未验证 提交 5cd04ae9 编写于 作者: D Double_V 提交者: GitHub

Merge pull request #6794 from mmglove/dygraph

fix benchmark
...@@ -54,6 +54,6 @@ random_infer_input:[{float32,[3,640,640]}];[{float32,[3,960,960]}] ...@@ -54,6 +54,6 @@ random_infer_input:[{float32,[3,640,640]}];[{float32,[3,960,960]}]
===========================train_benchmark_params========================== ===========================train_benchmark_params==========================
batch_size:8|16 batch_size:8|16
fp_items:fp32|fp16 fp_items:fp32|fp16
epoch:2 epoch:15
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile --profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096 flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
...@@ -54,5 +54,5 @@ random_infer_input:[{float32,[3,640,640]}];[{float32,[3,960,960]}] ...@@ -54,5 +54,5 @@ random_infer_input:[{float32,[3,640,640]}];[{float32,[3,960,960]}]
===========================train_benchmark_params========================== ===========================train_benchmark_params==========================
batch_size:8|16 batch_size:8|16
fp_items:fp32|fp16 fp_items:fp32|fp16
epoch:2 epoch:15
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile --profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
...@@ -54,5 +54,5 @@ random_infer_input:[{float32,[3,640,640]}];[{float32,[3,960,960]}] ...@@ -54,5 +54,5 @@ random_infer_input:[{float32,[3,640,640]}];[{float32,[3,960,960]}]
===========================train_benchmark_params========================== ===========================train_benchmark_params==========================
batch_size:8 batch_size:8
fp_items:fp32|fp16 fp_items:fp32|fp16
epoch:2 epoch:10
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile --profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册