models
Open weights, fine-tunes and adapters. Every listing here comes live from the Hugging Face Hub, attributed to it, and links back to the source.
FK_reversecowgirl_klein9bQwen2.5-VL-7B-OPSD-VisionZip-r010-Plain-ReverseKL-original10240Qwen2.5-VL-7B-OPSD-VisionZip-r010-BudgetJSD-Top10-ReverseKL-original10240Qwen2.5-VL-7B-OPSD-VisionZip-r010-TIP-Top20-ReverseKL-original10240Qwen2.5-VL-7B-OPSD-VisionZip-r010-BudgetJSD-Top10-ReverseKL-FullFT-original10240-FP16Qwen2.5-VL-7B-OPSD-VisionZip-r010-ReverseKL-FullFT-original10240-FP16Qwen2.5-VL-7B-OPSD-VisionZip-r010-BTIP-Top20-ReverseKL-gap01-original10240Qwen2.5-VL-7B-OPSD-VisionZip-r010-BTIP-Top20-ReverseKL-gap02-original10240Qwen2.5-VL-7B-OPSD-VisionZip-r010-KL-Top20-ReverseKL-original10240Qwen2.5-VL-7B-OPSD-VisionZip-r010-BudgetJSD-Top20-ReverseKL-gap01-original10240Qwen2.5-VL-7B-OPSD-VisionZip-r010-TeacherGT-ReverseKL-original10240sppo_reversekl-0.1-Gemma-2-2B-IT-RSPO-Iter1sppo_reversekl-0.1-Gemma-2-2B-IT-RSPO-Iter2FK_reversecowgirl_klein9bSMOL_DPO_REVERSE_KL_1_0-checkpoint-150Llama-3-8B-Instruct-SPPO-Iter1-gp-8b-gpm-reg0.05-sppo-reversekl-tablefeedback_disallowed_ema_Qwen2.5-7B-Instruct_reverse_kl_ema0p999_ep30ci-feedback_both_ema_Qwen2.5-7B-Instruct_reverse_kl_ema0p999_ep30SMOL_DPO_REVERSE_KL_0_05-checkpoint-100Llama-3-8B-Instruct-SPPO-Iter2-gp-8b-gpm-reg0.05-sppo-reversekl-tableci-feedback_disallowed_ema_Qwen2.5-7B-Instruct_reverse_kl_ema0p999_ep30feedback_both_ema_Qwen2.5-7B-Instruct_reverse_kl_ema0p999_ep30Qwen3-8B-on-policy-distill-reverse-kld-llava-pretrainDPO-reverse_kl_beta_1_0sppo_reversekl-0.1-Gemma-2-2B-IT-RSPO-Iter3feedback_disallowed_ema_DeepSeek-R1-Distill-Llama-8B_reverse_kl_ema0p999_ep30feedback_allowed_ema_Qwen2.5-7B-Instruct_reverse_kl_ema0p999_ep30feedback_both_ema_Qwen3-4B-Instruct-2507_reverse_kl_ema0p999_ep30feedback_disallowed_ema_Qwen3-4B_reverse_kl_ema0p999_ep30Qwen3-8B-on-policy-distill-reverse-kld
