1665 lines
278 KiB
Plaintext
1665 lines
278 KiB
Plaintext
|
|
|||
|
|
The following have been reloaded with a version change:
|
|||
|
|
1) GCCcore/.14.3.0 => GCCcore/14.3.0
|
|||
|
|
|
|||
|
|
|
|||
|
|
Lmod is automatically replacing "GCC/14.3.0" with
|
|||
|
|
"nvidia-compilers/25.9-CUDA-13".
|
|||
|
|
|
|||
|
|
Deactivating conda environment: /e/scratch/jureap59/feuer1/miniforge3/envs/otagent
|
|||
|
|
Activating RL environment: /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl
|
|||
|
|
Python executable: /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/bin/python
|
|||
|
|
Python path check: /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/bin/python
|
|||
|
|
[ray] RAY_TMPDIR=/tmp/ray/ray_496419
|
|||
|
|
[triton_cache] Triton cache: /tmp/triton_cache_feuer1_496419
|
|||
|
|
[triton_cache] TorchInductor cache: /tmp/torchinductor_cache_feuer1_496419
|
|||
|
|
[proxy] ✓ Found proxychains binary at /e/scratch/jureap59/feuer1/proxychains-ng-aarch64/bin/proxychains4
|
|||
|
|
[proxy] Setting up SSH tunnel to jpbl-s01-01
|
|||
|
|
[proxy] SSH key: /e/home/jusers/feuer1/jupiter/.ssh/authorized_keys/id_ed25519_jsc
|
|||
|
|
[proxy] Tunnel port: 7003
|
|||
|
|
[proxy] Node IP: 10.128.17.210 (workers will connect here)
|
|||
|
|
[proxy] ✓ SSH tunnel started successfully
|
|||
|
|
[proxy] ✓ Generated proxychains config at /e/home/jusers/feuer1/jupiter/.proxychains/proxychains_496419.conf
|
|||
|
|
[proxy] - Internal traffic (10.x.x.x, 172.x.x.x, 169.254.x.x) → DIRECT
|
|||
|
|
[proxy] - External traffic (internet) → PROXY via tunnel
|
|||
|
|
[proxy] ✓ Daytona timeout settings configured
|
|||
|
|
[proxy] Testing proxy connectivity...
|
|||
|
|
[proxychains] config file found: /e/home/jusers/feuer1/jupiter/.proxychains/proxychains_496419.conf
|
|||
|
|
[proxychains] preloading /e/scratch/jureap59/feuer1/proxychains-ng-aarch64/lib/libproxychains4.so
|
|||
|
|
[proxychains] DLL init: proxychains-ng 4.17-git-9-g78ead0e
|
|||
|
|
[proxy] ✓ Proxy connectivity test passed (huggingface.co reachable via wrapped binary)
|
|||
|
|
[proxy] ⚠ Tunnel not accessible at 10.128.17.210:7003 (workers may fail)
|
|||
|
|
[proxy] ✓ Proxy setup complete (using wrapped binary for Ray workers)
|
|||
|
|
[container_runtime] Using cloud backend: daytona (no local container setup)
|
|||
|
|
=== Universal RL Training Runner ===
|
|||
|
|
Config: /e/data1/datasets/playground/ot-baf/a3-rl-DCAgent_code-contests-noblock/configs/a3-rl-DCAgent_code-contests-noblock_rl_config.json
|
|||
|
|
Working directory: /e/scratch/jureap59/feuer1/OpenThoughts-Agent
|
|||
|
|
Python: /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/bin/python
|
|||
|
|
Python version: Python 3.12.12
|
|||
|
|
Proxy: DISABLED (direct internet or not configured)
|
|||
|
|
========================================
|
|||
|
|
=== RLJobRunner: a3-rl-DCAgent_code-contests-noblock ===
|
|||
|
|
[wandb_utils] Fixing permissions on: /e/data1/datasets/playground/ot-baf/a3-rl-DCAgent_code-contests-noblock/wandb
|
|||
|
|
[wandb_utils] WandB directory ready: /e/data1/datasets/playground/ot-baf/a3-rl-DCAgent_code-contests-noblock/wandb
|
|||
|
|
HF_TOKEN=****pDbg
|
|||
|
|
HF_HUB_CACHE=/e/data1/datasets/playground/ot-baf/hf_hub
|
|||
|
|
SUPABASE_URL=https://rpzmyuapoqilpghynmza.s... (direct Supabase config)
|
|||
|
|
Environment configured:
|
|||
|
|
TENSOR_PARALLEL_SIZE=1
|
|||
|
|
NUM_INFERENCE_ENGINES=56
|
|||
|
|
POLICY_NUM_NODES=14
|
|||
|
|
WANDB_DIR=/e/data1/datasets/playground/ot-baf/a3-rl-DCAgent_code-contests-noblock/wandb
|
|||
|
|
Starting Ray cluster with 14 nodes, 4 GPUs/node
|
|||
|
|
Cleaning up existing Ray instances...
|
|||
|
|
=== Starting Ray Cluster ===
|
|||
|
|
Nodes: 14
|
|||
|
|
GPUs per node: 4
|
|||
|
|
CPUs per node: 288
|
|||
|
|
Head node: jpbo-010-34 (10.128.17.210)
|
|||
|
|
Ray port: 6379
|
|||
|
|
============================
|
|||
|
|
Starting Ray head on jpbo-010-34 (logging to /e/data1/datasets/playground/ot-baf/experiments/_ray_logs/ray_head_jpbo-010-34.log)...
|
|||
|
|
Command: srun --export=ALL,VLLM_HOST_IP=10.128.17.210 --nodes=1 --ntasks=1 --gres=gpu:4 --gpu-bind=none --overlap --cpu-bind=none -w jpbo-010-34 bash -c /e/scratch/jureap59/feuer1/proxychains-ng-aarch64/bin/proxychains4 -f "$PROXYCHAINS_CONF_FILE" ray start --head --node-ip-address=10.128.17.210 --port=6379 --num-gpus=4 --num-cpus=288 --block --object-store-memory=42949672960
|
|||
|
|
Started Ray head on jpbo-010-34
|
|||
|
|
Starting Ray worker on jpbo-010-36 (logging to /e/data1/datasets/playground/ot-baf/experiments/_ray_logs/ray_worker_jpbo-010-36.log)...
|
|||
|
|
Command: srun --export=ALL,VLLM_HOST_IP=10.128.17.212 --nodes=1 --ntasks=1 --gres=gpu:4 --gpu-bind=none --overlap --cpu-bind=none -w jpbo-010-36 bash -c /e/scratch/jureap59/feuer1/proxychains-ng-aarch64/bin/proxychains4 -f "$PROXYCHAINS_CONF_FILE" ray start --address=10.128.17.210:6379 --node-ip-address=10.128.17.212 --num-gpus=4 --num-cpus=288 --block --object-store-memory=42949672960
|
|||
|
|
Started Ray worker 1 on jpbo-010-36
|
|||
|
|
Starting Ray worker on jpbo-010-38 (logging to /e/data1/datasets/playground/ot-baf/experiments/_ray_logs/ray_worker_jpbo-010-38.log)...
|
|||
|
|
Command: srun --export=ALL,VLLM_HOST_IP=10.128.17.214 --nodes=1 --ntasks=1 --gres=gpu:4 --gpu-bind=none --overlap --cpu-bind=none -w jpbo-010-38 bash -c /e/scratch/jureap59/feuer1/proxychains-ng-aarch64/bin/proxychains4 -f "$PROXYCHAINS_CONF_FILE" ray start --address=10.128.17.210:6379 --node-ip-address=10.128.17.214 --num-gpus=4 --num-cpus=288 --block --object-store-memory=42949672960
|
|||
|
|
Started Ray worker 2 on jpbo-010-38
|
|||
|
|
Starting Ray worker on jpbo-010-39 (logging to /e/data1/datasets/playground/ot-baf/experiments/_ray_logs/ray_worker_jpbo-010-39.log)...
|
|||
|
|
Command: srun --export=ALL,VLLM_HOST_IP=10.128.17.215 --nodes=1 --ntasks=1 --gres=gpu:4 --gpu-bind=none --overlap --cpu-bind=none -w jpbo-010-39 bash -c /e/scratch/jureap59/feuer1/proxychains-ng-aarch64/bin/proxychains4 -f "$PROXYCHAINS_CONF_FILE" ray start --address=10.128.17.210:6379 --node-ip-address=10.128.17.215 --num-gpus=4 --num-cpus=288 --block --object-store-memory=42949672960
|
|||
|
|
Started Ray worker 3 on jpbo-010-39
|
|||
|
|
Starting Ray worker on jpbo-010-40 (logging to /e/data1/datasets/playground/ot-baf/experiments/_ray_logs/ray_worker_jpbo-010-40.log)...
|
|||
|
|
Command: srun --export=ALL,VLLM_HOST_IP=10.128.17.216 --nodes=1 --ntasks=1 --gres=gpu:4 --gpu-bind=none --overlap --cpu-bind=none -w jpbo-010-40 bash -c /e/scratch/jureap59/feuer1/proxychains-ng-aarch64/bin/proxychains4 -f "$PROXYCHAINS_CONF_FILE" ray start --address=10.128.17.210:6379 --node-ip-address=10.128.17.216 --num-gpus=4 --num-cpus=288 --block --object-store-memory=42949672960
|
|||
|
|
Started Ray worker 4 on jpbo-010-40
|
|||
|
|
Starting Ray worker on jpbo-010-41 (logging to /e/data1/datasets/playground/ot-baf/experiments/_ray_logs/ray_worker_jpbo-010-41.log)...
|
|||
|
|
Command: srun --export=ALL,VLLM_HOST_IP=10.128.17.217 --nodes=1 --ntasks=1 --gres=gpu:4 --gpu-bind=none --overlap --cpu-bind=none -w jpbo-010-41 bash -c /e/scratch/jureap59/feuer1/proxychains-ng-aarch64/bin/proxychains4 -f "$PROXYCHAINS_CONF_FILE" ray start --address=10.128.17.210:6379 --node-ip-address=10.128.17.217 --num-gpus=4 --num-cpus=288 --block --object-store-memory=42949672960
|
|||
|
|
Started Ray worker 5 on jpbo-010-41
|
|||
|
|
Starting Ray worker on jpbo-010-43 (logging to /e/data1/datasets/playground/ot-baf/experiments/_ray_logs/ray_worker_jpbo-010-43.log)...
|
|||
|
|
Command: srun --export=ALL,VLLM_HOST_IP=10.128.17.219 --nodes=1 --ntasks=1 --gres=gpu:4 --gpu-bind=none --overlap --cpu-bind=none -w jpbo-010-43 bash -c /e/scratch/jureap59/feuer1/proxychains-ng-aarch64/bin/proxychains4 -f "$PROXYCHAINS_CONF_FILE" ray start --address=10.128.17.210:6379 --node-ip-address=10.128.17.219 --num-gpus=4 --num-cpus=288 --block --object-store-memory=42949672960
|
|||
|
|
Started Ray worker 6 on jpbo-010-43
|
|||
|
|
Starting Ray worker on jpbo-010-44 (logging to /e/data1/datasets/playground/ot-baf/experiments/_ray_logs/ray_worker_jpbo-010-44.log)...
|
|||
|
|
Command: srun --export=ALL,VLLM_HOST_IP=10.128.17.220 --nodes=1 --ntasks=1 --gres=gpu:4 --gpu-bind=none --overlap --cpu-bind=none -w jpbo-010-44 bash -c /e/scratch/jureap59/feuer1/proxychains-ng-aarch64/bin/proxychains4 -f "$PROXYCHAINS_CONF_FILE" ray start --address=10.128.17.210:6379 --node-ip-address=10.128.17.220 --num-gpus=4 --num-cpus=288 --block --object-store-memory=42949672960
|
|||
|
|
Started Ray worker 7 on jpbo-010-44
|
|||
|
|
Starting Ray worker on jpbo-010-45 (logging to /e/data1/datasets/playground/ot-baf/experiments/_ray_logs/ray_worker_jpbo-010-45.log)...
|
|||
|
|
Command: srun --export=ALL,VLLM_HOST_IP=10.128.17.221 --nodes=1 --ntasks=1 --gres=gpu:4 --gpu-bind=none --overlap --cpu-bind=none -w jpbo-010-45 bash -c /e/scratch/jureap59/feuer1/proxychains-ng-aarch64/bin/proxychains4 -f "$PROXYCHAINS_CONF_FILE" ray start --address=10.128.17.210:6379 --node-ip-address=10.128.17.221 --num-gpus=4 --num-cpus=288 --block --object-store-memory=42949672960
|
|||
|
|
Started Ray worker 8 on jpbo-010-45
|
|||
|
|
Starting Ray worker on jpbo-010-46 (logging to /e/data1/datasets/playground/ot-baf/experiments/_ray_logs/ray_worker_jpbo-010-46.log)...
|
|||
|
|
Command: srun --export=ALL,VLLM_HOST_IP=10.128.17.222 --nodes=1 --ntasks=1 --gres=gpu:4 --gpu-bind=none --overlap --cpu-bind=none -w jpbo-010-46 bash -c /e/scratch/jureap59/feuer1/proxychains-ng-aarch64/bin/proxychains4 -f "$PROXYCHAINS_CONF_FILE" ray start --address=10.128.17.210:6379 --node-ip-address=10.128.17.222 --num-gpus=4 --num-cpus=288 --block --object-store-memory=42949672960
|
|||
|
|
Started Ray worker 9 on jpbo-010-46
|
|||
|
|
Starting Ray worker on jpbo-022-02 (logging to /e/data1/datasets/playground/ot-baf/experiments/_ray_logs/ray_worker_jpbo-022-02.log)...
|
|||
|
|
Command: srun --export=ALL,VLLM_HOST_IP=10.128.24.50 --nodes=1 --ntasks=1 --gres=gpu:4 --gpu-bind=none --overlap --cpu-bind=none -w jpbo-022-02 bash -c /e/scratch/jureap59/feuer1/proxychains-ng-aarch64/bin/proxychains4 -f "$PROXYCHAINS_CONF_FILE" ray start --address=10.128.17.210:6379 --node-ip-address=10.128.24.50 --num-gpus=4 --num-cpus=288 --block --object-store-memory=42949672960
|
|||
|
|
Started Ray worker 10 on jpbo-022-02
|
|||
|
|
Starting Ray worker on jpbo-022-03 (logging to /e/data1/datasets/playground/ot-baf/experiments/_ray_logs/ray_worker_jpbo-022-03.log)...
|
|||
|
|
Command: srun --export=ALL,VLLM_HOST_IP=10.128.24.51 --nodes=1 --ntasks=1 --gres=gpu:4 --gpu-bind=none --overlap --cpu-bind=none -w jpbo-022-03 bash -c /e/scratch/jureap59/feuer1/proxychains-ng-aarch64/bin/proxychains4 -f "$PROXYCHAINS_CONF_FILE" ray start --address=10.128.17.210:6379 --node-ip-address=10.128.24.51 --num-gpus=4 --num-cpus=288 --block --object-store-memory=42949672960
|
|||
|
|
Started Ray worker 11 on jpbo-022-03
|
|||
|
|
Starting Ray worker on jpbo-022-04 (logging to /e/data1/datasets/playground/ot-baf/experiments/_ray_logs/ray_worker_jpbo-022-04.log)...
|
|||
|
|
Command: srun --export=ALL,VLLM_HOST_IP=10.128.24.52 --nodes=1 --ntasks=1 --gres=gpu:4 --gpu-bind=none --overlap --cpu-bind=none -w jpbo-022-04 bash -c /e/scratch/jureap59/feuer1/proxychains-ng-aarch64/bin/proxychains4 -f "$PROXYCHAINS_CONF_FILE" ray start --address=10.128.17.210:6379 --node-ip-address=10.128.24.52 --num-gpus=4 --num-cpus=288 --block --object-store-memory=42949672960
|
|||
|
|
Started Ray worker 12 on jpbo-022-04
|
|||
|
|
Starting Ray worker on jpbo-022-05 (logging to /e/data1/datasets/playground/ot-baf/experiments/_ray_logs/ray_worker_jpbo-022-05.log)...
|
|||
|
|
Command: srun --export=ALL,VLLM_HOST_IP=10.128.24.53 --nodes=1 --ntasks=1 --gres=gpu:4 --gpu-bind=none --overlap --cpu-bind=none -w jpbo-022-05 bash -c /e/scratch/jureap59/feuer1/proxychains-ng-aarch64/bin/proxychains4 -f "$PROXYCHAINS_CONF_FILE" ray start --address=10.128.17.210:6379 --node-ip-address=10.128.24.53 --num-gpus=4 --num-cpus=288 --block --object-store-memory=42949672960
|
|||
|
|
Started Ray worker 13 on jpbo-022-05
|
|||
|
|
Waiting for cluster (56 GPUs, 14 nodes)...
|
|||
|
|
Connecting to Ray at 10.128.17.210:6379 (expecting 14 nodes, 56.0 GPUs)
|
|||
|
|
Ray connection established, polling for resources...
|
|||
|
|
[Ray wait] nodes=14/14 GPUs=56.0/56.0 resources={'accelerator_type:GH200': 14.0, 'CPU': 4032.0, 'GPU': 56.0, 'node:10.128.17.216': 1.0, 'object_store_memory': 601295421440.0, 'memory': 10779218477056.0, 'node:10.128.17.222': 1.0, 'node:10.128.17.221': 1.0, 'node:10.128.24.50': 1.0, 'node:10.128.24.52': 1.0, 'node:10.128.17.215': 1.0, 'node:__internal_head__': 1.0, 'node:10.128.17.210': 1.0, 'node:10.128.17.214': 1.0, 'node:10.128.24.51': 1.0, 'node:10.128.24.53': 1.0, 'node:10.128.17.217': 1.0, 'node:10.128.17.220': 1.0, 'node:10.128.17.219': 1.0, 'node:10.128.17.212': 1.0}
|
|||
|
|
✓ Ray cluster ready
|
|||
|
|
=== Ray Cluster Ready ===
|
|||
|
|
Address: 10.128.17.210:6379
|
|||
|
|
Total GPUs: 56
|
|||
|
|
=========================
|
|||
|
|
Ray cluster ready at 10.128.17.210:6379
|
|||
|
|
Total GPUs available: 56
|
|||
|
|
[RLJobRunner] Pinggy check: url=False, token=False, needs_tunnel=False (agent=terminus-2, env=daytona)
|
|||
|
|
[RLJobRunner] No Pinggy tunnel needed, using local vLLM
|
|||
|
|
|
|||
|
|
Running SkyRL:
|
|||
|
|
Python: /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/bin/python
|
|||
|
|
Entrypoint: examples.terminal_bench.entrypoints.main_tbench
|
|||
|
|
Args: 119 Hydra arguments
|
|||
|
|
Working dir: /e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train
|
|||
|
|
Using proxychains binary: /e/scratch/jureap59/feuer1/proxychains-ng-aarch64/bin/proxychains4
|
|||
|
|
|
|||
|
|
Executing command with srun: /e/scratch/jureap59/feuer1/proxychains-ng-aarch64/bin/proxychains4 -f $PROXYCHAINS_CONF_FILE /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/bin/python -m examples.terminal_bench.entrypoints.main_tbench +terminal_bench_config=terminal_bench trainer.strategy=fsdp2 trainer.algorithm.advantage_estimator=rloo_n trainer.algorithm.use_kl_loss=false trainer.algorithm.kl_loss_coef=0.0 trainer.algorithm.eps_clip_low=0.2 trainer.algorithm.eps_clip_high=0.05 trainer.algorithm.loss_reduction=token_mean trainer.epochs=2 trainer.max_steps=80 trainer.update_epochs_per_batch=1 trainer.train_batch_size=64 trainer.policy_mini_batch_size=64 trainer.eval_batch_size=64 trainer.micro_forward_batch_size_per_gpu=4 trainer.micro_train_batch_size_per_gpu=1 trainer.max_prompt_length=999999 trainer.eval_interval=999999 trainer.eval_before_train=false trainer.ckpt_interval=2 trainer.resume_mode=latest trainer.hf_save_interval=5 ++trainer.hf_hub_repo_id=laion/a3-rl-DCAgent_code-contests-noblock ++trainer.hf_hub_private=false ++trainer.hf_hub_revision=main ++trainer.enable_db_registration=true trainer.project_name=OpenThoughts-Agent trainer.log_level=INFO trainer.tracker_commit_each_step=true trainer.logger=console trainer.run_name=a3-rl-DCAgent_code-contests-noblock trainer.ckpt_path=/e/data1/datasets/playground/ot-baf/a3-rl-DCAgent_code-contests-noblock/a3-rl-DCAgent_code-contests-noblock/checkpoints trainer.export_path=/e/data1/datasets/playground/ot-baf/a3-rl-DCAgent_code-contests-noblock/a3-rl-DCAgent_code-contests-noblock/exports trainer.policy.optimizer_config.lr=8e-6 trainer.policy.optimizer_config.weight_decay=0.0 trainer.policy.optimizer_config.adam_betas=[0.9,0.999] trainer.policy.optimizer_config.max_grad_norm=0.9 trainer.policy.fsdp_config.cpu_offload=false trainer.policy.fsdp_config.reshard_after_forward=true trainer.policy.fsdp_config.fsdp_size=4 trainer.policy.model.path=/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6 trainer.ref.fsdp_config.cpu_offload=false trainer.ref.fsdp_config.reshard_after_forward=true trainer.ref.fsdp_config.fsdp_size=4 trainer.placement.colocate_all=false trainer.placement.policy_num_nodes=2 trainer.placement.ref_num_nodes=2 trainer.placement.policy_num_gpus_per_node=4 trainer.placement.ref_num_gpus_per_node=4 trainer.fully_async.max_staleness_steps=16 trainer.fully_async.num_parallel_generation_workers=450 generator.backend=vllm generator.timeout_multiplier=1.0 generator.model_dtype=bfloat16 generator.inference_engine_tensor_parallel_size=1 generator.num_inference_engines=48 generator.n_samples_per_prompt=8 generator.eval_n_samples_per_prompt=8 generator.gpu_memory_utilization=0.75 generator.max_num_seqs=24 generator.max_num_batched_tokens=65536 generator.enable_prefix_caching=true generator.enable_chunked_prefill=true generator.run_engines_locally=true generator.weight_sync_backend=nccl generator.async_engine=true generator.batched=false generator.enable_http_endpoint=true generator.enable_ray_prometheus_stats=false generator.vllm_stats_interval=1 generator.append_eos_token_after_stop_str_in_multi_turn=true generator.max_turns=999999 generator.sampling_params.max_generate_length=4096 generator.sampling_params.temperature=0.7 generator.sampling_params.top_p=0.95 generator.sampling_params.top_k=20 ++generator.engine_init_kwargs.max_model_len=32768 ++generator.engine_init_kwargs.custom_chat_template_chat_completion_path=chat_templates/qwen3_thinking_acc.jinja2 ++generator.engine_init_kwargs.served_model_name=0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6 data.train_data=["/e/scratch/jureap59/feuer1/tasks/code-contests-noblock"] data.val_data=[] +terminal_bench_config.trials_dir=/e/data1/datasets/playground/ot-baf/a3-rl-DCAgent_code-contests-noblock/a3-rl-DCAgent_code-contests-noblock/trace_jobs +terminal_bench_config.harbor.name=terminus-2 +terminal_bench_config.harbor.max_episodes=999999 +terminal_bench_config.harbor.enable_summarize=false +t
|
|||
|
|
[proxychains] config file found: /e/home/jusers/feuer1/jupiter/.proxychains/proxychains_496419.conf
|
|||
|
|
[proxychains] preloading /e/scratch/jureap59/feuer1/proxychains-ng-aarch64/lib/libproxychains4.so
|
|||
|
|
[proxychains] DLL init: proxychains-ng 4.17-git-9-g78ead0e
|
|||
|
|
2026-05-23 15:16:29.498 | INFO | skyrl_train.utils.utils:prepare_runtime_environment:590 - Exporting wandb api key to ray runtime env
|
|||
|
|
2026-05-23 15:16:29.500 | INFO | skyrl_train.utils.utils:prepare_runtime_environment:609 - Exporting RAY_ADDRESS to ray runtime env
|
|||
|
|
2026-05-23 15:16:29,500 INFO worker.py:1680 -- Using address 10.128.17.210:6379 set in the environment variable RAY_ADDRESS
|
|||
|
|
2026-05-23 15:16:29,536 INFO worker.py:1821 -- Connecting to existing Ray cluster at address: 10.128.17.210:6379...
|
|||
|
|
2026-05-23 15:16:29,547 INFO worker.py:2007 -- Connected to Ray cluster.
|
|||
|
|
/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/ray/_private/worker.py:2046: FutureWarning: Tip: In future versions of Ray, Ray will no longer override accelerator visible devices env var if num_gpus=0 or num_gpus=None (default). To enable this behavior and turn off this error message, set RAY_ACCEL_ENV_VAR_OVERRIDE_ON_ZERO=0
|
|||
|
|
warnings.warn(
|
|||
|
|
[33m(raylet, ip=10.128.24.53)[0m [proxychains] DLL init: proxychains-ng 4.17-git-9-g78ead0e
|
|||
|
|
2026-05-23 15:16:32.151 | INFO | skyrl_train.utils.ppo_utils:sync_registries:546 - Synced registries to ray actor
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:16:41.392[0m | [1mINFO [0m | [36mskyrl_train.entrypoints.main_base[0m:[36m_configure_log_level[0m:[36m199[0m - [1mSkyRL log level set to: INFO[0m
|
|||
|
|
[33m(raylet)[0m [proxychains] DLL init: proxychains-ng 4.17-git-9-g78ead0e[32m [repeated 17x across cluster] (Ray deduplicates logs by default. Set RAY_DEDUP_LOGS=0 to disable log deduplication, or see https://docs.ray.io/en/master/ray-observability/user-guides/configure-logging.html#log-deduplication for more options.)[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m The tokenizer you are loading from '/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue.
