linuxid10t/G9v3-39A5B-GGUF
Same weights, converted vs ai9stars/G9v3-39A5B. Same weights stored in another dtype (BF16/F32 → BF16/F32), plus 186 new tensors.
Tensors changed: lm_head 0/1, Embeddings 0/1, Norms 0/77, MLP 0/151, Attention 0/114
Same weights, converted vs ai9stars/G9v3-39A5B. Same weights stored in another dtype (BF16/F32 → BF16/F32), plus 186 new tensors.
Tensors changed: lm_head 0/1, Embeddings 0/1, Norms 0/77, MLP 0/151, Attention 0/114