over_sample_rate: 0
+            profiler:
+                _target_: verl.utils.profiler.ProfilerConfig
+                all_ranks: false
+                enable: false
+                ranks: []
+                save_path: outputs/profile
+                tool: null
+                tool_config:
+                    npu:
+                        _target_: verl.utils.profiler.config.NPUToolConfig
+                        analysis: true
+                        contents: []
+                        discrete: false
+                        level: level1
+                    nsys:
+                        _target_: verl.utils.profiler.config.NsightToolConfig
+                        discrete: false
+                    torch:
+                        _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                        step_end: null
+                        step_start: 0
+                    torch_memory:
+                        _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                        stack_depth: 32
+                        trace_alloc_max_entries: 100000
+            prompt_length: 1
+            response_length: 400
+            rollout_filter_metric: reward_variance
+            rollout_filter_ratio: 0.25
+            rollout_filter_type: largest
+            skip_dump_dir: /tmp/rollout_dump
+            skip_rollout: false
+            skip_tokenizer_init: true
+            temperature: 1
+            tensor_model_parallel_size: 1
+            top_k: -1
+            top_p: 1
+            trace:
+                _target_: verl.workers.config.TraceConfig
+                backend: null
+                token2text: false
+            update_weights_bucket_megabytes: 512
+            val_kwargs:
+                _target_: verl.workers.config.SamplingConfig
+                do_sample: true
+                "n": 1
+                temperature: 0.5
+                top_k: -1
+                top_p: 1
+agent_proxy:
+    value:
+        action_sep: '||'
+        enable_think: true
+        max_actions_per_turn: 1
+        max_context_window: -1
+        max_turn: 25
+        reward_normalization:
+            grouping: state
+            method: identity
+        use_turn_scores: false
+algorithm:
+    value:
+        _target_: verl.trainer.config.AlgoConfig
+        adv_estimator: gae
+        bi_level_gae: false
+        gamma: 1
+        high_level_gamma: 0.95
+        kl_ctrl:
+            _target_: verl.trainer.config.KLControlConfig
+            horizon: 10000
+            kl_coef: 0
+            target_kl: 0.1
+            type: fixed
+        kl_penalty: kl
+        lam: 1
+        norm_adv_by_std_in_grpo: true
+        pf_ppo:
+            reweight_method: pow
+            weight_pow: 2
+        use_kl_in_reward: false
+        use_pf_ppo: false
+critic:
+    value:
+        _target_: verl.workers.config.FSDPCriticConfig
+        checkpoint:
+            _target_: verl.trainer.config.CheckpointConfig
+            async_save: false
+            load_contents:
+                - model
+                - optimizer
+                - extra
+            save_contents:
+                - model
+                - optimizer
+                - extra
+        cliprange_value: 0.5
+        enable: null
+        forward_max_token_len_per_gpu: 32768
+        forward_micro_batch_size: null
+        forward_micro_batch_size_per_gpu: 1
+        grad_clip: 1
+        loss_agg_mode: token-mean
+        model:
+            _target_: verl.workers.config.FSDPCriticModelCfg
+            enable_activation_offload: false
+            enable_gradient_checkpointing: true
+            external_lib: null
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                entropy_checkpointing: false
+                entropy_from_logits_with_chunking: false
+                forward_only: false
+                forward_prefetch: false
+                fsdp_size: -1
+                model_dtype: fp32
+                offload_policy: false
+                optimizer_offload: false
+                param_offload: false
+                reshard_after_forward: true
+                strategy: fsdp
+                ulysses_sequence_parallel_size: 1
+                use_orig_params: false
+                use_torch_compile: true
+                wrap_policy:
+                    min_num_params: 0
+            lora_alpha: 16
+            lora_rank: 0
+            path: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            target_modules: all-linear
+            tokenizer_path: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            trust_remote_code: false
+            use_remove_padding: false
+            use_shm: false
+        optim:
+            _target_: verl.workers.config.FSDPOptimizerConfig
+            betas:
+                - 0.9
+                - 0.999
+            clip_grad: 1
+            lr: 1e-05
+            lr_warmup_steps: -1
+            lr_warmup_steps_ratio: 0
+            min_lr_ratio: 0
+            num_cycles: 0.5
+            total_training_steps: 200
+            warmup_style: constant
+            weight_decay: 0.01
+        ppo_epochs: 1
+        ppo_max_token_len_per_gpu: 32768
+        ppo_micro_batch_size: null
+        ppo_micro_batch_size_per_gpu: 1
+        ppo_mini_batch_size: 32
+        profiler:
+            _target_: verl.utils.profiler.ProfilerConfig
+            all_ranks: false
+            enable: false
+            ranks: []
+            save_path: outputs/profile
+            tool: null
+            tool_config:
+                npu:
+                    _target_: verl.utils.profiler.config.NPUToolConfig
+                    analysis: true
+                    contents: []
+                    discrete: false
+                    level: level1
+                nsys:
+                    _target_: verl.utils.profiler.config.NsightToolConfig
+                    discrete: false
+                torch:
+                    _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                    step_end: null
+                    step_start: 0
+                torch_memory:
+                    _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                    stack_depth: 32
+                    trace_alloc_max_entries: 100000
+        rollout_n: 1
+        shuffle: false
+        strategy: fsdp
+        ulysses_sequence_parallel_size: 1
+        use_dynamic_bsz: false
+ctx_manager:
+    value:
+        generation:
+            gen_config:
+                kwargs: null
+                response_length: 400
+                temperature: 1
+                top_k: -1
+                top_p: 1
+custom_envs:
+    value:
+        Bandit:
+            env_config:
+                split: train
+            env_instruction: ""
+            env_type: bandit
+            max_actions_per_traj: 1
+            max_tokens: 100
+        BanditTest:
+            env_config:
+                split: test
+            env_instruction: ""
+            env_type: bandit
+            max_actions_per_traj: 1
+            max_tokens: 100
+        CoordFrozenLake:
+            env_config:
+                observation_format: grid_coord
+            env_instruction: |
+                You are solving the FrozenLake puzzle. The observation includes both a symbol grid and zero-indexed coordinates for the start, goal, player, and any holes.
+                Coordinates range from the top-left corner (0, 0) to the bottom-right corner (5, 5).
+                Beware that the ice is slippery, so the agent might slide and end up in an unintended tile.
+                Respond with a sequence of actions such as <answer>Left || Up || Up</answer>.
+            env_type: frozen_lake
+            max_actions_per_traj: 25
+            max_tokens: 120
+        CoordSokoban:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                max_steps: 100
+                num_boxes: 2
+                observation_format: grid_coord
+            env_instruction: "You are solving the Sokoban puzzle. You are the player and you need to push all boxes to targets.\nYou are provided with a symbol grid and the zero-indexed coordinates of the player, each box, and each target. \nCoordinates range from the top-left corner (0, 0) to the bottom-right corner (5, 5). \nWhen you are exactly next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box.\nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>.\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 120
+        Countdown:
+            env_config: null
+            env_instruction: 'You are solving the Countdown puzzle. You should use the num list to create an equation that equals the target. Example answer format: <think> To find an equation using [3, 5, 2] to get 4. Let''s check 2 + 5 = 7, 7 - 3 = 4. So the answer is 2 + 5 - 3 = 4. </think><answer>2 + 5 - 3</answer>'
+            env_type: countdown
+            max_actions_per_traj: 1
+            max_tokens: 100
+        FrozenLake:
+            env_config: null
+            env_instruction: 'You are solving the FrozenLake puzzle. Forbid the whole and go to the target. You may move to the unintended direction due to the slippery ice. Example answer format: <think>To forbid the hole and go to the target, I should go left then go up.</think><answer>Left || Up</answer>'
+            env_type: frozen_lake
+            max_actions_per_traj: 25
+            max_tokens: 100
+        LargerSokoban:
+            env_config:
+                dim_x: 8
+                dim_y: 8
+                max_steps: 100
+                num_boxes: 2
+                search_depth: 10
+            env_instruction: "You are solving the Sokoban puzzle. \nYou are the player and you need to push all boxes to targets. \nWhen you are right next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box. \nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        Lean:
+            env_config: null
+            env_instruction: You are a Lean theorem prover. Given a Lean theorem statement, propose a sequence of tactics that completes the proof. Think step by step about which tactics to apply next. Provide tactics separated by '||', for example <answer>intro || simp || rfl</answer>.
+            env_type: lean
+            max_actions_per_traj: 30
+            max_tokens: 512
+        MediumSudoku:
+            env_config:
+                difficulty: medium
+                grid_size: 9
+                max_steps: 81
+                render_format: with_feedback
+                show_conflicts: true
+                show_valid_numbers: true
+            env_instruction: |
+                You are solving a Sudoku puzzle. Fill in the grid so that every row, column, and 3x3 box contains the numbers 1-9 without repetition.
+                Initial cells are shown in [brackets] and cannot be modified. Empty cells are shown as dots (.).
+                Place numbers one at a time using the format: <answer>place 5 at row 2 col 3</answer> or <answer>2,3,5</answer>
+                The environment will provide feedback on valid/invalid moves and show conflicts if any occur.
+            env_type: sudoku
+            max_actions_per_traj: 30
+            max_tokens: 150
+            max_workers: 32
+            parallel_friendly: false
+        MetamathQA:
+            env_config: null
+            env_instruction: 'You are solving Math problems. '
+            env_type: metamathqa
+            max_actions_per_traj: 1
+            max_tokens: 100
+        SimpleSokoban:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                max_steps: 100
+                num_boxes: 2
