Gensyn/open-1b-sft

Full fine-tune vs Gensyn/open-1b-midtrained-93B. Fine-tune: 100% of the weights changed, across 24 of 24 layers.

Tensors changed: Attention 192/192, MLP 96/96, Norms 50/50, Embeddings 2/2

1 derivatives checked

• Gensyn/open-1b-sft-mlx-bf16: Quantization

On the Hub · Downloads on Model Pulse