Skip to content
GitLab
Menu
Projects
Groups
Snippets
Loading...
Help
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in
Toggle navigation
Menu
Open sidebar
OpenDAS
vllm_cscc
Commits
8b793f7e
Unverified
Commit
8b793f7e
authored
Mar 18, 2025
by
ekuznetsov139
Committed by
GitHub
Mar 18, 2025
Browse files
MI325 configs, fused_moe_kernel bugfix (#14987)
Signed-off-by:
Eugene Kuznetsov
<
eugene.kuznetsov@amd.com
>
parent
af35d3a3
Changes
54
Hide whitespace changes
Inline
Side-by-side
Showing
20 changed files
with
3302 additions
and
1 deletion
+3302
-1
vllm/model_executor/layers/fused_moe/configs/E=256,N=1024,device_name=AMD_Instinct_MI325X,block_shape=[128,128].json
...evice_name=AMD_Instinct_MI325X,block_shape=[128,128].json
+200
-0
vllm/model_executor/layers/fused_moe/configs/E=256,N=1024,device_name=AMD_Instinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
...tinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
+200
-0
vllm/model_executor/layers/fused_moe/configs/E=256,N=256,device_name=AMD_Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
...Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
+200
-0
vllm/model_executor/layers/fused_moe/configs/E=256,N=256,device_name=AMD_Instinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
...tinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
+200
-0
vllm/model_executor/layers/fused_moe/configs/E=256,N=512,device_name=AMD_Instinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
...tinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
+200
-0
vllm/model_executor/layers/fused_moe/fused_moe.py
vllm/model_executor/layers/fused_moe/fused_moe.py
+6
-1
vllm/model_executor/layers/quantization/utils/configs/N=1536,K=1536,device_name=AMD_Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
...Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
+164
-0
vllm/model_executor/layers/quantization/utils/configs/N=1536,K=1536,device_name=AMD_Instinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
...tinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
+164
-0
vllm/model_executor/layers/quantization/utils/configs/N=1536,K=7168,device_name=AMD_Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
...Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
+164
-0
vllm/model_executor/layers/quantization/utils/configs/N=1536,K=7168,device_name=AMD_Instinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
...tinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
+164
-0
vllm/model_executor/layers/quantization/utils/configs/N=2048,K=512,device_name=AMD_Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
...Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
+164
-0
vllm/model_executor/layers/quantization/utils/configs/N=2048,K=512,device_name=AMD_Instinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
...tinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
+164
-0
vllm/model_executor/layers/quantization/utils/configs/N=2304,K=7168,device_name=AMD_Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
...Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
+164
-0
vllm/model_executor/layers/quantization/utils/configs/N=2304,K=7168,device_name=AMD_Instinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
...tinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
+164
-0
vllm/model_executor/layers/quantization/utils/configs/N=24576,K=7168,device_name=AMD_Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
...Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
+164
-0
vllm/model_executor/layers/quantization/utils/configs/N=24576,K=7168,device_name=AMD_Instinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
...tinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
+164
-0
vllm/model_executor/layers/quantization/utils/configs/N=256,K=7168,device_name=AMD_Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
...Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
+164
-0
vllm/model_executor/layers/quantization/utils/configs/N=256,K=7168,device_name=AMD_Instinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
...tinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
+164
-0
vllm/model_executor/layers/quantization/utils/configs/N=3072,K=1536,device_name=AMD_Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
...Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
+164
-0
vllm/model_executor/layers/quantization/utils/configs/N=3072,K=1536,device_name=AMD_Instinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
...tinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
+164
-0
No files found.
