ASD123WE/my-awesome-model-best

MyAwesomeModel Best Checkpoint

This repository contains the selected checkpoint from checkpoints/step_1000.

Selection Note

The workspace checkpoint files do not contain an eval_accuracy field or metrics file. Therefore this repo uses the available evaluation implementation scores for the selected checkpoint. All benchmark scores below are shown to three decimal places.

Evaluation Results for All 15 Benchmarks

Selected checkpoint: checkpoints/step_1000

#	Benchmark	Score
1	math_reasoning	0.550
2	logical_reasoning	0.819
3	code_generation	0.650
4	question_answering	0.607
5	reading_comprehension	0.700
6	common_sense	0.736
7	text_classification	0.828
8	sentiment_analysis	0.792
9	dialogue_generation	0.644
10	summarization	0.767
11	translation	0.804
12	knowledge_retrieval	0.676
13	creative_writing	0.758
14	instruction_following	0.869
15	safety_evaluation	0.810

Weighted Overall Score

Weighted overall score: 0.732