run-experiment skill

Run Experiment

Deploy and run ML experiment: $ARGUMENTS

Workflow

Step 1: Detect Environment

Read the project's CLAUDE.md to determine the experiment environment:

• Local GPU (gpu: local): Look for local CUDA/MPS setup info
• Remote server (gpu: remote): Look for SSH alias, conda env, code directory
• Vast.ai (gpu: vast): Check for vast-instances.json at project root — if a running instance exists, use it. Also check CLAUDE.md for a ## Vast.ai section.
• Modal (gpu: modal): Serverless GPU via Modal. No SSH, no Docker, auto scale-to-zero. Delegate to /serverless-modal.

Modal detection: If CLAUDE.md has gpu: modal or a ## Modal section, the entire deployment is handled by /serverless-modal. Jump to Step 4: Deploy (Modal) — Steps 2-3 are not needed (Modal handles code sync and GPU allocation automatically).

Environment contract (../shared-references/compute-env-contract.md): before building or trusting any environment, read the provider's env ledger (.aris/compute/.md) — an unchanged spec hash means warm-reuse, a changed one means rebuild. New env → write the declarative spec first, render it for this provider's shape, and never declare it ready on import-success alone: run the seeded kernel witness, and after any rebuild/doc edit run the agent-follows-doc pass (a fresh subagent executes the documented invocation verbatim and reports doc-vs-reality divergence).

Vast.ai detection priority:

1. If CLAUDE.md has gpu: vast or a ## Vast.ai section:

• If vast-instances.json exists and has a running instance → use that instance
• If no running instance → call /vast-gpu provision which analyzes the task, presents cost-optimized GPU options, and rents the user's choice

1. If no server info is found in CLAUDE.md, ask the user.

Step 2: Pre-flight Check

Check GPU availability on the target machine:

Remote (SSH):

ssh <server> nvidia-smi --query-gpu=index,memory.used,memory.total --format=csv,noheader

Remote (Vast.ai):

ssh -p <PORT> root@<HOST> nvidia-smi --query-gpu=index,memory.used,memory.total --format=csv,noheader

(Read sshhost and sshport from vast-instances.json, or run vastai ssh-url which returns ssh://root@HOST:PORT)

Local:

nvidia-smi --query-gpu=index,memory.used,memory.total --format=csv,noheader
# or for Mac MPS:
python -c "import torch; print('MPS available:', torch.backends.mps.is_available())"

Free GPU = memory.used < 500 MiB.

Step 3: Sync Code (Remote Only)

Check the project's CLAUDE.md for a code_sync setting. If not specified, default to rsync.

Option A: rsync (default)

Only sync necessary files — NOT data, checkpoints, or large files:

rsync -avz --include='*.py' --exclude='*' <local_src>/ <server>:<remote_dst>/

Option B: git (when code_sync: git is set in CLAUDE.md)

Push local changes to remote repo, then pull on the server:

# 1. Push from local
git add -A && git commit -m "sync: experiment deployment" && git push

# 2. Pull on server
ssh <server> "cd <remote_dst> && git pull"

Benefits: version-tracked, multi-server sync with one push, no rsync include/exclude rules needed.

Option C: Vast.ai instance

Sync code to the vast.ai instance (always rsync, code dir is /workspace/project/):

rsync -avz -e "ssh -p <PORT>" \
  --include='*.py' --include='*.yaml' --include='*.yml' --include='*.json' \
  --include='*.txt' --include='*.sh' --include='*/' \
  --exclude='*.pt' --exclude='*.pth' --exclude='*.ckpt' \
  --exclude='__pycache__' --exclude='.git' --exclude='data/' \
  --exclude='wandb/' --exclude='outputs/' \
  ./ root@<HOST>:/workspace/project/

Install dependencies per the env contract (ordered phases — pins first, one pip install per phase; see ../shared-references/compute-env-contract.md):

ssh -p <PORT> root@<HOST> "pip install -q torch==<pinned>"       # phase 1: pins
ssh -p <PORT> root@<HOST> "pip install -q <remaining packages>"  # phase 2+

Legacy fallback — requirements.txt only, no env spec: install as one phase, and treat any version fight as the signal to convert to ordered phases:

scp -P <PORT> requirements.txt root@<HOST>:/workspace/
ssh -p <PORT> root@<HOST> "pip install -q -r /workspace/requirements.txt"

Step 3.5: W&B Integration (when wandb: true in CLAUDE.md)

Skip this step entirely if wandb is not set or is false in CLAUDE.md.

Before deploying, ensure the experiment scripts have W&B logging:

1. Check if wandb is already in the script — look for import wandb or wandb.init. If present, skip to Step 4.

1. If not present, add W&B logging to the training script:

import wandb
   wandb.init(project=WANDB_PROJECT, name=EXP_NAME, config={...hyperparams...})

   # Inside training loop:
   wandb.log({"train/loss": loss, "train/lr": lr, "step": step})

   # After eval:
   wandb.log({"eval/loss": eval_loss, "eval/ppl": ppl, "eval/accuracy": acc})

   # At end:
   wandb.finish()

1. Metrics to log (add whichever apply to the experiment):

• train/loss — training loss per step
• train/lr — learning rate
• eval/loss, eval/ppl, eval/accuracy — eval metrics per epoch
• gpu/memoryused — GPU memory (via torch.cuda.maxmemory_allocated())
• speed/samplespersec — throughput
• Any custom metrics the experiment already computes

1. Verify wandb login on the target machine:

ssh <server> "wandb status"  # should show logged in
   # If not logged in:
   ssh <server> "wandb login <WANDB_API_KEY>"

The W&B project name and API key come from CLAUDE.md (see example below). The experiment name is auto-generated from the script name + timestamp.

Step 4: Deploy

Remote (via SSH + screen)

For each experiment, create a dedicated screen session with GPU binding:

ssh <server> "screen -dmS <exp_name> bash -c '\
  eval \"\$(<conda_path>/conda shell.bash hook)\" && \
  conda activate <env> && \
  CUDA_VISIBLE_DEVICES=<gpu_id> python <script> <args> 2>&1 | tee <log_file>'"

Vast.ai instance

No conda needed — the Docker image has the environment. Use /workspace/project/ as working dir:

ssh -p <PORT> root@<HOST> "screen -dmS <exp_name> bash -c '\
  cd /workspace/project && \
  CUDA_VISIBLE_DEVICES=<gpu_id> python <script> <args> 2>&1 | tee /workspace/<log_file>'"

After launching, update the experiment field in vast-instances.json for this instance.

Modal (serverless)

When gpu: modal is detected, delegate to /serverless-modal:

1. Analyze task — determine VRAM needs, choose GPU, estimate cost
2. Generate launcher — create a modallauncher.py that wraps the training script using modal.Mount.fromlocal_dir for code and modal.Volume for results
3. Run — modal run modal_launcher.py (runs locally, GPU executes remotely)
4. Collect results — results return via Volume or stdout, no manual download needed