Auto-Train
Automated model training and optimization.
All methods are accessed via client.autotrain.
summarize_dataset() POST /v1/autotrain/summarize-dataset Summarize a dataset by uploading a file.
Parameters
file_path str Required
Path to the dataset file
team_id str default: None
Team ID (uses session team_id if not provided)
textgen_config TextGenConfig default: None
Text generation configuration
Returns
DatasetSummary — Dataset summary and metadata
Example
1 result = client . autotrain . summarize_dataset (
2 file_path = "./data.csv" ,
3 team_id = "team_abc123" ,
4 textgen_config = "..."
5 )
summarize_by_dataset_id() POST /v1/autotrain/summarize Summarize a dataset that's already on the xplainable platform. Downloads the dataset server-side and returns column statistics, types, distributions, and metadata. Use this to understand a dataset before training -- Claude never needs to see the raw data. Use datasets_list_team_datasets to find available dataset IDs.
Parameters
dataset_id str Required
ID of the dataset on the platform
team_id str default: None
Team ID (uses session team_id if not provided)
Returns
dict — Dataset summary with column statistics, types, and metadata
Example
1 result = client . autotrain . summarize_by_dataset_id (
2 dataset_id = "ds_abc123" ,
3 team_id = "team_abc123"
4 )
generate_goals() POST /v1/autotrain/generate_goals Generate training goals based on dataset summary.
Parameters
summary DatasetSummary Required
Dataset summary from summarize_dataset
team_id str default: None
Team ID (uses session team_id if not provided)
n int default: 5
Number of goals to generate
textgen_config TextGenConfig default: None
Text generation configuration
Returns
list — List of training goals
Example
1 result = client . autotrain . generate_goals (
2 summary = "..." ,
3 team_id = "team_abc123" ,
4 n = 5 ,
5 textgen_config = "..."
6 )
generate_labels() POST /v1/autotrain/generate_labels Generate label recommendations for training.
Parameters
summary DatasetSummary Required
Dataset summary from summarize_dataset
team_id str default: None
Team ID (uses session team_id if not provided)
textgen_config TextGenConfig default: None
Text generation configuration
Returns
list — List of label recommendations
Example
1 result = client . autotrain . generate_labels (
2 summary = "..." ,
3 team_id = "team_abc123" ,
4 textgen_config = "..."
5 )
generate_feature_engineering() POST /v1/autotrain/generate_feature_engineering Generate feature engineering recommendations.
Parameters
summary DatasetSummary Required
Dataset summary from summarize_dataset
team_id str default: None
Team ID (uses session team_id if not provided)
n int default: 5
Number of recommendations to generate
textgen_config TextGenConfig default: None
Text generation configuration
Returns
list — List of feature engineering recommendations
Example
1 result = client . autotrain . generate_feature_engineering (
2 summary = "..." ,
3 team_id = "team_abc123" ,
4 n = 5 ,
5 textgen_config = "..."
6 )
start_autotrain() POST /v1/autotrain/start_autotrain Start the autotrain process.
Parameters
model_name str Required
Name for the model
model_description str Required
Description of the model
summary DatasetSummary Required
Dataset summary from summarize_dataset
team_id str default: None
Team ID (uses session team_id if not provided)
textgen_config TextGenConfig default: None
Text generation configuration
Returns
AutotrainResponse — Training job information
Example
1 result = client . autotrain . start_autotrain (
2 model_name = "My Model" ,
3 model_description = "Predicts customer churn" ,
4 summary = "..." ,
5 team_id = "team_abc123" ,
6 textgen_config = "..."
7 )
check_training_status() GET /v1/autotrain/check_training_status Check the status of a training job.
Parameters
training_id str Required
Training job ID from start_autotrain
team_id str default: None
Team ID (uses session team_id if not provided)
Returns
TrainingStatus — Training status and progress information
Example
1 result = client . autotrain . check_training_status (
2 training_id = "train_abc123" ,
3 team_id = "team_abc123"
4 )
train_manual() POST /v1/autotrain/train_manual Train a model manually with specific parameters.
Parameters
label str Required
Target label column
model_name str Required
Name for the model
model_description str Required
Description of the model
preprocessor_id str Required
Preprocessor ID
version_id str Required
Preprocessor version ID
team_id str default: None
Team ID (uses session team_id if not provided)
drop_columns list default: None
Columns to drop
Returns
AutotrainResponse — Training job information
Example
1 result = client . autotrain . train_manual (
2 label = "target" ,
3 model_name = "My Model" ,
4 model_description = "Predicts customer churn" ,
5 preprocessor_id = "pp_abc123" ,
6 version_id = "version_xyz789" ,
7 team_id = "team_abc123"
8 )
visualize_data() Generate data visualizations.
Parameters
summary DatasetSummary Required
Dataset summary
goal dict Required
Visualization goal
team_id str default: None
Team ID (uses session team_id if not provided)
library str default: 'plotly'
Visualization library (plotly, matplotlib, seaborn)
textgen_config TextGenConfig default: None
Text generation configuration
Returns
VisualizationResponse — Visualization code and metadata
Example
1 result = client . autotrain . visualize_data (
2 summary = "..." ,
3 goal = { } ,
4 team_id = "team_abc123" ,
5 library = "plotly" ,
6 textgen_config = "..."
7 )
generate_insights() Generate insights about the dataset.
Parameters
goal dict Required
Analysis goal
summary DatasetSummary Required
Dataset summary
team_id str default: None
Team ID (uses session team_id if not provided)
textgen_config TextGenConfig default: None
Text generation configuration
Returns
dict — Generated insights and analysis
Example
1 result = client . autotrain . generate_insights (
2 goal = { } ,
3 summary = "..." ,
4 team_id = "team_abc123" ,
5 textgen_config = "..."
6 )
