Repository navigation
Conversation
…ggingFaceH4/ultrachat_200k Signed-off-by: chensuyue <suyue.chen@intel.com>
Contributor
There was a problem hiding this comment.
🟡 Changes recommended
The default path enables remote dataset code execution, and the indentation fix lacks regression coverage.
2 open findings
What changed in this PR
Updates AutoRound’s default calibration dataset to UltraChat and fixes chat-template preprocessing.
Changes:
- Replaces Pile-10k defaults across calibration, AutoScheme, export, and VLM templates.
- Corrects UltraChat chat-template indentation.
- Updates English and Chinese documentation.
| File | Description |
|---|---|
README.md |
Documents the new default. |
README_CN.md |
Mirrors README changes in Chinese. |
docs/step_by_step.md |
Updates dataset guidance. |
docs/step_by_step_CN.md |
Updates Chinese dataset guidance. |
auto_round/export/utils.py |
Updates export configuration filtering. |
auto_round/compressors/mllm/templates/phi3_v.json |
Updates Phi-3-Vision default. |
auto_round/compressors/mllm/templates/llava.json |
Updates LLaVA default. |
auto_round/compressors/mllm/templates/default.json |
Updates generic VLM default. |
auto_round/compressors/mllm/templates/cogvlm2.json |
Updates CogVLM2 default. |
auto_round/compressors/mllm/template.py |
Updates registered VLM defaults. |
auto_round/compressors/mllm/README.md |
Documents the VLM default. |
auto_round/compressors/diffusion_mixin.py |
Preserves diffusion dataset substitution. |
auto_round/compressors/base.py |
Changes the core calibration default. |
auto_round/calib_dataset.py |
Changes helper defaults and fixes indentation. |
auto_round/autoround.py |
Updates the internal builder default. |
auto_round/auto_scheme/gen_auto_scheme.py |
Updates AutoScheme’s dataset default. |
auto_round/auto_scheme/delta_loss.py |
Updates delta-loss dataset defaults. |
🧠 Review effort: Balanced
💡 Configure MCP servers for context-aware, tailored reviews. Learn more in the docs.
|
|
||
|
|
||
| def _get_dataset_impl(tokenizer, seqlen, dataset_name="NeelNanda/pile-10k", seed=42, nsamples=512): | ||
| def _get_dataset_impl(tokenizer, seqlen, dataset_name="HuggingFaceH4/ultrachat_200k", seed=42, nsamples=512): |
| elif not is_instruct and apply_chat_template: | ||
| logger.info("Tokenizer is not an instruct/chat model, but apply_chat_template=True. Setting to False.") | ||
| apply_chat_template = False | ||
| apply_chat_template = False |
…aset function Signed-off-by: chensuyue <suyue.chen@intel.com>
Contributor
|
We shouldn't change the default dataset at the code level unless it leads to accuracy improvements. If there are other issues involved, you should modify the dataset you use instead. |
Collaborator
|
/azp run Unit-Test-CUDA-AutoRound |
|
Azure Pipelines successfully started running 1 pipeline(s). |
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
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.
Description
Type of Change
Bug fix
Related Issues
Fixes or relates to #
Checklist Before Submitting
/azp run Unit-Test-CUDA-AutoRound.