Arun Patro

I first run the experiment on an `AWS p2.xlarge EC2` instance. It gave me these errors. I re-run the experiment later on a NVIDIA Titan X. It ran smoothly. Converging...

I definitely can use `generate_batch`, that works fast! But if I had access to the kv cache or HF like api then I could do fast guided generation like with...

Seconded! The developers at https://github.com/normal-computing/outlines would also benefit with such support.

Yes, the code is much easier to understand. Thank you so much. However I see that train.py is different than master branch. On Fri, Jan 25, 2019, 6:44 PM Gokhan...