Prerequisites

tracebloc is continually expanding supported data types and tasks to enable your use cases. In case your use case is not yet supported, please reach out to us at [email protected]. Before you can create a use case on the tracebloc website, make sure the following requirements are met:

You are registered as a user on the tracebloc platform
Your dataset is cleaned and preprocessed
You are familiar with the supported data types and tasks
To ingest the dataset, you have a secure environment running, locally or in the cloud. Preparing the files doesn’t need one.

Your path as a new user

Sign up for a tracebloc account.
Install your secure environment with the Quick Start. It also installs the tracebloc CLI.
Prepare your data against the dataset template for your task. You can do this before step 2; it doesn’t need a secure environment.
Ingest it with the tracebloc CLI, once as training data and once as test data.
Define the use case, including its evaluation metric.
Peers train models on it.
Evaluate the models on the leaderboard.

Supported Data Types and Tasks

The exact folder layout, CSV columns, ingest.yaml and validation rules for each task are on the dataset template pages. Where this overview and a template page differ, the template page is authoritative — it is derived from the ingestor’s own checks.

Image Data

Requirements for all image data tasks: Every image in a dataset has the same width and height and the same file type, for example all images as 256×256 RGB .jpg files. Any aspect ratio works; images don’t have to be square. tracebloc doesn’t resize or convert images when it ingests them, so convert, crop or resize them before you ingest. At training time, images are resized to the input size of the model a peer trains. The ingestor doesn’t check color mode or bit depth. If you combine datasets, declare color_mode (RGB or grayscale) and bit_depth (8 or 16) in the template so they can be checked for consistency.

Filenames in the label CSV: For all image and text tasks, the filename column may include the file extension or not: cat01 and cat01.jpg both resolve to images/cat01.jpg when the configured extension is .jpg. The extension is configured once for the whole dataset (file_options.extension in the template).

All images are validated before ingestion by the data ingestor. The ingestion process only starts when every file meets the requirements. Fix or remove any invalid images, then retry. One tracebloc data ingest takes at most 1 GiB in total and 500 MiB per file; see Size limit for larger datasets.

Task	Input file type	Color mode	Supported image size	Label file type	Requirements	Links
Classification	PNG, JPG, JPEG	RGB (3 channels) or grayscale (1 channel)	Any; one size for every image	CSV	Uniform image size and file type per dataset	Detailed structure Template
Keypoint Detection	PNG, JPG, JPEG	RGB (3 channels) or grayscale (1 channel)	Any; one size for every image	CSV	Uniform image sizes Same number of keypoints per image and class	Detailed structure Template
Object Detection	PNG, JPG, JPEG	RGB (3 channels) or grayscale (1 channel)	Any; one size for every image	Pascal VOC	Uniform image sizes, one xml per image	Detailed structure Template
Semantic Segmentation	PNG, JPG, JPEG	RGB (3 channels) or grayscale (1 channel)	Any; one size for every image	CSV + PNG masks	Uniform image and mask sizes	Detailed structure Template

Image Classification

train/
    labels.csv
    images/
        cat01.jpg
        dog02.jpg
        ...
test/
    ...

filename,label
cat01,cat
dog02,dog
...

The filename column may include the file extension or leave it out.

Image Keypoint Detection

The number of keypoints per image must be fixed: every row names the same keypoints, exactly as many as you declare for the dataset. You cannot mix classes with different keypoint counts (e.g., 16 for person and 32 for car) or annotate some images with fewer keypoints.

train/
    labels.csv
    images/
        image01.png
        image02.png
        ...
test/
    ...

Each row is one image. Annotation maps each keypoint name to its [x, y] coordinates, and Visibility uses the same keys with 1 (visible) or 0 (occluded or outside the image). Both are JSON objects inside a quoted CSV cell, with the inner quotes doubled:

filename,Annotation,Visibility,image_label
image01,"{""nose"": [100, 150], ""left_eye"": [92, 140]}","{""nose"": 1, ""left_eye"": 1}",person
image02,"{""nose"": [95, 155], ""left_eye"": [88, 146]}","{""nose"": 1, ""left_eye"": 0}",child

The image_label column needs at least two distinct values; the ingestor rejects a dataset with a single class. See the keypoint detection template for every rule the ingestor checks.

Image Object Detection

Images and their Pascal VOC XML annotations pair by file stem: images/street01.jpg belongs to annotations/street01.xml. There is no labels CSV; the image list and the classes are read from the XML files. See the object detection template for the full annotation format.

train/
    images/
        street01.jpg
        street02.jpg
        ...
    annotations/
        street01.xml
        street02.xml
        ...
test/
    ...

<annotation>
    <folder>images</folder>
    <filename>street01.jpg</filename>
    <source>
        <database>Unknown</database>
        <annotation>PASCAL VOC</annotation>
    </source>
    <size>
        <width>1920</width>
        <height>1080</height>
        <depth>3</depth>
    </size>
    <segmented>0</segmented>
    <object>
        <name>car</name>
        <pose>Unspecified</pose>
        <truncated>0</truncated>
        <difficult>0</difficult>
        <bndbox>
            <xmin>100</xmin>
            <ymin>200</ymin>
            <xmax>300</xmax>
            <ymax>400</ymax>
        </bndbox>
    </object>
    <!-- one <object> per box -->
</annotation>

Every element shown is required; the ingestor rejects a file that leaves one out.

Image Semantic Segmentation

Each image has one mask: a PNG named <image stem>_mask.png, with exactly the same width and height as the image, whose pixel values are class indices 0 to N-1. N is the number of distinct image_label values in the dataset, so if 0 is background, background must appear as an image_label value too. The labels CSV has one row per image: it links the image to its mask and names a class present in the image. Training reads each row as one sample, so list each image once. See the semantic segmentation template for the full rules.

train/
    labels.csv
    images/
        scene01.jpg
        scene02.jpg
        ...
    masks/
        scene01_mask.png
        scene02_mask.png
        ...
test/
    ...

filename,mask_id,image_label
scene01,scene01_mask,background
scene02,scene02_mask,road
scene03,scene03_mask,building
...

Tabular Data

Requirements for all tabular data tasks: Each dataset must be provided as a single CSV file with a header row. Every column must contain uniform data types, for example numeric values for features and a categorical or alphanumeric column for labels. Use UTF-8 encoding with comma separators and validate that your schema matches the expected types. Invalid rows are skipped by the ingestor.

Task	Data file type	Requirements	Links
Classification	CSV (features and label in one single file)	Uniform data formats per column. Feature columns: Numeric Label columns: Alphanumeric	Detailed structure Template
Regression	CSV (features and label in one single file)	Uniform data formats per column. Feature columns: Numeric Label column: Numeric (continuous target)	Detailed structure Template
Time Series Forecasting	CSV (timestamp, features and target in one single file)	A timestamp column in a parsable format (e.g. `YYYY-MM-DD` or ISO 8601). Feature columns: Numeric Target column: Numeric	Detailed structure Template
Time to Event Prediction	CSV (features, time and event in one single file)	A `time` column (duration until event or censoring, integer or numeric). An event column (binary 0/1 indicating whether the event occurred). Feature columns: Numeric	Detailed structure Template

Tabular Classification

Include a header row with clear column names, using a dedicated column for the labels. An id column is recommended but not required.

id,feature1,feature2,feature3,label
1,1.5,2.3,0.8,class_a
2,2.1,1.9,1.2,class_b
3,0.9,3.1,0.5,class_a
...

Tabular Regression

Same structure as Tabular Classification, but the label column holds a continuous numeric target (not a class).

id,square_feet,bedrooms,age,price
1,1668.08,3,15,285.50
2,1701.78,4,12,320.75
3,1697.01,2,8,245.30
...

Time Series Forecasting

Provide a single CSV with a timestamp column, one or more numeric feature columns, and the numeric target column you want to forecast. Rows must be ordered by time and use a consistent timestamp format.

timestamp,feature_1,feature_2,target
2023-10-01,7,1,125.50
2023-10-02,1,0,132.30
2023-10-03,2,0,128.75
...

Time to Event Prediction

Provide a single CSV with feature columns, a time column (duration to event or censoring), and a binary event column (1 = event occurred, 0 = censored).

age,feature_1,feature_2,time,event
75,0,1.9,4,1
55,0,1.1,6,1
65,0,1.3,7,0
...

Text Data

Task	Input files	Label file type	Requirements	Links
Classification	TXT	CSV	Text file may not be empty	Detailed structure Template

Text Classification

The filename column may include the file extension or leave it out. The extension is set once for the whole dataset (file_options.extension in the template).

train/
    labels.csv
    texts/
        review01.txt
        review02.txt
        ...
test/
    ...

filename,label
review01,positive
review02,negative
...

file example

# review01.txt
This product is amazing! I love it.

Next Steps

Prepare and ingest your dataset: Prepare Data

Need Help?

Email us at [email protected]

Overview

Create a Use Case

Join a Use Case

Environment Setup

Tools & Help

Releases

Your path as a new user

Supported Data Types and Tasks

Image Data

Image Classification

Image Keypoint Detection

Image Object Detection

Image Semantic Segmentation

Tabular Data

Tabular Classification

Tabular Regression

Time Series Forecasting

Time to Event Prediction

Text Data

Text Classification

Next Steps

Need Help?

​Your path as a new user

​Supported Data Types and Tasks

​Image Data

​Image Classification

​Image Keypoint Detection

​Image Object Detection

​Image Semantic Segmentation

​Tabular Data

​Tabular Classification

​Tabular Regression

​Time Series Forecasting

​Time to Event Prediction

​Text Data

​Text Classification

​Next Steps

​Need Help?

Your path as a new user

Supported Data Types and Tasks

Image Data

Image Classification

Image Keypoint Detection

Image Object Detection

Image Semantic Segmentation

Tabular Data

Tabular Classification

Tabular Regression

Time Series Forecasting

Time to Event Prediction

Text Data

Text Classification

Next Steps

Need Help?