Start Training
Start a new training job to fine-tune a model on your dataset. This creates your custom model or adapter based on one of the available base models.
Endpoint
POST https://api.platform.knowledgator.com/trainings/
Headers
| Name | Type | Description |
|---|---|---|
KnowledgatorAPI-Key | String | Your API key |
Content-Type | String | application/json |
Request Body
| Field | Type | Required | Description |
|---|---|---|---|
model_id | String | Yes | Base model ID to fine-tune (e.g. comprehend-it). See List Models (GET /models/) |
dataset_id | String | Yes | Dataset ID to use for training (2-64 chars, alphanumeric with hyphens) |
hyperparameters | Object | No | Optional training hyperparameters (e.g., learning_rate, batch_size) |
Response
{
"id": "78d379da228a4985ba19100f53245b79",
"model_id": "comprehend-it",
"dataset_id": "6babd48a-2256-4466-b16b-9fd45ba59688",
"status": "pending",
"hyperparameters": {
"learning_rate": 0.001,
"batch_size": 32
},
"created_at": "2024-01-15T10:30:00Z",
"updated_at": "2024-01-15T10:30:00Z",
"started_at": null,
"completed_at": null,
"error": null
}
Response Fields
id: Unique training job identifiermodel_id: Base model ID being traineddataset_id: Dataset ID used for trainingstatus: Training status (pending, running, successful, failed, stopped)hyperparameters: Training hyperparameters usedcreated_at: ISO 8601 timestamp of creationupdated_at: ISO 8601 timestamp of last updatestarted_at: ISO 8601 timestamp when training started (null if not started)completed_at: ISO 8601 timestamp when training completed (null if not completed)error: Error message if training failed (null if no error)
Error Codes
| Status Code | Description |
|---|---|
| 201 | Created – Training job created successfully. |
| 400 | Bad Request – Invalid input, validation failed, or fine-tuned model limit exceeded. |
| 401 | Unauthorized – Invalid API key or email verification required. |
| 403 | Forbidden – Invalid CSRF token. |
| 500 | Internal Server Error – Server-side issue. |
| 503 | Service Unavailable – Service temporarily unavailable. |
Example Usage
SDK Usage
💡 Dedicated SDK Guides: See Python SDK | TypeScript SDK
TypeScript/JavaScript
import { createClient } from '@knowledgator/typescript-sdk';
const kg = createClient('YOUR_API_KEY');
// Start training with default hyperparameters
const training = await kg.finetuning.startTraining({
model_id: 'comprehend-it',
dataset_id: '6babd48a-2256-4466-b16b-9fd45ba59688'
});
console.log(`Training started: ${training.id}`);
console.log(`Status: ${training.status}`);
// Start training with custom hyperparameters
const customTraining = await kg.finetuning.startTraining({
model_id: 'comprehend-it',
dataset_id: '6babd48a-2256-4466-b16b-9fd45ba59688',
hyperparameters: {
learning_rate: 0.001,
batch_size: 32,
epochs: 10
}
});
console.log(`Custom training started: ${customTraining.id}`);
Python
from knowledgator import Knowledgator
kg = Knowledgator(api_key="YOUR_API_KEY")
# Start training with default hyperparameters
training = kg.finetuning.start_training(
model_id="comprehend-it",
dataset_id="6babd48a-2256-4466-b16b-9fd45ba59688"
)
print(f"Training started: {training.id}")
print(f"Status: {training.status}")
# Start training with custom hyperparameters
custom_training = kg.finetuning.start_training(
model_id="comprehend-it",
dataset_id="6babd48a-2256-4466-b16b-9fd45ba59688",
hyperparameters={
"learning_rate": 0.001,
"batch_size": 32,
"epochs": 10
}
)
print(f"Custom training started: {custom_training.id}")
Expected SDK Output
Training started: 78d379da228a4985ba19100f53245b79
Status: pending
Custom training started: 89e248ab339c4912ca20211a64356c80
Python (requests)
import requests
url = "https://api.platform.knowledgator.com/trainings/"
headers = {
"KnowledgatorAPI-Key": "YOUR_API_KEY",
"Content-Type": "application/json",
}
data = {
"model_id": "comprehend-it",
"dataset_id": "6babd48a-2256-4466-b16b-9fd45ba59688",
"hyperparameters": {
"learning_rate": 0.001,
"batch_size": 32,
"epochs": 10
}
}
response = requests.post(url, headers=headers, json=data)
print(response.json())
cURL
curl -X POST "https://api.platform.knowledgator.com/trainings/" \
-H "KnowledgatorAPI-Key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model_id": "comprehend-it",
"dataset_id": "6babd48a-2256-4466-b16b-9fd45ba59688",
"hyperparameters": {
"learning_rate": 0.001,
"batch_size": 32,
"epochs": 10
}
}'
Expected Response (cURL & requests)
{
"id": "78d379da228a4985ba19100f53245b79",
"model_id": "comprehend-it",
"dataset_id": "6babd48a-2256-4466-b16b-9fd45ba59688",
"status": "pending",
"type": "model",
"created_at": "2024-01-15T10:30:00Z"
}
Additional Information
- Hyperparameters: Optional. If not provided, default values will be used
- Model Limits: Check your account limits for concurrent training jobs
- Use Cases:
- Fine-tune models on domain-specific data
- Experiment with different hyperparameter configurations
- Create specialized models for production use