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:16:41.632[0m | [1mINFO [0m | [36mexamples.terminal_bench.dataset[0m:[36m_load_data_files[0m:[36m40[0m - [1mLoading data from: /e/scratch/jureap59/feuer1/tasks/code-contests-noblock[0m
|
|||
|
|
[2026-05-23 15:16:59,779 E 3659268 3659688] core_worker_process.cc:842: Failed to establish connection to the metrics exporter agent. Metrics will not be exported. Exporter agent status: RpcError: Running out of retries to initialize the metrics agent. rpc_code: 14
|
|||
|
|
[36m(RegistryActor pid=1440387, ip=10.128.24.53)[0m [2026-05-23 15:17:00,782 E 1440387 1440427] core_worker_process.cc:842: Failed to establish connection to the metrics exporter agent. Metrics will not be exported. Exporter agent status: RpcError: Running out of retries to initialize the metrics agent. rpc_code: 14
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:18:34.306[0m | [1mINFO [0m | [36mexamples.terminal_bench.dataset[0m:[36m_load_data_files[0m:[36m50[0m - [1mFound 8728 valid task directories out of 8728 total directories[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:18:34.306[0m | [1mINFO [0m | [36mexamples.terminal_bench.dataset[0m:[36m__init__[0m:[36m27[0m - [1mTerminalBenchTaskDataset initialized with 8728 task paths[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:18:34.317[0m | [1mINFO [0m | [36mskyrl_train.entrypoints.main_base[0m:[36m_setup_trainer[0m:[36m352[0m - [1mdata:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m train_data:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m - /e/scratch/jureap59/feuer1/tasks/code-contests-noblock
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m val_data: []
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m trainer:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m placement:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m colocate_all: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m colocate_policy_ref: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m policy_num_nodes: 2
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m policy_num_gpus_per_node: 4
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m critic_num_nodes: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m critic_num_gpus_per_node: 4
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ref_num_nodes: 2
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ref_num_gpus_per_node: 4
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m sequence_parallel_backend: ulysses
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m strategy: fsdp2
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m policy:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m model:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m path: /e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m lora:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m rank: 0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m alpha: 16
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m dropout: 0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m lora_sync_path: /tmp/skyrl_lora_sync
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m target_modules: all-linear
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m exclude_modules: null
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m deepspeed_config: ${deepspeed_config.train}
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m optimizer_config:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m optimizer: AdamW
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m lr: 8.0e-06
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m adam_betas:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m - 0.9
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m - 0.999
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m weight_decay: 0.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m max_grad_norm: 0.9
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m offload_after_step: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m num_warmup_steps: 0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m scheduler: constant_with_warmup
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m optimizer_kwargs: {}
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m fsdp_config:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m cpu_offload: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m reshard_after_forward: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m fsdp_size: 4
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m sequence_parallel_size: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m use_torch_compile: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m record_memory: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m megatron_config:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m tensor_model_parallel_size: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m pipeline_model_parallel_size: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m context_parallel_size: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m expert_model_parallel_size: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m expert_tensor_parallel_size: null
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ddp_config:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m grad_reduce_in_fp32: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m overlap_grad_reduce: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m overlap_param_gather: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m average_in_collective: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m model_config_kwargs: {}
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m torch_profiler_config:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m enable: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ranks: []
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m save_path: null
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m optimizer_config_kwargs:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m overlap_cpu_optimizer_d2h_h2d: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m use_precision_aware_optimizer: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m optimizer_cpu_offload: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m optimizer_offload_fraction: 0.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m transformer_config_kwargs:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m recompute_granularity: full
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m recompute_modules:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m - core_attn
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m recompute_method: uniform
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m recompute_num_layers: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m empty_cuda_cache: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ref:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m model:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m path: ${trainer.policy.model.path}
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m sequence_parallel_size: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m deepspeed_config: ${deepspeed_config.eval}
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m fsdp_config:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m cpu_offload: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m reshard_after_forward: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m fsdp_size: 4
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m megatron_config:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m tensor_model_parallel_size: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m pipeline_model_parallel_size: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m context_parallel_size: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m expert_model_parallel_size: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m expert_tensor_parallel_size: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m model_config_kwargs: {}
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m transformer_config_kwargs: {}
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m critic:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m model:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m path: null
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m lora:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m rank: 0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m alpha: 16
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m dropout: 0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m target_modules: all-linear
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m exclude_modules: null
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m deepspeed_config: ${deepspeed_config.train}
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m optimizer_config:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m optimizer: AdamW
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m lr: 5.0e-06
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m adam_betas:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m - 0.9
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m - 0.999
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m weight_decay: 0.01
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m max_grad_norm: 1.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m offload_after_step: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m num_warmup_steps: 0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m scheduler: constant_with_warmup
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m optimizer_kwargs: {}
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m fsdp_config:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m cpu_offload: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m reshard_after_forward: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m fsdp_size: -1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m sequence_parallel_size: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m algorithm:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m advantage_estimator: rloo_n
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m kl_ctrl:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m type: fixed
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m kl_target: 0.1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m horizon: 10000
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m kl_estimator_type: k3
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m use_kl_estimator_k3: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m use_abs_kl: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m use_kl_in_reward: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m use_kl_loss: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m kl_loss_coef: 0.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m use_entropy_loss: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m entropy_loss_coef: 0.01
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m advantage_batch_normalize: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m value_head_prefix: value_head
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m policy_loss_type: regular
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m loss_reduction: token_mean
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m grpo_norm_by_std: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m rloo_n_min_group_size: 4
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m rloo_n_filter_zero_reward_groups: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m lambd: 1.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m gamma: 1.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m eps_clip_low: 0.2
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m eps_clip_high: 0.05
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m clip_ratio_c: 3.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m tis_imp_ratio_cap: -1.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m use_tis: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m sapo:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m tau_pos: 1.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m tau_neg: 1.05
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m value_clip: 0.2
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m dynamic_sampling:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m type: null
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m max_sample_batches: 30
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m min_replace_ratio: 0.3
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m clip_cov:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m clip_ratio: 0.0002
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m clip_cov_lb: 1.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m clip_cov_ub: 5.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m kl_cov:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m kl_cov_frac: 0.2
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ppo_kl_coef: 1.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m cispo:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m cispo_eps_clip_low: 0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m cispo_eps_clip_high: 5
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m z_clip:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m enabled: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m alpha: 0.97
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m z_thresh: 2.5
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m warmup_steps: 25
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m mode: zscore
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m clip_option: adaptive_scaling
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m clip_factor: 1.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m skip_update_on_spike: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m stale_clip:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m enabled: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m alpha: 0.3
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m entropy_threshold: 0.15
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m entropy_window: 10
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m min_lr_scale: 0.1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m max_seq_len: 1004095
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m fully_async:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m max_staleness_steps: 16
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m num_parallel_generation_workers: 450
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m gradient_checkpointing: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m gradient_checkpointing_use_reentrant: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m seed: 42
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m resume_mode: latest
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m resume_path: null
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ckpt_path: /e/data1/datasets/playground/ot-baf/a3-rl-DCAgent_code-contests-noblock/a3-rl-DCAgent_code-contests-noblock/checkpoints
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m max_ckpts_to_keep: -1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ckpt_interval: 2
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m hf_save_interval: 5
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m export_path: /e/data1/datasets/playground/ot-baf/a3-rl-DCAgent_code-contests-noblock/a3-rl-DCAgent_code-contests-noblock/exports
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m bf16: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m epochs: 2
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m max_steps: 80
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m update_epochs_per_batch: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m train_batch_size: 64
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m policy_mini_batch_size: 64
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m critic_mini_batch_size: 256
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m micro_train_batch_size_per_gpu: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m micro_forward_batch_size_per_gpu: 4
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m update_ref_every_epoch: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m use_sample_packing: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m eval_batch_size: 64
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m eval_before_train: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m eval_interval: 999999
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m max_prompt_length: 999999
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m flash_attn: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m disable_fast_tokenizer: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m target_modules: null
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m exclude_modules: null
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m project_name: OpenThoughts-Agent
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m run_name: a3-rl-DCAgent_code-contests-noblock
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m logger: console
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m tracker_commit_each_step: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m dump_data_batch: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m dump_eval_results: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m log_level: INFO
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m rope_scaling: null
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m rope_theta: null
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m step_wise_training: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m hf_hub_repo_id: laion/a3-rl-DCAgent_code-contests-noblock
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m hf_hub_private: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m hf_hub_revision: main
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m enable_db_registration: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m generator:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m model_name: ${trainer.policy.model.path}
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m model_dtype: bfloat16
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m timeout_multiplier: 1.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m run_engines_locally: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m num_inference_engines: 48
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m backend: vllm
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m weight_sync_backend: nccl
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m fuse_weights: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m weight_transfer_threshold_cuda_ipc_GB: 1.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m inference_engine_tensor_parallel_size: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m inference_engine_pipeline_parallel_size: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m inference_engine_expert_parallel_size: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m inference_engine_data_parallel_size: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m n_samples_per_prompt: 8
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m async_engine: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m batched: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m max_input_length: ${trainer.max_prompt_length}
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m vllm_v1_disable_multiproc: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m enable_prefix_caching: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m enable_chunked_prefill: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m max_num_batched_tokens: 65536
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m enforce_eager: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m fully_sharded_loras: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m enable_ray_prometheus_stats: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m vllm_stats_interval: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m gpu_memory_utilization: 0.75
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m max_num_seqs: 24
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m remote_inference_engine_urls:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m - 127.0.0.1:8001
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m enable_http_endpoint: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m http_endpoint_host: 127.0.0.1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m http_endpoint_port: 8000
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m max_turns: 999999
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m chat_template:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m source: name
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m name_or_path: null
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m chat_template_kwargs: {}
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m engine_init_kwargs:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m max_model_len: 32768
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m custom_chat_template_chat_completion_path: chat_templates/qwen3_thinking_acc.jinja2
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m served_model_name: 0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m override_existing_update_group: disable
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m sampling_params:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m max_generate_length: 4096
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m repetition_penalty: 1.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m temperature: 0.7
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m top_p: 0.95
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m min_p: 0.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m top_k: 20
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m logprobs: null
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m stop: null
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m use_conversation_multi_turn: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m append_eos_token_after_stop_str_in_multi_turn: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m eval_sampling_params:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m max_generate_length: ${generator.sampling_params.max_generate_length}
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m repetition_penalty: 1.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m temperature: 0.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m top_p: 1.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m min_p: 0.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m top_k: -1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m logprobs: null
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m stop: null
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m eval_n_samples_per_prompt: 8
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m zero_reward_on_non_stop: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m apply_overlong_filtering: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m rope_scaling: ${trainer.rope_scaling}
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m rope_theta: ${trainer.rope_theta}
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m teacher:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m model_path: null
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m top_k_logprobs: 256
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m num_inference_engines: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m inference_engine_tensor_parallel_size: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m inference_engine_pipeline_parallel_size: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m gpu_memory_utilization: 0.9
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m enforce_eager: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m backend: vllm
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m engine_init_kwargs: {}
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m environment:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m env_class: gsm8k
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m skyrl_gym:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m max_env_workers: 32
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m text2sql:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m db_path: /home/ray/default/sql_data
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m llm_as_a_judge:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m model: gpt-4o-mini
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m base_url: null
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m search:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m log_requests: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m search_url: http://127.0.0.1:8000/retrieve
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m topk: 3
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m timeout: 30
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m deepspeed_config:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m train:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m zero_optimization:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m stage: 3
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m offload_param:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m device: none
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m offload_optimizer:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m device: none
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m pin_memory: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m sub_group_size: auto
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m reduce_bucket_size: auto
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m stage3_param_persistence_threshold: auto
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m stage3_prefetch_bucket_size: auto
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m stage3_max_live_parameters: auto
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m stage3_max_reuse_distance: auto
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m round_robin_gradients: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m zero_hpz_partition_size: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m zero_quantized_weights: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m zero_quantized_gradients: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m torch_autocast:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m enabled: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m dtype: bfloat16
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m disable_trace_cache: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m data_types:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m grad_accum_dtype: fp32
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m gradient_clipping: 1.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m wall_clock_breakdown: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m prescale_gradient: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m eval:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m zero_optimization:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m stage: 3
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m stage3_param_persistence_threshold: auto
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m offload_param:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m device: cpu
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m pin_memory: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m torch_autocast:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m enabled: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m dtype: bfloat16
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m gradient_clipping: 1.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m prescale_gradient: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m wall_clock_breakdown: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m terminal_bench_config:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m trials_dir: /e/data1/datasets/playground/ot-baf/a3-rl-DCAgent_code-contests-noblock/a3-rl-DCAgent_code-contests-noblock/trace_jobs
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m harbor:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m name: terminus-2
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m max_episodes: 999999
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m enable_summarize: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m store_all_messages: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m trajectory_config:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m raw_content: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m enable_episode_logging: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m record_terminal_session: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m enable_pane_logging: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m strict_json_parser: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m interleaved_thinking: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m extra_body:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m chat_template_kwargs:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m enable_thinking: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m override_timeout_sec: 1800
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m override_cpus: 1
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m override_memory_mb: 2048
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m override_storage_mb: 2048
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m auto_snapshot: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m verifier_override_timeout_sec: 120
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m max_retries: 3
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m min_wait_sec: 60.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m max_wait_sec: 600.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m wait_multiplier: 2.0
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m exclude_exceptions:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m - VerifierTimeoutError
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m - VerifierRuntimeError
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m - RewardFileNotFoundError
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m - RewardFileEmptyError
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m - VerifierOutputParseError
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m n_concurrent_trials: 900
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m log_level: INFO
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m enable_reward_shaping: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m enable_error_classification: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m mask_exceptions:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m - DaytonaError
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m - EnvironmentStartTimeoutError
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m - NetworkError
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m - ConnectionError
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m - RewardFileNotFoundError
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m - RewardFileEmptyError
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m - AgentEnvironmentTimeoutError
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m default_error_treatment: zero
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m passthrough_exceptions:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m - AgentTimeoutError
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m - ContextLengthExceededError
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m model_info:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m max_input_tokens: 32768
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m max_output_tokens: 4096
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m archiving:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m enabled: false
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m trace_upload:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m enabled: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m repo_org: DCAgent
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m episodes: last
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m dataset_type: SFT
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m cleanup: true
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [2026-05-23 15:17:03,082 E 3659716 3659759] core_worker_process.cc:842: Failed to establish connection to the metrics exporter agent. Metrics will not be exported. Exporter agent status: RpcError: Running out of retries to initialize the metrics agent. rpc_code: 14[32m [repeated 2x across cluster][0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/__init__.py:7: RuntimeWarning: Failed to read commit hash:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m No module named 'vllm._version'
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m from .version import __version__, __version_tuple__ # isort:skip
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m W0523 15:18:39.734000 3659716 envs/rl/lib/python3.12/site-packages/torch/utils/cpp_extension.py:117] No CUDA runtime is found, using CUDA_HOME='/e/software/default/stages/2026/software/CUDA/13'
|
|||
|
|
[33m(raylet)[0m [proxychains] DLL init: proxychains-ng 4.17-git-9-g78ead0e
|
|||
|
|
[33m(raylet, ip=10.128.24.51)[0m [proxychains] DLL init: proxychains-ng 4.17-git-9-g78ead0e[32m [repeated 41x across cluster][0m
|
|||
|
|
[36m(pid=3294862, ip=10.128.17.219)[0m /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/__init__.py:7: RuntimeWarning: Failed to read commit hash:
|
|||
|
|
[36m(pid=3294862, ip=10.128.17.219)[0m No module named 'vllm._version'
|
|||
|
|
[36m(pid=3294862, ip=10.128.17.219)[0m from .version import __version__, __version_tuple__ # isort:skip
|
|||
|
|
[33m(raylet, ip=10.128.17.216)[0m [proxychains] DLL init: proxychains-ng 4.17-git-9-g78ead0e[32m [repeated 57x across cluster][0m
|
|||
|
|
[36m(pid=2397096, ip=10.128.24.51)[0m /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/__init__.py:7: RuntimeWarning: Failed to read commit hash:[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(pid=2397096, ip=10.128.24.51)[0m No module named 'vllm._version'[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(pid=2397096, ip=10.128.24.51)[0m from .version import __version__, __version_tuple__ # isort:skip[32m [repeated 8x across cluster][0m
|
|||
|
|
[33m(raylet, ip=10.128.17.220)[0m [proxychains] DLL init: proxychains-ng 4.17-git-9-g78ead0e[32m [repeated 101x across cluster][0m
|
|||
|
|
[36m(pid=3123494, ip=10.128.17.215)[0m /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/__init__.py:7: RuntimeWarning: Failed to read commit hash:[32m [repeated 15x across cluster][0m
|
|||
|
|
[36m(pid=3123494, ip=10.128.17.215)[0m No module named 'vllm._version'[32m [repeated 15x across cluster][0m
|
|||
|
|
[36m(pid=3123494, ip=10.128.17.215)[0m from .version import __version__, __version_tuple__ # isort:skip[32m [repeated 15x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m 2026-05-23 15:18:59.704 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:setup_envvars_for_vllm:123 - setup_envvars_for_vllm: distributed_executor_backend=uni, SKYRL_ENABLE_NUMA_AFFINITY=1, CUDA_VISIBLE_DEVICES=0, VLLM_ENABLE_V1_MULTIPROCESSING=<unset>
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m 2026-05-23 15:18:59.706 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:setup_envvars_for_vllm:133 - setup_envvars_for_vllm: numa_enabled=True
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m 2026-05-23 15:18:59.706 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:setup_envvars_for_vllm:136 - setup_envvars_for_vllm: set VLLM_ENABLE_V1_MULTIPROCESSING=0 for NUMA affinity
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m 2026-05-23 15:19:00.677 | INFO | skyrl_train.utils.numa:set_numa_affinity_for_gpu:340 - NUMA affinity: bound process to CPUs 0-71 (NUMA node 0) for GPU 0
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m 2026-05-23 15:19:00.999 | DEBUG | skyrl_train.utils.numa:_set_membind_via_libnuma:375 - NUMA affinity: set memory preferred to NUMA node 0
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m 2026-05-23 15:19:00.999 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:__init__:481 - BaseVLLMInferenceEngine: vllm_v1_disable_multiproc=True, vllm.__version__=dev, VLLM_ENABLE_V1_MULTIPROCESSING=0
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m 2026-05-23 15:19:00.999 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:__init__:489 - BaseVLLMInferenceEngine: set VLLM_ENABLE_V1_MULTIPROCESSING=0
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m 2026-05-23 15:19:00.999 | WARNING | skyrl_train.inference_engines.vllm.vllm_engine:_create_engine:1047 - OpenAI API sampling params overridden: temperature=0.7, top_p=0.95, top_k=20
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m 2026-05-23 15:19:01.027 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:_create_engine:1099 - Engine startup stagger: sleeping 1.88s (attempt 1/5) to avoid port collisions
|
|||
|
|
[33m(raylet, ip=10.128.24.52)[0m [proxychains] DLL init: proxychains-ng 4.17-git-9-g78ead0e[32m [repeated 64x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m The argument `trust_remote_code` is to be used with Auto classes. It has no effect here and is ignored.
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m The argument `trust_remote_code` is to be used with Auto classes. It has no effect here and is ignored.
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m The tokenizer you are loading from '/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue.
|
|||
|
|
[36m(pid=1440680, ip=10.128.24.53)[0m /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/__init__.py:7: RuntimeWarning: Failed to read commit hash:[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(pid=1440680, ip=10.128.24.53)[0m No module named 'vllm._version'[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(pid=1440680, ip=10.128.24.53)[0m from .version import __version__, __version_tuple__ # isort:skip[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523615, ip=10.128.17.222)[0m 2026-05-23 15:19:04.230 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:setup_envvars_for_vllm:123 - setup_envvars_for_vllm: distributed_executor_backend=uni, SKYRL_ENABLE_NUMA_AFFINITY=1, CUDA_VISIBLE_DEVICES=3, VLLM_ENABLE_V1_MULTIPROCESSING=<unset>[32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523615, ip=10.128.17.222)[0m 2026-05-23 15:19:04.288 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:setup_envvars_for_vllm:133 - setup_envvars_for_vllm: numa_enabled=True[32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523615, ip=10.128.17.222)[0m 2026-05-23 15:19:04.288 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:setup_envvars_for_vllm:136 - setup_envvars_for_vllm: set VLLM_ENABLE_V1_MULTIPROCESSING=0 for NUMA affinity[32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523615, ip=10.128.17.222)[0m 2026-05-23 15:19:05.504 | INFO | skyrl_train.utils.numa:set_numa_affinity_for_gpu:340 - NUMA affinity: bound process to CPUs 216-287 (NUMA node 3) for GPU 3[32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523615, ip=10.128.17.222)[0m 2026-05-23 15:19:05.792 | DEBUG | skyrl_train.utils.numa:_set_membind_via_libnuma:375 - NUMA affinity: set memory preferred to NUMA node 3[32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523615, ip=10.128.17.222)[0m 2026-05-23 15:19:05.793 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:__init__:481 - BaseVLLMInferenceEngine: vllm_v1_disable_multiproc=True, vllm.__version__=dev, VLLM_ENABLE_V1_MULTIPROCESSING=0[32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523615, ip=10.128.17.222)[0m 2026-05-23 15:19:05.793 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:__init__:489 - BaseVLLMInferenceEngine: set VLLM_ENABLE_V1_MULTIPROCESSING=0[32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523615, ip=10.128.17.222)[0m 2026-05-23 15:19:05.793 | WARNING | skyrl_train.inference_engines.vllm.vllm_engine:_create_engine:1047 - OpenAI API sampling params overridden: temperature=0.7, top_p=0.95, top_k=20[32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523615, ip=10.128.17.222)[0m 2026-05-23 15:19:05.823 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:_create_engine:1099 - Engine startup stagger: sleeping 2.65s (attempt 1/5) to avoid port collisions[32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [2026-05-23 15:19:06] INFO inference_engine_client_http_endpoint.py:350: Starting server on 127.0.0.1:8000
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [2026-05-23 15:19:06] INFO inference_engine_client_http_endpoint.py:242: Starting inference HTTP endpoint...
