Tiiny/SmallThinker-21BA3B-Instruct-GGUF

Same weights, converted vs Tiiny/SmallThinker-21BA3B-Instruct. Same weights stored in another dtype (BF16 → F16/F32), plus 156 new tensors.

Tensors changed: Embeddings 0/1, other 0/52, Norms 0/105, Attention 0/208, lm_head 0/1

On the Hub · Downloads on Model Pulse