katanemo/Arch-Function-7B
Full fine-tune vs Qwen/Qwen2.5-Coder-7B-Instruct. Fine-tune: 100% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 1/1, Norms 57/57, MLP 84/84, Attention 196/196, lm_head 1/1
7 derivatives checked
- bartowski/Arch-Function-7B-GGUF: Full fine-tune
- mradermacher/Arch-Function-7B-GGUF: Full fine-tune
- mradermacher/Arch-Function-7B-i1-GGUF: Full fine-tune
- katanemo/Arch-Function-7B.gguf: Quantization
- featherless-ai-quants/katanemo-Arch-Function-7B-GGUF: Quantization
- DevQuasar-7/katanemo.Arch-Function-7B-GGUF: Same weights, converted
- MaziyarPanahi/Arch-Function-7B-GGUF: Full fine-tune