+            env_instruction: "You are solving the Sokoban puzzle. \nYou are the player and you need to push all boxes to targets. \nWhen you are right next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box. \nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        SimpleSudoku:
+            env_config:
+                difficulty: easy
+                grid_size: 4
+                max_steps: 20
+                render_format: with_feedback
+                show_conflicts: false
+                show_valid_numbers: false
+            env_instruction: |
+                You are solving a Sudoku puzzle. Fill in the grid so that every row, column, and 2x2 box contains the numbers 1-4 without repetition.
+                Initial cells are shown in [brackets] and cannot be modified. Empty cells are shown as dots (.).
+                Place numbers one at a time using the format, for example: <answer>place 1 at row 2 col 3</answer> or <answer>1,2,3</answer>
+                The environment will provide feedback on valid/invalid moves and show conflicts if any occur.
+            env_type: sudoku
+            max_actions_per_traj: 20
+            max_tokens: 150
+            max_workers: 32
+            parallel_friendly: false
+        SokobanDifferentGridVocab:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                grid_lookup:
+                    "0": W
+                    "1": .
+                    "2": G
+                    "3": C
+                    "4": B
+                    "5": A
+                    "6": '@'
+                grid_vocab:
+                    .: empty
+                    '@': player on target
+                    A: player
+                    B: box
+                    C: box on target
+                    G: target
+                    W: wall
+                max_steps: 100
+                num_boxes: 1
+                search_depth: 30
+            env_instruction: "You are solving the Sokoban puzzle. \nYou are the player and you need to push all boxes to targets. \nWhen you are right next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box. \nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        VisualSimpleSokoban:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                max_steps: 100
+                num_boxes: 1
+                render_mode: rgb_array
+            env_instruction: You are solving the Sokoban puzzle. You are the player and you need to push all boxes to targets. When you are right next to a box, you can push it by moving in the same direction. You cannot push a box through a wall, and you cannot pull a box. The answer should be a sequence of actions, like <answer>Right || Right || Up</answer>
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        WebShop:
+            env_config:
+                dataset: small
+            env_instruction: 'You are browsing an online shop. Based on the instruction, buy a product that close to the production description. You need to search, read the search results, pick a product, choose the size and color and buy. You should only choose action from the available actions list provided later.  Example process: I need a gingko light and 20x20 pillow cover that is hand painted. First search[gingko light 20x20 pillow cover hand painted], answer format: <answer>search[blanket with fleece throw]</answer>. Valid answer is search[<keywords>] or click[<clickable>].'
+            env_type: webshop
+            max_actions_per_traj: 9
+            max_tokens: 200
+        blackjack:
+            env_config: null
+            env_instruction: |
+                You are playing Blackjack against a dealer. The dealer must hit on 16 or less and stand on 17 or more.
+                Choose either Stick or Hit. Respond with a single action.
+                Example: <answer>Hit</answer>
+            env_type: blackjack
+            max_actions_per_traj: 10
+            max_tokens: 64
+        game_2048:
+            env_config: null
+            env_instruction: |
+                You are playing the 2048 game on a 4x4 grid. Merge equal tiles by sliding Up, Right, Down, or Left.
+                If a move is invalid (no tiles move), a small penalty is applied. Respond with a single action.
+                Example: <answer>Up</answer>
+            env_type: game_2048
+            max_actions_per_traj: 700
+            max_tokens: 8192
+        rubikscube:
+            env_config:
+                max_steps: 20
+                render_mode: text
+                scramble_depth: 1
+            env_instruction: |
+                You are solving a 2x2 Rubik's Cube (Pocket Cube). The goal is to restore the cube so that each of the faces consists of a single, unique color.
+                Available actions use standard Singmaster notation for face rotations: U, U', D, D', L, L', R, R', F, F', B, B'.
+                - Faces: U (Up), D (Down), L (Left), R (Right), F (Front), B (Back).
+                - Modifiers: A letter alone means 90° clockwise (e.g., 'R'). A letter with prime (') means 90° counter-clockwise (e.g., "R'").
+                Respond with a sequence of actions separated by "||".
+                Example: <answer>U</answer>
+            env_type: rubikscube
+            max_actions_per_traj: 20
+            max_tokens: 96
+custom_reward_function:
+    value:
+        name: compute_score
+        path: null
+data:
+    value:
+        custom_cls:
+            name: null
+            path: null
+        datagen:
+            name: null
+            path: null
+        dataloader_num_workers: 8
+        filter_overlong_prompts: false
+        filter_overlong_prompts_workers: 1
+        image_key: images
+        max_prompt_length: null
+        max_response_length: null
+        prompt_key: prompt
+        return_full_prompt: false
+        return_multi_modal_inputs: true
+        return_raw_chat: false
+        return_raw_input_ids: false
+        reward_fn_key: data_source
+        sampler:
+            class_name: null
+            class_path: null
+        shuffle: true
+        tokenizer: null
+        train_batch_size: 128
+        train_files: ~/data/rlhf/gsm8k/train.parquet
+        truncation: error
+        trust_remote_code: false
+        use_shm: false
+        val_batch_size: null
+        val_files: ~/data/rlhf/gsm8k/test.parquet
+        validation_shuffle: false
+        video_key: videos
+enable_response_mask:
+    value: true
+es_manager:
+    value:
+        format_penalty: -0.1
+        train:
+            env_configs:
+                n_groups:
+                    - 8
+                tags:
+                    - CoordFrozenLake
+            env_groups: 8
+            group_size: 16
+        val:
+            env_configs:
+                n_groups:
+                    - 256
+                tags:
+                    - CoordFrozenLake
+            env_groups: 256
+            group_size: 1
+global_profiler:
+    value:
+        _target_: verl.utils.profiler.ProfilerConfig
+        global_tool_config:
+            nsys:
+                _target_: verl.utils.profiler.config.NsightToolConfig
+                controller_nsight_options:
+                    cuda-graph-trace: graph
+                    cuda-memory-usage: "true"
+                    trace: cuda,nvtx,cublas,ucx
+                discrete: false
+                worker_nsight_options:
+                    capture-range: cudaProfilerApi
+                    capture-range-end: null
+                    cuda-graph-trace: graph
+                    cuda-memory-usage: "true"
+                    kill: none
+                    trace: cuda,nvtx,cublas,ucx
+            torch_memory:
+                context: all
+                stack_depth: 32
+                stacks: all
+                trace_alloc_max_entries: 100000
+        profile_continuous_steps: false
+        save_path: outputs/profile
+        steps: null
+        tool: null
+grpo_advantage_length_weight:
+    value: false
+lora:
+    value:
+        alpha: 16
+        rank: 0
+        target_modules: all-linear
+micro_batch_size_per_gpu:
+    value: 1
+model_path:
+    value: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+ppo_mini_batch_size:
+    value: 32
+ray_kwargs:
+    value:
+        ray_init:
+            num_cpus: null
+        timeline_json_file: null
+reward_model:
+    value:
+        enable: false
+        enable_resource_pool: false
+        forward_max_token_len_per_gpu: 32768
+        launch_reward_fn_async: false
+        max_length: null
+        micro_batch_size: null
+        micro_batch_size_per_gpu: null
+        model:
+            external_lib: null
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                forward_prefetch: false
+                fsdp_size: -1
+                param_offload: false
+                reshard_after_forward: true
+                wrap_policy:
+                    min_num_params: 0
+            input_tokenizer: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            path: ~/models/FsfairX-LLaMA3-RM-v0.1
+            trust_remote_code: false
+            use_fused_kernels: false
+            use_remove_padding: false
+            use_shm: false
+        n_gpus_per_node: 0
+        nnodes: 0
+        profiler:
+            _target_: verl.utils.profiler.ProfilerConfig
+            all_ranks: false
+            enable: false
+            ranks: []
+            save_path: outputs/profile
+            tool: null
+            tool_config:
+                npu:
+                    _target_: verl.utils.profiler.config.NPUToolConfig
+                    analysis: true
+                    contents: []
+                    discrete: false
+                    level: level1
+                nsys:
+                    _target_: verl.utils.profiler.config.NsightToolConfig
+                    discrete: false
+                torch:
+                    _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                    step_end: null
+                    step_start: 0
+                torch_memory:
+                    _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                    stack_depth: 32
+                    trace_alloc_max_entries: 100000
+        reward_manager: naive
+        sandbox_fusion:
+            max_concurrent: 64
+            memory_limit_mb: 1024
+            url: null
+        strategy: fsdp
+        ulysses_sequence_parallel_size: 1
+        use_dynamic_bsz: false
+seed:
+    value:
+        train: 10000
+        val: 123
+system:
+    value:
+        CUDA_VISIBLE_DEVICES: 0,1,2,3,4,5,6,7
+trainer:
+    value:
+        balance_batch: true
+        critic_warmup: 0
+        default_hdfs_dir: null
+        default_local_dir: /mnt/general/wanghy/RAGEN/saves/
+        del_local_ckpt_after_load: false
+        device: cuda
+        esi_redundant_time: 0
+        experiment_name: frozen_lake-starpo-slippery
+        generations_to_log_to_wandb:
+            train: 128
+            val: 20
+        local_log_dir: results/
+        log_val_generations: 0
+        logger:
+            - console
+            - wandb
+        max_actor_ckpt_to_keep: 1
+        max_critic_ckpt_to_keep: 1
+        n_gpus_per_node: 8
+        nnodes: 1
+        project_name: ragen_latest_llama_1B_it
+        ray_wait_register_center_timeout: 300
+        resume_from_path: null
+        resume_mode: auto
+        rollout_data_dir: null
+        save_freq: -1
+        test_freq: 10
+        total_epochs: 30
+        total_training_steps: 200
+        use_legacy_worker_impl: auto
+        val_before_train: true
+        val_only: false
+        validation_data_dir: null
+        validation_steps: 1

