Multiple models #37
Merged
Multiple models #37
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Something odd happening with the rendering: |
d-lowl
added a commit
that referenced
this pull request
Aug 12, 2021
* Make init, get-endpoint and deploy actions take model name as the parameter * Add list, describe and remove commands for models * Fix get_model_dvc_pipeline invocations to have a model name
archena
added a commit
that referenced
this pull request
Nov 9, 2021
* Multiple models (#37) * Make init, get-endpoint and deploy actions take model name as the parameter * Add list, describe and remove commands for models * Fix get_model_dvc_pipeline invocations to have a model name * Generalised training (#46) * Make init, get-endpoint and deploy actions take model name as the parameter * Add list, describe and remove commands for models * Fix get_model_dvc_pipeline invocations to have a model name * WIP Generalise training on vertex script * WIP Generalise training on vertex script * WIP Generalise training on vertex script * WIP Generalise training on vertex script * Add setup.py * Fix config default paths * Add training utils * Do not get config, state or mongo when running on vertex * Fix is_vertex override * Fix model init * Ensure that Docker image can run stand-alone * Update edge script to invoke Docker correctly * Removed the examples. Now, this repo will contain just tool(s). Examples are to be found in github.com:fuzzylabs/vertex-edge-examples.git * README covering a rough guide to starting a model from scratch Co-authored-by: D. Lowl <[email protected]> * Hello world (#47) * Move vertex training decorator to vertex:edge * Add save results method * Fix training script path * Add joblib as a dependency * Unquote CLI sacred parameters * Add Google Cloud Storage as a dependency * Fix trained model path and add model name and path to the wrapper signature * Fix trained model path and add model name and path to the wrapper signature * Fix trained model path and add model name and path to the wrapper signature * Improve README introduction * Fix missing format string specifier * Further README step-by-step instructions (WIP) * README step-by-step instructions Co-authored-by: archena <[email protected]> * Add model template verb and template for an sklearn hello world pipeline (#48) Co-authored-by: archena <[email protected]> * Fixes: template, pin dependencies for grpcio and gcloud core * Take out experiment from template (temporarily). Update pip install package for the real deal * pip install vertex-edge * Fixing dependencies in setup.py * Fix experiment tracker init * WIP: experimental magical class that hides experiment tracking * Run using new boilerplate on Vertex * Support for local and remote training targets * Testing on Vertex * Testing on Vertex * Testing on Vertex * Testing on Vertex * Testing on Vertex * Downgrade Python to match Vertex * Outlined dev documentation; build fixes * Outlined dev documentation; build fixes * Fix issue where omniboard yaml file was not available inside Docker * Re-arrange packages: temporarily moved the original training package and replaced it with the new approach * Update Cookie Cutter template * Remove web app config from the tool; while this fits in as part of a reference example, it does not fit in to the tool * Removed unused Docker commands * Split requirements for dev and build * Remove broken training model reference in model deployment: to revisit later * Fail gracefully if there is no experiment tracker set up * Fail gracefully if there is no experiment tracker set up * Re-enable Git support for experiment tracking * Cleaned up old README content. Re-wrote introduction * Support TensorFlow * Standard path for saving model assets * Fix staging path * Fix import * Consistent model ID for Vertex hand-off * Re-introduce local model JSON file that allows us to deploy a model * Let there be models * Fix prediction image name * Update template to TensorFlow and remove DVC (for now) * README improvements * Documentation: sklearn -> tensorflow * Documentation: split out the developer documentation * README re-arranged * Quick-start guide * README TOC * README TOC * First draft of 'why'. Experiment tracker screenshot * More information on dashboards * Why: vision. More on buckets * DVC in README * Setup tutorial * Train and deploy tutorial * Train and deploy tutorial * Don't fail training if no model has been saved * Don't fail training if no model has been saved * Update link to vertex-edge package dependency for when we run on Vertex * Temporarily drop the README image Co-authored-by: D. Lowl <[email protected]> Co-authored-by: D. Lowl <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Make init, get-endpoint and deploy actions take model name as the parameter.
Add list, describe and remove commands for models.
Closes #3