arsoban/ocd-therapist-27b-v0.3
LoRA-style vs google/gemma-3-27b-it. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, fc1, fc2, gate_proj, k_proj, o_proj, out_proj, q_proj, up_proj, v_proj changed, in 62 of 62 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/4, Norms 0/484, MLP 214/294, Attention 318/464, other 0/1
2 derivatives checked
- mradermacher/ocd-therapist-27b-v0.3-i1-GGUF: Quantization (fewer layers)
- mradermacher/ocd-therapist-27b-v0.3-GGUF: Quantization (fewer layers)