[TorchACC]support batch flattening collator #2499
Merged
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
eedalong
force-pushed
the
support_batch_flattening
branch
from
396032c to
148e85e
Compare
Jintao-Huang
approved these changes
Nov 25, 2024
tastelikefeet
added a commit
that referenced
this pull request
Dec 3, 2024
…actor3 * commit '2bbc325ca789592197d2004bb0ffc47cc39c0317': (140 commits) fix fix update safe_ddp_context fix fix update row_processor support glm-edge & glm-edge-v (#2526) fix open-o1 support qwq-32b-preview (#2520) support mPLUG-Owl3 241101 (#2515) fix latex-ocr (#2510) support batch flattening collator (#2499) fix eval_dataset no (#2497) Support marco o1 (#2496) Fix preprocess num proc (#2492) fix awq quant device_map (#2488) Update Common QA (#2475) fix kto (#2478) update padding strategy for persistent cache (#2464) fix qwen2vl pt infer (#2463) ... # Conflicts: # docs/source/Instruction/命令行参数.md # docs/source/LLM/人类偏好对齐训练文档.md # docs/source/Multi-Modal/index.md # docs/source/Multi-Modal/qwen2-vl最佳实践.md # docs/source/Multi-Modal/人类偏好对齐训练文档.md # docs/source_en/Instruction/Command-line-parameters.md # docs/source_en/Instruction/Common-QA.md # docs/source_en/LLM/Human-Preference-Alignment-Training-Documentation.md # docs/source_en/Multi-Modal/qwen2-vl-best-practice.md
Jintao-Huang
changed the title
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Outdated suggestions cannot be applied.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.
PR type
PR information
Support batch flattening to avoid padding for better training performance and less memory usage. The function has the same behaviour with DataCollatorWithFlattening
Experiment results
Effectively reduce peak memory for ~25% and 50% maximum training throughput improvement.