mradermacher/Taiyi-CLIP-RoBERTa-102M-ViT-L-Chinese-GGUF

Same weights, converted vs IDEA-CCNL/Taiyi-CLIP-RoBERTa-102M-ViT-L-Chinese. Same weights stored in another dtype (F32/I64 → F16/F32).

Tensors changed: Embeddings 0/5, Norms 0/48, Attention 0/96, other 0/48

On the Hub · Downloads on Model Pulse