OpenBuddy/openbuddy-gemma-7b-v19.1-4k

Full fine-tune vs google/gemma-7b. Fine-tune: 100% of the weights changed, across 28 of 28 layers.

Tensors changed: Embeddings 1/1, Norms 56/57, MLP 84/84, Attention 112/112

2 derivatives checked

• mradermacher/openbuddy-gemma-7b-v19.1-4k-i1-GGUF: Quantization
• mradermacher/openbuddy-gemma-7b-v19.1-4k-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse