Skip to main content
Osmosis Platform is the web dashboard for managing reinforcement learning training of LLMs. It handles workspace setup, GitHub repository connection, GPU provisioning, training orchestration, metrics collection, and checkpoint deployment so you can focus on defining agent behavior and evaluation logic.

Core Capabilities

Workspaces

Organize your team, datasets, training runs, models, and workspace repository access with role-based permissions.

Training Runs

Submit, monitor, and manage RL training runs with configurable hyperparameters and checkpoint cadence.

Datasets

Upload and validate JSONL, CSV, or Parquet datasets up to 5 GB for training.

Models

List supported base models and deploy trained LoRA models for inference.

Monitoring

Track training run status, metrics, checkpoints, and outputs.

Git Integration

Create or connect a workspace repository to sync rollouts and configs automatically.

How It Works

The typical workflow from setup to deployed LoRA model follows five stages:
1

Complete onboarding

Start with Onboarding to create or join a platform workspace, connect GitHub, clone the workspace repository, install the CLI, and verify local workspace context.
2

Choose a first workflow

Run the included Multiply example for a known-good first training run, or use Create Your Own Rollout when you already have a task or dataset.
3

Push and submit an evaluation run

Push rollout changes to GitHub. Git Sync publishes the code version, then osmosis eval submit starts an evaluation run against a platform dataset to catch dataset, dependency, workflow, and grader issues before a training run.
4

Submit a training run

Once evaluation run results look healthy, osmosis train submit starts the training run.
5

Monitor and deploy

Track metrics, checkpoints, and outputs in the dashboard. When a run finishes, deploy a LoRA model.

Ready to Get Started?

Onboarding

Follow the full workspace setup flow for creators and invited members.

Quickstart

After onboarding, run the included example from evaluation run to training run.