start-run

name	start-run
description	How to launch prime-rl training runs — the `rl`, `sft`, `inference`, and `evals` entrypoints, their config classes, and single-node/SLURM/dry-run modes. Use when starting a run or picking the right entrypoint.

Start a run

All entrypoints run via uv run <command> and accept TOML configs via @ path/to.toml plus CLI overrides.

SLURM launches write generated scripts and coordination files under <run_dir>/launcher/, with batch logs under launcher/logs/. Local launches do not create this directory. Every launch writes configs and command.txt under configs/attempt_<n>/. configs/latest points to the current attempt. The command uses shell-safe quoting.

Run directories

output_dir (default outputs) groups related runs; each run writes all its artifacts (logs, configs, checkpoints, broadcasts, rollouts) to its own run directory <output_dir>/<run_name>. run.name auto-generates as <envs>--<model>--<short-id> (SFT: <dataset>--<model>--<short-id>), so every launch gets a fresh, readable run directory; run.dir overrides the directory leaf when it should differ from the name. Pass --run.name <name> to make the run directory predictable — required to resume the run later (--resume, or --resume.step N, reuses the named run directory; without [ckpt] it loads but saves no new checkpoints). Launching into a run directory that already contains artifacts fails unless resuming or --clean is set (which wipes only that run directory).

Config system at a glance

pydantic-config — Pydantic-based TOML + CLI loader. Highlights (see the configs skill for full mechanics):

• Config files via @ path (TOML / YAML / JSON); CLI args layer on top, deep-merged with class defaults.
• Nested groups via dotted CLI paths — kebab-case on the CLI, snake_case in TOML.
• Bool toggles: bare --flag enables, --no-flag disables (nested too).
• Lists: space-separated or JSON literal. Dicts: JSON literal, deep-merged with file values.
• Optional sub-configs (WandbMonitorConfig | None): bare --monitors.wandb enables defaults; --monitors.wandb @ wandb.toml enables from a file; --no-monitors.wandb disables.
• Discriminated unions are switched by the type tag (e.g. --optimizer.type muon).
• Validation aliases let renamed fields keep working; legacy keys can be remapped in a model_validator(mode="before").
• Auto-generated --help panels from Field(description=...) or PEP 224 docstrings.
• Friendly errors: required-field boxes, validator errors point at the offending flag, unknown flags get a "did you mean" hint.
• State-only optimizer offload remains enabled by default with model.optim_cpu_offload = true.
• For gradients, FP32 masters, optimizer state, and optimizer-in-backward CPU execution, set model.optim_cpu_offload = false and model.full_offload = true. This mode uses the native CPU optimizer kernel, only supports AdamW and SignSGD (SignSGD is stateless and halves the host RAM footprint), and disables gradient clipping. Use a [model.full_offload] table only to select the Torch debugging backend or disable NUMA binding.

rl — RL training

Launches inference server, orchestrator, and trainer as subprocesses.

uv run rl @ examples/basic/reverse-text/rl.toml
uv run rl @ examples/basic/reverse-text/rl.toml --dry-run                                # write scripts, don't run

• Config: RLConfig (packages/prime-rl-configs/src/prime_rl/configs/rl.py)
• Entrypoint: src/prime_rl/entrypoints/rl.py
• SLURM: single- and multi-node
• Multi-node SLURM stops after .trainer.done for trainer-only fake-data runs. Runs with inference stop after both .trainer.done and .orchestrator.done.
• NIXL on SLURM: install NIXL and ModelExpress with the provided scripts. The job starts ModelExpress and Redis unless slurm.launch_modelexpress = false.
• Environment packages: before launching a config with a non-core verifier env id, verify the package imports under uv run (for example uv run python -c "import importlib.util; print(importlib.util.find_spec('r2e_gym'))"). If a local env exists under deps/prime-envs/environments/ or deps/verifiers/environments/ but does not import, install the env workspace members with uv sync --all-extras --all-packages (all) or uv sync --all-extras --package prime-rl --package <env> (one) — they're auto-discovered, no pyproject.toml edit needed. Keep --all-extras for training so a targeted package sync does not prune accelerator dependencies from the environment.

sft — SFT training

Launches torchrun internally — never call torchrun directly.

uv run sft @ examples/basic/reverse-text/sft.toml
uv run sft @ examples/basic/reverse-text/sft.toml --slurm
uv run sft @ examples/basic/reverse-text/sft.toml --dry-run

• Config: SFTConfig (packages/prime-rl-configs/src/prime_rl/configs/sft.py)
• Entrypoint: src/prime_rl/entrypoints/sft.py
• SLURM: single- and multi-node
• Multi-node online evals use one SLURM job with num_train_nodes + num_infer_nodes nodes. The generated launcher/sft.sbatch assigns inference nodes first, then trainer nodes.

inference — vLLM server

OpenAI-compatible API plus prime-rl custom endpoints (/update_weights, /load_lora_adapter, /init_broadcaster). Always use this entrypoint — never vllm serve directly. It starts a vllm-router on server.port (default 8000, the client-facing URL) fronting the engine on backend_port (default 8100); admin endpoints must target the engine port directly.

uv run inference --vllm.model Qwen/Qwen3-0.6B
uv run inference --vllm.model Qwen/Qwen3-0.6B --vllm.enforce-eager

