Edit model card

Pythia-410m supervised finetuned with Anthropic-hh-rlhf dataset for 1 epoch (sft-model), before DPO (paper) with same dataset for 1 epoch.

wandb log

Benchmark evaluations included in repo done using lm-evaluation-harness.

See Pythia-410m for original model details (paper).

Downloads last month
3
Safetensors
Model size
506M params
Tensor type
F32
·
BOOL
·

Dataset used to train lomahony/eleuther-pythia410m-hh-dpo

Collection including lomahony/eleuther-pythia410m-hh-dpo