how to work with tensorboard? #686
Assignees
Labels
Comments
|
It's something we've been thinking about - but we don't have concrete plans. TFX announced a few exiting projects yesterday https://twitter.com/TensorFlow/status/1039244472050180096 and we'll probably sync with them + tensorboard integration for a TFJS solution. For now, we recommend rolling your own graphs and we have some examples using vega-lite. For example, here in the iris training example: https://github.com/tensorflow/tfjs-examples/blob/master/iris/ui.js#L40 |
|
I just want to add that when #631 lands, users will get a progress-bar logger when tf.Models are training in tfjs-node or tfjs-node-gpu, by default. Then users will be able to at least see the progress, the loss and metric values and measures of training speed in the console. |
|
Re-opening this issue as we are working on enabling scalar-summary writing from tfjs-node to TensorBoard now. |
caisq
added a commit
to tensorflow/tfjs-node
that referenced
this issue
Feb 5, 2019
- Add node.js-specific representation of int64-type tensors
- Add node.js-specific support for resource-type tensors
- Add the following node.js backend-specific op kernel binding:
- `summaryWriter`
- `createSummaryFileWriter`
- `writeScalarSummary`
- `flushSummaryWriter`
- Add the public Node.js JavaScript/TypeScrpit api in `tensorboard.ts`, under the API namespace `tf.node.*`
Usage example (TypeScript):
```typescript
import * as tf from '@tensorflow/tfjs-node';
const summaryWriter = tf.node.summaryFileWriter('/tmp/tfjs_tb_logdir');
for (let i = -1e3; i < 1e3; i += 10) {
summaryWriter.scalar('loss', i * i * i * i, i);
summaryWriter.scalar('acc', -i * i * i * i, i);
}
```
Open tensorboard:
```sh
pip install tensorboard # Unless you've already installed it.
tensorboard --logdir /tmp/tfjs_tb_logdir
```
Screenshot of result:
Towards tensorflow/tfjs#686
FEATURE
caisq
added a commit
to tensorflow/tfjs-node
that referenced
this issue
Feb 12, 2019
) FEATURE See example screenshot:  Fixes tensorflow/tfjs#686
caisq
added a commit
to caisq/tfjs-node
that referenced
this issue
Feb 28, 2019
…ensorflow#202) FEATURE See example screenshot:  Fixes tensorflow/tfjs#686
caisq
added a commit
to tensorflow/tfjs-node
that referenced
this issue
Feb 28, 2019
) (#213) FEATURE See example screenshot:  Fixes tensorflow/tfjs#686
nkreeger
added a commit
to tensorflow/tfjs-node
that referenced
this issue
Mar 12, 2019
* 0.3.x: Update the publish-npm script to allow publishing from the release branch. (#203) DEV * Upgrade 0.3.x to 0.15.3 (#210) <!-- Reviewable:start --> This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/tensorflow/tfjs-node/210) <!-- Reviewable:end --> * Fix win GPU packaging. (#208) (#211) Turns out that the windows GPU builds for TensorFlow 1.12 lack the directory structure and eager headers. A bug has been filed with core TF - but we should bake in some fixes for this. This PR simply refactors the downloading logic to a new file. I'd like to use this logic in the node-gles package as well (maybe worth releasing as a stand-alone package in the future). After the refactoring, I check the directory structure in Windows. If the folder structure is missing, but the required tensorflow.dll exists - I re-create the directory structure, move and download the proper header files. The screenshot below shows the contents of TF 1.12 Windows GPU:  <!-- Reviewable:start --> This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/tensorflow/tfjs-node/208) <!-- Reviewable:end --> * Bump 0.3.1 * Add TensorBoard callback for model training: tf.node.tensorBoard() (#202) (#213) FEATURE See example screenshot:  Fixes tensorflow/tfjs#686 * [0.3.x] Upgrade nyc package fo fix lodash security issue. (#218) (#219) Bump for 0.3.x so we can get a security release spun. https://github.com/tensorflow/tfjs-node/network/alert/yarn.lock/lodash/open <!-- Reviewable:start --> --- This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/tensorflow/tfjs-node/219) <!-- Reviewable:end --> * Bump to 0.3.2 * Upgrade TS libraries and change binding from typings file to plain TypeScript definition. * Upgrade TS dependencies * save * Fix deps-stage * save * Revert TS changes and keep binary staging fixes.
nkreeger
added a commit
to tensorflow/tfjs-node
that referenced
this issue
Mar 13, 2019
* 0.3.x: Update the publish-npm script to allow publishing from the release branch. (#203) DEV * Upgrade 0.3.x to 0.15.3 (#210) <!-- Reviewable:start --> This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/tensorflow/tfjs-node/210) <!-- Reviewable:end --> * Fix win GPU packaging. (#208) (#211) Turns out that the windows GPU builds for TensorFlow 1.12 lack the directory structure and eager headers. A bug has been filed with core TF - but we should bake in some fixes for this. This PR simply refactors the downloading logic to a new file. I'd like to use this logic in the node-gles package as well (maybe worth releasing as a stand-alone package in the future). After the refactoring, I check the directory structure in Windows. If the folder structure is missing, but the required tensorflow.dll exists - I re-create the directory structure, move and download the proper header files. The screenshot below shows the contents of TF 1.12 Windows GPU:  <!-- Reviewable:start --> This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/tensorflow/tfjs-node/208) <!-- Reviewable:end --> * Bump 0.3.1 * Add TensorBoard callback for model training: tf.node.tensorBoard() (#202) (#213) FEATURE See example screenshot:  Fixes tensorflow/tfjs#686 * [0.3.x] Upgrade nyc package fo fix lodash security issue. (#218) (#219) Bump for 0.3.x so we can get a security release spun. https://github.com/tensorflow/tfjs-node/network/alert/yarn.lock/lodash/open <!-- Reviewable:start --> --- This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/tensorflow/tfjs-node/219) <!-- Reviewable:end --> * Bump to 0.3.2 * Upgrade TS libraries and change binding from typings file to plain TypeScript definition. * Upgrade TS dependencies * save * Fix deps-stage * save * Revert TS changes and keep binary staging fixes. * Don't use a definition file for the bindings. This causes many issues and doesn't help with redistribution. It looks like exporting a local definition file on top of what else is exported is not a common supported TypeScript use case. This fix simply moves defnitions into a normal TypeScript file. This should fix: tensorflow/tfjs#1092 * save * Add typescript integration project. * save * Add license
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Is there any plan to support tf.summary?
The text was updated successfully, but these errors were encountered: