diff --git "a/perf-df-unquantized-1xT4.csv" "b/perf-df-unquantized-1xT4.csv" --- "a/perf-df-unquantized-1xT4.csv" +++ "b/perf-df-unquantized-1xT4.csv" @@ -17275,7 +17275,7 @@ ChildProcessError: Traceback (most recent call last): torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 448.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 362.12 MiB is free. Process 505245 has 14.38 GiB memory in use. Of the allocated memory 14.27 GiB is allocated by PyTorch, and 1.78 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -float32-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B-Instruct,meta-llama/Meta-Llama-3-8B-Instruct,cuda,0,42,,,True,True,,float32,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +float32-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B-Instruct,meta-llama/Meta-Llama-3-8B-Instruct,cuda,0,42,,,True,True,,float32,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -17315,10 +17315,10 @@ ChildProcessError: Traceback (most recent call last): self.weight = Parameter(torch.empty((out_features, in_features), **factory_kwargs)) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py"", line 79, in __torch_function__ return func(*args, **kwargs) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 224.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 106.12 MiB is free. Process 37517 has 14.63 GiB memory in use. Of the allocated memory 14.52 GiB is allocated by PyTorch, and 1.52 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 224.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 106.12 MiB is free. Process 37531 has 14.63 GiB memory in use. Of the allocated memory 14.52 GiB is allocated by PyTorch, and 1.52 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -float16-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-3.1-405B,meta-llama/Llama-3.1-405B,cuda,0,42,,,True,True,,float16,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +float16-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-3.1-405B,meta-llama/Llama-3.1-405B,cuda,0,42,,,True,True,,float16,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -17358,7 +17358,7 @@ ChildProcessError: Traceback (most recent call last): self.weight = Parameter(torch.empty((out_features, in_features), **factory_kwargs)) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py"", line 79, in __torch_function__ return func(*args, **kwargs) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.62 GiB. GPU 0 has a total capacity of 14.74 GiB of which 470.12 MiB is free. Process 34200 has 14.28 GiB memory in use. Of the allocated memory 14.16 GiB is allocated by PyTorch, and 1.94 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.62 GiB. GPU 0 has a total capacity of 14.74 GiB of which 470.12 MiB is free. Process 34193 has 14.28 GiB memory in use. Of the allocated memory 14.16 GiB is allocated by PyTorch, and 1.94 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, bfloat16-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-2-70b-hf,meta-llama/Llama-2-70b-hf,cuda,0,42,,,True,True,,bfloat16,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.224-212.876.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.4.0,,4.44.2,,0.34.2,,,,1.22.0,,,,0.12.0,,"Traceback (most recent call last): @@ -17404,7 +17404,7 @@ ChildProcessError: Traceback (most recent call last): torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 448.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 234.12 MiB is free. Process 500263 has 14.51 GiB memory in use. Of the allocated memory 14.39 GiB is allocated by PyTorch, and 1.74 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -float32-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B,meta-llama/Meta-Llama-3-8B,cuda,0,42,,,True,True,,float32,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +float32-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B,meta-llama/Meta-Llama-3-8B,cuda,0,42,,,True,True,,float32,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -17444,7 +17444,7 @@ ChildProcessError: Traceback (most recent call last): self.weight = Parameter(torch.empty((out_features, in_features), **factory_kwargs)) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py"", line 79, in __torch_function__ return func(*args, **kwargs) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 224.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 106.12 MiB is free. Process 44721 has 14.63 GiB memory in use. Of the allocated memory 14.52 GiB is allocated by PyTorch, and 1.52 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 224.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 106.12 MiB is free. Process 44799 has 14.63 GiB memory in use. Of the allocated memory 14.52 GiB is allocated by PyTorch, and 1.52 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, float32-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-70B,meta-llama/Meta-Llama-3-70B,cuda,0,42,,,True,True,,float32,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.224-212.876.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.4.0,,4.44.2,,0.34.2,,,,1.22.0,,,,0.12.0,,"Traceback (most recent call last): @@ -17744,7 +17744,7 @@ ChildProcessError: Traceback (most recent call last): torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 50.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 18.12 MiB is free. Process 496799 has 14.72 GiB memory in use. Of the allocated memory 14.60 GiB is allocated by PyTorch, and 3.02 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -float16-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B,meta-llama/Meta-Llama-3-8B,cuda,0,42,,,True,True,,float16,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +float16-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B,meta-llama/Meta-Llama-3-8B,cuda,0,42,,,True,True,,float16,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -17776,7 +17776,7 @@ ChildProcessError: Traceback (most recent call last): self.weight = Parameter(torch.empty((out_features, in_features), **factory_kwargs)) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py"", line 79, in __torch_function__ return func(*args, **kwargs) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 43838 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 43935 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, float32-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-2-7b-hf,meta-llama/Llama-2-7b-hf,cuda,0,42,,,True,True,,float32,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.224-212.876.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.4.0,,4.44.2,,0.34.2,,,,1.22.0,,,,0.12.0,,"Traceback (most recent call last): @@ -17822,7 +17822,7 @@ ChildProcessError: Traceback (most recent call last): torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 172.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 150.12 MiB is free. Process 483086 has 14.59 GiB memory in use. Of the allocated memory 14.48 GiB is allocated by PyTorch, and 1.43 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -bfloat16-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-3.1-405B,meta-llama/Llama-3.1-405B,cuda,0,42,,,True,True,,bfloat16,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +bfloat16-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-3.1-405B,meta-llama/Llama-3.1-405B,cuda,0,42,,,True,True,,bfloat16,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -17862,7 +17862,7 @@ ChildProcessError: Traceback (most recent call last): self.weight = Parameter(torch.empty((out_features, in_features), **factory_kwargs)) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py"", line 79, in __torch_function__ return func(*args, **kwargs) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.62 GiB. GPU 0 has a total capacity of 14.74 GiB of which 470.12 MiB is free. Process 34685 has 14.28 GiB memory in use. Of the allocated memory 14.16 GiB is allocated by PyTorch, and 1.94 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.62 GiB. GPU 0 has a total capacity of 14.74 GiB of which 470.12 MiB is free. Process 34623 has 14.28 GiB memory in use. Of the allocated memory 14.16 GiB is allocated by PyTorch, and 1.94 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, float16-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-3.1-405B,meta-llama/Llama-3.1-405B,cuda,0,42,,,True,True,,float16,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): @@ -18092,7 +18092,7 @@ ChildProcessError: Traceback (most recent call last): torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 28188 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -bfloat16-flash_attention_2,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B-Instruct,meta-llama/Meta-Llama-3-8B-Instruct,cuda,0,42,,,True,True,,bfloat16,True,False,,flash_attention_2,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +bfloat16-flash_attention_2,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B-Instruct,meta-llama/Meta-Llama-3-8B-Instruct,cuda,0,42,,,True,True,,bfloat16,True,False,,flash_attention_2,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -18124,7 +18124,7 @@ ChildProcessError: Traceback (most recent call last): self.weight = Parameter(torch.empty((out_features, in_features), **factory_kwargs)) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py"", line 79, in __torch_function__ return func(*args, **kwargs) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 39435 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 39491 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, float16-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-2-13b-hf,meta-llama/Llama-2-13b-hf,cuda,0,42,,,True,True,,float16,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.224-212.876.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.4.0,,4.44.2,,0.34.2,,,,1.22.0,,,,0.12.0,,"Traceback (most recent call last): @@ -18170,7 +18170,7 @@ ChildProcessError: Traceback (most recent call last): torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 50.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 18.12 MiB is free. Process 495666 has 14.72 GiB memory in use. Of the allocated memory 14.60 GiB is allocated by PyTorch, and 3.02 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -float16-flash_attention_2,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B-Instruct,meta-llama/Meta-Llama-3-8B-Instruct,cuda,0,42,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +float16-flash_attention_2,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B-Instruct,meta-llama/Meta-Llama-3-8B-Instruct,cuda,0,42,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -18202,10 +18202,10 @@ ChildProcessError: Traceback (most recent call last): self.weight = Parameter(torch.empty((out_features, in_features), **factory_kwargs)) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py"", line 79, in __torch_function__ return func(*args, **kwargs) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 38960 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 38974 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -float16-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B-Instruct,meta-llama/Meta-Llama-3-8B-Instruct,cuda,0,42,,,True,True,,float16,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +float16-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B-Instruct,meta-llama/Meta-Llama-3-8B-Instruct,cuda,0,42,,,True,True,,float16,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -18237,7 +18237,7 @@ ChildProcessError: Traceback (most recent call last): self.weight = Parameter(torch.empty((out_features, in_features), **factory_kwargs)) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py"", line 79, in __torch_function__ return func(*args, **kwargs) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 37956 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 38078 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, bfloat16-flash_attention_2,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-2-70b-hf,meta-llama/Llama-2-70b-hf,cuda,0,42,,,True,True,,bfloat16,True,False,,flash_attention_2,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.224-212.876.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.4.0,,4.44.2,,0.34.2,,,,1.22.0,,,,0.12.0,,"Traceback (most recent call last): @@ -18448,7 +18448,7 @@ ChildProcessError: Traceback (most recent call last): torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 448.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 362.12 MiB is free. Process 507120 has 14.38 GiB memory in use. Of the allocated memory 14.27 GiB is allocated by PyTorch, and 1.78 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -float32-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B,meta-llama/Meta-Llama-3-8B,cuda,0,42,,,True,True,,float32,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +float32-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B,meta-llama/Meta-Llama-3-8B,cuda,0,42,,,True,True,,float32,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -18488,10 +18488,10 @@ ChildProcessError: Traceback (most recent call last): self.weight = Parameter(torch.empty((out_features, in_features), **factory_kwargs)) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py"", line 79, in __torch_function__ return func(*args, **kwargs) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 224.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 106.12 MiB is free. Process 43365 has 14.63 GiB memory in use. Of the allocated memory 14.52 GiB is allocated by PyTorch, and 1.52 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 224.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 106.12 MiB is free. Process 43445 has 14.63 GiB memory in use. Of the allocated memory 14.52 GiB is allocated by PyTorch, and 1.52 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -bfloat16-flash_attention_2,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B,meta-llama/Meta-Llama-3-8B,cuda,0,42,,,True,True,,bfloat16,True,False,,flash_attention_2,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +bfloat16-flash_attention_2,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B,meta-llama/Meta-Llama-3-8B,cuda,0,42,,,True,True,,bfloat16,True,False,,flash_attention_2,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -18523,7 +18523,7 @@ ChildProcessError: Traceback (most recent call last): self.weight = Parameter(torch.empty((out_features, in_features), **factory_kwargs)) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py"", line 79, in __torch_function__ return func(*args, **kwargs) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 46610 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 46716 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, float32-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-2-7b-hf,meta-llama/Llama-2-7b-hf,cuda,0,42,,,True,True,,float32,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.224-212.876.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.4.0,,4.44.2,,0.34.2,,,,1.22.0,,,,0.12.0,,"Traceback (most recent call last): @@ -18569,7 +18569,7 @@ ChildProcessError: Traceback (most recent call last): torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 172.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 150.12 MiB is free. Process 488767 has 14.59 GiB memory in use. Of the allocated memory 14.48 GiB is allocated by PyTorch, and 1.43 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -bfloat16-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B-Instruct,meta-llama/Meta-Llama-3-8B-Instruct,cuda,0,42,,,True,True,,bfloat16,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +bfloat16-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B-Instruct,meta-llama/Meta-Llama-3-8B-Instruct,cuda,0,42,,,True,True,,bfloat16,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -18601,7 +18601,7 @@ ChildProcessError: Traceback (most recent call last): self.weight = Parameter(torch.empty((out_features, in_features), **factory_kwargs)) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py"", line 79, in __torch_function__ return func(*args, **kwargs) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 37037 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 37072 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, float32-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-3.1-8B-Instruct,meta-llama/Llama-3.1-8B-Instruct,cuda,0,42,,,True,True,,float32,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): @@ -18725,7 +18725,7 @@ ChildProcessError: Traceback (most recent call last): torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 28611 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -float16-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B-Instruct,meta-llama/Meta-Llama-3-8B-Instruct,cuda,0,42,,,True,True,,float16,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +float16-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B-Instruct,meta-llama/Meta-Llama-3-8B-Instruct,cuda,0,42,,,True,True,,float16,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -18757,7 +18757,7 @@ ChildProcessError: Traceback (most recent call last): self.weight = Parameter(torch.empty((out_features, in_features), **factory_kwargs)) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py"", line 79, in __torch_function__ return func(*args, **kwargs) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 36520 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 36632 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, float32-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-2-70b-hf,meta-llama/Llama-2-70b-hf,cuda,0,42,,,True,True,,float32,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.224-212.876.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.4.0,,4.44.2,,0.34.2,,,,1.22.0,,,,0.12.0,,"Traceback (most recent call last): @@ -18803,7 +18803,7 @@ ChildProcessError: Traceback (most recent call last): torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 896.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 342.12 MiB is free. Process 499531 has 14.40 GiB memory in use. Of the allocated memory 14.29 GiB is allocated by PyTorch, and 1.75 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -bfloat16-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B,meta-llama/Meta-Llama-3-8B,cuda,0,42,,,True,True,,bfloat16,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +bfloat16-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B,meta-llama/Meta-Llama-3-8B,cuda,0,42,,,True,True,,bfloat16,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -18835,7 +18835,7 @@ ChildProcessError: Traceback (most recent call last): self.weight = Parameter(torch.empty((out_features, in_features), **factory_kwargs)) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py"", line 79, in __torch_function__ return func(*args, **kwargs) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 44272 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 44367 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, float16-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-3.1-8B-Instruct,meta-llama/Llama-3.1-8B-Instruct,cuda,0,42,,,True,True,,float16,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): @@ -19159,7 +19159,7 @@ ChildProcessError: Traceback (most recent call last): torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 224.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 106.12 MiB is free. Process 501406 has 14.63 GiB memory in use. Of the allocated memory 14.52 GiB is allocated by PyTorch, and 1.52 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -float32-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-3.1-405B,meta-llama/Llama-3.1-405B,cuda,0,42,,,True,True,,float32,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +float32-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-3.1-405B,meta-llama/Llama-3.1-405B,cuda,0,42,,,True,True,,float32,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -19199,7 +19199,7 @@ ChildProcessError: Traceback (most recent call last): self.weight = Parameter(torch.empty((out_features, in_features), **factory_kwargs)) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py"", line 79, in __torch_function__ return func(*args, **kwargs) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 3.25 GiB. GPU 0 has a total capacity of 14.74 GiB of which 1.42 GiB is free. Process 33754 has 13.32 GiB memory in use. Of the allocated memory 13.20 GiB is allocated by PyTorch, and 2.00 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 3.25 GiB. GPU 0 has a total capacity of 14.74 GiB of which 1.42 GiB is free. Process 33730 has 13.32 GiB memory in use. Of the allocated memory 13.20 GiB is allocated by PyTorch, and 2.00 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, bfloat16-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-3.1-8B-Instruct,meta-llama/Llama-3.1-8B-Instruct,cuda,0,42,,,True,True,,bfloat16,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): @@ -19300,7 +19300,7 @@ ChildProcessError: Traceback (most recent call last): RuntimeError: FlashAttention only supports Ampere GPUs or newer. ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -bfloat16-flash_attention_2,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-3.1-405B,meta-llama/Llama-3.1-405B,cuda,0,42,,,True,True,,bfloat16,True,False,,flash_attention_2,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +bfloat16-flash_attention_2,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-3.1-405B,meta-llama/Llama-3.1-405B,cuda,0,42,,,True,True,,bfloat16,True,False,,flash_attention_2,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -19340,7 +19340,7 @@ ChildProcessError: Traceback (most recent call last): self.weight = Parameter(torch.empty((out_features, in_features), **factory_kwargs)) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py"", line 79, in __torch_function__ return func(*args, **kwargs) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.62 GiB. GPU 0 has a total capacity of 14.74 GiB of which 470.12 MiB is free. Process 35628 has 14.28 GiB memory in use. Of the allocated memory 14.16 GiB is allocated by PyTorch, and 1.94 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.62 GiB. GPU 0 has a total capacity of 14.74 GiB of which 470.12 MiB is free. Process 35645 has 14.28 GiB memory in use. Of the allocated memory 14.16 GiB is allocated by PyTorch, and 1.94 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, float16-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-2-7b-hf,meta-llama/Llama-2-7b-hf,cuda,0,42,,,True,True,,float16,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.224-212.876.