Welcome to Easy-ML
Select a task pipeline to build your model
Tabular Classification
Predict categories from structured CSV data.
Tabular Regression
Predict continuous numbers from structured CSV data.
NLP Text Classification
Text classification, sentiment analysis via natural language.
Image Classification
Train deep learning models on image datasets.
Requires the CV extra: {{ installHint('cv') }}
Settings
Manage your data storage and application preferences
📂 Data Storage
{{ settingsInfo.data_dir }}⚡ GPU / CUDA
{{ cudaInfo.torch_version }}
pip install torch torchvision --index-url https://download.pytorch.org/whl/cu118
🗑️ Clear Data
Delete all runs, databases, uploads, and logs. This cannot be undone.
About Easy-ML
An open-source, no-code ML platform
👥 Developers
Upload Dataset
Upload a CSV or Excel file for {{ taskLabel(config.task) }}
Drag & drop your CSV / Excel file here, or click to browse
{{ uploadedFileName }} loaded successfully
{{ uploadedData.rows }} rows × {{ uploadedData.cols }} columns
Uploading...
Select Image Dataset Folder
Choose a folder inside your Easy-ML Data Directory (e.g. uploads/) that contains subfolders for each image class.
Column Configuration
Run Details
A run name is required.
Data Preview
| {{ col }} TARGET |
|---|
| {{ truncate(row[col], 50) }} |
Pipeline Configuration
Configure preprocessing and features for {{ taskLabel(config.task) }}
Text Cleaning
Imbalanced Data Handling
Applied after TF-IDF / Embeddings step.
Feature Extraction
Sentence embeddings need the NLP extra: {{ installHint('embeddings') }}
Image Preprocessing
Data Transforms
Column Assignments
Hold Ctrl/Cmd to select multiple.
Select Models
Choose algorithms and tuning effort
The {{ isCV ? 'Computer Vision' : 'NLP' }} module is not fully installed in your current environment.
Please open your terminal, stop the server, and run:
pip install "easy-ml-studio[{{ isCV ? 'cv' : 'nlp' }}]"
XGBoost, LightGBM and CatBoost are available with {{ installHint('boost') }}
Training Config
Stop training if no improvement for N epochs.
Reduce LR if no improvement for N epochs.
Multiply LR by this factor (e.g. 0.5 = halve it).
🔧 Custom CNN Builder
Add layers to build your architecture. The model always starts with the image input and ends with a Linear layer for your classes.
Hyperparameter Tuning (GridSearch)
Training Pipeline
{{ trainingStatusText }}
{{ Math.round(trainingProgress) }}%
📈 Live Training Curves
Training Results
Performance metrics for your models
🔗 Pipeline Summary
{{ modelLabel(modelName) }}
{{ formatScore(perf.accuracy_test) }}| actual ↓ / predicted → | {{ lbl }} |
|---|---|
| {{ classLabels(perf)[ri] }} | {{ cell }} |
| Class | Precision | Recall | F1 |
|---|---|---|---|
| {{ clsK }} | {{ cls.precision?.toFixed(3) }} | {{ cls.recall?.toFixed(3) }} | {{ cls['f1-score']?.toFixed(3) }} |
Live Inference
Make predictions with Run #{{ inference.run_id }} · {{ taskLabel(inferenceRunType) }}
Enter feature values for prediction:
Upload a CSV containing the same features to get a bulk prediction file.
Feature Impact (SHAP Waterfall)
Run History
Manage previous training runs
| ID | Name | Task | Dataset | Models | Best Acc | Date | Actions | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| {{ run.id }} | {{ run.run_name }} | {{ taskLabel(run.task) }} | {{ run.dataset_name || '—' }} | {{ getRunModels(run) }} | {{ getRunBestAcc(run) }} | {{ formatTimestamp(run.timestamp) }} | |||||||||||||
{{ modelLabel(modelName) }}{{ (metrics.accuracy_test * 100).toFixed(2) }}%Train Acc
{{ (metrics.accuracy_train * 100).toFixed(2) }}%
Test Acc
{{ (metrics.accuracy_test * 100).toFixed(2) }}%
🎯 Best Params
{{ key }}
{{ val }}
📊 Test Metrics
{{ k.replace(/_/g, ' ') }}
{{ v.toFixed(4) }}
Confusion Matrix
Classification Report
No detailed metrics available for this run. |
|||||||||||||||||||
No runs found in database.