Skip to content
This repository was archived by the owner on Sep 10, 2025. It is now read-only.

Update test-tinystories-executorch: Disable config mobile-32, fix typo #1495

Merged
merged 2 commits into from
Feb 21, 2025
Merged

Update test-tinystories-executorch: Disable config mobile-32, fix typo #1495

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
3258c99
Update test-tinystories-executorch: Disable config mobile-32, fix typo
Jack-Khuu Feb 21, 2025
56c749a
Disable bitwidth 4 test
Jack-Khuu Feb 21, 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
21 changes: 10 additions & 11 deletions .github/workflows/pull.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -515,12 +515,11 @@ jobs:
python3 torchchat.py generate --checkpoint-path ${MODEL_PATH} --temperature 0 --pte-path ${MODEL_DIR}/${MODEL_NAME}.pte

echo "******************************************"
echo "*** can't test --quantize torchchat/quant_config/mobile.json ***"
echo "*** testing --quantize torchchat/quant_config/mobile-32.json ***"
echo "*** [TEST DISABLED] Can't test --quantize torchchat/quant_config/mobile.json ***"
echo "*** Testing --quantize torchchat/quant_config/mobile-32.json instead ***"
echo "******************************************"
python torchchat.py export --quantize torchchat/quant_config/mobile-32.json --checkpoint-path ${MODEL_PATH} --output-pte-path ${MODEL_DIR}/${MODEL_NAME}.pte
python3 torchchat.py generate --checkpoint-path ${MODEL_PATH} --temperature 0 --pte-path ${MODEL_DIR}/${MODEL_NAME}.pte

# python torchchat.py export --quantize torchchat/quant_config/mobile-32.json --checkpoint-path ${MODEL_PATH} --output-pte-path ${MODEL_DIR}/${MODEL_NAME}.pte
# python3 torchchat.py generate --checkpoint-path ${MODEL_PATH} --temperature 0 --pte-path ${MODEL_DIR}/${MODEL_NAME}.pte

echo "******************************************"
echo "******* Emb: channel-wise quantized ******"
Expand All @@ -535,16 +534,16 @@ jobs:
python3 torchchat.py generate --checkpoint-path ${MODEL_PATH} --temperature 0 --pte-path ${MODEL_DIR}/${MODEL_NAME}.pte

echo "******************************************"
echo "**** Emb 4bit: channel-wise quantized ****"
echo "**** [TEST DISABLED] Emb 4bit: channel-wise quantized ****"
echo "******************************************"
python torchchat.py export --quant '{"embedding" : {"bitwidth": 8, "groupsize": 0}}' --checkpoint-path ${MODEL_PATH} --output-pte-path ${MODEL_DIR}/${MODEL_NAME}.pte
python3 torchchat.py generate --checkpoint-path ${MODEL_PATH} --temperature 0 --pte-path ${MODEL_DIR}/${MODEL_NAME}.pte
# python torchchat.py export --quant '{"embedding" : {"bitwidth": 4, "groupsize": 0}}' --checkpoint-path ${MODEL_PATH} --output-pte-path ${MODEL_DIR}/${MODEL_NAME}.pte
# python3 torchchat.py generate --checkpoint-path ${MODEL_PATH} --temperature 0 --pte-path ${MODEL_DIR}/${MODEL_NAME}.pte

echo "******************************************"
echo "****** Emb 4bit: group-wise quantized ****"
echo "****** [TEST DISABLED] Emb 4bit: group-wise quantized ****"
echo "******************************************"
python torchchat.py export --quant '{"embedding" : {"bitwidth": 8, "groupsize": 8}}' --checkpoint-path ${MODEL_PATH} --output-pte-path ${MODEL_DIR}/${MODEL_NAME}.pte
python3 torchchat.py generate --checkpoint-path ${MODEL_PATH} --temperature 0 --pte-path ${MODEL_DIR}/${MODEL_NAME}.pte
# python torchchat.py export --quant '{"embedding" : {"bitwidth": 4, "groupsize": 8}}' --checkpoint-path ${MODEL_PATH} --output-pte-path ${MODEL_DIR}/${MODEL_NAME}.pte
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

any reason we need to change the args for commented command?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The test was meant to test bitwidth 4 (see the comment above it), but the code itself looks like a typo (copy and paste) from the tests that tests bitwidth8

# python3 torchchat.py generate --checkpoint-path ${MODEL_PATH} --temperature 0 --pte-path ${MODEL_DIR}/${MODEL_NAME}.pte

echo "******************************************"
echo "******* INT8 channel-wise quantized ******"
Expand Down
Loading