Vessl Docs
Search…
Release Notes

Upcoming updates

August, 2021

    Model Registry
    Improved experiment tracking dashboard
    Quota management

More to come in Q4

    Team dashboard with resource usage tracking)
    Fine-grained permission
    Managed Airflow / Airflow integration
    TPU Support
    Fractional GPU
    Dataset versioning - dvc integration
    Dataset, output directory NFS mount
    Save/load files with Vessl SDK
For feature requests, bug reports, and supports, please contact us at [email protected]

July 27, 2021 (v0.6.0)

Sweep (Hyperparameter Optimization)

You can now find the best hyperparameters using Vessl's automated model tuning. We support grid search, random search, and bayesian optimization. Choose the hyperparameters you want, and let us to optimize it.

Minor Fixes

    T4 GPU support
    Attach local volume to workspace

July 1, 2021 (v0.5.0)

Workspace

Now you can start your own development environment. Service page has been redesigned to Workspace. Start your workspace and connect with JupyterLab and SSH.
Only files stored in the home directory are persistently managed.

Minor Fixes

    Dataset versioning is available by dvc integration.

June 8, 2021 (v0.4.15)

Support for local datasets

You can now add local dataset. Local dataset can be mounted with NFS protocol in experiment.
Local dataset is only accessible in your cluster. Since Vessl does not have any access to local files, you cannot see your local datasets under dataset page.

Download logs and metrics

You can now download your logs and metrics in experiment detail page.

Minor fixes

    Workspace is renamed to Organization

May 25, 2021 (v0.4.10)

CLI-driven project

Now, your source code does not have to be on GitHub to run an experiment on Vessl. Create a CLI-driven project and run vessl experiment run on your local terminal without git push.
1
$ sv experiment run
2
[?] Select project: cli-driven-example
3
version-control-example
4
> cli-driven-example
5
6
[?] Experiment message:
7
[?] Start command: python main.py
8
[?] Please choose a cluster: [1] aws-apne2-prod1 (SavviHub)
9
> [1] aws-apne2-prod1 (SavviHub)
10
[2] on-premise-cluster (Custom)
11
12
[?] Please choose a resource: [11] v1.v100-1.mem-52.spot (GPU(V100) x 1 / CPU 8 Cores / Memory 52GB)
13
[1] v1.cpu-2.mem-6.spot (CPU 2 Cores / Memory 6GB)
14
[2] v1.cpu-2.mem-6 (CPU 2 Cores / Memory 6GB)
15
[3] v1.cpu-4.mem-13 (CPU 4 Cores / Memory 13GB)
16
[4] v1.k80-1.mem-52 (GPU(K80) x 1 / CPU 4 Cores / Memory 52GB)
17
[5] v1.k80-8.mem-480 (GPU(K80) x 8 / CPU 32 Cores / Memory 480GB)
18
[6] v1.v100-1.mem-52 (GPU(V100) x 1 / CPU 8 Cores / Memory 52GB)
19
[7] v1.v100-4.mem-232 (GPU(V100) x 4 / CPU 32 Cores / Memory 232GB)
20
[8] v1.cpu-0.mem-1 (CPU shared / Memory 1GB)
21
[9] v1.k80-1.mem-52.spot (GPU(K80) x 1 / CPU 4 Cores / Memory 52GB)
22
[10] v1.cpu-4.mem-13.spot (CPU 4 Cores / Memory 13GB)
23
> [11] v1.v100-1.mem-52.spot (GPU(V100) x 1 / CPU 8 Cores / Memory 52GB)
24
[12] v1.v100-8.mem-480 (GPU(V100) x 8 / CPU 96 Cores / Memory 480GB)
25
[13] v1.k80-16.mem-724 (GPU(K80) x 16 / CPU 64 Cores / Memory 724GB)
26
27
[?] Please choose a kernel image: [2] savvihub/kernels:py37.full-cpu (Python 3.7 (All Packages))
28
[1] savvihub/kernels:py36.full-cpu (Python 3.6 (All Packages))
29
> [2] savvihub/kernels:py37.full-cpu (Python 3.7 (All Packages))
30
[3] savvihub/kernels:py36.full-cpu.jupyter (Python 3.6 (JupyterLab))
31
[4] savvihub/kernels:py37.full-cpu.jupyter (Python 3.7 (JupyterLab))
32
[5] tensorflow/tensorflow:1.14.0-py3 (Tensorflow 1.14.0)
33
[6] tensorflow/tensorflow:1.15.5-py3 (Tensorflow 1.15.5)
34
[7] tensorflow/tensorflow:2.0.4-py3 (Tensorflow 2.0.4)
35
[8] tensorflow/tensorflow:2.2.1-py3 (Tensorflow 2.2.1)
36
[9] tensorflow/tensorflow:2.3.2 (Tensorflow 2.3.2)
37
[10] tensorflow/tensorflow:2.4.1 (Tensorflow 2.4.1)
38
[11] tensorflow/tensorflow:2.3.0 (TensorFlow 2.3.0 (Tensorboard))
39
40
Upload the zipped local project
41
Experiment 1 is running. Check the experiment status at below link
42
https://savvihub.com/example-workspace/cli-driven-example/experiments/1
Copied!

Service edit & reproduce

Reproduce and Edit have been added to service actions. A new service with an existing service's configuration by clicking Reproduce.
You can modify a stopped service's name, computing resource, start command, exposed ports, environment variables and ssh key. Name and ports can be updated even when running.

Minor fixes

May 9, 2021 (v0.4.0)

Billing

You can check your billing information based on on-demand instance usage on Under Settings -> Billing.

Continue training after spot instance termination

If you use spot instance in an experiment, it automatically continues after spot interruption. All you need to do is that make your experiment resist termination, such as save your checkpoint in every epoch, and start your experiment from saved checkpoint. You can check the details in here.

A100 / V100 spot instance in US West2

You can use A100 spot instances in us west2 (Oregon) region. Select default region as us-west2 in workspace create. You can only select one region per one workspace for now.

Minor Fixes

    Save & load checkpoint in savvihub/examples.
    Fix web terminal link broken in custom cluster.

Apr 27, 2021 (v0.3.0)

Support for private docker image

You can use your own private docker registry on DockerHub & AWS ECR. Register your own credentials under Workspace -> Settings -> Integrations.

SSH connection / terminal on web

Now you can connect to your experiment / service with terminal on web, and native SSH connection.

Experiment termination protection

You may want to access the experiment container after it runs. Termination protection allows you to do that. If you checked the checkbox, then the experiment will go to idle status upon completion.

Minor Fixes

    (Fix) Log collect on various docker/kubernetes runtime configuration (e.g. RKE)
      You can configure container log path withkubernetes.logContainerPath
      For RKE, install helm with --set kubernetes.logContainerPath=/var/log/containers. (Ref)
    (Fix) Remove prometheus dependency
      SavviHub does not install prometheus on savvihub agent installation.
Last modified 10d ago