microsoft/NextCoder-14B

Full fine-tune vs Qwen/Qwen2.5-Coder-14B-Instruct. Fine-tune: 99% of the weights changed, across 48 of 48 layers.

Tensors changed: Embeddings 1/1, Norms 96/97, MLP 142/144, Attention 336/336, lm_head 1/1

2 derivatives checked

• mradermacher/NextCoder-14B-GGUF: Quantization
• gabriellarson/NextCoder-14B-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse