mradermacher/Huihui-Qwen3-VL-8B-Instruct-Creative-v0.4-GGUF

Same weights, converted (fewer layers) vs WarlordHermes/Huihui-Qwen3-VL-8B-Instruct-Creative-v0.4. mradermacher/Huihui-Qwen3-VL-8B-Instruct-Creative-v0.4-GGUF matches part of WarlordHermes/Huihui-Qwen3-VL-8B-Instruct-Creative-v0.4 (399 of its 750 tensors; the rest of WarlordHermes/Huihui-Qwen3-VL-8B-Instruct-Creative-v0.4 has no counterpart). On the shared part: Same weights stored in another dtype (BF16 → F16/F32).

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

On the Hub · Downloads on Model Pulse