Skip to main content
Glama

rl_seed_sweep

Launch a training script across multiple seeds as separate runs, optionally on individual Hugging Face GPU jobs for parallel experiments.

Instructions

Launch the same training script across several seeds as separate runs.

seeds: comma-separated, e.g. "0,1,2". Each becomes a run '<run_name>_s' with GPU_BRIDGE_SEED set (read it as SEED in your script). Poll each run by name. hf_flavor: leave empty to run all seeds on the connected runtime. Set a Hugging Face machine type (e.g. "t4-small") to give every seed its own rented job, so they train in parallel; each is billed separately, up to hf_timeout.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "script": "<script>",
      "run_name": "<run_name>",
      "seeds": "<seeds>"
    },
    "name": "rl_seed_sweep"
  }
}

Input Schema

NameRequiredDescriptionDefault
seedsYes
scriptYes
run_nameYes
hf_flavorNo
hf_timeoutNo1h
requirementsNo

Output Schema

NameRequiredDescriptionDefault
resultYes

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv0.4.0

TDQS

A4.1/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the behavioral burden and does well: it discloses the generated run naming convention '<run_name>_s<seed>', the GPU_BRIDGE_SEED environment variable exposed as SEED, that runs must be polled individually by name, per-seed billing, and the hf_timeout cap. It omits failure behavior, partial-launch semantics, and what happens to seeds whose jobs time out.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Front-loaded with the one-line purpose, then per-parameter notes in a compact labeled style; every sentence adds information. The indented continuation lines for hf_flavor are slightly awkward but do not obscure meaning.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

An output schema exists, so return values need not be described, and the key operational facts (naming, seeding, parallelism, billing, timeout) are present for a six-parameter launch tool with no annotations. The remaining gap is the undocumented requirements and script parameters, which an agent must infer from the schema alone.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 0%, so the description must compensate and only partly does. It fully explains seeds (comma-separated with example), hf_flavor (empty vs. machine type, with billing consequence), and touches hf_timeout as a billing bound, but leaves script, requirements, and the run_name prefix convention without explicit semantics.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource: 'Launch the same training script across several seeds as separate runs.' The scope (parallel seed runs) and the artifact produced (one run per seed) are unambiguous, letting an agent distinguish it from single-run launch siblings such as hf_job_launch.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives clear guidance for the decision that matters most: leave hf_flavor empty to run all seeds on the connected runtime, or set a machine type to rent one job per seed for parallel training at separate billing. It doesn't explicitly exclude or route to sibling launch tools, so it stops short of the top score.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.