`training-check`¶

Pack: ARIS skills

Category: drafting

Field: —

License: MIT

Updated: 2026-05-18

Stages: paper-drafting

↗ view SKILL.md on source · GitHub stars

Training Check¶

Periodically read WandB metrics during training to catch problems early. Do not wait until training finishes to discover it was a waste of GPU time.

Context: $ARGUMENTS¶

Constants¶

WANDB_ENTITY and WANDB_PROJECT: read from CLAUDE.md or passed as argument (format: entity/project/run_id)
CHECK_INTERVAL: starts at 10 minutes, then gradually increases if consistently healthy: 10 min → 20 min → 30 min → 60 min (cap)
REVIEWER_MODEL = gpt-5.5 — used via Codex MCP for ambiguous cases only

When to Use¶

After training is confirmed running (session alive, loss decreasing for first few steps)
Set up via CronCreate to fire periodically during training
This skill checks training QUALITY, not process HEALTH. Process health (session alive, GPU utilization) is watchdog.py's job.

Workflow¶

Step 1: Read WandB Metrics¶

Python

import wandb
api = wandb.Api()
run = api.run("<entity>/<project>/<run_id>")
history = run.history()

If WandB is unreachable (API error, network issue), fall back to reading the log file directly via SSH:

Bash

ssh server "tail -100 /path/to/training.log"

Check these signals: - Loss trend: Is training loss decreasing over the last N steps? - Eval metrics: Are evaluation metrics improving (or at least not degrading)? - NaN / Inf: Any NaN or Inf values in loss or gradients? - Spikes: Sudden large jumps in loss (>10x normal variance)? - Learning rate: Is the schedule behaving as expected? - Gradient norm: Exploding or vanishing?

Step 2: Judgment¶

Signal	Judgment	Action
NaN/Inf in loss	Clearly bad	Stop training, investigate
Loss diverging (increasing for >N steps)	Clearly bad	Stop training, investigate
Eval metrics significantly worse than baseline	Clearly bad	Stop training, investigate
Loss decreasing, metrics improving	Clearly fine	Continue, increase check interval
Loss flat but not diverging	Unsure	→ Step 3 (Codex judgment)
Metrics noisy, can't tell trend	Unsure	→ Step 3 (Codex judgment)
Slightly worse than baseline but still early	Unsure	→ Step 3 (Codex judgment)

Step 3: Codex Judgment (only when unsure)¶

Only escalate to Codex when the signal is ambiguous. For clearly good or clearly bad signals, act directly.

Text Only

mcp__codex__codex:
  config: {"model_reasoning_effort": "high"}
  prompt: |
    TRAINING HEALTH CHECK — need your judgment on ambiguous metrics.

    Run: <entity>/<project>/<run_id>
    Current epoch/step: X / Y total
    Training loss (last 10 checkpoints): [values]
    Eval metrics (last 3 evals): [values]
    Baseline reference: [numbers from paper/reproduction]

    What I'm unsure about: [specific concern]

    Please respond with exactly one of:
    - STOP: clearly problematic, should kill training
    - CONTINUE: looks fine, check again next interval
    - WAIT: not enough data to judge, check again sooner

Step 4: Act¶

Decision	Action
Stop	Kill the training session. Save the WandB run URL, key metrics, and reason for stopping. Log to project notes for debugging.
Continue	Do nothing. Will be invoked again at next interval (increase interval if consistently healthy).
Wait	Do nothing but keep the current short interval (don't increase).

Integration with Watchdog¶

Training-check and watchdog.py operate at different levels:

Layer	Tool	What it checks	Frequency
Process health	watchdog.py	Session alive? GPU active?	Every 60s (continuous)
Training quality	training-check	Loss trend? Metrics improving?	Every 10-60 min (periodic)

Use both together: - Watchdog catches crashes and idle GPUs immediately - Training-check catches subtle quality issues (loss plateau, metric degradation)

Rules¶

Do not stop training on first sign of noise — some loss spikes are normal. Look at trends over multiple checkpoints.
When stopping training, always save the WandB run URL and key metrics as evidence.
If both WandB and log files are unreachable, report the connectivity issue and try again next interval. Do not assume training is broken.
Gradually increase check interval when healthy (10 → 20 → 30 → 60 min). Reset to 10 min after any anomaly.
This skill is meant to be automated via CronCreate — do not ask the user whether to set it up. Just set it.

CronCreate Setup Example¶