mradermacher/boomerang-qwen3-4.9B-GGUF

Same weights, converted vs Harvard-DCML/boomerang-qwen3-4.9B. Same weights stored in another dtype (F32 → F16/F32).

Tensors changed: Embeddings 0/1, Norms 0/77, MLP 0/57, Attention 0/76, lm_head 0/1

On the Hub · Downloads on Model Pulse