jina
jina copied to clipboard
jina-ai
☁️ Build multimodal AI applications with cloud-native stack
Goals: - ... - ... - [ ] check and update documentation. See [guide](https://github.com/jina-ai/jina/CONTRIBUTING.md#documentation-guidelines) and ask the team.
## Problem Endpoint discovery makes is so that requests do not reach executors where the endpoint doesn't match. For testing purposes, we often don't add an Executor, add `BasExecutor` will...
I am a member of the debian community and am interested in packaging jina for debian. This will make jina available in debian through the default package managing tool. In...
Just a random thought, how does this compare to other serving frameworks like Triton Inference Server, Torch Serve ... ? Thanks for this great project :)
**Describe the bug** ```python class FooGenerator(Executor): @requests(on='/') def generate(self, docs: DocumentArray, **kwargs): t = 'foo' self.logger.info(f'formatting "{t}"...') self.logger.info(f'formatting [{t}]...') print(f'formatting [{t}]...') f = Flow().add(uses=FooGenerator) with f: f.post(on='/') ``` above code...
**Describe the feature** We have seen that the simple `Request-Response` model with a static `request_size` is limiting in some cases. For instance some `Encoders` would rather accumulate `requests` until a...
# Custom Gateway We want to allow users to define their own Gateway. ## Context The point is that atm many users put a fastAI in front of our gateway...
# Context Right now we can override the mapping between method and endpoint of an Executor by using the `uses_request` parameters: Ex: ```python from jina import Flow with flow =...
**Describe the bug** As I talked to @samsja offline, Grafana metrics are sometimes questionable / confusing to me. For example in this screenshot: The Time elapsed between receiving a request...
