parkjo/Llama-3.2-3B-Instruct_grpo_ppl_adv_rollout_8_Use_KL_0.001_step580

Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:3.2BQuant:BF16Ctx Length:32kPublished:Apr 30, 2026Architecture:Transformer Warm

Loading preview...