-
Notifications
You must be signed in to change notification settings - Fork 35
Open
Labels
enhancementNew feature or requestNew feature or request
Description
🎯 Goal (What & Why)
Support chat template during dataset preparation to make it easier for SFT, DPO and other instruction finetuning methods. This takes away from the user the onus of preparing datasets and being careful about constructing loss masking spans, eos and bos tokens. e.g., OpenRLHF directly works with chat templates.
🚀 Execution Plan
- Work off of Combine GPTHuggingfaceDatasetConfig input sources into
source_schema
#255 or Concat prompt and completion cols for tokenizing #257 - Mask text from all sources except for
assistant
- Add an option to unroll turns i.e., multiple loss masking spans per conversation if
unroll=False
, else 1 prompt-response pair per conversation turn, with the prompt masked.
📌 Acceptance Criteria (Must-Haves for Completion)
- The feature must be functional and tested.
- The implementation must be documented in practical terms.
- No refactors unless directly necessary for feature completion.
🛠️ Project Management
- Assign the project to the Fast-LLM project.
- Set the
Estimate
field (in days) in the GitHub project. - Use the
Size
field to categorize the PR size (Small/Medium/Large). - Assign an owner when opening the issue.
Metadata
Metadata
Assignees
Labels
enhancementNew feature or requestNew feature or request