subhrokomol/Llama2-7B-Hindi-finetuned

Partial fine-tune vs meta-llama/Llama-2-7b-hf. Partial change: 20% of the weights changed (attention, embeddings, lm_head), in 32 of 32 layers.

Tensors changed: Attention 96/160, Embeddings 1/1, Norms 0/65, MLP 0/96, lm_head 1/1

1 derivatives checked

• mradermacher/Llama2-7B-Hindi-finetuned-GGUF: Quantization

On the Hub · Downloads on Model Pulse