vllm/model_executor/layers/fused_moe/configs/E=256,N=1024,device_name=AMD_Instinct_MI325X,block_shape=[128,128].json
0 → 100644
View file @
8b793f7e
{
"1"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
256
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"2"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
64
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"4"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"8"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"16"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"24"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"32"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
256
,
"GROUP_SIZE_M"
:
16
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"48"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
256
,
"GROUP_SIZE_M"
:
16
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"64"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
32
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"96"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
32
,
"num_warps"
:
1
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"128"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
32
,
"num_warps"
:
1
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"256"
:
{
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
32
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"512"
:
{
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"1024"
:
{
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"1536"
:
{
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"2048"
:
{
"BLOCK_SIZE_M"
:
128
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
64
,
"GROUP_SIZE_M"
:
16
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"3072"
:
{
"BLOCK_SIZE_M"
:
128
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
64
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"4096"
:
{
"BLOCK_SIZE_M"
:
128
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
64
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
}
}
vllm/model_executor/layers/fused_moe/configs/E=256,N=1024,device_name=AMD_Instinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
0 → 100644
View file @
8b793f7e
{
"1"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
256
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"2"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
64
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"4"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"8"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"16"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"24"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"32"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
256
,
"GROUP_SIZE_M"
:
16
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"48"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
256
,
"GROUP_SIZE_M"
:
16
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"64"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
32
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"96"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
32
,
"num_warps"
:
1
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"128"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
32
,
"num_warps"
:
1
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"256"
:
{
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
32
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"512"
:
{
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"1024"
:
{
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"1536"
:
{
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"2048"
:
{
"BLOCK_SIZE_M"
:
128
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
64
,
"GROUP_SIZE_M"
:
16
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"3072"
:
{
"BLOCK_SIZE_M"
:
128
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
64
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"4096"
:
{
"BLOCK_SIZE_M"
:
128
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
64
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
}
}
vllm/model_executor/layers/fused_moe/configs/E=256,N=256,device_name=AMD_Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
0 → 100644
View file @
8b793f7e
{
"1"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
16
,
"BLOCK_SIZE_K"
:
256
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"2"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"4"
:
{
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"8"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"16"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"24"
:
{
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
64
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"32"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"48"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
2
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"64"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
2
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"96"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
32
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
4
,
"num_warps"
:
2
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"128"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
32
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
4
,
"num_warps"
:
1
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"256"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
32
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
2
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"512"
:
{
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"1024"
:
{
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"1536"
:
{
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"2048"
:
{
"BLOCK_SIZE_M"
:
128
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
64
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"3072"
:
{
"BLOCK_SIZE_M"
:
128
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
64
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"4096"
:
{
"BLOCK_SIZE_M"
:
128
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
64
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
}
}
vllm/model_executor/layers/fused_moe/configs/E=256,N=256,device_name=AMD_Instinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
0 → 100644
View file @
8b793f7e
{
"1"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
16
,
"BLOCK_SIZE_K"
:
256
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"2"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"4"
:
{
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"8"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"16"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"24"
:
{
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
64
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"32"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"48"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
2
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"64"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
2
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"96"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
32
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
4
,
"num_warps"
:
2
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"128"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
32
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
4
,
"num_warps"
:
1
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"256"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
32
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
2
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"512"
:
{
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"1024"
:
{
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"1536"
:
{
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"2048"
:
{
"BLOCK_SIZE_M"
:
128
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
64
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"3072"
:
{
"BLOCK_SIZE_M"
:
128
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
64
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"4096"
:
{
"BLOCK_SIZE_M"
:
128
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
64
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
}
}
vllm/model_executor/layers/fused_moe/configs/E=256,N=512,device_name=AMD_Instinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
0 → 100644
View file @
8b793f7e
{
"1"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
16
,
"BLOCK_SIZE_K"
:
256
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"2"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"4"
:
{
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"8"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"16"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
1
},
"24"
:
{
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
64
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"32"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
4
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"48"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
2
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"64"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
2
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"96"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
32
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
4
,
"num_warps"
:
2
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"128"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
32
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
4
,
"num_warps"
:
1
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"256"
:
{
"BLOCK_SIZE_M"
:
16
,
"BLOCK_SIZE_N"
:
32
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
1
,
"num_warps"
:
2
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"512"
:
{
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
64
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"1024"
:
{
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"1536"
:
{
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
128
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"2048"
:
{
"BLOCK_SIZE_M"
:
128
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
64
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"3072"
:
{
"BLOCK_SIZE_M"
:
128
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
64
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
},
"4096"
:
{
"BLOCK_SIZE_M"
:
128
,
"BLOCK_SIZE_N"
:
128
,
"BLOCK_SIZE_K"
:
64
,
"GROUP_SIZE_M"
:
8
,
"num_warps"
:
8
,
"num_stages"
:
2
,
"waves_per_eu"
:
0
,
"matrix_instr_nonkdim"
:
16
,
"kpack"
:
2
}
}
vllm/model_executor/layers/fused_moe/fused_moe.py
View file @
8b793f7e
...
