Mattt
Mattt
Hey @anotherjesse. At the moment, Replicate's API doesn't have a way to filter predictions by version. But that's something I'd love to add.
@asingh9530 Thanks for your offer to help, but this is blocked by Replicate's API and there's nothing to be done client-side at the moment.
This is solved by #348
Hi @beatty. Thanks for letting us know. I'll take a look to see what's going on. Can you share an ID of a prediction that had truncated output? The `run`...
@beatty If you're still seeing this behavior, please let me know, and I'd be happy to take a look. Thanks!
Hi, @bertello72. Is this happening every time? Intermittently? Just a one-off?
@bertello72 Thanks for clarifying. Searching around for that error message, it sounds like the problem can be resolved by upgrading your installation of OpenSSL or nginx (if you're using that)....
@bertello72 According to https://github.com/python/cpython/issues/110467, this appears to be a problem with Python 3.10. I'd recommend upgrading to a newer version of Python. Since this doesn't appear to be a problem...
Hi @darhsu. The `timeout` parameter is passed to the underlying [httpx client instance](https://www.python-httpx.org/advanced/timeouts/), and configures the timeout for initially creating a prediction and each subsequent request to poll for its...
@darhsu Sorry for the late response. To cancel a prediction after a given timeout, you'll need to break up `replicate.async_run` into two steps: `replicate.predictions.async_create` and `replicate.async_wait`: ```python import asyncio import...