HumanSignal · robot-ci-heartex · Jan 29, 2025 · Feb 10, 2025 · Feb 21, 2025 · Apr 30, 2025
diff --git a/docs/source/guide/ml_tutorials.html b/docs/source/guide/ml_tutorials.html
@@ -7,7 +7,7 @@
   - Hugging Face
   hide_frontmatter_title: true
   hide_menu: true
-  image: /tutorials/bert.png
+  image: /guide/ml_tutorials/bert.png
   meta_description: Tutorial on how to use BERT-based text classification with your
     Label Studio project
   meta_title: BERT-based text classification
@@ -22,7 +22,7 @@
   - EasyOCR
   hide_frontmatter_title: true
   hide_menu: true
-  image: /tutorials/easyocr.png
+  image: /guide/ml_tutorials/easyocr.png
   meta_description: The EasyOCR model connection integrates the capabilities of EasyOCR
     with Label Studio to assist in machine learning labeling tasks involving Optical
     Character Recognition (OCR).
@@ -38,7 +38,7 @@
   - Flair
   hide_frontmatter_title: true
   hide_menu: true
-  image: /tutorials/flair.png
+  image: /guide/ml_tutorials/flair.png
   meta_description: Tutorial on how to use Label Studio and Flair for faster NER labeling
   meta_title: Use Flair with Label Studio
   order: 75
@@ -54,7 +54,7 @@
   - Hugging Face
   hide_frontmatter_title: true
   hide_menu: true
-  image: /tutorials/gliner.png
+  image: /guide/ml_tutorials/gliner.png
   meta_description: Tutorial on how to use GLiNER with your Label Studio project to
     complete NER tasks
   meta_title: Use GLiNER for NER annotation
@@ -70,7 +70,7 @@
   - Grounding DINO
   hide_frontmatter_title: true
   hide_menu: true
-  image: /tutorials/grounding-dino.png
+  image: /guide/ml_tutorials/grounding-dino.png
   meta_description: Label Studio tutorial for using Grounding DINO for zero-shot object
     detection in images
   meta_title: Image segmentation in Label Studio using a Grounding DINO backend
@@ -88,7 +88,7 @@
   - Segment Anything Model
   hide_frontmatter_title: true
   hide_menu: true
-  image: /tutorials/grounding-sam.png
+  image: /guide/ml_tutorials/grounding-sam.png
   meta_description: Label Studio tutorial for using Grounding DINO and SAM for zero-shot
     object detection in images
   meta_title: Image segmentation in Label Studio using a Grounding DINO backend and
@@ -106,7 +106,7 @@
   - Hugging Face
   hide_frontmatter_title: true
   hide_menu: true
-  image: /tutorials/hf-llm.png
+  image: /guide/ml_tutorials/hf-llm.png
   meta_description: This tutorial explains how to run Hugging Face Large Language
     model backend in Label Studio. Hugging Face Large Language Model Backend is a
     machine learning backend designed to work with Label Studio, providing a custom
@@ -123,7 +123,7 @@
   - Hugging Face
   hide_frontmatter_title: true
   hide_menu: true
-  image: /tutorials/hf-ner.png
+  image: /guide/ml_tutorials/hf-ner.png
   meta_description: This tutorial explains how to run a Hugging Face NER backend in
     Label Studio.
   meta_title: Label Studio tutorial to run Hugging Face NER backend
@@ -138,7 +138,7 @@
   - Interactive matching
   hide_frontmatter_title: true
   hide_menu: true
-  image: /tutorials/interactive-substring-matching.png
+  image: /guide/ml_tutorials/interactive-substring-matching.png
   meta_description: Use the interactive substring matching model for labeling NER
     tasks in Label Studio
   meta_title: Interactive substring matching for NER tasks
@@ -155,7 +155,7 @@
   - Langchain
   hide_frontmatter_title: true
   hide_menu: true
-  image: /tutorials/langchain.png
+  image: /guide/ml_tutorials/langchain.png
   meta_description: Use Langchain, OpenAI, and Google to generate responses based
     on Google search results.
   meta_title: RAG with a Langchain search agent
@@ -173,7 +173,7 @@
   - ChatGPT
   hide_frontmatter_title: true
   hide_menu: true
-  image: /tutorials/llm-interactive.png
+  image: /guide/ml_tutorials/llm-interactive.png
   meta_description: Label Studio tutorial for interactive LLM labeling with OpenAI,
     Azure, or Ollama
   meta_title: Interactive LLM labeling with OpenAI, Azure, or Ollama
@@ -190,7 +190,7 @@
   - MMDetection
   hide_frontmatter_title: true
   hide_menu: true
-  image: /tutorials/openmmlab.png
+  image: /guide/ml_tutorials/openmmlab.png
   meta_description: This is a tutorial on how to use the example MMDetection model
     backend with Label Studio for image segmentation tasks.
