Pythia-160m supervised finetuned with Anthropic-hh-rlhf dataset for 1 epoch.
Benchmark evaluations included in repo done using lm-evaluation-harness.
See Pythia-160m for model details (paper).
Pythia-160m supervised finetuned with Anthropic-hh-rlhf dataset for 1 epoch.
Benchmark evaluations included in repo done using lm-evaluation-harness.
See Pythia-160m for model details (paper).