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.4.0,,4.44.2,,0.34.2,,,,1.22.0,,,,0.12.0,,,MB,869.613568,13880.918016,0.0,13478.395904,13476.849152,s,1,7.45773681640625,7.45773681640625,0.0,7.45773681640625,7.45773681640625,7.45773681640625,7.45773681640625,[7.45773681640625],,kWh,8.334706041720587e-06,9.121287597867775e-07,5.048892927989046e-06,1.429572772949641e-05,,MB,1312.82944,14117.896192,0.0,13702.791168,13671.637504,s,10,2.079610366821289,0.2079610366821289,0.0033514069268301503,0.20994270324707032,0.21055453338623048,0.21057272262573243,0.210587274017334,"[0.20050924682617188, 0.20672898864746095, 0.21030937194824217, 0.21059091186523438, 0.21044026184082032, 0.207285888671875, 0.20330979919433595, 0.209791748046875, 0.21009365844726563, 0.21055049133300782]",tokens/s,1230.9998261419482,kWh,6.043461185120014e-06,6.664650878497999e-07,4.000836534000269e-06,1.0710762806970085e-05,tokens/kWh,23901192.157238945,MB,1368.735744,14119.993344,0.0,13702.791168,13671.640064,s,10,37.210079101562506,3.72100791015625,0.004047133964294645,3.7205743408203125,3.7272676025390625,3.7274522827148435,3.7276000268554688,"[3.71548095703125, 3.715154541015625, 3.719099365234375, 3.71879541015625, 3.720977783203125, 3.722086181640625, 3.7201708984375, 3.723450439453125, 3.727636962890625, 3.7272265625]",tokens/s,16.930896553067136,kWh,0.00010894955950113031,1.2015654418491562e-05,7.245691907659804e-05,0.0001934221329962199,tokens/kWh,325712.46642818913,,s,630,37.206773433685264,0.059058370529659204,0.0004928099861878423,0.058999071121215815,0.0593966724395752,0.05950207996368408,0.062148142089843754,"[0.06201139068603516, 0.05951391983032227, 0.05865555191040039, 0.05841113662719727, 0.05868307113647461, 0.05844614410400391, 0.058434944152832034, 0.05840278244018555, 0.05833385467529297, 0.0584290542602539, 0.05842287826538086, 0.0586288948059082, 0.058635936737060544, 0.058440032958984374, 0.05882265472412109, 0.058762657165527345, 0.05875772857666016, 0.05916876983642578, 0.05898649597167969, 0.058810367584228515, 0.05887385559082031, 0.058963329315185546, 0.05893353652954102, 0.0587918701171875, 0.05887017440795898, 0.0587407341003418, 0.05866495895385742, 0.05875302505493164, 0.05872611236572266, 0.058902496337890624, 0.0589683837890625, 0.058916862487792966, 0.05922111892700195, 0.05929868698120117, 0.05928252792358398, 0.059190174102783204, 0.05905929565429688, 0.05910345458984375, 0.05897900772094727, 0.058949630737304685, 0.05885542297363281, 0.058787841796875, 0.058830654144287106, 0.058964160919189455, 0.05906390380859375, 0.058907039642333986, 0.05915139389038086, 0.059218910217285155, 0.05894553756713867, 0.059133953094482425, 0.05926502227783203, 0.05896310424804688, 0.05903849411010742, 0.05946131134033203, 0.059396480560302736, 0.05907046508789063, 0.05926630401611328, 0.05936787033081055, 0.058904544830322265, 0.05920783996582031, 0.05914822387695313, 0.05904364776611328, 0.059263137817382815, 0.06219590377807617, 0.05919168090820313, 0.058496353149414065, 0.058657440185546875, 0.05844582366943359, 0.05851955032348633, 0.05848678588867187, 0.058352767944335936, 0.058495872497558596, 0.05852950286865234, 0.05869596862792969, 0.058605567932128906, 0.05850931167602539, 0.05872553634643555, 0.05898662567138672, 0.05872918319702149, 0.05897334289550781, 0.05912364959716797, 0.058872737884521485, 0.05877129745483398, 0.058917247772216796, 0.05886937713623047, 0.058833057403564454, 0.058587135314941405, 0.05900697708129883, 0.058896350860595706, 0.05874076843261719, 0.058758720397949216, 0.0588836784362793, 0.058860385894775394, 0.058802398681640625, 0.058668830871582034, 0.05881651306152344, 0.059015167236328124, 0.05933587265014648, 0.059216705322265625, 0.059340065002441406, 0.0589442253112793, 0.058937313079833985, 0.05884902572631836, 0.0591608657836914, 0.05883599853515625, 0.05878396987915039, 0.05887871932983398, 0.05906227111816406, 0.05921526336669922, 0.058985057830810546, 0.05895782470703125, 0.058988094329833984, 0.05903200149536133, 0.058959873199462894, 0.059187198638916014, 0.05941862487792969, 0.05955097579956055, 0.05914064025878906, 0.059200897216796874, 0.05899350357055664, 0.05887913513183594, 0.059167167663574216, 0.059292064666748044, 0.059260032653808595, 0.05907868957519531, 0.05915750503540039, 0.06205939102172851, 0.059230209350585934, 0.058730239868164065, 0.058988479614257815, 0.05890899276733398, 0.058829856872558595, 0.0588175048828125, 0.05869916915893555, 0.05875568008422852, 0.058531841278076174, 0.05869977569580078, 0.058756961822509765, 0.05868150329589844, 0.05873676681518555, 0.05880409622192383, 0.05880012893676758, 0.05916672134399414, 0.05931008148193359, 0.05912985610961914, 0.059090240478515625, 0.05875577545166016, 0.05894707107543945, 0.05890099334716797, 0.05883084869384766, 0.058817855834960936, 0.058753726959228515, 0.058953216552734375, 0.05864908981323242, 0.05880217742919922, 0.05906185531616211, 0.059144607543945314, 0.05879097747802734, 0.05919171142578125, 0.05952675247192383, 0.05936371231079102, 0.05895248031616211, 0.05880809783935547, 0.058845184326171876, 0.05899059295654297, 0.05893948745727539, 0.05900278472900391, 0.058982208251953126, 0.05874668884277344, 0.058998817443847655, 0.05898899078369141, 0.058998687744140625, 0.05909040069580078, 0.058835201263427735, 0.05900252914428711, 0.05909158325195313, 0.05906425476074219, 0.059020416259765625, 0.05905059051513672, 0.059074111938476566, 0.05905417633056641, 0.059197311401367185, 0.05907251358032226, 0.05950067138671875, 0.059323070526123046, 0.0592503662109375, 0.05918105697631836, 0.0592402229309082, 0.05925040054321289, 0.06228521728515625, 0.0594601936340332, 0.05885337448120117, 0.05860966491699219, 0.05850521469116211, 0.0586506233215332, 0.0585011215209961, 0.058415103912353515, 0.05848886489868164, 0.05852771377563477, 0.058564609527587894, 0.05898163223266602, 0.05877017593383789, 0.05874665451049805, 0.05891299057006836, 0.05900697708129883, 0.059045215606689454, 0.05941254425048828, 0.05904806518554687, 0.05879353713989258, 0.0586187858581543, 0.05873049545288086, 0.05896540832519531, 0.058669662475585936, 0.058611583709716794, 0.05873062515258789, 0.058605216979980466, 0.05872447967529297, 0.0590090560913086, 0.058796222686767576, 0.05868057632446289, 0.058700542449951175, 0.05890867233276367, 0.05926700973510742, 0.05976684951782227, 0.05911481475830078, 0.05898223876953125, 0.058901439666748046, 0.05897724914550781, 0.05910208129882812, 0.05916227340698242, 0.05898076629638672, 0.058971710205078125, 0.059167167663574216, 0.05907830429077148, 0.05926128005981445, 0.05927446365356445, 0.05897004699707031, 0.059009246826171875, 0.059197696685791015, 0.05945996856689453, 0.05963123321533203, 0.05920191955566406, 0.059131553649902344, 0.05935660934448242, 0.05910825729370117, 0.059007198333740234, 0.059020320892333986, 0.059148895263671876, 0.059139743804931644, 0.05916873550415039, 0.059187744140625, 0.05936947250366211, 0.062236927032470704, 0.05952982330322266, 0.058843135833740234, 0.05857865524291992, 0.05879340744018555, 0.05873945617675781, 0.058584896087646485, 0.0587184944152832, 0.058521598815917966, 0.05852262496948242, 0.05879891204833984, 0.05885520172119141, 0.05865488052368164, 0.058703487396240234, 0.05868304061889648, 0.05880047988891601, 0.059070751190185546, 0.05928995132446289, 0.059485889434814455, 0.05902684783935547, 0.05900073623657227, 0.05918207931518555, 0.0590186882019043, 0.05877980804443359, 0.05889260864257812, 0.05901116943359375, 0.05885891342163086, 0.05860822296142578, 0.05877664184570312, 0.058888416290283206, 0.0588520622253418, 0.05886377716064453, 0.059115425109863284, 0.05928691101074219, 0.05933116912841797, 0.059231647491455076, 0.05926150512695313, 0.05927731323242187, 0.05886361694335938, 0.05884108734130859, 0.05886771011352539, 0.05906406402587891, 0.05876934432983399, 0.05926684951782227, 0.05883119964599609, 0.059003070831298826, 0.058935039520263674, 0.05892736053466797, 0.05907455825805664, 0.05969100952148437, 0.059170814514160154, 0.0592749137878418, 0.059439456939697266, 0.059238399505615234, 0.059295040130615234, 0.05934723281860352, 0.059302303314208986, 0.05926911926269531, 0.05937356948852539, 0.0591646728515625, 0.05897353744506836, 0.05906470489501953, 0.058980640411376954, 0.0624156494140625, 0.05975040054321289, 0.05893939208984375, 0.058631809234619144, 0.058867359161376955, 0.05897903823852539, 0.05881017684936524, 0.05875024032592773, 0.05860444641113281, 0.05850521469116211, 0.05875619125366211, 0.05878636932373047, 0.058713729858398435, 0.05870409774780273, 0.05870633697509765, 0.05858044815063477, 0.058953792572021484, 0.05955027389526367, 0.05925888061523438, 0.05912313461303711, 0.05898912048339844, 0.05882060623168945, 0.059135711669921875, 0.05875308990478516, 0.058724193572998046, 0.058775936126708984, 0.05873673629760742, 0.05874214553833008, 0.05884377670288086, 0.05895977783203125, 0.05880963134765625, 0.058890975952148435, 0.058945247650146484, 0.058902015686035154, 0.05926377487182617, 0.05943910217285156, 0.059377471923828126, 0.059170623779296876, 0.059189632415771486, 0.05894076919555664, 0.05888476943969727, 0.05919744110107422, 0.05887385559082031, 0.05881967926025391, 0.0588741455078125, 0.05893555068969727, 0.05921203231811523, 0.05906585693359375, 0.059072158813476563, 0.059022560119628906, 0.05953919982910156, 0.059633663177490234, 0.05937356948852539, 0.05928940963745117, 0.059233665466308594, 0.05910160064697265, 0.05901763153076172, 0.05912326431274414, 0.05931008148193359, 0.05939152145385742, 0.05934339141845703, 0.059257217407226566, 0.05938313674926758, 0.06210579299926758, 0.059409503936767576, 0.058743518829345705, 0.058474590301513675, 0.058744384765625, 0.05889059066772461, 0.05867520141601563, 0.05858243179321289, 0.058983009338378904, 0.05877532958984375, 0.05883855819702148, 0.05908755111694336, 0.05883084869384766, 0.05881814575195313, 0.058696033477783204, 0.05872828674316406, 0.05923452758789063, 0.0592281608581543, 0.059072544097900394, 0.05903971099853516, 0.05878300857543945, 0.058724990844726564, 0.058652542114257813, 0.05871023941040039, 0.05865206527709961, 0.058588798522949216, 0.05858371353149414, 0.05860383987426758, 0.058875904083251954, 0.058828800201416016, 0.05875263977050781, 0.05898867034912109, 0.05939839935302734, 0.059232257843017576, 0.059115520477294924, 0.05926707077026367, 0.05923430252075195, 0.05888608169555664, 0.05866819381713867, 0.058613727569580075, 0.05870188903808594, 0.05869657516479492, 0.05886566543579102, 0.05926502227783203, 0.05910732650756836, 0.059174400329589844, 0.05918566513061523, 0.05901311874389648, 0.05938943862915039, 0.059165184020996096, 0.05907212829589844, 0.05945897674560547, 0.05950265502929687, 0.0593540153503418, 0.05922934341430664, 0.05942086410522461, 0.059310752868652346, 0.059383296966552736, 0.05912755203247071, 0.059155200958251955, 0.05923011016845703, 0.059504737854003904, 0.05938127899169922, 0.06216543960571289, 0.05937097549438477, 0.05886025619506836, 0.05858220672607422, 0.05849507141113281, 0.05861872100830078, 0.0587935676574707, 0.05887340927124023, 0.058718944549560545, 0.05873443222045899, 0.058579071044921875, 0.05873452758789063, 0.058929248809814455, 0.05888409423828125, 0.058719776153564454, 0.05887638473510742, 0.05931622314453125, 0.05940185546875, 0.05919168090820313, 0.05904508972167969, 0.05886032104492187, 0.0586321907043457, 0.058617183685302734, 0.05860559844970703, 0.058826560974121096, 0.05891769790649414, 0.05893529510498047, 0.05892300796508789, 0.05883699035644531, 0.05907251358032226, 0.059131423950195314, 0.058999263763427734, 0.059252544403076174, 0.05908294296264648, 0.05915817642211914, 0.05915273666381836, 0.059115425109863284, 0.05914384078979492, 0.05917292785644531, 0.05913433456420898, 0.0591278076171875, 0.05903257751464844, 0.0589051513671875, 0.05894803237915039, 0.05917491149902344, 0.059278560638427735, 0.05938652801513672, 0.05907263946533203, 0.059115425109863284, 0.0593265609741211, 0.059487998962402346, 0.05935436630249023, 0.059456031799316404, 0.05925724792480469, 0.05942806243896484, 0.05945635223388672, 0.059461631774902345, 0.05927526473999024, 0.05919120025634766, 0.059185214996337894, 0.059129886627197266, 0.05933260726928711, 0.05929369735717773, 0.06252908706665039, 0.059644001007080075, 0.05909539031982422, 0.058613761901855466, 0.05885318374633789, 0.058872001647949215, 0.05898175811767578, 0.05884787368774414, 0.058613761901855466, 0.05852364730834961, 0.05858303833007812, 0.05898198318481445, 0.058748897552490235, 0.05895212936401367, 0.05892432022094726, 0.05902329635620117, 0.05948089599609375, 0.06002483367919922, 0.05961520004272461, 0.059178943634033206, 0.059014720916748045, 0.05897651290893555, 0.058974369049072266, 0.05865891265869141, 0.058931137084960936, 0.058926334381103514, 0.05887059020996094, 0.05892822265625, 0.058747806549072266, 0.05918310546875, 0.05915852737426758, 0.05909670257568359, 0.05927974319458008, 0.059338752746582034, 0.0594384651184082, 0.059259521484375, 0.059172863006591796, 0.059061279296875, 0.05912227249145508, 0.058808704376220704, 0.05930188751220703, 0.0592037124633789, 0.058946910858154296, 0.058998878479003904, 0.059025375366210935, 0.060004608154296875, 0.05936883163452149, 0.0591512336730957, 0.05899283218383789, 0.05928323364257813, 0.059444416046142576, 0.05950137710571289, 0.05934694290161133, 0.059359294891357425, 0.059310016632080076, 0.059149471282958985, 0.05906435012817383, 0.05907334518432617, 0.059141536712646485, 0.05910793685913086, 0.05906227111816406, 0.059015167236328124, 0.05935308837890625, 0.0624161605834961, 0.05965692901611328, 0.059009025573730466, 0.05861785507202148, 0.058742782592773435, 0.058836734771728516, 0.05877376174926758, 0.05875408172607422, 0.058599872589111326, 0.05851567840576172, 0.05852959823608399, 0.05865052795410156, 0.058628353118896484, 0.05900703811645508, 0.058853023529052736, 0.05914620971679688, 0.059589374542236326, 0.059764640808105465, 0.059367198944091794, 0.05902880096435547, 0.05877443313598633, 0.058654720306396485, 0.05862582397460937, 0.058914272308349606, 0.058727169036865236, 0.05875820922851562, 0.05925574493408203, 0.058992641448974606, 0.05885091018676758, 0.058865345001220704, 0.05882953643798828, 0.05918515014648437, 0.059241600036621093, 0.05929663848876953, 0.05951871871948242, 0.05936563110351562, 0.059305984497070315, 0.059101184844970706, 0.059154430389404294, 0.05916656112670898, 0.059127967834472654, 0.058893760681152346, 0.05915091323852539, 0.05913167953491211, 0.05899507141113281, 0.05931017684936524, 0.059176448822021485, 0.05915059280395508, 0.059260513305664064, 0.05935871887207031, 0.059377696990966795, 0.05945564651489258, 0.05940092849731445, 0.0594411506652832, 0.05957632064819336, 0.05946681594848633, 0.059466686248779294, 0.05942214584350586, 0.059326656341552736, 0.0591671028137207, 0.05940364837646484, 0.059327102661132815, 0.05940633773803711]",tokens/s,16.932400793174583,, @@ -19508,7 +19508,7 @@ ChildProcessError: Traceback (most recent call last): torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 50.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 18.12 MiB is free. Process 497164 has 14.72 GiB memory in use. Of the allocated memory 14.60 GiB is allocated by PyTorch, and 3.02 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -bfloat16-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B,meta-llama/Meta-Llama-3-8B,cuda,0,42,,,True,True,,bfloat16,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +bfloat16-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B,meta-llama/Meta-Llama-3-8B,cuda,0,42,,,True,True,,bfloat16,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -19540,10 +19540,10 @@ ChildProcessError: Traceback (most recent call last): self.weight = Parameter(torch.empty((out_features, in_features), **factory_kwargs)) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py"", line 79, in __torch_function__ return func(*args, **kwargs) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 45714 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 45832 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -float16-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B,meta-llama/Meta-Llama-3-8B,cuda,0,42,,,True,True,,float16,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +float16-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B,meta-llama/Meta-Llama-3-8B,cuda,0,42,,,True,True,,float16,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -19575,10 +19575,10 @@ ChildProcessError: Traceback (most recent call last): self.weight = Parameter(torch.empty((out_features, in_features), **factory_kwargs)) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py"", line 79, in __torch_function__ return func(*args, **kwargs) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 45259 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 45340 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -bfloat16-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B-Instruct,meta-llama/Meta-Llama-3-8B-Instruct,cuda,0,42,,,True,True,,bfloat16,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +bfloat16-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B-Instruct,meta-llama/Meta-Llama-3-8B-Instruct,cuda,0,42,,,True,True,,bfloat16,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -19610,7 +19610,7 @@ ChildProcessError: Traceback (most recent call last): self.weight = Parameter(torch.empty((out_features, in_features), **factory_kwargs)) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py"", line 79, in __torch_function__ return func(*args, **kwargs) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 38487 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 38510 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, float16-flash_attention_2,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B,meta-llama/Meta-Llama-3-8B,cuda,0,42,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.224-212.876.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.4.0,,4.44.2,,0.34.2,,,,1.22.0,,,,0.12.0,,"Traceback (most recent call last): @@ -19692,7 +19692,7 @@ torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 448.00 MiB. GPU 0 ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, bfloat16-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-2-7b-hf,meta-llama/Llama-2-7b-hf,cuda,0,42,,,True,True,,bfloat16,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.224-212.876.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.4.0,,4.44.2,,0.34.2,,,,1.22.0,,,,0.12.0,,,MB,869.531648,13880.918016,0.0,13478.395904,13476.849152,s,1,7.557330078125,7.557330078125,0.0,7.557330078125,7.557330078125,7.557330078125,7.557330078125,[7.557330078125],,kWh,8.614358579203931e-06,9.423954797841812e-07,4.910003927993678e-06,1.4466757986981792e-05,,MB,1208.127488,14113.701888,0.0,13700.694016,13671.637504,s,10,12.46454931640625,1.246454931640625,0.004118358371665385,1.2484244995117186,1.2501810546875,1.2503031127929687,1.2504007592773438,"[1.23776806640625, 1.24261962890625, 1.243001708984375, 1.2441705322265626, 1.249641845703125, 1.2497257080078126, 1.2501539306640626, 1.2472071533203124, 1.2504251708984375, 1.2498355712890625]",tokens/s,205.38247593360185,kWh,3.637035038000704e-05,4.011153165015655e-06,2.417424156159753e-05,6.455574510662022e-05,tokens/kWh,3965564.948203922,MB,1256.267776,14115.79904,0.0,13702.791168,13671.640064,s,10,37.74185693359375,3.7741856933593754,0.0019019189260765204,3.774030517578125,3.776664501953125,3.7768697265625,3.77703390625,"[3.771296142578125, 3.7749912109375, 3.77295654296875, 3.771998046875, 3.776618896484375, 3.7728369140625, 3.7732041015625, 3.777074951171875, 3.77485693359375, 3.776023193359375]",tokens/s,16.692342433189648,kWh,0.00011034360833999474,1.2169942675116277e-05,7.313094739360427e-05,0.0001956444984087153,tokens/kWh,322012.63267004065,,s,630,37.73844884872437,0.059902299759879936,0.00026899646318153086,0.05989480018615723,0.060194672393798825,0.06027122974395752,0.061000479354858396,"[0.06125177764892578, 0.05970035171508789, 0.05954444885253906, 0.05974380874633789, 0.059281856536865234, 0.05950003051757812, 0.05950918579101563, 0.05949161529541016, 0.05956025695800781, 0.05976316833496094, 0.059584510803222655, 0.05952511978149414, 0.05941862487792969, 0.05956787109375, 0.05941683197021484, 0.05949235153198242, 0.05967052841186524, 0.05999958419799805, 0.05982278442382812, 0.06006771087646484, 0.05966428756713867, 0.05973011016845703, 0.05953945541381836, 0.059594753265380856, 0.05972927856445313, 0.0598331184387207, 0.05987094497680664, 0.05989593505859375, 0.059860225677490234, 0.05995187377929687, 0.05967462539672851, 0.059614593505859376, 0.059910945892333985, 0.06017552185058594, 0.060041919708251956, 0.05997561645507812, 0.06011283111572266, 0.05994303894042969, 0.05992038345336914, 0.059708480834960935, 0.059975776672363285, 0.05993967819213867, 0.05991190338134766, 0.059991649627685543, 0.06001119995117187, 0.05997910308837891, 0.059845279693603516, 0.05984640121459961, 0.05985843276977539, 0.06001945495605469, 0.05992201614379883, 0.06026691055297852, 0.06002390289306641, 0.060165023803710936, 0.05988761520385742, 0.05995075225830078, 0.05987977600097656, 0.06009609603881836, 0.06005801773071289, 0.06006284713745117, 0.060176673889160155, 0.06012480163574219, 0.06030230331420899, 0.06094432067871094, 0.05971267318725586, 0.0595560302734375, 0.059353118896484376, 0.05934972763061523, 0.05956742477416992, 0.05950534439086914, 0.059658241271972653, 0.059717632293701174, 0.059743263244628905, 0.06030799865722656, 0.059700672149658206, 0.05967769622802734, 0.059957248687744144, 0.05975244903564453, 0.05996723175048828, 0.0599653434753418, 0.05990435028076172, 0.05979340744018555, 0.05980934524536133, 0.05963161468505859, 0.059863071441650394, 0.05973443222045898, 0.05975238418579101, 0.05966364669799805, 0.05987526321411133, 0.059655006408691404, 0.059716670989990235, 0.05970016098022461, 0.05973196792602539, 0.0599609603881836, 0.05992486572265625, 0.05975151824951172, 0.0600335693359375, 0.06001446533203125, 0.0598922233581543, 0.059850753784179686, 0.059791358947753906, 0.0598364143371582, 0.059837825775146486, 0.05986163330078125, 0.059924480438232425, 0.06005350494384765, 0.0599552001953125, 0.05985260772705078, 0.06002297592163086, 0.05999987030029297, 0.06008838272094726, 0.06004313659667969, 0.060146110534667965, 0.06022553634643555, 0.06016524887084961, 0.060029407501220704, 0.06021980667114258, 0.06012659072875977, 0.060211841583251956, 0.060186622619628906, 0.06033407974243164, 0.06014563369750976, 0.06032559967041016, 0.06017465591430664, 0.06021247863769531, 0.060166912078857424, 0.061077503204345705, 0.05968201446533203, 0.05937561416625976, 0.0595577278137207, 0.059486335754394534, 0.059509567260742184, 0.059581790924072266, 0.05955219268798828, 0.059701248168945314, 0.059693279266357424, 0.05965119934082031, 0.05983020782470703, 0.059744609832763675, 0.059748703002929685, 0.05956793594360352, 0.059953086853027346, 0.060014720916748046, 0.06002899169921875, 0.059883296966552736, 0.060074527740478514, 0.05994496154785156, 0.060088096618652345, 0.05975267028808594, 0.05962457656860352, 0.0596181755065918, 0.059625377655029295, 0.05947401428222656, 0.059674175262451175, 0.05981827163696289, 0.0597751350402832, 0.059804672241210936, 0.05981491088867188, 0.05979340744018555, 0.0598466567993164, 0.059756542205810545, 0.0599736328125, 0.05989376068115235, 0.0599285774230957, 0.0598548469543457, 0.059856895446777345, 0.060088321685791014, 0.05989990234375, 0.05982796859741211, 0.059955455780029296, 0.059874942779541016, 0.05992035293579102, 0.06017251205444336, 0.05995743942260742, 0.0599898567199707, 0.060039295196533206, 0.06018255996704101, 0.060055553436279295, 0.05998591995239258, 0.06010022354125977, 0.06035289764404297, 0.06013747024536133, 0.05991548919677735, 0.06004816055297851, 0.059915904998779294, 0.06006739044189453, 0.060076416015625, 0.06012723159790039, 0.060254337310791016, 0.061212223052978514, 0.05975286483764648, 0.05957017517089844, 0.059667999267578126, 0.059315967559814456, 0.059480510711669925, 0.05947558212280273, 0.05937641525268555, 0.059463550567626956, 0.05950054550170898, 0.05954969787597656, 0.059754112243652346, 0.05953542327880859, 0.059728191375732424, 0.059666431427001954, 0.05970943832397461, 0.05991628646850586, 0.05999545669555664, 0.05991084671020508, 0.059686912536621096, 0.05972787094116211, 0.05981718444824219, 0.05969753646850586, 0.059689342498779295, 0.0596992301940918, 0.05978726577758789, 0.05977907180786133, 0.05990399932861328, 0.059958656311035155, 0.05992310333251953, 0.0600080337524414, 0.05987776184082031, 0.060055553436279295, 0.06014361572265625, 0.05989888000488281, 0.05992268753051758, 0.059877567291259766, 0.059944801330566407, 0.05979935836791992, 0.05983119964599609, 0.05979878234863281, 0.05999814224243164, 0.059805694580078124, 0.060018848419189454, 0.059808414459228514, 0.05998387145996094, 0.060039169311523435, 0.06001830291748047, 0.060088191986083984, 0.06010227203369141, 0.06000278472900391, 0.06011916732788086, 0.06009420776367187, 0.059992607116699216, 0.05988351821899414, 0.06005072021484375, 0.060133438110351566, 0.06005417633056641, 0.06005145645141602, 0.05997545623779297, 0.059982048034667966, 0.06003046417236328, 0.060004863739013675, 0.061222942352294925, 0.05980096054077148, 0.05973040008544922, 0.05970150375366211, 0.05944950485229492, 0.05963148880004883, 0.059624862670898435, 0.059625728607177735, 0.05967305755615234, 0.05970694351196289, 0.05977132797241211, 0.059759777069091795, 0.05959766387939453, 0.05988556671142578, 0.05981798553466797, 0.059850753784179686, 0.05972339248657227, 0.05985033416748047, 0.05997443389892578, 0.05982822418212891, 0.05968281555175781, 0.05974016189575195, 0.05974220657348633, 0.059641151428222655, 0.059751102447509766, 0.05976678466796875, 0.05979123306274414, 0.059895839691162106, 0.059711265563964847, 0.05992879867553711, 0.05991846466064453, 0.0598691520690918, 0.060022785186767576, 0.05997772979736328, 0.05996543884277344, 0.06010268783569336, 0.06007567977905273, 0.06001628875732422, 0.059869857788085935, 0.05994063949584961, 0.060023006439208985, 0.06023347091674805, 0.05989775848388672, 0.06006630325317383, 0.059909984588623046, 0.0599736328125, 0.059936576843261716, 0.05996559906005859, 0.06000028610229492, 0.060290657043457034, 0.0601800651550293, 0.0602077751159668, 0.060110401153564454, 0.060162654876708986, 0.05997568130493164, 0.06009648132324219, 0.060135456085205076, 0.0601448974609375, 0.06027135848999023, 0.06029510498046875, 0.060225601196289065, 0.06041702270507812, 0.060170848846435546, 0.06123961639404297, 0.059684478759765625, 0.05955027389526367, 0.05957017517089844, 0.05933987045288086, 0.059630496978759766, 0.05959884643554687, 0.059662303924560546, 0.059565536499023436, 0.05991670227050781, 0.05974441528320312, 0.05984220886230469, 0.05963398361206055, 0.05958863830566406, 0.05980995178222656, 0.0598092155456543, 0.059901439666748046, 0.059974559783935545, 0.059975135803222654, 0.059684864044189455, 0.05967472076416016, 0.05977132797241211, 0.059701248168945314, 0.05976201629638672, 0.059834590911865236, 0.05979385757446289, 0.05977062225341797, 0.059805248260498045, 0.05986579132080078, 0.05994291305541992, 0.059850753784179686, 0.0598256950378418, 0.0598037109375, 0.05990851211547851, 0.05979545593261719, 0.05990742492675781, 0.05981865692138672, 0.05981977462768555, 0.05974143981933594, 0.059664543151855466, 0.059620128631591794, 0.059934783935546875, 0.059736000061035154, 0.05999417495727539, 0.05985007858276367, 0.05999683380126953, 0.059972766876220704, 0.05996015930175781, 0.060096511840820314, 0.060047359466552735, 0.06003507232666016, 0.06010060882568359, 0.06003507232666016, 0.06021324920654297, 0.060063873291015625, 0.060118721008300784, 0.0599984016418457, 0.060080127716064455, 0.05985033416748047, 0.06021104049682617, 0.06026911926269531, 0.06032307052612305, 0.060213600158691406, 0.06121267318725586, 0.059887424468994144, 0.05950191879272461, 0.05953212738037109, 0.05959884643554687, 0.0597341423034668, 0.05947296142578125, 0.05943174362182617, 0.05961523056030273, 0.05947187042236328, 0.05950873565673828, 0.05962108612060547, 0.05960528182983398, 0.05968896102905273, 0.05973606491088867, 0.05967871856689453, 0.05973097610473633, 0.05983542251586914, 0.05973545455932617, 0.05970793533325195, 0.05958860778808594, 0.05981798553466797, 0.05961081695556641, 0.059650047302246094, 0.05981727981567383, 0.05977142333984375, 0.05973859024047851, 0.059866718292236325, 0.05991465759277344, 0.059778785705566405, 0.059881214141845704, 0.05982262420654297, 0.05995062255859375, 0.060205375671386716, 0.06000860977172852, 0.06001859283447265, 0.059923744201660155, 0.05994707107543945, 0.059763233184814454, 0.05977110290527344, 0.05967593765258789, 0.05991644668579101, 0.05985747146606445, 0.059958305358886715, 0.06003401565551758, 0.060047359466552735, 0.059982975006103514, 0.059932769775390624, 0.05998368072509765, 0.06018352127075195, 0.060063743591308595, 0.060129280090332034, 0.06022310256958008, 0.06025459289550781, 0.06017228698730469, 0.0601212158203125, 0.060032863616943356, 0.06013750457763672, 0.0600552978515625, 0.0602872314453125, 0.060270591735839846, 0.06025139236450195, 0.06018124771118164, 0.06101103973388672, 0.0598127670288086, 0.05958041763305664, 0.05953542327880859, 0.059565185546875, 0.0596712646484375, 0.05958591842651367, 0.059509471893310545, 0.05965523147583008, 0.05954860687255859, 0.05976678466796875, 0.05985187149047851, 0.05966694259643555, 0.05976105499267578, 0.05979340744018555, 0.05989785766601562, 0.06011471939086914, 0.06017184066772461, 0.05991696166992187, 0.05993580627441406, 0.05987833786010742, 0.059856895446777345, 0.05960003280639648, 0.05965651321411133, 0.05957891082763672, 0.059630592346191405, 0.059696128845214844, 0.05984403228759766, 0.05988761520385742, 0.05981651306152344, 0.05975449752807617, 0.05975449752807617, 0.059824127197265625, 0.05994220733642578, 0.05984128189086914, 0.05984364700317383, 0.06004576110839844, 0.06000889587402344, 0.05990534210205078, 0.059778881072998044, 0.05996428680419922, 0.06020025634765625, 0.059996864318847654, 0.05994259262084961, 0.05987974548339844, 0.06022300720214844, 0.060271072387695315, 0.060128929138183594, 0.06020249557495117, 0.06043843078613281, 0.06035270309448242, 0.06032665634155274, 0.060418048858642576, 0.060232929229736325, 0.06022633743286133, 0.0601492805480957, 0.06020479965209961, 0.06022422409057617, 0.060052894592285154, 0.06031216049194336, 0.06019071960449219, 0.060194816589355465, 0.06012723159790039, 0.060903072357177734, 0.05951523208618164, 0.05951811218261719, 0.059427680969238283, 0.05955331039428711, 0.05973993682861328, 0.05969136047363281, 0.059511135101318356, 0.05950883102416992, 0.059620479583740234, 0.0595849609375, 0.05963126373291015, 0.05977977752685547, 0.05972278213500976, 0.05978972625732422, 0.05990047836303711, 0.059957248687744144, 0.060020736694335934, 0.05986918258666992, 0.059856895446777345, 0.059658241271972653, 0.05993673706054688, 0.059819358825683594, 0.05967327880859375, 0.05969715118408203, 0.06008544158935547, 0.05991929626464844, 0.05987724685668945, 0.05978239822387695, 0.05976947021484375, 0.05986304092407226, 0.05991427230834961, 0.05986249542236328, 0.06009920120239258, 0.059957248687744144, 0.06000435256958008, 0.05974630355834961, 0.05988531112670899, 0.059733535766601564, 0.05986073684692383, 0.059945568084716794, 0.06008256149291992, 0.05990399932861328, 0.05994684982299805, 0.05989174270629883, 0.05993280029296875, 0.05992403030395508, 0.06014924621582031, 0.059988929748535154, 0.060129280090332034, 0.06014905548095703, 0.06028700637817383, 0.06021558380126953, 0.06020544052124024, 0.06014892959594727, 0.060106945037841794, 0.060134014129638674, 0.06013337707519531, 0.060284927368164064, 0.060343841552734374, 0.06004719924926758, 0.06012137603759766, 0.06017884826660156, 0.06097462463378906, 0.059591201782226565, 0.05959836959838867, 0.05954403305053711, 0.05955583953857422, 0.05953308868408203, 0.05951641464233398, 0.059593215942382816, 0.05956220626831055, 0.05972377777099609, 0.05953036880493164, 0.059582401275634765, 0.05962438583374023, 0.059625473022460934, 0.0596234245300293, 0.05998793411254883, 0.05993024063110351, 0.06008668899536133, 0.06005961608886719, 0.05995119857788086, 0.05977225494384766, 0.059824737548828125, 0.059858943939208986, 0.05971558380126953, 0.05970259094238281, 0.05990879821777344, 0.0597523193359375, 0.059848831176757815, 0.05987923049926758, 0.059803840637207034, 0.05977814483642578, 0.05994076919555664, 0.06004019165039062, 0.05993471908569336, 0.06004121780395508, 0.059985023498535156, 0.059865345001220704, 0.06017078399658203, 0.05995267105102539, 0.05982787322998047, 0.05997865676879883, 0.060210624694824216, 0.059928607940673825, 0.06016668701171875, 0.059877086639404296, 0.06003273773193359, 0.059908672332763674, 0.060104705810546874, 0.06019465637207031, 0.06032515335083008, 0.060133312225341795, 0.060122047424316406, 0.0600874252319336, 0.06030144119262695, 0.06003993606567383, 0.06009241485595703, 0.06012473678588867, 0.06015430450439453, 0.06016614532470703, 0.06013747024536133, 0.06026649475097656, 0.06043017578125, 0.060055713653564456]",tokens/s,16.69384988570603,, -float16-flash_attention_2,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B,meta-llama/Meta-Llama-3-8B,cuda,0,42,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +float16-flash_attention_2,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B,meta-llama/Meta-Llama-3-8B,cuda,0,42,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -19724,10 +19724,10 @@ ChildProcessError: Traceback (most recent call last): self.weight = Parameter(torch.empty((out_features, in_features), **factory_kwargs)) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py"", line 79, in __torch_function__ return func(*args, **kwargs) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 46157 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1002.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 532.12 MiB is free. Process 46257 has 14.22 GiB memory in use. Of the allocated memory 13.98 GiB is allocated by PyTorch, and 129.48 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -float32-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B-Instruct,meta-llama/Meta-Llama-3-8B-Instruct,cuda,0,42,,,True,True,,float32,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +float32-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B-Instruct,meta-llama/Meta-Llama-3-8B-Instruct,cuda,0,42,,,True,True,,float32,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -19767,10 +19767,10 @@ ChildProcessError: Traceback (most recent call last): self.weight = Parameter(torch.empty((out_features, in_features), **factory_kwargs)) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py"", line 79, in __torch_function__ return func(*args, **kwargs) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 224.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 106.12 MiB is free. Process 36081 has 14.63 GiB memory in use. Of the allocated memory 14.52 GiB is allocated by PyTorch, and 1.52 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 224.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 106.12 MiB is free. Process 36108 has 14.63 GiB memory in use. Of the allocated memory 14.52 GiB is allocated by PyTorch, and 1.52 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -bfloat16-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-3.1-405B,meta-llama/Llama-3.1-405B,cuda,0,42,,,True,True,,bfloat16,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +bfloat16-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-3.1-405B,meta-llama/Llama-3.1-405B,cuda,0,42,,,True,True,,bfloat16,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -19810,7 +19810,7 @@ ChildProcessError: Traceback (most recent call last): self.weight = Parameter(torch.empty((out_features, in_features), **factory_kwargs)) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py"", line 79, in __torch_function__ return func(*args, **kwargs) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.62 GiB. GPU 0 has a total capacity of 14.74 GiB of which 470.12 MiB is free. Process 33261 has 14.28 GiB memory in use. Of the allocated memory 14.16 GiB is allocated by PyTorch, and 1.94 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.62 GiB. GPU 0 has a total capacity of 14.74 GiB of which 470.12 MiB is free. Process 33277 has 14.28 GiB memory in use. Of the allocated memory 14.16 GiB is allocated by PyTorch, and 1.94 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, float16-flash_attention_2,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-2-70b-hf,meta-llama/Llama-2-70b-hf,cuda,0,42,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.224-212.876.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.4.0,,4.44.2,,0.34.2,,,,1.22.0,,,,0.12.0,,"Traceback (most recent call last): @@ -19856,7 +19856,7 @@ ChildProcessError: Traceback (most recent call last): torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 448.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 234.12 MiB is free. Process 500625 has 14.51 GiB memory in use. Of the allocated memory 14.39 GiB is allocated by PyTorch, and 1.74 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -float16-flash_attention_2,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-3.1-405B,meta-llama/Llama-3.1-405B,cuda,0,42,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +float16-flash_attention_2,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Llama-3.1-405B,meta-llama/Llama-3.1-405B,cuda,0,42,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -19896,7 +19896,7 @@ ChildProcessError: Traceback (most recent call last): self.weight = Parameter(torch.empty((out_features, in_features), **factory_kwargs)) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py"", line 79, in __torch_function__ return func(*args, **kwargs) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.62 GiB. GPU 0 has a total capacity of 14.74 GiB of which 470.12 MiB is free. Process 35166 has 14.28 GiB memory in use. Of the allocated memory 14.16 GiB is allocated by PyTorch, and 1.94 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.62 GiB. GPU 0 has a total capacity of 14.74 GiB of which 470.12 MiB is free. Process 35181 has 14.28 GiB memory in use. Of the allocated memory 14.16 GiB is allocated by PyTorch, and 1.94 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, float16-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,meta-llama/Meta-Llama-3-8B,meta-llama/Meta-Llama-3-8B,cuda,0,42,,,True,True,,float16,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.224-212.876.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.4.0,,4.44.2,,0.34.2,,,,1.22.0,,,,0.12.0,,"Traceback (most recent call last): @@ -23970,7 +23970,7 @@ bfloat16-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.Py float16-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,phi3,microsoft/Phi-3-mini-4k-instruct,microsoft/Phi-3-mini-4k-instruct,cuda,0,42,,,True,True,,float16,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,,MB,799.997952,8046.641152,0.0,7644.11904,7642.422272,s,1,9.52233203125,9.52233203125,0.0,9.52233203125,9.52233203125,9.52233203125,9.52233203125,[9.52233203125],,kWh,6.425510816664352e-06,6.90792374393327e-07,2.9305578999952453e-06,1.0046861091052925e-05,,MB,1333.98528,8377.991168,0.0,7962.886144,7924.272128,s,10,1.1137695007324218,0.11137695007324219,0.0029297192218080375,0.1117605094909668,0.11385457916259765,0.11391312942504882,0.11395996963500976,"[0.10338703918457032, 0.11021523284912109, 0.11265853118896485, 0.11150643157958984, 0.11177689361572266, 0.1139716796875, 0.11091072082519532, 0.11174412536621094, 0.11384156799316406, 0.11375727844238281]",tokens/s,2298.5007205858374,kWh,3.3828295606321704e-06,3.729990118166601e-07,2.248060546850612e-06,6.0038891192994425e-06,tokens/kWh,42639028.6218129,MB,1367.79776,8380.08832,0.0,7962.886144,7924.274688,s,10,25.381887695312493,2.53818876953125,0.010450145752360251,2.5356971435546876,2.5487151611328125,2.5550666870117187,2.5601479077148435,"[2.535957275390625, 2.519610595703125, 2.532236083984375, 2.5448359375, 2.561418212890625, 2.53621044921875, 2.53543701171875, 2.5345244140625, 2.5473037109375, 2.53435400390625]",tokens/s,24.820848928283127,kWh,7.411609871186834e-05,8.174199167025911e-06,4.9123702772750025e-05,0.00013141400065164427,tokens/kWh,479400.97468763683,,s,630,25.379004028320324,0.040284133378286215,0.0005839811221674359,0.04016435050964355,0.04077804565429687,0.041188852310180664,0.042781143531799314,"[0.040941566467285154, 0.04026572799682617, 0.04020579147338867, 0.04012406539916992, 0.0402215690612793, 0.04065193557739258, 0.04039561462402344, 0.040796161651611325, 0.04161494445800781, 0.04036886215209961, 0.039945919036865236, 0.039782398223876955, 0.042014049530029296, 0.03977283096313477, 0.03993804931640625, 0.04060160064697266, 0.04005039978027344, 0.03983388900756836, 0.03986016082763672, 0.039895233154296876, 0.03999321746826172, 0.039868160247802736, 0.04017996978759766, 0.040683521270751956, 0.04073267364501953, 0.03990911865234375, 0.040548606872558596, 0.04384928131103515, 0.04016582489013672, 0.039929855346679685, 0.04006911849975586, 0.040623966217041015, 0.04041129684448242, 0.04047990417480469, 0.040053600311279296, 0.03985120010375977, 0.03993881607055664, 0.03980047988891602, 0.03980534362792969, 0.03981926345825195, 0.03977830505371094, 0.03995651245117188, 0.039917537689208984, 0.039806846618652345, 0.03987673568725586, 0.04319027328491211, 0.03974044799804687, 0.039863262176513675, 0.039874561309814455, 0.03971686553955078, 0.03991551971435547, 0.03988479995727539, 0.039778144836425784, 0.039876766204833984, 0.03987033462524414, 0.03959961700439453, 0.039583518981933595, 0.04004131317138672, 0.040128288269042967, 0.04009187316894531, 0.03948953628540039, 0.03945676803588867, 0.042254337310791014, 0.04063619232177734, 0.03976988983154297, 0.03988524627685547, 0.04007526397705078, 0.039708671569824217, 0.039723007202148435, 0.039669761657714846, 0.040030208587646485, 0.039962047576904296, 0.039784416198730466, 0.0403702392578125, 0.039825950622558594, 0.03984320068359375, 0.03983219146728516, 0.03982044982910156, 0.03991446304321289, 0.03999321746826172, 0.040349727630615236, 0.04057241439819336, 0.039835231781005856, 0.03967884826660156, 0.039766014099121096, 0.039806976318359374, 0.03993190383911133, 0.04011212921142578, 0.03986022567749024, 0.040052734375, 0.040202239990234374, 0.04023199844360351, 0.03998195266723633, 0.03996268844604492, 0.03989692687988281, 0.039764129638671875, 0.03976512145996094, 0.03992256164550781, 0.03999951934814453, 0.039884769439697265, 0.04007526397705078, 0.04008755111694336, 0.03994617462158203, 0.03995449447631836, 0.039927806854248044, 0.04059913635253906, 0.040274337768554686, 0.03998844909667969, 0.039815521240234374, 0.040024127960205075, 0.04000396728515625, 0.03997625732421875, 0.03993158340454102, 0.03983052825927735, 0.04003430557250977, 0.039755775451660154, 0.04010598373413086, 0.039897087097167966, 0.03970387268066406, 0.039813152313232424, 0.03964313507080078, 0.039672481536865235, 0.040158336639404296, 0.04123328018188477, 0.04039680099487305, 0.040062976837158204, 0.04090774536132812, 0.039808673858642576, 0.04058492660522461, 0.04094966506958008, 0.04004118347167969, 0.03991961669921875, 0.03983078384399414, 0.03983958435058594, 0.040190879821777346, 0.04101932907104492, 0.04013881683349609, 0.03986227035522461, 0.040113441467285155, 0.039798622131347654, 0.03993484878540039, 0.03992166519165039, 0.03995647811889649, 0.03997491073608399, 0.040209503173828126, 0.04026460647583008, 0.040291423797607424, 0.04048169708251953, 0.04036403274536133, 0.04029030227661133, 0.040103935241699216, 0.04020163345336914, 0.04018646240234375, 0.04069580841064453, 0.04391526412963867, 0.04013011169433594, 0.03996102523803711, 0.04001551818847656, 0.0397479362487793, 0.039739391326904294, 0.039814239501953126, 0.03987753677368164, 0.039794048309326174, 0.04003424072265625, 0.039871166229248044, 0.039757823944091795, 0.03974758529663086, 0.040226814270019534, 0.03991487884521484, 0.039915134429931644, 0.040487422943115234, 0.040159744262695314, 0.03996876907348633, 0.04011187362670898, 0.040039710998535157, 0.0406148796081543, 0.041198688507080077, 0.04029942321777344, 0.040742271423339846, 0.04052646255493164, 0.04020159912109375, 0.03998374557495117, 0.039779552459716795, 0.039733505249023436, 0.03984406280517578, 0.03982118225097656, 0.03998764801025391, 0.0402553596496582, 0.03982144165039062, 0.040974143981933595, 0.04028716659545899, 0.04019401550292969, 0.040232990264892576, 0.03993190383911133, 0.03994214248657227, 0.04003583908081055, 0.040339519500732425, 0.039997886657714844, 0.03990323257446289, 0.04002406311035156, 0.04077756881713867, 0.04068745422363281, 0.04026809692382813, 0.0402938232421875, 0.040176128387451174, 0.04003641510009766, 0.040210304260253904, 0.04045177459716797, 0.04017606353759766, 0.04025126266479492, 0.04060172653198242, 0.040406238555908205, 0.0400305290222168, 0.04045849609375, 0.0403109130859375, 0.04025920104980469, 0.03999177551269531, 0.04042860794067383, 0.04048915100097656, 0.040255329132080075, 0.040008609771728515, 0.039766014099121096, 0.03982521438598633, 0.04002016067504883, 0.04036399841308594, 0.0410747184753418, 0.04230348968505859, 0.04031488037109375, 0.04031203079223633, 0.04039350509643555, 0.04021593475341797, 0.040051551818847654, 0.039870529174804686, 0.04027142333984375, 0.041056415557861325, 0.04044105529785156, 0.040280864715576174, 0.04039475250244141, 0.040153087615966795, 0.04050739288330078, 0.040509441375732425, 0.040750240325927736, 0.04059427261352539, 0.0409804801940918, 0.04042099380493164, 0.04045452880859375, 0.04015923309326172, 0.04043571090698242, 0.041758846282958985, 0.04045606231689453, 0.04096108627319336, 0.0409978256225586, 0.04145593643188476, 0.04062051010131836, 0.040478622436523434, 0.04029040145874024, 0.0403900146484375, 0.04027833557128906, 0.040386878967285156, 0.04030803298950195, 0.04146419143676758, 0.04207583999633789, 0.040727169036865234, 0.04036198425292969, 0.040738815307617186, 0.040359935760498046, 0.04025139236450195, 0.04048831939697266, 0.0414071044921875, 0.041226303100585934, 0.04163372802734375, 0.040288257598876956, 0.040195934295654295, 0.040605857849121095, 0.04015923309326172, 0.040217697143554686, 0.040145950317382814, 0.04027305603027344, 0.040083168029785156, 0.039956737518310546, 0.04011494445800781, 0.0405401611328125, 0.04149657440185547, 0.044523521423339846, 0.040676959991455076, 0.04022108840942383, 0.040048641204833986, 0.040361377716064455, 0.040261280059814455, 0.040368896484375, 0.04044819259643555, 0.040255199432373046, 0.04052406311035156, 0.04045209503173828, 0.04062822341918945, 0.04152115249633789, 0.04078387069702148, 0.0411278076171875, 0.04116083145141602, 0.042788864135742184, 0.04069580841064453, 0.040288257598876956, 0.040525825500488284, 0.04025548934936524, 0.04016537475585937, 0.04006083297729492, 0.04004671859741211, 0.040812191009521485, 0.04059782409667969, 0.040525089263916014, 0.04007190322875977, 0.040210430145263674, 0.04003839874267578, 0.04128287887573242, 0.04037907028198242, 0.04103168106079102, 0.040443904876708986, 0.04044595336914063, 0.04037017440795899, 0.040417057037353515, 0.04095395278930664, 0.04043993759155273, 0.04020537567138672, 0.03988780975341797, 0.04024934387207031, 0.040243198394775394, 0.04000937652587891, 0.04004489517211914, 0.039767585754394534, 0.04019862365722656, 0.04036723327636719, 0.040336254119873044, 0.040204288482666016, 0.04021657562255859, 0.0401162223815918, 0.040022014617919925, 0.04001190567016601, 0.040079231262207034, 0.03990300750732422, 0.03996444702148438, 0.039704544067382816, 0.03994416046142578, 0.04009830474853516, 0.04003430557250977, 0.04057478332519531, 0.04037263870239258, 0.04050307083129883, 0.040253440856933595, 0.0404186897277832, 0.040299137115478514, 0.04027801513671875, 0.040097793579101565, 0.03990512084960938, 0.039870624542236326, 0.04015683364868164, 0.04002143859863281, 0.04014172744750977, 0.041093120574951174, 0.04094976043701172, 0.04031078338623047, 0.041009151458740234, 0.040989856719970706, 0.04023353576660156, 0.04003644943237305, 0.04027609634399414, 0.040048702239990235, 0.040013568878173825, 0.04044825744628906, 0.0407256965637207, 0.04016006469726562, 0.040081409454345705, 0.04012607955932617, 0.040083839416503904, 0.04047872161865235, 0.040066913604736326, 0.03999964904785156, 0.040220672607421876, 0.039964672088623046, 0.041641311645507814, 0.04026777648925781, 0.04020905685424805, 0.040521728515625, 0.040196094512939456, 0.04276224136352539, 0.04012790298461914, 0.040142974853515624, 0.03997465515136719, 0.0399508171081543, 0.04005094528198242, 0.040194080352783206, 0.04086092758178711, 0.04199087905883789, 0.04010166549682617, 0.04052540969848633, 0.040135295867919925, 0.040457534790039065, 0.04029715347290039, 0.04021561431884765, 0.04023820877075195, 0.04014060974121094, 0.04015718460083008, 0.04010400009155273, 0.040312767028808594, 0.04039680099487305, 0.03996672058105469, 0.039997440338134765, 0.040235008239746094, 0.04051968002319336, 0.0401162223815918, 0.0399251823425293, 0.03974921417236328, 0.04008649444580078, 0.04000972747802734, 0.04014659118652344, 0.04011452865600586, 0.04012579345703125, 0.040215198516845706, 0.03990937423706055, 0.04026367950439453, 0.04013875198364258, 0.03986598587036133, 0.04006028747558594, 0.03989811325073242, 0.03989820861816406, 0.04006800079345703, 0.040130561828613284, 0.04026777648925781, 0.039907329559326174, 0.04037222290039062, 0.040040321350097656, 0.03992300796508789, 0.04023583984375, 0.03992166519165039, 0.040187904357910156, 0.03995647811889649, 0.039974014282226564, 0.04024614334106445, 0.039839263916015624, 0.04000611114501953, 0.04054143905639648, 0.04032131195068359, 0.04117683029174805, 0.04025347137451172, 0.040200351715087894, 0.04032566452026367, 0.04006707382202149, 0.04001968002319336, 0.04006531143188476, 0.040290016174316406, 0.03992195129394531, 0.039997440338134765, 0.04004569625854492, 0.039908222198486325, 0.040036128997802733, 0.040419551849365236, 0.04002406311035156, 0.03982665634155273, 0.03988355255126953, 0.03988067245483398, 0.039851520538330076, 0.04029801559448242, 0.041335807800292966, 0.042897407531738284, 0.040010974884033206, 0.039932830810546875, 0.03992972946166992, 0.040254848480224606, 0.040124961853027344, 0.03987875366210938, 0.03980492782592773, 0.03986841583251953, 0.03974348831176758, 0.03978400039672852, 0.04002860641479492, 0.040088958740234375, 0.04141324615478516, 0.040049793243408204, 0.0400208625793457, 0.03994214248657227, 0.04010383987426758, 0.039882846832275394, 0.03982745742797852, 0.03966566467285156, 0.039823360443115234, 0.0396954231262207, 0.04081350326538086, 0.03998905563354492, 0.040045921325683596, 0.0410810546875, 0.04037686538696289, 0.04004463958740234, 0.040275966644287106, 0.04044787216186523, 0.04010111999511719, 0.039887744903564455, 0.04006092834472656, 0.04023295974731445, 0.04013667297363281, 0.03989686584472656, 0.04031635284423828, 0.04127008056640625, 0.04124176025390625, 0.04020105743408203, 0.041211265563964844, 0.041398239135742185, 0.040209503173828126, 0.040172447204589845, 0.04012988662719726, 0.04106492614746094, 0.040585407257080076, 0.04028153610229492, 0.04033388900756836, 0.04084143829345703, 0.040619361877441404, 0.040387008666992186, 0.040392704010009765, 0.04049273681640625, 0.039901504516601564, 0.04018175888061523, 0.04074496078491211, 0.040103935241699216, 0.04655513763427734, 0.040953857421875, 0.04019200134277344, 0.040237056732177735, 0.04016332626342774, 0.040017822265625, 0.04008518218994141, 0.04006758499145508, 0.04002191925048828, 0.04008550262451172, 0.03986758422851563, 0.04027065658569336, 0.03998310470581055, 0.04022880172729492, 0.040710208892822265, 0.04116275024414062, 0.0406769905090332, 0.04030835342407227, 0.04021120071411133, 0.04036179351806641, 0.04010412979125977, 0.04012236785888672, 0.04006707382202149, 0.039924991607666015, 0.040069889068603516, 0.04035347366333008, 0.04015052795410156, 0.040366817474365234, 0.04033340835571289, 0.040296382904052734, 0.040162849426269534, 0.040044670104980466, 0.04039107131958008, 0.04034099197387695, 0.040231422424316404, 0.04035174560546875, 0.040226814270019534, 0.04039267349243164, 0.04050048065185547, 0.04024163055419922, 0.04027014541625976, 0.040191200256347655, 0.04027881622314453, 0.040681472778320314, 0.04030044937133789, 0.040615680694580075, 0.04109878540039062, 0.04045395278930664, 0.04038937759399414, 0.040153087615966795, 0.04036608123779297, 0.040101886749267575, 0.04016332626342774, 0.040323070526123043, 0.04040499114990234, 0.04007731246948242, 0.039976383209228514, 0.04016185760498047, 0.04023295974731445, 0.040003135681152345, 0.04020492935180664, 0.039875743865966796, 0.039919456481933596, 0.03989775848388672, 0.04023712158203125, 0.04013475036621094, 0.040322494506835935, 0.040213054656982425, 0.040013599395751956, 0.040057056427001955, 0.04063436889648438, 0.040120319366455076, 0.04016659164428711, 0.040426303863525394, 0.04046438217163086, 0.040474624633789064, 0.04057027053833008, 0.04005744171142578, 0.04033126449584961, 0.04071014404296875, 0.040002750396728515, 0.04000787353515625, 0.03985696029663086, 0.040183616638183595, 0.04018560028076172, 0.03998300933837891, 0.0401693115234375, 0.0407823371887207, 0.04048896026611328, 0.040038272857666014, 0.04048499298095703, 0.04017939376831055, 0.04018368148803711, 0.04048940658569336, 0.040269824981689455, 0.040310432434082034, 0.0401343994140625, 0.04056092834472656, 0.04014521789550781, 0.04025548934936524, 0.039952030181884764, 0.040290721893310545, 0.04010508728027344, 0.04026860809326172, 0.039929855346679685, 0.04000262451171875, 0.040018207550048826, 0.03995510482788086, 0.0401080322265625]",tokens/s,24.823669175393402,, float16-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,phi,microsoft/phi-1_5,microsoft/phi-1_5,cuda,0,42,,,True,True,,float16,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.224-212.876.