|
|||
|
|
[33m(raylet, ip=10.128.17.214)[0m [proxychains] DLL init: proxychains-ng 4.17-git-9-g78ead0e[32m [repeated 118x across cluster][0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [2026-05-23 15:19:07] INFO inference_engine_client_http_endpoint.py:229: Server ready after 2 attempts (2 seconds)
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:07.125[0m | [1mINFO [0m | [36mskyrl_train.inference_engines.inference_engine_client[0m:[36m_spin_up_http_endpoint[0m:[36m960[0m - [1mInferenceEngineClient HTTP endpoint started on 127.0.0.1:8000[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:07.125[0m | [1mINFO [0m | [36mskyrl_train.inference_engines.inference_engine_client[0m:[36m__init__[0m:[36m61[0m - [1mInferenceEngineClient initialized with 48 engines.[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:07.127[0m | [1mINFO [0m | [36mexamples.terminal_bench.terminal_bench_generator[0m:[36m_configure_harbor_logging[0m:[36m213[0m - [1mHarbor logging level set to INFO[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:07.128[0m | [1mINFO [0m | [36mexamples.terminal_bench.terminal_bench_generator[0m:[36m__init__[0m:[36m113[0m - [1mTerminalBenchGenerator initialized with HarborConfigBuilder. Exposed fields: ['name', 'max_episodes', 'enable_summarize', 'store_all_messages', 'trajectory_config', 'enable_episode_logging', 'record_terminal_session', 'enable_pane_logging', 'strict_json_parser', 'interleaved_thinking', 'extra_body', 'override_timeout_sec', 'override_cpus', 'override_memory_mb', 'override_storage_mb', 'auto_snapshot', 'verifier_override_timeout_sec', 'max_retries', 'min_wait_sec', 'max_wait_sec', 'wait_multiplier', 'exclude_exceptions', 'n_concurrent_trials', 'log_level', 'enable_reward_shaping', 'enable_error_classification', 'mask_exceptions', 'default_error_treatment', 'passthrough_exceptions']. Retry config: max_retries=3, backoff=60.0-600.0s. Concurrent trials: 900. Reward shaping: enabled=False, shaper=pass_ratio. Error classification: enabled=True[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:07.130[0m | [1mINFO [0m | [36mexamples.terminal_bench.terminal_bench_generator[0m:[36m__init__[0m:[36m129[0m - [1mTerminalBenchGenerator initialized with custom chat template read from: chat_templates/qwen3_thinking_acc.jinja2[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:07.131[0m | [1mINFO [0m | [36mskyrl_train.utils.trainer_utils[0m:[36mbuild_dataloader[0m:[36m656[0m - [1mTotal steps: 272[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:07.131[0m | [1mINFO [0m | [36mskyrl_train.fully_async_trainer[0m:[36m_build_train_dataloader_and_compute_training_steps[0m:[36m351[0m - [1mLength of train_dataloader: 8728[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:07.131[0m | [1mINFO [0m | [36mskyrl_train.fully_async_trainer[0m:[36m_build_train_dataloader_and_compute_training_steps[0m:[36m352[0m - [1mNumber of steps per epoch: 136[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:07.131[0m | [1mINFO [0m | [36mskyrl_train.fully_async_trainer[0m:[36m_build_train_dataloader_and_compute_training_steps[0m:[36m353[0m - [1mTotal training steps: 80[0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523614, ip=10.128.17.222)[0m The argument `trust_remote_code` is to be used with Auto classes. It has no effect here and is ignored.[32m [repeated 18x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523614, ip=10.128.17.222)[0m The tokenizer you are loading from '/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue.[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523614, ip=10.128.17.222)[0m /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/__init__.py:7: RuntimeWarning: Failed to read commit hash:[32m [repeated 22x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523614, ip=10.128.17.222)[0m No module named 'vllm._version'[32m [repeated 22x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523614, ip=10.128.17.222)[0m from .version import __version__, __version_tuple__ # isort:skip[32m [repeated 22x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3222225, ip=10.128.17.212)[0m 2026-05-23 15:19:10.098 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:setup_envvars_for_vllm:123 - setup_envvars_for_vllm: distributed_executor_backend=uni, SKYRL_ENABLE_NUMA_AFFINITY=1, CUDA_VISIBLE_DEVICES=0, VLLM_ENABLE_V1_MULTIPROCESSING=<unset>[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3222225, ip=10.128.17.212)[0m 2026-05-23 15:19:10.100 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:setup_envvars_for_vllm:133 - setup_envvars_for_vllm: numa_enabled=True[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3222225, ip=10.128.17.212)[0m 2026-05-23 15:19:10.100 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:setup_envvars_for_vllm:136 - setup_envvars_for_vllm: set VLLM_ENABLE_V1_MULTIPROCESSING=0 for NUMA affinity[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(pid=2753436, ip=10.128.17.217)[0m Using blocking ray.get inside async actor. This blocks the event loop. Please use `await` on object ref with asyncio.gather if you want to yield execution to the event loop instead.
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909165, ip=10.128.17.216)[0m 2026-05-23 15:19:08.832 | INFO | skyrl_train.utils.numa:set_numa_affinity_for_gpu:340 - NUMA affinity: bound process to CPUs 216-287 (NUMA node 3) for GPU 3[32m [repeated 4x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200625, ip=10.128.17.221)[0m 2026-05-23 15:19:10.876 | DEBUG | skyrl_train.utils.numa:_set_membind_via_libnuma:375 - NUMA affinity: set memory preferred to NUMA node 0[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200625, ip=10.128.17.221)[0m 2026-05-23 15:19:10.877 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:__init__:481 - BaseVLLMInferenceEngine: vllm_v1_disable_multiproc=True, vllm.__version__=dev, VLLM_ENABLE_V1_MULTIPROCESSING=0[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200625, ip=10.128.17.221)[0m 2026-05-23 15:19:10.877 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:__init__:489 - BaseVLLMInferenceEngine: set VLLM_ENABLE_V1_MULTIPROCESSING=0[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200625, ip=10.128.17.221)[0m 2026-05-23 15:19:10.877 | WARNING | skyrl_train.inference_engines.vllm.vllm_engine:_create_engine:1047 - OpenAI API sampling params overridden: temperature=0.7, top_p=0.95, top_k=20[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200625, ip=10.128.17.221)[0m 2026-05-23 15:19:10.898 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:_create_engine:1099 - Engine startup stagger: sleeping 2.48s (attempt 1/5) to avoid port collisions[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(bundle_reservation_check_func pid=3659816)[0m [2026-05-23 15:19:11,518 E 3659816 3659856] core_worker_process.cc:842: Failed to establish connection to the metrics exporter agent. Metrics will not be exported. Exporter agent status: RpcError: Running out of retries to initialize the metrics agent. rpc_code: 14
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909165, ip=10.128.17.216)[0m [proxychains] DLL init: proxychains-ng 4.17-git-9-g78ead0e[32m [repeated 21x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3222354, ip=10.128.17.212)[0m The argument `trust_remote_code` is to be used with Auto classes. It has no effect here and is ignored.[32m [repeated 17x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3123358, ip=10.128.17.215)[0m The tokenizer you are loading from '/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue.[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909166, ip=10.128.17.216)[0m /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/__init__.py:7: RuntimeWarning: Failed to read commit hash:[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909166, ip=10.128.17.216)[0m No module named 'vllm._version'[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909166, ip=10.128.17.216)[0m from .version import __version__, __version_tuple__ # isort:skip[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396962, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397192) /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/__init__.py:1617: UserWarning: Please use the new API settings to control TF32 behavior, such as torch.backends.cudnn.conv.fp32_precision = 'tf32' or torch.backends.cuda.matmul.fp32_precision = 'ieee'. Old settings, e.g, torch.backends.cuda.matmul.allow_tf32 = True, torch.backends.cudnn.allow_tf32 = True, allowTF32CuDNN() and allowTF32CuBLAS() will be deprecated after Pytorch 2.9. Please see https://pytorch.org/docs/main/notes/cuda.html#tensorfloat-32-tf32-on-ampere-and-later-devices (Triggered internally at /pytorch/aten/src/ATen/Context.cpp:80.)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396962, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397192) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440680, ip=10.128.24.53)[0m 2026-05-23 15:19:14.386 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:setup_envvars_for_vllm:123 - setup_envvars_for_vllm: distributed_executor_backend=uni, SKYRL_ENABLE_NUMA_AFFINITY=1, CUDA_VISIBLE_DEVICES=3, VLLM_ENABLE_V1_MULTIPROCESSING=<unset>[32m [repeated 19x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440680, ip=10.128.24.53)[0m 2026-05-23 15:19:14.387 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:setup_envvars_for_vllm:133 - setup_envvars_for_vllm: numa_enabled=True[32m [repeated 19x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440680, ip=10.128.24.53)[0m 2026-05-23 15:19:14.387 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:setup_envvars_for_vllm:136 - setup_envvars_for_vllm: set VLLM_ENABLE_V1_MULTIPROCESSING=0 for NUMA affinity[32m [repeated 19x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440680, ip=10.128.24.53)[0m 2026-05-23 15:19:15.611 | INFO | skyrl_train.utils.numa:set_numa_affinity_for_gpu:340 - NUMA affinity: bound process to CPUs 216-287 (NUMA node 3) for GPU 3[32m [repeated 20x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396963, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397201) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=372514, ip=10.128.17.220)[0m 2026-05-23 15:19:14.277 | DEBUG | skyrl_train.utils.numa:_set_membind_via_libnuma:375 - NUMA affinity: set memory preferred to NUMA node 3[32m [repeated 15x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=372514, ip=10.128.17.220)[0m 2026-05-23 15:19:14.277 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:__init__:481 - BaseVLLMInferenceEngine: vllm_v1_disable_multiproc=True, vllm.__version__=dev, VLLM_ENABLE_V1_MULTIPROCESSING=0[32m [repeated 15x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=372514, ip=10.128.17.220)[0m 2026-05-23 15:19:14.277 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:__init__:489 - BaseVLLMInferenceEngine: set VLLM_ENABLE_V1_MULTIPROCESSING=0[32m [repeated 15x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=372514, ip=10.128.17.220)[0m 2026-05-23 15:19:14.277 | WARNING | skyrl_train.inference_engines.vllm.vllm_engine:_create_engine:1047 - OpenAI API sampling params overridden: temperature=0.7, top_p=0.95, top_k=20[32m [repeated 15x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=372514, ip=10.128.17.220)[0m 2026-05-23 15:19:14.309 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:_create_engine:1099 - Engine startup stagger: sleeping 1.81s (attempt 1/5) to avoid port collisions[32m [repeated 15x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396835, ip=10.128.24.51)[0m [2026-05-23 15:19:16,018 E 2396835 2396937] core_worker_process.cc:842: Failed to establish connection to the metrics exporter agent. Metrics will not be exported. Exporter agent status: RpcError: Running out of retries to initialize the metrics agent. rpc_code: 14[32m [repeated 6x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=372386, ip=10.128.17.220)[0m [proxychains] DLL init: proxychains-ng 4.17-git-9-g78ead0e[32m [repeated 34x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2397096, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397216) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396963, ip=10.128.24.51)[0m [W523 15:19:17.519222234 socket.cpp:767] [c10d] The client socket cannot be initialized to connect to [jpbo-022-03-interconnect-1.jupiter.internal]:40421 (errno: 97 - Address family not supported by protocol).
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396963, ip=10.128.24.51)[0m [W523 15:19:17.519601842 Utils.hpp:166] Warning: Environment variable NCCL_BLOCKING_WAIT is deprecated; use TORCH_NCCL_BLOCKING_WAIT instead (function operator())
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396963, ip=10.128.24.51)[0m [rank0]:[W523 15:19:17.521578185 Utils.hpp:137] Warning: Environment variable TORCH_NCCL_TRACE_BUFFER_SIZE is deprecated; use TORCH_FR_BUFFER_SIZE instead (function operator())
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396835, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397197) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m 2026-05-23 15:19:18 INFO [ipv4-debug] hostname=jpbo-010-41.jupiter.internal
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m 2026-05-23 15:19:18 INFO [ipv4-debug] _global_node.node_ip_address=10.128.17.217
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m 2026-05-23 15:19:18 INFO [ipv4-debug] get_node_ip_address()=10.128.17.217
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:18.138[0m | [1m[32mINFO [0m | [36mskyrl_train.workers.worker[0m:[36m_initiate_actors[0m:[36m496[0m - [1m[32mInitializing process group for RayActorGroup[0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440679, ip=10.128.24.53)[0m The argument `trust_remote_code` is to be used with Auto classes. It has no effect here and is ignored.[32m [repeated 29x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440681, ip=10.128.24.53)[0m The tokenizer you are loading from '/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue.[32m [repeated 15x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=372386, ip=10.128.17.220)[0m /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/__init__.py:7: RuntimeWarning: Failed to read commit hash:[32m [repeated 16x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=372386, ip=10.128.17.220)[0m No module named 'vllm._version'[32m [repeated 16x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=372386, ip=10.128.17.220)[0m from .version import __version__, __version_tuple__ # isort:skip[32m [repeated 16x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294996, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295247) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294998, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295229) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295221) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396963, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397201)
Loading safetensors checkpoint shards: 0% Completed | 0/4 [00:00<?, ?it/s]
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) Process EngineCore_DP0:
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) Traceback (most recent call last):
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/multiprocessing/process.py", line 314, in _bootstrap
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) self.run()
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/multiprocessing/process.py", line 108, in run
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) self._target(*self._args, **self._kwargs)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core.py", line 1010, in run_engine_core
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) raise e
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core.py", line 996, in run_engine_core
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) engine_core = EngineCoreProc(*args, engine_index=dp_rank, **kwargs)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core.py", line 740, in __init__
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) super().__init__(
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core.py", line 106, in __init__
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) self.model_executor = executor_class(vllm_config)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/executor/abstract.py", line 103, in __init__
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) self._init_executor()
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/executor/uniproc_executor.py", line 47, in _init_executor
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) self.driver_worker.init_device()
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/worker/worker_base.py", line 332, in init_device
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) self.worker.init_device() # type: ignore
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/worker/gpu_worker.py", line 232, in init_device
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) init_worker_distributed_environment(
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/worker/gpu_worker.py", line 1040, in init_worker_distributed_environment
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) init_distributed_environment(
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/distributed/parallel_state.py", line 1256, in init_distributed_environment
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) torch.distributed.init_process_group(
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/distributed/c10d_logger.py", line 81, in wrapper
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) return func(*args, **kwargs)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/distributed/c10d_logger.py", line 95, in wrapper
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) func_return = func(*args, **kwargs)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py", line 1762, in init_process_group
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) store, rank, world_size = next(rendezvous_iterator)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/distributed/rendezvous.py", line 230, in _tcp_rendezvous_handler
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) store = _create_c10d_store(
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/distributed/rendezvous.py", line 198, in _create_c10d_store
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) return TCPStore(
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) torch.distributed.DistNetworkError: The server socket has failed to listen on any local network address. port: 34697, useIpv6: false, code: -98, name: EADDRINUSE, message: address already in use
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) Exception raised from makeWithPort at /pytorch/torch/csrc/distributed/c10d/TCPStoreLibUvBackend.cpp:307 (most recent call first):
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) + 0xb0 (0x40003896c700 in /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/lib/libc10.so)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #1: <unknown function> + 0x5f29220 (0x400018b09220 in /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #2: <unknown function> + 0x5f4326c (0x400018b2326c in /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #3: <unknown function> + 0x5f49074 (0x400018b29074 in /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #4: <unknown function> + 0x5f49138 (0x400018b29138 in /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #5: <unknown function> + 0x5f2ccc4 (0x400018b0ccc4 in /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #6: c10d::TCPStore::TCPStore(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, c10d::TCPStoreOptions const&) + 0x104 (0x400018b11564 in /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #7: <unknown function> + 0x109a094 (0x4000128ba094 in /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/lib/libtorch_python.so)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #8: <unknown function> + 0x113236c (0x40001295236c in /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/lib/libtorch_python.so)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #9: <unknown function> + 0x5d6d60 (0x400011df6d60 in /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/lib/libtorch_python.so)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #10: <unknown function> + 0x1b7a38 (0xaaaaca047a38 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #11: _PyObject_MakeTpCall + 0x98 (0xaaaac9ff5db8 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #12: <unknown function> + 0x169f50 (0xaaaac9ff9f50 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #13: <unknown function> + 0x1682e4 (0xaaaac9ff82e4 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #14: <unknown function> + 0x1e0ce8 (0xaaaaca070ce8 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #15: <unknown function> + 0x1d7ddc (0xaaaaca067ddc in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #16: <unknown function> + 0x646b0c (0x400011e66b0c in /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/lib/libtorch_python.so)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #17: _PyObject_MakeTpCall + 0x98 (0xaaaac9ff5db8 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #18: _PyEval_EvalFrameDefault + 0x280c (0xaaaaca0fae54 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #19: <unknown function> + 0x1808c0 (0xaaaaca0108c0 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #20: <unknown function> + 0x182bf8 (0xaaaaca012bf8 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #21: <unknown function> + 0x25fd30 (0xaaaaca0efd30 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #22: <unknown function> + 0x1b7d20 (0xaaaaca047d20 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #23: PyObject_Vectorcall + 0x54 (0xaaaac9ff60e4 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #24: _PyEval_EvalFrameDefault + 0x280c (0xaaaaca0fae54 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #25: _PyObject_FastCallDictTstate + 0x80 (0xaaaac9ff7f20 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #26: _PyObject_Call_Prepend + 0x140 (0xaaaac9ff81ec in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #27: <unknown function> + 0x1e0d80 (0xaaaaca070d80 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #28: <unknown function> + 0x1d7ddc (0xaaaaca067ddc in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #29: _PyObject_MakeTpCall + 0x98 (0xaaaac9ff5db8 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #30: _PyEval_EvalFrameDefault + 0x280c (0xaaaaca0fae54 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #31: _PyObject_FastCallDictTstate + 0x10c (0xaaaac9ff7fac in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #32: _PyObject_Call_Prepend + 0x140 (0xaaaac9ff81ec in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #33: <unknown function> + 0x1e0d80 (0xaaaaca070d80 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #34: <unknown function> + 0x1d7ddc (0xaaaaca067ddc in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #35: _PyObject_Call + 0x68 (0xaaaac9ff8488 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #36: _PyEval_EvalFrameDefault + 0x52ac (0xaaaaca0fd8f4 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #37: PyEval_EvalCode + 0xb4 (0xaaaaca102eb4 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #38: <unknown function> + 0x2ccdcc (0xaaaaca15cdcc in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #39: <unknown function> + 0x2ccef4 (0xaaaaca15cef4 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #40: PyRun_StringFlags + 0x90 (0xaaaaca161050 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #41: PyRun_SimpleStringFlags + 0x58 (0xaaaaca1610f8 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #42: Py_RunMain + 0x2c8 (0xaaaaca189190 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #43: Py_BytesMain + 0x64 (0xaaaaca189fb4 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #44: <unknown function> + 0x27540 (0x400000e67540 in /lib64/libc.so.6)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #45: __libc_start_main + 0x98 (0x400000e67618 in /lib64/libc.so.6)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) frame #46: <unknown function> + 0x10e0c0 (0xaaaac9f9e0c0 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/__init__.py:1617: UserWarning: Please use the new API settings to control TF32 behavior, such as torch.backends.cudnn.conv.fp32_precision = 'tf32' or torch.backends.cuda.matmul.fp32_precision = 'ieee'. Old settings, e.g, torch.backends.cuda.matmul.allow_tf32 = True, torch.backends.cudnn.allow_tf32 = True, allowTF32CuDNN() and allowTF32CuBLAS() will be deprecated after Pytorch 2.9. Please see https://pytorch.org/docs/main/notes/cuda.html#tensorfloat-32-tf32-on-ampere-and-later-devices (Triggered internally at /pytorch/aten/src/ATen/Context.cpp:80.)[32m [repeated 7x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172774, ip=10.128.24.52)[0m 2026-05-23 15:19:16.152 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:setup_envvars_for_vllm:123 - setup_envvars_for_vllm: distributed_executor_backend=uni, SKYRL_ENABLE_NUMA_AFFINITY=1, CUDA_VISIBLE_DEVICES=3, VLLM_ENABLE_V1_MULTIPROCESSING=<unset>[32m [repeated 4x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172774, ip=10.128.24.52)[0m 2026-05-23 15:19:16.152 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:setup_envvars_for_vllm:133 - setup_envvars_for_vllm: numa_enabled=True[32m [repeated 4x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172774, ip=10.128.24.52)[0m 2026-05-23 15:19:16.152 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:setup_envvars_for_vllm:136 - setup_envvars_for_vllm: set VLLM_ENABLE_V1_MULTIPROCESSING=0 for NUMA affinity[32m [repeated 4x across cluster][0m
|
|||
|
|
[36m(pid=2753511, ip=10.128.17.217)[0m Using blocking ray.get inside async actor. This blocks the event loop. Please use `await` on object ref with asyncio.gather if you want to yield execution to the event loop instead.
