Quick Links

Dashboard Namespaces Pull Requests Issues Starred

Models Datasets Papers Spaces Collections Experiments Notebooks

Quick Links

Dashboard Namespaces Pull Requests Issues Starred

Models Datasets Papers Spaces Collections Experiments Notebooks

Back to Experiments

llama-finetuning-v3

running

language-models · research-team

Fine-tuning Llama 3.1 on custom instruction dataset

Accuracy

0.924

Loss

0.234

Epoch

12/50

Duration

2h 34m

GPU

A100 x 4

Training Metrics

Metrics visualization (accuracy, loss over time)

Latest Metrics

Train Accuracy

Step 12000.924

Val Accuracy

Step 12000.887

Train Loss

Step 12000.234

Val Loss

Step 12000.289

Learning Rate

Step 12002e-5