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

Update pull.yml to test snapshot saving and loading #1486

Merged
merged 6 commits into from
Feb 18, 2025
Merged

Update pull.yml to test snapshot saving and loading #1486

Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
343c94d
Update pull.yml to test snapshot saving and loading
mikekgfb Jan 31, 2025
30a0d31
Update pull.yml
mikekgfb Jan 31, 2025
2afca1e
Update pull.yml
mikekgfb Feb 6, 2025
cb24aec
Merge branch 'pytorch:main' into patch-48
mikekgfb Feb 6, 2025
e733606
Update pull.yml
mikekgfb Feb 8, 2025
5d098bd
Update pull.yml
mikekgfb Feb 18, 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
29 changes: 25 additions & 4 deletions .github/workflows/pull.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -291,6 +291,16 @@ jobs:
bash .ci/scripts/validate.sh "./checkpoints/${REPO_NAME}/model.pth" "cuda" "aoti-bfloat16"
echo "::endgroup::"

echo "::group::Run inference with quantize file"
for DEVICE in cpu; do # cuda
# cuda - fails because `AttributeError: 'Linear' object has no attribute '_linear_extra_repr'`
# follow up with torchao as a separate PR
echo "saving snapshot for device ${DEVICE} and dtype bfloat16, and reloading as snapshot"
python3 torchchat.py export --device ${DEVICE} --output-snap model.tc --dtype bfloat16 --quantize torchchat/quant_config/cuda-32.json --checkpoint "./checkpoints/${REPO_NAME}/model.pth"
python3 torchchat.py generate --device ${DEVICE} --snap model.tc --dtype bfloat16 --checkpoint "./checkpoints/${REPO_NAME}/model.pth"
done
echo "::endgroup::"

test-gpu-aoti-float32:
permissions:
id-token: write
Expand Down Expand Up @@ -335,6 +345,11 @@ jobs:
fi
echo "::endgroup::"

# echo "::group::Run inference with quantize file"
# python3 torchchat.py export --output-snap model.tc --dtype float32 --quantize torchchat/quant_config/cuda-32.json --checkpoint "./checkpoints/${REPO_NAME}/model.pth"
# python3 torchchat.py generate --snap model.tc --dtype float32 --checkpoint "./checkpoints/${REPO_NAME}/model.pth"
# echo "::endgroup::"

test-gpu-aoti-float16:
permissions:
id-token: write
Expand Down Expand Up @@ -376,10 +391,15 @@ jobs:
echo "::group::Run inference with quantize file"
if [ $(uname -s) == Darwin ]; then
python3 torchchat.py export --output-aoti-package-path /tmp/model.pt2 --quantize torchchat/quant_config/cuda.json --checkpoint "./checkpoints/${REPO_NAME}/model.pth"
python3 torchchat.py generate --aoti-package-path /tmp/model.pt2 --checkpoint "./checkpoints/${REPO_NAME}/model.pth"~
python3 torchchat.py generate --aoti-package-path /tmp/model.pt2 --checkpoint "./checkpoints/${REPO_NAME}/model.pth"~
fi
echo "::endgroup::"

# echo "::group::Run inference with quantize file"
# python3 torchchat.py export --output-snap model.tc --dtype float16 --quantize torchchat/quant_config/cuda-32.json --checkpoint "./checkpoints/${REPO_NAME}/model.pth"
# python3 torchchat.py generate --snap model.tc --dtype float16 --checkpoint "./checkpoints/${REPO_NAME}/model.pth"
# echo "::endgroup::"

test-gpu-eval-sanity-check:
permissions:
id-token: write
Expand Down Expand Up @@ -495,10 +515,11 @@ jobs:
python3 torchchat.py generate --checkpoint-path ${MODEL_PATH} --temperature 0 --pte-path ${MODEL_DIR}/${MODEL_NAME}.pte

echo "******************************************"
echo "*** --quantize torchchat/quant_config/mobile.json ***"
echo "*** can't test --quantize torchchat/quant_config/mobile.json ***"
echo "*** testing --quantize torchchat/quant_config/mobile-32.json ***"
echo "******************************************"
# python torchchat.py export --quantize torchchat/quant_config/mobile.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 "******************************************"
Expand Down
Loading