unsloth/ERNIE-Image-Turbo-GGUF

Same weights, converted vs baidu/ERNIE-Image-Turbo. Same weights stored in another dtype (BF16 → F16/BF16).

Tensors changed: other 0/41, MLP 0/144, Norms 0/74, Attention 0/144, Embeddings 0/6

On the Hub · Downloads on Model Pulse