Jackrong/Qwopus3.6-35B-A3B-v1
LoRA-style vs unsloth/Qwen3.6-35B-A3B. 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 39 of 40 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/5, Attention 40/392, MLP 120/399, Norms 0/247, lm_head 0/1, other 0/1
11 derivatives checked
- Jackrong/Qwopus3.6-35B-A3B-Coder-MTP-GGUF: Full fine-tune
- mudler/Qwopus3.6-35B-A3B-v1-APEX-MTP-GGUF: Same weights, converted
- mudler/Qwopus3.6-35B-A3B-v1-APEX-GGUF: Quantization
- Jackrong/Qwopus3.6-35B-A3B-Coder: LoRA-style
- barozp/Qwopus3.6-35B-A3B-v1-GGUF: Same weights, converted
- mradermacher/Qwopus3.6-35B-A3B-v1-i1-GGUF: Quantization
- mradermacher/Qwopus3.6-35B-A3B-v1-GGUF: Quantization
- Jackrong/Qwopus3.6-35B-A3B-Coder-FP8: Quantization
- noctrex/Qwopus3.6-35B-A3B-v1-MXFP4_MOE-GGUF: Same weights, converted
- cyburn/Qwopus3.6-35B-A3B-v1-PrismaSCOUT-Blackwell-NVFP4-BF16-vllm-4.75bits: Quantization
- jamiefutch/Qwopus3.6-35B-A3B-v1-MTP-MXFP4_MOE-GGUF: Same weights, converted
1 other repos hold byte-for-byte identical weight files.