wandb/run-20251222_203657-pjvd64ya/files/requirements.txt

@@ -0,0 +1,316 @@
+ragen==0.1
+ragen==0.1
+colorama==0.4.6
+psutil==7.1.2
+ragen==0.1
+setuptools==80.9.0
+wheel==0.45.1
+pip==25.3
+ragen==0.1
+pytz==2025.2
+nvidia-cusparselt-cu12==0.6.2
+mpmath==1.3.0
+typing_extensions==4.15.0
+sympy==1.13.1
+pylatexenc==2.10
+nvidia-cusolver-cu12==11.6.1.9
+numpy==1.26.4
+Markdown==3.10
+absl-py==2.3.1
+pyasn1_modules==0.4.2
+Werkzeug==3.1.3
+networkx==3.3
+MarkupSafe==2.1.5
+fsspec==2025.9.0
+filelock==3.19.1
+nvidia-cudnn-cu12==9.1.0.70
+Jinja2==3.1.6
+tensorboard==2.20.0
+triton==3.2.0
+pillow==11.3.0
+xxhash==3.6.0
+nvidia-nvtx-cu12==12.4.127
+nvidia-nvjitlink-cu12==12.4.127
+nvidia-nccl-cu12==2.21.5
+nvidia-curand-cu12==10.3.5.147
+wrapt==2.0.1
+nvidia-cufft-cu12==11.2.1.3
+pyjnius==1.7.0
+nvidia-cuda-runtime-cu12==12.4.127
+nvidia-cuda-nvrtc-cu12==12.4.127
+websockets==15.0.1
+nvidia-cuda-cupti-cu12==12.4.127
+zipp==3.23.0
+nvidia-cublas-cu12==12.4.5.8
+nvidia-cusparse-cu12==12.3.1.170
+flash_attn==2.8.3
+rsa==4.9.1
+einops==0.8.1
+py-cpuinfo==9.0.0
+pure_eval==0.2.3
+ptyprocess==0.7.0
+gym-notices==0.1.0
+flatbuffers==25.9.23
+fastrlock==0.8.3
+Farama-Notifications==0.0.4
+faiss-gpu==1.7.2
+antlr4-python3-runtime==4.9.3
+wcwidth==0.2.14
+wasabi==1.1.3
+uvloop==0.22.1
+urllib3==2.5.0
+tzdata==2025.2
+typing-inspection==0.4.2
+traitlets==5.14.3
+tqdm==4.67.1
+tomli==2.3.0
+threadpoolctl==3.6.0
+tenacity==9.1.2
+tabulate==0.9.0
+spacy-loggers==1.0.5
+spacy-legacy==3.0.12
+soupsieve==2.8
+sniffio==1.3.1
+smmap==5.0.2
+six==1.17.0
+shellingham==1.5.4
+sentencepiece==0.2.1
+scipy==1.15.3
+safetensors==0.6.2
+rpds-py==0.29.0
+rignore==0.7.6
+regex==2025.11.3
+RapidFuzz==3.14.3
+rank-bm25==0.2.2
+pyzmq==27.1.0
+PyYAML==6.0.3
+pytokens==0.3.0
+python-multipart==0.0.20
+python-json-logger==4.0.0
+python-dotenv==1.2.1
+PySocks==1.7.1
+pyparsing==3.2.5
+PyJWT==2.10.1
+Pygments==2.19.2
+pygame==2.6.1
+pydantic_core==2.41.5
+pycparser==2.23
+pycountry==24.6.1
+pybind11==3.0.1
+pyarrow==22.0.0
+psutil==7.1.3
+protobuf==4.25.8
+propcache==0.4.1
+prometheus_client==0.23.1
+platformdirs==4.5.0
+pexpect==4.9.0
+pathspec==0.12.1
+partial-json-parser==0.2.1.1.post6
+parso==0.8.5
+packaging==25.0
+orjson==3.11.4
+opentelemetry-semantic-conventions-ai==0.4.13
+opencv-python-headless==4.12.0.88
+ninja==1.13.0
+nest-asyncio==1.6.0
+mypy_extensions==1.1.0
+murmurhash==1.0.15
+multidict==6.7.0
+msgspec==0.19.0
+msgpack==1.1.2
+mdurl==0.1.2
+loguru==0.7.3
+llvmlite==0.44.0
+llguidance==0.7.30
+lark==1.2.2
+kiwisolver==1.4.9
+joblib==1.5.2
+jiter==0.12.0
+itsdangerous==2.2.0
+interegular==0.3.3
+ImageIO==2.37.2
+idna==3.11
+humanfriendly==10.0
+httpx-sse==0.4.3
+httptools==0.7.1
+html2text==2025.4.15
+hf-xet==1.2.0
+h11==0.16.0
+grpcio==1.76.0
+frozenlist==1.8.0
+fonttools==4.60.1
+executing==2.2.1
+exceptiongroup==1.3.0
+eval_type_backport==0.2.2
+docstring_parser==0.17.0
+dnspython==2.8.0
+distro==1.9.0
+diskcache==5.6.3
+dill==0.4.0
+decorator==5.2.1
+debugpy==1.8.17
+Cython==3.2.1
+cymem==2.0.13
+cycler==0.12.1
+cupy-cuda12x==13.6.0
+contourpy==1.3.2
+colorama==0.4.6
+codetiming==1.4.0
+cloudpickle==3.1.2
+cloudpathlib==0.23.0
+click==8.3.1
+charset-normalizer==3.4.4
+certifi==2025.11.12
+catalogue==2.0.10
+cachetools==6.2.2
+blis==1.3.2
+blinker==1.9.0
+blake3==1.0.8
+attrs==25.4.0
+async-timeout==5.0.1
+asttokens==3.0.1
+astor==0.8.1
+annotated-types==0.7.0
+annotated-doc==0.0.4
+airportsdata==20250909
+aiohappyeyeballs==2.6.1
+yarl==1.22.0
+uvicorn==0.38.0
+typer-slim==0.20.0
+thefuzz==0.22.1
+stack-data==0.6.3
+srsly==2.5.1
+smart_open==7.5.0
+sentry-sdk==2.44.0
+scikit-learn==1.7.2
+requests==2.32.5
+referencing==0.37.0
+python-dateutil==2.9.0.post0
+pydantic==2.12.4
+prompt_toolkit==3.0.52
+preshed==3.0.11
+opentelemetry-proto==1.26.0
+omegaconf==2.3.0
+numba==0.61.2
+nltk==3.9.2
+multiprocess==0.70.18
+matplotlib-inline==0.2.1
+markdown-it-py==4.0.0
+jedi==0.19.2
+importlib_metadata==8.0.0
+httpcore==1.0.9
+gymnasium==1.2.2
+gym==0.26.2
+googleapis-common-protos==1.72.0
+gitdb==4.0.12
+gguf==0.17.1
+Flask==3.1.2
+faiss-cpu==1.13.0
+email-validator==2.3.0
+depyf==0.18.0
+Deprecated==1.3.1
+coloredlogs==15.0.1
+cffi==2.0.0
+black==25.11.0
+beautifulsoup4==4.14.2
+anyio==4.11.0
+aiosignal==1.4.0
+watchfiles==1.1.1
+tiktoken==0.12.0
+starlette==0.49.3
+sse-starlette==3.0.3
+rich==14.2.0
+pydantic-settings==2.12.0
+pydantic-extra-types==2.10.6
+pandas==2.3.3
+opentelemetry-exporter-otlp-proto-common==1.26.0
+onnxruntime==1.23.2
+matplotlib==3.10.7
+lm-format-enforcer==0.10.12
+jsonschema-specifications==2025.9.1
+ipython==8.37.0
+hydra-core==1.3.2
+huggingface-hub==0.36.0
+httpx==0.28.1
+gym-sokoban==0.0.6
+GitPython==3.1.45
+cryptography==46.0.3
+confection==0.1.5
+cleantext==1.1.4
+aiohttp==3.13.2
+weasel==0.4.3
+wandb==0.23.0
+typer==0.19.2
+virtualenv==20.35.4
+pyasn1==0.6.1
+torch==2.6.0
+tokenizers==0.22.1
+thinc==8.3.9
+rich-toolkit==0.15.1
+prometheus-fastapi-instrumentator==7.1.0
+openai==2.8.0
+jsonschema==4.25.1
+gdown==5.2.0
+fastapi==0.121.2
+anthropic==0.73.0
+xformers==0.0.29.post2
+transformers==4.57.1
+proto-plus==1.26.1
+torchvision==0.21.0
+torchdata==0.11.0
+torchaudio==2.6.0
+together==1.5.30
+tensordict==0.8.3
+spacy==3.8.9
+ray==2.51.1
+outlines_core==0.1.26
+mistral_common==1.8.5
+mcp==1.21.1
+fastapi-cloud-cli==0.3.1
+fastapi-cli==0.0.16
+datasets==4.4.1
+accelerate==1.11.0
+xgrammar==0.1.18
+pyserini==1.2.0
+peft==0.18.0
+outlines==0.1.11
+opentelemetry-exporter-otlp-proto-http==1.26.0
+opentelemetry-exporter-otlp-proto-grpc==1.26.0
+kimina-client==0.2.1
+compressed-tensors==0.9.3
+opentelemetry-exporter-otlp==1.26.0
+vllm==0.8.5
+py-spy==0.4.1
+opencensus-context==0.1.3
+distlib==0.4.0
+colorful==0.5.8
+tensorboard-data-server==0.7.2
+google-auth==2.43.0
+google-api-core==2.28.1
+aiohttp-cors==0.8.1
+opentelemetry-exporter-prometheus==0.59b0
+opencensus==0.11.4
+verl==0.5.0.dev0
+opentelemetry-api==1.38.0
+opentelemetry-semantic-conventions==0.59b0
+opentelemetry-sdk==1.38.0
+typeguard==4.4.4
+shtab==1.8.0
+tyro==0.9.35
+verl==0.5.0.dev0
+autocommand==2.2.2
+backports.tarfile==1.2.0
+importlib_metadata==8.0.0
+inflect==7.3.1
+jaraco.collections==5.1.0
+jaraco.context==5.3.0
+jaraco.functools==4.0.1
+jaraco.text==3.12.1
+more-itertools==10.3.0
+packaging==24.2
+platformdirs==4.2.2
+tomli==2.0.1
+typeguard==4.3.0
+typing_extensions==4.12.2
+wheel==0.45.1
+zipp==3.19.2

wandb/run-20251222_203657-pjvd64ya/files/wandb-metadata.json

@@ -0,0 +1,107 @@
+{
+  "os": "Linux-5.14.0-284.25.1.el9_2.x86_64-x86_64-with-glibc2.35",
+  "python": "CPython 3.10.19",
+  "startedAt": "2025-12-22T12:36:57.314412Z",
+  "args": [
+    "--node-ip-address=10.119.27.78",
+    "--node-manager-port=33093",
+    "--object-store-name=/tmp/ray/session_2025-12-22_20-35-38_490044_519970/sockets/plasma_store",
+    "--raylet-name=/tmp/ray/session_2025-12-22_20-35-38_490044_519970/sockets/raylet",
+    "--redis-address=None",
+    "--metrics-agent-port=64448",
+    "--logging-rotate-bytes=536870912",
+    "--logging-rotate-backup-count=5",
+    "--runtime-env-agent-port=61324",
+    "--gcs-address=10.119.27.78:59659",
+    "--session-name=session_2025-12-22_20-35-38_490044_519970",
+    "--temp-dir=/tmp/ray",
+    "--webui=127.0.0.1:8265",
+    "--cluster-id=0dc698b87906cc59debdc15a5634e568861d5234a9cc80ce98248a23",
+    "--startup-token=112",
+    "--worker-launch-time-ms=1766406950439",
+    "--node-id=6fb8b54e88b3c54a306633a87e28c055cbbe44a1a5b5852b0d26078f",
+    "--runtime-env-hash=-839022692"
+  ],
+  "program": "/root/local/miniconda3/envs/ragen/lib/python3.10/site-packages/ray/_private/workers/default_worker.py",
+  "git": {
+    "remote": "https://github.com/mll-lab-nu/RAGEN.git",
+    "commit": "8d73639b99b38265453f898b8d6af7d4af50d56e"
+  },
+  "email": "haoyu-wa22@mails.tsinghua.edu.cn",
+  "root": "/mnt/general/wanghy/RAGEN",
+  "host": "pt-4d654cf4576f4d23ad3d3919f12932fe-worker-0",
+  "executable": "/root/local/miniconda3/envs/ragen/bin/python",
+  "cpu_count": 64,
+  "cpu_count_logical": 128,
+  "gpu": "NVIDIA H100 80GB HBM3",
+  "gpu_count": 8,
+  "disk": {
+    "/": {
+      "total": "5153960755200",
+      "used": "471741583360"
+    }
+  },
+  "memory": {
+    "total": "2163642122240"
+  },
+  "gpu_nvidia": [
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-35e2d43d-4067-82ce-90d4-def9e389bf28"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-af4135e3-88f2-e9ac-518d-502c75a85429"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-d7fdeeba-fe9b-ec03-d9f7-6724fe4266b5"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-ccc4f668-3882-5a8e-2c07-c5cd08f6f666"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-7b73c0cf-d3d5-e10c-7176-a43be1e41001"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-81b58d94-5d1f-8ec2-f9d2-fd56172ed177"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-03e8bc66-3b44-6794-49fd-5392fbdda6d1"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-b7cf0ec6-7c29-1179-dceb-09565da51890"
+    }
+  ],
+  "cudaVersion": "12.4",
+  "writerId": "vece8z0wtyrd6guwhtkw2pg8e4uzzajh"
+}

