lunahr/thea-pro-2b-100r
LoRA-style vs LGAI-EXAONE/EXAONE-3.5-2.4B-Instruct. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of c_fc_0, c_fc_1, c_proj, k_proj, out_proj, q_proj, v_proj changed, in 30 of 30 layers; norms, biases and embeddings are untouched.
Tensors changed: Attention 120/120, Norms 0/61, MLP 90/90, Embeddings 0/1
2 derivatives checked
- mradermacher/thea-pro-2b-100r-i1-GGUF: Quantization
- mradermacher/thea-pro-2b-100r-GGUF: Same weights, converted