We train the models, not just the prompts.
Most teams consume models. We build them. Our Jugnu family is 5 compact language models pretrained from scratch on 4× Blackwell GPUs — and the same discipline drives the custom models, fine-tunes, and distillations we deliver for clients.
Pretraining from scratch
We build models from the tokenizer up — architecture, data pipeline, and training loop — not just prompt wrappers.
Fine-tuning & distillation
Adapt frontier or compact models to your domain, then distill into something small enough to run cheaply in production.
Evaluation & benchmarking
Reproducible harnesses, honest baselines, and the eval discipline to know a model is actually better before it ships.
The Jugnu research family
An open, reproducible line of sub-150M models exploring how much capability fits into a small budget. Our 110M flagship reaches 81.25% on BLiMP — matching a larger baseline with fewer parameters — and our 53M ranks #3 on the sub-150M Tiny-ML Leaderboard.
These are base research models — Tiny-ML leaderboard entries under a permissive license, not instruction-tuned production assistants. Client systems use whichever model best meets their quality, cost, latency, and privacy needs.
JugnuLM-110M
Researchmatches a 125M baseline at ~12% fewer parameters
Deep-thin Qwen3 decoder with QK-Norm, grouped-query attention, RoPE, and tied embeddings. The family flagship.
- Params
- 109.7M
- Tokens
- ~8B
- Context
- 2,048
- Type
- Base (R0)
- Data
- FineWeb-Edu (sample-10BT)
JugnuLM-110M-R1
ResearchAdds value residuals to the base recipe — testing whether richer residual pathways lift sample efficiency.
- Params
- 109.7M
- Tokens
- ~8B
- Context
- 2,048
- Type
- Base + value residuals
- Data
- FineWeb-Edu (sample-10BT)
JugnuLM-110M-R2
ResearchSwaps the optimizer for Muon — an ablation on how second-order-style updates change small-model training dynamics.
- Params
- 109.7M
- Tokens
- ~8B
- Context
- 2,048
- Type
- Base + Muon optimizer
- Data
- FineWeb-Edu (sample-10BT)
JugnuLM-110M-R3
ResearchRe-weights the training data mixture — isolating how corpus composition alone moves downstream quality.
- Params
- 109.7M
- Tokens
- ~8B
- Context
- 2,048
- Type
- Base + data blend
- Data
- Blended corpus
JugnuLM-53M
ResearchOur sub-150M leaderboard entry — how much capability fits into 53M parameters trained from scratch.
- Params
- 53.5M
- Tokens
- ~12B
- Context
- 2,048
- Type
- Base
- Data
- FineWeb-Edu (sample-10BT)
Need a model tuned to your domain?
From fine-tuning a frontier model to shipping a distilled one on your own hardware.