@@ -783,8 +783,12 @@ def invoke_fused_moe_kernel(A: torch.Tensor,
...
@@ -783,8 +783,12 @@ def invoke_fused_moe_kernel(A: torch.Tensor,
use_int8_w8a16
=
use_int8_w8a16
,
use_int8_w8a16
=
use_int8_w8a16
,
**
config
,
**
config
,
)
)
else
:
else
:
config
=
config
.
copy
()
BLOCK_SIZE_K
=
config
.
pop
(
"BLOCK_SIZE_K"
)
if
block_shape
is
not
None
:
BLOCK_SIZE_K
=
min
(
BLOCK_SIZE_K
,
min
(
block_shape
[
0
],
block_shape
[
1
]))
fused_moe_kernel
[
grid
](
fused_moe_kernel
[
grid
](
A
,
A
,
B
,
B
,
...
@@ -823,6 +827,7 @@ def invoke_fused_moe_kernel(A: torch.Tensor,
...
@@ -823,6 +827,7 @@ def invoke_fused_moe_kernel(A: torch.Tensor,
compute_type
=
compute_type
,
compute_type
=
compute_type
,
use_fp8_w8a8
=
use_fp8_w8a8
,
use_fp8_w8a8
=
use_fp8_w8a8
,
use_int8_w8a16
=
use_int8_w8a16
,
use_int8_w8a16
=
use_int8_w8a16
,
BLOCK_SIZE_K
=
BLOCK_SIZE_K
,
**
config
,
**
config
,
)
)
...
...
vllm/model_executor/layers/quantization/utils/configs/N=1536,K=1536,device_name=AMD_Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
0 → 100644
View file @
8b793f7e
{
"1"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"8"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"16"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"24"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"32"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"48"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"64"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"96"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"128"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"256"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"512"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1024"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1536"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2048"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"3072"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4096"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
}
}
\ No newline at end of file
vllm/model_executor/layers/quantization/utils/configs/N=1536,K=1536,device_name=AMD_Instinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
0 → 100644
View file @
8b793f7e
{
"1"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"8"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"16"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"24"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"32"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"48"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"64"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"96"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"128"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"256"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"512"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1024"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1536"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2048"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"3072"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4096"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
}
}
\ No newline at end of file
vllm/model_executor/layers/quantization/utils/configs/N=1536,K=7168,device_name=AMD_Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
0 → 100644
View file @
8b793f7e
{
"1"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"8"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"16"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"24"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"32"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"48"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"64"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"96"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"128"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"256"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"512"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1024"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1536"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2048"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"3072"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4096"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
}
}
\ No newline at end of file
vllm/model_executor/layers/quantization/utils/configs/N=1536,K=7168,device_name=AMD_Instinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
0 → 100644
View file @
8b793f7e
{
"1"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"8"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"16"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"24"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"32"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"48"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"64"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"96"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"128"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"256"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"512"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1024"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1536"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2048"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"3072"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4096"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
}
}
\ No newline at end of file
vllm/model_executor/layers/quantization/utils/configs/N=2048,K=512,device_name=AMD_Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
0 → 100644
View file @
8b793f7e
{
"1"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"8"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"16"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"24"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"32"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"48"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"64"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"96"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"128"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"256"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"512"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1024"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
64
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1536"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
64
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2048"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"3072"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4096"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
}
}
\ No newline at end of file
vllm/model_executor/layers/quantization/utils/configs/N=2048,K=512,device_name=AMD_Instinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
0 → 100644
View file @
8b793f7e
{
"1"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"8"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"16"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"24"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"32"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"48"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"64"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"96"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"128"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"256"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"512"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1024"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
64
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1536"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
64
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2048"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"3072"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4096"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
}
}
\ No newline at end of file
vllm/model_executor/layers/quantization/utils/configs/N=2304,K=7168,device_name=AMD_Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
0 → 100644
View file @
8b793f7e
{
"1"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"8"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"16"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"24"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"32"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"48"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"64"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"96"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"128"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"256"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"512"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
64
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1024"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1536"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2048"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"3072"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4096"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
