(sober) wty@ubuntu49:~/sober-reasoning$ python main.py \
> --model /data/wty/models/deepscaler_1.5b_grpo_replica_step500 \
> --task "custom|aime24|0|0" \
> --temperature 0.8 \
> --top_p 0.9 \
> --seed 0 \
> --output_dir /path/to/output \
> --max_new_tokens 4096 \
> --max_model_length 2048 \
> --custom_tasks_directory lighteval_tasks.py \
> --use_chat_template
INFO 09-08 15:04:59 __init__.py:190] Automatically detected platform cuda.
`torch_dtype` is deprecated! Use `dtype` instead!
Loading safetensors checkpoint shards: 0% Completed | 0/2 [00:00<?, ?it/s]
Loading safetensors checkpoint shards: 50% Completed | 1/2 [00:01<00:01, 1.45s/it]
Loading safetensors checkpoint shards: 100% Completed | 2/2 [00:02<00:00, 1.04s/it]
Loading safetensors checkpoint shards: 100% Completed | 2/2 [00:02<00:00, 1.10s/it]
Capturing CUDA graph shapes: 100%|██████████████████████████████████| 35/35 [00:16<00:00, 2.16it/s]
[rank0]: Traceback (most recent call last):
[rank0]: File "/home/wty/sober-reasoning/main.py", line 193, in <module>
[rank0]: main()
[rank0]: File "/home/wty/sober-reasoning/main.py", line 123, in main
[rank0]: pipeline = Pipeline(
[rank0]: File "/home/wty/miniconda3/envs/sober/lib/python3.10/site-packages/lighteval/pipeline.py", line 162, in __init__
[rank0]: self._init_tasks_and_requests(tasks=tasks)
[rank0]: File "/home/wty/miniconda3/envs/sober/lib/python3.10/site-packages/lighteval/pipeline.py", line 220, in _init_tasks_and_requests
[rank0]: task_names_list, fewshots_dict = taskinfo_selector(tasks, registry)
[rank0]: File "/home/wty/miniconda3/envs/sober/lib/python3.10/site-packages/lighteval/tasks/registry.py", line 299, in taskinfo_selector
[rank0]: expanded_tasks = task_registry.task_groups_dict.get(maybe_task_group, [maybe_task_group])
[rank0]: File "/home/wty/miniconda3/envs/sober/lib/python3.10/site-packages/lighteval/tasks/registry.py", line 190, in task_groups_dict
[rank0]: custom_tasks_module = create_custom_tasks_module(custom_tasks=self._custom_tasks)
[rank0]: File "/home/wty/miniconda3/envs/sober/lib/python3.10/site-packages/lighteval/tasks/registry.py", line 258, in create_custom_tasks_module
[rank0]: dataset_module = dataset_module_factory(str(custom_tasks), trust_remote_code=True)
[rank0]: File "/home/wty/miniconda3/envs/sober/lib/python3.10/site-packages/datasets/load.py", line 878, in dataset_module_factory
[rank0]: download_config = DownloadConfig(**download_kwargs)
[rank0]: TypeError: DownloadConfig.__init__() got an unexpected keyword argument 'trust_remote_code'
[rank0]:[W908 15:05:32.355204775 ProcessGroupNCCL.cpp:1250] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator())
I install the same packages in the dockerfile...
but it is not working
I install the same packages in the dockerfile...
but it is not working