   meta_title: Object detection in images with Label Studio and MMDetection
@@ -206,7 +206,7 @@
   - NVidia
   hide_frontmatter_title: true
   hide_menu: true
-  image: /tutorials/nvidia.png
+  image: /guide/ml_tutorials/nvidia.png
   meta_description: Tutorial on how to use set up Nvidia NeMo to use for ASR tasks
     in Label Studio
   meta_title: Automatic Speech Recognition with NeMo
@@ -215,14 +215,30 @@
   title: Automatic Speech Recognition with NVidia NeMo
   type: guide
   url: /tutorials/nemo_asr.html
+- categories:
+  - Computer Vision
+  - Optical Character Recognition
+  - PaddleOCR
+  hide_frontmatter_title: true
+  hide_menu: true
+  image: /guide/ml_tutorials/ppocr.png
+  meta_description: The PP-OCR model connection integrates PaddleX's OCR pipeline
+    with Label Studio for high-accuracy text detection and recognition supporting
+    100+ languages.
+  meta_title: PP-OCR model connection for text detection and recognition
+  order: 41
+  tier: all
+  title: PP-OCR Text Detection and Recognition
+  type: guide
+  url: /tutorials/ppocr.html
 - categories:
   - Computer Vision
   - Image Annotation
   - Object Detection
   - Segment Anything Model
   hide_frontmatter_title: true
   hide_menu: true
-  image: /tutorials/sam2-images.png
+  image: /guide/ml_tutorials/sam2-images.png
   meta_title: Using SAM2 with Label Studio for Image Annotation
   order: 15
   tier: all
@@ -236,7 +252,7 @@
   - Segment Anything Model
   hide_frontmatter_title: true
   hide_menu: true
-  image: /tutorials/sam2-video.png
+  image: /guide/ml_tutorials/sam2-video.png
   meta_title: Using SAM2 with Label Studio for Video Annotation
   order: 15
   tier: all
@@ -252,7 +268,7 @@
   - ONNX
   hide_frontmatter_title: true
   hide_menu: true
-  image: /tutorials/segment-anything.png
+  image: /guide/ml_tutorials/segment-anything.png
   meta_description: Label Studio tutorial for labeling images with MobileSAM or ONNX
     SAM.
   meta_title: Interactive annotation in Label Studio with Segment Anything Model (SAM)
@@ -267,7 +283,7 @@
   - Scikit-learn
   hide_frontmatter_title: true
   hide_menu: true
-  image: /tutorials/scikit-learn.png
+  image: /guide/ml_tutorials/scikit-learn.png
   meta_description: Tutorial on how to use an example ML backend for Label Studio
     with Scikit-learn logistic regression
   meta_title: Sklearn Text Classifier model for Label Studio
@@ -282,7 +298,7 @@
   - SpaCy
   hide_frontmatter_title: true
   hide_menu: true
-  image: /tutorials/spacy.png
+  image: /guide/ml_tutorials/spacy.png
   meta_description: Tutorial on how to use Label Studio and spaCy for faster NER and
     POS labeling
   meta_title: Use spaCy models with Label Studio
@@ -297,7 +313,7 @@
   - Tesseract
   hide_frontmatter_title: true
   hide_menu: true
-  image: /tutorials/tesseract.png
+  image: /guide/ml_tutorials/tesseract.png
   meta_description: Tutorial for how to use Label Studio and Tesseract to assist with
     your OCR projects
   meta_title: Interactive bounding boxes OCR in Label Studio with a Tesseract backend
@@ -312,7 +328,7 @@
   - WatsonX
   hide_frontmatter_title: true
   hide_menu: true
-  image: /tutorials/watsonx.png
+  image: /guide/ml_tutorials/watsonx.png
   meta_title: Integrate WatsonX with Label Studio
   order: 15
   tier: all
@@ -326,7 +342,7 @@
   - YOLO
   hide_frontmatter_title: true
   hide_menu: true
-  image: /tutorials/yolo.png
+  image: /guide/ml_tutorials/yolo.png
   meta_description: Tutorial on how to use an example ML backend for Label Studio
     with YOLO
   meta_title: YOLO ML Backend for Label Studio
@@ -335,22 +351,6 @@
   title: YOLO ML Backend for Label Studio
   type: guide
   url: /tutorials/yolo.html
-- categories:
-  - Computer Vision
-  - Video Classification
-  - Temporal Labeling
-  - LSTM
-  hide_frontmatter_title: true
-  hide_menu: true
-  image: /tutorials/yolo-video-classification.png
-  meta_description: Tutorial on how to use an example ML backend for Label Studio
-    with TimelineLabels
-  meta_title: TimelineLabels ML Backend for Label Studio
-  order: 51
-  tier: all
-  title: TimelineLabels ML Backend for Label Studio
-  type: guide
-  url: /tutorials/yolo_timeline_labels.html
 layout: templates
 meta_description: Tutorial documentation for setting up a machine learning model with
   predictions using PyTorch, GPT2, Sci-kit learn, and other popular frameworks.

