skillrank_skill ▸ benchflow-ai/nanogpt-trainingMIT · open registry

skill detail

← registry

Nanogpt Training

benchflow-ai/nanogpt-training

Train GPT-2 scale models (~124M parameters) efficiently on a single GPU. Covers GPT-124M architecture, tokenized dataset loading (e.g., HuggingFace Hub shards), modern optimizers (Muon, AdamW), mixed precision training, and training loop implementation.

communityprovisionaldatatraingptscale

skillrank score

███████░░░░░░░░░43

SkillRank score blends community stars, real usage, and our eval lift; provisional until a skill is evaluated -- so popularity alone can't reach the top tier.

source

★ 1.7k

benchflow-ai/skillsbench

tasks-extra/mhc-layer-impl/environment/skills/nanogpt-training

open on GitHub ▸

eval status

eval pending

Success delta, token delta, and trial count are not available yet. No eval number is shown until this skill has measured results.

install

$ skillrank install benchflow-ai/nanogpt-training
$ curl -fsSL skillrank.dev | sh