wandb/run-20251222_203657-pjvd64ya/files/wandb-summary.json

@@ -0,0 +1 @@
+{"val-aux/unknown/reward/best@32/std":0.03206727325668882,"response_length_non_aborted/min":341,"val-core/unknown/reward/best@256/std":0,"critic/rewards/mean":-1.0812499523162842,"val-aux/unknown/reward/worst@256/std":0,"train/CoordFrozenLake/num_actions":2.7109375,"rollout/chosen_in_group_max":1,"critic/returns/min":-2.5,"train/CoordFrozenLake/non-zero/action_is_effective":0.592994388149771,"rollout/chosen_in_group_reward_max":1,"response_length_non_aborted/mean":4889.15625,"actor/grad_norm":31.720598220825195,"_step":1,"val-aux/unknown/reward/best@4/std":0.657955346810967,"actor/ppo_kl":0,"val-env/CoordFrozenLake/non-zero/pass@1":1,"actor/pg_loss":-0.19302269211038947,"val-aux/unknown/reward/worst@16/mean":-2.4975,"val-env/CoordFrozenLake/action_is_valid":0.5985451406689071,"val-aux/unknown/reward/worst@8/mean":-2.44549999987334,"timing_per_token_ms/update_actor":0.011858465277438484,"response_length_non_aborted/max":8000,"val-env/CoordFrozenLake/non-zero/num_actions":10.80379746835443,"critic/vf_clipfrac":0,"train/CoordFrozenLake/success":0.25,"val-aux/unknown/reward/best@64/mean":1,"val-aux/unknown/reward/worst@256/mean":-2.5,"val-env/CoordFrozenLake/episodic_return":0,"val-aux/unknown/reward/best@32/mean":0.9986999999284745,"rollout/chosen_in_group_mean":-1.0812499523162842,"timing_s/update_actor":1.8554110527038574,"critic/vf_loss":1.2944875285029411,"response/aborted_ratio":0,"actor/pg_clipfrac":0,"critic/returns/max":1,"val-aux/unknown/reward/best@2/mean":-0.07910000048577785,"critic/values/min":-6.03125,"critic/advantages/max":9.022233963012695,"val-env/CoordFrozenLake/action_is_effective":0.485667827494044,"val-env/CoordFrozenLake/success":0.1875,"critic/returns/mean":-2.400373697280884,"_timestamp":1.7664071744306018e+09,"val-aux/unknown/reward/worst@4/mean":-2.161399999588728,"val-aux/unknown/reward/worst@8/std":0.35625798245775797,"perf/cpu_memory_used_gb":103.15261459350586,"critic/grad_norm":343.27099609375,"perf/total_num_tokens":156463,"critic/advantages/mean":-2.9771680942758394e-07,"global_seqlen/balanced_min":17383,"global_seqlen/min":9398,"timing_per_token_ms/adv":0.0020857759500862924,"critic/lr":1e-05,"train/CoordFrozenLake/episodic_return":0,"rollout/ref_log_prob":-9.413333892822266,"val-aux/unknown/reward/best@8/mean":0.8094999986886978,"train/charts/avg_episode_return":0,"val-aux/unknown/reward/worst@64/std":0,"train/response_length":2703.71875,"train/CoordFrozenLake/non-zero/pass@16":1,"prompt_length/mean":0.3125,"critic/values/max":9.1875,"perf/mfu/critic":0.19184528346404042,"train/CoordFrozenLake/pass@16":1,"val-core/unknown/reward/best@256/mean":1,"rollout/in_group_reward_std":1.4240182638168335,"val-aux/unknown/reward/worst@32/std":0,"val-aux/unknown/reward/worst@128/mean":-2.5,"val-aux/unknown/reward/worst@16/std":0.07901740314639555,"critic/rewards/min":-2.5,"global_seqlen/minmax_diff":15779,"val-aux/unknown/reward/worst@2/mean":-1.508399999655783,"val-env/charts/avg_episode_return":0,"perf/time_per_step":78.07747589796782,"perf/mfu/actor":0.12817975413308264,"timing_s/ref":1.0796662643551826,"response_length/max":8000,"perf/throughput":250.49317713034634,"val-aux/unknown/reward/best@128/std":0,"timing_s/step":78.07747589796782,"rollout/in_group_reward_mean":-1.3992186784744263,"train/CoordFrozenLake/non-zero/action_is_valid":0.7137810502989075,"val-aux/unknown/reward/worst@64/mean":-2.5,"rollout/in_group_max":0.987500011920929,"val/generations":{"path":"media/table/val/generations_0_594e349d07b7ffe22904.table.json","nrows":1,"ncols":61,"sha256":"594e349d07b7ffe2290462d6785cc4e4ca9309c1dd98c69dc3daf56490e5873e","log_mode":"IMMUTABLE","size":227684,"_latest_artifact_path":"wandb-client-artifact://h22dnvryh5c3wtlik3fzw91r8y1kodk9mb6r4yogjs3pxqvj5x1z0yblou1m38f2n4nuzpn0nkseit0i8zz2l3jyrkp8a3azt93y9ri2eurgx98vkkhdy1j9sa5q7ghr:latest/val/generations.table.json","_type":"table-file","artifact_path":"wandb-client-artifact://uigmf1gfgahd330fe0u02mpw5drld9hxr5u1ry0uz50yl30m5qu4nkwpfow245ha7d1y8rvln2823zxgu6at7rs1okmv2ob9n6r3pnp7ylswvca0ea6n7mipm49zb3ai/val/generations.table.json"},"actor/entropy":9.3908052444458,"response_length_non_aborted/clip_ratio":0.34375,"critic/vf_explained_var":-3.5359854698181152,"critic/score/max":1,"actor/lr":1e-06,"response_length/clip_ratio":0.34375,"timing_per_token_ms/values":0.002038286004694548,"timing_s/values":0.31891634315252304,"actor/pg_clipfrac_lower":0,"val-env/CoordFrozenLake/non-zero/success":1,"prompt_length/max":1,"_wandb":{"runtime":210},"perf/max_memory_allocated_gb":76.34673595428467,"val-aux/unknown/reward/worst@128/std":0,"val-aux/unknown/reward/best@128/mean":1,"timing_s/gen":72.03112238645554,"train/CoordFrozenLake/action_is_valid":0.31227920950577204,"global_seqlen/mean":19557.875,"prompt_length/min":0,"critic/score/mean":-1.0812499523162842,"val-env/CoordFrozenLake/num_actions":6.66796875,"timing_per_token_ms/update_critic":0.007993092987621086,"train/CoordFrozenLake/non-zero/success":1,"val-aux/unknown/reward/best@16/mean":0.9716999994516373,"train/CoordFrozenLake/non-zero/num_actions":5.507936507936508,"val-aux/unknown/reward/best@16/std":0.15946507479268063,"rollout/chosen_in_group_reward_std":1.7277626991271973,"val-env/CoordFrozenLake/pass@1":0.1875,"critic/values/mean":1.0937836170196533,"perf/max_memory_reserved_gb":84.759765625,"timing_s/adv":0.3263467624783516,"response_length/mean":4889.15625,"val-env/CoordFrozenLake/non-zero/action_is_effective":0.7919169671240464,"rollout/in_group_mean":-1.3992186784744263,"global_seqlen/max":25177,"rollout/in_group_reward_max":0.987500011920929,"val-aux/unknown/reward/std@256":1.410370480957647,"val-env/CoordFrozenLake/non-zero/action_is_valid":0.9759716943391096,"timing_per_token_ms/gen":0.4604010302548084,"val-aux/unknown/reward/best@64/std":0,"val-core/unknown/reward/mean@256":-0.8085937499126885,"timing_s/old_log_prob":1.2117513045668602,"val-aux/unknown/reward/worst@4/std":0.8475907265562409,"val-aux/unknown/reward/worst@2/std":1.2613997930657996,"train/CoordFrozenLake/raw_reward":0,"timing_s/update_critic":1.250623308122158,"critic/score/min":-2.5,"response_length/min":341,"critic/vpred_mean":1.214583158493042,"_runtime":210.172523788,"rollout/chosen_in_group_std":1.7277626991271973,"val-aux/unknown/reward/worst@32/mean":-2.5,"critic/rewards/max":1,"prompt_length/clip_ratio":0.3125,"timing_s/total":78.08085680007935,"rollout/chosen_in_group_reward_mean":-1.0812499523162842,"val-aux/unknown/reward/best@2/std":1.1311556860113463,"val-env/CoordFrozenLake/raw_reward":0,"global_seqlen/balanced_max":23621,"rollout/in_group_std":1.4240182638168335,"critic/advantages/min":-8.670817375183105,"train/CoordFrozenLake/action_is_effective":0.24553673884326457,"val-aux/unknown/reward/best@4/mean":0.4925999987050891,"val-aux/unknown/reward/best@8/std":0.38634149371075105,"timing_per_token_ms/ref":0.006900457388361355,"val-env/response_length":1397.296875}

wandb/run-20251222_203657-pjvd64ya/run-pjvd64ya.wandb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2ff1c6ea711407debefce4ba6552e54d743f912a557fffe3ea83da103d64414b
+size 108386

