Synchronous (with “async” support) client for Yandex's Foundation Models API (YandexGPT and YandexART). Fire off long-running “AI” generation tasks without sweating the wait.
Stars
2
Forks
1
Watchers
2
Open Issues
0
Overall repository health assessment
No package.json found
This might not be a Node.js project
57
commits
feat: add support for embeddings api; fix[gpt]: context manager enter did not override; chore(yafma): export classes from submodules too
1ad3938View on GitHubrefactor: move core client functionality to a `base` class
f1adb65View on GitHubfix(ci): omg gitlab is cool at not giving the branch info on tag
2703bb1View on GitHubfix(example): reflect recent need for explicit `stream` value specification; fix[gpt]: untype args and kwargs (ma brain exploded)
0eca7d4View on GitHubfeat[gpt]: replace response processing with new generic'd method + minor fixes
08affdcView on GitHubfeat: update error handling and docs in gpt client and models
6a58d12View on GitHubrefactor: restructure modules, rename classes and attributes, update import statements, docstrings overhaul
7b3f564View on GitHubfix(module): properly export class from the module as per pyright recommendations
6fa475bView on GitHubchore(deps): pin down versions for the building packages
870d438View on GitHubfix(ci): renaming to yafma should also be reflected in ruff checker
9ba0c9aView on GitHubfeat: relicense to lgpl + rename to `yafma` + rename llm client to `YandexGPTClient` + add pillow as dependecy for future yandexart support
9822dceView on GitHubchore(deps): update httpx 'till a major version; chore: bump to 0.4.3
ac61ac9View on GitHubfix: add content_filtered status for the choice status
15e302cView on GitHub