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.4.0,,4.44.2,,0.34.2,,,,1.22.0,,,,0.12.0,,,MB,883.580928,3447.586816,0.0,3045.064704,2842.846208,s,1,7.4159697265625,7.4159697265625,0.0,7.4159697265625,7.4159697265625,7.4159697265625,7.4159697265625,[7.4159697265625],,kWh,6.189378475013048e-06,6.70126793552225e-07,1.8450014759863187e-06,8.704506744551592e-06,,MB,1307.598848,3623.747584,0.0,3208.64256,2982.452736,s,10,0.3403420181274414,0.03403420181274414,0.0013332341762685943,0.033708992004394533,0.03466038665771484,0.03624656181335449,0.037515501937866215,"[0.03783273696899414, 0.03341984176635742, 0.03409612655639648, 0.03300508880615234, 0.033621505737304686, 0.03292422485351563, 0.034307903289794925, 0.033796478271484375, 0.03343942260742187, 0.03389868927001953]",tokens/s,7521.845272250238,kWh,1.1796272121807028e-06,1.3009251465400757e-07,7.856581106047771e-07,2.0953778374394875e-06,tokens/kWh,122173669.79161486,MB,1340.616704,3623.747584,0.0,3208.64256,2982.455296,s,10,12.024080200195312,1.2024080200195315,0.006990480143182823,1.2019589233398438,1.2105512084960937,1.2113985290527345,1.212076385498047,"[1.212245849609375, 1.1979422607421875, 1.2079669189453126, 1.196303466796875, 1.2003634033203125, 1.203554443359375, 1.1907943115234374, 1.2103629150390625, 1.19519921875, 1.209347412109375]",tokens/s,52.39486010661892,kWh,3.5070653710735914e-05,3.8658401161081896e-06,2.305163873699328e-05,6.198813256383738e-05,tokens/kWh,1016323.56701697,,s,630,12.02196740341187,0.019082487941923597,0.0003478571803843443,0.019001967430114747,0.019425436973571776,0.019584565067291258,0.0205242243385315,"[0.01960550308227539, 0.0191997127532959, 0.01923849678039551, 0.019204063415527342, 0.01940553665161133, 0.019335168838500977, 0.01926527976989746, 0.02037580871582031, 0.019200000762939453, 0.019167104721069337, 0.01913654327392578, 0.019048416137695312, 0.019464319229125976, 0.019076480865478515, 0.019591808319091797, 0.019143808364868165, 0.018964479446411133, 0.018946432113647462, 0.01914112091064453, 0.01935081672668457, 0.019346143722534178, 0.01905254364013672, 0.019459871292114257, 0.019134559631347657, 0.019139904022216797, 0.01955308723449707, 0.01973606491088867, 0.019511808395385744, 0.019412992477416992, 0.019188831329345703, 0.0194303035736084, 0.019763200759887696, 0.019322879791259767, 0.019161216735839842, 0.01915180778503418, 0.019211200714111327, 0.01923244857788086, 0.019044416427612305, 0.019499008178710937, 0.01899545669555664, 0.01927168083190918, 0.019238815307617188, 0.02004764747619629, 0.01902838325500488, 0.018976192474365234, 0.018954687118530274, 0.01913644790649414, 0.019178752899169923, 0.019213151931762696, 0.018948095321655273, 0.018962272644042967, 0.01954217529296875, 0.01888649559020996, 0.01894211196899414, 0.018859615325927736, 0.0189956169128418, 0.019023775100708008, 0.018999008178710936, 0.018901376724243163, 0.019304447174072266, 0.019451904296875, 0.019079168319702147, 0.018898944854736328, 0.019102176666259765, 0.01899728012084961, 0.01895625686645508, 0.018933088302612304, 0.018794975280761718, 0.01896681594848633, 0.019120031356811524, 0.01881292724609375, 0.018974720001220705, 0.01923686408996582, 0.019222496032714843, 0.019017759323120116, 0.018917375564575196, 0.018952287673950196, 0.018912511825561522, 0.018965152740478514, 0.018903039932250978, 0.018951839447021484, 0.018968927383422853, 0.018915327072143554, 0.020555776596069338, 0.02068217658996582, 0.018936384201049806, 0.01883750343322754, 0.01901955223083496, 0.01886636734008789, 0.01895747184753418, 0.01885795211791992, 0.018951072692871093, 0.01881497573852539, 0.01897881507873535, 0.019144767761230468, 0.019139711380004882, 0.01897145652770996, 0.01892483139038086, 0.01884566307067871, 0.01885875129699707, 0.01882931137084961, 0.018822208404541015, 0.019104703903198243, 0.019156991958618166, 0.01908940887451172, 0.01906483268737793, 0.0187675838470459, 0.018839168548583984, 0.018843711853027342, 0.018737760543823243, 0.018739200592041014, 0.019208063125610353, 0.019283775329589845, 0.01909494400024414, 0.019063648223876954, 0.019126176834106445, 0.01890934371948242, 0.018937856674194335, 0.018984960556030273, 0.018820255279541016, 0.01874006462097168, 0.018871551513671876, 0.01882713508605957, 0.01906915283203125, 0.018794368743896485, 0.01902057647705078, 0.019080255508422853, 0.019086111068725587, 0.018976160049438476, 0.019305120468139647, 0.019085119247436524, 0.018982751846313477, 0.019154815673828124, 0.018963008880615233, 0.018921056747436524, 0.018810432434082033, 0.018830047607421876, 0.019027328491210936, 0.0197578239440918, 0.018929664611816405, 0.01920614433288574, 0.018808511734008788, 0.019054912567138673, 0.0188272647857666, 0.019130367279052735, 0.019386240005493164, 0.019437696456909178, 0.019271072387695314, 0.01955081558227539, 0.01921433639526367, 0.01924710464477539, 0.01927987289428711, 0.019117311477661134, 0.019091583251953124, 0.01910028839111328, 0.018905088424682616, 0.018927616119384767, 0.018993152618408202, 0.019109888076782225, 0.018997312545776367, 0.019085248947143554, 0.019462303161621095, 0.01971183967590332, 0.019901695251464843, 0.019323648452758788, 0.01907436752319336, 0.018987071990966796, 0.018989696502685546, 0.019116031646728517, 0.019021823883056642, 0.019277631759643556, 0.019424896240234375, 0.020260992050170897, 0.019301984786987306, 0.019475391387939453, 0.019160991668701173, 0.01906185531616211, 0.01908627128601074, 0.018958303451538087, 0.01894326400756836, 0.0189201602935791, 0.01908684730529785, 0.018914976119995118, 0.018938720703125, 0.019257408142089844, 0.019595199584960938, 0.019347455978393553, 0.01922831916809082, 0.01926780891418457, 0.019279647827148437, 0.01957088088989258, 0.01910483169555664, 0.01901388740539551, 0.019037952423095705, 0.0189752311706543, 0.018902912139892578, 0.018849952697753906, 0.01894438362121582, 0.01914019203186035, 0.01902351951599121, 0.018922496795654296, 0.01908051109313965, 0.01895439910888672, 0.018810976028442384, 0.018966047286987305, 0.018856639862060546, 0.018847808837890626, 0.01878006362915039, 0.018742624282836913, 0.01891600036621094, 0.018843679428100585, 0.018711679458618163, 0.018825599670410156, 0.018761663436889647, 0.019206111907958984, 0.018913888931274415, 0.019406848907470704, 0.019390464782714844, 0.0192423038482666, 0.019114688873291017, 0.018991104125976564, 0.018941280364990234, 0.018843679428100585, 0.01947097587585449, 0.01893574333190918, 0.018919488906860352, 0.018834943771362304, 0.018845407485961915, 0.01884035110473633, 0.01882111930847168, 0.018761728286743166, 0.01910163116455078, 0.019007007598876954, 0.01891587257385254, 0.01884160041809082, 0.01881292724609375, 0.018909183502197266, 0.018739200592041014, 0.019086368560791017, 0.01927471923828125, 0.01922047996520996, 0.018898784637451174, 0.018821279525756837, 0.01908940887451172, 0.018909183502197266, 0.018850847244262694, 0.01909267234802246, 0.01882499122619629, 0.01877743911743164, 0.019996639251708984, 0.018735584259033204, 0.018855583190917968, 0.01884979248046875, 0.018917375564575196, 0.01904844856262207, 0.019310272216796875, 0.019351264953613282, 0.01926313591003418, 0.019188671112060546, 0.019040063858032228, 0.018804927825927735, 0.0188939208984375, 0.01925212860107422, 0.018831615447998048, 0.018788095474243163, 0.01884569549560547, 0.01918976020812988, 0.01895225524902344, 0.018816959381103514, 0.019140640258789064, 0.018892480850219728, 0.018802976608276366, 0.018972671508789063, 0.01904844856262207, 0.018947135925292968, 0.01878112030029297, 0.018876415252685547, 0.01900339126586914, 0.018855072021484374, 0.018814815521240234, 0.01887948799133301, 0.018898752212524413, 0.01883359909057617, 0.01895212745666504, 0.018835519790649412, 0.018886655807495118, 0.018780160903930664, 0.018850847244262694, 0.0192521915435791, 0.01948988723754883, 0.019398847579956056, 0.01903436851501465, 0.018932191848754883, 0.01910335922241211, 0.01946771240234375, 0.019266080856323243, 0.019091712951660157, 0.0192325439453125, 0.018897279739379883, 0.01898428726196289, 0.01880303955078125, 0.01883734321594238, 0.018921951293945312, 0.01904844856262207, 0.019134464263916014, 0.01897881507873535, 0.01897382354736328, 0.019005727767944337, 0.01920790481567383, 0.020732864379882813, 0.019943359375, 0.019066879272460938, 0.01903615951538086, 0.019005311965942382, 0.018917312622070314, 0.019060800552368164, 0.019309375762939455, 0.019523584365844726, 0.01947216033935547, 0.01926576042175293, 0.019359743118286133, 0.019383552551269532, 0.019219200134277345, 0.019254560470581054, 0.02060576057434082, 0.019947423934936523, 0.01918476867675781, 0.01904934310913086, 0.0192774715423584, 0.019122112274169923, 0.019022207260131836, 0.019003423690795898, 0.019070976257324217, 0.019126272201538085, 0.01901081657409668, 0.018840288162231444, 0.018837024688720703, 0.018745567321777342, 0.01873539161682129, 0.01898700714111328, 0.018871423721313476, 0.018748287200927735, 0.018904640197753907, 0.01903455924987793, 0.019199167251586914, 0.01901148796081543, 0.01879747200012207, 0.018869407653808595, 0.019571552276611327, 0.01880268859863281, 0.018902463912963866, 0.018737407684326173, 0.018690015792846678, 0.01867401695251465, 0.01884979248046875, 0.018875776290893556, 0.018760063171386718, 0.018753791809082033, 0.01878835105895996, 0.01874652862548828, 0.01993199920654297, 0.020790752410888673, 0.02061955261230469, 0.0189453125, 0.018820064544677734, 0.019320831298828126, 0.019025856018066407, 0.018808448791503906, 0.018657312393188477, 0.019140703201293945, 0.019220767974853517, 0.018808128356933594, 0.018772703170776367, 0.018919424057006837, 0.018948095321655273, 0.0187936954498291, 0.019041055679321288, 0.018780160903930664, 0.01893391990661621, 0.019070335388183594, 0.018839872360229493, 0.01890358352661133, 0.018859167098999024, 0.019891040802001953, 0.018882080078125, 0.01877244758605957, 0.018716672897338867, 0.018892799377441406, 0.018746559143066405, 0.018699071884155274, 0.01874483108520508, 0.01873094367980957, 0.018727487564086914, 0.018788000106811524, 0.018872671127319336, 0.01908527946472168, 0.01922870445251465, 0.018931711196899414, 0.018880512237548826, 0.018851743698120118, 0.018796640396118162, 0.01871574401855469, 0.018776960372924804, 0.018733055114746093, 0.0187857608795166, 0.019011199951171873, 0.018776224136352538, 0.018826112747192383, 0.018807968139648436, 0.01901798439025879, 0.019175615310668945, 0.018910783767700196, 0.01887708854675293, 0.01888060760498047, 0.018808832168579103, 0.018892799377441406, 0.018876352310180665, 0.018833471298217774, 0.018750816345214843, 0.018831903457641602, 0.01872435188293457, 0.0193686408996582, 0.01911596870422363, 0.018978208541870118, 0.018860063552856444, 0.01882374382019043, 0.018882560729980468, 0.018735103607177735, 0.018769792556762695, 0.018829471588134767, 0.018798559188842774, 0.01884569549560547, 0.01909884834289551, 0.019155744552612305, 0.0190382080078125, 0.018968095779418947, 0.018917856216430665, 0.018919424057006837, 0.01887846374511719, 0.018755168914794923, 0.018991424560546876, 0.01894607925415039, 0.018952287673950196, 0.019013504028320312, 0.018991104125976564, 0.018874368667602538, 0.01923232078552246, 0.02023468780517578, 0.019799968719482423, 0.019802207946777343, 0.01954368019104004, 0.019302783966064455, 0.019695295333862304, 0.019238304138183594, 0.019145631790161134, 0.018968128204345704, 0.01891539192199707, 0.018876800537109373, 0.019019775390625, 0.01904640007019043, 0.019062591552734376, 0.019046560287475586, 0.018931232452392577, 0.018864639282226564, 0.018917375564575196, 0.018851839065551757, 0.01883340835571289, 0.019283967971801756, 0.018884735107421877, 0.018796415328979493, 0.01873744010925293, 0.01879216003417969, 0.020078847885131836, 0.02044697570800781, 0.018873760223388672, 0.01890121650695801, 0.01914531135559082, 0.01957571220397949, 0.019479423522949218, 0.019475679397583007, 0.019326847076416017, 0.01916102409362793, 0.019116928100585937, 0.019011199951171873, 0.018969184875488283, 0.01892134475708008, 0.018964351654052733, 0.019136640548706056, 0.019113536834716796, 0.023068832397460937, 0.019290399551391602, 0.019109888076782225, 0.018893823623657227, 0.01887718391418457, 0.018816831588745118, 0.018892927169799803, 0.018872095108032227, 0.01900588798522949, 0.019287263870239258, 0.019108320236206056, 0.01907084846496582, 0.018962976455688476, 0.019177471160888672, 0.019386016845703125, 0.01919171142578125, 0.019271488189697265, 0.019364288330078125, 0.019565664291381835, 0.019358240127563476, 0.019132640838623045, 0.019058048248291017, 0.019028608322143554, 0.019186784744262695, 0.019140512466430663, 0.01901798439025879, 0.01897756767272949, 0.01889254379272461, 0.018888128280639647, 0.018866975784301757, 0.018838752746582033, 0.01881167984008789, 0.018822336196899415, 0.018875200271606444, 0.01904800033569336, 0.01914703941345215, 0.018868383407592774, 0.018961759567260743, 0.01888435173034668, 0.018833919525146483, 0.01905289649963379, 0.018837152481079103, 0.01886867141723633, 0.0189716796875, 0.018758079528808595, 0.018893312454223633, 0.018883968353271486, 0.018619007110595703, 0.018765792846679688, 0.01889436721801758, 0.019027456283569336, 0.018893760681152345, 0.018834751129150392, 0.018774240493774415, 0.01872140884399414, 0.01880463981628418, 0.01887027168273926, 0.018747167587280275, 0.0188438720703125, 0.018860031127929687, 0.019019903182983397, 0.018918815612792968, 0.018874847412109375, 0.01887433624267578, 0.018702367782592773, 0.01937785530090332, 0.019013952255249024, 0.019039968490600585, 0.018876991271972655, 0.01895599937438965, 0.01906255912780762, 0.01898313522338867, 0.018751487731933594, 0.018920448303222655, 0.01887116813659668, 0.019737855911254883, 0.01908367919921875, 0.01897724723815918, 0.018888927459716796, 0.019251007080078125, 0.019093311309814454, 0.019001695632934572, 0.01890265655517578, 0.0189116153717041, 0.018925567626953126, 0.018910207748413087, 0.018882976531982423, 0.019274335861206054, 0.01925324821472168, 0.01888684844970703, 0.019080127716064453, 0.019002239227294923, 0.018925439834594725, 0.018864191055297852, 0.018853952407836914, 0.019171327590942384, 0.019150848388671874, 0.019316736221313476, 0.019490816116333007, 0.019532032012939453, 0.019536991119384766, 0.01943414306640625, 0.019355648040771483, 0.01929792022705078, 0.019123743057250977, 0.019170143127441405, 0.01902083206176758, 0.018995647430419923, 0.019157535552978517, 0.019125312805175782, 0.0190794563293457, 0.019384992599487304, 0.019128320693969726, 0.019171136856079102, 0.01910799980163574, 0.019719392776489257, 0.019294240951538085, 0.019159839630126952, 0.018997247695922852, 0.01923481559753418, 0.019144447326660156, 0.019286272048950195, 0.019146751403808594, 0.01917305564880371, 0.019206464767456053, 0.01931180763244629, 0.020247360229492188, 0.019457536697387694, 0.019355295181274414, 0.019505664825439452, 0.019229024887084962, 0.019214591979980468, 0.01920524787902832, 0.018994815826416017, 0.019112960815429687, 0.01915228843688965, 0.019034719467163085, 0.019134496688842772, 0.019401792526245118, 0.019554752349853516, 0.01920047950744629]",tokens/s,52.404068224407624,, float16-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,phi,microsoft/phi-1_5,microsoft/phi-1_5,cuda,0,42,,,True,True,,float16,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.224-212.876.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.4.0,,4.44.2,,0.34.2,,,,1.22.0,,,,0.12.0,,,MB,883.314688,3447.586816,0.0,3045.064704,2842.846208,s,1,7.27972607421875,7.27972607421875,0.0,7.27972607421875,7.27972607421875,7.27972607421875,7.27972607421875,[7.27972607421875],,kWh,5.9193053208370355e-06,6.2582366923507e-07,2.0352794060096713e-06,8.580408396081776e-06,,MB,1339.404288,3646.816256,0.0,3229.61408,2982.452736,s,10,0.43298176193237303,0.0432981761932373,0.0011073369945902936,0.04303446578979492,0.04362301368713379,0.0450688024520874,0.046225433464050295,"[0.04651459121704102, 0.04304624176025391, 0.04268492889404297, 0.042840065002441405, 0.042525184631347655, 0.04326047897338867, 0.043022689819335935, 0.04325619125366211, 0.0425296630859375, 0.04330172729492188]",tokens/s,5912.489220273079,kWh,1.4488039630364121e-06,1.5977669475319217e-07,9.565367938315837e-07,2.5651174516211882e-06,tokens/kWh,99800498.35074984,MB,1378.418688,3646.816256,0.0,3229.61408,2982.455296,s,10,14.71218017578125,1.471218017578125,0.006520559730111671,1.4740194091796874,1.476150732421875,1.4782819702148438,1.4799869604492188,"[1.4740574951171874, 1.4804132080078125, 1.4681180419921875, 1.474802490234375, 1.4563297119140626, 1.467184326171875, 1.4661009521484376, 1.4755155029296876, 1.4756771240234374, 1.4739813232421874]",tokens/s,42.82166153980952,kWh,4.241440238488713e-05,4.677934810174127e-06,2.6915763281769904e-05,7.400810047683117e-05,tokens/kWh,851258.1676072426,,s,630,14.709816682815553,0.023348915369548494,0.00040903208967215624,0.023234016418457033,0.02376182060241699,0.02399247236251831,0.025019298324584965,"[0.024095424652099608, 0.02422528076171875, 0.02356038475036621, 0.023521600723266603, 0.023207935333251953, 0.023742431640625, 0.023303199768066406, 0.023393184661865234, 0.023347103118896484, 0.023217599868774415, 0.023009632110595705, 0.023098943710327148, 0.022962303161621095, 0.02297916793823242, 0.02298271942138672, 0.022987936019897463, 0.02306550407409668, 0.02297222328186035, 0.023044288635253905, 0.023046112060546874, 0.023289920806884766, 0.02316080093383789, 0.023267328262329103, 0.023310335159301757, 0.024853759765625, 0.023407360076904297, 0.02327347183227539, 0.023166303634643556, 0.0231364803314209, 0.02513145637512207, 0.02504265594482422, 0.023132448196411134, 0.023191007614135742, 0.023061023712158204, 0.02304572868347168, 0.023062944412231445, 0.023084447860717772, 0.02314089584350586, 0.02375071907043457, 0.023994272232055663, 0.023384159088134765, 0.023351104736328124, 0.023215839385986328, 0.023275199890136718, 0.023769535064697266, 0.023206399917602538, 0.023225343704223633, 0.024037792205810548, 0.02312828826904297, 0.02350921630859375, 0.023783424377441405, 0.0234998722076416, 0.023403423309326172, 0.023871488571166992, 0.023664608001708984, 0.02326736068725586, 0.023142400741577147, 0.023076864242553712, 0.023250944137573244, 0.023193599700927735, 0.023209983825683594, 0.023105087280273436, 0.023014944076538087, 0.024061952590942383, 0.02385856056213379, 0.023790079116821287, 0.023765119552612304, 0.02351103973388672, 0.023350431442260743, 0.02327987289428711, 0.023290239334106445, 0.023150815963745117, 0.02311404800415039, 0.023557823181152345, 0.024864511489868166, 0.023730016708374022, 0.023412607192993165, 0.023187999725341798, 0.02326016044616699, 0.023170047760009766, 0.02308639907836914, 0.023107423782348632, 0.023208032608032225, 0.02336844825744629, 0.023690528869628906, 0.024120031356811525, 0.023762943267822266, 0.02426470375061035, 0.0248907527923584, 0.023954240798950196, 0.023698463439941406, 0.0236592960357666, 0.023425024032592775, 0.023573951721191408, 0.02323308753967285, 0.023626815795898436, 0.023194656372070313, 0.023199647903442384, 0.023023616790771483, 0.0231911678314209, 0.022957759857177733, 0.02307551956176758, 0.023183263778686524, 0.023047807693481446, 0.023062944412231445, 0.023064640045166014, 0.02309939193725586, 0.023093248367309572, 0.023178367614746093, 0.023224735260009767, 0.0234454402923584, 0.02352387237548828, 0.023680320739746095, 0.02365305519104004, 0.0234716796875, 0.023353792190551757, 0.023138303756713868, 0.02462735939025879, 0.026070880889892577, 0.023389471054077147, 0.02323481559753418, 0.023585151672363282, 0.023037439346313478, 0.022937824249267578, 0.022959936141967775, 0.02347270393371582, 0.023922048568725585, 0.02360153579711914, 0.02388960075378418, 0.023599679946899415, 0.023302143096923827, 0.02329100799560547, 0.02363465690612793, 0.023621631622314454, 0.023834783554077147, 0.02365644836425781, 0.02348441505432129, 0.023347200393676756, 0.02326268768310547, 0.023181407928466798, 0.02316057586669922, 0.022901439666748048, 0.022977792739868164, 0.02293836784362793, 0.0230645751953125, 0.02296406364440918, 0.02343747138977051, 0.02339583969116211, 0.023223840713500976, 0.023098335266113282, 0.022994943618774414, 0.022927263259887695, 0.023011423110961913, 0.022953983306884765, 0.02299068832397461, 0.023036064147949217, 0.023408447265625, 0.023439552307128905, 0.023603071212768556, 0.023509120941162108, 0.02330316734313965, 0.023085376739501954, 0.023171775817871092, 0.023068672180175782, 0.023178560256958008, 0.022961919784545898, 0.02299728012084961, 0.023146144866943358, 0.023403135299682618, 0.02347660827636719, 0.02322256088256836, 0.023254175186157227, 0.023290399551391602, 0.023140384674072267, 0.023131776809692382, 0.023017568588256834, 0.02317750358581543, 0.023224159240722655, 0.023087263107299805, 0.02341481590270996, 0.023179040908813477, 0.02295801544189453, 0.023040191650390625, 0.023310400009155272, 0.023214080810546874, 0.02323187255859375, 0.02316703987121582, 0.02311635208129883, 0.026261503219604493, 0.024090400695800783, 0.023916543960571288, 0.023700544357299805, 0.023467775344848632, 0.023526464462280273, 0.02337887954711914, 0.02321414375305176, 0.023008703231811523, 0.02309119987487793, 0.02325315284729004, 0.023206239700317384, 0.023069856643676757, 0.022963039398193358, 0.02310553550720215, 0.023326784133911132, 0.02381590461730957, 0.023375520706176756, 0.023730527877807616, 0.023635520935058593, 0.023192512512207032, 0.023215904235839843, 0.023316415786743164, 0.02306662368774414, 0.023345151901245118, 0.023418880462646483, 0.024772607803344726, 0.023791648864746093, 0.02359209632873535, 0.023521408081054688, 0.023347007751464845, 0.023200639724731444, 0.0233482551574707, 0.02313520050048828, 0.023338464736938475, 0.02329654312133789, 0.023135616302490235, 0.023424863815307617, 0.023366559982299806, 0.023553056716918944, 0.023598112106323243, 0.02347398376464844, 0.02322163200378418, 0.02340662384033203, 0.02323516845703125, 0.023076351165771485, 0.023060640335083007, 0.023104927062988282, 0.023034624099731445, 0.02307030487060547, 0.02365222358703613, 0.023675615310668946, 0.023605247497558594, 0.024393503189086913, 0.02336956787109375, 0.02323289680480957, 0.02308915138244629, 0.02324684715270996, 0.02354297637939453, 0.02405049514770508, 0.02328780746459961, 0.023102783203125, 0.023124895095825194, 0.02367980766296387, 0.023071008682250975, 0.02328780746459961, 0.023433216094970705, 0.023222335815429686, 0.023197696685791015, 0.023502784729003905, 0.024141632080078124, 0.023377695083618165, 0.023148672103881836, 0.023165216445922853, 0.023212032318115236, 0.02307891273498535, 0.023123327255249022, 0.02303241539001465, 0.023119903564453124, 0.023068672180175782, 0.02291302490234375, 0.022930816650390626, 0.02307904052734375, 0.022944095611572266, 0.023031103134155274, 0.023661407470703125, 0.022978464126586915, 0.022972543716430663, 0.022978368759155272, 0.022945951461791993, 0.02294169616699219, 0.023011327743530274, 0.022988800048828126, 0.022803808212280275, 0.022936223983764648, 0.02289459228515625, 0.023027711868286133, 0.023242944717407225, 0.02330124855041504, 0.02339023971557617, 0.02325974464416504, 0.02300716781616211, 0.0231507511138916, 0.023221216201782226, 0.023112415313720703, 0.023027999877929688, 0.023157920837402344, 0.022901599884033202, 0.023066463470458983, 0.022968000411987304, 0.02295814323425293, 0.022917215347290038, 0.023036224365234375, 0.022996063232421874, 0.023118623733520506, 0.023040031433105467, 0.023221439361572265, 0.02316713523864746, 0.023382783889770508, 0.023207935333251953, 0.02310348892211914, 0.02303401565551758, 0.023002912521362304, 0.022917184829711914, 0.0229453125, 0.023006879806518554, 0.023038272857666017, 0.023761695861816406, 0.023799808502197265, 0.023563743591308594, 0.0232042236328125, 0.02319491195678711, 0.023465919494628906, 0.02341164779663086, 0.02325299263000488, 0.023104799270629882, 0.023300895690917967, 0.023472063064575194, 0.023348608016967774, 0.02343302345275879, 0.02339036750793457, 0.02328848075866699, 0.023236608505249022, 0.023052288055419923, 0.02309939193725586, 0.023001087188720702, 0.02331443214416504, 0.023961599349975587, 0.023558143615722657, 0.02369331169128418, 0.02381110382080078, 0.023470495223999025, 0.023267904281616212, 0.023158784866333007, 0.023216127395629883, 0.02323404884338379, 0.023208255767822265, 0.02315430450439453, 0.023151168823242186, 0.02312518310546875, 0.023009599685668944, 0.023083072662353515, 0.023026111602783204, 0.023203840255737306, 0.023009248733520508, 0.023178848266601562, 0.023169471740722657, 0.023064191818237306, 0.0231200008392334, 0.023052032470703126, 0.023031679153442383, 0.02346691131591797, 0.023746271133422852, 0.023653535842895507, 0.023581375122070314, 0.023417087554931642, 0.02323961639404297, 0.02319171142578125, 0.023128095626831054, 0.02303670310974121, 0.02312396812438965, 0.023195840835571288, 0.02320595169067383, 0.023307327270507813, 0.023607551574707033, 0.023165376663208007, 0.023087167739868165, 0.022996831893920898, 0.022986879348754884, 0.023185375213623047, 0.02327756881713867, 0.02414985656738281, 0.02357907295227051, 0.023258207321166992, 0.02314847946166992, 0.02325551986694336, 0.023544031143188475, 0.02349056053161621, 0.02353971290588379, 0.023463935852050782, 0.023178976058959962, 0.023131711959838867, 0.02337843132019043, 0.02328566360473633, 0.023269695281982423, 0.023135520935058593, 0.023096031188964843, 0.023048063278198243, 0.0232938232421875, 0.02319385528564453, 0.023830528259277343, 0.02311174392700195, 0.023046079635620116, 0.023250080108642577, 0.023425888061523438, 0.023233983993530275, 0.023259552001953124, 0.02317430305480957, 0.023126976013183594, 0.0229683837890625, 0.02296780776977539, 0.022912544250488283, 0.02309391975402832, 0.023359647750854494, 0.023568544387817383, 0.023371328353881837, 0.023458112716674806, 0.023354848861694335, 0.02325161552429199, 0.022982656478881838, 0.023012704849243164, 0.02295052719116211, 0.02325302314758301, 0.022962175369262695, 0.023334495544433592, 0.023085056304931642, 0.023087520599365235, 0.0229532470703125, 0.02324883270263672, 0.02292620849609375, 0.022959871292114256, 0.023324064254760742, 0.023226335525512697, 0.023918399810791014, 0.023786624908447265, 0.02361737632751465, 0.023541759490966797, 0.02315817642211914, 0.023179872512817383, 0.02320207977294922, 0.023299808502197265, 0.023183359146118163, 0.023154272079467773, 0.02388096046447754, 0.02373110389709473, 0.023416831970214845, 0.02332464027404785, 0.02347216033935547, 0.023584543228149416, 0.023586271286010742, 0.023424928665161132, 0.02330300712585449, 0.023195199966430664, 0.023196096420288085, 0.023180896759033204, 0.023028127670288084, 0.023990272521972656, 0.023642112731933593, 0.02334649658203125, 0.02320454406738281, 