wandb/run-20251222_204216-q200utyn/files/requirements.txt

@@ -0,0 +1,316 @@
+ragen==0.1
+ragen==0.1
+colorama==0.4.6
+psutil==7.1.2
+ragen==0.1
+setuptools==80.9.0
+wheel==0.45.1
+pip==25.3
+ragen==0.1
+pytz==2025.2
+nvidia-cusparselt-cu12==0.6.2
+mpmath==1.3.0
+typing_extensions==4.15.0
+sympy==1.13.1
+pylatexenc==2.10
+nvidia-cusolver-cu12==11.6.1.9
+numpy==1.26.4
+Markdown==3.10
+absl-py==2.3.1
+pyasn1_modules==0.4.2
+Werkzeug==3.1.3
+networkx==3.3
+MarkupSafe==2.1.5
+fsspec==2025.9.0
+filelock==3.19.1
+nvidia-cudnn-cu12==9.1.0.70
+Jinja2==3.1.6
+tensorboard==2.20.0
+triton==3.2.0
+pillow==11.3.0
+xxhash==3.6.0
+nvidia-nvtx-cu12==12.4.127
+nvidia-nvjitlink-cu12==12.4.127
+nvidia-nccl-cu12==2.21.5
+nvidia-curand-cu12==10.3.5.147
+wrapt==2.0.1
+nvidia-cufft-cu12==11.2.1.3
+pyjnius==1.7.0
+nvidia-cuda-runtime-cu12==12.4.127
+nvidia-cuda-nvrtc-cu12==12.4.127
+websockets==15.0.1
+nvidia-cuda-cupti-cu12==12.4.127
+zipp==3.23.0
+nvidia-cublas-cu12==12.4.5.8
+nvidia-cusparse-cu12==12.3.1.170
+flash_attn==2.8.3
+rsa==4.9.1
+einops==0.8.1
+py-cpuinfo==9.0.0
+pure_eval==0.2.3
+ptyprocess==0.7.0
+gym-notices==0.1.0
+flatbuffers==25.9.23
+fastrlock==0.8.3
+Farama-Notifications==0.0.4
+faiss-gpu==1.7.2
+antlr4-python3-runtime==4.9.3
+wcwidth==0.2.14
+wasabi==1.1.3
+uvloop==0.22.1
+urllib3==2.5.0
+tzdata==2025.2
+typing-inspection==0.4.2
+traitlets==5.14.3
+tqdm==4.67.1
+tomli==2.3.0
+threadpoolctl==3.6.0
+tenacity==9.1.2
+tabulate==0.9.0
+spacy-loggers==1.0.5
+spacy-legacy==3.0.12
+soupsieve==2.8
+sniffio==1.3.1
+smmap==5.0.2
+six==1.17.0
+shellingham==1.5.4
+sentencepiece==0.2.1
+scipy==1.15.3
+safetensors==0.6.2
+rpds-py==0.29.0
+rignore==0.7.6
+regex==2025.11.3
+RapidFuzz==3.14.3
+rank-bm25==0.2.2
+pyzmq==27.1.0
+PyYAML==6.0.3
+pytokens==0.3.0
+python-multipart==0.0.20
+python-json-logger==4.0.0
+python-dotenv==1.2.1
+PySocks==1.7.1
+pyparsing==3.2.5
+PyJWT==2.10.1
+Pygments==2.19.2
+pygame==2.6.1
+pydantic_core==2.41.5
+pycparser==2.23
+pycountry==24.6.1
+pybind11==3.0.1
+pyarrow==22.0.0
+psutil==7.1.3
+protobuf==4.25.8
+propcache==0.4.1
+prometheus_client==0.23.1
+platformdirs==4.5.0
+pexpect==4.9.0
+pathspec==0.12.1
+partial-json-parser==0.2.1.1.post6
+parso==0.8.5
+packaging==25.0
+orjson==3.11.4
+opentelemetry-semantic-conventions-ai==0.4.13
+opencv-python-headless==4.12.0.88
+ninja==1.13.0
+nest-asyncio==1.6.0
+mypy_extensions==1.1.0
+murmurhash==1.0.15
+multidict==6.7.0
+msgspec==0.19.0
+msgpack==1.1.2
+mdurl==0.1.2
+loguru==0.7.3
+llvmlite==0.44.0
+llguidance==0.7.30
+lark==1.2.2
+kiwisolver==1.4.9
+joblib==1.5.2
+jiter==0.12.0
+itsdangerous==2.2.0
+interegular==0.3.3
+ImageIO==2.37.2
+idna==3.11
+humanfriendly==10.0
+httpx-sse==0.4.3
+httptools==0.7.1
+html2text==2025.4.15
+hf-xet==1.2.0
+h11==0.16.0
+grpcio==1.76.0
+frozenlist==1.8.0
+fonttools==4.60.1
+executing==2.2.1
+exceptiongroup==1.3.0
+eval_type_backport==0.2.2
+docstring_parser==0.17.0
+dnspython==2.8.0
+distro==1.9.0
+diskcache==5.6.3
+dill==0.4.0
+decorator==5.2.1
+debugpy==1.8.17
+Cython==3.2.1
+cymem==2.0.13
+cycler==0.12.1
+cupy-cuda12x==13.6.0
+contourpy==1.3.2
+colorama==0.4.6
+codetiming==1.4.0
+cloudpickle==3.1.2
+cloudpathlib==0.23.0
+click==8.3.1
+charset-normalizer==3.4.4
+certifi==2025.11.12
+catalogue==2.0.10
+cachetools==6.2.2
+blis==1.3.2
+blinker==1.9.0
+blake3==1.0.8
+attrs==25.4.0
+async-timeout==5.0.1
+asttokens==3.0.1
+astor==0.8.1
+annotated-types==0.7.0
+annotated-doc==0.0.4
+airportsdata==20250909
+aiohappyeyeballs==2.6.1
+yarl==1.22.0
+uvicorn==0.38.0
+typer-slim==0.20.0
+thefuzz==0.22.1
+stack-data==0.6.3
+srsly==2.5.1
+smart_open==7.5.0
+sentry-sdk==2.44.0
+scikit-learn==1.7.2
+requests==2.32.5
+referencing==0.37.0
+python-dateutil==2.9.0.post0
+pydantic==2.12.4
+prompt_toolkit==3.0.52
+preshed==3.0.11
+opentelemetry-proto==1.26.0
+omegaconf==2.3.0
+numba==0.61.2
+nltk==3.9.2
+multiprocess==0.70.18
+matplotlib-inline==0.2.1
+markdown-it-py==4.0.0
+jedi==0.19.2
+importlib_metadata==8.0.0
+httpcore==1.0.9
+gymnasium==1.2.2
+gym==0.26.2
+googleapis-common-protos==1.72.0
+gitdb==4.0.12
+gguf==0.17.1
+Flask==3.1.2
+faiss-cpu==1.13.0
+email-validator==2.3.0
+depyf==0.18.0
+Deprecated==1.3.1
+coloredlogs==15.0.1
+cffi==2.0.0
+black==25.11.0
+beautifulsoup4==4.14.2
+anyio==4.11.0
+aiosignal==1.4.0
+watchfiles==1.1.1
+tiktoken==0.12.0
+starlette==0.49.3
+sse-starlette==3.0.3
+rich==14.2.0
+pydantic-settings==2.12.0
+pydantic-extra-types==2.10.6
+pandas==2.3.3
+opentelemetry-exporter-otlp-proto-common==1.26.0
+onnxruntime==1.23.2
+matplotlib==3.10.7
+lm-format-enforcer==0.10.12
+jsonschema-specifications==2025.9.1
+ipython==8.37.0
+hydra-core==1.3.2
+huggingface-hub==0.36.0
+httpx==0.28.1
+gym-sokoban==0.0.6
+GitPython==3.1.45
+cryptography==46.0.3
+confection==0.1.5
+cleantext==1.1.4
+aiohttp==3.13.2
+weasel==0.4.3
+wandb==0.23.0
+typer==0.19.2
+virtualenv==20.35.4
+pyasn1==0.6.1
+torch==2.6.0
+tokenizers==0.22.1
+thinc==8.3.9
+rich-toolkit==0.15.1
+prometheus-fastapi-instrumentator==7.1.0
+openai==2.8.0
+jsonschema==4.25.1
+gdown==5.2.0
+fastapi==0.121.2
+anthropic==0.73.0
+xformers==0.0.29.post2
+transformers==4.57.1
+proto-plus==1.26.1
+torchvision==0.21.0
+torchdata==0.11.0
+torchaudio==2.6.0
+together==1.5.30
+tensordict==0.8.3
+spacy==3.8.9
+ray==2.51.1
+outlines_core==0.1.26
+mistral_common==1.8.5
+mcp==1.21.1
+fastapi-cloud-cli==0.3.1
+fastapi-cli==0.0.16
+datasets==4.4.1
+accelerate==1.11.0
+xgrammar==0.1.18
+pyserini==1.2.0
+peft==0.18.0
+outlines==0.1.11
+opentelemetry-exporter-otlp-proto-http==1.26.0
+opentelemetry-exporter-otlp-proto-grpc==1.26.0
+kimina-client==0.2.1
+compressed-tensors==0.9.3
+opentelemetry-exporter-otlp==1.26.0
+vllm==0.8.5
+py-spy==0.4.1
+opencensus-context==0.1.3
+distlib==0.4.0
+colorful==0.5.8
+tensorboard-data-server==0.7.2
+google-auth==2.43.0
+google-api-core==2.28.1
+aiohttp-cors==0.8.1
+opentelemetry-exporter-prometheus==0.59b0
+opencensus==0.11.4
+verl==0.5.0.dev0
+opentelemetry-api==1.38.0
+opentelemetry-semantic-conventions==0.59b0
+opentelemetry-sdk==1.38.0
+typeguard==4.4.4
+shtab==1.8.0
+tyro==0.9.35
+verl==0.5.0.dev0
+autocommand==2.2.2
+backports.tarfile==1.2.0
+importlib_metadata==8.0.0
+inflect==7.3.1
+jaraco.collections==5.1.0
+jaraco.context==5.3.0
+jaraco.functools==4.0.1
+jaraco.text==3.12.1
+more-itertools==10.3.0
+packaging==24.2
+platformdirs==4.2.2
+tomli==2.0.1
+typeguard==4.3.0
+typing_extensions==4.12.2
+wheel==0.45.1
+zipp==3.19.2

wandb/run-20251222_204216-q200utyn/files/wandb-metadata.json

@@ -0,0 +1,107 @@
+{
+  "os": "Linux-5.14.0-284.25.1.el9_2.x86_64-x86_64-with-glibc2.35",
+  "python": "CPython 3.10.19",
+  "startedAt": "2025-12-22T12:42:16.666376Z",
+  "args": [
+    "--node-ip-address=10.119.27.78",
+    "--node-manager-port=37267",
+    "--object-store-name=/tmp/ray/session_2025-12-22_20-41-01_001225_536662/sockets/plasma_store",
+    "--raylet-name=/tmp/ray/session_2025-12-22_20-41-01_001225_536662/sockets/raylet",
+    "--redis-address=None",
+    "--metrics-agent-port=56422",
+    "--logging-rotate-bytes=536870912",
+    "--logging-rotate-backup-count=5",
+    "--runtime-env-agent-port=52385",
+    "--gcs-address=10.119.27.78:55346",
+    "--session-name=session_2025-12-22_20-41-01_001225_536662",
+    "--temp-dir=/tmp/ray",
+    "--webui=127.0.0.1:8265",
+    "--cluster-id=1a734f20bd020fc91173be3c2aa3032c81ef270e176e31a8146936e5",
+    "--startup-token=112",
+    "--worker-launch-time-ms=1766407269624",
+    "--node-id=39777b12b546dbda3babc455bdda2766a0d9ef51686b058776139b51",
+    "--runtime-env-hash=-839022692"
+  ],
+  "program": "/root/local/miniconda3/envs/ragen/lib/python3.10/site-packages/ray/_private/workers/default_worker.py",
+  "git": {
+    "remote": "https://github.com/mll-lab-nu/RAGEN.git",
+    "commit": "8d73639b99b38265453f898b8d6af7d4af50d56e"
+  },
+  "email": "haoyu-wa22@mails.tsinghua.edu.cn",
+  "root": "/mnt/general/wanghy/RAGEN",
+  "host": "pt-4d654cf4576f4d23ad3d3919f12932fe-worker-0",
+  "executable": "/root/local/miniconda3/envs/ragen/bin/python",
+  "cpu_count": 64,
+  "cpu_count_logical": 128,
+  "gpu": "NVIDIA H100 80GB HBM3",
+  "gpu_count": 8,
+  "disk": {
+    "/": {
+      "total": "5153960755200",
+      "used": "471749394432"
+    }
+  },
+  "memory": {
+    "total": "2163642122240"
+  },
+  "gpu_nvidia": [
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-35e2d43d-4067-82ce-90d4-def9e389bf28"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-af4135e3-88f2-e9ac-518d-502c75a85429"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-d7fdeeba-fe9b-ec03-d9f7-6724fe4266b5"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-ccc4f668-3882-5a8e-2c07-c5cd08f6f666"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-7b73c0cf-d3d5-e10c-7176-a43be1e41001"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-81b58d94-5d1f-8ec2-f9d2-fd56172ed177"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-03e8bc66-3b44-6794-49fd-5392fbdda6d1"
+    },
+    {
+      "name": "NVIDIA H100 80GB HBM3",
+      "memoryTotal": "85520809984",
+      "cudaCores": 16896,
+      "architecture": "Hopper",
+      "uuid": "GPU-b7cf0ec6-7c29-1179-dceb-09565da51890"
+    }
+  ],
+  "cudaVersion": "12.4",
+  "writerId": "m926dm4v97v6ptajgmaaghx3lzy12pre"
+}

