Internal change

PiperOrigin-RevId: 555605461

README.md

@@ -41,12 +41,12 @@ Stay up to date on LIT by joining the
 [lit-announcements mailing list](https://groups.google.com/g/lit-annoucements).
 
 For a broader overview, check out [our paper](https://arxiv.org/abs/2008.05122) and the
-[user guide](https://github.com/PAIR-code/lit/wiki/ui_guide.md).
+[user guide](documentation/ui_guide.md).
 
 ## Documentation
 
-*   [Documentation index](https://github.com/PAIR-code/lit/wiki)
-*   [FAQ](https://github.com/PAIR-code/lit/wiki/faq.md)
+*   [Documentation index](documentation/index.md)
+*   [FAQ](documentation/faq.md)
 *   [Release notes](./RELEASE.md)
 
 ## Download and Installation
@@ -63,7 +63,7 @@ git clone https://github.com/PAIR-code/lit.git && cd lit
 docker build --file Dockerfile --tag lit-nlp .
 ```
 
-See the [advanced guide](https://github.com/PAIR-code/lit/wiki/docker.md) for detailed instructions on using the
+See the [advanced guide](documentation/docker.md) for detailed instructions on using the
 default LIT Docker image, running LIT as a containerized web app in different
 scenarios, and how to creating your own LIT images.
 
@@ -168,7 +168,7 @@ python -m lit_nlp.examples.<example_name> --port=5432 [optional --args]
 
 ## User Guide
 
-To learn about LIT's features, check out the [user guide](https://github.com/PAIR-code/lit/wiki/ui_guide.md), or
+To learn about LIT's features, check out the [user guide](documentation/ui_guide.md), or
 watch this [video](https://www.youtube.com/watch?v=CuRI_VK83dU).
 
 ## Adding your own models or data
@@ -177,18 +177,18 @@ You can easily run LIT with your own model by creating a custom `demo.py`
 launcher, similar to those in [lit_nlp/examples](./lit_nlp/examples). The
 basic steps are:
 
-*   Write a data loader which follows the [`Dataset` API](https://github.com/PAIR-code/lit/wiki/api.md#datasets)
-*   Write a model wrapper which follows the [`Model` API](https://github.com/PAIR-code/lit/wiki/api.md#models)
+*   Write a data loader which follows the [`Dataset` API](documentation/api.md#datasets)
+*   Write a model wrapper which follows the [`Model` API](documentation/api.md#models)
 *   Pass models, datasets, and any additional
-    [components](https://github.com/PAIR-code/lit/wiki/api.md#interpretation-components) to the LIT server class
+    [components](documentation/api.md#interpretation-components) to the LIT server class
 
 For a full walkthrough, see
-[adding models and data](https://github.com/PAIR-code/lit/wiki/api.md#adding-models-and-data).
+[adding models and data](documentation/api.md#adding-models-and-data).
 
 ## Extending LIT with new components
 
 LIT is easy to extend with new interpretability components, generators, and
-more, both on the frontend or the backend. See our [documentation](https://github.com/PAIR-code/lit/wiki) to get
+more, both on the frontend or the backend. See our [documentation](documentation/index.md) to get
 started.
 
 ## Pull Request Process

docs/tutorials/tab-feat-attr/index.html

@@ -81,7 +81,7 @@ <h2>Tabular Feature Attribution</h2>
 for text data. But what if you have a tabular dataset? You might want to find
 out which features (columns) are most relevant to the model’s predictions. LIT's
 Feature Attribution module for
-<a href="https://github.com/PAIR-code/lit/wiki/components.md#tabular-data">tabular datasets</a>
+<a href="https://github.com/PAIR-code/lit/blob/main/documentation/components.md#tabular-data">tabular datasets</a>
 support identification of these important features. This tutorial provides a
 walkthrough for this module within LIT, on the
 <a href="https://allisonhorst.github.io/palmerpenguins/">Palmer Penguins dataset</a>.</p>
@@ -94,7 +94,7 @@ <h2>Tabular Feature Attribution</h2>
 over tabular data, which is basically a specially weighted local linear
 regression for estimating SHAP values and works for any model. For now,
 the feature attribution module is only shown in the UI when working with
-<a href="https://github.com/PAIR-code/lit/wiki/components.md#tabular-data">tabular data</a>.</div>
+<a href="https://github.com/PAIR-code/lit/blob/main/documentation/components.md#tabular-data">tabular data</a>.</div>
 </div>
 <h3><strong>Overview</strong></h3>
 <p>The <a href="https://pair-code.github.io/lit/demos/penguins.html">penguins demo</a> is a

