3 ms·
> To run inference with the model in full bfloat16 precision you need approximately 8xA100 80GB or equivalent. https://huggingface.co/tiiuae/falcon-180B https:
by sodality2 3y ago
> To run inference with the model in full bfloat16 precision you need approximately 8xA100 80GB or equivalent.
https://huggingface.co/tiiuae/falcon-180B https://huggingface.co/tiiuae/falcon-180B