AllSpark-Research/Iris-mini
Full fine-tune (fewer layers) vs Qwen/Qwen3.6-35B-A3B. Qwen/Qwen3.6-35B-A3B matches part of AllSpark-Research/Iris-mini (1043 of its 1811 tensors; the rest of AllSpark-Research/Iris-mini has no counterpart). On the shared part: Fine-tune: 99% of the weights changed, across 40 of 40 layers.
Tensors changed: Embeddings 1/5, Attention 245/392, MLP 285/397, Norms 101/247, lm_head 1/1, other 1/1
10 derivatives checked
- IsValorum/Iris-mini-APEX-I-MiniPlus-V2.1-GGUF: Quantization
- IsValorum/Iris-mini-APEX-I-MiniPlus-V1-GGUF: Quantization
- nightmedia/Qwen3.6-35B-A3B-Brainwaves-Nex-Uncensored: Full fine-tune
- mradermacher/Iris-mini-GGUF: Quantization
- mradermacher/Iris-mini-i1-GGUF: Quantization
- nightmedia/Qwen3.6-35B-A3B-Brainwaves-Nex-Uncensored-qx64-hi-mlx: Quantization
- felkf/Iris-mini-oQ6e-fp16-mtp: Quantization
- nightmedia/Qwen3.6-35B-A3B-Brainwaves-Nex-qx64-hi-mlx: Quantization
- nightmedia/Qwen3.6-35B-A3B-Brainwaves-qx64-hi-mlx: Quantization
- nightmedia/Qwen3.6-35B-A3B-Brainwaves-Nex: Full fine-tune