TeeZee/Qra-13b-instruct
LoRA-style vs OPI-PG/Qra-13b. 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 40 of 40 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/81, MLP 120/120, Attention 160/160, lm_head 0/1
4 derivatives checked
- TeeZee/Qra-13B-chat: LoRA-style
- mradermacher/Qra-13b-instruct-i1-GGUF: Quantization
- mradermacher/Qra-13b-instruct-GGUF: Quantization
- featherless-ai-quants/TeeZee-Qra-13b-instruct-GGUF: Quantization