Skip to content
This repository was archived by the owner on Sep 10, 2025. It is now read-only.
Merged
Changes from all commits
Commits
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
29 changes: 15 additions & 14 deletions .github/workflows/compile.yml
Original file line number Diff line number Diff line change
Expand Up @@ -51,9 +51,9 @@ jobs:
python generate.py --checkpoint-path ${MODEL_PATH} --temperature 0 --dso-path ${MODEL_DIR}/${MODEL_NAME}.so > ./output_aoti
cat ./output_aoti

# echo "******************************************"
# echo "******* Emb: channel-wise quantized ******"
# echo "******************************************"
echo "******************************************"
echo "******* Emb: channel-wise quantized ******"
echo "******************************************"
python generate.py --quant '{"embedding" : {"bitwidth": 8, "group_size": 0}}' --checkpoint-path ${MODEL_PATH} --temperature 0 > ./output_eager
cat ./output_eager
python generate.py --compile --quant '{"embedding" : {"bitwidth": 8, "group_size": 0}}' --checkpoint-path ${MODEL_PATH} --temperature 0 > ./output_compiled
Expand All @@ -62,20 +62,20 @@ jobs:
python generate.py --checkpoint-path ${MODEL_PATH} --temperature 0 --dso-path ${MODEL_DIR}/${MODEL_NAME}.so > ./output_aoti
cat ./output_aoti

# echo "******************************************"
# echo "******** Emb: group-wise quantized *******"
# echo "******************************************"
echo "******************************************"
echo "******** Emb: group-wise quantized *******"
echo "******************************************"
python generate.py --quant '{"embedding" : {"bitwidth": 8, "group_size": 8}}' --checkpoint-path ${MODEL_PATH} --temperature 0 > ./output_eager
cat ./output_eager
python generate.py --compile --quant '{"embedding" : {"bitwidth": 8, "group_size": 8}}' --checkpoint-path ${MODEL_PATH} --temperature 0 > ./output_compiled
cat ./output_compiled
python export.py --quant "embedding" : {"bitwidth": 8, "group_size": 8}}' --checkpoint-path ${MODEL_PATH} --output-dso-path ${MODEL_DIR}/${MODEL_NAME}.so
python export.py --quant '{"embedding" : {"bitwidth": 8, "group_size": 8}}' --checkpoint-path ${MODEL_PATH} --output-dso-path ${MODEL_DIR}/${MODEL_NAME}.so
python generate.py --checkpoint-path ${MODEL_PATH} --temperature 0 --dso-path ${MODEL_DIR}/${MODEL_NAME}.so > ./output_aoti
cat ./output_aoti

# echo "******************************************"
# echo "******* INT8 channel-wise quantized ******"
# echo "******************************************"
echo "******************************************"
echo "******* INT8 channel-wise quantized ******"
echo "******************************************"
python generate.py --quant '{"linear:int8" : {"bitwidth": 8, "group_size": 0}}' --checkpoint-path ${MODEL_PATH} --temperature 0 > ./output_eager
cat ./output_eager
python generate.py --compile --quant '{"linear:int8" : {"bitwidth": 8, "group_size": 0}}' --checkpoint-path ${MODEL_PATH} --temperature 0 > ./output_compiled
Expand All @@ -84,9 +84,9 @@ jobs:
python generate.py --checkpoint-path ${MODEL_PATH} --temperature 0 --dso-path ${MODEL_DIR}/${MODEL_NAME}.so > ./output_aoti
cat ./output_aoti

# echo "******************************************"
# echo "******** INT8 group-wise quantized *******"
# echo "******************************************"
echo "******************************************"
echo "******** INT8 group-wise quantized *******"
echo "******************************************"
python generate.py --quant '{"linear:int8" : {"bitwidth": 8, "group_size": 8}}' --checkpoint-path ${MODEL_PATH} --temperature 0 > ./output_eager
cat ./output_eager
python generate.py --compile --quant '{"linear:int8" : {"bitwidth": 8, "group_size": 8}}' --checkpoint-path ${MODEL_PATH} --temperature 0 > ./output_compiled
Expand All @@ -95,7 +95,8 @@ jobs:
python generate.py --checkpoint-path ${MODEL_PATH} --temperature 0 --dso-path ${MODEL_DIR}/${MODEL_NAME}.so > ./output_aoti
cat ./output_aoti

# echo "******************************************"
echo "tests complete"
echo "******************************************"
# echo "********* EAGER vs TORCH.COMPILE *********"
# echo "******************************************"
# diff output_eager output_compiled
Expand Down