Our great sponsors
-
WorkOS
The modern identity platform for B2B SaaS. The APIs are flexible and easy-to-use, supporting authentication, user identity, and complex enterprise features like SSO and SCIM provisioning.
-
self-hosted
Sentry, feature-complete and packaged up for low-volume deployments and proofs-of-concept
DocArray to manipulate data and interact with the storage backend using document store.
Right now, Jina only supports the export mechanism to push telemetry signals to external systems. It uses OpenTelemetry Collector Contrib as the unified component to collect telemetry signals before exporting to enhanced components that transform the data for visualization and analysis. The collector setup is very basic and functions only as the uniform intermediary for collecting and exporting data.
OpenTelemetry is a project incubated by CNCF. It brings distributed tracing and metrics. Pamela has brought us on to implement this telemetry into her Jina Flow. Before getting started we have to learn a few new terms:
💡In this post we’re just building out a backend, and not touching on a frontend. To build your own low-code backend+frontend neural search solution, check out Jina NOW.
We use the self-hosted Sentry application landscape to set up the actual APM or SPM. We'll only explore a small set of features supported by Sentry to preserve the focus of this post. Refer to the documentation for more details.
We derived the dataset by pre-processing the deepfashion dataset using Finetuner. The image label generated by Finetuner is extracted and formatted to produce the text attribute of each product.
Related posts
- Improving Search Quality for Non-English Queries with Fine-tuned Multilingual CLIP Models
- Most Trending Open Source MLOps Tools of 2022
- DocArray – Represent, send, and store multimodal data for ML
- Some questions about multimodal data.
- Trying to create an AI recommender system that’s also ad-free video streaming.