mfirth/agi-ds

Full fine-tune vs deepseek-ai/DeepSeek-Coder-V2-Lite-Instruct. Fine-tune: 98% of the weights changed, across 27 of 27 layers.

Tensors changed: Embeddings 0/1, Norms 29/82, lm_head 1/1, Attention 108/108, MLP 2808/2808

1 derivatives checked

• mradermacher/agi-ds-GGUF: Quantization

On the Hub · Downloads on Model Pulse