<h2 id="qwen-caption-(6gb)-uncensored">QWEN Caption (6GB) Uncensored</h2><ul><li>Full FP32 Training (NO 8BIT Optimizers)</li><li>1024px vision input with 256-1024 token length captions (4096 token limit)</li><li>If you wish to use the full BF16 or FP32 model change the script target to "Felldude/Qwen3-VL-4B-Instruct-Uncensored"</li></ul><hr />Note: This model would likely work well with Krea.2 but is untested

Files with SHA256: 7dd782dfe35df5313818788ca6e5c78dd53324234d06ba0028133aea8df92f0c