support multiple device evaluation for activation quantized model #1394
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Contributor
There was a problem hiding this comment.
Pull request overview
This PR addresses test and code improvements by commenting out multiple test cases and adding new device management functionality.
Changes:
- Commented out extensive test cases in the scheme test file to focus on a single
test_set_schemetest - Added a new
dispatch_model_block_wiseutility function for multi-device model dispatching - Updated evaluation code to use the new device dispatch mechanism
- Changed a logging level from warning to info and added
tie_weights()calls in multiple locations
Reviewed changes
Copilot reviewed 5 out of 5 changed files in this pull request and generated 3 comments.
Show a summary per file
| File | Description |
|---|---|
| test/test_cpu/schemes/test_scheme.py | Commented out multiple test cases, leaving only test_set_scheme active |
| auto_round/utils/device.py | Added new dispatch_model_block_wise function for block-wise model dispatching across devices |
| auto_round/eval/evaluation.py | Updated prepare_model_for_eval to use new device dispatch utility and changed parameter name |
| auto_round/compressors/base.py | Changed logger level from warning to info, added tie_weights() call, and modified error raising |
| auto_round/auto_scheme/utils.py | Added tie_weights() call before device map inference |
wenhuach21
changed the title
for more information, see https://pre-commit.ci
…ix_0204 # Conflicts: # auto_round/eval/eval_cli.py
for more information, see https://pre-commit.ci
for more information, see https://pre-commit.ci
Signed-off-by: n1ck-guo <heng.guo@intel.com>
xin3he
reviewed
Feb 10, 2026
| # temporary tensors, other processes, and allocator fragmentation, reducing | ||
| # the chance of runtime OOM while still utilizing most available memory. | ||
| new_max_memory[device] = max_memory[device] * max_mem_ratio | ||
| new_max_memory = get_balanced_memory( |
Contributor
There was a problem hiding this comment.
In my memory, this will use all CUDA_VISIBLE_DEVICES. Setting CUDA_VISIBLE_DEVICES with device_map information might help.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
Please briefly describe your main changes, the motivation.
Type of Change
Related Issues
Fixes or relates to #
Checklist Before Submitting