wandb/run-20251222_204216-q200utyn/run-q200utyn.wandb

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f306e2553e250face9c74780cd43c1d0b9e3faf183dff7a9ee1654b45bda1a8a
+size 131072

wandb/run-20251222_204856-wk1bayyz/files/config.yaml

@@ -0,0 +1,962 @@
+_wandb:
+    value:
+        cli_version: 0.23.0
+        e:
+            ofgsigoz1pj2x4y546ges4yar7yl57fh:
+                args:
+                    - --node-ip-address=10.119.27.78
+                    - --node-manager-port=37599
+                    - --object-store-name=/tmp/ray/session_2025-12-22_20-47-31_599398_568116/sockets/plasma_store
+                    - --raylet-name=/tmp/ray/session_2025-12-22_20-47-31_599398_568116/sockets/raylet
+                    - --redis-address=None
+                    - --metrics-agent-port=57678
+                    - --logging-rotate-bytes=536870912
+                    - --logging-rotate-backup-count=5
+                    - --runtime-env-agent-port=51183
+                    - --gcs-address=10.119.27.78:54756
+                    - --session-name=session_2025-12-22_20-47-31_599398_568116
+                    - --temp-dir=/tmp/ray
+                    - --webui=127.0.0.1:8265
+                    - --cluster-id=f28ead16b4b4ceb793bc3b1e0e8ddd20dc8e86b4c690db66d6ef9de7
+                    - --startup-token=112
+                    - --worker-launch-time-ms=1766407665960
+                    - --node-id=d06e28f20c1442859c21d048348647f0d46a0b4c3f514005d0c7b5a6
+                    - --runtime-env-hash=-839022692
+                cpu_count: 64
+                cpu_count_logical: 128
+                cudaVersion: "12.4"
+                disk:
+                    /:
+                        total: "5153960755200"
+                        used: "471765520384"
+                email: haoyu-wa22@mails.tsinghua.edu.cn
+                executable: /root/local/miniconda3/envs/ragen/bin/python
+                git:
+                    commit: 8d73639b99b38265453f898b8d6af7d4af50d56e
+                    remote: https://github.com/mll-lab-nu/RAGEN.git
+                gpu: NVIDIA H100 80GB HBM3
+                gpu_count: 8
+                gpu_nvidia:
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-35e2d43d-4067-82ce-90d4-def9e389bf28
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-af4135e3-88f2-e9ac-518d-502c75a85429
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-d7fdeeba-fe9b-ec03-d9f7-6724fe4266b5
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-ccc4f668-3882-5a8e-2c07-c5cd08f6f666
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-7b73c0cf-d3d5-e10c-7176-a43be1e41001
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-81b58d94-5d1f-8ec2-f9d2-fd56172ed177
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-03e8bc66-3b44-6794-49fd-5392fbdda6d1
+                    - architecture: Hopper
+                      cudaCores: 16896
+                      memoryTotal: "85520809984"
+                      name: NVIDIA H100 80GB HBM3
+                      uuid: GPU-b7cf0ec6-7c29-1179-dceb-09565da51890
+                host: pt-4d654cf4576f4d23ad3d3919f12932fe-worker-0
+                memory:
+                    total: "2163642122240"
+                os: Linux-5.14.0-284.25.1.el9_2.x86_64-x86_64-with-glibc2.35
+                program: /root/local/miniconda3/envs/ragen/lib/python3.10/site-packages/ray/_private/workers/default_worker.py
+                python: CPython 3.10.19
+                root: /mnt/general/wanghy/RAGEN
+                startedAt: "2025-12-22T12:48:56.627630Z"
+                writerId: ofgsigoz1pj2x4y546ges4yar7yl57fh
+        m: []
+        python_version: 3.10.19
+        t:
+            "1":
+                - 1
+                - 5
+                - 11
+                - 30
+                - 41
+                - 49
+                - 50
+                - 51
+                - 53
+                - 71
+                - 95
+                - 97
+                - 98
+                - 105
+            "2":
+                - 1
+                - 5
+                - 11
+                - 30
+                - 41
+                - 49
+                - 50
+                - 51
+                - 53
+                - 71
+                - 95
+                - 97
+                - 98
+                - 105
+            "3":
+                - 2
+                - 13
+                - 16
+                - 61
+            "4": 3.10.19
+            "5": 0.23.0
+            "6": 4.57.1
+            "12": 0.23.0
+            "13": linux-x86_64
+actor_rollout_ref:
+    value:
+        actor:
+            _target_: verl.workers.config.FSDPActorConfig
+            checkpoint:
+                _target_: verl.trainer.config.CheckpointConfig
+                async_save: false
+                load_contents:
+                    - model
+                    - optimizer
+                    - extra
+                save_contents:
+                    - model
+                    - optimizer
+                    - extra
+            clip_ratio: 0.2
+            clip_ratio_c: 3
+            clip_ratio_high: 0.28
+            clip_ratio_low: 0.2
+            entropy_checkpointing: false
+            entropy_coeff: 0.001
+            entropy_from_logits_with_chunking: false
+            freeze_vision_tower: false
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                entropy_checkpointing: false
+                entropy_from_logits_with_chunking: false
+                forward_only: false
+                forward_prefetch: false
+                fsdp_size: -1
+                model_dtype: fp32
+                offload_policy: false
+                optimizer_offload: false
+                param_offload: false
+                reshard_after_forward: true
+                strategy: fsdp
+                ulysses_sequence_parallel_size: 1
+                use_orig_params: false
+                use_torch_compile: true
+                wrap_policy:
+                    min_num_params: 0
+            grad_clip: 1
+            grpo_advantage_length_weight: false
+            kl_loss_coef: 0
+            kl_loss_type: kl
+            loss_agg_mode: token-mean
+            optim:
+                _target_: verl.workers.config.FSDPOptimizerConfig
+                betas:
+                    - 0.9
+                    - 0.999
+                clip_grad: 1
+                lr: 1e-06
+                lr_warmup_steps: -1
+                lr_warmup_steps_ratio: 0
+                min_lr_ratio: 0
+                num_cycles: 0.5
+                total_training_steps: 200
+                warmup_style: constant
+                weight_decay: 0.01
+            policy_loss:
+                _target_: verl.workers.config.PolicyLossConfig
+                clip_cov_lb: 1
+                clip_cov_ratio: 0.0002
+                clip_cov_ub: 5
+                kl_cov_ratio: 0.0002
+                loss_mode: vanilla
+                ppo_kl_coef: 0.1
+            ppo_epochs: 1
+            ppo_max_token_len_per_gpu: 16384
+            ppo_micro_batch_size: null
+            ppo_micro_batch_size_per_gpu: 1
+            ppo_mini_batch_size: 32
+            profiler:
+                _target_: verl.utils.profiler.ProfilerConfig
+                all_ranks: false
+                enable: false
+                ranks: []
+                save_path: outputs/profile
+                tool: null
+                tool_config:
+                    npu:
+                        _target_: verl.utils.profiler.config.NPUToolConfig
+                        analysis: true
+                        contents: []
+                        discrete: false
+                        level: level1
+                    nsys:
+                        _target_: verl.utils.profiler.config.NsightToolConfig
+                        discrete: false
+                    torch:
+                        _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                        step_end: null
+                        step_start: 0
+                    torch_memory:
+                        _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                        stack_depth: 32
+                        trace_alloc_max_entries: 100000
+            shuffle: false
+            strategy: fsdp
+            tis_imp_ratio_cap: -1
+            ulysses_sequence_parallel_size: 1
+            use_dynamic_bsz: false
+            use_fused_kernels: false
+            use_kl_loss: false
+            use_ref: true
+            use_remove_padding: false
+            use_torch_compile: true
+        hybrid_engine: true
+        model:
+            _target_: verl.workers.config.HFModelConfig
+            custom_chat_template: null
+            enable_activation_offload: false
+            enable_gradient_checkpointing: true
+            exclude_modules: null
+            external_lib: null
+            fused_kernel_options:
+                impl_backend: torch
+            hf_config_path: null
+            lora_alpha: 16
+            lora_rank: 0
+            path: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            target_modules: all-linear
+            tokenizer_path: null
+            torch_dtype: bfloat16
+            trust_remote_code: false
+            use_fused_kernels: false
+            use_liger: false
+            use_remove_padding: false
+            use_shm: false
+        nccl_timeout: 600
+        ref:
+            entropy_checkpointing: false
+            entropy_from_logits_with_chunking: false
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                entropy_checkpointing: false
+                entropy_from_logits_with_chunking: false
+                forward_only: false
+                forward_prefetch: false
+                fsdp_size: -1
+                model_dtype: fp32
+                offload_policy: false
+                optimizer_offload: false
+                param_offload: false
+                reshard_after_forward: true
+                strategy: fsdp
+                ulysses_sequence_parallel_size: 1
+                use_orig_params: false
+                use_torch_compile: true
+                wrap_policy:
+                    min_num_params: 0
+            log_prob_max_token_len_per_gpu: 16384
+            log_prob_micro_batch_size: null
+            log_prob_micro_batch_size_per_gpu: 1
+            log_prob_use_dynamic_bsz: false
+            model: null
+            profiler:
+                _target_: verl.utils.profiler.ProfilerConfig
+                all_ranks: false
+                enable: false
+                ranks: []
+                save_path: outputs/profile
+                tool: null
+                tool_config:
+                    npu:
+                        _target_: verl.utils.profiler.config.NPUToolConfig
+                        analysis: true
+                        contents: []
+                        discrete: false
+                        level: level1
+                    nsys:
+                        _target_: verl.utils.profiler.config.NsightToolConfig
+                        discrete: false
+                    torch:
+                        _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                        step_end: null
+                        step_start: 0
+                    torch_memory:
+                        _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                        stack_depth: 32
+                        trace_alloc_max_entries: 100000
+            strategy: fsdp
+            ulysses_sequence_parallel_size: 1
+            use_torch_compile: true
+        rollout:
+            _target_: verl.workers.config.RolloutConfig
+            agent:
+                _target_: verl.workers.config.AgentLoopConfig
+                agent_loop_config_path: null
+                custom_async_server:
+                    _target_: verl.workers.config.CustomAsyncServerConfig
+                    name: null
+                    path: null
+                num_workers: 8
+            calculate_log_probs: false
+            cudagraph_capture_sizes: null
+            data_parallel_size: 1
+            disable_log_stats: true
+            do_sample: true
+            dtype: bfloat16
+            enable_chunked_prefill: true
+            enable_prefix_caching: true
+            enforce_eager: true
+            expert_parallel_size: 1
