aisingapore/Llama-SEA-LION-v3.5-70B-R

Full fine-tune vs aisingapore/Llama-SEA-LION-v3-70B-IT. Fine-tune: 100% of the weights changed, across 80 of 80 layers.

Tensors changed: Norms 161/161, MLP 240/240, Attention 320/320, lm_head 1/1, Embeddings 1/1

2 derivatives checked

• mradermacher/Llama-SEA-LION-v3.5-70B-R-i1-GGUF: Quantization
• mradermacher/Llama-SEA-LION-v3.5-70B-R-GGUF: Quantization

On the Hub · Downloads on Model Pulse