hum-ma/flux.1-lite-8B-GGUF

Quantization vs Freepik/flux.1-lite-8B. hum-ma/flux.1-lite-8B-GGUF is a GGUF quantization (Q8_0) of the same weights: 322 tensors match exactly and 194 within quantization error.

Tensors changed: Norms 0/108, Attention 0/64, MLP 0/64, other 0/280

On the Hub · Downloads on Model Pulse