Chen Qian
Chen Qian
Earlier we raise warning on micro version mismatch for python and cloudpickle, with this PR we only check the major version.
In order to support high-concurrency workflows, we are using `cache.FanoutCache` (not supported by litellm cache yet) to replace the diskcache.Cache. The following changes are included in this PR: - Unified...
This PR adds output streaming support in DSPy, which covers: - Only string-type output field is qualified for streaming. - Users can stream string output field from any submodule of...
As one step towards stable package (DSPy 3.0), we will enforce the python style for easy maintenance and scalability. This PR just contains the desired ruff rules, and we will...
Original idea from https://arxiv.org/pdf/2402.09723, made a few tweaks to make it applicable for multi-module program. Still in progress.
Use `contextvar` instead of `threading.local()` so that both async and multithreading are safe.
Make it possible to do `module.inspect_history`, which gives the LM history of the module. In nesting case, the same LM history message will be available to all the ancestor callers....
Based on rule: #7885
Based on rule: #7885
Based on rule: #7885