Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.benchgen.com/llms.txt

Use this file to discover all available pages before exploring further.

Train Overview

Train is BenchGen’s fine-tuning module. It gives you a simple, configuration-driven way to improve a model on your specific task — without managing infrastructure.

What Train Does

Upload a dataset, choose a base model, configure a LoRA adapter, and Train handles the rest. When the run completes you download a merged model ready for deployment in Agents or evaluation in Eval. You get:
  • Managed fine-tuning runs (no GPU provisioning)
  • LoRA adapter training with configurable hyperparameters
  • One-click adapter merging
  • Inference endpoint for immediate testing

When to Use Train

SituationWhat to do
Eval shows a cluster of failuresExport those cases and fine-tune on them
You have domain-specific data to improve qualityUpload a custom dataset and run a targeted fine-tune
A merged model is deployed but driftingRe-run training with updated data
You want to test a smaller, cheaper modelFine-tune a compact base model on your task

What Train Hands Off

  • → Eval: run a benchmark against the fine-tuned model to measure improvement.
  • → Agents: connect the merged model as the LLM inside an agent.

Next Steps