|
Error executing job with overrides: ['backend.model=NousResearch/Hermes-3-Llama-3.1-8B', 'backend.processor=NousResearch/Hermes-3-Llama-3.1-8B'] |
|
Traceback (most recent call last): |
|
File "/optimum-benchmark/optimum_benchmark/cli.py", line 65, in benchmark_cli |
|
benchmark_report: BenchmarkReport = launch(experiment_config=experiment_config) |
|
File "/optimum-benchmark/optimum_benchmark/experiment.py", line 102, in launch |
|
raise error |
|
File "/optimum-benchmark/optimum_benchmark/experiment.py", line 90, in launch |
|
report = launcher.launch(run, experiment_config.benchmark, experiment_config.backend) |
|
File "/optimum-benchmark/optimum_benchmark/launchers/process/launcher.py", line 47, in launch |
|
while not process_context.join(): |
|
File "/opt/conda/lib/python3.9/site-packages/torch/multiprocessing/spawn.py", line 189, in join |
|
raise ProcessRaisedException(msg, error_index, failed_process.pid) |
|
torch.multiprocessing.spawn.ProcessRaisedException: |
|
|
|
|
|
Traceback (most recent call last): |
|
File "/opt/conda/lib/python3.9/site-packages/torch/multiprocessing/spawn.py", line 76, in _wrap |
|
fn(i, *args) |
|
File "/optimum-benchmark/optimum_benchmark/launchers/process/launcher.py", line 63, in entrypoint |
|
worker_output = worker(*worker_args) |
|
File "/optimum-benchmark/optimum_benchmark/experiment.py", line 55, in run |
|
backend: Backend = backend_factory(backend_config) |
|
File "/optimum-benchmark/optimum_benchmark/backends/pytorch/backend.py", line 81, in __init__ |
|
self.load_model_with_no_weights() |
|
File "/optimum-benchmark/optimum_benchmark/backends/pytorch/backend.py", line 246, in load_model_with_no_weights |
|
self.load_model_from_pretrained() |
|
File "/optimum-benchmark/optimum_benchmark/backends/pytorch/backend.py", line 204, in load_model_from_pretrained |
|
self.pretrained_model = self.automodel_class.from_pretrained( |
|
File "/opt/conda/lib/python3.9/site-packages/transformers/models/auto/auto_factory.py", line 564, in from_pretrained |
|
return model_class.from_pretrained( |
|
File "/opt/conda/lib/python3.9/site-packages/transformers/modeling_utils.py", line 3738, in from_pretrained |
|
state_dict = load_state_dict(resolved_archive_file) |
|
File "/opt/conda/lib/python3.9/site-packages/transformers/modeling_utils.py", line 556, in load_state_dict |
|
return safe_load_file(checkpoint_file) |
|
File "/opt/conda/lib/python3.9/site-packages/safetensors/torch.py", line 315, in load_file |
|
result[k] = f.get_tensor(k) |
|
File "/opt/conda/lib/python3.9/site-packages/torch/utils/_device.py", line 79, in __torch_function__ |
|
return func(*args, **kwargs) |
|
RuntimeError: CUDA error: out of memory |
|
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. |
|
For debugging consider passing CUDA_LAUNCH_BLOCKING=1 |
|
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions. |
|
|
|
|
|
|
|
Set the environment variable HYDRA_FULL_ERROR=1 for a complete stack trace. |
|
|
