Chime Ogbuji 95e1f22812 Incorporate use of response template for completion masking ... Follow example of trl's DataCollatorForCompletionOnlyLM to use response template to identify beginning of completion/continuation tokens for the purpose of masking out the other tokens during loss calculation		2025-02-09 07:43:04 -08:00
..
__init__.py	LoRA: Extract small function (#614)	2024-06-02 06:38:42 -07:00
datasets.py	Add response template (or token) argument	2025-02-09 07:43:01 -08:00
dora.py	Feature: QDoRA (#891)	2024-09-30 08:01:11 -07:00
lora.py	Allow the entire model to be targed for LoRA and DoRA fine tuning: LoRA and DoRA embeddings with small DoRALinear bug fix (#914)	2024-08-16 07:38:36 -07:00
trainer.py	Incorporate use of response template for completion masking	2025-02-09 07:43:04 -08:00
utils.py	Add IBM granite model (#1265)	2025-02-08 15:46:15 -08:00