mradermacher/Qwen-3-VL-8B-Instruct-heretic-GGUF

Same weights, converted (fewer layers) vs heretic-org/Qwen-3-VL-8B-Instruct-heretic. mradermacher/Qwen-3-VL-8B-Instruct-heretic-GGUF matches part of heretic-org/Qwen-3-VL-8B-Instruct-heretic (399 of its 750 tensors; the rest of heretic-org/Qwen-3-VL-8B-Instruct-heretic has no counterpart). On the shared part: Same weights stored in another dtype (BF16 → F16/F32).

Tensors changed: lm_head 0/1, Embeddings 0/1, Norms 0/145, MLP 0/108, Attention 0/144

On the Hub · Downloads on Model Pulse