Add original FP32 Krea 2 Real and HD VAE alternatives with provenance

Remove redundant renamed Quanto text encoder

Add language-only BF16 and Quanto Krea2 text encoders

Add standalone BF16 Qwen3-VL vision tower

Update README.md

Upload Krea2_TextFusion_Refusal_Reduction.safetensors

Add full-rank Krea 2 Identity Edit v1.2 LoRA

Add Qwen3-VL image processor config for Krea Edit

Upload krea2bypass_balanced.safetensors

Create loras/readme.txt

Upload README.md with huggingface_hub

Upload krea2/krea2_transformer_config.json with huggingface_hub

Upload Qwen3-VL-4B-Instruct/chat_template.jinja with huggingface_hub

Upload Qwen3-VL-4B-Instruct/tokenizer_config.json with huggingface_hub

Upload Qwen3-VL-4B-Instruct/tokenizer.json with huggingface_hub

Upload Qwen3-VL-4B-Instruct/config.json with huggingface_hub

Upload Qwen3-VL-4B-Instruct/Qwen3-VL-4B-Instruct_quanto_bf16_int8.safetensors with huggingface_hub

Upload Qwen3-VL-4B-Instruct/Qwen3-VL-4B-Instruct_bf16.safetensors with huggingface_hub

Upload Krea2Turbo_quanto_bf16_int8.safetensors with huggingface_hub

Upload Krea2Turbo_bf16.safetensors with huggingface_hub

Upload Krea2Raw_quanto_bf16_int8.safetensors with huggingface_hub

Upload Krea2Raw_bf16.safetensors with huggingface_hub

initial commit