iHore/vLLM
8
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

Add tuned selective_state_update config for AMD Instinct MI355 (#47767)

Browse Source 
Signed-off-by: vanshbhatia-amd <210711135+vanshbhatia-amd@users.noreply.github.com>
Co-authored-by: vanshbhatia-amd <210711135+vanshbhatia-amd@users.noreply.github.com>
This commit is contained in:
vanshbhatia-amd
2026-07-07 22:19:08 +00:00
committed by GitHub
co-authored by vanshbhatia-amd
parent dd0d74cd92
commit c8c2f838e7
1 changed files with 35 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
vllm/model_executor/layers/mamba/ops/configs/selective_state_update/headdim=64,dstate=128,device_name=AMD_Instinct_MI355_OAM,cache_dtype=float16.json
+35
View File
@@ -0,0 +1,35 @@
{
"triton_version": "3.6.0",
"10": {
"BLOCK_SIZE_M": 16,
"num_warps": 4
},
"80": {
"BLOCK_SIZE_M": 16,
"num_warps": 4
},
"160": {
"BLOCK_SIZE_M": 16,
"num_warps": 4
},
"320": {
"BLOCK_SIZE_M": 16,
"num_warps": 4
},
"640": {
"BLOCK_SIZE_M": 32,
"num_warps": 4
},
"1280": {
"BLOCK_SIZE_M": 16,
"num_warps": 1
},
"2560": {
"BLOCK_SIZE_M": 16,
"num_warps": 1
},
"5120": {
"BLOCK_SIZE_M": 16,
"num_warps": 1
}
}