Python · MIT
FareedKhan-dev/train-llm-from-scratch
A straightforward method for training your LLM, from downloading data to generating text.
★ 13.1k
Stars
1.8k forks · 2 open issues
+36/day
Growth · measured
Stars counted between two of our own snapshots roughly a week apart. The repository page gives the exact span.
54.1
Momentum / 100
see the breakdown below
22mo ago
Created
last push 36d ago
Star history
How this score was built
| Component | Weight | Earned | Why |
|---|---|---|---|
| Velocity | 55 | 30.0 | +36/day, log-scaled against a 750/day ceiling |
| Acceleration | 15 | 0.0 | last week ran 0.85× the monthly pace |
| Freshness | 15 | 12.2 | last push 36d ago; decays to zero at 180 days |
| Scale | 15 | 11.9 | 13.1k stars, log-scaled against 150k — stops a 3-star repo topping the board |
| Momentum | 100 | 54.1 | sum of the above |
Detail
| Topics | geminilarge-language-modelsllmopenaitrainingtransformers |
| Matched on | LLM & Inference (large-language-models) |
| Short window | +252 stars over 7 days = 36.0/day |
| Long window | +1.1k stars over 26 days = 42.2/day |
| Lifetime average | 19.84 stars/day over 662 days |
| First seen here | 2026-10-11 |
| Last read | 2d ago |