Skip to main content
flash env setup scaffolds a starter project into the current directory. A run is fully described by what lands on disk: an environment (your task and how it’s scored) and a config (how to train on it). Every file is plain text you can read, diff, and version-control. Rerunning is safe: any file that already exists is left untouched. 
./
├── environment.py        # the task + reward (a Freesolo environment)
├── dataset/
│   └── train.jsonl       # a tiny starter dataset (input/output rows)
├── configs/
│   ├── sft.toml          # an SFT (supervised) training config
│   └── rl.toml           # a GRPO (RL) training config
└── TRAINING.md           # a playbook for the AI agent driving your runs
What it is: your environment, the single source of truth for what the model practices on and how it’s graded. It defines load_environment(), which returns a Freesolo EnvironmentSingleTurn (or EnvironmentMultiTurn) carrying a dataset and a score_response reward. This is the file you edit first.When Flash reads it: flash env push packages and uploads it; flash train --cost may import it locally to count training examples; and on every run the worker imports it and calls load_environment(**params). The scaffolded starter loads its rows from dataset/train.jsonl.See the scaffolded file in full — StarterEnv with build_prompt_messages and score_response — in Environments.
What it is: a tiny starter dataset of input/output rows that the scaffolded environment.py loads. Replace it with your real training rows before a real run. See Datasets.When Flash reads it: your environment.py reads it on the worker at run time (and locally when flash train --cost counts its rows; --dry-run validates only the config and does not read it); flash env push uploads the dataset/ folder with the environment.
dataset/train.jsonl
{"input":"What is 2 + 2?","output":"4"}
{"input":"What is 3 + 5?","output":"8"}
What it is: an SFT training config for supervised fine-tuning on the input/output pairs in your environment’s dataset. You set model, the [environment] id, and the [train] knobs (epochs, lora_rank); the training infrastructure and artifact storage are managed for you. Copy it per experiment.When Flash reads it: every flash train, --dry-run, and --cost parses this file into the resolved job spec.
configs/sft.toml
model = "Qwen/Qwen3.5-4B"
algorithm = "sft"

[environment]
id = ""   # paste the id returned by `flash env push --name my-env .`

[train]
epochs = 3
max_examples = 1000
lora_rank = 32
What it is: the same config shape with algorithm = "grpo", for GRPO (RL) training that optimizes against your environment’s score_response reward. Keep both and pick one at train time.When Flash reads it: same as sft.toml; pass it to flash train to run GRPO instead of SFT.
flash train configs/rl.toml
What it is: a playbook for the AI coding agent you point at this project — how to design the reward, what to read in a run’s output, and how to decide a run actually improved the model. It includes current CLI usage and common Flash issue mitigations.When it travels: if you publish the whole scaffolded folder, flash env push includes .md sidecars, so TRAINING.md can travel with the environment source in the Hub for humans and coding agents.

​
Packaging an environment as a folder

The scaffolded environment.py is enough to publish on its own. Once your task grows data files or helper modules, move it into a folder with environment.py at the root and publish the whole folder. 
math/
├── environment.py        # defines load_environment()
├── helpers.py            # optional sibling modules, imported by environment.py
└── dataset/
    ├── train.jsonl       # input/output records
    └── eval.jsonl
What it is: any sibling Python modules your environment.py imports. Keep imports either from installed packages or from files in the folder you publish.When Flash reads it: uploaded with the folder by flash env push and importable on the worker.
What it is: sidecar data files (.jsonl, .json, .csv, and more) that your environment loads. See Datasets for the full list.When Flash reads it: your environment code reads it (for example via load_task_examples(...)); flash env push includes the dataset/ folder in the uploaded artifact.
dataset/train.jsonl
{"input":"What is 2 + 2?","output":"4"}
{"input":"What is 3 + 5?","output":"8"}