valerielucro
/

mistral_gsm8k_preference_dataset_v2_beta_3

Inference Endpoints

Model card Files Files and versions Community

mistral_gsm8k_preference_dataset_v2_beta_3 / tokenizer.json

Commit History

second iteration Qlora with DPO on full gsm8k preference dataset version 2.1 and 1 epoch and rank 64, beta 0.3

a6e46a4
verified

valerielucro commited on Jun 29