train
- Repo stars 1,155
- License BSD-2-Clause
- Author updated Live
- Author repo LearningHumanoidWalking
- Domain
- AI
- Compatible agents
-
- Claude Code
- Cursor
- Cline
- Codex
- Windsurf
- Gemini CLI
- +20
- Trust score
- 94 / 100 · audit passed
- Author / version / license
- @rohanpsingh · BSD-2-Clause
- Token usage
- Lean
- Setup complexity
- Plug-and-play
- External API key
- Not required
- Operating systems
- Unspecified (assume cross-platform)
- Runtime requirements
- Python
- Permissions
-
- Read-only
- Write / modify
- Network behavior
- Local-only
- Install commands
- 26 variants
Profile is derived at build time from SKILL.md and install vectors. Subject to drift from author intent.
Heads up: 未限定 allowed-tools，默认拥有全部工具权限。
---
name: train
description: Launch a training run for a robot environment using PPO Parse the user's request from $ARGUMENTS…
category: ai
runtime: Python
---
# train output preview
## PART A: Task fit
- Use case: Launch a training run for a robot environment using PPO Parse the user's request from $ARGUMENTS and construct a training command. RAYADDRESS= uv run python runexperiment.py train --env <ENV> --logdir <LOGDIR> [OPTIONS...] runs entirely locally; runs on Python. Works with Claude Code, Cursor, Cline and 23 more..
- Inputs: target material, constraints, expected output, and acceptance criteria.
- Evidence boundary: follow “Command Template / Available Environments / Hyperparameters (defaults)” and do not present inference as author intent.
## PART B: Execution result
- **01** The card summarizes the use case; runtime output centers on “Launch a training run for a robot environment using PPO Parse the user's request from $ARGUMENTS and construct a training command. RAYADDRESS= uv run python runexperiment.py train --env <ENV> --logdir <LOGDIR> [OPTIONS...] runs entirely locally; runs on Python. Works with Claude Code, Cursor, Cline and 23 more.”.
- **02** When the source has headings, the agent prioritizes “Command Template / Available Environments / Hyperparameters (defaults)” so the result follows the author’s structure.
- **03** Typical output includes task judgment, concrete steps, required commands or file edits, validation, and follow-up options.
- **04** Risk context follows the fingerprint: read files, write/modify files; mostly runs locally; usually needs no extra API key.
## Running Rules
- read files, write/modify files; mostly runs locally; usually needs no extra API key.
- Validate with a small sample before expanding scope.
- Return the result, validation criteria, and next iteration options. Choose the trigger
The source mentions slash commands such as `/train`, `/tmp`; use them first when your agent supports command triggers.
Provide inputs and boundaries
Name target files or source material, expected output, forbidden changes, and whether network or shell access is allowed. Permission fingerprint: read files, write/modify files.
Validate with a small run
Start with a small task and check whether the result follows “Command Template / Available Environments / Hyperparameters (defaults)”. Inspect diffs, logs, previews, or tests before expanding scope.
Review before delivery
Confirm the final output includes a concrete result, evidence, and next action. If it stays generic, tighten inputs, boundaries, and acceptance criteria.
---
name: train
description: Launch a training run for a robot environment using PPO Parse the user's request from $ARGUMENTS…
category: ai
source: rohanpsingh/LearningHumanoidWalking
---
# train
## When to use
- Launch a training run for a robot environment using PPO Parse the user's request from $ARGUMENTS and construct a train…
- Use it when the task has clear inputs, repeatable steps, and validation criteria.
## What to provide
- Target material, scope, expected result, and forbidden changes.
- Whether network, commands, file writes, or external services are allowed.
## Execution rules
- Organize steps around “Command Template / Available Environments / Hyperparameters (defaults)” and keep inference separate from source facts.
- read files, write/modify files; mostly runs locally; usually needs no extra API key.
- Validate with a small sample before expanding the task.
## Output requirements
- Return the deliverable, key evidence, validation method, and next action.
- Mark missing information as unknown; do not invent commands, platforms, or dependencies. Evidence boundary and execution chain
The author source anchors workflow facts; repository files anchor sources and commands; Fluxly only adds fit, limitations, and quality judgment.
skill "train" {
input -> user goal + target files + boundaries + acceptance criteria
context -> Command Template / Available Environments / Hyperparameters (defaults)
rules -> SKILL.md triggers / order / output contract
runtime -> Python | read files, write/modify files | mostly runs locally
guardrails -> usually needs no extra API key + small-sample validation + diff/log review
output -> copyable result + checklist + next iteration
}/train — Launch a PPO Training Run
Parse the user's request from $ARGUMENTS and construct a training command.
Command Template
RAY_ADDRESS= uv run python run_experiment.py train --env <ENV> --logdir <LOGDIR> [OPTIONS...]Available Environments
| Name | Description |
|---|---|
cartpole |
Cartpole swing-up (simplest, good for testing) |
h1 |
Unitree H1 standing task |
jvrc_walk |
JVRC humanoid basic walking |
jvrc_step |
JVRC humanoid stepping with planned footsteps |
Hyperparameters (defaults)
| Flag | Default | Description |
|---|---|---|
--n-itr |
20000 | Training iterations |
--lr |
1e-4 | Learning rate |
--gamma |
0.99 | Discount factor |
--std-dev |
0.223 | Action noise |
--learn-std |
off | Learn action noise (flag) |
--entropy-coeff |
0.0 | Entropy regularization |
--clip |
0.2 | PPO clipping |
--minibatch-size |
64 | Minibatch size |
--epochs |
3 | Optimization epochs per update |
--num-procs |
12 | Parallel workers |
--num-envs-per-worker |
1 | Vectorized envs per worker |
--max-grad-norm |
0.05 | Gradient clipping |
--max-traj-len |
400 | Episode horizon |
--eval-freq |
100 | Eval every N iterations |
--seed |
None | Random seed |
--device |
auto | Training device (auto/cpu/cuda) |
--no-mirror |
off | Disable symmetry wrapper (flag) |
--recurrent |
off | Use LSTM policy (flag) |
--continued |
None | Path to pretrained weights |
Instructions
- Determine the environment name from the user's request. If ambiguous, ask.
- Use
--logdir /tmp/training_runsunless the user specifies a different path. - Only include flags that differ from defaults — keep the command clean.
- Show the user the full command you're about to run.
- Run the command in the background using
run_in_background: trueon the Bash tool. Set a generous timeout (600000ms). - After launching, tell the user the logdir path and how to check progress (you can tail the output using the task ID).
- If the user asks to check on training, use
TaskOutputwithblock: falseto check the latest output.
Cartpole-Specific Defaults
For cartpole, these settings are known to work well with the current defaults
(--lr 3e-4 --max-grad-norm 0.5 --lam 0.95 --gamma 0.99):
--minibatch-size 256--std-dev 0.15 --learn-std --entropy-coeff 0.01--max-traj-len 500 --n-itr 500 --num-procs 12--no-mirror(cartpole has no body symmetry)
Suggest these defaults when the user trains cartpole, but let them override.
Decide Fit First
Design Intent
How To Use It
Boundaries And Review