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523614, ip=10.128.17.222)[0m (EngineCore_DP0 pid=523846) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172774, ip=10.128.24.52)[0m 2026-05-23 15:19:17.465 | INFO | skyrl_train.utils.numa:set_numa_affinity_for_gpu:340 - NUMA affinity: bound process to CPUs 216-287 (NUMA node 3) for GPU 3[32m [repeated 4x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172774, ip=10.128.24.52)[0m 2026-05-23 15:19:17.766 | DEBUG | skyrl_train.utils.numa:_set_membind_via_libnuma:375 - NUMA affinity: set memory preferred to NUMA node 3[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172774, ip=10.128.24.52)[0m 2026-05-23 15:19:17.766 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:__init__:481 - BaseVLLMInferenceEngine: vllm_v1_disable_multiproc=True, vllm.__version__=dev, VLLM_ENABLE_V1_MULTIPROCESSING=0[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172774, ip=10.128.24.52)[0m 2026-05-23 15:19:17.766 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:__init__:489 - BaseVLLMInferenceEngine: set VLLM_ENABLE_V1_MULTIPROCESSING=0[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172774, ip=10.128.24.52)[0m 2026-05-23 15:19:17.766 | WARNING | skyrl_train.inference_engines.vllm.vllm_engine:_create_engine:1047 - OpenAI API sampling params overridden: temperature=0.7, top_p=0.95, top_k=20[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172774, ip=10.128.24.52)[0m 2026-05-23 15:19:17.796 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:_create_engine:1099 - Engine startup stagger: sleeping 2.12s (attempt 1/5) to avoid port collisions[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m Exception raised in creation task: The actor died because of an error raised in its creation task, [36mray::AsyncVLLMInferenceEngine.__init__()[39m (pid=3294997, ip=10.128.17.219, actor_id=3099a3967d816aa5fa00928502000000, repr=<skyrl_train.inference_engines.vllm.vllm_engine.AsyncVLLMInferenceEngine object at 0x400bfefd3080>)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/concurrent/futures/_base.py", line 456, in result
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m return self.__get_result()
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m ^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/concurrent/futures/_base.py", line 401, in __get_result
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m raise self._exception
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m ^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/vllm/vllm_engine.py", line 1028, in __init__
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m super().__init__(*args, **kwargs)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/vllm/vllm_engine.py", line 500, in __init__
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m self.llm = self._create_engine(*args, **kwargs)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/vllm/vllm_engine.py", line 1105, in _create_engine
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m engine = vllm.AsyncLLMEngine.from_engine_args(engine_args, stat_loggers=stat_loggers)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/async_llm.py", line 251, in from_engine_args
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m return cls(
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m ^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/async_llm.py", line 148, in __init__
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m self.engine_core = EngineCoreClient.make_async_mp_client(
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core_client.py", line 124, in make_async_mp_client
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m return AsyncMPClient(*client_args)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core_client.py", line 835, in __init__
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m super().__init__(
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core_client.py", line 490, in __init__
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m with launch_core_engines(vllm_config, executor_class, log_stats) as (
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/contextlib.py", line 144, in __exit__
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m next(self.gen)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/utils.py", line 936, in launch_core_engines
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m wait_for_engine_startup(
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/utils.py", line 995, in wait_for_engine_startup
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m raise RuntimeError(
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m RuntimeError: Engine core initialization failed. See root cause above. Failed core proc(s): {}
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909038, ip=10.128.17.216)[0m [2026-05-23 15:19:20,076 E 1909038 1909140] core_worker_process.cc:842: Failed to establish connection to the metrics exporter agent. Metrics will not be exported. Exporter agent status: RpcError: Running out of retries to initialize the metrics agent. rpc_code: 14[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523482, ip=10.128.17.222)[0m (EngineCore_DP0 pid=523842) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523613, ip=10.128.17.222)[0m (EngineCore_DP0 pid=523858) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m [proxychains] DLL init: proxychains-ng 4.17-git-9-g78ead0e[32m [repeated 58x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523613, ip=10.128.17.222)[0m [W523 15:19:22.212466561 socket.cpp:767] [c10d] The client socket cannot be initialized to connect to [jpbo-010-46.jupiter.internal]:58805 (errno: 97 - Address family not supported by protocol).[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523613, ip=10.128.17.222)[0m [W523 15:19:22.215869301 Utils.hpp:166] Warning: Environment variable NCCL_BLOCKING_WAIT is deprecated; use TORCH_NCCL_BLOCKING_WAIT instead (function operator())[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523613, ip=10.128.17.222)[0m [rank0]:[W523 15:19:22.220727209 Utils.hpp:137] Warning: Environment variable TORCH_NCCL_TRACE_BUFFER_SIZE is deprecated; use TORCH_FR_BUFFER_SIZE instead (function operator())[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523615, ip=10.128.17.222)[0m (EngineCore_DP0 pid=523862) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396835, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397197)
Loading safetensors checkpoint shards: 25% Completed | 1/4 [00:03<00:11, 3.75s/it]
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m The argument `trust_remote_code` is to be used with Auto classes. It has no effect here and is ignored.[32m [repeated 14x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3123358, ip=10.128.17.215)[0m (EngineCore_DP0 pid=3123723) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m The tokenizer you are loading from '/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue.[32m [repeated 6x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/__init__.py:7: RuntimeWarning: Failed to read commit hash:[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m No module named 'vllm._version'[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m from .version import __version__, __version_tuple__ # isort:skip[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294996, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295247)
Loading safetensors checkpoint shards: 0% Completed | 0/4 [00:00<?, ?it/s][32m [repeated 6x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200625, ip=10.128.17.221)[0m (EngineCore_DP0 pid=200983) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3123494, ip=10.128.17.215)[0m (EngineCore_DP0 pid=3123731) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3222354, ip=10.128.17.212)[0m (EngineCore_DP0 pid=3222582) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3222354, ip=10.128.17.212)[0m (EngineCore_DP0 pid=3222582) /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/__init__.py:1617: UserWarning: Please use the new API settings to control TF32 behavior, such as torch.backends.cudnn.conv.fp32_precision = 'tf32' or torch.backends.cuda.matmul.fp32_precision = 'ieee'. Old settings, e.g, torch.backends.cuda.matmul.allow_tf32 = True, torch.backends.cudnn.allow_tf32 = True, allowTF32CuDNN() and allowTF32CuBLAS() will be deprecated after Pytorch 2.9. Please see https://pytorch.org/docs/main/notes/cuda.html#tensorfloat-32-tf32-on-ampere-and-later-devices (Triggered internally at /pytorch/aten/src/ATen/Context.cpp:80.)[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051741, ip=10.128.17.214)[0m 2026-05-23 15:19:22.722 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:setup_envvars_for_vllm:123 - setup_envvars_for_vllm: distributed_executor_backend=uni, SKYRL_ENABLE_NUMA_AFFINITY=1, CUDA_VISIBLE_DEVICES=3, VLLM_ENABLE_V1_MULTIPROCESSING=<unset>[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051741, ip=10.128.17.214)[0m 2026-05-23 15:19:22.724 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:setup_envvars_for_vllm:133 - setup_envvars_for_vllm: numa_enabled=True[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051741, ip=10.128.17.214)[0m 2026-05-23 15:19:22.724 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:setup_envvars_for_vllm:136 - setup_envvars_for_vllm: set VLLM_ENABLE_V1_MULTIPROCESSING=0 for NUMA affinity[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3222225, ip=10.128.17.212)[0m (EngineCore_DP0 pid=3222586) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(pid=3659917)[0m Using blocking ray.get inside async actor. This blocks the event loop. Please use `await` on object ref with asyncio.gather if you want to yield execution to the event loop instead.[32m [repeated 6x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200756, ip=10.128.17.221)[0m (EngineCore_DP0 pid=200991) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3123493, ip=10.128.17.215)[0m (EngineCore_DP0 pid=3123736) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051741, ip=10.128.17.214)[0m 2026-05-23 15:19:23.980 | INFO | skyrl_train.utils.numa:set_numa_affinity_for_gpu:340 - NUMA affinity: bound process to CPUs 216-287 (NUMA node 3) for GPU 3[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051741, ip=10.128.17.214)[0m 2026-05-23 15:19:24.008 | DEBUG | skyrl_train.utils.numa:_set_membind_via_libnuma:375 - NUMA affinity: set memory preferred to NUMA node 3[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051741, ip=10.128.17.214)[0m 2026-05-23 15:19:24.008 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:__init__:481 - BaseVLLMInferenceEngine: vllm_v1_disable_multiproc=True, vllm.__version__=dev, VLLM_ENABLE_V1_MULTIPROCESSING=0[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051741, ip=10.128.17.214)[0m 2026-05-23 15:19:24.008 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:__init__:489 - BaseVLLMInferenceEngine: set VLLM_ENABLE_V1_MULTIPROCESSING=0[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051741, ip=10.128.17.214)[0m 2026-05-23 15:19:24.008 | WARNING | skyrl_train.inference_engines.vllm.vllm_engine:_create_engine:1047 - OpenAI API sampling params overridden: temperature=0.7, top_p=0.95, top_k=20[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051740, ip=10.128.17.214)[0m 2026-05-23 15:19:24.038 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:_create_engine:1099 - Engine startup stagger: sleeping 1.65s (attempt 1/5) to avoid port collisions[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909165, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909404) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909166, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909420) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3222352, ip=10.128.17.212)[0m (EngineCore_DP0 pid=3222606) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3123492, ip=10.128.17.215)[0m (EngineCore_DP0 pid=3123744) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200625, ip=10.128.17.221)[0m [2026-05-23 15:19:25,287 E 200625 200728] core_worker_process.cc:842: Failed to establish connection to the metrics exporter agent. Metrics will not be exported. Exporter agent status: RpcError: Running out of retries to initialize the metrics agent. rpc_code: 14[32m [repeated 16x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909038, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909396) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m [proxychains] DLL init: proxychains-ng 4.17-git-9-g78ead0e[32m [repeated 16x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3222353, ip=10.128.17.212)[0m (EngineCore_DP0 pid=3222598) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200758, ip=10.128.17.221)[0m (EngineCore_DP0 pid=201001) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=372387, ip=10.128.17.220)[0m (EngineCore_DP0 pid=372619) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909167, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909400) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909167, ip=10.128.17.216)[0m [W523 15:19:27.784140679 socket.cpp:767] [c10d] The client socket cannot be initialized to connect to [jpbo-010-40-interconnect-1.jupiter.internal]:45605 (errno: 97 - Address family not supported by protocol).[32m [repeated 19x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909167, ip=10.128.17.216)[0m [W523 15:19:27.784665083 Utils.hpp:166] Warning: Environment variable NCCL_BLOCKING_WAIT is deprecated; use TORCH_NCCL_BLOCKING_WAIT instead (function operator())[32m [repeated 19x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909167, ip=10.128.17.216)[0m [rank0]:[W523 15:19:27.786653454 Utils.hpp:137] Warning: Environment variable TORCH_NCCL_TRACE_BUFFER_SIZE is deprecated; use TORCH_FR_BUFFER_SIZE instead (function operator())[32m [repeated 16x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200757, ip=10.128.17.221)[0m (EngineCore_DP0 pid=200995) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2397096, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397216)
Loading safetensors checkpoint shards: 50% Completed | 2/4 [00:08<00:08, 4.15s/it][32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m The argument `trust_remote_code` is to be used with Auto classes. It has no effect here and is ignored.[32m [repeated 16x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=372514, ip=10.128.17.220)[0m (EngineCore_DP0 pid=372624) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m The tokenizer you are loading from '/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue.[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/__init__.py:7: RuntimeWarning: Failed to read commit hash:[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m No module named 'vllm._version'[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m from .version import __version__, __version_tuple__ # isort:skip[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=372258, ip=10.128.17.220)[0m (EngineCore_DP0 pid=372628) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909165, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909404)
Loading safetensors checkpoint shards: 0% Completed | 0/4 [00:00<?, ?it/s][32m [repeated 20x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440679, ip=10.128.24.53)[0m (EngineCore_DP0 pid=1440908) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=372386, ip=10.128.17.220)[0m (EngineCore_DP0 pid=372633) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440681, ip=10.128.24.53)[0m (EngineCore_DP0 pid=1440912) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440681, ip=10.128.24.53)[0m (EngineCore_DP0 pid=1440912) /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/__init__.py:1617: UserWarning: Please use the new API settings to control TF32 behavior, such as torch.backends.cudnn.conv.fp32_precision = 'tf32' or torch.backends.cuda.matmul.fp32_precision = 'ieee'. Old settings, e.g, torch.backends.cuda.matmul.allow_tf32 = True, torch.backends.cudnn.allow_tf32 = True, allowTF32CuDNN() and allowTF32CuBLAS() will be deprecated after Pytorch 2.9. Please see https://pytorch.org/docs/main/notes/cuda.html#tensorfloat-32-tf32-on-ampere-and-later-devices (Triggered internally at /pytorch/aten/src/ATen/Context.cpp:80.)[32m [repeated 18x across cluster][0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:30.331[0m | [1m[32mINFO [0m | [36mskyrl_train.workers.worker[0m:[36m_initiate_actors[0m:[36m498[0m - [1m[32mInitialized process group for RayActorGroup[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:30.341[0m | [1m[32mINFO [0m | [36mskyrl_train.workers.worker[0m:[36m_initiate_actors[0m:[36m500[0m - [1m[32mMesh Ranks: [MeshRank(dp=0, sp=0, tp=0, pp=0, world_size=8, dp_size=8, pp_size=1), MeshRank(dp=1, sp=0, tp=0, pp=0, world_size=8, dp_size=8, pp_size=1), MeshRank(dp=2, sp=0, tp=0, pp=0, world_size=8, dp_size=8, pp_size=1), MeshRank(dp=3, sp=0, tp=0, pp=0, world_size=8, dp_size=8, pp_size=1), MeshRank(dp=4, sp=0, tp=0, pp=0, world_size=8, dp_size=8, pp_size=1), MeshRank(dp=5, sp=0, tp=0, pp=0, world_size=8, dp_size=8, pp_size=1), MeshRank(dp=6, sp=0, tp=0, pp=0, world_size=8, dp_size=8, pp_size=1), MeshRank(dp=7, sp=0, tp=0, pp=0, world_size=8, dp_size=8, pp_size=1)][0m
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m `torch_dtype` is deprecated! Use `dtype` instead!
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=3659918)[0m
Loading checkpoint shards: 0%| | 0/4 [00:00<?, ?it/s]
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m
Loading checkpoint shards: 50%|█████ | 2/4 [00:00<00:00, 16.35it/s]
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m
Loading checkpoint shards: 100%|██████████| 4/4 [00:00<00:00, 18.60it/s]
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=3659915)[0m
Loading checkpoint shards: 75%|███████▌ | 3/4 [00:00<00:00, 20.16it/s]
Loading checkpoint shards: 100%|██████████| 4/4 [00:00<00:00, 25.72it/s]
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=3659916)[0m
Loading checkpoint shards: 100%|██████████| 4/4 [00:00<00:00, 15.36it/s]
Loading checkpoint shards: 100%|██████████| 4/4 [00:00<00:00, 15.01it/s]
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440680, ip=10.128.24.53)[0m (EngineCore_DP0 pid=1440929) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172774, ip=10.128.24.52)[0m (EngineCore_DP0 pid=2173007) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172645, ip=10.128.24.52)[0m [2026-05-23 15:19:31,089 E 2172645 2172743] core_worker_process.cc:842: Failed to establish connection to the metrics exporter agent. Metrics will not be exported. Exporter agent status: RpcError: Running out of retries to initialize the metrics agent. rpc_code: 14[32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440552, ip=10.128.24.53)[0m (EngineCore_DP0 pid=1440918) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172645, ip=10.128.24.52)[0m (EngineCore_DP0 pid=2173002) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440552, ip=10.128.24.53)[0m [W523 15:19:32.269254945 socket.cpp:767] [c10d] The client socket cannot be initialized to connect to [jpbo-022-05.jupiter.internal]:39961 (errno: 97 - Address family not supported by protocol).[32m [repeated 13x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440552, ip=10.128.24.53)[0m [W523 15:19:32.269727257 Utils.hpp:166] Warning: Environment variable NCCL_BLOCKING_WAIT is deprecated; use TORCH_NCCL_BLOCKING_WAIT instead (function operator())[32m [repeated 14x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440552, ip=10.128.24.53)[0m [rank0]:[W523 15:19:32.271950002 Utils.hpp:137] Warning: Environment variable TORCH_NCCL_TRACE_BUFFER_SIZE is deprecated; use TORCH_FR_BUFFER_SIZE instead (function operator())[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396835, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397197)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396963, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397201)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396962, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397192)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2397096, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397216)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m (EngineCore_DP0 pid=2173022) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2397096, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397216)
Loading safetensors checkpoint shards: 100% Completed | 4/4 [00:13<00:00, 3.38s/it][32m [repeated 42x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295221)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294998, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295229)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294996, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295247)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172775, ip=10.128.24.52)[0m (EngineCore_DP0 pid=2173018) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503354, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503581) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440680, ip=10.128.24.53)[0m (EngineCore_DP0 pid=1440929)
Loading safetensors checkpoint shards: 0% Completed | 0/4 [00:00<?, ?it/s][32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503354, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503581) /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/__init__.py:1617: UserWarning: Please use the new API settings to control TF32 behavior, such as torch.backends.cudnn.conv.fp32_precision = 'tf32' or torch.backends.cuda.matmul.fp32_precision = 'ieee'. Old settings, e.g, torch.backends.cuda.matmul.allow_tf32 = True, torch.backends.cudnn.allow_tf32 = True, allowTF32CuDNN() and allowTF32CuBLAS() will be deprecated after Pytorch 2.9. Please see https://pytorch.org/docs/main/notes/cuda.html#tensorfloat-32-tf32-on-ampere-and-later-devices (Triggered internally at /pytorch/aten/src/ATen/Context.cpp:80.)[32m [repeated 7x across cluster][0m
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=3659917)[0m `torch_dtype` is deprecated! Use `dtype` instead![32m [repeated 7x across cluster][0m
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=3659915)[0m
Loading checkpoint shards: 0%| | 0/4 [00:00<?, ?it/s][32m [repeated 7x across cluster][0m
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=3659917)[0m
Loading checkpoint shards: 50%|█████ | 2/4 [00:00<00:00, 11.90it/s][32m [repeated 6x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396835, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397197) The tokenizer you are loading from '/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue.
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753509, ip=10.128.17.217)[0m
Loading checkpoint shards: 100%|██████████| 4/4 [00:00<00:00, 17.05it/s][32m [repeated 3x across cluster][0m
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=3659917)[0m
Loading checkpoint shards: 100%|██████████| 4/4 [00:00<00:00, 14.54it/s]
Loading checkpoint shards: 100%|██████████| 4/4 [00:00<00:00, 14.07it/s][32m [repeated 2x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396835, ip=10.128.24.51)[0m 2026-05-23 15:19:36.781 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:_create_engine:1156 - Initializing OpenAIServingChat with custom_chat_template read from: chat_templates/qwen3_thinking_acc.jinja2
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503353, ip=10.128.24.50)[0m [2026-05-23 15:19:35,200 E 1503353 1503520] core_worker_process.cc:842: Failed to establish connection to the metrics exporter agent. Metrics will not be exported. Exporter agent status: RpcError: Running out of retries to initialize the metrics agent. rpc_code: 14[32m [repeated 14x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200625, ip=10.128.17.221)[0m (EngineCore_DP0 pid=200983)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200756, ip=10.128.17.221)[0m (EngineCore_DP0 pid=200991)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200757, ip=10.128.17.221)[0m (EngineCore_DP0 pid=200995)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200758, ip=10.128.17.221)[0m (EngineCore_DP0 pid=201001)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503353, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503586) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503354, ip=10.128.24.50)[0m [W523 15:19:35.878750956 socket.cpp:767] [c10d] The client socket cannot be initialized to connect to [jpbo-022-02.jupiter.internal]:55605 (errno: 97 - Address family not supported by protocol).[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503354, ip=10.128.24.50)[0m [W523 15:19:35.883184374 Utils.hpp:166] Warning: Environment variable NCCL_BLOCKING_WAIT is deprecated; use TORCH_NCCL_BLOCKING_WAIT instead (function operator())[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503354, ip=10.128.24.50)[0m [rank0]:[W523 15:19:35.885823523 Utils.hpp:137] Warning: Environment variable TORCH_NCCL_TRACE_BUFFER_SIZE is deprecated; use TORCH_FR_BUFFER_SIZE instead (function operator())[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3052003) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440680, ip=10.128.24.53)[0m (EngineCore_DP0 pid=1440929)
Loading safetensors checkpoint shards: 25% Completed | 1/4 [00:04<00:12, 4.05s/it][32m [repeated 42x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051741, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051990) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503355, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503590) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523482, ip=10.128.17.222)[0m (EngineCore_DP0 pid=523842)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523614, ip=10.128.17.222)[0m (EngineCore_DP0 pid=523846)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523613, ip=10.128.17.222)[0m (EngineCore_DP0 pid=523858)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523615, ip=10.128.17.222)[0m (EngineCore_DP0 pid=523862)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3123358, ip=10.128.17.215)[0m (EngineCore_DP0 pid=3123723)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3123493, ip=10.128.17.215)[0m (EngineCore_DP0 pid=3123736)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3123492, ip=10.128.17.215)[0m (EngineCore_DP0 pid=3123744)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3123494, ip=10.128.17.215)[0m (EngineCore_DP0 pid=3123731)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503222, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503605) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051740, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051998) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503355, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503590)
Loading safetensors checkpoint shards: 0% Completed | 0/4 [00:00<?, ?it/s][32m [repeated 6x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051740, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051998) /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/__init__.py:1617: UserWarning: Please use the new API settings to control TF32 behavior, such as torch.backends.cudnn.conv.fp32_precision = 'tf32' or torch.backends.cuda.matmul.fp32_precision = 'ieee'. Old settings, e.g, torch.backends.cuda.matmul.allow_tf32 = True, torch.backends.cudnn.allow_tf32 = True, allowTF32CuDNN() and allowTF32CuBLAS() will be deprecated after Pytorch 2.9. Please see https://pytorch.org/docs/main/notes/cuda.html#tensorfloat-32-tf32-on-ampere-and-later-devices (Triggered internally at /pytorch/aten/src/ATen/Context.cpp:80.)[32m [repeated 6x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051742, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051996) _C._set_float32_matmul_precision(precision)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200758, ip=10.128.17.221)[0m (EngineCore_DP0 pid=201001) The tokenizer you are loading from '/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue.[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200758, ip=10.128.17.221)[0m 2026-05-23 15:19:40.963 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:_create_engine:1156 - Initializing OpenAIServingChat with custom_chat_template read from: chat_templates/qwen3_thinking_acc.jinja2[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m [2026-05-23 15:19:38,106 E 2753436 2753476] core_worker_process.cc:842: Failed to establish connection to the metrics exporter agent. Metrics will not be exported. Exporter agent status: RpcError: Running out of retries to initialize the metrics agent. rpc_code: 14[32m [repeated 4x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909038, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909396)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909166, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909420)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909167, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909400)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909165, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909404)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051742, ip=10.128.17.214)[0m [W523 15:19:40.328829609 socket.cpp:767] [c10d] The client socket cannot be initialized to connect to [jpbo-010-38-interconnect-1.jupiter.internal]:50715 (errno: 97 - Address family not supported by protocol).[32m [repeated 7x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051742, ip=10.128.17.214)[0m [W523 15:19:40.329285632 Utils.hpp:166] Warning: Environment variable NCCL_BLOCKING_WAIT is deprecated; use TORCH_NCCL_BLOCKING_WAIT instead (function operator())[32m [repeated 7x across cluster][0m
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753509, ip=10.128.17.217)[0m [rank2]:[W523 15:19:42.418333463 Utils.hpp:137] Warning: Environment variable TORCH_NCCL_TRACE_BUFFER_SIZE is deprecated; use TORCH_FR_BUFFER_SIZE instead (function operator())[32m [repeated 15x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172645, ip=10.128.24.52)[0m (EngineCore_DP0 pid=2173002)
Loading safetensors checkpoint shards: 50% Completed | 2/4 [00:08<00:08, 4.16s/it][32m [repeated 57x across cluster][0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:45.040[0m | [1m[32mINFO [0m | [36mskyrl_train.trainer[0m:[36mbuild_models[0m:[36m802[0m - [1m[32minit policy/ref/critic models done[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:45.042[0m | [1m[32mINFO [0m | [36mharbor.orchestrators.queue[0m:[36mstart[0m:[36m262[0m - [1m[32m[terminal_bench_generator:236] Started 900 workers (status every 120.0s, 0.75s launch grace period)[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:45.042[0m | [1m[32mINFO [0m | [36mexamples.terminal_bench.terminal_bench_generator[0m:[36m_create_orchestrator[0m:[36m257[0m - [1m[32mQueueOrchestrator created and started with n_concurrent_trials=900, rollback_hook registered for ContextLengthExceededError/AgentTimeoutError[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:45.042[0m | [1m[32mINFO [0m | [36mexamples.terminal_bench.terminal_bench_generator[0m:[36mstartup[0m:[36m226[0m - [1m[32mTerminalBenchGenerator startup complete. Shared orchestrator ready with n_concurrent_trials=900[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:45.043[0m | [1m[32mINFO [0m | [36mskyrl_train.fully_async_trainer[0m:[36mtrain[0m:[36m428[0m - [1m[32mGenerator startup complete[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:45.043[0m | [1m[32mINFO [0m | [36mskyrl_train.fully_async_trainer[0m:[36m_train_loop[0m:[36m453[0m - [1m[32mStarted: 'load_checkpoints'[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:45.088[0m | [1m[32mINFO [0m | [36mskyrl_train.trainer[0m:[36mload_checkpoints[0m:[36m1595[0m - [1m[32mLoading checkpoint from: /e/data1/datasets/playground/ot-baf/a3-rl-DCAgent_code-contests-noblock/a3-rl-DCAgent_code-contests-noblock/checkpoints/global_step_18[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:45.088[0m | [1m[32mINFO [0m | [36mskyrl_train.trainer[0m:[36mload_checkpoints[0m:[36m1601[0m - [1m[32mResuming from global_step: 18[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:45.128[0m | [1m[32mINFO [0m | [36mskyrl_train.trainer[0m:[36mload_checkpoints[0m:[36m1617[0m - [1m[32mSuccessfully loaded trainer state[0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051740, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051998)
Loading safetensors checkpoint shards: 0% Completed | 0/4 [00:00<?, ?it/s][32m [repeated 6x across cluster][0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:45.171[0m | [1m[32mINFO [0m | [36mskyrl_train.trainer[0m:[36mload_checkpoints[0m:[36m1627[0m - [1m[32mSuccessfully loaded dataloader state[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:19:45.171[0m | [1m[32mINFO [0m | [36mskyrl_train.trainer[0m:[36mload_checkpoints[0m:[36m1636[0m - [1m[32mLoading policy checkpoint from /e/data1/datasets/playground/ot-baf/a3-rl-DCAgent_code-contests-noblock/a3-rl-DCAgent_code-contests-noblock/checkpoints/global_step_18/policy[0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=372258, ip=10.128.17.220)[0m (EngineCore_DP0 pid=372628)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051742, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051996) /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/__init__.py:1617: UserWarning: Please use the new API settings to control TF32 behavior, such as torch.backends.cudnn.conv.fp32_precision = 'tf32' or torch.backends.cuda.matmul.fp32_precision = 'ieee'. Old settings, e.g, torch.backends.cuda.matmul.allow_tf32 = True, torch.backends.cudnn.allow_tf32 = True, allowTF32CuDNN() and allowTF32CuBLAS() will be deprecated after Pytorch 2.9. Please see https://pytorch.org/docs/main/notes/cuda.html#tensorfloat-32-tf32-on-ampere-and-later-devices (Triggered internally at /pytorch/aten/src/ATen/Context.cpp:80.)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=372386, ip=10.128.17.220)[0m (EngineCore_DP0 pid=372633)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=372387, ip=10.128.17.220)[0m (EngineCore_DP0 pid=372619)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=372514, ip=10.128.17.220)[0m (EngineCore_DP0 pid=372624)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909166, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909420) The tokenizer you are loading from '/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue.[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909038, ip=10.128.17.216)[0m 2026-05-23 15:19:46.919 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:_create_engine:1156 - Initializing OpenAIServingChat with custom_chat_template read from: chat_templates/qwen3_thinking_acc.jinja2[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440552, ip=10.128.24.53)[0m (EngineCore_DP0 pid=1440918)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440681, ip=10.128.24.53)[0m (EngineCore_DP0 pid=1440912)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440679, ip=10.128.24.53)[0m (EngineCore_DP0 pid=1440908)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440680, ip=10.128.24.53)[0m (EngineCore_DP0 pid=1440929)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503222, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503605)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503354, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503581)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503355, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503590)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503353, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503586)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3222225, ip=10.128.17.212)[0m (EngineCore_DP0 pid=3222586)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3222352, ip=10.128.17.212)[0m (EngineCore_DP0 pid=3222606)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3222353, ip=10.128.17.212)[0m (EngineCore_DP0 pid=3222598)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3222354, ip=10.128.17.212)[0m (EngineCore_DP0 pid=3222582)
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=3659915)[0m [2026-05-23 15:19:48,908 E 3659915 3660061] core_worker_process.cc:842: Failed to establish connection to the metrics exporter agent. Metrics will not be exported. Exporter agent status: RpcError: Running out of retries to initialize the metrics agent. rpc_code: 14
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3222354, ip=10.128.17.212)[0m (EngineCore_DP0 pid=3222582)
Loading safetensors checkpoint shards: 100% Completed | 4/4 [00:19<00:00, 4.98s/it][32m [repeated 59x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172645, ip=10.128.24.52)[0m (EngineCore_DP0 pid=2173002)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172775, ip=10.128.24.52)[0m (EngineCore_DP0 pid=2173018)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m (EngineCore_DP0 pid=2173022)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172774, ip=10.128.24.52)[0m (EngineCore_DP0 pid=2173007)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503354, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503581) The tokenizer you are loading from '/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue.[32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440681, ip=10.128.24.53)[0m 2026-05-23 15:19:51.235 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:_create_engine:1156 - Initializing OpenAIServingChat with custom_chat_template read from: chat_templates/qwen3_thinking_acc.jinja2[32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=3659917)[0m [2026-05-23 15:19:49,079 E 3659917 3660145] core_worker_process.cc:842: Failed to establish connection to the metrics exporter agent. Metrics will not be exported. Exporter agent status: RpcError: Running out of retries to initialize the metrics agent. rpc_code: 14[32m [repeated 6x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3052003)
Loading safetensors checkpoint shards: 75% Completed | 3/4 [00:13<00:04, 4.43s/it][32m [repeated 13x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3052003)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172774, ip=10.128.24.52)[0m (EngineCore_DP0 pid=2173007) The tokenizer you are loading from '/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue.[32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051742, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051996)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051740, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051998)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051741, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051990)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m 2026-05-23 15:19:53.103 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:_create_engine:1156 - Initializing OpenAIServingChat with custom_chat_template read from: chat_templates/qwen3_thinking_acc.jinja2[32m [repeated 12x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051741, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051990)
Loading safetensors checkpoint shards: 100% Completed | 4/4 [00:15<00:00, 3.76s/it][32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py:4876: UserWarning: barrier(): using the device under current context. You can specify `device_id` in `init_process_group` to mute this warning.