+            free_cache_engine: true
+            gpu_memory_utilization: 0.7
+            ignore_eos: false
+            layered_summon: false
+            load_format: dummy
+            log_prob_max_token_len_per_gpu: 16384
+            log_prob_micro_batch_size: null
+            log_prob_micro_batch_size_per_gpu: 1
+            log_prob_use_dynamic_bsz: false
+            max_model_len: 16384
+            max_num_batched_tokens: 16384
+            max_num_seqs: 1024
+            mode: sync
+            multi_stage_wake_up: false
+            multi_turn:
+                _target_: verl.workers.config.MultiTurnConfig
+                enable: false
+                format: hermes
+                interaction_config_path: null
+                max_assistant_turns: null
+                max_parallel_calls: 1
+                max_tool_response_length: 256
+                max_user_turns: null
+                num_repeat_rollouts: null
+                tokenization_sanity_check_mode: strict
+                tool_config_path: null
+                tool_response_truncate_side: middle
+                use_inference_chat_template: false
+            "n": 1
+            name: vllm
+            over_sample_rate: 0
+            profiler:
+                _target_: verl.utils.profiler.ProfilerConfig
+                all_ranks: false
+                enable: false
+                ranks: []
+                save_path: outputs/profile
+                tool: null
+                tool_config:
+                    npu:
+                        _target_: verl.utils.profiler.config.NPUToolConfig
+                        analysis: true
+                        contents: []
+                        discrete: false
+                        level: level1
+                    nsys:
+                        _target_: verl.utils.profiler.config.NsightToolConfig
+                        discrete: false
+                    torch:
+                        _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                        step_end: null
+                        step_start: 0
+                    torch_memory:
+                        _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                        stack_depth: 32
+                        trace_alloc_max_entries: 100000
+            prompt_length: 1
+            response_length: 400
+            rollout_filter_metric: reward_variance
+            rollout_filter_ratio: 0.25
+            rollout_filter_type: largest
+            skip_dump_dir: /tmp/rollout_dump
+            skip_rollout: false
+            skip_tokenizer_init: true
+            temperature: 1
+            tensor_model_parallel_size: 1
+            top_k: -1
+            top_p: 1
+            trace:
+                _target_: verl.workers.config.TraceConfig
+                backend: null
+                token2text: false
+            update_weights_bucket_megabytes: 512
+            val_kwargs:
+                _target_: verl.workers.config.SamplingConfig
+                do_sample: true
+                "n": 1
+                temperature: 0.5
+                top_k: -1
+                top_p: 1
+agent_proxy:
+    value:
+        action_sep: '||'
+        enable_think: true
+        max_actions_per_turn: 1
+        max_context_window: -1
+        max_turn: 25
+        reward_normalization:
+            grouping: state
+            method: identity
+        use_turn_scores: false
+algorithm:
+    value:
+        _target_: verl.trainer.config.AlgoConfig
+        adv_estimator: gae
+        bi_level_gae: false
+        gamma: 1
+        high_level_gamma: 0.95
+        kl_ctrl:
+            _target_: verl.trainer.config.KLControlConfig
+            horizon: 10000
+            kl_coef: 0
+            target_kl: 0.1
+            type: fixed
+        kl_penalty: kl
+        lam: 1
+        norm_adv_by_std_in_grpo: true
+        pf_ppo:
+            reweight_method: pow
+            weight_pow: 2
+        use_kl_in_reward: false
+        use_pf_ppo: false
+critic:
+    value:
+        _target_: verl.workers.config.FSDPCriticConfig
+        checkpoint:
+            _target_: verl.trainer.config.CheckpointConfig
+            async_save: false
+            load_contents:
+                - model
+                - optimizer
+                - extra
+            save_contents:
+                - model
+                - optimizer
+                - extra
+        cliprange_value: 0.5
+        enable: null
+        forward_max_token_len_per_gpu: 32768
+        forward_micro_batch_size: null
+        forward_micro_batch_size_per_gpu: 1
+        grad_clip: 1
+        loss_agg_mode: token-mean
+        model:
+            _target_: verl.workers.config.FSDPCriticModelCfg
+            enable_activation_offload: false
+            enable_gradient_checkpointing: true
+            external_lib: null
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                entropy_checkpointing: false
+                entropy_from_logits_with_chunking: false
+                forward_only: false
+                forward_prefetch: false
+                fsdp_size: -1
+                model_dtype: fp32
+                offload_policy: false
+                optimizer_offload: false
+                param_offload: false
+                reshard_after_forward: true
+                strategy: fsdp
+                ulysses_sequence_parallel_size: 1
+                use_orig_params: false
+                use_torch_compile: true
+                wrap_policy:
+                    min_num_params: 0
+            lora_alpha: 16
+            lora_rank: 0
+            path: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            target_modules: all-linear
+            tokenizer_path: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            trust_remote_code: false
+            use_remove_padding: false
+            use_shm: false
+        optim:
+            _target_: verl.workers.config.FSDPOptimizerConfig
+            betas:
+                - 0.9
+                - 0.999
+            clip_grad: 1
+            lr: 1e-05
+            lr_warmup_steps: -1
+            lr_warmup_steps_ratio: 0
+            min_lr_ratio: 0
+            num_cycles: 0.5
+            total_training_steps: 200
+            warmup_style: constant
+            weight_decay: 0.01
+        ppo_epochs: 1
+        ppo_max_token_len_per_gpu: 32768
+        ppo_micro_batch_size: null
+        ppo_micro_batch_size_per_gpu: 1
+        ppo_mini_batch_size: 32
+        profiler:
+            _target_: verl.utils.profiler.ProfilerConfig
+            all_ranks: false
+            enable: false
+            ranks: []
+            save_path: outputs/profile
+            tool: null
+            tool_config:
+                npu:
+                    _target_: verl.utils.profiler.config.NPUToolConfig
+                    analysis: true
+                    contents: []
+                    discrete: false
+                    level: level1
+                nsys:
+                    _target_: verl.utils.profiler.config.NsightToolConfig
+                    discrete: false
+                torch:
+                    _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                    step_end: null
+                    step_start: 0
+                torch_memory:
+                    _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                    stack_depth: 32
+                    trace_alloc_max_entries: 100000
+        rollout_n: 1
+        shuffle: false
+        strategy: fsdp
+        ulysses_sequence_parallel_size: 1
+        use_dynamic_bsz: false
+ctx_manager:
+    value:
+        generation:
+            gen_config:
+                kwargs: null
+                response_length: 400
+                temperature: 1
+                top_k: -1
+                top_p: 1
+custom_envs:
+    value:
+        Bandit:
+            env_config:
+                split: train
+            env_instruction: ""
+            env_type: bandit
+            max_actions_per_traj: 1
+            max_tokens: 100
+        BanditTest:
+            env_config:
+                split: test
+            env_instruction: ""
+            env_type: bandit
+            max_actions_per_traj: 1
+            max_tokens: 100
+        CoordFrozenLake:
+            env_config:
+                observation_format: grid_coord
+            env_instruction: |
+                You are solving the FrozenLake puzzle. The observation includes both a symbol grid and zero-indexed coordinates for the start, goal, player, and any holes.
+                Coordinates range from the top-left corner (0, 0) to the bottom-right corner (5, 5).
+                Beware that the ice is slippery, so the agent might slide and end up in an unintended tile.
+                Respond with a sequence of actions such as <answer>Left || Up || Up</answer>.
+            env_type: frozen_lake
+            max_actions_per_traj: 25
+            max_tokens: 120
+        CoordSokoban:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                max_steps: 100
+                num_boxes: 2
+                observation_format: grid_coord
+            env_instruction: "You are solving the Sokoban puzzle. You are the player and you need to push all boxes to targets.\nYou are provided with a symbol grid and the zero-indexed coordinates of the player, each box, and each target. \nCoordinates range from the top-left corner (0, 0) to the bottom-right corner (5, 5). \nWhen you are exactly next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box.\nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>.\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 120
+        Countdown:
+            env_config: null
+            env_instruction: 'You are solving the Countdown puzzle. You should use the num list to create an equation that equals the target. Example answer format: <think> To find an equation using [3, 5, 2] to get 4. Let''s check 2 + 5 = 7, 7 - 3 = 4. So the answer is 2 + 5 - 3 = 4. </think><answer>2 + 5 - 3</answer>'
+            env_type: countdown
+            max_actions_per_traj: 1
+            max_tokens: 100
+        FrozenLake:
+            env_config: null
+            env_instruction: 'You are solving the FrozenLake puzzle. Forbid the whole and go to the target. You may move to the unintended direction due to the slippery ice. Example answer format: <think>To forbid the hole and go to the target, I should go left then go up.</think><answer>Left || Up</answer>'
+            env_type: frozen_lake
+            max_actions_per_traj: 25
+            max_tokens: 100
+        LargerSokoban:
+            env_config:
+                dim_x: 8
+                dim_y: 8
+                max_steps: 100
+                num_boxes: 2
+                search_depth: 10
+            env_instruction: "You are solving the Sokoban puzzle. \nYou are the player and you need to push all boxes to targets. \nWhen you are right next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box. \nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        Lean:
+            env_config: null
+            env_instruction: You are a Lean theorem prover. Given a Lean theorem statement, propose a sequence of tactics that completes the proof. Think step by step about which tactics to apply next. Provide tactics separated by '||', for example <answer>intro || simp || rfl</answer>.
+            env_type: lean
+            max_actions_per_traj: 30
+            max_tokens: 512
+        MediumSudoku:
+            env_config:
+                difficulty: medium
+                grid_size: 9
+                max_steps: 81
+                render_format: with_feedback
+                show_conflicts: true
+                show_valid_numbers: true
+            env_instruction: |
+                You are solving a Sudoku puzzle. Fill in the grid so that every row, column, and 3x3 box contains the numbers 1-9 without repetition.
+                Initial cells are shown in [brackets] and cannot be modified. Empty cells are shown as dots (.).
+                Place numbers one at a time using the format: <answer>place 5 at row 2 col 3</answer> or <answer>2,3,5</answer>
+                The environment will provide feedback on valid/invalid moves and show conflicts if any occur.
+            env_type: sudoku
+            max_actions_per_traj: 30
+            max_tokens: 150
