-
Notifications
You must be signed in to change notification settings - Fork 101
Add script for for finetuning a code llm #69
base: main
Are you sure you want to change the base?
Conversation
@@ -83,13 +84,18 @@ def main(): | |||
retrieved = retriever.invoke(item[args.question_field]) | |||
item["retrieved"] = [] | |||
for doc_idx, doc in enumerate(retrieved): | |||
pprint.pprint(doc) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
These documents can get extremely long. Could you just use logging.info() here, and then set the appropriate logging level whenever you want to see the documents?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
stray print. removing
@@ -0,0 +1,447 @@ | |||
import argparse |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Could you leave a comment at the top of this file saying where it was originally forked from and how it's different from the fork?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I had it below the imports but i can move up
Also, it seems like my fix for the github action didn't work. Could you run the following before I get round to fixing it:
|
What does this PR do?
This PR provides basic support for finetuning a code LLM. It is adapted from the following source: https://huggingface.co/learn/cookbook/en/fine_tuning_code_llm_on_single_gpu
NOTE: we still need to add appropriate eval metrics (of which there aren't very many standard ones for finetuning code llms)