RecursiveMAS/Mixture-Summarizer-Qwen3.5-2B
Same weights, converted (fewer layers) vs Qwen/Qwen3.5-2B. RecursiveMAS/Mixture-Summarizer-Qwen3.5-2B matches part of Qwen/Qwen3.5-2B (320 of its 632 tensors; the rest of Qwen/Qwen3.5-2B has no counterpart). On the shared part: Same weights stored in another dtype (BF16/F32 → BF16).
Tensors changed: Attention 0/168, Norms 0/79, Embeddings 0/1, MLP 0/72
2 derivatives checked
- mradermacher/Mixture-Summarizer-Qwen3.5-2B-i1-GGUF: Quantization
- mradermacher/Mixture-Summarizer-Qwen3.5-2B-GGUF: Same weights, converted
4 other repos hold byte-for-byte identical weight files.