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m warnings.warn( # warn only once
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051741, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051990) The tokenizer you are loading from '/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6' with an incorrect regex pattern: https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503/discussions/84#69121093e8b480e709447d5e. This will lead to incorrect tokenization. You should set the `fix_mistral_regex=True` flag when loading this tokenizer to fix this issue.[32m [repeated 4x across cluster][0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:20:03.601[0m | [1m[32mINFO [0m | [36mskyrl_train.trainer[0m:[36mload_checkpoints[0m:[36m1646[0m - [1m[32mSuccessfully loaded policy checkpoint[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:20:03.601[0m | [1m[32mINFO [0m | [36mskyrl_train.trainer[0m:[36mload_checkpoints[0m:[36m1662[0m - [1m[32mSuccessfully loaded complete checkpoint state from global_step_18[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:20:03.601[0m | [1m[32mINFO [0m | [36mskyrl_train.fully_async_trainer[0m:[36m_train_loop[0m:[36m455[0m - [1m[32mResumed training from global_step 18[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:20:03.623[0m | [1m[32mINFO [0m | [36mskyrl_train.utils.data_tracker[0m:[36mload_state[0m:[36m97[0m - [1m[32mLoaded data tracker state: epoch=0, consumed_in_epoch=1144, total_consumed=1144[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:20:03.624[0m | [33m[1mWARNING [0m | [36mskyrl_train.fully_async_trainer[0m:[36m_train_loop[0m:[36m480[0m - [33m[1mData consumption count mismatch on resume: expected 1152, got 1144. This can happen after epoch boundary transitions or error recovery.[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:20:03.624[0m | [1m[32mINFO [0m | [36mskyrl_train.fully_async_trainer[0m:[36m_train_loop[0m:[36m453[0m - [1m[32mFinished: 'load_checkpoints', time cost: 18.58s[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:20:03.624[0m | [1m[32mINFO [0m | [36mskyrl_train.fully_async_trainer[0m:[36m_train_loop[0m:[36m487[0m - [1m[32mStarted: 'init_weight_sync_state'[0m
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/__init__.py:7: RuntimeWarning: Failed to read commit hash:
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m No module named 'vllm._version'
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m from .version import __version__, __version_tuple__ # isort:skip
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m [32m2026-05-23 15:20:09.761[0m | [1m[32mINFO [0m | [36mlogging[0m:[36minfo[0m:[36m2216[0m - [1m[32m[weight-sync] Using master_addr=10.128.17.217, master_port=47881[0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051741, ip=10.128.17.214)[0m 2026-05-23 15:20:00.782 | INFO | skyrl_train.inference_engines.vllm.vllm_engine:_create_engine:1156 - Initializing OpenAIServingChat with custom_chat_template read from: chat_templates/qwen3_thinking_acc.jinja2[32m [repeated 4x across cluster][0m
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m [32m2026-05-23 15:20:09.760[0m | [1m[32mINFO [0m | [36mlogging[0m:[36minfo[0m:[36m2216[0m - [1m[32m[weight-sync] get_node_ip_address()=10.128.17.217[0m[32m [repeated 3x across cluster][0m
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=3659917)[0m /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/__init__.py:7: RuntimeWarning: Failed to read commit hash:[32m [repeated 7x across cluster][0m
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=3659917)[0m No module named 'vllm._version'[32m [repeated 7x across cluster][0m
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=3659917)[0m from .version import __version__, __version_tuple__ # isort:skip[32m [repeated 7x across cluster][0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:20:09.842[0m | [1m[32mINFO [0m | [36mskyrl_train.fully_async_trainer[0m:[36m_train_loop[0m:[36m487[0m - [1m[32mFinished: 'init_weight_sync_state', time cost: 6.22s[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m --- Logging error in Loguru Handler #2 ---
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m Record was: {'elapsed': datetime.timedelta(seconds=212, microseconds=858264), 'exception': (type=<class 'ray.exceptions.RayTaskError(ActorDiedError)'>, value=RayTaskError(ActorDiedError)(ActorDiedError(RayTaskError('__init__', 'Traceback (most recent call last):\n File "python/ray/_raylet.pyx", line 1722, in ray._raylet.execute_task\n File "python/ray/_raylet.pyx", line 1659, in ray._raylet.execute_task.function_executor\n File "python/ray/_raylet.pyx", line 4342, in ray._raylet.CoreWorker.run_async_func_or_coro_in_event_loop\n File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/concurrent/futures/_base.py", line 456, in result\n return self.__get_result()\n ^^^^^^^^^^^^^^^^^^^\n File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/concurrent/futures/_base.py", line 401, in __get_result\n raise self._exception\n File "python/ray/_raylet.pyx", line 4329, in async_func\n File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/ray/_private/async_compat.py", line 52, in wrapper\n return func(*args, **kwargs)\n ^^^^^^^^^^^^^^^^^^^^^\n File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/ray/_private/function_manager.py", line 693, in actor_method_executor\n return method(__ray_actor, *args, **kwargs)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/ray/util/tracing/tracing_helper.py", line 461, in _resume_span\n return method(self, *_args, **_kwargs)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/vllm/vllm_engine.py", line 1028, in __init__\n super().__init__(*args, **kwargs)\n File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/vllm/vllm_engine.py", line 500, in __init__\n self.llm = self._create_engine(*args, **kwargs)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/ray/util/tracing/tracing_helper.py", line 461, in _resume_span\n return method(self, *_args, **_kwargs)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/vllm/vllm_engine.py", line 1105, in _create_engine\n engine = vllm.AsyncLLMEngine.from_engine_args(engine_args, stat_loggers=stat_loggers)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/async_llm.py", line 251, in from_engine_args\n return cls(\n ^^^^\n File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/async_llm.py", line 148, in __init__\n self.engine_core = EngineCoreClient.make_async_mp_client(\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core_client.py", line 124, in make_async_mp_client\n return AsyncMPClient(*client_args)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^\n File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core_client.py", line 835, in __init__\n super().__init__(\n File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core_client.py", line 490, in __init__\n with launch_core_engines(vllm_config, executor_class, log_stats) as (\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/contextlib.py", line 144, in __exit__\n next(self.gen)\n File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/utils.py", line 936, in la
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m [rank0]:[W523 15:20:09.372907026 socket.cpp:767] [c10d] The client socket cannot be initialized to connect to [jpbo-010-41.jupiter.internal]:47881 (errno: 97 - Address family not supported by protocol).
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m Traceback (most recent call last):
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/fully_async_trainer.py", line 434, in train
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m await self._train_loop()
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/fully_async_trainer.py", line 488, in _train_loop
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m self.init_weight_sync_state()
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/trainer.py", line 808, in init_weight_sync_state
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ray.get(
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/ray/_private/auto_init_hook.py", line 22, in auto_init_wrapper
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m return fn(*args, **kwargs)
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/ray/_private/client_mode_hook.py", line 104, in wrapper
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m return func(*args, **kwargs)
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/ray/_private/worker.py", line 2967, in get
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m values, debugger_breakpoint = worker.get_objects(
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/ray/_private/worker.py", line 1015, in get_objects
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m raise value.as_instanceof_cause()
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ray.exceptions.RayTaskError(ActorDiedError): [36mray::FSDPPolicyWorkerBase.init_weight_sync_state()[39m (pid=2753436, ip=10.128.17.217, actor_id=7804ef248912326214f6164a02000000, repr=<skyrl_train.workers.fsdp.fsdp_worker.FSDPPolicyWorkerBase object at 0x400df41fe420>)
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/concurrent/futures/_base.py", line 449, in result
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m return self.__get_result()
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/concurrent/futures/_base.py", line 401, in __get_result
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m raise self._exception
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/workers/worker.py", line 299, in init_weight_sync_state
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m results = await asyncio.gather(*tasks)
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/inference_engine_client.py", line 655, in init_weight_update_communicator
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m rank_offset_count += engine.tp_size() * engine.pp_size()
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/ray_wrapped_inference_engine.py", line 26, in tp_size
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m return ray.get(self.inference_engine_actor.tp_size.remote())
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ray.exceptions.ActorDiedError: The actor died because of an error raised in its creation task, [36mray::AsyncVLLMInferenceEngine.__init__()[39m (pid=3294997, ip=10.128.17.219, actor_id=3099a3967d816aa5fa00928502000000, repr=<skyrl_train.inference_engines.vllm.vllm_engine.AsyncVLLMInferenceEngine object at 0x400bfefd3080>)
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/concurrent/futures/_base.py", line 456, in result
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m return self.__get_result()
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/concurrent/futures/_base.py", line 401, in __get_result
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m raise self._exception
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/vllm/vllm_engine.py", line 1028, in __init__
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m super().__init__(*args, **kwargs)
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/vllm/vllm_engine.py", line 500, in __init__
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m self.llm = self._create_engine(*args, **kwargs)
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/vllm/vllm_engine.py", line 1105, in _create_engine
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m engine = vllm.AsyncLLMEngine.from_engine_args(engine_args, stat_loggers=stat_loggers)
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/async_llm.py", line 251, in from_engine_args
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m return cls(
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/async_llm.py", line 148, in __init__
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m self.engine_core = EngineCoreClient.make_async_mp_client(
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core_client.py", line 124, in make_async_mp_client
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m return AsyncMPClient(*client_args)
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core_client.py", line 835, in __init__
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m super().__init__(
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core_client.py", line 490, in __init__
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m with launch_core_engines(vllm_config, executor_class, log_stats) as (
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/contextlib.py", line 144, in __exit__
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m next(self.gen)
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/utils.py", line 936, in launch_core_engines
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m wait_for_engine_startup(
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/utils.py", line 995, in wait_for_engine_startup
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m raise RuntimeError(
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m RuntimeError: Engine core initialization failed. See root cause above. Failed core proc(s): {}
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m During handling of the above exception, another exception occurred:
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m Traceback (most recent call last):
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/loguru/_handler.py", line 204, in emit
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m self._queue.put(str_record)
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/multiprocessing/queues.py", line 393, in put
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m obj = _ForkingPickler.dumps(obj)
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/multiprocessing/reduction.py", line 51, in dumps
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m cls(buf, protocol).dump(obj)
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m _pickle.PicklingError: Can't pickle <class 'ray.exceptions.RayTaskError(ActorDiedError)'>: attribute lookup RayTaskError(ActorDiedError) on ray.exceptions failed
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m --- End of logging error ---
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:20:10.907[0m | [1m[32mINFO [0m | [36mskyrl_train.inference_engines.inference_engine_client_http_endpoint[0m:[36mshutdown_server[0m:[36m203[0m - [1m[32mServer shut down after 2 seconds[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:20:10.908[0m | [1m[32mINFO [0m | [36mskyrl_train.trainer[0m:[36m_guarded_sync[0m:[36m226[0m - [1m[32mHTTP endpoint shutdown complete[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:20:10.908[0m | [1m[32mINFO [0m | [36mexamples.terminal_bench.terminal_bench_generator[0m:[36mshutdown[0m:[36m328[0m - [1m[32mShutting down shared QueueOrchestrator...[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:20:10.910[0m | [1m[32mINFO [0m | [36mharbor.orchestrators.queue[0m:[36mshutdown[0m:[36m377[0m - [1m[32m[terminal_bench_generator:236] Shutdown complete. Total completed: 0[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:20:10.910[0m | [1m[32mINFO [0m | [36mexamples.terminal_bench.terminal_bench_generator[0m:[36mshutdown[0m:[36m330[0m - [1m[32mQueueOrchestrator shutdown complete[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:20:10.910[0m | [1m[32mINFO [0m | [36mskyrl_train.trainer[0m:[36m_guarded_async[0m:[36m215[0m - [1m[32mGenerator shutdown complete[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:20:10.920[0m | [33m[1mWARNING [0m | [36mskyrl_train.trainer[0m:[36m_guarded_async[0m:[36m219[0m - [33m[1mInference engine teardown error (non-fatal): The actor died because of an error raised in its creation task, [36mray::AsyncVLLMInferenceEngine.__init__()[39m (pid=3294997, ip=10.128.17.219, actor_id=3099a3967d816aa5fa00928502000000, repr=<skyrl_train.inference_engines.vllm.vllm_engine.AsyncVLLMInferenceEngine object at 0x400bfefd3080>)
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/concurrent/futures/_base.py", line 456, in result
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m return self.__get_result()
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/concurrent/futures/_base.py", line 401, in __get_result
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m raise self._exception
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/vllm/vllm_engine.py", line 1028, in __init__
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m super().__init__(*args, **kwargs)
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/vllm/vllm_engine.py", line 500, in __init__
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m self.llm = self._create_engine(*args, **kwargs)
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/vllm/vllm_engine.py", line 1105, in _create_engine
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m engine = vllm.AsyncLLMEngine.from_engine_args(engine_args, stat_loggers=stat_loggers)
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/async_llm.py", line 251, in from_engine_args
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m return cls(
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/async_llm.py", line 148, in __init__
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m self.engine_core = EngineCoreClient.make_async_mp_client(
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core_client.py", line 124, in make_async_mp_client
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m return AsyncMPClient(*client_args)
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core_client.py", line 835, in __init__
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m super().__init__(
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core_client.py", line 490, in __init__
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m with launch_core_engines(vllm_config, executor_class, log_stats) as (
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/contextlib.py", line 144, in __exit__
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m next(self.gen)
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/utils.py", line 936, in launch_core_engines
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m wait_for_engine_startup(
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/utils.py", line 995, in wait_for_engine_startup
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m raise RuntimeError(
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m RuntimeError: Engine core initialization failed. See root cause above. Failed core proc(s): {}[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:20:10.920[0m | [1m[32mINFO [0m | [36mskyrl_train.trainer[0m:[36m_kill_ray_actors[0m:[36m248[0m - [1m[32mKilling policy_model actors...[0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295221) /e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/vllm/vllm_engine.py:470: UserWarning: No model update group to destroy
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295221) warnings.warn("No model update group to destroy")
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:20:10.936[0m | [1m[32mINFO [0m | [36mskyrl_train.trainer[0m:[36m_kill_ray_actors[0m:[36m267[0m - [1m[32mKilled 48 inference engine actor(s)[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:20:10.936[0m | [1m[32mINFO [0m | [36mskyrl_train.trainer[0m:[36m_guarded_sync[0m:[36m226[0m - [1m[32mRay actor cleanup complete[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:20:10.937[0m | [1m[32mINFO [0m | [36mskyrl_train.trainer[0m:[36m_kill_ray_actors[0m:[36m248[0m - [1m[32mKilling policy_model actors...[0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m [32m2026-05-23 15:20:10.941[0m | [1m[32mINFO [0m | [36mskyrl_train.trainer[0m:[36m_kill_ray_actors[0m:[36m267[0m - [1m[32mKilled 48 inference engine actor(s)[0m
|
|||
|
|
2026-05-23 15:20:10.953 | ERROR | __main__:main:125 - Training failed: [36mray::skyrl_entrypoint()[39m (pid=3659716, ip=10.128.17.210)
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/examples/terminal_bench/entrypoints/main_tbench.py", line 96, in skyrl_entrypoint
|
|||
|
|
exp.run()
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/entrypoints/main_base.py", line 403, in run
|
|||
|
|
asyncio.run(trainer.train())
|
|||
|
|
File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/asyncio/runners.py", line 195, in run
|
|||
|
|
return runner.run(main)
|
|||
|
|
^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/asyncio/runners.py", line 118, in run
|
|||
|
|
return self._loop.run_until_complete(task)
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "uvloop/loop.pyx", line 1518, in uvloop.loop.Loop.run_until_complete
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/fully_async_trainer.py", line 434, in train
|
|||
|
|
await self._train_loop()
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/fully_async_trainer.py", line 488, in _train_loop
|
|||
|
|
self.init_weight_sync_state()
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/trainer.py", line 808, in init_weight_sync_state
|
|||
|
|
ray.get(
|
|||
|
|
^^^^^^^^^^^^^^^^^^^
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
^^^^^^^^^^^^^^^^^^^
|
|||
|
|
ray.exceptions.RayTaskError(ActorDiedError): [36mray::FSDPPolicyWorkerBase.init_weight_sync_state()[39m (pid=2753436, ip=10.128.17.217, actor_id=7804ef248912326214f6164a02000000, repr=<skyrl_train.workers.fsdp.fsdp_worker.FSDPPolicyWorkerBase object at 0x400df41fe420>)
|
|||
|
|
File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/concurrent/futures/_base.py", line 449, in result
|
|||
|
|
return self.__get_result()
|
|||
|
|
^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/concurrent/futures/_base.py", line 401, in __get_result
|
|||
|
|
raise self._exception
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/workers/worker.py", line 299, in init_weight_sync_state
|
|||
|
|
results = await asyncio.gather(*tasks)
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/inference_engine_client.py", line 655, in init_weight_update_communicator
|
|||
|
|
rank_offset_count += engine.tp_size() * engine.pp_size()
|
|||
|
|
^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/ray_wrapped_inference_engine.py", line 26, in tp_size
|
|||
|
|
return ray.get(self.inference_engine_actor.tp_size.remote())
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
^^^^^^^^^^^^^^^^^^^
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
^^^^^^^^^^^^^^^^^^^
|
|||
|
|
ray.exceptions.ActorDiedError: The actor died because of an error raised in its creation task, [36mray::AsyncVLLMInferenceEngine.__init__()[39m (pid=3294997, ip=10.128.17.219, actor_id=3099a3967d816aa5fa00928502000000, repr=<skyrl_train.inference_engines.vllm.vllm_engine.AsyncVLLMInferenceEngine object at 0x400bfefd3080>)
|
|||
|
|
File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/concurrent/futures/_base.py", line 456, in result
|
|||
|
|
return self.__get_result()
|
|||
|
|
^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/concurrent/futures/_base.py", line 401, in __get_result
|
|||
|
|
raise self._exception
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/vllm/vllm_engine.py", line 1028, in __init__
|
|||
|
|
super().__init__(*args, **kwargs)
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/vllm/vllm_engine.py", line 500, in __init__
|
|||
|
|
self.llm = self._create_engine(*args, **kwargs)
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/vllm/vllm_engine.py", line 1105, in _create_engine
|
|||
|
|
engine = vllm.AsyncLLMEngine.from_engine_args(engine_args, stat_loggers=stat_loggers)
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/async_llm.py", line 251, in from_engine_args
|
|||
|
|
return cls(
|
|||
|
|
^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/async_llm.py", line 148, in __init__
|
|||
|
|
self.engine_core = EngineCoreClient.make_async_mp_client(
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core_client.py", line 124, in make_async_mp_client
|
|||
|
|
return AsyncMPClient(*client_args)
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core_client.py", line 835, in __init__
|
|||
|
|
super().__init__(
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core_client.py", line 490, in __init__
|
|||
|
|
with launch_core_engines(vllm_config, executor_class, log_stats) as (
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/contextlib.py", line 144, in __exit__
|
|||
|
|
next(self.gen)
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/utils.py", line 936, in launch_core_engines
|
|||
|
|
wait_for_engine_startup(
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/utils.py", line 995, in wait_for_engine_startup
|
|||
|
|
raise RuntimeError(
|
|||
|
|
RuntimeError: Engine core initialization failed. See root cause above. Failed core proc(s): {}
|
|||
|
|
2026-05-23 15:20:10.954 | INFO | __main__:main:128 - Shutting down Ray on head node...
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909166, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909420) /e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/vllm/vllm_engine.py:470: UserWarning: No model update group to destroy[32m [repeated 37x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909166, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909420) warnings.warn("No model update group to destroy")[32m [repeated 37x across cluster][0m
|
|||
|
|
[36m(skyrl_entrypoint pid=3659716)[0m ⚙️ Running in WANDB offline mode
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m WARNING 05-23 15:19:02 [arg_utils.py:1256] The global random seed is set to 42. Since VLLM_ENABLE_V1_MULTIPROCESSING is set to False, this may affect the random state of the Python process that launched vLLM.
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m INFO 05-23 15:19:03 [model.py:529] Resolved architecture: Qwen3ForCausalLM
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m INFO 05-23 15:19:03 [model.py:1549] Using max model len 32768
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m INFO 05-23 15:19:03 [arg_utils.py:1469] Using ray runtime env (env vars redacted): {'env_vars': {'NCCL_CUMEM_ENABLE': '***', 'RAY_ADDRESS': '***', 'VLLM_ALLOW_INSECURE_SERIALIZATION': '***', 'VLLM_ALLOW_RUNTIME_LORA_UPDATING': '***', 'VLLM_DISABLE_COMPILE_CACHE': '***', 'WANDB_API_KEY': '***'}}
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m INFO 05-23 15:19:03 [scheduler.py:224] Chunked prefill is enabled with max_num_batched_tokens=65536.
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m INFO 05-23 15:19:03 [vllm.py:690] Asynchronous scheduling is enabled.