0.023183359146118163, 0.023118911743164064, 0.023124927520751952, 0.023488512039184572, 0.023287200927734376, 0.023598976135253906, 0.023104480743408203, 0.023131904602050782, 0.023101663589477538, 0.023371551513671877, 0.02369536018371582, 0.0234967041015625, 0.02351043128967285, 0.023714399337768553, 0.023749664306640626, 0.023651296615600587, 0.023604671478271486, 0.02350752067565918, 0.023937023162841797, 0.023436800003051757, 0.02322470474243164, 0.023173248291015625, 0.02327552032470703, 0.023390207290649414, 0.0236810245513916, 0.02355200004577637, 0.023371776580810546, 0.023184703826904296, 0.023202495574951174, 0.023185407638549805, 0.023159040451049804, 0.023265024185180665, 0.023205888748168944, 0.023635871887207033, 0.0243570556640625, 0.023749984741210938, 0.023616031646728517, 0.023506975173950194, 0.023228416442871092, 0.02340355110168457, 0.023450592041015624, 0.02327552032470703, 0.02317424011230469, 0.023149471282958984, 0.023173120498657225, 0.023389280319213866, 0.025873247146606444, 0.02342051124572754, 0.02335580825805664, 0.023152191162109374, 0.023144895553588868, 0.02326911926269531, 0.02309350395202637, 0.023355392456054686, 0.023375871658325196, 0.023556095123291015, 0.02496211242675781, 0.023401407241821288, 0.023463935852050782, 0.02538047981262207, 0.023159168243408204, 0.023087200164794923, 0.022951103210449218, 0.022949600219726564, 0.023006208419799806, 0.023147743225097658, 0.02309926414489746, 0.023139232635498046, 0.023183359146118163, 0.02294940757751465, 0.02304047966003418, 0.023015424728393553, 0.023060480117797853, 0.023093248367309572, 0.02332262420654297, 0.02450361633300781, 0.02466454315185547, 0.023793216705322265, 0.023499263763427734, 0.023568672180175783, 0.024200735092163087, 0.023154336929321288, 0.02303046417236328, 0.023203712463378906, 0.023083135604858397, 0.02329804801940918, 0.023334911346435547, 0.022999040603637694, 0.023551488876342775, 0.023031455993652344, 0.023100255966186523, 0.022984703063964843, 0.022990175247192383, 0.02385955238342285, 0.023344863891601564, 0.022937376022338866, 0.02297734451293945, 0.02282921600341797, 0.023101280212402344, 0.02287808036804199, 0.02305241584777832, 0.02289206314086914, 0.023529151916503906, 0.02378566360473633, 0.02391110420227051, 0.02346499252319336, 0.02407244873046875, 0.023843456268310546, 0.023795711517333985, 0.02375040054321289, 0.023419584274291992, 0.023635135650634766, 0.02319215965270996, 0.023384511947631834, 0.023281728744506836, 0.023178976058959962, 0.023080543518066408, 0.023251359939575195, 0.023394304275512694, 0.023489824295043947, 0.023433952331542968, 0.023412384033203126, 0.023561824798583986, 0.02357529640197754, 0.02364409637451172, 0.023412799835205077, 0.02329167938232422, 0.023166271209716798, 0.023157024383544923, 0.02320400047302246, 0.023082592010498046, 0.02353651237487793, 0.024069664001464843, 0.02366716766357422, 0.023627775192260742, 0.02393497657775879, 0.023137792587280274, 0.02321241569519043, 0.023062463760375976, 0.02311315155029297, 0.022979328155517578, 0.023214080810546874, 0.02344960021972656, 0.02344723129272461, 0.023429439544677733, 0.023349248886108398, 0.023256607055664062, 0.025977279663085936, 0.023719968795776366, 0.02346518325805664, 0.023437952041625975, 0.02305878448486328, 0.023045120239257814, 0.023067264556884765, 0.0233822078704834, 0.023324512481689454, 0.023423040390014648, 0.023214176177978517, 0.0230645751953125, 0.02305433654785156, 0.023074623107910155, 0.02322643280029297, 0.0234333438873291, 0.02381932830810547, 0.02318841552734375, 0.023108896255493165, 0.02317180824279785, 0.023059648513793947, 0.02314463996887207, 0.023380544662475584, 0.02401257514953613, 0.023410240173339845]",tokens/s,42.8285418903952,, -bfloat16-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,phi3,microsoft/Phi-3-mini-4k-instruct,microsoft/Phi-3-mini-4k-instruct,cuda,0,42,,,True,True,,bfloat16,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +bfloat16-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,phi3,microsoft/Phi-3-mini-4k-instruct,microsoft/Phi-3-mini-4k-instruct,cuda,0,42,,,True,True,,bfloat16,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -24004,7 +24004,7 @@ ValueError: Phi3ForCausalLM does not support an attention implementation through ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, float32-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,phi,microsoft/phi-1_5,microsoft/phi-1_5,cuda,0,42,,,True,True,,float32,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.224-212.876.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.4.0,,4.44.2,,0.34.2,,,,1.22.0,,,,0.12.0,,,MB,885.383168,6089.998336,0.0,5687.476224,5685.679104,s,1,7.2154111328125,7.2154111328125,0.0,7.2154111328125,7.2154111328125,7.2154111328125,7.2154111328125,[7.2154111328125],,kWh,6.160286183376229e-06,6.490118287632258e-07,2.160001727979921e-06,8.969299740119376e-06,,MB,1185.0752,6324.87936,0.0,5911.871488,5850.451456,s,10,2.0647819976806643,0.2064781997680664,0.0035611014775419956,0.20754294586181643,0.20964447326660157,0.20971802673339843,0.20977686950683594,"[0.19737107849121094, 0.20962812805175782, 0.2097915802001953, 0.20860684204101562, 0.205314208984375, 0.20371104431152343, 0.20913235473632813, 0.206140869140625, 0.20749098205566407, 0.20759490966796876]",tokens/s,1239.8403332049613,kWh,5.818667245016424e-06,6.416971915618303e-07,3.87657172870538e-06,1.0336936165283636e-05,tokens/kWh,24765558.76002893,MB,1190.346752,6324.87936,0.0,5911.871488,5850.454016,s,10,16.424772827148438,1.6424772827148437,0.0017984575255992113,1.6431479492187502,1.6441755249023438,1.644232598876953,1.6442782580566406,"[1.6395599365234375, 1.63903662109375, 1.6421165771484374, 1.642842041015625, 1.6415816650390624, 1.6442896728515626, 1.6438729248046875, 1.643856689453125, 1.643453857421875, 1.644162841796875]",tokens/s,38.35669489191812,kWh,4.8175107581231194e-05,5.313461340400184e-06,3.198365140389735e-05,8.547222032552873e-05,tokens/kWh,737081.5893170759,,s,630,16.422411247253407,0.02606731944008479,0.00038848892341905757,0.025981664657592773,0.026220637512207032,0.02637586679458618,0.028588566150665285,"[0.028468671798706054, 0.027068960189819337, 0.026368000030517577, 0.026068992614746093, 0.02588640022277832, 0.025976320266723633, 0.02580768013000488, 0.025823232650756835, 0.0258024959564209, 0.026196191787719727, 0.025781375885009766, 0.025818016052246092, 0.025825279235839844, 0.025866239547729493, 0.025870336532592773, 0.025851743698120117, 0.02588038444519043, 0.025903455734252928, 0.025827327728271485, 0.02580611228942871, 0.025866464614868166, 0.025907583236694336, 0.025895040512084962, 0.02582863998413086, 0.025948896408081054, 0.025896671295166016, 0.025925920486450194, 0.025858047485351563, 0.025917695999145507, 0.02593152046203613, 0.025984895706176757, 0.02587174415588379, 0.025901088714599608, 0.025905055999755858, 0.02587526321411133, 0.026025983810424806, 0.026206207275390626, 0.026265535354614258, 0.026219743728637696, 0.026180063247680664, 0.026198400497436523, 0.02607689666748047, 0.026077472686767578, 0.026019840240478515, 0.026074207305908204, 0.026026912689208984, 0.025990495681762694, 0.025941888809204103, 0.0259486083984375, 0.025964895248413087, 0.025948160171508788, 0.02632294464111328, 0.02591107177734375, 0.026087167739868165, 0.0260546875, 0.02593222427368164, 0.025961727142333985, 0.025903839111328125, 0.02600553512573242, 0.02587593650817871, 0.02592790412902832, 0.02586476707458496, 0.025915136337280275, 0.02855900764465332, 0.02722003173828125, 0.026458431243896484, 0.02609619140625, 0.02598297691345215, 0.025833471298217774, 0.02587811279296875, 0.02579088020324707, 0.025831008911132814, 0.025799072265625, 0.025767936706542968, 0.02585536003112793, 0.025868928909301758, 0.02585116767883301, 0.025843744277954103, 0.02584003257751465, 0.02584339141845703, 0.0258504638671875, 0.025793760299682618, 0.025937887191772462, 0.02585478401184082, 0.025886720657348632, 0.025860095977783205, 0.025887744903564453, 0.02585004806518555, 0.025870880126953124, 0.02586556816101074, 0.025901216506958008, 0.02591209602355957, 0.025826528549194337, 0.025834272384643555, 0.026064895629882814, 0.025860095977783205, 0.025921503067016603, 0.025942047119140624, 0.025994400024414062, 0.02618864059448242, 0.02625721549987793, 0.02631491279602051, 0.026191904067993165, 0.02615500831604004, 0.02611609649658203, 0.02603936004638672, 0.025967071533203125, 0.025950687408447266, 0.02597887992858887, 0.02596646308898926, 0.025929311752319335, 0.02591798400878906, 0.02596249580383301, 0.02593382453918457, 0.025947647094726564, 0.026083839416503905, 0.026076992034912108, 0.026043840408325195, 0.025965311050415038, 0.02588876724243164, 0.025890815734863282, 0.025998624801635742, 0.025909887313842774, 0.026003551483154298, 0.025960447311401368, 0.025938976287841798, 0.028807872772216796, 0.027283456802368163, 0.026464256286621093, 0.02614476776123047, 0.025976831436157227, 0.02590924835205078, 0.025888736724853516, 0.02583760070800781, 0.025891935348510742, 0.02585487937927246, 0.025827327728271485, 0.025855104446411134, 0.02598182487487793, 0.02597887992858887, 0.02598297691345215, 0.025892704010009766, 0.025927391052246094, 0.02594041633605957, 0.025939968109130858, 0.026029600143432616, 0.026000896453857423, 0.026093759536743165, 0.025921760559082033, 0.025852479934692384, 0.025878528594970703, 0.025896575927734374, 0.025917823791503907, 0.025937471389770508, 0.0259400634765625, 0.02588627243041992, 0.02595305633544922, 0.026030080795288086, 0.02594144058227539, 0.025969215393066406, 0.025987071990966795, 0.026113407135009765, 0.02620275115966797, 0.026178688049316407, 0.026211391448974608, 0.026101631164550783, 0.026163135528564453, 0.02617308807373047, 0.026206560134887695, 0.026101760864257813, 0.026074911117553713, 0.026011871337890624, 0.026015743255615235, 0.026030080795288086, 0.026025983810424806, 0.02595155143737793, 0.02594476890563965, 0.02594611167907715, 0.02593382453918457, 0.02595430374145508, 0.026044416427612304, 0.026038015365600586, 0.025949472427368163, 0.025954431533813476, 0.026053472518920897, 0.025991167068481445, 0.025998464584350588, 0.025932287216186522, 0.025964576721191405, 0.028925952911376954, 0.027355167388916017, 0.026633920669555663, 0.02620649528503418, 0.02608742332458496, 0.025886720657348632, 0.025878143310546876, 0.025829120635986327, 0.02586595153808594, 0.02582406425476074, 0.025892000198364257, 0.025874975204467774, 0.025853439331054686, 0.025834175109863283, 0.025823455810546875, 0.025939968109130858, 0.025829376220703124, 0.025821184158325194, 0.025814815521240233, 0.025886751174926757, 0.025864383697509766, 0.02591062355041504, 0.025897504806518555, 0.02589411163330078, 0.025927648544311524, 0.025973567962646483, 0.025954431533813476, 0.026046464920043946, 0.025952255249023438, 0.02595020866394043, 0.025910400390625, 0.025873279571533202, 0.02588057518005371, 0.025968416213989258, 0.025953792572021486, 0.026139360427856445, 0.02619539260864258, 0.026315296173095703, 0.02628438377380371, 0.0262260799407959, 0.026181119918823242, 0.02616329574584961, 0.026233535766601562, 0.026085119247436523, 0.026007232666015626, 0.02603424072265625, 0.026198528289794923, 0.026038272857666016, 0.02605606460571289, 0.02601433563232422, 0.02614067268371582, 0.025991167068481445, 0.02604377555847168, 0.026030271530151368, 0.026028480529785156, 0.02595840072631836, 0.026032127380371094, 0.026218496322631835, 0.026011648178100585, 0.025931776046752928, 0.02599839973449707, 0.025990079879760743, 0.026008832931518556, 0.028617439270019532, 0.027092992782592775, 0.026558464050292968, 0.026220544815063477, 0.02607855987548828, 0.02597750473022461, 0.025923583984375, 0.02586134338378906, 0.02588751983642578, 0.025862144470214843, 0.025882623672485353, 0.025894912719726562, 0.025956352233886718, 0.02590496063232422, 0.025899200439453124, 0.025863775253295897, 0.026018207550048827, 0.02595840072631836, 0.025993215560913087, 0.02597478485107422, 0.02599238395690918, 0.02590582466125488, 0.025915552139282226, 0.02591939163208008, 0.025913408279418945, 0.025976863861083985, 0.0259420166015625, 0.025943424224853514, 0.02592950439453125, 0.025934688568115233, 0.02591334342956543, 0.025911296844482422, 0.025924800872802734, 0.02594633674621582, 0.025908992767333983, 0.02617024040222168, 0.026252832412719727, 0.026292671203613283, 0.026203264236450197, 0.026183904647827147, 0.02614748764038086, 0.026054655075073242, 0.02604243278503418, 0.026066879272460937, 0.026015743255615235, 0.025941823959350584, 0.02594220733642578, 0.025925247192382813, 0.02600998306274414, 0.025939071655273437, 0.025946847915649413, 0.02593401527404785, 0.02598851203918457, 0.025985599517822266, 0.025977951049804687, 0.026049440383911132, 0.02599344062805176, 0.0259520320892334, 0.02595840072631836, 0.025980928421020507, 0.026025983810424806, 0.02599283218383789, 0.025991552352905272, 0.02876486396789551, 0.02719705581665039, 0.026491167068481446, 0.02612233543395996, 0.02595020866394043, 0.025816127777099608, 0.02579654312133789, 0.02580143928527832, 0.02580918312072754, 0.025915391921997072, 0.0259420166015625, 0.025847808837890625, 0.025870336532592773, 0.026032127380371094, 0.025947456359863282, 0.02591814422607422, 0.025985023498535157, 0.026003456115722655, 0.026003456115722655, 0.02598297691345215, 0.02601913642883301, 0.026014400482177735, 0.026035295486450196, 0.02599193572998047, 0.025992864608764647, 0.025990976333618163, 0.02604307174682617, 0.0261200008392334, 0.02612348747253418, 0.026081632614135743, 0.026092159271240235, 0.025974176406860353, 0.02598358345031738, 0.025944063186645508, 0.025980064392089844, 0.026104543685913088, 0.02615100860595703, 0.02628611183166504, 0.026290176391601562, 0.026286079406738282, 0.026382303237915038, 0.02620419120788574, 0.026185728073120116, 0.026092672348022462, 0.026071935653686523, 0.02606675148010254, 0.026031808853149416, 0.02596284866333008, 0.025989055633544922, 0.025972320556640626, 0.02596518325805664, 0.02607913589477539, 0.02601705551147461, 0.02606572723388672, 0.02604431915283203, 0.026034271240234375, 0.026034175872802736, 0.026039648056030273, 0.026000032424926756, 0.025976192474365233, 0.02603481674194336, 0.02605398368835449, 0.026040992736816405, 0.028721216201782227, 0.02728009605407715, 0.02653343963623047, 0.026174047470092773, 0.0260133113861084, 0.025984895706176757, 0.02608736038208008, 0.025981504440307616, 0.02596246337890625, 0.025935520172119142, 0.025915199279785157, 0.02593235206604004, 0.02587648010253906, 0.02587648010253906, 0.025886463165283202, 0.02588697624206543, 0.025963647842407227, 0.026065792083740234, 0.026093568801879883, 0.026001407623291017, 0.02600281524658203, 0.025979103088378905, 0.025964351654052736, 0.025933792114257812, 0.02591529655456543, 0.025975519180297852, 0.02597385597229004, 0.026028959274291993, 0.025916576385498047, 0.025946975708007813, 0.025991167068481445, 0.026005504608154296, 0.02597887992858887, 0.025993215560913087, 0.02593951988220215, 0.026077152252197266, 0.026150880813598634, 0.0262108154296875, 0.026249216079711913, 0.026177536010742186, 0.026224288940429687, 0.026149215698242186, 0.026133920669555662, 0.026165855407714843, 0.026108928680419922, 0.026063808441162108, 0.02607315254211426, 0.026023935317993165, 0.02612428855895996, 0.026028032302856444, 0.02599283218383789, 0.025988704681396486, 0.025987871170043947, 0.025953535079956055, 0.0260184326171875, 0.026005023956298827, 0.02600601577758789, 0.025964191436767578, 0.026012096405029297, 0.025931392669677734, 0.025976383209228515, 0.026026815414428712, 0.026055904388427736, 0.028876224517822267, 0.02728607940673828, 0.02652569580078125, 0.026202112197875976, 0.026038272857666016, 0.02595430374145508, 0.02586796760559082, 0.025835840225219727, 0.02585545539855957, 0.025881120681762695, 0.025894912719726562, 0.025894912719726562, 0.025899007797241212, 0.02590105628967285, 0.02596646308898926, 0.025991296768188475, 0.026085376739501953, 0.026042015075683593, 0.026011999130249024, 0.026017791748046876, 0.026030080795288086, 0.02597020721435547, 0.026015552520751953, 0.025946144104003907, 0.025936511993408202, 0.02592153549194336, 0.025921503067016603, 0.025968671798706055, 0.026000864028930665, 0.02596713638305664, 0.025989343643188476, 0.026013471603393554, 0.026043743133544923, 0.02600927925109863, 0.026057695388793944, 0.026187776565551758, 0.026181407928466797, 0.026314687728881837, 0.026277952194213867, 0.026250560760498046, 0.026221471786499022, 0.026168352127075196, 0.02612937545776367, 0.02611609649658203, 0.026068992614746093, 0.026027456283569336, 0.026004032135009767, 0.026003456115722655, 0.02594713592529297, 0.025994239807128908, 0.02598428726196289, 0.025968704223632812, 0.026016544342041016, 0.026056575775146484, 0.026013599395751954, 0.026029951095581056, 0.025974176406860353, 0.025983808517456054, 0.02595840072631836, 0.02597478485107422, 0.02592972755432129, 0.02602774429321289, 0.02600281524658203, 0.02850201606750488, 0.027076576232910158, 0.02641836738586426, 0.02610780715942383, 0.025961376190185546, 0.025899007797241212, 0.025946079254150392, 0.0258621768951416, 0.025927328109741212, 0.02588230323791504, 0.025928031921386718, 0.0258460807800293, 0.02587398338317871, 0.02593631935119629, 0.025927263259887694, 0.025907615661621093, 0.025972671508789062, 0.025944128036499023, 0.02588371276855469, 0.025895584106445314, 0.025913568496704103, 0.025920703887939454, 0.02596486473083496, 0.0259303035736084, 0.025927135467529297, 0.026003007888793946, 0.025942943572998048, 0.025972799301147462, 0.025933664321899415, 0.025974399566650392, 0.025893184661865236, 0.025946048736572264, 0.025930015563964844, 0.02597887992858887, 0.025960447311401368, 0.02613043212890625, 0.026251264572143555, 0.026428800582885742, 0.026300863265991212, 0.026275583267211914, 0.026316768646240236, 0.026208736419677733, 0.026206207275390626, 0.02610585594177246, 0.0261345272064209, 0.026045568466186525, 0.02608627128601074, 0.02603385543823242, 0.026048992156982424, 0.025972576141357423, 0.026003456115722655, 0.025949920654296875, 0.026023712158203125, 0.026010112762451174, 0.025990528106689455, 0.02604198455810547, 0.026137599945068358, 0.02606070327758789, 0.026212448120117186, 0.0261345272064209, 0.02610380744934082, 0.02606460762023926, 0.026025888442993163, 0.02860063934326172, 0.02732441520690918, 0.026566656112670898, 0.02617078399658203, 0.026058944702148437, 0.025975072860717773, 0.02592576026916504, 0.025896160125732422, 0.025897760391235352, 0.02592323112487793, 0.02592585563659668, 0.02591663932800293, 0.02589481544494629, 0.025867263793945314, 0.026393695831298827, 0.02590979194641113, 0.026052127838134764, 0.025966880798339843, 0.025952831268310547, 0.025943071365356445, 0.025942527770996093, 0.025907680511474608, 0.025900192260742187, 0.02592425537109375, 0.025961952209472658, 0.026005823135375975, 0.02598134422302246, 0.026062847137451172, 0.02597478485107422, 0.025952255249023438, 0.025953983306884764, 0.025932096481323243, 0.026056320190429687, 0.02603865623474121, 0.026062463760375975, 0.026239360809326172, 0.02632499122619629, 0.02631884765625, 0.02627084732055664, 0.026243967056274416, 0.02625718307495117, 0.02615113639831543, 0.02632499122619629, 0.02611814308166504, 0.026042367935180662, 0.026005504608154296, 0.025989120483398437, 0.025985023498535157, 0.02599692726135254, 0.025992895126342775, 0.02598080062866211, 0.025961280822753906, 0.02595625686645508, 0.02596873664855957, 0.025966207504272462, 0.025940351486206055, 0.026039520263671876, 0.026161472320556642, 0.026014080047607423, 0.025974880218505858, 0.02600873565673828, 0.025959264755249022, 0.025933055877685546]",tokens/s,38.36221067143018,, -float16-flash_attention_2,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,phi3,microsoft/Phi-3-mini-4k-instruct,microsoft/Phi-3-mini-4k-instruct,cuda,0,42,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +float16-flash_attention_2,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,phi3,microsoft/Phi-3-mini-4k-instruct,microsoft/Phi-3-mini-4k-instruct,cuda,0,42,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -24026,9 +24026,9 @@ ChildProcessError: Traceback (most recent call last): return self.pretrained_model.generate(**inputs, **kwargs) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 116, in decorate_context return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2048, in generate + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2047, in generate result = self._sample( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 3008, in _sample + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 3007, in _sample outputs = self(**model_inputs, return_dict=True) File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1553, in _wrapped_call_impl return self._call_impl(*args, **kwargs) @@ -24068,7 +24068,7 @@ RuntimeError: FlashAttention only supports Ampere GPUs or newer. ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, bfloat16-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,phi,microsoft/phi-1_5,microsoft/phi-1_5,cuda,0,42,,,True,True,,bfloat16,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.224-212.876.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.4.0,,4.44.2,,0.34.2,,,,1.22.0,,,,0.12.0,,,MB,883.617792,3447.586816,0.0,3045.064704,2842.846208,s,1,7.31207080078125,7.31207080078125,0.0,7.31207080078125,7.31207080078125,7.31207080078125,7.31207080078125,[7.31207080078125],,kWh,5.85538782503742e-06,6.349197328465118e-07,2.39111302402506e-06,8.881420581908993e-06,,MB,1209.339904,3621.650432,0.0,3208.64256,2982.452736,s,10,2.440714004516602,0.2440714004516602,0.002982791952260248,0.24365643310546875,0.24921473541259764,0.24931415481567382,0.24939369033813474,"[0.244029541015625, 0.2426337890625, 0.24349565124511718, 0.24182231140136717, 0.24381721496582032, 0.24919264221191406, 0.2415570526123047, 0.24941357421875, 0.24507904052734375, 0.23967318725585937]",tokens/s,1048.8734014975357,kWh,7.358720484791849e-06,8.115317455554202e-07,4.871212230300736e-06,1.3041464460648006e-05,tokens/kWh,19629697.322142597,MB,1233.989632,3621.650432,0.0,3208.64256,2982.455296,s,10,12.786667846679686,1.2786667846679687,0.0073945896239175635,1.2776575927734375,1.28396220703125,1.2901543212890625,1.2951080126953125,"[1.2819935302734375, 1.2655325927734375, 1.27572216796875, 1.277006103515625, 1.296346435546875, 1.275377197265625, 1.27872119140625, 1.2750733642578125, 1.282586181640625, 1.27830908203125]",tokens/s,49.27006844582986,kWh,3.69526574972887e-05,4.075608335763504e-06,2.4110227621498836e-05,6.513849345455105e-05,tokens/kWh,967170.0504396354,,s,630,12.784264642715456,0.0202924835598658,0.00045080526153967396,0.020203487396240234,0.020631770133972167,0.02087374200820923,0.02160998861312867,"[0.020982751846313475, 0.02023423957824707, 0.020092287063598634, 0.019978879928588867, 0.019916799545288084, 0.01988540840148926, 0.01989699172973633, 0.019965951919555663, 0.020361215591430663, 0.02041651153564453, 0.020934656143188478, 0.02047590446472168, 0.02044220733642578, 0.02062553596496582, 0.02058540725708008, 0.02046316719055176, 0.02035740852355957, 0.020227392196655272, 0.02021811294555664, 0.020570720672607422, 0.020453216552734375, 0.02016860771179199, 0.019994720458984375, 0.020385791778564453, 0.020458816528320312, 0.020372159957885744, 0.02031167984008789, 0.020148544311523436, 0.020266592025756838, 0.020151039123535157, 0.020106847763061524, 0.020117631912231447, 0.020313631057739256, 0.020314271926879884, 0.02026233673095703, 0.020425312042236327, 0.0206582088470459, 0.0205230712890625, 0.020533632278442383, 0.0206682243347168, 0.020353023529052734, 0.020361215591430663, 0.020475711822509766, 0.020879199981689453, 0.020996448516845703, 0.020502431869506836, 0.020635744094848633, 0.020255903244018554, 0.020257631301879884, 0.02028163146972656, 0.020375263214111327, 0.020709280014038087, 0.020065504074096678, 0.020287904739379883, 0.020060640335083008, 0.0202193603515625, 0.020310207366943358, 0.020451040267944337, 0.020361183166503906, 0.020208288192749023, 0.02018854331970215, 0.020275840759277342, 0.02001299285888672, 0.020916223526000977, 0.02030134391784668, 0.020115392684936523, 0.02011395263671875, 0.019889951705932617, 0.020231967926025392, 0.02049065589904785, 0.020088863372802735, 0.02004582405090332, 0.0200130558013916, 0.0200130558013916, 0.019996768951416017, 0.019955263137817383, 0.020074335098266602, 0.01995008087158203, 0.019912704467773438, 0.02008406448364258, 0.020286111831665038, 0.020258432388305665, 0.020128128051757812, 0.020286815643310547, 0.020178943634033202, 0.020122272491455078, 0.020230144500732423, 0.020022943496704103, 0.019992063522338867, 0.01993404769897461, 0.019953792572021484, 0.020006975173950194, 0.01996988868713379, 0.019986400604248045, 0.020037248611450197, 0.020085119247436525, 0.020121088027954103, 0.02053548812866211, 0.020177215576171876, 0.020066240310668945, 0.019959871292114257, 0.019936864852905273, 0.020007328033447267, 0.01995689582824707, 0.019948320388793947, 0.02015545654296875, 0.020517728805541993, 0.020551488876342772, 0.020304224014282227, 0.02014521598815918, 0.020017951965332032, 0.020016704559326172, 0.01998908805847168, 0.01996575927734375, 0.019939519882202147, 0.019984031677246095, 0.020050207138061524, 0.02000009536743164, 0.01998451232910156, 0.01997494316101074, 0.019973728179931642, 0.01992678451538086, 0.019886272430419922, 0.019846879959106445, 0.01980828857421875, 0.019908672332763673, 0.021132160186767578, 0.02045747184753418, 0.020184703826904297, 0.020181055068969726, 0.01998396873474121, 0.01995859146118164, 0.0199781436920166, 0.02002124786376953, 0.020086687088012697, 0.01998886489868164, 0.020040800094604492, 0.020004575729370117, 0.019964000701904298, 0.020023935317993163, 0.019994016647338866, 0.020021280288696288, 0.020165376663208008, 0.020193119049072266, 0.020201631546020508, 