diff --git a/docs/source/tutorials/deepgram.md b/docs/source/tutorials/deepgram.md
@@ -0,0 +1,34 @@
+Using Deepgram with Label Studio for Text to Speech annotation
+
+This backend uses the Deepgram API to take the input text from the user, do text to speech, and return the output audio for annotation in Label Studio.
+
+https://github.com/user-attachments/assets/9569a955-0baf-4a95-9e8a-d08250a0a298
+
+
+IMPORTANT NOTE: YOU MUST REFRESH THE PAGE AFTER SUBMITTING THE TEXT TO SEE THE AUDIO APPEAR.
+
+## Prerequistes 
+1. [Deepgram API Key](https://deepgram.com/) -- create an account and follow the instructions to get an api key with default permissions. Store this key as `DEEPGRAM_API_KEY` in `docker_compose.yml`
+2. AWS Storage -- make sure you configure the following parameters in `docker_compose.yml`: 
+      - `AWS_ACCESS_KEY_ID` -- your AWS access key id
+      - `AWS_SECRET_ACCESS_KEY` -- your AWS secret access key
+      - `AWS_SESSION_TOKEN` -- your AWS session token
+      - `AWS_DEFAULT_REGION` - the region you want to use for S3
+      - `S3_BUCKET` -- the name of the bucket where you'd like to store the created audio files
+      - `S3_FOLDER` -- the name of the folder within the specified bucket where you'd like to store the audio files. 
+3. Label Studio -- make sure you set your `LABEL_STUDIO_URL` and your `LABEL_STUDIO_API_KEY` in `docker_compose.yml`. As of 11/12/25, you must use the LEGACY TOKEN. 
+
+## Labeling Config 
+This is the base labeling config to be used with this backend. Note that you may add additional annotations to the document after the audio without breaking anything!
+```
+<View>
+  <Header value="What would you like to TTS?"/>
+  <TextArea name="text" toName="audio" placeholder="What do you want to tts?" value="$text" valrows="4" maxSubmissions="1"/>
+  <Audio name="audio" value="$audio" zoom="true" hotkey="ctrl+enter"/>
+</View>
+```
+## A Data Note 
+Note that in order for this to work, you need to upload dummy data (i.e. empty text and audio) so that the tasks populate. You can use `dummy_data.json` as this data. 
+
+## Configuring the backend 
+When you attach the model to Label Studio in your model settings, make sure to toggle ON interactive preannotations!
diff --git a/docs/source/tutorials/gliner.md b/docs/source/tutorials/gliner.md
@@ -82,4 +82,17 @@ The following common parameters are available:
 - `WORKERS` - Specify the number of workers for the model server.
 - `THREADS` - Specify the number of threads for the model server.
 - `LABEL_STUDIO_URL` - Specify the URL of your Label Studio instance. Note that this might need to be `http://host.docker.internal:8080` if you are running Label Studio on another Docker container.
-- `LABEL_STUDIO_API_KEY`- Specify the API key for authenticating your Label Studio instance. You can find this by logging into Label Studio and and [going to the **Account & Settings** page](https://labelstud.io/guide/user_account#Access-token).
+- `LABEL_STUDIO_API_KEY`- Specify the API key for authenticating your Label Studio instance. You can find this by logging into Label Studio and and [going to the **Account & Settings** page](https://labelstud.io/guide/user_account#Access-token). 
+
+## A Note on Model Training 
+
+If you plan to use a webhook to train this model on "Start Training", note that you do 
+not need to configure a separate webhook. Instead, go to the three dots next to your model
+on the Model tab in your project settings and click "start training". 
+
+Additionally, note that this container has been set for a **VERY SMALL** demo set, with only 1
+non-eval sample (we expect the first 10 data samples to be for evaluation.)
+
+If you're working with a larger dataset, be sure to: 
+1. update num_steps and batch size to the number of training steps you want and the batch size that works for your dataset. 
+2. change the uploaded model after training (line 239 of `model.py`) to the highest checkpoint that you have.
diff --git a/docs/source/tutorials/mmdetection-3.md b/docs/source/tutorials/mmdetection-3.md
@@ -23,7 +23,7 @@ https://mmdetection.readthedocs.io/en/latest/
 This example demonstrates how to use the MMDetection model with Label Studio to annotate images with bounding boxes. 
 The model is based on the YOLOv3 architecture with a MobileNetV2 backbone and trained on the COCO dataset. 
 
