temaq-org/Tema_Q-X7-Thinking
Partial fine-tune vs ornith-ai/Ornith-1.5-35B-A3B. Partial change: 3% of the weights changed (attention, embeddings, mlp, norms, other), in 35 of 40 layers.
Tensors changed: Embeddings 1/5, Norms 6/247, Attention 56/392, MLP 777/1165, lm_head 0/1, other 1/1
2 derivatives checked
- mradermacher/Tema_Q-X7-Thinking-i1-GGUF: Quantization
- mradermacher/Tema_Q-X7-Thinking-GGUF: Quantization