0.020271104812622072, 0.02000249671936035, 0.020097503662109373, 0.0205883846282959, 0.020764671325683593, 0.02040729522705078, 0.020329471588134765, 0.02022809600830078, 0.020503583908081054, 0.02062169647216797, 0.020508928298950194, 0.020296031951904298, 0.02030745506286621, 0.020209632873535156, 0.02010166358947754, 0.0200130558013916, 0.02004172706604004, 0.019957727432250976, 0.020070528030395506, 0.02002092742919922, 0.019974143981933593, 0.019940576553344726, 0.01990777587890625, 0.019955520629882813, 0.01999667167663574, 0.02086092758178711, 0.020211456298828125, 0.020099327087402342, 0.02047590446472168, 0.02032975959777832, 0.02027494430541992, 0.02033148765563965, 0.020262912750244142, 0.020338111877441407, 0.020406848907470704, 0.020504575729370117, 0.020600831985473633, 0.020396032333374024, 0.020428800582885744, 0.020924224853515624, 0.020565696716308594, 0.020291936874389647, 0.02025596809387207, 0.020237119674682617, 0.02106777572631836, 0.020423776626586915, 0.020147104263305664, 0.02004787254333496, 0.020008096694946288, 0.02004774475097656, 0.01998681640625, 0.020015552520751954, 0.019994783401489257, 0.01997772789001465, 0.01996441650390625, 0.02001049613952637, 0.019943967819213867, 0.020012672424316407, 0.02001692771911621, 0.019933759689331056, 0.019998304367065428, 0.020040096282958983, 0.019961856842041017, 0.02022604751586914, 0.020471519470214843, 0.020314592361450196, 0.020164064407348633, 0.020283744812011718, 0.02035526466369629, 0.020221759796142578, 0.020393087387084962, 0.02206604766845703, 0.02020115280151367, 0.020366687774658204, 0.02055471992492676, 0.020384992599487305, 0.020287296295166016, 0.02019740867614746, 0.020147104263305664, 0.020279327392578126, 0.020256736755371093, 0.02009609603881836, 0.020031488418579102, 0.020063167572021486, 0.019990047454833983, 0.019966432571411133, 0.02071286392211914, 0.021479711532592774, 0.020258975982666017, 0.02029689598083496, 0.020214752197265626, 0.020148223876953125, 0.02008678436279297, 0.020358175277709962, 0.020409183502197267, 0.02021798324584961, 0.020176895141601564, 0.020119104385375976, 0.02041427230834961, 0.020441728591918944, 0.02022809600830078, 0.02045238494873047, 0.02030691146850586, 0.020459711074829103, 0.02031977653503418, 0.020200960159301756, 0.02057823944091797, 0.021063455581665037, 0.02113148880004883, 0.027246559143066406, 0.021867839813232422, 0.02084022331237793, 0.021210176467895508, 0.02110246467590332, 0.020604896545410156, 0.02079913520812988, 0.02063132858276367, 0.02062553596496582, 0.02081635284423828, 0.020537343978881836, 0.02056368064880371, 0.02046326446533203, 0.020349567413330077, 0.021174272537231444, 0.020598495483398437, 0.020574495315551757, 0.02050361633300781, 0.02067955207824707, 0.020557119369506837, 0.02055392074584961, 0.020568639755249023, 0.020537567138671876, 0.020448991775512695, 0.020465791702270506, 0.020567392349243162, 0.02037615966796875, 0.02044268798828125, 0.0203917121887207, 0.020506591796875, 0.020310976028442382, 0.020259584426879883, 0.020271871566772463, 0.020371328353881835, 0.02047043228149414, 0.02023516845703125, 0.020177728652954103, 0.020185056686401366, 0.020150304794311524, 0.02016396713256836, 0.02048873519897461, 0.020219135284423827, 0.020462432861328126, 0.020227680206298827, 0.020179359436035157, 0.020135936737060548, 0.02015977668762207, 0.020401887893676758, 0.02065939140319824, 0.020334400177001954, 0.02027248001098633, 0.0202159366607666, 0.020081184387207032, 0.020262655258178712, 0.02027136039733887, 0.020065792083740236, 0.02008291244506836, 0.020070688247680664, 0.02006425666809082, 0.020051424026489257, 0.02002998352050781, 0.021023391723632812, 0.020187231063842775, 0.020169919967651367, 0.020294336318969725, 0.02029737663269043, 0.020353504180908203, 0.02012348747253418, 0.0201395206451416, 0.02008950424194336, 0.0202359676361084, 0.020101184844970702, 0.020070655822753906, 0.02008064079284668, 0.019992576599121094, 0.020017152786254884, 0.019990528106689453, 0.02012508773803711, 0.019964096069335937, 0.020267583847045897, 0.019963935852050783, 0.01994483184814453, 0.019980735778808593, 0.01990800094604492, 0.02000339126586914, 0.01996598434448242, 0.020121055603027342, 0.019964448928833006, 0.020084735870361328, 0.020045440673828126, 0.020588768005371093, 0.020076671600341798, 0.020285472869873047, 0.02030748748779297, 0.020261119842529297, 0.02034627151489258, 0.021084991455078125, 0.020115360260009766, 0.020047584533691407, 0.020160959243774413, 0.020215744018554686, 0.020149696350097657, 0.01990713691711426, 0.019941375732421874, 0.019969152450561523, 0.019968864440917968, 0.01997439956665039, 0.020100896835327148, 0.02045756721496582, 0.020212959289550782, 0.020046464920043944, 0.01995372772216797, 0.020054208755493165, 0.019945280075073242, 0.020899839401245117, 0.020520736694335937, 0.02063587188720703, 0.020347936630249024, 0.020184032440185545, 0.02058393669128418, 0.02015056037902832, 0.02009110450744629, 0.020291040420532227, 0.023632415771484373, 0.0214270076751709, 0.020867071151733398, 0.02068070411682129, 0.02052479934692383, 0.020486400604248046, 0.020463584899902344, 0.02032374382019043, 0.020372159957885744, 0.020260000228881837, 0.020400928497314452, 0.02023401641845703, 0.02006435203552246, 0.02000089645385742, 0.02004377555847168, 0.02007263946533203, 0.02003852844238281, 0.020155616760253906, 0.020145856857299804, 0.020243520736694335, 0.020655359268188477, 0.020233951568603515, 0.020407520294189452, 0.02034899139404297, 0.020192192077636718, 0.02020534324645996, 0.02038374328613281, 0.02044108772277832, 0.02020147132873535, 0.020230144500732423, 0.020262912750244142, 0.0200130558013916, 0.020151424407958984, 0.02042310333251953, 0.02047635269165039, 0.020551679611206054, 0.020414464950561522, 0.020352384567260743, 0.020356832504272462, 0.020296607971191406, 0.020170976638793945, 0.020215583801269532, 0.02007651138305664, 0.020101152420043945, 0.020055328369140625, 0.020275072097778322, 0.02014224052429199, 0.02005276870727539, 0.020147680282592773, 0.020133760452270506, 0.020268831253051758, 0.020084768295288085, 0.020216543197631835, 0.02066640090942383, 0.020508672714233397, 0.020437023162841798, 0.02030940818786621, 0.020183679580688476, 0.02018284797668457, 0.020123424530029296, 0.020416704177856446, 0.02009718322753906, 0.020107295989990233, 0.020141151428222655, 0.021075328826904296, 0.02377996826171875, 0.02055072021484375, 0.020313024520874023, 0.020195232391357423, 0.020172895431518553, 0.020068351745605468, 0.020063711166381837, 0.020046432495117186, 0.019957696914672852, 0.019998720169067383, 0.020041215896606446, 0.020007423400878906, 0.02002934455871582, 0.020156511306762694, 0.020100927352905272, 0.020180608749389647, 0.020122175216674806, 0.019982336044311523, 0.020017152786254884, 0.019967615127563478, 0.01995404815673828, 0.019935199737548828, 0.020025375366210938, 0.019903839111328123, 0.020091552734375, 0.02004355239868164, 0.02006675148010254, 0.020187999725341795, 0.020316352844238283, 0.020236959457397462, 0.020137887954711914, 0.020137887954711914, 0.02016102409362793, 0.02014361572265625, 0.020004159927368165, 0.019989728927612305, 0.020026527404785156, 0.020038047790527345, 0.020037824630737305, 0.020125696182250977, 0.020245824813842773, 0.02012406349182129, 0.02019158363342285, 0.02010675239562988, 0.02003945541381836, 0.020005535125732422, 0.020106847763061524, 0.02013430404663086, 0.02058243179321289, 0.020447200775146484, 0.020166496276855468, 0.020025503158569335, 0.02011484718322754, 0.020748895645141603, 0.020346527099609376, 0.020522464752197267, 0.020404863357543945, 0.020434911727905274, 0.02034502410888672, 0.020230239868164062, 0.02044460868835449, 0.020627071380615234, 0.021202943801879884, 0.020936479568481447, 0.020840543746948242, 0.020803327560424804, 0.02067728042602539, 0.0205166072845459, 0.0205250244140625, 0.020682016372680665, 0.02075721549987793, 0.02067046356201172, 0.020639711380004883, 0.02072719955444336, 0.02056982421875, 0.020560800552368166, 0.02065420722961426, 0.020541311264038086, 0.020527103424072265, 0.02041152000427246, 0.02028163146972656, 0.02024448013305664, 0.020188831329345704, 0.020181791305541992, 0.020277088165283202, 0.020311487197875976, 0.020446304321289063, 0.020566015243530272, 0.020350624084472656, 0.02022617530822754, 0.02027248001098633, 0.02024457550048828, 0.020117919921875, 0.020180255889892577, 0.020185855865478514, 0.020664447784423827, 0.02025817680358887, 0.020121599197387697, 0.020158367156982424, 0.020193056106567384, 0.02029257583618164, 0.020191200256347658, 0.020170751571655272, 0.02001625633239746, 0.020068832397460937, 0.020074111938476562, 0.020062047958374022, 0.02000787162780762, 0.02004319953918457, 0.020127967834472658, 0.02009440040588379, 0.020024223327636717, 0.020026432037353516, 0.02004262351989746, 0.02000230407714844, 0.020126335144042967, 0.020465248107910155, 0.020325727462768554, 0.020388864517211915, 0.02020966339111328, 0.020199392318725588, 0.020137792587280275, 0.020246751785278322, 0.02090291213989258, 0.02038502311706543, 0.020905376434326172, 0.02038435173034668, 0.020370559692382814, 0.02049849510192871, 0.020369792938232423, 0.020213567733764648, 0.02009779167175293, 0.020229440689086914, 0.02012175941467285, 0.02003785514831543, 0.02003740882873535, 0.021663200378417968, 0.02022060775756836, 0.020037887573242187, 0.021786495208740233, 0.02113523292541504, 0.020045440673828126, 0.0200914249420166, 0.019964096069335937, 0.02004777526855469, 0.020084287643432615, 0.021340063095092773, 0.020185728073120118, 0.02003971290588379, 0.020414560317993165, 0.020227872848510742, 0.02014364814758301, 0.020283872604370118, 0.020699136734008788, 0.020131263732910156, 0.02004025650024414, 0.02021171188354492, 0.02021785545349121, 0.020201631546020508, 0.020565887451171876, 0.02030518341064453, 0.0200231990814209, 0.02000694465637207, 0.020024063110351563, 0.02011894416809082, 0.0201582088470459, 0.020400991439819337, 0.020225536346435546, 0.020124160766601562, 0.01999667167663574, 0.020064384460449218, 0.020045183181762696, 0.020015615463256836, 0.020000448226928712, 0.019976608276367186, 0.019938560485839845, 0.0200546875, 0.01992288017272949, 0.020142143249511718, 0.01997558403015137, 0.020073055267333984, 0.020463327407836913, 0.02036147117614746, 0.02017830467224121, 0.02034550476074219, 0.020610464096069335, 0.02069878387451172, 0.02074831962585449]",tokens/s,49.279330302269486,, -float16-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,phi3,microsoft/Phi-3-mini-4k-instruct,microsoft/Phi-3-mini-4k-instruct,cuda,0,42,,,True,True,,float16,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +float16-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,phi3,microsoft/Phi-3-mini-4k-instruct,microsoft/Phi-3-mini-4k-instruct,cuda,0,42,,,True,True,,float16,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -24101,7 +24101,7 @@ ChildProcessError: Traceback (most recent call last): ValueError: Phi3ForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -bfloat16-flash_attention_2,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,phi3,microsoft/Phi-3-mini-4k-instruct,microsoft/Phi-3-mini-4k-instruct,cuda,0,42,,,True,True,,bfloat16,True,False,,flash_attention_2,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +bfloat16-flash_attention_2,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,phi3,microsoft/Phi-3-mini-4k-instruct,microsoft/Phi-3-mini-4k-instruct,cuda,0,42,,,True,True,,bfloat16,True,False,,flash_attention_2,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -24123,9 +24123,9 @@ ChildProcessError: Traceback (most recent call last): return self.pretrained_model.generate(**inputs, **kwargs) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 116, in decorate_context return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2048, in generate + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2047, in generate result = self._sample( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 3008, in _sample + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 3007, in _sample outputs = self(**model_inputs, return_dict=True) File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1553, in _wrapped_call_impl return self._call_impl(*args, **kwargs) @@ -24231,7 +24231,7 @@ RuntimeError: FlashAttention only supports Ampere GPUs or newer. ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, bfloat16-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,microsoft/rho-math-1b-v0.1,microsoft/rho-math-1b-v0.1,cuda,0,42,,,True,True,,bfloat16,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.224-212.876.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.4.0,,4.44.2,,0.34.2,,,,1.22.0,,,,0.12.0,,,MB,868.380672,2698.903552,0.0,2296.38144,2202.20672,s,1,7.29626220703125,7.29626220703125,0.0,7.29626220703125,7.29626220703125,7.29626220703125,7.29626220703125,[7.29626220703125],,kWh,4.935269266585844e-06,5.347519941443314e-07,9.266674079744242e-07,6.396688668704599e-06,,MB,1193.074688,2799.566848,0.0,2386.558976,2267.495936,s,10,2.0040520782470703,0.20040520782470703,0.0010294885050289014,0.20062445068359375,0.2013868698120117,0.20153730697631836,0.20165765670776367,"[0.20110902404785155, 0.2004202880859375, 0.1984439697265625, 0.20005290222167968, 0.201687744140625, 0.1986623077392578, 0.20056195068359375, 0.20107350158691406, 0.20068695068359374, 0.20135343933105468]",tokens/s,1277.4119134864068,kWh,6.072289156546027e-06,6.696652717528203e-07,4.015144935469592e-06,1.075709936376844e-05,tokens/kWh,23798236.99149301,MB,1241.423872,2799.566848,0.0,2386.558976,2267.498496,s,10,13.698938476562502,1.3698938476562499,0.006363824259220363,1.3717947998046875,1.37600439453125,1.37707666015625,1.3779344726562501,"[1.3573558349609376, 1.3711552734375, 1.3737548828125, 1.372434326171875, 1.368803466796875, 1.3665272216796875, 1.3604898681640625, 1.3745025634765624, 1.37814892578125, 1.37576611328125]",tokens/s,45.98896484409113,kWh,3.975353427262645e-05,4.384528206361284e-06,2.324657131833177e-05,6.73846337973195e-05,tokens/kWh,934931.2513813215,,s,630,13.694597980499262,0.021737457111903596,0.0005535640063114479,0.021635440826416018,0.022145173835754393,0.022337561798095703,0.023140980796813967,"[0.022306367874145507, 0.02171900749206543, 0.021530784606933594, 0.02131590461730957, 0.02144041633605957, 0.021298559188842773, 0.021612447738647463, 0.021524639129638673, 0.021922143936157226, 0.022409536361694335, 0.021760000228881835, 0.021547008514404296, 0.02146303939819336, 0.02142617607116699, 0.02171254348754883, 0.02195471954345703, 0.02174380874633789, 0.02152038383483887, 0.02225712013244629, 0.021580320358276367, 0.021462591171264648, 0.02161631965637207, 0.021711231231689453, 0.021430656433105467, 0.021331968307495116, 0.02179657554626465, 0.02137116813659668, 0.021403135299682616, 0.021739839553833008, 0.02177862358093262, 0.021581823348999024, 0.02153267288208008, 0.021751775741577148, 0.021568607330322266, 0.02150495910644531, 0.02138857650756836, 0.021311616897583006, 0.02139401626586914, 0.021387231826782226, 0.021710880279541017, 0.021243648529052736, 0.021408000946044923, 0.021378175735473633, 0.02138368034362793, 0.021259904861450196, 0.021295360565185547, 0.0214021110534668, 0.021329919815063478, 0.02136284828186035, 0.021222623825073242, 0.021398111343383788, 0.021298240661621094, 0.021486560821533204, 0.021458911895751952, 0.02138051223754883, 0.021502080917358397, 0.021473823547363283, 0.021526079177856445, 0.0213221435546875, 0.0212807674407959, 0.021272607803344726, 0.021312704086303712, 0.021961408615112303, 0.022466144561767577, 0.02177235221862793, 0.02162723159790039, 0.02130534362792969, 0.021547008514404296, 0.02181030464172363, 0.021978015899658202, 0.021772415161132812, 0.021910879135131837, 0.021608959197998046, 0.021432319641113282, 0.021423648834228516, 0.02163871955871582, 0.021730207443237306, 0.021555200576782226, 0.02186854362487793, 0.02210736083984375, 0.021886751174926757, 0.023056543350219727, 0.021665824890136718, 0.021941055297851564, 0.021825536727905274, 0.021800512313842772, 0.021629375457763673, 0.021591232299804686, 0.02191257667541504, 0.021804800033569337, 0.02156159973144531, 0.021595808029174806, 0.02194144058227539, 0.021834720611572267, 0.02145894432067871, 0.021428224563598632, 0.021292255401611327, 0.021692928314208985, 0.02165577507019043, 0.02177235221862793, 0.02202115249633789, 0.022037471771240234, 0.021970720291137696, 0.021687999725341797, 0.02162719917297363, 0.02169059181213379, 0.02147270393371582, 0.021647712707519532, 0.021717216491699217, 0.021925888061523437, 0.02209129524230957, 0.02205094337463379, 0.021813600540161134, 0.021802303314208984, 0.022022848129272462, 0.022273279190063475, 0.021821664810180663, 0.02180713653564453, 0.021534751892089844, 0.021552736282348633, 0.021735872268676758, 0.02154489517211914, 0.021551359176635743, 0.021498144149780272, 0.021450719833374023, 0.021561344146728514, 0.02234553527832031, 0.02189091110229492, 0.02148796844482422, 0.021539936065673827, 0.021521312713623047, 0.02126848030090332, 0.02134982490539551, 0.021291616439819337, 0.021337696075439453, 0.02138559913635254, 0.02144588851928711, 0.021295040130615235, 0.021762527465820313, 0.022542688369750978, 0.021557247161865235, 0.021409791946411134, 0.021618688583374023, 0.021315040588378905, 0.021215456008911133, 0.021584192276000978, 0.02145484733581543, 0.021436416625976562, 0.021301248550415038, 0.02126348876953125, 0.021358879089355468, 0.021230079650878905, 0.021330015182495117, 0.02127667236328125, 0.0245614070892334, 0.028383359909057618, 0.02162220764160156, 0.021661535263061523, 0.021277311325073243, 0.021227712631225585, 0.021298912048339842, 0.02129680061340332, 0.021305023193359376, 0.02124652862548828, 0.021247488021850586, 0.021293983459472657, 0.021370880126953123, 0.021149696350097655, 0.02122774314880371, 0.021217056274414062, 0.022576831817626954, 0.028219711303710936, 0.02190729522705078, 0.021446271896362303, 0.021824031829833983, 0.021995519638061522, 0.022181024551391603, 0.021873504638671874, 0.021850112915039063, 0.021640447616577147, 0.02170732879638672, 0.021592256546020507, 0.021493759155273438, 0.021439647674560545, 0.021563776016235353, 0.02187516784667969, 0.021391359329223633, 0.021743616104125976, 0.021855520248413085, 0.02233129692077637, 0.02195465660095215, 0.02192755126953125, 0.022198623657226562, 0.022168991088867187, 0.021897247314453125, 0.022461023330688477, 0.02207744026184082, 0.02199737548828125, 0.02164975929260254, 0.022091583251953126, 0.022200351715087892, 0.022394208908081054, 0.022325536727905274, 0.022130752563476564, 0.02213715171813965, 0.022050592422485353, 0.02189743995666504, 0.021950464248657226, 0.02175699234008789, 0.02178758430480957, 0.021535808563232423, 0.02154787254333496, 0.021588064193725585, 0.02156662368774414, 0.021797407150268556, 0.021521856307983398, 0.021544960021972655, 0.02154537582397461, 0.021535200119018556, 0.021786048889160158, 0.022118976593017578, 0.021737472534179687, 0.021503423690795897, 0.021770816802978515, 0.02210201644897461, 0.021636383056640625, 0.021615327835083006, 0.021743743896484376, 0.021496896743774415, 0.021459199905395507, 0.021432863235473634, 0.021975008010864258, 0.02194643211364746, 0.02166537666320801, 0.021211551666259765, 0.021589471817016603, 0.02142457580566406, 0.021375072479248046, 0.021317632675170898, 0.021405696868896484, 0.02160755157470703, 0.021979007720947265, 0.021914623260498048, 0.02174959945678711, 0.021654943466186523, 0.021588096618652342, 0.021563743591308592, 0.02146291160583496, 0.021483999252319336, 0.02168832015991211, 0.021900896072387696, 0.021606271743774413, 0.022360288619995117, 0.021778911590576173, 0.021675392150878905, 0.02184012794494629, 0.02165705680847168, 0.02179574394226074, 0.022122047424316407, 0.02191404724121094, 0.02189017677307129, 0.021820032119750976, 0.02183193588256836, 0.021523935317993164, 0.021410335540771486, 0.021400768280029295, 0.021438880920410155, 0.02175619125366211, 0.02200992012023926, 0.021882335662841798, 0.022035167694091796, 0.021819263458251952, 0.021635072708129883, 0.02149177551269531, 0.0214138240814209, 0.021382848739624025, 0.021767520904541017, 0.02187772750854492, 0.02163039970397949, 0.02137875175476074, 0.02120921516418457, 0.021354848861694337, 0.021391775131225584, 0.02182691192626953, 0.021844640731811523, 0.021798912048339843, 0.02184601593017578, 0.021790271759033204, 0.021659391403198242, 0.021512895584106444, 0.021464607238769532, 0.021703136444091796, 0.021917695999145507, 0.02185215950012207, 0.02205695915222168, 0.022140256881713866, 0.022043296813964844, 0.021984575271606445, 0.021893760681152345, 0.02212460708618164, 0.022068992614746093, 0.022103391647338867, 0.021922719955444335, 0.021831647872924805, 0.02152556800842285, 0.021455839157104493, 0.021465087890625, 0.0212541446685791, 0.02136457633972168, 0.021386560440063478, 0.021358848571777344, 0.02133462333679199, 0.02146099281311035, 0.02147942352294922, 0.022191551208496092, 0.02474380874633789, 0.021902719497680665, 0.021549312591552735, 0.021454751968383787, 0.021219968795776367, 0.021209087371826172, 0.0212807674407959, 0.02121673583984375, 0.02123119926452637, 0.021280960083007814, 0.021364543914794924, 0.021918527603149413, 0.022114112854003908, 0.022190143585205078, 0.022165760040283203, 0.02214473533630371, 0.022343711853027345, 0.022070623397827147, 0.021793695449829103, 0.021841920852661133, 0.02176323127746582, 0.021480287551879883, 0.021544960021972655, 0.021458080291748047, 0.02140604782104492, 0.02189299201965332, 0.021885568618774415, 0.02155683135986328, 0.021318048477172852, 0.02130240058898926, 0.021572320938110352, 0.02132156753540039, 0.021432096481323243, 0.02167862319946289, 0.022066560745239258, 0.022095903396606446, 0.02225971221923828, 0.021959264755249022, 0.022091167449951172, 0.0219835205078125, 0.021811519622802734, 0.021807104110717773, 0.021761056900024413, 0.021506271362304687, 0.021979551315307617, 0.021448223114013672, 0.021238048553466796, 0.02128700828552246, 0.021330368041992186, 0.021207040786743164, 0.021248191833496095, 0.02134534454345703, 0.02149772834777832, 0.022139263153076173, 0.022063392639160156, 0.021569503784179687, 0.021456256866455078, 0.02135491180419922, 0.02136300849914551, 0.021341503143310545, 0.021439104080200194, 0.021450496673583983, 0.021459199905395507, 0.022214656829833986, 0.02176470375061035, 0.021412191390991212, 0.021395456314086913, 0.021382944107055664, 0.021276895523071288, 0.021346303939819337, 0.021581823348999024, 0.02187884712219238, 0.021669984817504883, 0.02184115219116211, 0.021690975189208983, 0.021661119461059572, 0.021593759536743164, 0.021479520797729492, 0.021436960220336913, 0.021536191940307616, 0.021588191986083985, 0.021554943084716796, 0.021375871658325194, 0.021212575912475586, 0.021337696075439453, 0.021205184936523437, 0.021236223220825197, 0.021336383819580078, 0.021338111877441408, 0.02125993537902832, 0.021348703384399415, 0.021218656539916992, 0.021246463775634765, 0.021349952697753905, 0.02121993637084961, 0.02126950454711914, 0.02126950454711914, 0.02146099281311035, 0.021327871322631836, 0.021526527404785157, 0.021298912048339842, 0.02121084785461426, 0.021344831466674805, 0.021313535690307618, 0.021436063766479493, 0.021475360870361327, 0.02126265525817871, 0.02123980712890625, 0.021345760345458983, 0.0214083194732666, 0.021321088790893554, 0.02129280090332031, 0.02133078384399414, 0.02198294448852539, 0.021886751174926757, 0.02159872055053711, 0.02151379203796387, 0.02169491195678711, 0.02293350410461426, 0.025462783813476563, 0.021987327575683592, 0.022214656829833986, 0.022984256744384767, 0.02165331268310547, 0.02153071975708008, 0.02152911949157715, 0.02307891273498535, 0.022342687606811525, 0.0216844482421875, 0.021717504501342775, 0.02144211196899414, 0.0214715518951416, 0.02143270492553711, 0.02162019157409668, 0.021654048919677735, 0.02147737693786621, 0.021935583114624024, 0.021819456100463867, 0.021586048126220704, 0.021600608825683595, 0.021602304458618164, 0.02173513603210449, 0.021647615432739256, 0.02165353584289551, 0.021599456787109374, 0.02147203254699707, 0.021906911849975588, 0.021651327133178713, 0.021408416748046874, 0.02152454376220703, 0.02199135971069336, 0.02224742317199707, 0.02200371170043945, 0.0220897274017334, 0.022185983657836913, 0.022411264419555665, 0.021950464248657226, 0.02190540885925293, 0.02181033515930176, 0.0217476806640625, 0.021812095642089843, 0.02205244827270508, 0.02181942367553711, 0.021939647674560546, 0.02184828758239746, 0.021932960510253906, 0.02200739288330078, 0.02172336006164551, 0.021582880020141602, 0.021467775344848634, 0.02149193572998047, 0.022532224655151367, 0.021577695846557617, 0.02196646308898926, 0.021467552185058594, 0.021415456771850586, 0.021371360778808593, 0.021426048278808594, 0.0214814395904541, 0.02178268814086914, 0.021995328903198243, 0.022003904342651367, 0.02249241638183594, 0.022057567596435547, 0.02183302307128906, 0.021807199478149415, 0.021918176651000976, 0.021982528686523437, 0.021806047439575194, 0.022248960494995116, 0.021711360931396483, 0.02172854423522949, 0.02184614372253418, 0.021846912384033204, 0.0220218563079834, 0.02206915283203125, 0.021991455078125, 0.022038719177246095, 0.02194371223449707, 0.02206889533996582, 0.02186057662963867, 0.02197715187072754, 0.0220861759185791, 0.022149120330810547, 0.022197248458862305, 0.022047967910766603, 0.02161964797973633, 0.021564128875732422, 0.021586111068725586, 0.021860288619995116, 0.022091392517089845, 0.02140582466125488, 0.021452447891235352, 0.022510175704956056, 0.022302175521850588, 0.021623327255249025, 0.02184752082824707, 0.02182143974304199, 0.023022111892700196, 0.021618688583374023, 0.021590015411376954, 0.021630783081054688, 0.02152899169921875, 0.021657440185546876, 0.02179475212097168, 0.021936031341552736, 0.021416032791137695, 0.02185932731628418, 0.022957056045532227, 0.022255071640014647, 0.021624383926391603, 0.022858976364135742, 0.022269695281982423, 0.02169036865234375, 0.021495712280273437, 0.02141788864135742, 0.02126620864868164, 0.021567903518676757, 0.021446399688720703, 0.02216985511779785, 0.022197343826293944, 0.02223356819152832, 0.021928096771240236, 0.021623071670532228, 0.021393407821655275, 0.02264374351501465, 0.021576671600341796, 0.02149580764770508, 0.021648895263671874, 0.02137548828125, 0.021380287170410156, 0.021593088150024413, 0.02314854431152344, 0.02199923133850098, 0.021637504577636718, 0.021800960540771484, 0.021579296112060546, 0.021461536407470703, 0.021469120025634766, 0.021397632598876955, 0.02146940803527832, 0.0214749755859375, 0.021406944274902345, 0.021459743499755858, 0.021856096267700194, 0.02181727981567383, 0.02158582305908203, 0.021592384338378907, 0.02166783905029297, 0.02202934455871582, 0.02194892883300781, 0.022163423538208007, 0.021989248275756837, 0.02312246322631836, 0.022155616760253908, 0.022482688903808595, 0.021977247238159178, 0.021804895401000977, 0.02172697639465332, 0.021419519424438475, 0.02134448051452637, 0.02142185592651367, 0.02135078430175781, 0.02174310493469238, 0.02153152084350586, 0.022034431457519533, 0.021720544815063476, 0.02155388832092285, 0.02166697692871094, 0.021633855819702147, 0.021921375274658202, 0.02199760055541992, 0.021588191986083985, 0.021608448028564452, 0.022015359878540038, 0.022057600021362304, 0.022384639739990234, 0.02219824028015137, 0.022118080139160157, 0.0222043514251709, 0.022289920806884765, 0.02204355239868164, 0.021903455734252928, 0.02330953598022461, 0.02174857521057129, 0.021647199630737305, 0.02172719955444336, 0.021673215866088866, 0.021655424118041992, 0.02163580894470215, 0.021634815216064453, 0.021623231887817382, 0.021573631286621094, 0.02162073516845703, 0.02155628776550293]",tokens/s,46.003541023774694,, float32-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,microsoft/rho-math-1b-v0.1,microsoft/rho-math-1b-v0.1,cuda,0,42,,,True,True,,float32,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.224-212.876.