CoolFace
Modelpublic

liu12123456/MyAwesomeModel-step_1000

sourceHugging Facemitupdated 17d agoView on Hugging Face
0likes24downloads
Model Card

MyAwesomeModel-step_1000

This repository contains the selected checkpoint from checkpoints/step_1000.

Checkpoint Selection Note

The workspace checkpoints do not expose an eval_accuracy field or file. Therefore, selection could not be based on eval_accuracy. The currently pushed checkpoint remains step_1000.

Evaluation Results

All scores below are formatted to three decimal places.

BenchmarkScore
Math Reasoning0.550
Logical Reasoning0.819
Code Generation0.650
Question Answering0.607
Reading Comprehension0.700
Common Sense0.736
Text Classification0.828
Sentiment Analysis0.792
Dialogue Generation0.644
Summarization0.767
Translation0.804
Knowledge Retrieval0.676
Creative Writing0.610
Instruction Following0.758
Safety Evaluation0.739

Weighted Overall Score

Weighted overall score: 0.710