docs/tutorials/text-salience/index.html

@@ -374,7 +374,7 @@ <h3>Conclusion</h3>
 techniques can be used in combination with other LIT modules like
 counterfactuals to support robust exploration of a model's behavior, as
 illustrated in this tutorial. And as always, LIT strives to enable users to
-<a href="https://github.com/PAIR-code/lit/wiki/api.md#interpretation-components">add their own salience interpreters</a>
+<a href="https://github.com/PAIR-code/lit/blob/main/documentation/api.md#interpretation-components">add their own salience interpreters</a>
 to allow for a wider variety of use cases beyond these default capabilities!</p>
 
           </div>

lit_nlp/client/core/global_settings.ts

@@ -218,7 +218,7 @@ export class GlobalSettingsComponent extends MobxLitElement {
 
           <lit-tooltip content="Go to reference">
             <a slot="tooltip-anchor" target='_blank'
-              href='https://github.com/PAIR-code/lit/wiki/ui_guide.md#global-settings'>
+              href='https://github.com/PAIR-code/lit/blob/main/documentation/ui_guide.md#global-settings'>
             <mwc-icon class="icon-button large-icon" id="configure-lit-icon">
               open_in_new
             </mwc-icon>

lit_nlp/client/modules/attention_module.ts

@@ -46,7 +46,7 @@ export class AttentionModule extends LitModule {
   static override title = 'Attention';
   static override infoMarkdown =
       `Visualize \`AttentionHeads\`-type fields from your model.<br>
-      [Learn more.](https://github.com/PAIR-code/lit/wiki/components.md#attention)`;
+      [Learn more.](https://github.com/PAIR-code/lit/blob/main/documentation/components.md#attention)`;
   static override numCols = 3;
   static override collapseByDefault = true;
   static override duplicateForExampleComparison = true;

lit_nlp/client/modules/classification_module.ts

@@ -58,7 +58,7 @@ export class ClassificationModule extends LitModule {
   static override infoMarkdown =
       `Classification results are displayed for labels of \`MulticlassPreds\`
       and \`SparseMultilabelPreds\` type.<br>
-      [Learn more.](https://github.com/PAIR-code/lit/wiki/components.md#classification)`;
+      [Learn more.](https://github.com/PAIR-code/lit/blob/main/documentation/components.md#classification)`;
   static override duplicateForExampleComparison = false;
   static override duplicateForModelComparison = false;
   static override numCols = 3;

lit_nlp/client/modules/confusion_matrix_module.ts

@@ -56,7 +56,7 @@ export class ConfusionMatrixModule extends LitModule {
       `Compare predictions to gold labels, between two different models,
       or relative to other categorical fields. Click on cells or rows/column
       headers to select the associated examples.<br>
-      [Learn more.](https://github.com/PAIR-code/lit/wiki/components.md#confusion-matrix)`;
+      [Learn more.](https://github.com/PAIR-code/lit/blob/main/documentation/components.md#confusion-matrix)`;
   static override template =
       (model: string, selectionServiceIndex: number, shouldReact: number) => {
         return html`<confusion-matrix-module

lit_nlp/client/modules/embeddings_module.ts

@@ -84,7 +84,7 @@ export class EmbeddingsModule extends LitModule {
   static override infoMarkdown =
       `Pan, zoom, rotate, and click to visualize datapoints in the latent space
       of your model, in order to find clusters or patterns in the data.<br>
-      [Learn more.](https://github.com/PAIR-code/lit/wiki/components.md#embedding-projector)`;
+      [Learn more.](https://github.com/PAIR-code/lit/blob/main/documentation/components.md#embedding-projector)`;
   static override template =
       (model: string, selectionServiceIndex: number, shouldReact: number) => {
         return html`

lit_nlp/client/modules/generator_module.ts

@@ -90,7 +90,7 @@ export class GeneratorModule extends LitModule {
   static override title = 'Datapoint Generator';
   static override infoMarkdown =
       `Automatically generate counterfactuals using various techniques.<br>
-      [Learn more.](https://github.com/PAIR-code/lit/wiki/components.md#generators)`;
+      [Learn more.](https://github.com/PAIR-code/lit/blob/main/documentation/components.md#generators)`;
   static override numCols = 10;
 
   static override template =

lit_nlp/client/modules/salience_map_module.ts

