ukisai/Swift-Qwen3.8-27B-GGUF

Same weights, converted vs ukisai/Swift-Qwen3.8-27b. Same weights stored in another dtype (BF16 → F16/F32), plus 63 new tensors.

Tensors changed: Norms 0/161, MLP 0/192, Attention 0/64, Embeddings 0/1, lm_head 0/1

1 derivatives checked

• jpetrina/Swift-Qwen3.8-27B-IQ4_XS-pure-GGUF: Quantization

On the Hub · Downloads on Model Pulse