-![screenshot.png](/tutorials/screenshot.png)
+![screenshot.png](screenshot.png)
 
 ## Before you begin
 
@@ -43,7 +43,7 @@ docker-compose up -d
 
 See the tutorial in the documentation for building your own image and advanced usage:
 
-https://github.com/HumanSignal/label-studio/blob/develop/docs/source/tutorials/object-detector.md
+https://github.com/HumanSignal/label-studio/blob/master/docs/source/tutorials/object-detector.md
 
 
 ## Labeling config
@@ -85,7 +85,7 @@ In this example, you can combine multiple labels into one Label Studio annotatio
 1. Clone the Label Studio ML Backend repository in your directory of choice:
 
 ```
-git clone https://github.com/HumanSignal/label-studio-ml-backend
+git clone https://github.com/heartexlabs/label-studio-ml-backend
 cd label-studio-ml-backend/label_studio_ml/examples/mmdetection-3
 ```
 
@@ -166,4 +166,4 @@ gunicorn --preload --bind :9090 --workers 1 --threads 1 --timeout 0 _wsgi:app
 ```
 
 * Use this guide to find out your access token: https://labelstud.io/guide/api.html
-* You can use and increased value of `SCORE_THRESHOLD` parameter when you see a lot of unwanted detections or lower its value if you don't see any detections.
+* You can use and increased value of `SCORE_THRESHOLD` parameter when you see a lot of unwanted detections or lower its value if you don't see any detections.
diff --git a/docs/source/tutorials/segment_anything_2_image.md b/docs/source/tutorials/segment_anything_2_image.md
@@ -131,19 +131,32 @@ cd label_studio_ml/examples/segment_anything_2_image
 pip install -r requirements.txt
 ```
 
-2. Download [`segment-anything-2` repo](https://github.com/facebookresearch/segment-anything-2) into the root directory. Install SegmentAnything model and download checkpoints using [the official Meta documentation](https://github.com/facebookresearch/segment-anything-2?tab=readme-ov-file#installation)
+2. Download [`segment-anything-2` repo](https://github.com/facebookresearch/sam2) into the root directory. Install SegmentAnything model and download checkpoints using [the official Meta documentation](https://github.com/facebookresearch/sam2?tab=readme-ov-file#installation)
+You should now have the following folder structure: 
+
+
+    | root directory 
+        | label-studio-ml-backend 
+            | label-studio-ml
+                | examples 
+                    | segment_anything_2_image
+        | sam2
+            | sam2
+            | checkpoints
 
 
 3. Then you can start the ML backend on the default port `9090`:
 
 ```bash
-cd ../
-label-studio-ml start ./segment_anything_2_image
+cd ~/sam2
+label-studio-ml start ../label-studio-ml-backend/label_studio_ml/examples/segment_anything_2_image
 ```
 
+Due to breaking changes from Meta [HERE](https://github.com/facebookresearch/sam2/blob/c2ec8e14a185632b0a5d8b161928ceb50197eddc/sam2/build_sam.py#L20), it is CRUCIAL that you run this command from the sam2 directory at your root directory. 
+
 4. Connect running ML backend server to Label Studio: go to your project `Settings -> Machine Learning -> Add Model` and specify `http://localhost:9090` as a URL. Read more in the official [Label Studio documentation](https://labelstud.io/guide/ml#Connect-the-model-to-Label-Studio).
 
-## Running with Docker (coming soon)
+## Running with Docker
 
 1. Start Machine Learning backend on `http://localhost:9090` with prebuilt image:
 

diff --git a/docs/source/tutorials/segment_anything_2_video.md b/docs/source/tutorials/segment_anything_2_video.md
@@ -21,7 +21,7 @@ This guide describes the simplest way to start using **SegmentAnything 2** with
 This repository is specifically for working with object tracking in videos. For working with images, 
 see the [segment_anything_2_image repository](https://github.com/HumanSignal/label-studio-ml-backend/tree/master/label_studio_ml/examples/segment_anything_2_image)
 
-![sam2](/tutorials/Sam2Video.gif)
+![sam2](./Sam2Video.gif)
 
 ## Before you begin
 
@@ -83,4 +83,4 @@ If you want to contribute to this repository to help with some of these limitati
 
 ## Customization
 
-The ML backend can be customized by adding your own models and logic inside the `./segment_anything_2_video` directory.
+The ML backend can be customized by adding your own models and logic inside the `./segment_anything_2_video` directory.