+            max_workers: 32
+            parallel_friendly: false
+        MetamathQA:
+            env_config: null
+            env_instruction: 'You are solving Math problems. '
+            env_type: metamathqa
+            max_actions_per_traj: 1
+            max_tokens: 100
+        SimpleSokoban:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                max_steps: 100
+                num_boxes: 2
+            env_instruction: "You are solving the Sokoban puzzle. \nYou are the player and you need to push all boxes to targets. \nWhen you are right next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box. \nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        SimpleSudoku:
+            env_config:
+                difficulty: easy
+                grid_size: 4
+                max_steps: 20
+                render_format: with_feedback
+                show_conflicts: false
+                show_valid_numbers: false
+            env_instruction: |
+                You are solving a Sudoku puzzle. Fill in the grid so that every row, column, and 2x2 box contains the numbers 1-4 without repetition.
+                Initial cells are shown in [brackets] and cannot be modified. Empty cells are shown as dots (.).
+                Place numbers one at a time using the format, for example: <answer>place 1 at row 2 col 3</answer> or <answer>1,2,3</answer>
+                The environment will provide feedback on valid/invalid moves and show conflicts if any occur.
+            env_type: sudoku
+            max_actions_per_traj: 20
+            max_tokens: 150
+            max_workers: 32
+            parallel_friendly: false
+        SokobanDifferentGridVocab:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                grid_lookup:
+                    "0": W
+                    "1": .
+                    "2": G
+                    "3": C
+                    "4": B
+                    "5": A
+                    "6": '@'
+                grid_vocab:
+                    .: empty
+                    '@': player on target
+                    A: player
+                    B: box
+                    C: box on target
+                    G: target
+                    W: wall
+                max_steps: 100
+                num_boxes: 1
+                search_depth: 30
+            env_instruction: "You are solving the Sokoban puzzle. \nYou are the player and you need to push all boxes to targets. \nWhen you are right next to a box, you can push it by moving in the same direction. \nYou cannot push a box through a wall, and you cannot pull a box. \nThe answer should be a sequence of actions, like <answer>Right || Right || Up</answer>\n"
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        VisualSimpleSokoban:
+            env_config:
+                dim_x: 6
+                dim_y: 6
+                max_steps: 100
+                num_boxes: 1
+                render_mode: rgb_array
+            env_instruction: You are solving the Sokoban puzzle. You are the player and you need to push all boxes to targets. When you are right next to a box, you can push it by moving in the same direction. You cannot push a box through a wall, and you cannot pull a box. The answer should be a sequence of actions, like <answer>Right || Right || Up</answer>
+            env_type: sokoban
+            max_actions_per_traj: 10
+            max_tokens: 100
+        WebShop:
+            env_config:
+                dataset: small
+            env_instruction: 'You are browsing an online shop. Based on the instruction, buy a product that close to the production description. You need to search, read the search results, pick a product, choose the size and color and buy. You should only choose action from the available actions list provided later.  Example process: I need a gingko light and 20x20 pillow cover that is hand painted. First search[gingko light 20x20 pillow cover hand painted], answer format: <answer>search[blanket with fleece throw]</answer>. Valid answer is search[<keywords>] or click[<clickable>].'
+            env_type: webshop
+            max_actions_per_traj: 9
+            max_tokens: 200
+        blackjack:
+            env_config: null
+            env_instruction: |
+                You are playing Blackjack against a dealer. The dealer must hit on 16 or less and stand on 17 or more.
+                Choose either Stick or Hit. Respond with a single action.
+                Example: <answer>Hit</answer>
+            env_type: blackjack
+            max_actions_per_traj: 10
+            max_tokens: 64
+        game_2048:
+            env_config: null
+            env_instruction: |
+                You are playing the 2048 game on a 4x4 grid. Merge equal tiles by sliding Up, Right, Down, or Left.
+                If a move is invalid (no tiles move), a small penalty is applied. Respond with a single action.
+                Example: <answer>Up</answer>
+            env_type: game_2048
+            max_actions_per_traj: 700
+            max_tokens: 8192
+        rubikscube:
+            env_config:
+                max_steps: 20
+                render_mode: text
+                scramble_depth: 1
+            env_instruction: |
+                You are solving a 2x2 Rubik's Cube (Pocket Cube). The goal is to restore the cube so that each of the faces consists of a single, unique color.
+                Available actions use standard Singmaster notation for face rotations: U, U', D, D', L, L', R, R', F, F', B, B'.
+                - Faces: U (Up), D (Down), L (Left), R (Right), F (Front), B (Back).
+                - Modifiers: A letter alone means 90° clockwise (e.g., 'R'). A letter with prime (') means 90° counter-clockwise (e.g., "R'").
+                Respond with a sequence of actions separated by "||".
+                Example: <answer>U</answer>
+            env_type: rubikscube
+            max_actions_per_traj: 20
+            max_tokens: 96
+custom_reward_function:
+    value:
+        name: compute_score
+        path: null
+data:
+    value:
+        custom_cls:
+            name: null
+            path: null
+        datagen:
+            name: null
+            path: null
+        dataloader_num_workers: 8
+        filter_overlong_prompts: false
+        filter_overlong_prompts_workers: 1
+        image_key: images
+        max_prompt_length: null
+        max_response_length: null
+        prompt_key: prompt
+        return_full_prompt: false
+        return_multi_modal_inputs: true
+        return_raw_chat: false
+        return_raw_input_ids: false
+        reward_fn_key: data_source
+        sampler:
+            class_name: null
+            class_path: null
+        shuffle: true
+        tokenizer: null
+        train_batch_size: 128
+        train_files: ~/data/rlhf/gsm8k/train.parquet
+        truncation: error
+        trust_remote_code: false
+        use_shm: false
+        val_batch_size: null
+        val_files: ~/data/rlhf/gsm8k/test.parquet
+        validation_shuffle: false
+        video_key: videos
+enable_response_mask:
+    value: true
+es_manager:
+    value:
+        format_penalty: -0.1
+        train:
+            env_configs:
+                n_groups:
+                    - 8
+                tags:
+                    - CoordFrozenLake
+            env_groups: 8
+            group_size: 16
+        val:
+            env_configs:
+                n_groups:
+                    - 256
+                tags:
+                    - CoordFrozenLake
+            env_groups: 256
+            group_size: 1
+global_profiler:
+    value:
+        _target_: verl.utils.profiler.ProfilerConfig
+        global_tool_config:
+            nsys:
+                _target_: verl.utils.profiler.config.NsightToolConfig
+                controller_nsight_options:
+                    cuda-graph-trace: graph
+                    cuda-memory-usage: "true"
+                    trace: cuda,nvtx,cublas,ucx
+                discrete: false
+                worker_nsight_options:
+                    capture-range: cudaProfilerApi
+                    capture-range-end: null
+                    cuda-graph-trace: graph
+                    cuda-memory-usage: "true"
+                    kill: none
+                    trace: cuda,nvtx,cublas,ucx
+            torch_memory:
+                context: all
+                stack_depth: 32
+                stacks: all
+                trace_alloc_max_entries: 100000
+        profile_continuous_steps: false
+        save_path: outputs/profile
+        steps: null
+        tool: null
+grpo_advantage_length_weight:
+    value: false
+lora:
+    value:
+        alpha: 16
+        rank: 0
+        target_modules: all-linear
+micro_batch_size_per_gpu:
+    value: 1
+model_path:
+    value: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+ppo_mini_batch_size:
+    value: 32
+ray_kwargs:
+    value:
+        ray_init:
+            num_cpus: null
+        timeline_json_file: null
+reward_model:
+    value:
+        enable: false
+        enable_resource_pool: false
+        forward_max_token_len_per_gpu: 32768
+        launch_reward_fn_async: false
+        max_length: null
+        micro_batch_size: null
+        micro_batch_size_per_gpu: null
+        model:
+            external_lib: null
+            fsdp_config:
+                _target_: verl.workers.config.FSDPEngineConfig
+                forward_prefetch: false
+                fsdp_size: -1
+                param_offload: false
+                reshard_after_forward: true
+                wrap_policy:
+                    min_num_params: 0
+            input_tokenizer: /mnt/general/share/model/meta-llama/Llama-3.2-1B-Instruct
+            path: ~/models/FsfairX-LLaMA3-RM-v0.1
+            trust_remote_code: false
+            use_fused_kernels: false
+            use_remove_padding: false
+            use_shm: false
+        n_gpus_per_node: 0
+        nnodes: 0
+        profiler:
+            _target_: verl.utils.profiler.ProfilerConfig
+            all_ranks: false
+            enable: false
+            ranks: []
+            save_path: outputs/profile
+            tool: null
+            tool_config:
+                npu:
+                    _target_: verl.utils.profiler.config.NPUToolConfig
+                    analysis: true
+                    contents: []
+                    discrete: false
+                    level: level1
+                nsys:
+                    _target_: verl.utils.profiler.config.NsightToolConfig
+                    discrete: false
+                torch:
+                    _target_: verl.utils.profiler.config.TorchProfilerToolConfig
+                    step_end: null
+                    step_start: 0
+                torch_memory:
+                    _target_: verl.utils.profiler.config.TorchMemoryToolConfig
+                    stack_depth: 32
+                    trace_alloc_max_entries: 100000
+        reward_manager: naive
+        sandbox_fusion:
+            max_concurrent: 64
+            memory_limit_mb: 1024
+            url: null
+        strategy: fsdp
+        ulysses_sequence_parallel_size: 1
+        use_dynamic_bsz: false
+seed:
+    value:
+        train: 10000
+        val: 123
+system:
+    value:
+        CUDA_VISIBLE_DEVICES: 0,1,2,3,4,5,6,7
+trainer:
+    value:
+        balance_batch: true
+        critic_warmup: 0
+        default_hdfs_dir: null
+        default_local_dir: /mnt/general/wanghy/RAGEN/saves/
+        del_local_ckpt_after_load: false
+        device: cuda
+        esi_redundant_time: 0
+        experiment_name: frozen_lake-starpo-slippery
+        generations_to_log_to_wandb:
+            train: 128
+            val: 20
+        local_log_dir: results/
+        log_val_generations: 0
+        logger:
+            - console
+            - wandb
+        max_actor_ckpt_to_keep: 1
+        max_critic_ckpt_to_keep: 1
+        n_gpus_per_node: 8
+        nnodes: 1
+        project_name: ragen_latest_llama_1B_it
+        ray_wait_register_center_timeout: 300
+        resume_from_path: null
+        resume_mode: auto
+        rollout_data_dir: null
+        save_freq: -1
+        test_freq: 10
+        total_epochs: 30
+        total_training_steps: 200
+        use_legacy_worker_impl: auto
+        val_before_train: true
+        val_only: false
+        validation_data_dir: null
+        validation_steps: 1