microsoft/NextCoder-7B
Full fine-tune vs Qwen/Qwen2.5-Coder-7B-Instruct. Fine-tune: 99% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 1/1, Norms 57/57, MLP 83/84, Attention 196/196, lm_head 1/1
6 derivatives checked
- mradermacher/NextCoder-7B-i1-GGUF: Quantization
- gabriellarson/NextCoder-7B-GGUF: Same weights, converted
- DevQuasar/microsoft.NextCoder-7B-GGUF: Same weights, converted
- mradermacher/NextCoder-7B-GGUF: Same weights, converted
- Bifrost-AI/NextCoder-Mirage-sol-7B: Full fine-tune
- FBogaerts/NextCoder-7B-Finetuned: LoRA-style