akumaburn/Swift-Qwen3.8-27b-heretic
LoRA-style vs Qwen/Qwen3.8-27B. 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, out_proj, q_proj, up_proj, v_proj changed, in 63 of 64 layers; norms, biases and embeddings are untouched.
Tensors changed: Norms 0/325, Attention 76/560, MLP 192/307, Embeddings 0/5, lm_head 0/1, other 0/1
5 derivatives checked
- mradermacher/Swift-Qwen3.8-27b-heretic-GGUF: Quantization
- mradermacher/Swift-Qwen3.8-27b-heretic-i1-GGUF: Quantization
- akumaburn/Swift-Qwen3.8-27b-heretic-W8A8-DFlash2: Quantization
- akumaburn/Swift-Qwen3.8-27b-heretic-W4A16: Quantization
- akumaburn/Swift-Qwen3.8-27b-heretic-SmoothQuant-W8A8-INT8: Quantization