Read news on llama-70b with our app.
Read more in the app
Post-transformer inference: 224× compression of Llama-70B with improved accuracy
