Skip to content

Add a comment on the use of float16 and set some EVs explicitly #69

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
functionstackx merged 4 commits into from
Sep 28, 2025
Merged

Add a comment on the use of float16 and set some EVs explicitly #69

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
76d5071
Explain the use of float16 for the dtype option
qcolombet Sep 28, 2025
95a499e
[mi355][70b] Explicitly set VLLM_ROCM_USE_AITER_MHA
qcolombet Sep 28, 2025
eca8d22
Update 70b_fp4_mi355x_docker.sh envvars
JArnoldAMD Sep 28, 2025
39dafc4
Update 70b_fp4_mi355x_slurm.sh envvars
JArnoldAMD Sep 28, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 7 additions & 0 deletions benchmarks/70b_fp4_mi355x_docker.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -28,6 +28,13 @@ elif [[ "$ISL" == "1024" && "$OSL" == "8192" ]]; then
elif [[ "$ISL" == "8192" && "$OSL" == "1024" ]]; then
if [[ "$CONC" -gt "16" ]]; then
export VLLM_ROCM_USE_AITER_MHA=1
else
export VLLM_ROCM_USE_AITER_MHA=0
fi
if [[ "$CONC" -lt "16" && "$TP" -gt "1" ]]; then
export VLLM_ROCM_USE_AITER_TRITON_BF16_GEMM=0
else
export VLLM_TRITON_FP4_GEMM_USE_ASM=1
fi
fi

Expand Down
13 changes: 10 additions & 3 deletions benchmarks/70b_fp4_mi355x_slurm.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -33,9 +33,16 @@ elif [[ "$ISL" == "1024" && "$OSL" == "8192" ]]; then
export VLLM_TRITON_FP4_GEMM_USE_ASM=1
fi
elif [[ "$ISL" == "8192" && "$OSL" == "1024" ]]; then
if [[ "$CONC" -gt "16" ]]; then
export VLLM_ROCM_USE_AITER_MHA=1
fi
if [[ "$CONC" -gt "16" ]]; then
export VLLM_ROCM_USE_AITER_MHA=1
else
export VLLM_ROCM_USE_AITER_MHA=0
fi
if [[ "$CONC" -lt "16" && "$TP" -gt "1" ]]; then
export VLLM_ROCM_USE_AITER_TRITON_BF16_GEMM=0
else
export VLLM_TRITON_FP4_GEMM_USE_ASM=1
fi
fi


Expand Down
2 changes: 2 additions & 0 deletions benchmarks/70b_fp8_mi300x_docker.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -34,6 +34,8 @@ elif [[ "$ISL" == "8192" && "$OSL" == "1024" ]]; then
fi
fi

# In this specific case, float16 performs better than the datatype
# picked by vllm when using auto for --dtype (bfloat16).
set -x
vllm serve $MODEL --port=$PORT \
--swap-space=64 \
Expand Down
2 changes: 2 additions & 0 deletions benchmarks/70b_fp8_mi300x_slurm.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -45,6 +45,8 @@ elif [[ "$ISL" == "8192" && "$OSL" == "1024" ]]; then
fi
fi

# In this specific case, float16 performs better than the datatype
# picked by vllm when using auto for --dtype (bfloat16).
set -x
vllm serve $MODEL --port=$PORT \
--swap-space=64 \
Expand Down
2 changes: 2 additions & 0 deletions benchmarks/70b_fp8_mi325x_docker.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -22,6 +22,8 @@ elif [[ "$ISL" == "8192" && "$OSL" == "1024" ]]; then
fi
fi

# In this specific case, float16 performs better than the datatype
# picked by vllm when using auto for --dtype (bfloat16).
set -x
vllm serve $MODEL --port=$PORT \
--swap-space=64 \
Expand Down
2 changes: 2 additions & 0 deletions benchmarks/70b_fp8_mi325x_slurm.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -34,6 +34,8 @@ elif [[ "$ISL" == "8192" && "$OSL" == "1024" ]]; then
fi
fi

# In this specific case, float16 performs better than the datatype
# picked by vllm when using auto for --dtype (bfloat16).
set -x
vllm serve $MODEL --port=$PORT \
--swap-space=64 \
Expand Down