Smoke checks:

curl http://<host>:<port>/health
curl http://<host>:<port>/v1/models
curl http://localhost:8000/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{"model": "Qwen/Qwen3-0.6B", "messages": [{"role": "user", "content": "Hi"}], "max_tokens": 50}'

• Config: InferenceConfig (packages/prime-rl-configs/src/prime_rl/configs/inference.py)
• Entrypoint: src/prime_rl/entrypoints/inference.py
• SLURM: single-node, multi-node, and disaggregated deployments

evals — multi-env evals

Runs the configured eval sources against a live inference server. Standalone (no [online] block): one epoch of every source against the served weights, then exit. Add [ckpt] (interval counts completed task groups in the eval-source order) to make the run interruptible, then use --resume, --resume.step N, or --resume.dir path/to/checkpoints/step_N; for standalone evals, checkpoint step N means resume from task cursor N. Checkpoints store only the cursor, not generated episode records; partially completed groups and groups beyond the durable cursor are retried. Resume loads without [ckpt] but does not save new checkpoints. Checkpoint/resume is rejected with [online] because that process is coupled to the trainer's live broadcast handshake. With [online] (broadcasts_dir, max_steps, resume_step): watch the broadcasts dir for stable step_{n} weight broadcasts and evaluate each — the sft launcher writes this config for online evals. By default a newer checkpoint cancels unfinished episodes from the prior eval. Set eval.cancel_on_new_checkpoint = false to drain every epoch. The trainer can idle while it waits for slow evals. Launcher-managed SFT evals use NCCL weight broadcast by default, including multi-node SLURM deployments. LoRA and external inference use filesystem broadcast.

uv run inference --vllm.model Qwen/Qwen3-4B   # start inference separately
uv run evals @ eval.toml

Minimal standalone eval.toml:

model = "Qwen/Qwen3-4B"

[eval.client]
base_url = "http://localhost:8000/v1"

[eval.concurrency]  # adaptive; same controller as [orchestrator.concurrency]
min_inflight = 8
max_inflight = 128

[[eval.source]]
num_examples = 32   # always cap eval size for smokes
group_size = 4
env.taskset.id = "aime25"
env.agent.harness.id = "null"
env.agent.runtime.type = "subprocess"

[ckpt]               # optional: make a standalone eval resumable
interval = 10        # completed task groups between cursor saves

• Env servers: spawned by the evals process, one per source without an explicit serve.address, at tcp://127.0.0.1:<eval.env_server_base_port + index>; logs at {output_dir}/logs/latest/envs/eval/{name}.log.
• External inference APIs (no vLLM /metrics, e.g. Prime Inference) have no load signal for adaptive concurrency: the startup /metrics probe fails fast unless the band is pinned (min_inflight = max_inflight). Full example: examples/evals/swe.toml (SWE-bench Verified + Terminal-Bench 2 on Prime Inference, agent.timeout.rollout = 3600).
• Config: EvalsConfig (packages/prime-rl-configs/src/prime_rl/configs/evals.py)
• Entrypoint: src/prime_rl/entrypoints/evals.py (implementation: src/prime_rl/evals/evals.py)

Exporting checkpoints

Trainer checkpoints are DCP-sharded (<run_dir>/checkpoints/step_{n}/trainer). Convert to HF safetensors with uv run python tools/convert_dcp_to_bf16.py <run_dir>/checkpoints/step_{n} (writes <ckpt_dir>/weights, serveable via uv run inference --vllm.model <dir>; model config auto-read from the run’s configs/latest/resolved/trainer.json/sft.json; multi-rank via torchrun --nproc-per-node N; full fine-tunes only, LoRA rejected). Quantize a bf16 HF dir to blockwise FP8 with tools/convert_bf16_to_fp8.py <dir> (vLLM-native format), or straight from a checkpoint with tools/convert_dcp_to_fp8.py <ckpt_dir> (rank-parallel, writes only <ckpt_dir>/weights-FP8, no bf16 on disk); dequantize fp8-only releases with tools/convert_fp8_to_bf16.py <dir>. Caveat: on SM120 GPUs (RTX PRO 6000) vLLM 0.26 picks CutlassFp8BlockScaledMMKernel for blockwise-fp8 checkpoints and it silently degrades outputs — serve with VLLM_DISABLED_KERNELS=CutlassFp8BlockScaledMMKernel,MarlinFP8ScaledMMLinearKernel to fall back to the Triton kernel.

Summary

Command	Purpose	Typical use
rl	Full RL pipeline	Production RL training
sft	Supervised fine-tuning	SFT and hard-distill
inference	vLLM server	Standalone serving / debugging
evals	Multi-env evals	Standalone evals / SFT online evals

Key paths

• src/prime_rl/entrypoints/ — rl, sft, inference (+ trainer, orchestrator for direct launches)
• packages/prime-rl-configs/src/prime_rl/configs/ — all config classes
• configs/debug/ — minimal debug configs
• examples/ — full example configs (e.g. reverse-text/)

Dashboard

Interactive launches auto-start one shared dashboard daemon per user (process title PRL::Dashboard) and end startup with a Dashboard · <url> banner. Relay that URL to the researcher. Discovery: ~/.cache/prime-rl/dashboard/daemon.json holds the live url (the port can differ from 7788 when it was taken). --no-dashboard opts a run out.