Benchmarks · Project 4: Fine-tune a model with LoRA
Project 4: Fine-tune a model with LoRA — lora-r4-forbattrar
The measure: ppl_gain_r4 (>= = higher is better). One entry per user (the best run).
| # | Name | Value | When |
|---|---|---|---|
| No entries yet. | |||
Submit a run
Sign in and run the lab in the sandbox.