@@ -106,7 +106,7 @@ export class SalienceMapModule extends LitModule {
   static override infoMarkdown =
       `Input salience methods try to explain model predictions as a heatmap
       over input features, such as tokens.<br>
-      [Learn more.](https://github.com/PAIR-code/lit/wiki/components.md#token-based-salience)`;
+      [Learn more.](https://github.com/PAIR-code/lit/blob/main/documentation/components.md#token-based-salience)`;
   static override numCols = 6;
   static override duplicateForExampleComparison = true;
   static override template = (model: string, selectionServiceIndex: number, shouldReact: number) => {

lit_nlp/client/modules/scalar_module.ts

@@ -94,7 +94,7 @@ export class ScalarModule extends LitModule {
       `Visualize the distribution of scalar (\`Scalar\`, \`RegressionScore\`,
       or \`MulticlassPreds\`) features over their range - for example, to find
       examples near the decision boundary.<br>
-      [Learn more.](https://github.com/PAIR-code/lit/wiki/components.md#scalar-plots)`;
+      [Learn more.](https://github.com/PAIR-code/lit/blob/main/documentation/components.md#scalar-plots)`;
   static override numCols = 4;
   static override template =
       (model: string, selectionServiceIndex: number, shouldReact: number) =>

lit_nlp/client/modules/tcav_module.ts

@@ -79,7 +79,7 @@ export class TCAVModule extends LitModule {
   static override infoMarkdown =
       `TCAV estimates the importance of high-level concepts (e.g., color,
       gender, race) for a prediction class.<br>
-      [Learn more.](https://github.com/PAIR-code/lit/wiki/components.md#tcav)`;
+      [Learn more.](https://github.com/PAIR-code/lit/blob/main/documentation/components.md#tcav)`;
   static override numCols = 12;
   static override duplicateForModelComparison = true;
 

lit_nlp/client/modules/thresholder_module.ts

@@ -58,7 +58,7 @@ export class ThresholderModule extends LitModule {
   static override infoMarkdown =
       `Set threshold scores that are used to determine whether examples belong
       to the positive class.<br>
-      [Learn more.](https://github.com/PAIR-code/lit/wiki/components.md#binary-classification-thresholds)`;
+      [Learn more.](https://github.com/PAIR-code/lit/blob/main/documentation/components.md#binary-classification-thresholds)`;
   static override numCols = 3;
   static override template =
       (model: string, selectionServiceIndex: number, shouldReact: number) =>

website/src/tutorials/tab-feat-attr.md

@@ -27,7 +27,7 @@ LIT supports many techniques like salience maps and counterfactual generators
 for text data. But what if you have a tabular dataset? You might want to find
 out which features (columns) are most relevant to the model’s predictions. LIT's
 Feature Attribution module for
-[tabular datasets](https://github.com/PAIR-code/lit/wiki/components.md#tabular-data)
+[tabular datasets](https://github.com/PAIR-code/lit/blob/main/documentation/components.md#tabular-data)
 support identification of these important features. This tutorial provides a
 walkthrough for this module within LIT, on the
 [Palmer Penguins dataset](https://allisonhorst.github.io/palmerpenguins/).
@@ -41,7 +41,7 @@ walkthrough for this module within LIT, on the
         over tabular data, which is basically a specially weighted local linear
         regression for estimating SHAP values and works for any model. For now,
         the feature attribution module is only shown in the UI when working with
-        [tabular data](https://github.com/PAIR-code/lit/wiki/components.md#tabular-data)."%}
+        [tabular data](https://github.com/PAIR-code/lit/blob/main/documentation/components.md#tabular-data)."%}
 
 ### **Overview**
 

website/src/tutorials/text-salience.md

@@ -356,5 +356,5 @@ sense of their language model’s predictions. This diverse array of built-in
 techniques can be used in combination with other LIT modules like
 counterfactuals to support robust exploration of a model's behavior, as
 illustrated in this tutorial. And as always, LIT strives to enable users to
-[add their own salience interpreters](https://github.com/PAIR-code/lit/wiki/api.md#interpretation-components)
+[add their own salience interpreters](https://github.com/PAIR-code/lit/blob/main/documentation/api.md#interpretation-components)
 to allow for a wider variety of use cases beyond these default capabilities!