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m WARNING 05-23 15:19:03 [vllm.py:728] Enforce eager set, overriding optimization level to -O0
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m INFO 05-23 15:19:03 [vllm.py:846] Cudagraph is disabled under eager mode
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m WARNING 05-23 15:19:03 [serial_utils.py:57] Allowing insecure serialization using pickle due to VLLM_ALLOW_INSECURE_SERIALIZATION=1
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294862, ip=10.128.17.219)[0m WARNING 05-23 15:19:03 [system_utils.py:140] We must use the `spawn` multiprocessing start method. Overriding VLLM_WORKER_MULTIPROC_METHOD to 'spawn'. See https://docs.vllm.ai/en/latest/usage/troubleshooting.html#python-multiprocessing for more information. Reasons: In a Ray actor and can only be spawned
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523613, ip=10.128.17.222)[0m WARNING 05-23 15:19:08 [arg_utils.py:1256] The global random seed is set to 47. Since VLLM_ENABLE_V1_MULTIPROCESSING is set to False, this may affect the random state of the Python process that launched vLLM.[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523613, ip=10.128.17.222)[0m INFO 05-23 15:19:08 [model.py:529] Resolved architecture: Qwen3ForCausalLM[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523613, ip=10.128.17.222)[0m INFO 05-23 15:19:08 [model.py:1549] Using max model len 32768[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523613, ip=10.128.17.222)[0m INFO 05-23 15:19:08 [arg_utils.py:1469] Using ray runtime env (env vars redacted): {'env_vars': {'NCCL_CUMEM_ENABLE': '***', 'RAY_ADDRESS': '***', 'VLLM_ALLOW_INSECURE_SERIALIZATION': '***', 'VLLM_ALLOW_RUNTIME_LORA_UPDATING': '***', 'VLLM_DISABLE_COMPILE_CACHE': '***', 'WANDB_API_KEY': '***'}}[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523613, ip=10.128.17.222)[0m INFO 05-23 15:19:08 [scheduler.py:224] Chunked prefill is enabled with max_num_batched_tokens=65536.[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523613, ip=10.128.17.222)[0m INFO 05-23 15:19:08 [vllm.py:690] Asynchronous scheduling is enabled.[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523613, ip=10.128.17.222)[0m WARNING 05-23 15:19:08 [vllm.py:728] Enforce eager set, overriding optimization level to -O0[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523613, ip=10.128.17.222)[0m INFO 05-23 15:19:08 [vllm.py:846] Cudagraph is disabled under eager mode[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523614, ip=10.128.17.222)[0m WARNING 05-23 15:19:07 [serial_utils.py:57] Allowing insecure serialization using pickle due to VLLM_ALLOW_INSECURE_SERIALIZATION=1[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523614, ip=10.128.17.222)[0m WARNING 05-23 15:19:07 [system_utils.py:140] We must use the `spawn` multiprocessing start method. Overriding VLLM_WORKER_MULTIPROC_METHOD to 'spawn'. See https://docs.vllm.ai/en/latest/usage/troubleshooting.html#python-multiprocessing for more information. Reasons: In a Ray actor and can only be spawned[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396962, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397192) INFO 05-23 15:19:12 [core.py:97] Initializing a V1 LLM engine (vdev) with config: model='/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6', speculative_config=None, tokenizer='/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6', skip_tokenizer_init=False, tokenizer_mode=auto, revision=None, tokenizer_revision=None, trust_remote_code=True, dtype=torch.bfloat16, max_seq_len=32768, download_dir=None, load_format=auto, tensor_parallel_size=1, pipeline_parallel_size=1, data_parallel_size=1, disable_custom_all_reduce=False, quantization=None, enforce_eager=True, enable_return_routed_experts=False, kv_cache_dtype=auto, device_config=cuda, structured_outputs_config=StructuredOutputsConfig(backend='auto', disable_fallback=False, disable_any_whitespace=False, disable_additional_properties=False, reasoning_parser='', reasoning_parser_plugin='', enable_in_reasoning=False), observability_config=ObservabilityConfig(show_hidden_metrics_for_version=None, otlp_traces_endpoint=None, collect_detailed_traces=None, kv_cache_metrics=False, kv_cache_metrics_sample=0.01, cudagraph_metrics=False, enable_layerwise_nvtx_tracing=False, enable_mfu_metrics=False, enable_mm_processor_stats=False, enable_logging_iteration_details=False), seed=45, served_model_name=0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6, enable_prefix_caching=True, enable_chunked_prefill=True, pooler_config=None, compilation_config={'level': None, 'mode': <CompilationMode.NONE: 0>, 'debug_dump_path': None, 'cache_dir': '', 'compile_cache_save_format': 'binary', 'backend': 'inductor', 'custom_ops': ['all'], 'splitting_ops': [], 'compile_mm_encoder': False, 'compile_sizes': [], 'compile_ranges_split_points': [65536], 'inductor_compile_config': {'enable_auto_functionalized_v2': False, 'combo_kernels': True, 'benchmark_combo_kernel': True}, 'inductor_passes': {}, 'cudagraph_mode': <CUDAGraphMode.NONE: 0>, 'cudagraph_num_of_warmups': 0, 'cudagraph_capture_sizes': [], 'cudagraph_copy_inputs': False, 'cudagraph_specialize_lora': True, 'use_inductor_graph_partition': False, 'pass_config': {'fuse_norm_quant': False, 'fuse_act_quant': False, 'fuse_attn_quant': False, 'eliminate_noops': False, 'enable_sp': False, 'fuse_gemm_comms': False, 'fuse_allreduce_rms': False, 'fuse_act_padding': False}, 'max_cudagraph_capture_size': 0, 'dynamic_shapes_config': {'type': <DynamicShapesType.BACKED: 'backed'>, 'evaluate_guards': False, 'assume_32_bit_indexing': False}, 'local_cache_dir': None, 'fast_moe_cold_start': True, 'static_all_moe_layers': []}
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3123358, ip=10.128.17.215)[0m WARNING 05-23 15:19:13 [arg_utils.py:1256] The global random seed is set to 62. Since VLLM_ENABLE_V1_MULTIPROCESSING is set to False, this may affect the random state of the Python process that launched vLLM.[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3123358, ip=10.128.17.215)[0m INFO 05-23 15:19:13 [model.py:529] Resolved architecture: Qwen3ForCausalLM[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3123358, ip=10.128.17.215)[0m INFO 05-23 15:19:13 [model.py:1549] Using max model len 32768[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3123358, ip=10.128.17.215)[0m INFO 05-23 15:19:13 [arg_utils.py:1469] Using ray runtime env (env vars redacted): {'env_vars': {'NCCL_CUMEM_ENABLE': '***', 'RAY_ADDRESS': '***', 'VLLM_ALLOW_INSECURE_SERIALIZATION': '***', 'VLLM_ALLOW_RUNTIME_LORA_UPDATING': '***', 'VLLM_DISABLE_COMPILE_CACHE': '***', 'WANDB_API_KEY': '***'}}[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3123358, ip=10.128.17.215)[0m INFO 05-23 15:19:13 [scheduler.py:224] Chunked prefill is enabled with max_num_batched_tokens=65536.[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3123358, ip=10.128.17.215)[0m INFO 05-23 15:19:13 [vllm.py:690] Asynchronous scheduling is enabled.[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3123358, ip=10.128.17.215)[0m WARNING 05-23 15:19:13 [vllm.py:728] Enforce eager set, overriding optimization level to -O0[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3123358, ip=10.128.17.215)[0m INFO 05-23 15:19:13 [vllm.py:846] Cudagraph is disabled under eager mode[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3123358, ip=10.128.17.215)[0m WARNING 05-23 15:19:13 [serial_utils.py:57] Allowing insecure serialization using pickle due to VLLM_ALLOW_INSECURE_SERIALIZATION=1[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3123358, ip=10.128.17.215)[0m WARNING 05-23 15:19:13 [system_utils.py:140] We must use the `spawn` multiprocessing start method. Overriding VLLM_WORKER_MULTIPROC_METHOD to 'spawn'. See https://docs.vllm.ai/en/latest/usage/troubleshooting.html#python-multiprocessing for more information. Reasons: In a Ray actor and can only be spawned[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396962, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397192) INFO 05-23 15:19:15 [worker_base.py:289] Injected <class 'skyrl_train.inference_engines.vllm.vllm_engine.WorkerWrap'> into <class 'vllm.v1.worker.gpu_worker.Worker'> for extended collective_rpc calls ['_apply_fp8_weight_loader_patches', '_is_fp8_model', '_quantize_weights_for_fp8', '_restore_param_subclasses', '_undo_param_subclasses', 'begin_weight_update', 'destroy_weights_update_group', 'end_weight_update', 'init_weight_update_communicator', 'load_weights', 'set_numa_affinity', 'test_rpc']
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523614, ip=10.128.17.222)[0m (EngineCore_DP0 pid=523846) INFO 05-23 15:19:17 [core.py:97] Initializing a V1 LLM engine (vdev) with config: model='/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6', speculative_config=None, tokenizer='/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6', skip_tokenizer_init=False, tokenizer_mode=auto, revision=None, tokenizer_revision=None, trust_remote_code=True, dtype=torch.bfloat16, max_seq_len=32768, download_dir=None, load_format=auto, tensor_parallel_size=1, pipeline_parallel_size=1, data_parallel_size=1, disable_custom_all_reduce=False, quantization=None, enforce_eager=True, enable_return_routed_experts=False, kv_cache_dtype=auto, device_config=cuda, structured_outputs_config=StructuredOutputsConfig(backend='auto', disable_fallback=False, disable_any_whitespace=False, disable_additional_properties=False, reasoning_parser='', reasoning_parser_plugin='', enable_in_reasoning=False), observability_config=ObservabilityConfig(show_hidden_metrics_for_version=None, otlp_traces_endpoint=None, collect_detailed_traces=None, kv_cache_metrics=False, kv_cache_metrics_sample=0.01, cudagraph_metrics=False, enable_layerwise_nvtx_tracing=False, enable_mfu_metrics=False, enable_mm_processor_stats=False, enable_logging_iteration_details=False), seed=48, served_model_name=0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6, enable_prefix_caching=True, enable_chunked_prefill=True, pooler_config=None, compilation_config={'level': None, 'mode': <CompilationMode.NONE: 0>, 'debug_dump_path': None, 'cache_dir': '', 'compile_cache_save_format': 'binary', 'backend': 'inductor', 'custom_ops': ['all'], 'splitting_ops': [], 'compile_mm_encoder': False, 'compile_sizes': [], 'compile_ranges_split_points': [65536], 'inductor_compile_config': {'enable_auto_functionalized_v2': False, 'combo_kernels': True, 'benchmark_combo_kernel': True}, 'inductor_passes': {}, 'cudagraph_mode': <CUDAGraphMode.NONE: 0>, 'cudagraph_num_of_warmups': 0, 'cudagraph_capture_sizes': [], 'cudagraph_copy_inputs': False, 'cudagraph_specialize_lora': True, 'use_inductor_graph_partition': False, 'pass_config': {'fuse_norm_quant': False, 'fuse_act_quant': False, 'fuse_attn_quant': False, 'eliminate_noops': False, 'enable_sp': False, 'fuse_gemm_comms': False, 'fuse_allreduce_rms': False, 'fuse_act_padding': False}, 'max_cudagraph_capture_size': 0, 'dynamic_shapes_config': {'type': <DynamicShapesType.BACKED: 'backed'>, 'evaluate_guards': False, 'assume_32_bit_indexing': False}, 'local_cache_dir': None, 'fast_moe_cold_start': True, 'static_all_moe_layers': []}[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396962, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397192) INFO 05-23 15:19:17 [parallel_state.py:1234] world_size=1 rank=0 local_rank=0 distributed_init_method=tcp://10.128.24.51:59091 backend=nccl
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396963, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397201) INFO 05-23 15:19:17 [parallel_state.py:1445] rank 0 in world size 1 is assigned as DP rank 0, PP rank 0, PCP rank 0, TP rank 0, EP rank N/A
|
|||
|
|
[36m(pid=2753436, ip=10.128.17.217)[0m ⚙️ Running in WANDB offline mode
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396963, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397201) INFO 05-23 15:19:18 [gpu_model_runner.py:4125] Starting to load model /e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6...
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440552, ip=10.128.24.53)[0m WARNING 05-23 15:19:18 [arg_utils.py:1256] The global random seed is set to 70. Since VLLM_ENABLE_V1_MULTIPROCESSING is set to False, this may affect the random state of the Python process that launched vLLM.[32m [repeated 16x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440552, ip=10.128.24.53)[0m INFO 05-23 15:19:18 [model.py:529] Resolved architecture: Qwen3ForCausalLM[32m [repeated 16x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440552, ip=10.128.24.53)[0m INFO 05-23 15:19:18 [model.py:1549] Using max model len 32768[32m [repeated 16x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440552, ip=10.128.24.53)[0m INFO 05-23 15:19:18 [arg_utils.py:1469] Using ray runtime env (env vars redacted): {'env_vars': {'NCCL_CUMEM_ENABLE': '***', 'RAY_ADDRESS': '***', 'VLLM_ALLOW_INSECURE_SERIALIZATION': '***', 'VLLM_ALLOW_RUNTIME_LORA_UPDATING': '***', 'VLLM_DISABLE_COMPILE_CACHE': '***', 'WANDB_API_KEY': '***'}}[32m [repeated 16x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440552, ip=10.128.24.53)[0m INFO 05-23 15:19:18 [scheduler.py:224] Chunked prefill is enabled with max_num_batched_tokens=65536.[32m [repeated 16x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440552, ip=10.128.24.53)[0m INFO 05-23 15:19:18 [vllm.py:690] Asynchronous scheduling is enabled.[32m [repeated 16x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440552, ip=10.128.24.53)[0m WARNING 05-23 15:19:18 [vllm.py:728] Enforce eager set, overriding optimization level to -O0[32m [repeated 16x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440552, ip=10.128.24.53)[0m INFO 05-23 15:19:18 [vllm.py:846] Cudagraph is disabled under eager mode[32m [repeated 16x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440681, ip=10.128.24.53)[0m WARNING 05-23 15:19:18 [serial_utils.py:57] Allowing insecure serialization using pickle due to VLLM_ALLOW_INSECURE_SERIALIZATION=1[32m [repeated 15x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440681, ip=10.128.24.53)[0m WARNING 05-23 15:19:18 [system_utils.py:140] We must use the `spawn` multiprocessing start method. Overriding VLLM_WORKER_MULTIPROC_METHOD to 'spawn'. See https://docs.vllm.ai/en/latest/usage/troubleshooting.html#python-multiprocessing for more information. Reasons: In a Ray actor and can only be spawned[32m [repeated 15x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396963, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397201) INFO 05-23 15:19:19 [cuda.py:367] Using FLASH_ATTN attention backend out of potential backends: ['FLASH_ATTN', 'FLASHINFER', 'TRITON_ATTN', 'FLEX_ATTENTION'].
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] EngineCore failed to start.
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] Traceback (most recent call last):
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core.py", line 996, in run_engine_core
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] engine_core = EngineCoreProc(*args, engine_index=dp_rank, **kwargs)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core.py", line 740, in __init__
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] super().__init__(
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core.py", line 106, in __init__
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] self.model_executor = executor_class(vllm_config)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] ^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/executor/abstract.py", line 103, in __init__
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] self._init_executor()
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/executor/uniproc_executor.py", line 47, in _init_executor
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] self.driver_worker.init_device()
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/worker/worker_base.py", line 332, in init_device
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] self.worker.init_device() # type: ignore
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] ^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/worker/gpu_worker.py", line 232, in init_device
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] init_worker_distributed_environment(
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/worker/gpu_worker.py", line 1040, in init_worker_distributed_environment
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] init_distributed_environment(
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/distributed/parallel_state.py", line 1256, in init_distributed_environment
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] torch.distributed.init_process_group(
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/distributed/c10d_logger.py", line 81, in wrapper
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] return func(*args, **kwargs)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] ^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/distributed/c10d_logger.py", line 95, in wrapper
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] func_return = func(*args, **kwargs)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] ^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/distributed/distributed_c10d.py", line 1762, in init_process_group
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] store, rank, world_size = next(rendezvous_iterator)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] ^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/distributed/rendezvous.py", line 230, in _tcp_rendezvous_handler
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] store = _create_c10d_store(
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] ^^^^^^^^^^^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/distributed/rendezvous.py", line 198, in _create_c10d_store
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] return TCPStore(
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] ^^^^^^^^^
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] torch.distributed.DistNetworkError: The server socket has failed to listen on any local network address. port: 34697, useIpv6: false, code: -98, name: EADDRINUSE, message: address already in use
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] Exception raised from makeWithPort at /pytorch/torch/csrc/distributed/c10d/TCPStoreLibUvBackend.cpp:307 (most recent call first):
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) + 0xb0 (0x40003896c700 in /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/lib/libc10.so)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #1: <unknown function> + 0x5f29220 (0x400018b09220 in /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #2: <unknown function> + 0x5f4326c (0x400018b2326c in /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #3: <unknown function> + 0x5f49074 (0x400018b29074 in /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #4: <unknown function> + 0x5f49138 (0x400018b29138 in /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #5: <unknown function> + 0x5f2ccc4 (0x400018b0ccc4 in /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #6: c10d::TCPStore::TCPStore(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, c10d::TCPStoreOptions const&) + 0x104 (0x400018b11564 in /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #7: <unknown function> + 0x109a094 (0x4000128ba094 in /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/lib/libtorch_python.so)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #8: <unknown function> + 0x113236c (0x40001295236c in /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/lib/libtorch_python.so)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #9: <unknown function> + 0x5d6d60 (0x400011df6d60 in /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/lib/libtorch_python.so)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #10: <unknown function> + 0x1b7a38 (0xaaaaca047a38 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #11: _PyObject_MakeTpCall + 0x98 (0xaaaac9ff5db8 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #12: <unknown function> + 0x169f50 (0xaaaac9ff9f50 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #13: <unknown function> + 0x1682e4 (0xaaaac9ff82e4 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #14: <unknown function> + 0x1e0ce8 (0xaaaaca070ce8 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #15: <unknown function> + 0x1d7ddc (0xaaaaca067ddc in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #16: <unknown function> + 0x646b0c (0x400011e66b0c in /e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/torch/lib/libtorch_python.so)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #17: _PyObject_MakeTpCall + 0x98 (0xaaaac9ff5db8 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #18: _PyEval_EvalFrameDefault + 0x280c (0xaaaaca0fae54 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #19: <unknown function> + 0x1808c0 (0xaaaaca0108c0 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #20: <unknown function> + 0x182bf8 (0xaaaaca012bf8 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #21: <unknown function> + 0x25fd30 (0xaaaaca0efd30 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #22: <unknown function> + 0x1b7d20 (0xaaaaca047d20 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #23: PyObject_Vectorcall + 0x54 (0xaaaac9ff60e4 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #24: _PyEval_EvalFrameDefault + 0x280c (0xaaaaca0fae54 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #25: _PyObject_FastCallDictTstate + 0x80 (0xaaaac9ff7f20 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #26: _PyObject_Call_Prepend + 0x140 (0xaaaac9ff81ec in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #27: <unknown function> + 0x1e0d80 (0xaaaaca070d80 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #28: <unknown function> + 0x1d7ddc (0xaaaaca067ddc in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #29: _PyObject_MakeTpCall + 0x98 (0xaaaac9ff5db8 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #30: _PyEval_EvalFrameDefault + 0x280c (0xaaaaca0fae54 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #31: _PyObject_FastCallDictTstate + 0x10c (0xaaaac9ff7fac in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #32: _PyObject_Call_Prepend + 0x140 (0xaaaac9ff81ec in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #33: <unknown function> + 0x1e0d80 (0xaaaaca070d80 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #34: <unknown function> + 0x1d7ddc (0xaaaaca067ddc in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #35: _PyObject_Call + 0x68 (0xaaaac9ff8488 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #36: _PyEval_EvalFrameDefault + 0x52ac (0xaaaaca0fd8f4 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #37: PyEval_EvalCode + 0xb4 (0xaaaaca102eb4 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #38: <unknown function> + 0x2ccdcc (0xaaaaca15cdcc in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #39: <unknown function> + 0x2ccef4 (0xaaaaca15cef4 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #40: PyRun_StringFlags + 0x90 (0xaaaaca161050 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #41: PyRun_SimpleStringFlags + 0x58 (0xaaaaca1610f8 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #42: Py_RunMain + 0x2c8 (0xaaaaca189190 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #43: Py_BytesMain + 0x64 (0xaaaaca189fb4 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #44: <unknown function> + 0x27540 (0x400000e67540 in /lib64/libc.so.6)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #45: __libc_start_main + 0x98 (0x400000e67618 in /lib64/libc.so.6)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006] frame #46: <unknown function> + 0x10e0c0 (0xaaaac9f9e0c0 in VLLM::EngineCore)
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) ERROR 05-23 15:19:19 [core.py:1006]