}
}
\ No newline at end of file
vllm/model_executor/layers/quantization/utils/configs/N=2304,K=7168,device_name=AMD_Instinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
0 → 100644
View file @
8b793f7e
{
"1"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"8"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"16"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"24"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"32"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"48"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"64"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"96"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"128"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"256"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"512"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
64
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1024"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1536"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2048"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"3072"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4096"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
}
}
\ No newline at end of file
vllm/model_executor/layers/quantization/utils/configs/N=24576,K=7168,device_name=AMD_Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
0 → 100644
View file @
8b793f7e
{
"1"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"8"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"16"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"24"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"32"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"48"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
64
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"64"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
64
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"96"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"128"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"256"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"512"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1024"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1536"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2048"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"3072"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4096"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
}
}
\ No newline at end of file
vllm/model_executor/layers/quantization/utils/configs/N=24576,K=7168,device_name=AMD_Instinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
0 → 100644
View file @
8b793f7e
{
"1"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"8"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"16"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"24"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"32"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"48"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
64
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"64"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
64
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"96"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"128"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"256"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"512"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1024"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1536"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2048"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"3072"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4096"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
}
}
\ No newline at end of file
vllm/model_executor/layers/quantization/utils/configs/N=256,K=7168,device_name=AMD_Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
0 → 100644
View file @
8b793f7e
{
"1"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"8"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"16"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"24"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"32"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"48"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"64"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"96"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"128"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"256"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"512"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1024"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1536"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2048"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"3072"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4096"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
}
}
\ No newline at end of file
vllm/model_executor/layers/quantization/utils/configs/N=256,K=7168,device_name=AMD_Instinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
0 → 100644
View file @
8b793f7e
{
"1"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"8"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"16"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"24"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"32"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"48"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"64"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"96"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"128"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"256"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"512"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1024"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1536"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2048"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"3072"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4096"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
}
}
\ No newline at end of file
vllm/model_executor/layers/quantization/utils/configs/N=3072,K=1536,device_name=AMD_Instinct_MI325X,dtype=fp8_w8a8,block_shape=[128,128].json
0 → 100644
View file @
8b793f7e
{
"1"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"8"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"16"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"24"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"32"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"48"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"64"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"96"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"128"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"256"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"512"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1024"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1536"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2048"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"3072"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4096"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
}
}
\ No newline at end of file
vllm/model_executor/layers/quantization/utils/configs/N=3072,K=1536,device_name=AMD_Instinct_MI325_OAM,dtype=fp8_w8a8,block_shape=[128,128].json
0 → 100644
View file @
8b793f7e
{
"1"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"8"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"16"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"24"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"32"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"48"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
8
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"64"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"96"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"128"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
16
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"256"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
32
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"512"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1024"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
32
,
"GROUP_SIZE_M"
:
1
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"1536"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"2048"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"3072"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
},
"4096"
:
{
"BLOCK_SIZE_K"
:
128
,
"BLOCK_SIZE_M"
:
64
,
"BLOCK_SIZE_N"
:
128
,
"GROUP_SIZE_M"
:
32
,
"kpack"
:
1
,
"matrix_instr_nonkdim"
:
16
,
"num_warps"
:
4
}
}
\ No newline at end of file
Prev
1
2
3
Next
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
.
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment