Fix: Disable Sparse Decompression for Dense Compressors #237
Conversation
This file contains 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
rahul-tuli
changed the title
mgoin
previously approved these changes
Jan 10, 2025
src/compressed_tensors/compressors/model_compressors/model_compressor.py
Outdated
Show resolved
Hide resolved
kylesayrs
requested changes
Jan 10, 2025
src/compressed_tensors/compressors/model_compressors/model_compressor.py
Outdated
Show resolved
Hide resolved
rahul-tuli
force-pushed
the
turn-off-sparse-decompression-when-dense
branch
from
ff2b26a to
51cee4e
Compare
horheynm
approved these changes
Jan 10, 2025
rahul-tuli
force-pushed
the
turn-off-sparse-decompression-when-dense
branch
from
51cee4e to
cc4f78e
Compare
kylesayrs
approved these changes
Jan 10, 2025
dsikka
added a commit
to vllm-project/llm-compressor
that referenced
this pull request
Jan 23, 2025
~~Contingent on merge of huggingface/transformers#34719 ~~ ^ has been merged not yet released ~~ ^ has been released Blocked on neuralmagic/compressed-tensors#237 SUMMARY: * In multiple optimization tests, automatically decompress model if provided as optimized model * Fix recipe stage length * Revive old code * When running multiple optimizations (ex. oneshot then finetune, oneshot and oneshot), the recipes needs to be added to the session using `initialize_recipe`. Example here https://github.com/vllm-project/llm-compressor/pull/971/files#diff-c9ae8b3ad24d13abeea5b649a5fd6d0b0925f5c9cc40220cbfbe21ae81242f8dR63-R65 TEST PLAN: ran the test using transformers main Must pass tests/llmcompressor/transformers/obcq/test_consecutive_runs.py --------- Co-authored-by: Dipika Sikka <[email protected]> Co-authored-by: Rahul Tuli <[email protected]>
rahul-tuli
added a commit
to vllm-project/llm-compressor
that referenced
this pull request
Jan 28, 2025
~~Contingent on merge of huggingface/transformers#34719 ~~ ^ has been merged not yet released ~~ ^ has been released Blocked on neuralmagic/compressed-tensors#237 SUMMARY: * In multiple optimization tests, automatically decompress model if provided as optimized model * Fix recipe stage length * Revive old code * When running multiple optimizations (ex. oneshot then finetune, oneshot and oneshot), the recipes needs to be added to the session using `initialize_recipe`. Example here https://github.com/vllm-project/llm-compressor/pull/971/files#diff-c9ae8b3ad24d13abeea5b649a5fd6d0b0925f5c9cc40220cbfbe21ae81242f8dR63-R65 TEST PLAN: ran the test using transformers main Must pass tests/llmcompressor/transformers/obcq/test_consecutive_runs.py --------- Co-authored-by: Dipika Sikka <[email protected]> Co-authored-by: Rahul Tuli <[email protected]> Signed-off-by: Rahul Tuli <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Problem
When the sparse compressor is set to
"dense", sparse decompression is incorrectly triggered, causing uninitialized weights and downstream errors.Example CI failure: [GitHub Actions Log](https://github.com/vllm-project/llm-compressor/actions/runs/12659596814/job/35326229412).
Solution
Added a condition to skip sparse decompression when the sparsity configuration format is
"dense".Testing
llm-compressormain commit:03e21770.