win10/Blue-Rose-Coder-12.3B-Instruct
Full fine-tune (fewer layers) vs Qwen/Qwen2.5-Coder-7B-Instruct. Qwen/Qwen2.5-Coder-7B-Instruct matches part of win10/Blue-Rose-Coder-12.3B-Instruct (339 of its 579 tensors; the rest of win10/Blue-Rose-Coder-12.3B-Instruct has no counterpart). On the shared part: Fine-tune: 93% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 1/1, Norms 56/57, MLP 84/84, Attention 196/196, lm_head 0/1
2 derivatives checked
- mradermacher/Blue-Rose-Coder-12.3B-Instruct-i1-GGUF: Quantization
- mradermacher/Blue-Rose-Coder-12.3B-Instruct-GGUF: Quantization