neighborwang/ModeliCo-15B
LoRA-style vs bigcode/starcoder2-15b. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of c_fc, c_proj, k_proj, o_proj, q_proj, v_proj changed, in 40 of 40 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/162, MLP 80/160, Attention 160/320, lm_head 0/1
2 derivatives checked
- mradermacher/ModeliCo-15B-i1-GGUF: Quantization
- mradermacher/ModeliCo-15B-GGUF: Quantization