Unverified Commit 2f0f6541 authored by Ziyue Yang's avatar Ziyue Yang Committed by GitHub
Browse files

Benchmarks: Fix Bug - Fix wrong parameters for gpu-sm-copy-bw in configuration examples (#203)

**Description**
This commit fixes wrong parameters for gpu-sm-copy-bw call in configuration examples.
parent 53e972c3
...@@ -75,15 +75,16 @@ superbench: ...@@ -75,15 +75,16 @@ superbench:
parameters: parameters:
block_devices: [] block_devices: []
gpu-sm-copy-bw: gpu-sm-copy-bw:
enable: false enable: true
modes: modes:
- name: local - name: local
proc_num: 32 proc_num: 32
prefix: CUDA_VISIBLE_DEVICES=$(({proc_rank}%8)) numactl -N $(({proc_rank}%4)) -m $(({proc_rank}%4)) prefix: CUDA_VISIBLE_DEVICES=$(({proc_rank}%8)) numactl -N $(({proc_rank}%4)) -m $(({proc_rank}%4))
parallel: no parallel: no
parameters: parameters:
dtoh: true mem_type:
htod: true - dtoh
- htod
gpt_models: gpt_models:
<<: *default_pytorch_mode <<: *default_pytorch_mode
models: models:
......
...@@ -35,6 +35,25 @@ superbench: ...@@ -35,6 +35,25 @@ superbench:
<<: *default_local_mode <<: *default_local_mode
gemm-flops: gemm-flops:
<<: *default_local_mode <<: *default_local_mode
disk-benchmark:
enable: false
modes:
- name: local
proc_num: 1
parallel: no
parameters:
block_devices: []
gpu-sm-copy-bw:
enable: true
modes:
- name: local
proc_num: 32
prefix: CUDA_VISIBLE_DEVICES=$(({proc_rank}%8)) numactl -N $(({proc_rank}%4)) -m $(({proc_rank}%4))
parallel: no
parameters:
mem_type:
- dtoh
- htod
cudnn-function: cudnn-function:
<<: *default_local_mode <<: *default_local_mode
cublas-function: cublas-function:
......
...@@ -61,15 +61,16 @@ superbench: ...@@ -61,15 +61,16 @@ superbench:
prefix: CUDA_VISIBLE_DEVICES={proc_rank} numactl -c $(({proc_rank}/2)) prefix: CUDA_VISIBLE_DEVICES={proc_rank} numactl -c $(({proc_rank}/2))
parallel: yes parallel: yes
gpu-sm-copy-bw: gpu-sm-copy-bw:
enable: false enable: true
modes: modes:
- name: local - name: local
proc_num: 32 proc_num: 32
prefix: CUDA_VISIBLE_DEVICES=$(({proc_rank}%8)) numactl -N $(({proc_rank}%4)) -m $(({proc_rank}%4)) prefix: CUDA_VISIBLE_DEVICES=$(({proc_rank}%8)) numactl -N $(({proc_rank}%4)) -m $(({proc_rank}%4))
parallel: no parallel: no
parameters: parameters:
dtoh: true mem_type:
htod: true - dtoh
- htod
kernel-launch: kernel-launch:
<<: *default_local_mode <<: *default_local_mode
gemm-flops: gemm-flops:
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment