Note that this coding model is trained on programming use cases, and is also not tuned for multi-turn chat. You can ask it to implement an algorithm; we provid

Note that this coding model is trained on programming use cases, and is also not tuned for multi-turn chat. You can ask it to implement an algorithm; we provide suggested prompts you can test. Also, this tech preview is really about the speed of the inference engine (not the model itself) so I'm glad you got 3.4k tok/s!