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294997, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295225) INFO 05-23 15:19:19 [worker_base.py:289] Injected <class 'skyrl_train.inference_engines.vllm.vllm_engine.WorkerWrap'> into <class 'vllm.v1.worker.gpu_worker.Worker'> for extended collective_rpc calls ['_apply_fp8_weight_loader_patches', '_is_fp8_model', '_quantize_weights_for_fp8', '_restore_param_subclasses', '_undo_param_subclasses', 'begin_weight_update', 'destroy_weights_update_group', 'end_weight_update', 'init_weight_update_communicator', 'load_weights', 'set_numa_affinity', 'test_rpc'][32m [repeated 7x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3222354, ip=10.128.17.212)[0m (EngineCore_DP0 pid=3222582) INFO 05-23 15:19:22 [core.py:97] Initializing a V1 LLM engine (vdev) with config: model='/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6', speculative_config=None, tokenizer='/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6', skip_tokenizer_init=False, tokenizer_mode=auto, revision=None, tokenizer_revision=None, trust_remote_code=True, dtype=torch.bfloat16, max_seq_len=32768, download_dir=None, load_format=auto, tensor_parallel_size=1, pipeline_parallel_size=1, data_parallel_size=1, disable_custom_all_reduce=False, quantization=None, enforce_eager=True, enable_return_routed_experts=False, kv_cache_dtype=auto, device_config=cuda, structured_outputs_config=StructuredOutputsConfig(backend='auto', disable_fallback=False, disable_any_whitespace=False, disable_additional_properties=False, reasoning_parser='', reasoning_parser_plugin='', enable_in_reasoning=False), observability_config=ObservabilityConfig(show_hidden_metrics_for_version=None, otlp_traces_endpoint=None, collect_detailed_traces=None, kv_cache_metrics=False, kv_cache_metrics_sample=0.01, cudagraph_metrics=False, enable_layerwise_nvtx_tracing=False, enable_mfu_metrics=False, enable_mm_processor_stats=False, enable_logging_iteration_details=False), seed=61, served_model_name=0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6, enable_prefix_caching=True, enable_chunked_prefill=True, pooler_config=None, compilation_config={'level': None, 'mode': <CompilationMode.NONE: 0>, 'debug_dump_path': None, 'cache_dir': '', 'compile_cache_save_format': 'binary', 'backend': 'inductor', 'custom_ops': ['all'], 'splitting_ops': [], 'compile_mm_encoder': False, 'compile_sizes': [], 'compile_ranges_split_points': [65536], 'inductor_compile_config': {'enable_auto_functionalized_v2': False, 'combo_kernels': True, 'benchmark_combo_kernel': True}, 'inductor_passes': {}, 'cudagraph_mode': <CUDAGraphMode.NONE: 0>, 'cudagraph_num_of_warmups': 0, 'cudagraph_capture_sizes': [], 'cudagraph_copy_inputs': False, 'cudagraph_specialize_lora': True, 'use_inductor_graph_partition': False, 'pass_config': {'fuse_norm_quant': False, 'fuse_act_quant': False, 'fuse_attn_quant': False, 'eliminate_noops': False, 'enable_sp': False, 'fuse_gemm_comms': False, 'fuse_allreduce_rms': False, 'fuse_act_padding': False}, 'max_cudagraph_capture_size': 0, 'dynamic_shapes_config': {'type': <DynamicShapesType.BACKED: 'backed'>, 'evaluate_guards': False, 'assume_32_bit_indexing': False}, 'local_cache_dir': None, 'fast_moe_cold_start': True, 'static_all_moe_layers': []}[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523613, ip=10.128.17.222)[0m (EngineCore_DP0 pid=523858) INFO 05-23 15:19:22 [parallel_state.py:1234] world_size=1 rank=0 local_rank=0 distributed_init_method=tcp://10.128.17.222:58805 backend=nccl[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523613, ip=10.128.17.222)[0m (EngineCore_DP0 pid=523858) INFO 05-23 15:19:22 [parallel_state.py:1445] rank 0 in world size 1 is assigned as DP rank 0, PP rank 0, PCP rank 0, TP rank 0, EP rank N/A[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523615, ip=10.128.17.222)[0m (EngineCore_DP0 pid=523862) INFO 05-23 15:19:23 [gpu_model_runner.py:4125] Starting to load model /e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6...[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m WARNING 05-23 15:19:20 [arg_utils.py:1256] The global random seed is set to 76. Since VLLM_ENABLE_V1_MULTIPROCESSING is set to False, this may affect the random state of the Python process that launched vLLM.[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m INFO 05-23 15:19:20 [model.py:529] Resolved architecture: Qwen3ForCausalLM[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m INFO 05-23 15:19:20 [model.py:1549] Using max model len 32768[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m INFO 05-23 15:19:20 [arg_utils.py:1469] Using ray runtime env (env vars redacted): {'env_vars': {'NCCL_CUMEM_ENABLE': '***', 'RAY_ADDRESS': '***', 'VLLM_ALLOW_INSECURE_SERIALIZATION': '***', 'VLLM_ALLOW_RUNTIME_LORA_UPDATING': '***', 'VLLM_DISABLE_COMPILE_CACHE': '***', 'WANDB_API_KEY': '***'}}[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m INFO 05-23 15:19:20 [scheduler.py:224] Chunked prefill is enabled with max_num_batched_tokens=65536.[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m INFO 05-23 15:19:20 [vllm.py:690] Asynchronous scheduling is enabled.[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m WARNING 05-23 15:19:20 [vllm.py:728] Enforce eager set, overriding optimization level to -O0[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m INFO 05-23 15:19:20 [vllm.py:846] Cudagraph is disabled under eager mode[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m WARNING 05-23 15:19:20 [serial_utils.py:57] Allowing insecure serialization using pickle due to VLLM_ALLOW_INSECURE_SERIALIZATION=1[32m [repeated 6x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m WARNING 05-23 15:19:20 [system_utils.py:140] We must use the `spawn` multiprocessing start method. Overriding VLLM_WORKER_MULTIPROC_METHOD to 'spawn'. See https://docs.vllm.ai/en/latest/usage/troubleshooting.html#python-multiprocessing for more information. Reasons: In a Ray actor and can only be spawned[32m [repeated 6x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3294996, ip=10.128.17.219)[0m (EngineCore_DP0 pid=3295247) INFO 05-23 15:19:21 [cuda.py:367] Using FLASH_ATTN attention backend out of potential backends: ['FLASH_ATTN', 'FLASHINFER', 'TRITON_ATTN', 'FLEX_ATTENTION'].[32m [repeated 6x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3222354, ip=10.128.17.212)[0m (EngineCore_DP0 pid=3222582) INFO 05-23 15:19:24 [worker_base.py:289] Injected <class 'skyrl_train.inference_engines.vllm.vllm_engine.WorkerWrap'> into <class 'vllm.v1.worker.gpu_worker.Worker'> for extended collective_rpc calls ['_apply_fp8_weight_loader_patches', '_is_fp8_model', '_quantize_weights_for_fp8', '_restore_param_subclasses', '_undo_param_subclasses', 'begin_weight_update', 'destroy_weights_update_group', 'end_weight_update', 'init_weight_update_communicator', 'load_weights', 'set_numa_affinity', 'test_rpc'][32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(pid=2753509, ip=10.128.17.217)[0m ⚙️ Running in WANDB offline mode
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440681, ip=10.128.24.53)[0m (EngineCore_DP0 pid=1440912) INFO 05-23 15:19:27 [core.py:97] Initializing a V1 LLM engine (vdev) with config: model='/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6', speculative_config=None, tokenizer='/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6', skip_tokenizer_init=False, tokenizer_mode=auto, revision=None, tokenizer_revision=None, trust_remote_code=True, dtype=torch.bfloat16, max_seq_len=32768, download_dir=None, load_format=auto, tensor_parallel_size=1, pipeline_parallel_size=1, data_parallel_size=1, disable_custom_all_reduce=False, quantization=None, enforce_eager=True, enable_return_routed_experts=False, kv_cache_dtype=auto, device_config=cuda, structured_outputs_config=StructuredOutputsConfig(backend='auto', disable_fallback=False, disable_any_whitespace=False, disable_additional_properties=False, reasoning_parser='', reasoning_parser_plugin='', enable_in_reasoning=False), observability_config=ObservabilityConfig(show_hidden_metrics_for_version=None, otlp_traces_endpoint=None, collect_detailed_traces=None, kv_cache_metrics=False, kv_cache_metrics_sample=0.01, cudagraph_metrics=False, enable_layerwise_nvtx_tracing=False, enable_mfu_metrics=False, enable_mm_processor_stats=False, enable_logging_iteration_details=False), seed=72, served_model_name=0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6, enable_prefix_caching=True, enable_chunked_prefill=True, pooler_config=None, compilation_config={'level': None, 'mode': <CompilationMode.NONE: 0>, 'debug_dump_path': None, 'cache_dir': '', 'compile_cache_save_format': 'binary', 'backend': 'inductor', 'custom_ops': ['all'], 'splitting_ops': [], 'compile_mm_encoder': False, 'compile_sizes': [], 'compile_ranges_split_points': [65536], 'inductor_compile_config': {'enable_auto_functionalized_v2': False, 'combo_kernels': True, 'benchmark_combo_kernel': True}, 'inductor_passes': {}, 'cudagraph_mode': <CUDAGraphMode.NONE: 0>, 'cudagraph_num_of_warmups': 0, 'cudagraph_capture_sizes': [], 'cudagraph_copy_inputs': False, 'cudagraph_specialize_lora': True, 'use_inductor_graph_partition': False, 'pass_config': {'fuse_norm_quant': False, 'fuse_act_quant': False, 'fuse_attn_quant': False, 'eliminate_noops': False, 'enable_sp': False, 'fuse_gemm_comms': False, 'fuse_allreduce_rms': False, 'fuse_act_padding': False}, 'max_cudagraph_capture_size': 0, 'dynamic_shapes_config': {'type': <DynamicShapesType.BACKED: 'backed'>, 'evaluate_guards': False, 'assume_32_bit_indexing': False}, 'local_cache_dir': None, 'fast_moe_cold_start': True, 'static_all_moe_layers': []}[32m [repeated 15x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909165, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909404) INFO 05-23 15:19:27 [parallel_state.py:1234] world_size=1 rank=0 local_rank=0 distributed_init_method=tcp://10.128.17.216:50115 backend=nccl[32m [repeated 13x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200758, ip=10.128.17.221)[0m (EngineCore_DP0 pid=201001) INFO 05-23 15:19:27 [parallel_state.py:1445] rank 0 in world size 1 is assigned as DP rank 0, PP rank 0, PCP rank 0, TP rank 0, EP rank N/A[32m [repeated 16x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200758, ip=10.128.17.221)[0m (EngineCore_DP0 pid=201001) INFO 05-23 15:19:28 [gpu_model_runner.py:4125] Starting to load model /e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6...[32m [repeated 15x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m WARNING 05-23 15:19:26 [arg_utils.py:1256] The global random seed is set to 86. Since VLLM_ENABLE_V1_MULTIPROCESSING is set to False, this may affect the random state of the Python process that launched vLLM.[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m INFO 05-23 15:19:26 [model.py:529] Resolved architecture: Qwen3ForCausalLM[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m INFO 05-23 15:19:26 [model.py:1549] Using max model len 32768[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m INFO 05-23 15:19:26 [arg_utils.py:1469] Using ray runtime env (env vars redacted): {'env_vars': {'NCCL_CUMEM_ENABLE': '***', 'RAY_ADDRESS': '***', 'VLLM_ALLOW_INSECURE_SERIALIZATION': '***', 'VLLM_ALLOW_RUNTIME_LORA_UPDATING': '***', 'VLLM_DISABLE_COMPILE_CACHE': '***', 'WANDB_API_KEY': '***'}}[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m INFO 05-23 15:19:26 [scheduler.py:224] Chunked prefill is enabled with max_num_batched_tokens=65536.[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m INFO 05-23 15:19:26 [vllm.py:690] Asynchronous scheduling is enabled.[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m WARNING 05-23 15:19:26 [vllm.py:728] Enforce eager set, overriding optimization level to -O0[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m INFO 05-23 15:19:26 [vllm.py:846] Cudagraph is disabled under eager mode[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m WARNING 05-23 15:19:26 [serial_utils.py:57] Allowing insecure serialization using pickle due to VLLM_ALLOW_INSECURE_SERIALIZATION=1[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m WARNING 05-23 15:19:26 [system_utils.py:140] We must use the `spawn` multiprocessing start method. Overriding VLLM_WORKER_MULTIPROC_METHOD to 'spawn'. See https://docs.vllm.ai/en/latest/usage/troubleshooting.html#python-multiprocessing for more information. Reasons: In a Ray actor and can only be spawned[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909165, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909404) INFO 05-23 15:19:29 [cuda.py:367] Using FLASH_ATTN attention backend out of potential backends: ['FLASH_ATTN', 'FLASHINFER', 'TRITON_ATTN', 'FLEX_ATTENTION'].[32m [repeated 20x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440681, ip=10.128.24.53)[0m (EngineCore_DP0 pid=1440912) INFO 05-23 15:19:30 [worker_base.py:289] Injected <class 'skyrl_train.inference_engines.vllm.vllm_engine.WorkerWrap'> into <class 'vllm.v1.worker.gpu_worker.Worker'> for extended collective_rpc calls ['_apply_fp8_weight_loader_patches', '_is_fp8_model', '_quantize_weights_for_fp8', '_restore_param_subclasses', '_undo_param_subclasses', 'begin_weight_update', 'destroy_weights_update_group', 'end_weight_update', 'init_weight_update_communicator', 'load_weights', 'set_numa_affinity', 'test_rpc'][32m [repeated 18x across cluster][0m
|
|||
|
|
[36m(pid=3659917)[0m ⚙️ Running in WANDB offline mode[32m [repeated 6x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503354, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503581) INFO 05-23 15:19:31 [core.py:97] Initializing a V1 LLM engine (vdev) with config: model='/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6', speculative_config=None, tokenizer='/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6', skip_tokenizer_init=False, tokenizer_mode=auto, revision=None, tokenizer_revision=None, trust_remote_code=True, dtype=torch.bfloat16, max_seq_len=32768, download_dir=None, load_format=auto, tensor_parallel_size=1, pipeline_parallel_size=1, data_parallel_size=1, disable_custom_all_reduce=False, quantization=None, enforce_eager=True, enable_return_routed_experts=False, kv_cache_dtype=auto, device_config=cuda, structured_outputs_config=StructuredOutputsConfig(backend='auto', disable_fallback=False, disable_any_whitespace=False, disable_additional_properties=False, reasoning_parser='', reasoning_parser_plugin='', enable_in_reasoning=False), observability_config=ObservabilityConfig(show_hidden_metrics_for_version=None, otlp_traces_endpoint=None, collect_detailed_traces=None, kv_cache_metrics=False, kv_cache_metrics_sample=0.01, cudagraph_metrics=False, enable_layerwise_nvtx_tracing=False, enable_mfu_metrics=False, enable_mm_processor_stats=False, enable_logging_iteration_details=False), seed=83, served_model_name=0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6, enable_prefix_caching=True, enable_chunked_prefill=True, pooler_config=None, compilation_config={'level': None, 'mode': <CompilationMode.NONE: 0>, 'debug_dump_path': None, 'cache_dir': '', 'compile_cache_save_format': 'binary', 'backend': 'inductor', 'custom_ops': ['all'], 'splitting_ops': [], 'compile_mm_encoder': False, 'compile_sizes': [], 'compile_ranges_split_points': [65536], 'inductor_compile_config': {'enable_auto_functionalized_v2': False, 'combo_kernels': True, 'benchmark_combo_kernel': True}, 'inductor_passes': {}, 'cudagraph_mode': <CUDAGraphMode.NONE: 0>, 'cudagraph_num_of_warmups': 0, 'cudagraph_capture_sizes': [], 'cudagraph_copy_inputs': False, 'cudagraph_specialize_lora': True, 'use_inductor_graph_partition': False, 'pass_config': {'fuse_norm_quant': False, 'fuse_act_quant': False, 'fuse_attn_quant': False, 'eliminate_noops': False, 'enable_sp': False, 'fuse_gemm_comms': False, 'fuse_allreduce_rms': False, 'fuse_act_padding': False}, 'max_cudagraph_capture_size': 0, 'dynamic_shapes_config': {'type': <DynamicShapesType.BACKED: 'backed'>, 'evaluate_guards': False, 'assume_32_bit_indexing': False}, 'local_cache_dir': None, 'fast_moe_cold_start': True, 'static_all_moe_layers': []}[32m [repeated 7x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440552, ip=10.128.24.53)[0m (EngineCore_DP0 pid=1440918) INFO 05-23 15:19:32 [parallel_state.py:1234] world_size=1 rank=0 local_rank=0 distributed_init_method=tcp://10.128.24.53:39961 backend=nccl[32m [repeated 12x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396835, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397197) INFO 05-23 15:19:32 [default_loader.py:293] Loading weights took 13.48 seconds
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440552, ip=10.128.24.53)[0m (EngineCore_DP0 pid=1440918) INFO 05-23 15:19:32 [parallel_state.py:1445] rank 0 in world size 1 is assigned as DP rank 0, PP rank 0, PCP rank 0, TP rank 0, EP rank N/A[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396835, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397197) INFO 05-23 15:19:33 [gpu_model_runner.py:4222] Model loading took 15.27 GiB memory and 14.270659 seconds
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440552, ip=10.128.24.53)[0m (EngineCore_DP0 pid=1440918) INFO 05-23 15:19:32 [gpu_model_runner.py:4125] Starting to load model /e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6...[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440680, ip=10.128.24.53)[0m (EngineCore_DP0 pid=1440929) INFO 05-23 15:19:33 [cuda.py:367] Using FLASH_ATTN attention backend out of potential backends: ['FLASH_ATTN', 'FLASHINFER', 'TRITON_ATTN', 'FLEX_ATTENTION'].[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396963, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397201) INFO 05-23 15:19:35 [gpu_worker.py:373] Available KV cache memory: 49.32 GiB
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396963, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397201) INFO 05-23 15:19:35 [kv_cache_utils.py:1307] GPU KV cache size: 359,104 tokens
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396963, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397201) INFO 05-23 15:19:35 [kv_cache_utils.py:1312] Maximum concurrency for 32,768 tokens per request: 10.96x
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503354, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503581) INFO 05-23 15:19:34 [worker_base.py:289] Injected <class 'skyrl_train.inference_engines.vllm.vllm_engine.WorkerWrap'> into <class 'vllm.v1.worker.gpu_worker.Worker'> for extended collective_rpc calls ['_apply_fp8_weight_loader_patches', '_is_fp8_model', '_quantize_weights_for_fp8', '_restore_param_subclasses', '_undo_param_subclasses', 'begin_weight_update', 'destroy_weights_update_group', 'end_weight_update', 'init_weight_update_communicator', 'load_weights', 'set_numa_affinity', 'test_rpc'][32m [repeated 7x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396835, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397197) INFO 05-23 15:19:36 [kernel_warmup.py:44] Skipping FlashInfer autotune because it is disabled.
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2397096, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397216) INFO 05-23 15:19:36 [core.py:278] init engine (profile, create kv cache, warmup model) took 2.76 seconds
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396835, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397197) WARNING 05-23 15:19:36 [serial_utils.py:57] Allowing insecure serialization using pickle due to VLLM_ALLOW_INSECURE_SERIALIZATION=1
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396835, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397197) INFO 05-23 15:19:36 [vllm.py:690] Asynchronous scheduling is enabled.
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396835, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397197) WARNING 05-23 15:19:36 [vllm.py:735] Inductor compilation was disabled by user settings, optimizations settings that are only active during inductor compilation will be ignored.
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396835, ip=10.128.24.51)[0m (EngineCore_DP0 pid=2397197) INFO 05-23 15:19:36 [vllm.py:846] Cudagraph is disabled under eager mode
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2396835, ip=10.128.24.51)[0m WARNING 05-23 15:19:36 [model.py:1350] Default vLLM sampling parameters have been overridden by the model's `generation_config.json`: `{'temperature': 0.6, 'top_k': 20, 'top_p': 0.95}`. If this is not intended, please relaunch vLLM instance with `--generation-config vllm`.
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051742, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051996) INFO 05-23 15:19:37 [core.py:97] Initializing a V1 LLM engine (vdev) with config: model='/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6', speculative_config=None, tokenizer='/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6', skip_tokenizer_init=False, tokenizer_mode=auto, revision=None, tokenizer_revision=None, trust_remote_code=True, dtype=torch.bfloat16, max_seq_len=32768, download_dir=None, load_format=auto, tensor_parallel_size=1, pipeline_parallel_size=1, data_parallel_size=1, disable_custom_all_reduce=False, quantization=None, enforce_eager=True, enable_return_routed_experts=False, kv_cache_dtype=auto, device_config=cuda, structured_outputs_config=StructuredOutputsConfig(backend='auto', disable_fallback=False, disable_any_whitespace=False, disable_additional_properties=False, reasoning_parser='', reasoning_parser_plugin='', enable_in_reasoning=False), observability_config=ObservabilityConfig(show_hidden_metrics_for_version=None, otlp_traces_endpoint=None, collect_detailed_traces=None, kv_cache_metrics=False, kv_cache_metrics_sample=0.01, cudagraph_metrics=False, enable_layerwise_nvtx_tracing=False, enable_mfu_metrics=False, enable_mm_processor_stats=False, enable_logging_iteration_details=False), seed=87, served_model_name=0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6, enable_prefix_caching=True, enable_chunked_prefill=True, pooler_config=None, compilation_config={'level': None, 'mode': <CompilationMode.NONE: 0>, 'debug_dump_path': None, 'cache_dir': '', 'compile_cache_save_format': 'binary', 'backend': 'inductor', 'custom_ops': ['all'], 'splitting_ops': [], 'compile_mm_encoder': False, 'compile_sizes': [], 'compile_ranges_split_points': [65536], 'inductor_compile_config': {'enable_auto_functionalized_v2': False, 'combo_kernels': True, 'benchmark_combo_kernel': True}, 'inductor_passes': {}, 'cudagraph_mode': <CUDAGraphMode.NONE: 0>, 'cudagraph_num_of_warmups': 0, 'cudagraph_capture_sizes': [], 'cudagraph_copy_inputs': False, 'cudagraph_specialize_lora': True, 'use_inductor_graph_partition': False, 'pass_config': {'fuse_norm_quant': False, 'fuse_act_quant': False, 'fuse_attn_quant': False, 'eliminate_noops': False, 'enable_sp': False, 'fuse_gemm_comms': False, 'fuse_allreduce_rms': False, 'fuse_act_padding': False}, 'max_cudagraph_capture_size': 0, 'dynamic_shapes_config': {'type': <DynamicShapesType.BACKED: 'backed'>, 'evaluate_guards': False, 'assume_32_bit_indexing': False}, 'local_cache_dir': None, 'fast_moe_cold_start': True, 'static_all_moe_layers': []}[32m [repeated 7x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503354, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503581) INFO 05-23 15:19:35 [parallel_state.py:1234] world_size=1 rank=0 local_rank=0 distributed_init_method=tcp://10.128.24.50:55605 backend=nccl[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200758, ip=10.128.17.221)[0m (EngineCore_DP0 pid=201001) INFO 05-23 15:19:37 [default_loader.py:293] Loading weights took 8.24 seconds[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503354, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503581) INFO 05-23 15:19:35 [parallel_state.py:1445] rank 0 in world size 1 is assigned as DP rank 0, PP rank 0, PCP rank 0, TP rank 0, EP rank N/A[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200758, ip=10.128.17.221)[0m (EngineCore_DP0 pid=201001) INFO 05-23 15:19:37 [gpu_model_runner.py:4222] Model loading took 15.27 GiB memory and 8.907134 seconds[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503354, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503581) INFO 05-23 15:19:35 [gpu_model_runner.py:4125] Starting to load model /e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6...[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503354, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503581) INFO 05-23 15:19:36 [cuda.py:367] Using FLASH_ATTN attention backend out of potential backends: ['FLASH_ATTN', 'FLASHINFER', 'TRITON_ATTN', 'FLEX_ATTENTION'].[32m [repeated 5x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200758, ip=10.128.17.221)[0m (EngineCore_DP0 pid=201001) INFO 05-23 15:19:39 [gpu_worker.py:373] Available KV cache memory: 49.32 GiB[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200758, ip=10.128.17.221)[0m (EngineCore_DP0 pid=201001) INFO 05-23 15:19:39 [kv_cache_utils.py:1307] GPU KV cache size: 359,104 tokens[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200758, ip=10.128.17.221)[0m (EngineCore_DP0 pid=201001) INFO 05-23 15:19:39 [kv_cache_utils.py:1312] Maximum concurrency for 32,768 tokens per request: 10.96x[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051740, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051998) INFO 05-23 15:19:39 [worker_base.py:289] Injected <class 'skyrl_train.inference_engines.vllm.vllm_engine.WorkerWrap'> into <class 'vllm.v1.worker.gpu_worker.Worker'> for extended collective_rpc calls ['_apply_fp8_weight_loader_patches', '_is_fp8_model', '_quantize_weights_for_fp8', '_restore_param_subclasses', '_undo_param_subclasses', 'begin_weight_update', 'destroy_weights_update_group', 'end_weight_update', 'init_weight_update_communicator', 'load_weights', 'set_numa_affinity', 'test_rpc'][32m [repeated 6x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200757, ip=10.128.17.221)[0m (EngineCore_DP0 pid=200995) INFO 05-23 15:19:40 [kernel_warmup.py:44] Skipping FlashInfer autotune because it is disabled.[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200758, ip=10.128.17.221)[0m (EngineCore_DP0 pid=201001) INFO 05-23 15:19:40 [core.py:278] init engine (profile, create kv cache, warmup model) took 2.82 seconds[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200758, ip=10.128.17.221)[0m (EngineCore_DP0 pid=201001) WARNING 05-23 15:19:40 [serial_utils.py:57] Allowing insecure serialization using pickle due to VLLM_ALLOW_INSECURE_SERIALIZATION=1[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200758, ip=10.128.17.221)[0m (EngineCore_DP0 pid=201001) INFO 05-23 15:19:40 [vllm.py:690] Asynchronous scheduling is enabled.[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200758, ip=10.128.17.221)[0m (EngineCore_DP0 pid=201001) WARNING 05-23 15:19:40 [vllm.py:735] Inductor compilation was disabled by user settings, optimizations settings that are only active during inductor compilation will be ignored.[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200758, ip=10.128.17.221)[0m (EngineCore_DP0 pid=201001) INFO 05-23 15:19:40 [vllm.py:846] Cudagraph is disabled under eager mode[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=200758, ip=10.128.17.221)[0m WARNING 05-23 15:19:40 [model.py:1350] Default vLLM sampling parameters have been overridden by the model's `generation_config.json`: `{'temperature': 0.6, 'top_k': 20, 'top_p': 0.95}`. If this is not intended, please relaunch vLLM instance with `--generation-config vllm`.[32m [repeated 10x across cluster][0m
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m NCCL version 2.27.7+cuda13.0
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m [rank-0]: Loading model from /e/data1/datasets/playground/ot-baf/a3-rl-DCAgent_code-contests-noblock/a3-rl-DCAgent_code-contests-noblock/checkpoints/global_step_18/policy/model_world_size_8_rank_0.pt
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m [rank-0]: Loading extra_state from /e/data1/datasets/playground/ot-baf/a3-rl-DCAgent_code-contests-noblock/a3-rl-DCAgent_code-contests-noblock/checkpoints/global_step_18/policy/extra_state_world_size_8_rank_0.pt
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m [rank-0]: Loading optim from /e/data1/datasets/playground/ot-baf/a3-rl-DCAgent_code-contests-noblock/a3-rl-DCAgent_code-contests-noblock/checkpoints/global_step_18/policy/optim_world_size_8_rank_0.pt
