This issue lists Renovate updates and detected dependencies. Read the Dependency Dashboard docs to learn more.

Rate-Limited

These updates are currently rate-limited. Click on a checkbox below to force their creation now.

• Update dependency MarkupSafe to v2.1.5
• Update dependency Werkzeug to v2.3.8
• Update dependency absl-py to v0.15.0
• Update dependency click to v8.1.7
• Update dependency cloudevents to v1.10.1
• Update dependency gast to v0.5.4
• Update dependency google-api-core to v1.34.1
• Update dependency google-api-python-client to v2.122.0
• Update dependency google-auth to v1.35.0
• Update dependency google-auth-httplib2 to v0.2.0
• Update dependency google-auth-oauthlib to v0.8.0
• Update dependency googleapis-common-protos to v1.63.0
• Update dependency h5py to v3.10.0
• Update dependency httplib2 to v0.22.0
• Update dependency importlib-metadata to v4.13.0
• Update dependency itsdangerous to v2.1.2
• Update dependency numpy to v1.26.4
• Update dependency oauthlib to v3.2.2
• Update dependency opencv-python to v4.9.0.80
• Update dependency protobuf to v3.20.3
• Update dependency pyasn1 to v0.5.1
• Update dependency pyasn1-modules to v0.3.0
• Update dependency pytz to v2021.3
• Update dependency rsa to v4.9
• Update dependency sympy to v1.12
• Update dependency wrapt to v1.16.0
• Update dependency zipp to v3.18.1
• Update golang Docker tag to v1.20
• Update module github.com/aws/aws-sdk-go to v1.51.1
• Update python Docker tag to v3.12
• Update dependency Flask to v3
• Update dependency Werkzeug to v3
• Update dependency absl-py to v2
• Update dependency cachetools to v5
• Update dependency chardet to v5
• Update dependency google-api-core to v2
• Update dependency google-auth to v2
• Update dependency google-auth-oauthlib to v1
• Update dependency idna to v3
• Update dependency importlib-metadata to v7
• Update dependency packaging to v24
• Update dependency protobuf to v5
• Update dependency pyparsing to v3
• Update dependency pytz to v2024
• Update dependency tensorboard to v2
• Update dependency tensorflow-estimator to v2
• Update dependency termcolor to v2
• Update dependency typing-extensions to v4
• Update dependency uritemplate to v4
• Update tensorflow/tensorflow Docker tag
• 🔐 Create all rate-limited PRs at once 🔐

Open

These updates have all been created already. Click a checkbox below to force a retry/rebase of any.

• Update dependency urllib3 to v1.26.18 [SECURITY]
• Update dependency grpcio to v1.53.0 [SECURITY]
• Update dependency mpmath to v1.3.0 [SECURITY]
• Update dependency requests to v2.31.0 [SECURITY]
• Update module github.com/cloudevents/sdk-go/v2 to v2.15.2 [SECURITY]
• Update dependency certifi to v2023 [SECURITY]
• Update dependency tensorflow to v2 [SECURITY]
• Update dependency cachetools to v4.2.4
• Update dependency tensorflow-estimator to v1.15.2
• Update dependency typing-extensions to v3.10.0.2
• Update tensorflow/tensorflow Docker tag to v1.15.5
• Update dependency Flask to v2.3.3
• Update dependency Jinja2 to v3.1.3
• Update dependency Markdown to v3.6
• Update dependency requests-oauthlib to v1.4.0
• Click on this checkbox to rebase all open PRs at once

Detected dependencies

• Check this box to trigger a request for Renovate to run again on this repository

I really do not like io.triggermesh.transformations.s3-tensorflow.response I would like to see this replaced with io.triggermesh.transformations.tensformation.response. However I want to wait for the demo to be recorded to avoid causing any havoc :)

TF Serving does not respond to SIGTERM, meaning when running in Docker/K8s it will timeout and be ended after the timeout by a SIGKILL:

• tensorflow/serving#356
• tensorflow/serving#1853

Implement a wrapper script for the container entrypoint to take SIGTERM and automatically SIGKILL the child process as an interim step to having SIGTERM support added to TF Serving server upstream.

ref implementations:

• https://unix.stackexchange.com/questions/146756/forward-sigterm-to-child-in-bash
• https://medium.com/@kavehmz/how-to-handle-an-ungraceful-in-kubernetes-by-bash-ef93a0bf5140
• https://betterprogramming.pub/how-to-deploy-your-tensorflow-model-in-a-production-environment-23a9572e94d3
• https://github.com/krallin/tini

Model warmup and pausing the container ready notification until the model has warmed up has the potential to increase performance by decreasing the latency per inference, and in particular on the first instance of the API's use per tf-inference container.

Need to generate model warmup data (which we can do with: https://github.com/mylesagray/tensorflow-anpr/tree/master/dataset_prep/artificial), include in the anpr-serving container image and configure TF Serving to load the warmup data on start.

ref:

• https://www.tensorflow.org/tfx/serving/saved_model_warmup

Currently we are using a set of numbers for batching in our tf-inference server that we have arrived at through trial and error.

There must be a more comprehensive way to look at these with a view to improving throughput and reliability.

refs:

• https://github.com/tensorflow/serving/blob/master/tensorflow_serving/batching/README.md
• https://mux.com/blog/tuning-performance-of-tensorflow-serving-pipeline/
• https://techblog.realtor.com/building-tensorflow-serving-aws-gpu-instances/
• https://stackoverflow.com/questions/42519010/how-to-do-batching-in-tensorflow-serving