unslothai/unsloth

Single token output / binary classification loss goes to 0

Open

#946 opened on Aug 22, 2024

View on GitHub
 (14 comments) (0 reactions) (0 assignees)Python (5,658 forks)batch import
good first issue

Repository metrics

Stars
 (64,271 stars)
PR merge metrics
 (Avg merge 3d 15h) (525 merged PRs in 30d)

Description

Hello, I get the error in the title when finetuning Phi3.5.
I believe I'm on the latest unsloth (installed from git wit pip).

Context: finetuning Phi3.5 with code that already works with other unsloth models.

Let me know if I can help and how to provide more information as needed.

Contributor guide

Research direction
Investigate why the model outputs a single token; check the model's forward pass and loss computation. Verify that the dataset provides proper labels and that the model's head is configured correctly.
Tech stack
python
Domain
machine learningai
Issue type
Bug
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
3
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
1-2 days
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
Active
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Needs investigation
Prerequisites
PythonPyTorch
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
55