lm-provers/QED-Nano

Full fine-tune vs lm-provers/QED-Nano-SFT. Fine-tune: 100% of the weights changed, across 36 of 36 layers.

Tensors changed: Embeddings 1/1, Norms 132/145, MLP 108/108, Attention 144/144

2 derivatives checked

• mlx-community/QED-Nano-8bit: Quantization
• DevQuasar/lm-provers.QED-Nano-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse