…ow (nv-morpheus#429)

Fixes an issue where without specifying the `nvidia/label/dev` channel conda/mamba will install an old version of tensorflow (2.4.1) which is unable to load the model file.

fixes nv-morpheus#324

Authors:
  - David Gardner (https://github.com/dagardner-nv)

Approvers:
  - Christopher Harris (https://github.com/cwharris)
  - Pete MacKinnon (https://github.com/pdmack)
  - Michael Demoret (https://github.com/mdemoret-nv)

URL: nv-morpheus#429

…v-morpheus#419)

* `MultiInferenceMessage` & `MultiResponseMessage` now inherit a new class `MultiTensorMessage` reducing the duplicated code in each
* `MultiResponseProbsMessage` class now uses `ResponseMemoryProbs`
* Add missing docstrings
* Common CSV & Json reading code moved to deserializers.[hpp|cpp]

fixes nv-morpheus#313 
fixes nv-morpheus#311

Authors:
  - David Gardner (https://github.com/dagardner-nv)

Approvers:
  - Michael Demoret (https://github.com/mdemoret-nv)

URL: nv-morpheus#419

Builds two different images: `morpheus-ci-driver` and `morpheus-ci-test`.
The driver image is 15.9GB and the test image is 14.2GB. 

In this setup only the wheel is transferred from the build to the test and documentation stages(nv-morpheus#112).

Base image contains:
- Pre-populated conda env
- libnuma1

Driver image contains: 
- libnvidia-compute-495
- ibcublas-dev-11-5
- libcufft-dev-11-5
- libcurand-dev-11-5
- libcusolver-dev-11-5 

Test image contains
- pytest-kafka
- Kafka
- openjdk-11 (needed for kafka)
- nodejs
- npm
- camouflage

fixed nv-morpheus#112 
fixes nv-morpheus#400

Authors:
  - David Gardner (https://github.com/dagardner-nv)

Approvers:
  - Michael Demoret (https://github.com/mdemoret-nv)

URL: nv-morpheus#408

If a test crashes or a developer stops a test in the middle of debugging, Camouflage will be left open. Any subsequent runs of pytest will fail to start.

This fixes the issue by binding the child process to the pytest process, ensuring that it will always be closed. Also added support for using already running Camouflage servers.

Closes nv-morpheus#329

Authors:
  - Michael Demoret (https://github.com/mdemoret-nv)

Approvers:
  - David Gardner (https://github.com/dagardner-nv)

URL: nv-morpheus#436

…json

* Avoid unnecessary copies in file source stage
* When we do a copy, perform it before emitting
* Skip benchmarks by default
* Adds a `benchmark` pytest marker and associated `--run_benchmark` flag

fixes nv-morpheus#443

Authors:
  - David Gardner (https://github.com/dagardner-nv)

Approvers:
  - Michael Demoret (https://github.com/mdemoret-nv)

URL: nv-morpheus#444

This requires ensuring that all newly built Tensors using the `create` method have their strides specified in elements not bytes.

Authors:
  - David Gardner (https://github.com/dagardner-nv)
  - Michael Demoret (https://github.com/mdemoret-nv)

Approvers:
  - Michael Demoret (https://github.com/mdemoret-nv)

URL: nv-morpheus#438