ClearML/clearml-serving
1
0
Fork 0
mirror of https://github.com/clearml/clearml-serving synced 2025-01-31 02:46:54 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

Merge pull request #10 from pollfly/main

Browse Source 
link to clear.ml docs
This commit is contained in:
Jake Henning 2021-09-14 15:45:14 +03:00 committed by GitHub
commit 7c1c02c9ea
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 3 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
README.md
View File

@ -48,7 +48,7 @@ pip install clearml-serving
Clearml-Serving will automatically serve *published* models from your ClearML model repository, so the first step is getting a model into your ClearML model repository.
Background: When using `clearml` in your training code, any model stored by your python code is automatically registered (and, optionally, uploaded) to the model repository. This auto-magic logging is key for continuous model deployment.
To learn more on training models and the ClearML model repository, see the [ClearML documentation](https://allegro.ai/clearml/docs)
To learn more on training models and the ClearML model repository, see the [ClearML documentation](https://clear.ml/docs/latest/docs/)
### Training a toy model with Keras (about 2 minutes on a laptop)
@ -115,7 +115,7 @@ clearml-serving triton --endpoint "keras_mnist" --model-project "examples" --mo
3. Launch the serving service.
The service will be launched on your "services" queue, which by default runs services on the ClearML server machine.
(Read more on services queue [here](https://allegro.ai/clearml/docs/docs/concepts_fundamentals/concepts_fundamentals_clearml_server.html#clearml-agent-services-container))
(Read more on services queue [here](https://clear.ml/docs/latest/docs/clearml_agent#services-mode))
We set our serving-engine to launch on the "default" queue,
```bash
clearml-serving launch --queue default
@ -123,7 +123,7 @@ clearml-serving launch --queue default
4. Optional: If you do not have a machine connected to your ClearML cluster, either read more on our Kubernetes integration, or spin a bare-metal worker and connect it with your ClearML Server.
`clearml-serving` is leveraging the orchestration capabilities of `ClearML` to launch the serving engine on the cluster.
Read more on the [ClearML Agent](https://github.com/allegroai/clearml-agent) orchestration module [here](https://allegro.ai/clearml/docs/docs/concepts_fundamentals/concepts_fundamentals_clearml_agent.html)
Read more on the [ClearML Agent](https://github.com/allegroai/clearml-agent) orchestration module [here](https://clear.ml/docs/latest/docs/clearml_agent)
If you have not yet setup a ClearML worker connected to your `clearml` account, you can do this now using:
```bash
pip install clearml-agent