zeroentropy/zerank-2-reranker
LoRA-style vs Qwen/Qwen3-4B. 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 36 of 36 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/145, MLP 108/108, Attention 144/144
9 derivatives checked
- mradermacher/zerank-2-GGUF: Same weights, converted
- MarshallHD/zerank-2-reranker-NVFP4: Quantization
- seamon67/Zerank-2-GGUF: Same weights, converted
- liudev/zerank-2-reranker-fp8: Quantization
- k8smee/zerank-2-reranker-mlx-8Bit: Quantization
- fcmeyer/zerank-2-reranker-MLX-8bit: Quantization
- lexrivera/zerank-2-reranker-mlx-6Bit: Quantization
- fcmeyer/zerank-2-reranker-MLX-bf16: Same weights, repacked
- baseten-admin/zerank-2-reranker-seq: Same weights, repacked
1 other repos hold byte-for-byte identical weight files.