Per-PL perplexity vs. pass@k rate

bigcode-project / bigcode-evaluation-harness

A framework for the evaluation of autoregressive code generation language models.

Apache License 2.0

771 stars 201 forks source link

Closed arjunguha closed 1 year ago

arjunguha commented 1 year ago

For @canders1, @ytzi, and me

loubnabnl commented 1 year ago

closing this issue as we finished the scaling laws