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051742, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051996) INFO 05-23 15:19:40 [parallel_state.py:1234] world_size=1 rank=0 local_rank=0 distributed_init_method=tcp://10.128.17.214:50715 backend=nccl[32m [repeated 7x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909165, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909404) INFO 05-23 15:19:43 [default_loader.py:293] Loading weights took 13.72 seconds[32m [repeated 12x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051742, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051996) INFO 05-23 15:19:40 [parallel_state.py:1445] rank 0 in world size 1 is assigned as DP rank 0, PP rank 0, PCP rank 0, TP rank 0, EP rank N/A[32m [repeated 7x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909165, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909404) INFO 05-23 15:19:43 [gpu_model_runner.py:4222] Model loading took 15.27 GiB memory and 14.989906 seconds[32m [repeated 12x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051742, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051996) INFO 05-23 15:19:41 [gpu_model_runner.py:4125] Starting to load model /e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6...[32m [repeated 7x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051740, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051998) INFO 05-23 15:19:41 [cuda.py:367] Using FLASH_ATTN attention backend out of potential backends: ['FLASH_ATTN', 'FLASHINFER', 'TRITON_ATTN', 'FLEX_ATTENTION'].[32m [repeated 7x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909038, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909396) INFO 05-23 15:19:45 [gpu_worker.py:373] Available KV cache memory: 49.32 GiB[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909038, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909396) INFO 05-23 15:19:45 [kv_cache_utils.py:1307] GPU KV cache size: 359,104 tokens[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909038, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909396) INFO 05-23 15:19:45 [kv_cache_utils.py:1312] Maximum concurrency for 32,768 tokens per request: 10.96x[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051742, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051996) INFO 05-23 15:19:40 [worker_base.py:289] Injected <class 'skyrl_train.inference_engines.vllm.vllm_engine.WorkerWrap'> into <class 'vllm.v1.worker.gpu_worker.Worker'> for extended collective_rpc calls ['_apply_fp8_weight_loader_patches', '_is_fp8_model', '_quantize_weights_for_fp8', '_restore_param_subclasses', '_undo_param_subclasses', 'begin_weight_update', 'destroy_weights_update_group', 'end_weight_update', 'init_weight_update_communicator', 'load_weights', 'set_numa_affinity', 'test_rpc']
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909038, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909396) INFO 05-23 15:19:46 [kernel_warmup.py:44] Skipping FlashInfer autotune because it is disabled.[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909038, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909396) INFO 05-23 15:19:46 [core.py:278] init engine (profile, create kv cache, warmup model) took 2.73 seconds[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909038, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909396) WARNING 05-23 15:19:46 [serial_utils.py:57] Allowing insecure serialization using pickle due to VLLM_ALLOW_INSECURE_SERIALIZATION=1[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909038, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909396) INFO 05-23 15:19:46 [vllm.py:690] Asynchronous scheduling is enabled.[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909038, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909396) WARNING 05-23 15:19:46 [vllm.py:735] Inductor compilation was disabled by user settings, optimizations settings that are only active during inductor compilation will be ignored.[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1909038, ip=10.128.17.216)[0m (EngineCore_DP0 pid=1909396) INFO 05-23 15:19:46 [vllm.py:846] Cudagraph is disabled under eager mode[32m [repeated 9x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3123494, ip=10.128.17.215)[0m WARNING 05-23 15:19:42 [model.py:1350] Default vLLM sampling parameters have been overridden by the model's `generation_config.json`: `{'temperature': 0.6, 'top_k': 20, 'top_p': 0.95}`. If this is not intended, please relaunch vLLM instance with `--generation-config vllm`.[32m [repeated 8x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172774, ip=10.128.24.52)[0m (EngineCore_DP0 pid=2173007) INFO 05-23 15:19:49 [default_loader.py:293] Loading weights took 13.85 seconds[32m [repeated 20x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172774, ip=10.128.24.52)[0m (EngineCore_DP0 pid=2173007) INFO 05-23 15:19:49 [gpu_model_runner.py:4222] Model loading took 15.27 GiB memory and 14.998130 seconds[32m [repeated 20x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503222, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503605) INFO 05-23 15:19:51 [gpu_worker.py:373] Available KV cache memory: 49.32 GiB[32m [repeated 12x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503222, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503605) INFO 05-23 15:19:51 [kv_cache_utils.py:1307] GPU KV cache size: 359,104 tokens[32m [repeated 12x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503222, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503605) INFO 05-23 15:19:51 [kv_cache_utils.py:1312] Maximum concurrency for 32,768 tokens per request: 10.96x[32m [repeated 12x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440680, ip=10.128.24.53)[0m (EngineCore_DP0 pid=1440929) INFO 05-23 15:19:50 [kernel_warmup.py:44] Skipping FlashInfer autotune because it is disabled.[32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440680, ip=10.128.24.53)[0m (EngineCore_DP0 pid=1440929) INFO 05-23 15:19:50 [core.py:278] init engine (profile, create kv cache, warmup model) took 2.82 seconds[32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503354, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503581) WARNING 05-23 15:19:51 [serial_utils.py:57] Allowing insecure serialization using pickle due to VLLM_ALLOW_INSECURE_SERIALIZATION=1[32m [repeated 12x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503354, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503581) INFO 05-23 15:19:51 [vllm.py:690] Asynchronous scheduling is enabled.[32m [repeated 12x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503354, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503581) WARNING 05-23 15:19:51 [vllm.py:735] Inductor compilation was disabled by user settings, optimizations settings that are only active during inductor compilation will be ignored.[32m [repeated 12x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1503354, ip=10.128.24.50)[0m (EngineCore_DP0 pid=1503581) INFO 05-23 15:19:51 [vllm.py:846] Cudagraph is disabled under eager mode[32m [repeated 12x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=1440681, ip=10.128.24.53)[0m WARNING 05-23 15:19:51 [model.py:1350] Default vLLM sampling parameters have been overridden by the model's `generation_config.json`: `{'temperature': 0.6, 'top_k': 20, 'top_p': 0.95}`. If this is not intended, please relaunch vLLM instance with `--generation-config vllm`.[32m [repeated 12x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3052003) INFO 05-23 15:19:56 [default_loader.py:293] Loading weights took 14.98 seconds
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m (EngineCore_DP0 pid=2173022) INFO 05-23 15:19:52 [gpu_worker.py:373] Available KV cache memory: 49.32 GiB[32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m (EngineCore_DP0 pid=2173022) INFO 05-23 15:19:52 [kv_cache_utils.py:1307] GPU KV cache size: 359,104 tokens[32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m (EngineCore_DP0 pid=2173022) INFO 05-23 15:19:52 [kv_cache_utils.py:1312] Maximum concurrency for 32,768 tokens per request: 10.96x[32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m (EngineCore_DP0 pid=2173022) INFO 05-23 15:19:52 [kernel_warmup.py:44] Skipping FlashInfer autotune because it is disabled.[32m [repeated 12x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172774, ip=10.128.24.52)[0m (EngineCore_DP0 pid=2173007) INFO 05-23 15:19:52 [core.py:278] init engine (profile, create kv cache, warmup model) took 2.78 seconds[32m [repeated 12x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051742, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051996) INFO 05-23 15:19:56 [default_loader.py:293] Loading weights took 14.91 seconds
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3052003) INFO 05-23 15:19:57 [gpu_model_runner.py:4222] Model loading took 15.27 GiB memory and 16.066204 seconds
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172774, ip=10.128.24.52)[0m (EngineCore_DP0 pid=2173007) WARNING 05-23 15:19:52 [serial_utils.py:57] Allowing insecure serialization using pickle due to VLLM_ALLOW_INSECURE_SERIALIZATION=1[32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172774, ip=10.128.24.52)[0m (EngineCore_DP0 pid=2173007) INFO 05-23 15:19:52 [vllm.py:690] Asynchronous scheduling is enabled.[32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172774, ip=10.128.24.52)[0m (EngineCore_DP0 pid=2173007) WARNING 05-23 15:19:52 [vllm.py:735] Inductor compilation was disabled by user settings, optimizations settings that are only active during inductor compilation will be ignored.[32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172774, ip=10.128.24.52)[0m (EngineCore_DP0 pid=2173007) INFO 05-23 15:19:52 [vllm.py:846] Cudagraph is disabled under eager mode[32m [repeated 11x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=2172773, ip=10.128.24.52)[0m WARNING 05-23 15:19:53 [model.py:1350] Default vLLM sampling parameters have been overridden by the model's `generation_config.json`: `{'temperature': 0.6, 'top_k': 20, 'top_p': 0.95}`. If this is not intended, please relaunch vLLM instance with `--generation-config vllm`.[32m [repeated 12x across cluster][0m
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m [rank-0]: Successfully loaded model state dict
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3052003) INFO 05-23 15:19:59 [gpu_worker.py:373] Available KV cache memory: 49.32 GiB[32m [repeated 4x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3052003) INFO 05-23 15:19:59 [kv_cache_utils.py:1307] GPU KV cache size: 359,104 tokens[32m [repeated 4x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3052003) INFO 05-23 15:19:59 [kv_cache_utils.py:1312] Maximum concurrency for 32,768 tokens per request: 10.96x[32m [repeated 4x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051741, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051990) INFO 05-23 15:20:00 [kernel_warmup.py:44] Skipping FlashInfer autotune because it is disabled.[32m [repeated 4x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051741, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051990) INFO 05-23 15:20:00 [core.py:278] init engine (profile, create kv cache, warmup model) took 2.82 seconds[32m [repeated 4x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051741, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051990) INFO 05-23 15:19:56 [default_loader.py:293] Loading weights took 15.06 seconds[32m [repeated 2x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051741, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3051990) INFO 05-23 15:19:57 [gpu_model_runner.py:4222] Model loading took 15.27 GiB memory and 16.070195 seconds[32m [repeated 3x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3052003) WARNING 05-23 15:20:00 [serial_utils.py:57] Allowing insecure serialization using pickle due to VLLM_ALLOW_INSECURE_SERIALIZATION=1[32m [repeated 4x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3052003) INFO 05-23 15:20:00 [vllm.py:690] Asynchronous scheduling is enabled.[32m [repeated 4x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3052003) WARNING 05-23 15:20:00 [vllm.py:735] Inductor compilation was disabled by user settings, optimizations settings that are only active during inductor compilation will be ignored.[32m [repeated 4x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051613, ip=10.128.17.214)[0m (EngineCore_DP0 pid=3052003) INFO 05-23 15:20:00 [vllm.py:846] Cudagraph is disabled under eager mode[32m [repeated 4x across cluster][0m
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3051741, ip=10.128.17.214)[0m WARNING 05-23 15:20:00 [model.py:1350] Default vLLM sampling parameters have been overridden by the model's `generation_config.json`: `{'temperature': 0.6, 'top_k': 20, 'top_p': 0.95}`. If this is not intended, please relaunch vLLM instance with `--generation-config vllm`.[32m [repeated 4x across cluster][0m
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m [rank-0]: Successfully loaded optimizer state
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m [rank-0]: Successfully loaded scheduler state
|
|||
|
|
[36m(FSDPPolicyWorkerBase pid=2753436, ip=10.128.17.217)[0m [rank-0]: Checkpoint loaded successfully from /e/data1/datasets/playground/ot-baf/a3-rl-DCAgent_code-contests-noblock/a3-rl-DCAgent_code-contests-noblock/checkpoints/global_step_18/policy
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=523482, ip=10.128.17.222)[0m ERROR 05-23 15:20:10 [core_client.py:616] Engine core proc EngineCore_DP0 died unexpectedly, shutting down client.
|
|||
|
|
[36m(AsyncVLLMInferenceEngine pid=3123358, ip=10.128.17.215)[0m ERROR 05-23 15:20:10 [core_client.py:616] Engine core proc EngineCore_DP0 died unexpectedly, shutting down client.[32m [repeated 12x across cluster][0m
|
|||
|
|
Error executing job with overrides: ['+terminal_bench_config=terminal_bench', 'trainer.strategy=fsdp2', 'trainer.algorithm.advantage_estimator=rloo_n', 'trainer.algorithm.use_kl_loss=false', 'trainer.algorithm.kl_loss_coef=0.0', 'trainer.algorithm.eps_clip_low=0.2', 'trainer.algorithm.eps_clip_high=0.05', 'trainer.algorithm.loss_reduction=token_mean', 'trainer.epochs=2', 'trainer.max_steps=80', 'trainer.update_epochs_per_batch=1', 'trainer.train_batch_size=64', 'trainer.policy_mini_batch_size=64', 'trainer.eval_batch_size=64', 'trainer.micro_forward_batch_size_per_gpu=4', 'trainer.micro_train_batch_size_per_gpu=1', 'trainer.max_prompt_length=999999', 'trainer.eval_interval=999999', 'trainer.eval_before_train=false', 'trainer.ckpt_interval=2', 'trainer.resume_mode=latest', 'trainer.hf_save_interval=5', '++trainer.hf_hub_repo_id=laion/a3-rl-DCAgent_code-contests-noblock', '++trainer.hf_hub_private=false', '++trainer.hf_hub_revision=main', '++trainer.enable_db_registration=true', 'trainer.project_name=OpenThoughts-Agent', 'trainer.log_level=INFO', 'trainer.tracker_commit_each_step=true', 'trainer.logger=console', 'trainer.run_name=a3-rl-DCAgent_code-contests-noblock', 'trainer.ckpt_path=/e/data1/datasets/playground/ot-baf/a3-rl-DCAgent_code-contests-noblock/a3-rl-DCAgent_code-contests-noblock/checkpoints', 'trainer.export_path=/e/data1/datasets/playground/ot-baf/a3-rl-DCAgent_code-contests-noblock/a3-rl-DCAgent_code-contests-noblock/exports', 'trainer.policy.optimizer_config.lr=8e-6', 'trainer.policy.optimizer_config.weight_decay=0.0', 'trainer.policy.optimizer_config.adam_betas=[0.9,0.999]', 'trainer.policy.optimizer_config.max_grad_norm=0.9', 'trainer.policy.fsdp_config.cpu_offload=false', 'trainer.policy.fsdp_config.reshard_after_forward=true', 'trainer.policy.fsdp_config.fsdp_size=4', 'trainer.policy.model.path=/e/data1/datasets/playground/ot-baf/hf_hub/models--laion--GLM-4_7-swesmith-sandboxes-with_tests-oracle_verified_120s-maxeps-131k-fixthink/snapshots/0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6', 'trainer.ref.fsdp_config.cpu_offload=false', 'trainer.ref.fsdp_config.reshard_after_forward=true', 'trainer.ref.fsdp_config.fsdp_size=4', 'trainer.placement.colocate_all=false', 'trainer.placement.policy_num_nodes=2', 'trainer.placement.ref_num_nodes=2', 'trainer.placement.policy_num_gpus_per_node=4', 'trainer.placement.ref_num_gpus_per_node=4', 'trainer.fully_async.max_staleness_steps=16', 'trainer.fully_async.num_parallel_generation_workers=450', 'generator.backend=vllm', 'generator.timeout_multiplier=1.0', 'generator.model_dtype=bfloat16', 'generator.inference_engine_tensor_parallel_size=1', 'generator.num_inference_engines=48', 'generator.n_samples_per_prompt=8', 'generator.eval_n_samples_per_prompt=8', 'generator.gpu_memory_utilization=0.75', 'generator.max_num_seqs=24', 'generator.max_num_batched_tokens=65536', 'generator.enable_prefix_caching=true', 'generator.enable_chunked_prefill=true', 'generator.run_engines_locally=true', 'generator.weight_sync_backend=nccl', 'generator.async_engine=true', 'generator.batched=false', 'generator.enable_http_endpoint=true', 'generator.enable_ray_prometheus_stats=false', 'generator.vllm_stats_interval=1', 'generator.append_eos_token_after_stop_str_in_multi_turn=true', 'generator.max_turns=999999', 'generator.sampling_params.max_generate_length=4096', 'generator.sampling_params.temperature=0.7', 'generator.sampling_params.top_p=0.95', 'generator.sampling_params.top_k=20', '++generator.engine_init_kwargs.max_model_len=32768', '++generator.engine_init_kwargs.custom_chat_template_chat_completion_path=chat_templates/qwen3_thinking_acc.jinja2', '++generator.engine_init_kwargs.served_model_name=0e3bff0c4e51f6b9ec0713b98b9eec36efb91cc6', 'data.train_data=["/e/scratch/jureap59/feuer1/tasks/code-contests-noblock"]', 'data.val_data=[]', '+terminal_bench_config.trials_dir=/e/data1/datasets/playground/ot-baf/a3-rl-DCAgent_code-contests-noblock/a3-rl-DCAgent_code-contests-noblock/trace_jobs', '+terminal_bench_config.harbor.name=terminus-2', '+terminal_bench_config.harbor.max_episodes=999999', '+te
|
|||
|
|
Traceback (most recent call last):
|
|||
|
|
File "<frozen runpy>", line 198, in _run_module_as_main
|
|||
|
|
File "<frozen runpy>", line 88, in _run_code
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/examples/terminal_bench/entrypoints/main_tbench.py", line 133, in <module>
|
|||
|
|
main()
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/hydra/main.py", line 94, in decorated_main
|
|||
|
|
_run_hydra(
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/hydra/_internal/utils.py", line 394, in _run_hydra
|
|||
|
|
_run_app(
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/hydra/_internal/utils.py", line 457, in _run_app
|
|||
|
|
run_and_report(
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/hydra/_internal/utils.py", line 223, in run_and_report
|
|||
|
|
raise ex
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/hydra/_internal/utils.py", line 220, in run_and_report
|
|||
|
|
return func()
|
|||
|
|
^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/hydra/_internal/utils.py", line 458, in <lambda>
|
|||
|
|
lambda: hydra.run(
|
|||
|
|
^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/hydra/_internal/hydra.py", line 132, in run
|
|||
|
|
_ = ret.return_value
|
|||
|
|
^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/hydra/core/utils.py", line 260, in return_value
|
|||
|
|
raise self._return_value
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/hydra/core/utils.py", line 186, in run_job
|
|||
|
|
ret.return_value = task_function(task_cfg)
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/examples/terminal_bench/entrypoints/main_tbench.py", line 123, in main
|
|||
|
|
ray.get(skyrl_entrypoint.remote(cfg))
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/ray/_private/auto_init_hook.py", line 22, in auto_init_wrapper
|
|||
|
|
return fn(*args, **kwargs)
|
|||
|
|
^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/ray/_private/client_mode_hook.py", line 104, in wrapper
|
|||
|
|
return func(*args, **kwargs)
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/ray/_private/worker.py", line 2967, in get
|
|||
|
|
values, debugger_breakpoint = worker.get_objects(
|
|||
|
|
^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/ray/_private/worker.py", line 1015, in get_objects
|
|||
|
|
raise value.as_instanceof_cause()
|
|||
|
|
ray.exceptions.RayTaskError(ActorDiedError): [36mray::skyrl_entrypoint()[39m (pid=3659716, ip=10.128.17.210)
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/examples/terminal_bench/entrypoints/main_tbench.py", line 96, in skyrl_entrypoint
|
|||
|
|
exp.run()
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/entrypoints/main_base.py", line 403, in run
|
|||
|
|
asyncio.run(trainer.train())
|
|||
|
|
File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/asyncio/runners.py", line 195, in run
|
|||
|
|
return runner.run(main)
|
|||
|
|
^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/asyncio/runners.py", line 118, in run
|
|||
|
|
return self._loop.run_until_complete(task)
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "uvloop/loop.pyx", line 1518, in uvloop.loop.Loop.run_until_complete
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/fully_async_trainer.py", line 434, in train
|
|||
|
|
await self._train_loop()
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/fully_async_trainer.py", line 488, in _train_loop
|
|||
|
|
self.init_weight_sync_state()
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/trainer.py", line 808, in init_weight_sync_state
|
|||
|
|
ray.get(
|
|||
|
|
^^^^^^^^^^^^^^^^^^^
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
^^^^^^^^^^^^^^^^^^^
|
|||
|
|
ray.exceptions.RayTaskError(ActorDiedError): [36mray::FSDPPolicyWorkerBase.init_weight_sync_state()[39m (pid=2753436, ip=10.128.17.217, actor_id=7804ef248912326214f6164a02000000, repr=<skyrl_train.workers.fsdp.fsdp_worker.FSDPPolicyWorkerBase object at 0x400df41fe420>)
|
|||
|
|
File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/concurrent/futures/_base.py", line 449, in result
|
|||
|
|
return self.__get_result()
|
|||
|
|
^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/concurrent/futures/_base.py", line 401, in __get_result
|
|||
|
|
raise self._exception
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/workers/worker.py", line 299, in init_weight_sync_state
|
|||
|
|
results = await asyncio.gather(*tasks)
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/inference_engine_client.py", line 655, in init_weight_update_communicator
|
|||
|
|
rank_offset_count += engine.tp_size() * engine.pp_size()
|
|||
|
|
^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/ray_wrapped_inference_engine.py", line 26, in tp_size
|
|||
|
|
return ray.get(self.inference_engine_actor.tp_size.remote())
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
^^^^^^^^^^^^^^^^^^^
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
^^^^^^^^^^^^^^^^^^^
|
|||
|
|
ray.exceptions.ActorDiedError: The actor died because of an error raised in its creation task, [36mray::AsyncVLLMInferenceEngine.__init__()[39m (pid=3294997, ip=10.128.17.219, actor_id=3099a3967d816aa5fa00928502000000, repr=<skyrl_train.inference_engines.vllm.vllm_engine.AsyncVLLMInferenceEngine object at 0x400bfefd3080>)
|
|||
|
|
File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/concurrent/futures/_base.py", line 456, in result
|
|||
|
|
return self.__get_result()
|
|||
|
|
^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/concurrent/futures/_base.py", line 401, in __get_result
|
|||
|
|
raise self._exception
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/vllm/vllm_engine.py", line 1028, in __init__
|
|||
|
|
super().__init__(*args, **kwargs)
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/vllm/vllm_engine.py", line 500, in __init__
|
|||
|
|
self.llm = self._create_engine(*args, **kwargs)
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/SkyRL/skyrl-train/skyrl_train/inference_engines/vllm/vllm_engine.py", line 1105, in _create_engine
|
|||
|
|
engine = vllm.AsyncLLMEngine.from_engine_args(engine_args, stat_loggers=stat_loggers)
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/async_llm.py", line 251, in from_engine_args
|
|||
|
|
return cls(
|
|||
|
|
^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/async_llm.py", line 148, in __init__
|
|||
|
|
self.engine_core = EngineCoreClient.make_async_mp_client(
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core_client.py", line 124, in make_async_mp_client
|
|||
|
|
return AsyncMPClient(*client_args)
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core_client.py", line 835, in __init__
|
|||
|
|
super().__init__(
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/core_client.py", line 490, in __init__
|
|||
|
|
with launch_core_engines(vllm_config, executor_class, log_stats) as (
|
|||
|
|
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|
|||
|
|
File "/e/scratch/jureap59/feuer1/miniforge3/lib/python3.12/contextlib.py", line 144, in __exit__
|
|||
|
|
next(self.gen)
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/utils.py", line 936, in launch_core_engines
|
|||
|
|
wait_for_engine_startup(
|
|||
|
|
File "/e/scratch/jureap59/feuer1/OpenThoughts-Agent/envs/rl/lib/python3.12/site-packages/vllm/v1/engine/utils.py", line 995, in wait_for_engine_startup
|
|||
|
|
raise RuntimeError(
|
|||
|
|
RuntimeError: Engine core initialization failed. See root cause above. Failed core proc(s): {}
|
|||
|
|
Stopping Ray cluster...
|
|||
|
|
Ray cluster stopped
|
|||
|
|
[RLJobRunner] Launching trace upload (training exit code: 1):
|
|||
|
|
repo_id: DCAgent/a3-rl-DCAgent_code-contests-noblock
|
|||
|
|
job_dir: /e/data1/datasets/playground/ot-baf/a3-rl-DCAgent_code-contests-noblock/a3-rl-DCAgent_code-contests-noblock
|
|||
|
|
episodes: last
|
|||
|
|
log: /e/data1/datasets/playground/ot-baf/a3-rl-DCAgent_code-contests-noblock/logs/a3-rl-DCAgent_code-contests-noblock_trace_upload.log
|
|||
|
|
[RLJobRunner] Waiting for trace upload to complete...
|
|||
|
|
[RLJobRunner] Trace upload failed with exit code 1.
|
|||
|
|
Preserving Ray logs to /e/data1/datasets/playground/ot-baf/a3-rl-DCAgent_code-contests-noblock/ray_logs/
|
|||
|
|
Collecting Ray logs from worker jpbo-010-36...
|
|||
|
|
Collecting Ray logs from worker jpbo-010-38...
|
|||
|
|
Collecting Ray logs from worker jpbo-010-39...
|
|||
|
|
Collecting Ray logs from worker jpbo-010-40...
|
|||
|
|
Collecting Ray logs from worker jpbo-010-41...
|
|||
|
|
Collecting Ray logs from worker jpbo-010-43...
|
|||
|
|
Collecting Ray logs from worker jpbo-010-44...
|
|||
|
|
Collecting Ray logs from worker jpbo-010-45...
|
|||
|
|
Collecting Ray logs from worker jpbo-010-46...
|
|||
|
|
Collecting Ray logs from worker jpbo-022-02...
|
|||
|
|
Collecting Ray logs from worker jpbo-022-03...
|
|||
|
|
Collecting Ray logs from worker jpbo-022-04...
|
|||
|
|
Collecting Ray logs from worker jpbo-022-05...
|
|||
|
|
Ray log preservation complete
|