Qwen/Qwen3-Reranker-8B
Full fine-tune vs Qwen/Qwen3-8B-Base. Fine-tune: 100% of the weights changed, across 36 of 36 layers.
Tensors changed: Norms 143/145, MLP 108/108, Attention 144/144
12 derivatives checked
- mradermacher/Qwen3-Reranker-8B-GGUF: Same weights, converted
- furiosa-ai/Qwen3-Reranker-8B: Identical copy
- Voodisss/Qwen3-Reranker-8B-GGUF-llama_cpp: Same weights, converted
- mradermacher/Qwen3-Reranker-8B-i1-GGUF: Quantization
- giladgd/Qwen3-Reranker-8B-GGUF: Same weights, converted
- DevQuasar/Qwen.Qwen3-Reranker-8B-GGUF: Same weights, converted
- ethicalabs/Qwen3-Reranker-8B-GGUF: Quantization
- PXY107kWqS/Qwen3-Reranker-8B-bnb-4bit: Quantization
- drawais/Qwen3-Reranker-8B-AWQ-INT4: Quantization
- ebasKing/Qwen3-Reranker-8B-GGUF-llama_cpp: Same weights, converted
- Horkios/Qwen3-Reranker-8B-mlx-4Bit: Quantization
- galaxycore/Qwen3-Reranker-8B-MLX-4bit: Quantization
6 other repos hold byte-for-byte identical weight files.