Skip to content

Commit cb0a13c

Browse files
Requested changes for quick-start Update
1 parent 1f9994f commit cb0a13c

File tree

1 file changed

+9
-2
lines changed

1 file changed

+9
-2
lines changed

Quick-Start.md

Lines changed: 9 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -146,7 +146,6 @@ If you are working in an on-prem environment, and for the purposes of following
146146
cluster_config = ClusterConfiguration(
147147
name='jobtest',
148148
namespace="default",
149-
# machine_types = ["m4.xlarge", "g4dn.xlarge"]
150149
num_workers=2,
151150
min_cpus=1,
152151
max_cpus=1,
@@ -184,6 +183,8 @@ We are going to use the CodeFlare SDK to submit batch jobs via TorchX, either to
184183
185184
First, let's begin by submitting to Ray, training a basic NN on the MNIST dataset:
186185
186+
The `mnist.py` file used comes from [here](https://github.com/opendatahub-io/distributed-workloads/blob/main/tests/resources/mnist.py), which is accessed in your jupyter notebook under `codeflare-sdk/demo-notebooks/guided-demos/mnist.py`
187+
187188
```python
188189
jobdef = DDPJobDefinition(
189190
name="mnisttest",
@@ -193,10 +194,16 @@ jobdef = DDPJobDefinition(
193194
job = jobdef.submit(cluster)
194195
```
195196
196-
Once the job is submitted you can follow it on the Ray dashboard using the following commands to output the job status directly into you're notebook:
197+
Once the job is submitted you can follow it on the Ray dashboard using the following commands to output the job status directly into your notebook:
197198
```python
198199
cluster.cluster_dashboard_uri()
200+
```
201+
202+
```python
199203
job.status()
204+
```
205+
206+
```python
200207
job.logs()
201208
```
202209

0 commit comments

Comments
 (0)