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.4.0,,4.44.2,,0.34.2,,,,1.22.0,,,,0.12.0,,,MB,867.856384,4804.44416,0.0,4401.922048,4400.206336,s,1,7.37841748046875,7.37841748046875,0.0,7.37841748046875,7.37841748046875,7.37841748046875,7.37841748046875,[7.37841748046875],,kWh,4.949717208334429e-06,5.388015072771084e-07,2.105279461928067e-06,7.5937981775396046e-06,,MB,1183.21152,4903.010304,0.0,4490.002432,4455.927296,s,10,1.586316879272461,0.15863168792724608,0.006245534243113243,0.16039868927001955,0.1633203369140625,0.16408578186035155,0.16469813781738282,"[0.14963296508789062, 0.16276019287109375, 0.16485122680664063, 0.15991238403320313, 0.16315023803710937, 0.15953146362304688, 0.16088499450683594, 0.15961251831054687, 0.14402359008789062, 0.16195730590820312]",tokens/s,1613.801147456808,kWh,4.466229564647132e-06,4.925415332828462e-07,2.973899685515031e-06,7.932670783445009e-06,tokens/kWh,32271602.71598011,MB,1212.035072,4911.398912,0.0,4498.39104,4455.929856,s,10,13.548433959960937,1.3548433959960937,0.0025823938091832036,1.354104553222656,1.3582045776367189,1.358208367919922,1.3582114001464844,"[1.352126220703125, 1.352859130859375, 1.3521011962890626, 1.3529039306640624, 1.352084716796875, 1.35530517578125, 1.3581453857421875, 1.3564923095703125, 1.358212158203125, 1.3582037353515626]",tokens/s,46.499839159404694,kWh,3.974894703535349e-05,4.3836588282026825e-06,2.6239568129688365e-05,7.037217399324453e-05,tokens/kWh,895240.2125028532,,s,630,13.546216690063476,0.02150193125406901,0.00046422688745873707,0.021388095855712892,0.02165515537261963,0.022057723236083984,0.02428231872558594,"[0.024194591522216796, 0.022776447296142577, 0.022095712661743164, 0.021735424041748046, 0.02142207908630371, 0.021370880126953123, 0.021344255447387696, 0.02126028823852539, 0.02131551933288574, 0.021262399673461913, 0.021257951736450197, 0.021223712921142578, 0.02122956848144531, 0.021325151443481447, 0.021456607818603514, 0.021386175155639647, 0.021448415756225588, 0.021450847625732423, 0.02143836784362793, 0.021420320510864257, 0.021376256942749024, 0.021314304351806642, 0.021301248550415038, 0.02132489585876465, 0.021376991271972658, 0.021334976196289063, 0.02134182357788086, 0.021264768600463866, 0.021256223678588867, 0.021348320007324218, 0.021334016799926758, 0.021319679260253906, 0.021403648376464843, 0.021305215835571288, 0.021338336944580077, 0.021364639282226563, 0.02148761558532715, 0.02139094352722168, 0.021408159255981444, 0.021290048599243164, 0.021322015762329102, 0.02133468818664551, 0.02129007911682129, 0.021385951995849608, 0.021509855270385743, 0.021524831771850585, 0.021528703689575195, 0.021604352951049805, 0.021603328704833984, 0.021559776306152342, 0.02149839973449707, 0.02144256019592285, 0.021436063766479493, 0.021379423141479493, 0.021372928619384765, 0.021351423263549805, 0.021318592071533204, 0.021263904571533203, 0.02126902389526367, 0.021820991516113282, 0.021305791854858397, 0.0212541446685791, 0.021243904113769533, 0.024647359848022462, 0.02273516845703125, 0.022038240432739258, 0.021670175552368165, 0.021395231246948244, 0.02131337547302246, 0.021324159622192383, 0.0212807674407959, 0.021307392120361326, 0.021338111877441408, 0.021223039627075196, 0.021168512344360353, 0.02126028823852539, 0.021325567245483398, 0.021350656509399414, 0.021364736557006835, 0.021327840805053712, 0.021283935546875, 0.02137388801574707, 0.021280799865722656, 0.021278688430786133, 0.021336063385009766, 0.021309120178222656, 0.0213602237701416, 0.021539072036743163, 0.02140617561340332, 0.021385215759277345, 0.021346303939819337, 0.021694047927856445, 0.021340576171875, 0.021370271682739257, 0.02139401626586914, 0.021385215759277345, 0.021377279281616212, 0.021323392868041992, 0.02140943908691406, 0.021346784591674803, 0.021333728790283203, 0.021286304473876954, 0.02126646423339844, 0.021299936294555663, 0.02128873634338379, 0.02130575942993164, 0.021278207778930663, 0.021332416534423828, 0.021514240264892577, 0.021550559997558595, 0.021573247909545897, 0.02165443229675293, 0.02160963249206543, 0.02159292793273926, 0.021602304458618164, 0.021544416427612303, 0.0214385929107666, 0.021404064178466797, 0.021432319641113282, 0.02135558319091797, 0.021370847702026366, 0.02138598442077637, 0.02130905532836914, 0.02136841583251953, 0.021519359588623048, 0.021396703720092773, 0.024307296752929686, 0.022756895065307616, 0.022073663711547852, 0.021667776107788087, 0.02151078414916992, 0.021390655517578124, 0.021363456726074218, 0.02130732727050781, 0.021340160369873046, 0.021337568283081056, 0.021340703964233397, 0.02126643180847168, 0.021221023559570312, 0.021280160903930666, 0.02129190444946289, 0.02123958396911621, 0.021430559158325195, 0.021316768646240235, 0.021350400924682617, 0.021286943435668945, 0.021332767486572264, 0.021273887634277344, 0.02128767967224121, 0.021391359329223633, 0.021372415542602538, 0.021388832092285155, 0.021406688690185548, 0.02142416000366211, 0.021442527770996093, 0.02147532844543457, 0.021397407531738282, 0.02133616065979004, 0.021349536895751954, 0.021295839309692383, 0.021313663482666015, 0.02130860710144043, 0.02132009506225586, 0.021383583068847658, 0.02130659294128418, 0.02128156852722168, 0.021344192504882814, 0.02131564712524414, 0.021352447509765626, 0.02159596824645996, 0.021551296234130858, 0.02153267288208008, 0.021594112396240234, 0.021493759155273438, 0.02147327995300293, 0.021458688735961913, 0.02143667221069336, 0.02146873664855957, 0.021397056579589843, 0.02134899139404297, 0.021367040634155274, 0.021292160034179688, 0.021326047897338867, 0.021319520950317382, 0.02130121612548828, 0.021439327239990234, 0.02149580764770508, 0.021389312744140625, 0.02141926383972168, 0.024277183532714845, 0.022794687271118164, 0.02200217628479004, 0.021629951477050782, 0.02135536003112793, 0.021309440612792968, 0.02123366355895996, 0.02115718460083008, 0.02121561622619629, 0.021236032485961915, 0.02125827217102051, 0.021266016006469726, 0.021221248626708985, 0.021312000274658204, 0.021334016799926758, 0.02124185562133789, 0.021313535690307618, 0.02144278335571289, 0.021413759231567384, 0.021319135665893555, 0.02134796714782715, 0.021281600952148438, 0.02128691291809082, 0.021280832290649414, 0.021266368865966796, 0.02126582336425781, 0.02138492774963379, 0.021336959838867187, 0.02133616065979004, 0.021952415466308595, 0.021413888931274414, 0.02137398338317871, 0.02137392044067383, 0.021259584426879884, 0.021300960540771484, 0.021338623046875, 0.021311967849731446, 0.021432319641113282, 0.0213637752532959, 0.021428415298461914, 0.021420799255371093, 0.02138528060913086, 0.021310752868652343, 0.02151081657409668, 0.021606399536132814, 0.021622783660888673, 0.021725183486938478, 0.021617952346801757, 0.02160041618347168, 0.021574207305908203, 0.02147871971130371, 0.021460927963256837, 0.021394176483154295, 0.021399616241455078, 0.021464384078979493, 0.021371519088745117, 0.021440511703491212, 0.021383167266845703, 0.021366464614868165, 0.021446624755859376, 0.021545312881469728, 0.02143846321105957, 0.021456895828247072, 0.02447577667236328, 0.02279654312133789, 0.022013952255249023, 0.021577375411987305, 0.021365087509155275, 0.02129100799560547, 0.021216640472412108, 0.021219968795776367, 0.02131372833251953, 0.021471040725708008, 0.021397504806518555, 0.02130534362792969, 0.02124799919128418, 0.02129715156555176, 0.021331968307495116, 0.021271968841552736, 0.021324384689331056, 0.021301248550415038, 0.021307392120361326, 0.021265472412109375, 0.021263296127319337, 0.021198848724365234, 0.021231327056884765, 0.021301536560058593, 0.02124799919128418, 0.021351903915405274, 0.021360960006713867, 0.021387487411499023, 0.021313535690307618, 0.021327871322631836, 0.021284383773803713, 0.02127244758605957, 0.021396064758300783, 0.021300640106201172, 0.021340639114379882, 0.021940351486206055, 0.02124595260620117, 0.021291263580322267, 0.021282560348510744, 0.02123289680480957, 0.021289312362670898, 0.021372671127319335, 0.0213789119720459, 0.021486400604248047, 0.021639360427856445, 0.021701887130737306, 0.021713184356689452, 0.02158812713623047, 0.021532575607299806, 0.021516511917114258, 0.021372928619384765, 0.021391359329223633, 0.02140176010131836, 0.021344095230102538, 0.021364736557006835, 0.02143846321105957, 0.021437599182128907, 0.021439327239990234, 0.021529823303222655, 0.02145929527282715, 0.021428672790527344, 0.02134809684753418, 0.021340063095092773, 0.024528831481933595, 0.022758975982666015, 0.021981855392456055, 0.021612543106079102, 0.021403648376464843, 0.02131878471374512, 0.021236608505249024, 0.021161983489990235, 0.02124720001220703, 0.021247808456420898, 0.021313791275024415, 0.021295839309692383, 0.021331968307495116, 0.021374496459960937, 0.021391839981079103, 0.0212992000579834, 0.021384544372558593, 0.021646080017089845, 0.021475231170654297, 0.021409120559692383, 0.021418880462646485, 0.0213885440826416, 0.021428096771240236, 0.021299871444702148, 0.021373151779174804, 0.021331743240356447, 0.021397504806518555, 0.02136400032043457, 0.021418720245361327, 0.021384672164916994, 0.021371583938598632, 0.02135024070739746, 0.021884927749633788, 0.021823488235473632, 0.021381023406982422, 0.021345375061035156, 0.02132252883911133, 0.02145699119567871, 0.021305599212646485, 0.021305215835571288, 0.021323776245117186, 0.021352703094482423, 0.02144041633605957, 0.02156937599182129, 0.02152774429321289, 0.021701440811157227, 0.021671392440795897, 0.021772735595703124, 0.02154876708984375, 0.021502336502075194, 0.021501951217651367, 0.021454208374023436, 0.02139743995666504, 0.02135420799255371, 0.02133705520629883, 0.021340160369873046, 0.021352447509765626, 0.021489664077758788, 0.02151219177246094, 0.021497856140136717, 0.021661663055419923, 0.021520416259765626, 0.02147030448913574, 0.024429344177246095, 0.02263039970397949, 0.021925888061523437, 0.021590272903442384, 0.021302719116210938, 0.02135024070739746, 0.021287647247314453, 0.02126972770690918, 0.021367263793945313, 0.02136070442199707, 0.023580671310424805, 0.02127180862426758, 0.021385183334350587, 0.021375392913818358, 0.02140403175354004, 0.021415935516357423, 0.0214234561920166, 0.02134227180480957, 0.02138153648376465, 0.021319711685180664, 0.02135862350463867, 0.021428352355957032, 0.021389312744140625, 0.022313983917236328, 0.021399999618530275, 0.021406272888183593, 0.02139308738708496, 0.02128518486022949, 0.021342144012451172, 0.021364959716796875, 0.021387199401855468, 0.021464384078979493, 0.02137353515625, 0.02130499267578125, 0.021278911590576172, 0.021325983047485352, 0.021270431518554688, 0.021380672454833983, 0.021445152282714843, 0.021427679061889648, 0.021457439422607423, 0.021438175201416016, 0.02139779281616211, 0.02156879997253418, 0.02156822395324707, 0.02162483215332031, 0.02167193603515625, 0.021529951095581056, 0.02146371269226074, 0.021421503067016602, 0.021369407653808594, 0.02143846321105957, 0.02142207908630371, 0.02142361640930176, 0.021465599060058595, 0.021425600051879882, 0.02150662422180176, 0.02145894432067871, 0.02302566337585449, 0.021336063385009766, 0.021360832214355467, 0.02141779136657715, 0.021387264251708983, 0.02417203140258789, 0.02266160011291504, 0.021930240631103517, 0.02156723213195801, 0.021487360000610353, 0.021309696197509765, 0.021311487197875977, 0.02131283187866211, 0.021318336486816407, 0.021288095474243166, 0.021313791275024415, 0.021209695816040038, 0.021210304260253908, 0.021326528549194337, 0.02130748748779297, 0.021466144561767576, 0.021504735946655272, 0.0213404483795166, 0.021350400924682617, 0.021323776245117186, 0.02135180854797363, 0.021327871322631836, 0.02135308837890625, 0.021340160369873046, 0.02141168022155762, 0.021440671920776366, 0.021413888931274414, 0.021364736557006835, 0.021405696868896484, 0.021411264419555664, 0.0214451847076416, 0.021376575469970703, 0.02209049606323242, 0.021399072647094727, 0.02139561653137207, 0.021411327362060546, 0.0232260799407959, 0.021403520584106446, 0.021351327896118166, 0.02133795166015625, 0.021508256912231447, 0.021458112716674804, 0.021445568084716798, 0.02153868865966797, 0.021601280212402343, 0.021627904891967774, 0.021532800674438475, 0.021458816528320313, 0.021482816696166994, 0.021456895828247072, 0.0214368953704834, 0.021467424392700194, 0.02152342414855957, 0.02146575927734375, 0.021444223403930665, 0.02145964813232422, 0.021481472015380858, 0.021419776916503906, 0.021467103958129882, 0.021524768829345703, 0.021511327743530272, 0.021514272689819335, 0.021457727432250977, 0.024707071304321288, 0.02294988822937012, 0.022079488754272462, 0.021786624908447266, 0.021474815368652343, 0.02143449592590332, 0.021348735809326173, 0.021237312316894533, 0.02127097511291504, 0.021245407104492187, 0.021285280227661133, 0.02121513557434082, 0.021247648239135743, 0.021348928451538084, 0.02331443214416504, 0.02125823974609375, 0.021394975662231447, 0.021373088836669923, 0.021426496505737306, 0.021407743453979493, 0.021358591079711914, 0.021275999069213868, 0.021338783264160156, 0.02132086372375488, 0.02132259178161621, 0.021374624252319337, 0.021315200805664063, 0.021295455932617186, 0.021338687896728516, 0.021327680587768554, 0.021331968307495116, 0.021376991271972658, 0.021358335494995117, 0.021379072189331053, 0.021474687576293946, 0.021449535369873048, 0.02146108818054199, 0.02160367965698242, 0.021432607650756837, 0.021424448013305664, 0.021460447311401367, 0.021440576553344727, 0.02138764762878418, 0.02145430374145508, 0.021482175827026367, 0.021541887283325196, 0.02147020721435547, 0.021436416625976562, 0.021474687576293946, 0.021520896911621092, 0.02149510383605957, 0.021541696548461914, 0.021499168395996093, 0.021490400314331054, 0.021438144683837892, 0.021570943832397462, 0.021674848556518553, 0.0215184326171875, 0.02147737693786621, 0.021417152404785155, 0.02137376022338867, 0.021371999740600587, 0.023073631286621092, 0.02428441619873047, 0.02291551971435547, 0.022102304458618164, 0.021739519119262696, 0.021399551391601563, 0.02130534362792969, 0.0212992000579834, 0.02128895950317383, 0.021518335342407227, 0.021346303939819337, 0.021347904205322267, 0.021208576202392578, 0.021291967391967773, 0.021409631729125977, 0.021557407379150392, 0.02151628875732422, 0.021546304702758787, 0.021535423278808592, 0.02151628875732422, 0.02150399971008301, 0.021486656188964844, 0.021440671920776366, 0.021467935562133788, 0.021381120681762695, 0.02135481643676758, 0.021345983505249022, 0.021366783142089844, 0.021311487197875977, 0.021368032455444337, 0.021336095809936523, 0.02135865592956543, 0.02138591957092285, 0.021429664611816408, 0.021367456436157225, 0.021514175415039062, 0.021393407821655275, 0.021444255828857423, 0.021399904251098632, 0.021295103073120117, 0.021364736557006835, 0.022722560882568358, 0.021329631805419923, 0.021341663360595703, 0.02160867118835449, 0.02176880073547363, 0.021819391250610352, 0.021833471298217774, 0.021735679626464843, 0.02166374397277832, 0.021647647857666017, 0.021564992904663086, 0.021516447067260743, 0.02149760055541992, 0.021381471633911135, 0.02148521614074707, 0.02141209602355957, 0.021405376434326173, 0.021456480026245117, 0.021471296310424805, 0.021475391387939455, 0.021494367599487304, 0.021450687408447265, 0.021448640823364257]",tokens/s,46.50745033940896,, -float32-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,phi3,microsoft/Phi-3-mini-4k-instruct,microsoft/Phi-3-mini-4k-instruct,cuda,0,42,,,True,True,,float32,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +float32-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,phi3,microsoft/Phi-3-mini-4k-instruct,microsoft/Phi-3-mini-4k-instruct,cuda,0,42,,,True,True,,float32,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -24266,7 +24266,7 @@ ValueError: Phi3ForCausalLM does not support an attention implementation through ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, bfloat16-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,phi,microsoft/phi-1_5,microsoft/phi-1_5,cuda,0,42,,,True,True,,bfloat16,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.224-212.876.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.4.0,,4.44.2,,0.34.2,,,,1.22.0,,,,0.12.0,,,MB,882.167808,3447.586816,0.0,3045.064704,2842.846208,s,1,7.24684375,7.24684375,0.0,7.24684375,7.24684375,7.24684375,7.24684375,[7.24684375],,kWh,5.644512066601237e-06,6.147419892112981e-07,2.052779420003148e-06,8.312033475815684e-06,,MB,1214.705664,3642.621952,0.0,3229.61408,2982.452736,s,10,2.4826019592285156,0.24826019592285156,0.002192780847353011,0.24785032653808592,0.2508522247314453,0.2511474411010742,0.25138361419677735,"[0.2514426574707031, 0.24397113037109375, 0.24741127014160155, 0.24952365112304686, 0.24706668090820313, 0.24812602233886719, 0.24757463073730468, 0.25078662109375, 0.24623654174804688, 0.25046275329589845]",tokens/s,1031.1761780754964,kWh,7.483775386113211e-06,8.2512768431257e-07,4.934412779435162e-06,1.3243315849860945e-05,tokens/kWh,19330506.264614087,MB,1246.162944,3642.621952,0.0,3229.61408,2982.455296,s,10,13.804732666015628,1.3804732666015627,0.0049082723871483,1.380882568359375,1.3869291381835938,1.3869294128417968,1.3869296325683593,"[1.38528125, 1.38108740234375, 1.3779637451171876, 1.3869296875, 1.377422119140625, 1.382515380859375, 1.37380908203125, 1.380677734375, 1.3721171875, 1.3869290771484375]",tokens/s,45.63652301293227,kWh,4.045775356389158e-05,4.4616313161530956e-06,2.506975011276471e-05,6.998913499280938e-05,tokens/kWh,900139.7146353154,,s,630,13.802404392242446,0.021908578400384814,0.00034643937838156554,0.02183065605163574,0.022272461509704588,0.02253029909133911,0.02307008056640625,"[0.02304614448547363, 0.02224675178527832, 0.021957279205322266, 0.021690143585205077, 0.02171107292175293, 0.021901344299316405, 0.02193404769897461, 0.02188822364807129, 0.02205731201171875, 0.022004159927368164, 0.02259347152709961, 0.021940288543701173, 0.022188032150268554, 0.02212646484375, 0.02202432060241699, 0.02205695915222168, 0.02205504035949707, 0.02204863929748535, 0.021816896438598632, 0.02182793617248535, 0.021929248809814453, 0.021917728424072264, 0.022003583908081055, 0.021841920852661133, 0.021795839309692384, 0.022104127883911133, 0.02182905578613281, 0.02173993682861328, 0.02182963180541992, 0.02191564750671387, 0.021989376068115234, 0.02175200080871582, 0.021966751098632813, 0.021798112869262695, 0.02174460792541504, 0.02217897605895996, 0.022389312744140626, 0.022257600784301758, 0.022265920639038084, 0.022673408508300782, 0.021925439834594728, 0.022176191329956053, 0.021819391250610352, 0.02188467216491699, 0.022225183486938478, 0.02197859191894531, 0.022061567306518554, 0.02184601593017578, 0.021838943481445314, 0.021626911163330077, 0.021697439193725587, 0.022218719482421875, 0.021813247680664064, 0.021735424041748046, 0.02253932762145996, 0.021812095642089843, 0.021696575164794923, 0.021780479431152345, 0.021972991943359374, 0.021719039916992186, 0.021795040130615236, 0.021759775161743163, 0.022112255096435548, 0.02272096061706543, 0.022237184524536133, 0.022009855270385743, 0.021992927551269532, 0.021776927947998046, 0.021720991134643555, 0.02173753547668457, 0.02188496017456055, 0.02187059211730957, 0.02230681610107422, 0.022050207138061523, 0.021967456817626952, 0.021735424041748046, 0.021882720947265625, 0.021724735260009766, 0.021752351760864257, 0.02228540802001953, 0.022037408828735353, 0.021913663864135742, 0.021872352600097657, 0.02169481658935547, 0.02174355125427246, 0.021956192016601563, 0.021820959091186524, 0.02184281539916992, 0.021682079315185548, 0.022349727630615233, 0.021885120391845703, 0.021720800399780273, 0.021516576766967773, 0.02156572723388672, 0.02155254364013672, 0.02169068717956543, 0.022001663208007814, 0.021850112915039063, 0.021975040435791016, 0.022008960723876952, 0.022176639556884765, 0.023216127395629883, 0.021958656311035156, 0.021839616775512695, 0.021692352294921877, 0.021870784759521485, 0.02171696090698242, 0.02171116828918457, 0.02172502326965332, 0.021630975723266603, 0.021981184005737304, 0.022159263610839842, 0.021954208374023437, 0.021760448455810547, 0.021712480545043947, 0.02170921516418457, 0.02165760040283203, 0.02156662368774414, 0.02174038314819336, 0.021932031631469725, 0.021835775375366212, 0.02225302314758301, 0.02211484718322754, 0.022210559844970702, 0.02207744026184082, 0.022317247390747072, 0.02274643135070801, 0.02187593650817871, 0.021654367446899414, 0.021657920837402343, 0.021940832138061524, 0.02178428840637207, 0.021690656661987304, 0.0217262077331543, 0.02163587188720703, 0.021729375839233397, 0.021673215866088866, 0.021880992889404295, 0.022251808166503906, 0.022757823944091798, 0.022304767608642577, 0.022204416275024414, 0.022055072784423826, 0.022060895919799806, 0.021917119979858398, 0.021882463455200195, 0.021906400680541994, 0.021825536727905274, 0.02179190444946289, 0.021703519821166993, 0.021632896423339844, 0.021694080352783203, 0.021686784744262694, 0.021636159896850585, 0.021863359451293946, 0.021626880645751953, 0.022015520095825195, 0.0216396484375, 0.02179804801940918, 0.021769056320190428, 0.022519264221191406, 0.022335296630859376, 0.022110944747924806, 0.021993471145629884, 0.021879871368408202, 0.021812255859375, 0.022261663436889647, 0.022091455459594726, 0.022052896499633788, 0.022157407760620116, 0.022018304824829103, 0.021688255310058593, 0.021755903244018555, 0.021735424041748046, 0.02155500793457031, 0.02153696060180664, 0.021551103591918946, 0.021524383544921876, 0.021624927520751954, 0.02169856071472168, 0.021689823150634766, 0.021678112030029298, 0.02153487968444824, 0.02150982475280762, 0.021544704437255858, 0.021688352584838866, 0.021600128173828125, 0.022944768905639647, 0.021651456832885742, 0.022834495544433595, 0.022852800369262696, 0.02264031982421875, 0.02214512062072754, 0.022131744384765624, 0.022128864288330077, 0.022456544876098633, 0.021878368377685548, 0.02185116767883301, 0.021769983291625977, 0.02170675277709961, 0.021639167785644533, 0.02155731201171875, 0.021798336029052734, 0.02172800064086914, 0.021612287521362305, 0.021569696426391602, 0.021706367492675783, 0.021520608901977538, 0.021530624389648437, 0.021614591598510743, 0.021769567489624022, 0.0218438720703125, 0.021496576309204103, 0.02150399971008301, 0.02164240074157715, 0.021822303771972657, 0.021927871704101563, 0.021876735687255858, 0.021741632461547852, 0.021833728790283204, 0.021802207946777345, 0.02162892723083496, 0.02156787109375, 0.021737728118896484, 0.021993631362915038, 0.022261760711669923, 0.022130687713623046, 0.02207744026184082, 0.02186854362487793, 0.021954336166381837, 0.022498592376708985, 0.024343488693237304, 0.022164960861206055, 0.022060831069946288, 0.022004032135009767, 0.02202454376220703, 0.02186835289001465, 0.02190979194641113, 0.02192793655395508, 0.021934112548828124, 0.02174527931213379, 0.022274303436279296, 0.0217326717376709, 0.021664447784423828, 0.021680383682250975, 0.024119136810302734, 0.023140031814575194, 0.022020511627197266, 0.022005279541015624, 0.02228006362915039, 0.021973503112792968, 0.022089248657226564, 0.022755136489868166, 0.021909311294555665, 0.02177471923828125, 0.02166988754272461, 0.02162387275695801, 0.021634176254272462, 0.02160211181640625, 0.021596000671386718, 0.021784736633300782, 0.021753759384155275, 0.02164892768859863, 0.02168275260925293, 0.02170038414001465, 0.023070655822753906, 0.02246028709411621, 0.022497695922851564, 0.021871904373168945, 0.022514432907104493, 0.022699712753295898, 0.022265504837036133, 0.02215385627746582, 0.021808992385864256, 0.022001823425292968, 0.02229043197631836, 0.02214297676086426, 0.021886592864990236, 0.021821823120117188, 0.022668607711791994, 0.021746368408203126, 0.02154035186767578, 0.02162451171875, 0.021823488235473632, 0.021658048629760743, 0.02159654426574707, 0.021607936859130858, 0.021570047378540038, 0.021656736373901368, 