mradermacher/Qwen3.8-2B-Heretic-Balanced-GGUF

Same weights, converted vs MihaiPopa-1/Qwen3.8-2B-Heretic-Balanced. Same weights stored in another dtype (BF16 → F16/F32), plus 18 new tensors.

Tensors changed: Embeddings 0/1, Norms 0/61, MLP 0/72, Attention 0/24

On the Hub · Downloads on Model Pulse