Skip to main content

Create fine-tuning job

POST 
/finetuning/jobs

Starts a new fine-tuning job that trains a LoRA adapter on top of a supported base model using your training logs. The resulting model can be deployed for inference once training completes. Supported base models are returned by List base models. Training data source is determined by the combination of tags, custom_dataset, custom_logs_filename, or defaults to all logs. Fine-tuning is billed per hour of compute used.

Request​

Responses​

Created