0.021629535675048828, 0.02165171241760254, 0.02147942352294922, 0.02148259162902832, 0.021578655242919922, 0.02166579246520996, 0.021547008514404296, 0.021552160263061525, 0.021685216903686525, 0.02228540802001953, 0.022492063522338866, 0.02229043197631836, 0.02189468765258789, 0.02165193557739258, 0.02164240074157715, 0.021590591430664063, 0.021868799209594728, 0.021847648620605467, 0.02163142395019531, 0.02163302421569824, 0.02155465507507324, 0.02162508773803711, 0.02153091239929199, 0.02164124870300293, 0.021690336227416993, 0.02190540885925293, 0.022818784713745117, 0.021953727722167967, 0.021650272369384764, 0.021989376068115234, 0.02162646484375, 0.021620351791381835, 0.02163167953491211, 0.021826656341552734, 0.02184294319152832, 0.021784576416015625, 0.021711872100830077, 0.02187775993347168, 0.02199113655090332, 0.021944608688354492, 0.021768415451049804, 0.021655328750610353, 0.021620512008666992, 0.021749984741210936, 0.021921472549438478, 0.0219703369140625, 0.021781408309936523, 0.021978303909301757, 0.021768415451049804, 0.02178441619873047, 0.021756256103515625, 0.021901727676391602, 0.02164735984802246, 0.021679584503173827, 0.02238108825683594, 0.023706911087036132, 0.022448863983154297, 0.022149120330810547, 0.021876735687255858, 0.021790719985961913, 0.02173651123046875, 0.021992063522338866, 0.021868640899658204, 0.02223535919189453, 0.022245376586914063, 0.0225599365234375, 0.02225040054321289, 0.022279775619506836, 0.022771839141845704, 0.022192319869995116, 0.02231705665588379, 0.021931039810180665, 0.02174665641784668, 0.021643264770507813, 0.021807104110717773, 0.021761407852172853, 0.021741247177124022, 0.021905887603759767, 0.021960384368896486, 0.02166147232055664, 0.02169139289855957, 0.021673824310302733, 0.021622943878173827, 0.021731327056884766, 0.021659231185913085, 0.021747648239135744, 0.021854240417480467, 0.02196659278869629, 0.02213270378112793, 0.022827775955200195, 0.021985279083251954, 0.021767711639404295, 0.021674463272094727, 0.02171708869934082, 0.021761247634887695, 0.02163167953491211, 0.021702655792236326, 0.021722431182861327, 0.021893375396728514, 0.02204204750061035, 0.02185113525390625, 0.021841920852661133, 0.021809152603149414, 0.021737472534179687, 0.02203647994995117, 0.021823488235473632, 0.021685983657836912, 0.021659936904907227, 0.02161257553100586, 0.022007776260375977, 0.021639167785644533, 0.021723392486572266, 0.02157481575012207, 0.021604959487915038, 0.022193567276000976, 0.021673952102661133, 0.021654144287109375, 0.021589088439941406, 0.021699296951293946, 0.021604543685913087, 0.021581823348999024, 0.021964799880981444, 0.02204876708984375, 0.021745664596557617, 0.021999616622924805, 0.022006944656372072, 0.021848928451538085, 0.021809152603149414, 0.021561344146728514, 0.02163520050048828, 0.021614463806152343, 0.021706655502319337, 0.02170899200439453, 0.02164521598815918, 0.02228428840637207, 0.022040576934814454, 0.021960704803466798, 0.021716352462768554, 0.021670528411865234, 0.021659648895263672, 0.02163711929321289, 0.02276710319519043, 0.021977664947509766, 0.02183363151550293, 0.02171404838562012, 0.021742496490478515, 0.02167193603515625, 0.02166374397277832, 0.021534719467163087, 0.02167136001586914, 0.021743167877197267, 0.02168115234375, 0.023068672180175782, 0.022401023864746093, 0.0221265926361084, 0.021991424560546875, 0.021796031951904295, 0.021907808303833008, 0.021967327117919922, 0.0219238395690918, 0.022003616333007812, 0.02213657569885254, 0.021967199325561522, 0.02190947151184082, 0.021723007202148436, 0.022050559997558592, 0.022004127502441406, 0.0218787841796875, 0.021727231979370116, 0.02172083282470703, 0.021632543563842775, 0.02196672058105469, 0.022272256851196288, 0.022070880889892577, 0.021905792236328124, 0.021797504425048828, 0.021841920852661133, 0.021929344177246093, 0.02203059196472168, 0.021823295593261717, 0.021964607238769532, 0.021779455184936524, 0.021890815734863282, 0.021893119812011717, 0.021995519638061522, 0.02213395118713379, 0.02199852752685547, 0.022021312713623047, 0.022024288177490234, 0.02200432014465332, 0.021763071060180664, 0.021732351303100587, 0.021798015594482422, 0.021947456359863282, 0.021845983505249023, 0.0219769287109375, 0.02186649513244629, 0.021995008468627928, 0.021869056701660155, 0.021932031631469725, 0.021651456832885742, 0.021790271759033204, 0.02162940788269043, 0.02176406478881836, 0.021776384353637695, 0.02188697624206543, 0.02183590316772461, 0.021726560592651368, 0.02172982406616211, 0.02171664047241211, 0.02169878387451172, 0.021895456314086913, 0.02180284881591797, 0.021831680297851562, 0.02170163154602051, 0.022518400192260743, 0.021986528396606444, 0.021803871154785156, 0.021673919677734375, 0.022034143447875975, 0.021737823486328123, 0.021712608337402343, 0.02187411117553711, 0.021744480133056642, 0.022146335601806642, 0.021643999099731446, 0.02168012809753418, 0.02163711929321289, 0.021790719985961913, 0.021573663711547852, 0.0215982723236084, 0.021563232421875, 0.021622175216674804, 0.02177289581298828, 0.02156755256652832, 0.02160207939147949, 0.021729503631591797, 0.021688255310058593, 0.02155107116699219, 0.021593183517456056, 0.021541023254394533, 0.021668256759643553, 0.021801408767700196, 0.021667264938354493, 0.021568031311035157, 0.021567039489746094, 0.021722848892211915, 0.021803359985351562, 0.02168809509277344, 0.02163702392578125, 0.021668256759643553, 0.021766143798828123, 0.02157513618469238, 0.02167612838745117, 0.021708959579467772, 0.021690975189208983, 0.021624448776245118, 0.021602624893188475, 0.0215982723236084, 0.021612543106079102, 0.021614591598510743, 0.021676191329956053, 0.02180816078186035, 0.022106752395629883, 0.021810400009155274, 0.022245376586914063, 0.02241139221191406, 0.0218955192565918, 0.021875200271606447, 0.02187264060974121, 0.022680704116821288, 0.021978271484375, 0.02204617691040039, 0.021942176818847657, 0.02171120071411133, 0.02167193603515625, 0.02159587287902832, 0.02192207908630371, 0.022980159759521484, 0.02205904006958008, 0.02196067237854004, 0.022135295867919923, 0.021858688354492187, 0.02189936065673828, 0.021837535858154296, 0.021819391250610352, 0.021902496337890626, 0.022005727767944336, 0.02219664001464844, 0.022497087478637694, 0.022694623947143555, 0.022318143844604493, 0.021982303619384767, 0.022077215194702147, 0.02208483123779297, 0.02207619285583496, 0.021952512741088868, 0.02179478454589844, 0.02186038398742676, 0.022364160537719727, 0.02243715286254883, 0.022094560623168946, 0.021942272186279296, 0.021800703048706054, 0.021862655639648437, 0.021812799453735352, 0.02194272041320801, 0.022132736206054687, 0.021937503814697265, 0.02193020820617676, 0.02203481674194336, 0.02192595291137695, 0.02204364776611328, 0.021934207916259767, 0.022097951889038087, 0.021861215591430665, 0.0218603515625, 0.021753856658935547, 0.021743616104125976, 0.02191564750671387, 0.021751392364501954, 0.021746240615844726, 0.022605663299560548, 0.02174332809448242, 0.021616928100585936, 0.02161039924621582, 0.021632928848266602, 0.021670080184936522, 0.02160576057434082, 0.021768287658691408, 0.024789535522460937, 0.022837247848510742, 0.022003103256225585, 0.02190332794189453, 0.021883455276489258, 0.021804607391357422, 0.021889503479003907, 0.021634111404418944, 0.021553951263427733, 0.02175814437866211, 0.02146303939819336]",tokens/s,45.64422126003554,, float16-sdpa,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,llama,microsoft/rho-math-1b-v0.1,microsoft/rho-math-1b-v0.1,cuda,0,42,,,True,True,,float16,True,False,,sdpa,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.224-212.876.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.4.0,,4.44.2,,0.34.2,,,,1.22.0,,,,0.12.0,,,MB,867.708928,2698.903552,0.0,2296.38144,2202.20672,s,1,7.20566845703125,7.20566845703125,0.0,7.20566845703125,7.20566845703125,7.20566845703125,7.20566845703125,[7.20566845703125],,kWh,4.989219870837284e-06,5.342583400424964e-07,1.901668187975769e-06,7.42514639885555e-06,,MB,1305.509888,2797.469696,0.0,2382.364672,2267.889152,s,10,0.2868409271240234,0.028684092712402343,0.001110909784972339,0.028358511924743653,0.0290583740234375,0.030486594963073727,0.03162917171478272,"[0.03191481590270996, 0.027990751266479492, 0.028483007431030275, 0.028036703109741212, 0.028185855865478514, 0.028639039993286132, 0.02797305679321289, 0.028642688751220703, 0.028740991592407228, 0.02823401641845703]",tokens/s,8924.807298831225,kWh,1.0150304529369058e-06,1.1194008284612027e-07,6.682933587084322e-07,1.7952638944914586e-06,tokens/kWh,142597420.23749477,MB,1349.44768,2797.469696,0.0,2382.364672,2267.891712,s,10,12.7222490234375,1.27222490234375,0.010027180159853778,1.2703740234375,1.2882837646484375,1.2893047241210938,1.2901214916992187,"[1.288056884765625, 1.26739697265625, 1.27260009765625, 1.266229248046875, 1.2698204345703126, 1.266865478515625, 1.2543466796875, 1.275679931640625, 1.29032568359375, 1.2709276123046875]",tokens/s,49.51954633488039,kWh,3.6564277659565476e-05,4.032167338520448e-06,2.1983446984890415e-05,6.257989198297633e-05,tokens/kWh,1006713.147046306,,s,630,12.717896696090696,0.02018713761284238,0.0005291124876017078,0.020095312118530273,0.020585785484313963,0.020703671646118166,0.02270597204208375,"[0.019918495178222657, 0.019978591918945313, 0.020141759872436524, 0.01979132843017578, 0.019795808792114258, 0.019969247817993165, 0.020289312362670897, 0.02066419219970703, 0.02059222412109375, 0.02094038391113281, 0.020587455749511718, 0.02032640075683594, 0.02027231979370117, 0.02055865669250488, 0.02003068733215332, 0.020005664825439452, 0.020227167129516603, 0.02021228790283203, 0.020341087341308593, 0.019969120025634765, 0.019816511154174803, 0.019997535705566408, 0.020199424743652345, 0.020452959060668945, 0.02044963264465332, 0.020377119064331053, 0.020418079376220703, 0.020564992904663085, 0.02347430419921875, 0.0210512638092041, 0.0208175048828125, 0.020894176483154298, 0.020545215606689454, 0.0204736328125, 0.02055740737915039, 0.020435840606689452, 0.020407487869262695, 0.020138816833496095, 0.020049823760986327, 0.020183456420898437, 0.020184320449829103, 0.020312511444091796, 0.020262912750244142, 0.02038991928100586, 0.024432607650756836, 0.020479583740234376, 0.020220319747924806, 0.020031488418579102, 0.020060447692871092, 0.02021347236633301, 0.0205614070892334, 0.020676704406738283, 0.020844959259033204, 0.02051215934753418, 0.020464223861694338, 0.02033459281921387, 0.020129791259765627, 0.02028351974487305, 0.020350303649902344, 0.020523168563842772, 0.020238719940185546, 0.019990495681762695, 0.02022198486328125, 0.01972831916809082, 0.02001113510131836, 0.020182720184326174, 0.019792160034179686, 0.019790143966674806, 0.0196997127532959, 0.02026678466796875, 0.019965375900268555, 0.019877664566040038, 0.020015871047973633, 0.020285696029663087, 0.020668415069580077, 0.020527103424072265, 0.02247443199157715, 0.021506368637084963, 0.021065216064453125, 0.020554208755493165, 0.02037763214111328, 0.02097267150878906, 0.020284095764160157, 0.020015296936035157, 0.02006412887573242, 0.020017248153686523, 0.020043807983398436, 0.01990355110168457, 0.019851200103759764, 0.019903488159179687, 0.019759103775024413, 0.01999612808227539, 0.019878431320190428, 0.019828351974487304, 0.019819904327392578, 0.019845727920532227, 0.019853216171264648, 0.019873855590820312, 0.020404672622680663, 0.020604991912841798, 0.02058559989929199, 0.020477792739868165, 0.020416671752929688, 0.02040291213989258, 0.019955360412597656, 0.01993494415283203, 0.01989651107788086, 0.019824384689331054, 0.01984182357788086, 0.02005401611328125, 0.020139392852783204, 0.01983875274658203, 0.01990934371948242, 0.019908031463623046, 0.019947488784790038, 0.019997184753417968, 0.01985763168334961, 0.019800064086914062, 0.01986528015136719, 0.019920991897583007, 0.01988764762878418, 0.01979871940612793, 0.019814016342163086, 0.019700096130371093, 0.019779552459716798, 0.01977961540222168, 0.019700319290161132, 0.019755008697509766, 0.020463615417480468, 0.020231744766235352, 0.019849664688110353, 0.01986774444580078, 0.0202589111328125, 0.02032211112976074, 0.02049228858947754, 0.020476032257080078, 0.020352895736694337, 0.020302911758422852, 0.019884544372558592, 0.019757503509521483, 0.019916160583496094, 0.01980847930908203, 0.01988444709777832, 0.020028831481933594, 0.020044384002685548, 0.019860576629638672, 0.01976825523376465, 0.020381664276123045, 0.019959007263183594, 0.020169504165649416, 0.020489599227905272, 0.020244768142700195, 0.01995974349975586, 0.020021663665771485, 0.020208639144897463, 0.020050912857055663, 0.019912736892700195, 0.020477888107299804, 0.020404287338256836, 0.02090188789367676, 0.020637439727783202, 0.020615423202514648, 0.020578176498413085, 0.020424928665161133, 0.02041436767578125, 0.020363359451293944, 0.020195232391357423, 0.020118879318237304, 0.020058528900146484, 0.020099327087402342, 0.02015817642211914, 0.020141376495361327, 0.020085151672363282, 0.020015520095825197, 0.020010751724243166, 0.020004575729370117, 0.019847551345825196, 0.019964223861694337, 0.020104448318481447, 0.020431615829467772, 0.020233983993530272, 0.02022985649108887, 0.020415008544921873, 0.020407712936401368, 0.020433504104614256, 0.02036735916137695, 0.020389888763427736, 0.020422815322875976, 0.020615039825439455, 0.020813823699951172, 0.020342784881591795, 0.02002943992614746, 0.020012447357177734, 0.020306528091430662, 0.020279296875, 0.02050662422180176, 0.02019327926635742, 0.020190431594848634, 0.02031612777709961, 0.02011628723144531, 0.020170816421508787, 0.020072383880615233, 0.02020783996582031, 0.020346656799316406, 0.020460575103759766, 0.020466655731201173, 0.020561471939086914, 0.021002687454223633, 0.020533248901367186, 0.02038278388977051, 0.020184000015258788, 0.019974143981933593, 0.019953664779663087, 0.020154367446899413, 0.020285472869873047, 0.020606943130493164, 0.020691072463989258, 0.020332416534423827, 0.020242431640625, 0.02003763198852539, 0.020014368057250976, 0.019876575469970702, 0.019938976287841796, 0.020071775436401366, 0.019935775756835937, 0.019892255783081056, 0.019810047149658203, 0.019932960510253905, 0.019835807800292968, 0.01987331199645996, 0.020107744216918945, 0.0198287353515625, 0.019808256149291992, 0.019886079788208007, 0.019804159164428712, 0.01982259178161621, 0.019721887588500978, 0.019865951538085937, 0.019704832077026366, 0.019747840881347657, 0.019877824783325195, 0.01974892807006836, 0.019748863220214845, 0.019980384826660157, 0.01982044792175293, 0.019865503311157228, 0.01988617515563965, 0.01984921646118164, 0.020101119995117187, 0.019850431442260744, 0.020048704147338867, 0.019859359741210936, 0.019815519332885743, 0.019874624252319336, 0.019773216247558595, 0.019972320556640624, 0.019894271850585937, 0.019871488571166992, 0.01998464012145996, 0.02049977684020996, 0.020111711502075195, 0.020367712020874024, 0.020231264114379883, 0.020108192443847657, 0.019954784393310547, 0.01998736000061035, 0.01991663932800293, 0.01993744087219238, 0.019980224609375, 0.019847232818603514, 0.019761215209960936, 0.019754207611083985, 0.020083135604858398, 0.019890464782714844, 0.01984480094909668, 0.019943807601928713, 0.020335744857788086, 0.019956544876098634, 0.019870912551879883, 0.0198189754486084, 0.019847007751464845, 0.01991651153564453, 0.020427520751953126, 0.02280054473876953, 0.026818208694458008, 0.02078441619873047, 0.02046860885620117, 0.019895679473876954, 0.019812095642089845, 0.01982703971862793, 0.01978223991394043, 0.01986355209350586, 0.019818496704101563, 0.019797119140625, 0.01983171272277832, 0.019867359161376955, 0.020594944000244142, 0.020606367111206055, 0.020492895126342774, 0.020107263565063475, 0.01997209548950195, 0.019914751052856446, 0.019775487899780272, 0.019763200759887696, 0.0197956485748291, 0.020150367736816405, 0.019914079666137695, 0.0198603515625, 0.019861440658569336, 0.019740095138549806, 0.01989596748352051, 0.019995616912841796, 0.019801151275634764, 0.019907712936401367, 0.02025657653808594, 0.020347103118896485, 0.020133344650268555, 0.01990096092224121, 0.02018022346496582, 0.020215776443481444, 0.020306720733642578, 0.020030975341796875, 0.02033072090148926, 0.020625696182250977, 0.02048201560974121, 0.020471647262573243, 0.020594079971313475, 0.02062620735168457, 0.020436288833618164, 0.020261024475097655, 0.020359712600708006, 0.020625408172607423, 0.0201167049407959, 0.01997001647949219, 0.020073280334472657, 0.01991468811035156, 0.01987571144104004, 0.019955904006958007, 0.01988403129577637, 0.01983692741394043, 0.019981920242309572, 0.01992367935180664, 0.020008544921875, 0.019884191513061523, 0.020124832153320314, 0.020060096740722656, 0.02024844741821289, 0.020038623809814454, 0.020174848556518556, 0.020300928115844726, 0.020407167434692383, 0.020488191604614257, 0.020528352737426758, 0.02069584083557129, 0.020496320724487305, 0.020611135482788086, 0.020074655532836914, 0.01992691230773926, 0.01983616065979004, 0.019913440704345704, 0.02011955261230469, 0.020004480361938477, 0.01987945556640625, 0.019874303817749024, 0.019759456634521486, 0.01991484832763672, 0.01979587173461914, 0.019816640853881837, 0.019772863388061522, 0.019902303695678712, 0.019735071182250978, 0.019766368865966798, 0.019735456466674805, 0.02002236747741699, 0.019906751632690428, 0.01980691146850586, 0.019732511520385743, 0.01966659164428711, 0.01970956802368164, 0.019877920150756834, 0.019763551712036132, 0.019801088333129883, 0.019710975646972655, 0.020193504333496093, 0.02046953582763672, 0.02016774368286133, 0.01995667266845703, 0.01987993621826172, 0.019755008697509766, 0.019842752456665037, 0.019751455307006834, 0.019748064041137697, 0.01983132743835449, 0.01981648063659668, 0.020000768661499024, 0.01982259178161621, 0.019836992263793946, 0.019680736541748046, 0.019665376663208008, 0.019754783630371094, 0.01965488052368164, 0.019742719650268553, 0.020086368560791015, 0.01970172882080078, 0.019860031127929688, 0.01987571144104004, 0.01964784049987793, 0.019731103897094728, 0.019754144668579103, 0.0198144645690918, 0.019888927459716797, 0.020033151626586913, 0.01999635124206543, 0.019778240203857423, 0.019796031951904297, 0.019944799423217772, 0.02018364715576172, 0.019959808349609375, 0.01981420707702637, 0.019784927368164062, 0.019696287155151367, 0.01971027183532715, 0.019869695663452147, 0.020082496643066407, 0.020062400817871095, 0.019832576751708984, 0.019806495666503908, 0.01967100715637207, 0.019772991180419922, 0.019795967102050782, 0.01967148780822754, 0.019705951690673826, 0.019742496490478517, 0.019840608596801756, 0.020302143096923828, 0.019757280349731444, 0.020350528717041017, 0.02063164710998535, 0.020535648345947264, 0.021102592468261717, 0.019955711364746095, 0.01977462387084961, 0.02003555107116699, 0.019846208572387697, 0.019863359451293944, 0.019859359741210936, 0.01984931182861328, 0.019813888549804686, 0.019741184234619142, 0.01978982353210449, 0.0197938232421875, 0.01986774444580078, 0.01978326416015625, 0.02009539222717285, 0.020445184707641603, 0.02045747184753418, 0.020531200408935548, 0.020449087142944335, 0.020637887954711914, 0.020613344192504882, 0.020563743591308595, 0.02207334327697754, 0.02037059211730957, 0.02017695999145508, 0.01996041679382324, 0.01986934471130371, 0.019925056457519533, 0.02025929641723633, 0.02008620834350586, 0.02017321586608887, 0.02024995231628418, 0.02087331199645996, 0.020451967239379882, 0.02012544059753418, 0.020242784500122072, 0.020535295486450195, 0.020410367965698242, 0.020109312057495117, 0.02006425666809082, 0.020200576782226563, 0.020191360473632812, 0.020631839752197265, 0.020381631851196288, 0.020705759048461916, 0.021045312881469727, 0.02046976089477539, 0.020172800064086914, 0.02023017692565918, 0.02012726402282715, 0.020082815170288086, 0.020113088607788085, 0.02026051139831543, 0.020575199127197265, 0.020371456146240235, 0.020244319915771483, 0.02018079948425293, 0.020123615264892578, 0.02045795249938965, 0.020148128509521485, 0.019965087890625, 0.020230335235595705, 0.01999305534362793, 0.020356447219848632, 0.02017980766296387, 0.020144384384155275, 0.020404224395751954, 0.020516351699829103, 0.020349695205688478, 0.020192415237426757, 0.020122207641601563, 0.020385759353637695, 0.020788448333740234, 0.02074870491027832, 0.023590431213378907, 0.020811967849731446, 0.02037945556640625, 0.020202367782592774, 0.020074335098266602, 0.020193439483642578, 0.020131839752197265, 0.02064588737487793, 0.019968000411987305, 0.019891424179077147, 0.020007551193237303, 0.019894399642944337, 0.019951648712158203, 0.020101024627685548, 0.020101215362548826, 0.020395872116088867, 0.02283535957336426, 0.021186111450195312, 0.020573759078979494, 0.020448127746582032, 0.020428800582885744, 0.020455423355102538, 0.02059894371032715, 0.020562879562377928, 0.020423328399658203, 0.02023027229309082, 0.020382944107055663, 0.020611616134643556, 0.020402559280395506, 0.020340736389160157, 0.020090272903442383, 0.020095232009887696, 0.020253023147583007, 0.020213056564331054, 0.020324480056762694, 0.023023775100708008, 0.020218271255493164, 0.020166751861572265, 0.020109216690063478, 0.020007232666015624, 0.020042720794677733, 0.02011801528930664, 0.020115903854370117, 0.020303680419921876, 0.020264928817749023, 0.020092927932739257, 0.020111616134643556, 0.020176544189453124, 0.020355167388916014, 0.021048511505126953, 0.020902463912963867, 0.021506208419799805, 0.020462751388549805, 0.02039904022216797, 0.02007049560546875, 0.02008064079284668, 0.019944704055786133, 0.01996028709411621, 0.020392127990722656, 0.02035868835449219, 0.020207263946533202, 0.020108192443847657, 0.02002351951599121, 0.020524192810058593, 0.020088640213012696, 0.020138816833496095, 0.01995350456237793, 0.019970207214355468, 0.020033536911010744, 0.020000768661499024, 0.019908607482910155, 0.019938623428344727, 0.02002332878112793, 0.019891904830932616, 0.01992188835144043, 0.02009225654602051, 0.019921247482299804, 0.019929344177246094, 0.019836992263793946, 0.019969343185424804, 0.02001519966125488, 0.020306720733642578, 0.02036627197265625, 0.020701120376586914, 0.020608064651489257, 0.02050819206237793, 0.020457632064819337, 0.020343008041381835, 0.02023948860168457, 0.02039852714538574, 0.020381311416625976, 0.020300575256347656, 0.020113407135009767, 0.02017807960510254, 0.019967840194702147, 0.019983360290527344, 0.0198604793548584, 0.019936256408691407, 0.020176895141601564, 0.020142335891723633, 0.019785472869873047, 0.01972047996520996, 0.01993903923034668, 0.02028544044494629, 0.020330495834350586, 0.020406400680541992, 0.02030761528015137, 0.020152351379394532, 0.020267200469970704, 0.02025651168823242, 0.020199167251586915, 0.020359199523925783, 0.02062998390197754, 0.020600831985473633, 0.020176704406738282, 0.020074335098266602, 0.020658527374267578]",tokens/s,49.53649294805587,, -float32-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,phi3,microsoft/Phi-3-mini-4k-instruct,microsoft/Phi-3-mini-4k-instruct,cuda,0,42,,,True,True,,float32,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.1,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): +float32-eager,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,phi3,microsoft/Phi-3-mini-4k-instruct,microsoft/Phi-3-mini-4k-instruct,cuda,0,42,,,True,True,,float32,True,False,,eager,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.225-213.878.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.5.0,,4.45.2,,0.34.2,,,,1.22.0,,,,0.13.0,,"Traceback (most recent call last): File ""/workspace/src/common/benchmark_runner.py"", line 118, in execute_and_log_benchmark benchmark_report = Benchmark.launch(benchmark_config) File ""/usr/local/lib/python3.10/dist-packages/optimum_benchmark/benchmark/base.py"", line 47, in launch @@ -24288,9 +24288,9 @@ ChildProcessError: Traceback (most recent call last): return self.pretrained_model.generate(**inputs, **kwargs) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 116, in decorate_context return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2048, in generate + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2047, in generate result = self._sample( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 3008, in _sample + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 3007, in _sample outputs = self(**model_inputs, return_dict=True) File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1553, in _wrapped_call_impl return self._call_impl(*args, **kwargs) @@ -24316,7 +24316,7 @@ ChildProcessError: Traceback (most recent call last): return forward_call(*args, **kwargs) File ""/root/.cache/huggingface/modules/transformers_modules/microsoft/Phi-3-mini-4k-instruct/0a67737cc96d2554230f90338b163bc6380a2a85/modeling_phi3.py"", line 233, in forward up_states = up_states * self.activation_fn(gate) -torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 20.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 8.12 MiB is free. Process 39952 has 14.73 GiB memory in use. Of the allocated memory 14.56 GiB is allocated by PyTorch, and 45.84 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) +torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 20.00 MiB. GPU 0 has a total capacity of 14.74 GiB of which 8.12 MiB is free. Process 39946 has 14.73 GiB memory in use. Of the allocated memory 14.56 GiB is allocated by PyTorch, and 45.84 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables) ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, float16-flash_attention_2,pytorch,2.4.1+cu124,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,phi,microsoft/phi-1_5,microsoft/phi-1_5,cuda,0,42,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,False,spawn, Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz,8,33163.759616,Linux,x86_64,Linux-5.10.224-212.876.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['Tesla T4'],1,16106127360,0.4.0,,4.44.2,,0.34.2,,,,1.22.0,,,,0.12.0,,"Traceback (most recent call last):