paulml/NMTOB-7B

Full fine-tune vs Kukedlc/NeuTrixOmniBe-7B-model-remix. Fine-tune: 94% of the weights changed, across 32 of 32 layers.

Tensors changed: MLP 90/96, Norms 65/65, Attention 120/128, lm_head 1/1, Embeddings 1/1

2 derivatives checked

• mradermacher/NMTOB-7B-i1-GGUF: Quantization
• mradermacher/NMTOB-7B-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse