mirror of https://github.com/kubeflow/examples.git
68 lines
3.5 KiB
Markdown
68 lines
3.5 KiB
Markdown
# Jupyter Notebook
|
||
> Please put Cornell-1000-nltk.ipynb and Twitter-5000-nltk.ipynb into the folder of Jupyter Notebook first. \
|
||
> If you use Minikube to install Kubeflow, the folder of Jupyter Notebook will usually be in：
|
||
```Bash
|
||
/tmp/hostpath-provisioner/kubeflow-user-example-com/workspace-<your Jupyter name>
|
||
```
|
||
|
||
## Pipeline
|
||
> Cornell-1000.zip and twitter-5000.zip are compressed files generated after executing Cornell-1000-nltk.ipynb and Twitter-5000-nltk.ipynb. \
|
||
> The content of the compressed file is the yaml file of the pipeline.
|
||
|
||
<img src="https://github.com/dfm871002/examples/blob/master/Natural-Language-Processing/4.%20Image/pipeline.png" alt="pipeline"/><br/>
|
||
|
||
## Custom data
|
||
> Twitter-5000-nltk and Cornell-1000-nltk use similar code, and the difference is in downloading and reading data. \
|
||
> If you want to use other data, you only need to classify the data and save it in str format into pos_tweets and neg_tweets.
|
||
|
||
<img src="https://github.com/dfm871002/examples/blob/master/Natural-Language-Processing/4.%20Image/data%20list.png" alt="data list"/><br/>
|
||
|
||
# Port Forward
|
||
|
||
### Step 1：Find the pod name of Http port
|
||
<img src="https://github.com/dfm871002/examples/blob/master/Natural-Language-Processing/4.%20Image/nltk.jpg" alt="nltk pod"/><br/>
|
||
|
||
### Step 2：Port-forward
|
||
```Bash
|
||
kubectl port-forward -n kubeflow-user-example-com <pod name> 3000:5000
|
||
```
|
||
<img src="https://github.com/dfm871002/examples/blob/master/Natural-Language-Processing/4.%20Image/port%20forward.png" alt="nltk pod port forward"/><br/>
|
||
|
||
### Step 3：Input in the browser
|
||
```Bash
|
||
http://localhost:3000/
|
||
```
|
||
or
|
||
```Bash
|
||
127.0.0.1:3000
|
||
```
|
||
<img src="https://github.com/dfm871002/examples/blob/master/Natural-Language-Processing/4.%20Image/NLP.png" alt="NLP"/><br/>
|
||
|
||
### Step 4：Predict
|
||
<img src="https://github.com/dfm871002/examples/blob/master/Natural-Language-Processing/4.%20Image/nice%20to%20meet%20you.png" alt="nice to meet you"/><br/>
|
||
<img src="https://github.com/dfm871002/examples/blob/master/Natural-Language-Processing/4.%20Image/NLP%20N.png" alt="i hate you"/><br/>
|
||
|
||
# Accuracy
|
||
You can confirm the accuracy of the NLP individually, \
|
||
<img src="https://github.com/dfm871002/examples/blob/master/Natural-Language-Processing/4.%20Image/twitter-5000%20accuracy.png" alt="twitter"/>
|
||
<img src="https://github.com/dfm871002/examples/blob/master/Natural-Language-Processing/4.%20Image/cornell-1000%20accuracy.png" alt="cornell"/><br/>
|
||
or you can use a comparison run for comparison. \
|
||
<br>
|
||
<img src="https://github.com/dfm871002/examples/blob/master/Natural-Language-Processing/4.%20Image/compare%20runs.png" alt="compare"/><br/>
|
||
|
||
# Disabling caching in your Kubeflow Pipelines deployment
|
||
> If you delete the pvc and execute the pipeline again, you find that it does not work properly, it may be a cache problem. \
|
||
> The following command can be executed to disable the cache.
|
||
```Bash
|
||
export NAMESPACE=kubeflow
|
||
kubectl patch mutatingwebhookconfiguration cache-webhook-${NAMESPACE} --type='json' -p='[{"op":"replace", "path": "/webhooks/0/rules/0/operations/0", "value": "DELETE"}]'
|
||
```
|
||
* [Kubeflow Caching](https://www.kubeflow.org/docs/components/pipelines/caching/)
|
||
|
||
# Relevant part
|
||
|
||
* [About Version](https://github.com/dfm871002/examples/blob/master/Natural-Language-Processing/README.md)
|
||
* [About Install](https://github.com/dfm871002/examples/blob/master/Natural-Language-Processing/1.%20Install/Install.md)
|
||
* [About Docker](https://github.com/dfm871002/examples/blob/master/Natural-Language-Processing/2.%20Docker/Docker.md)
|
||
|