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.
sweep-nep-sft_ministral-3-3b_lora_r32_lr2e-4_wd0.01_adamw-8bit_linearunsloth_Qwen2.5-VL-7B-Instruct_lr_0.0002_linear_adamw_torch_fused_True_fp16_v1_16_32_2_linearblip_50k_5epoch_4batch_5e-05lr_AdamWblip_10k_deduped_10epoch_6batch_1e-05lr_AdamW_1e-2wdblip_10k_deduped_5epoch_6batch_5e-05lr_AdamW_2e-2wdblip_10k_deduped_10epoch_6batch_5e-05lr_5ss_2e-1g_AdamW_1e-2wdblip_10k_deduped_horizontalflip_20epoch_6batch_5e-05mlr_AdamW_1e-2wddonut-base-demo-new-3e-05_AdamW_1876code-nep_qwen3.5-4b_lora_r32a32_lr2e4_wd0.01_adamw-torch-fused_cosine_stage1code-nep_qwen3.5-4b_lora_r32a32_lr2e4_wd0.01_adamw-torch-fused_cosine_stage1-95kblip_50k_deduped_5epoch_6batch_5e-05lr_AdamW_1e-2wdblip_10k_deduped_5epoch_6batch_5e-05lr_AdamW_1e-2wdblip_10k_deduped_horizontalflip_5epoch_6batch_5e-05mlr_AdamW_1e-2wdblip_10k_deduped_horizontalflip_222sd2_20epoch_6batch_5e-05mlr_AdamW_1e-2wdblip_10k_deduped_horizontalflip_1106sd2_20epoch_6batch_5e-05mlr_AdamW_1e-2wddonut-base-demo-new-3e-05_AdamW_938donut-base-demo-final_3e-05_AdamWdonut-base-demo-new-v2-3e-05_AdamW_938donut-base-demo-final_v23e-05_AdamWdonut-base-demo-new-v2-3e-05_AdamW_3752donut-base-demo-new-v2-3e-05_AdamW_5159donut-base-demo-new-v2-3e-05_AdamW_6566donut-base-demo-new-v2-3e-05_AdamW_1876sweep-nep-sft_ministral-3-3b_lora_r32_lr2e4_wd0.01_adamw-torch-fused_cosinezeta-recreations_qwen3.5-4b_lora_r128a128_lr0.0002_wd0.01_adamw_torch_fused_cosine_sftunsloth_Qwen2.5-VL-7B-Instruct_lr_0.0002_cosine_adamw_torch_fused_True_fp16_v1_16_32_2
