AltSlate Labs

Efficient, honest AI — small models trained from scratch, with the full recipe published.

We build compact language models that punch above their weight, and we release everything needed to trust and reproduce them: code, weights, evaluations, and reports.

🪰 The Jugnu family

Tiny language models pretrained from scratch (jugnu — जुगनू — means "firefly": small, but it glows).

• JugnuLM-53M — a 53.5M-parameter base model. BLiMP 78.1% · ARC-Easy 51.4% · WikiText-2 byte-perplexity 2.04, competitive with models 2–3× larger. Trained on 4× Blackwell GPUs in under a day. Training code: github.com/AltSlate-Labs/jugnu

What we care about

• Efficiency per parameter — the sub-150M regime: on-device, low-latency, cheap-to-serve models.
• Reproducibility — public code, open weights, and one-command evals.
• Honest reporting — clean baselines and measured ablations over unvalidated clever tricks.

Links

• 💻 Code — https://github.com/AltSlate-Labs
• 🌐 Web — https://altslate.com
• ✉️ hello@altslate.com