Llama3-Taiwan-70B-Instruct-128K-AWQ-4bits leverages 4-bit quantized weights, processed with AutoAWQ, to significantly reduce GPU memory requirements.

References

For more information and detailed documentation, please refer to the links provided.

Safetensors

Model size

11.3B params

Tensor type

I32

FP16

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support