prithivMLmods/Raptor-X3
LoRA-style vs prithivMLmods/Viper-OneCoder-UIGEN. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, gate_proj, k_proj, o_proj, q_proj, up_proj, v_proj changed, in 48 of 48 layers; norms, biases and embeddings are untouched.
Tensors changed: lm_head 0/1, Embeddings 0/1, Norms 0/97, MLP 141/144, Attention 188/336
2 derivatives checked
- mradermacher/Raptor-X3-GGUF: Quantization
- DevQuasar-2/prithivMLmods.Raptor-X3-GGUF: Same weights, converted