LilMoo-v0.1 is a decoder-transformer natively pretrained in Hindi. LilMoo is part of the Polygl0t initiative to advance language models for low-resource languages.

Details

Architecture: a Transformer-based model (llama)
Size: 670,127,616 parameters
Context length: 4096 tokens
Dataset(s): Polygl0t/gigalekh-v1
Language: Hindi
Batch size: 2,097,152 tokens
Number of steps: 179,590
GPU: 8 NVIDIA A100-SXM4-80GB
Training time: ~ 347 hours
Emissions: 538 KgCO2 (Germany)
Total energy consumption: 1414 kWh

This repository has the source code used to train this model. The full configuration used for training is available in this configuration file.

Learning Rate

This model was trained with a cosine learning rate schedule with a linear warmup over the first 1,000 steps to a peak learning rate of 7e-4, followed by a cosine decay to 7e-5. The minimum learning rate was reached and held constant for the last 10% of training. Checkpoints were saved every 5,000 steps, which equates to approximately 10 billion tokens.

The main branch of this repository contains the final checkpoint saved at step 179,590. All other checkpoints are available as separate branches. To load a specific checkpoint, you can use the following code snippet:

python

from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "Polygl0t/LilMoo-v0.1"
revision = "step_5000"  # Change this to the desired checkpoint branch
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, revision=revision)

Or, you can access all the revisions for the models via the following code snippet:

python

from huggingface_hub import list_repo_refs
out = list_repo_refs("Polygl0t/LilMoo-v0.1")
branches = [b.name for b in out.branches]
print(branches)

Learning Curves

Gradient Norm (L2)

Intended Uses

The primary intended use LilMoo is to serve as foundations for research and development involving native Hindi language modeling. Checkpoints saved during training are designed to provide a controlled setting for performing comparative experiments, specifically regarding the effects of active pretraining on the performance of currently available benchmarks. You may also fine-tune and adapt LilMoo for deployment if your use follows the Apache 2.0 license. If you decide to use LilMoo as a basis for your fine-tuned model, please conduct your own risk and bias assessment.

Out-of-scope Use

LilMoo is not intended for deployment. It is not an out-of-the-box product and should not be used for human-facing interactions.
LilMoo is for the Hindi language only and is unsuitable for text generation tasks in other languages.
LilMoo has not been fine-tuned for downstream tasks.

Basic usage

python

from transformers import GenerationConfig, TextGenerationPipeline, AutoTokenizer, AutoModelForCausalLM
import torch

# Specify the model and tokenizer
model_id = "Polygl0t/LilMoo-v0.1"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)

# Specify the generation parameters as you like
generation_config = GenerationConfig(
    **{
    "do_sample": True,
    "max_new_tokens": 150,
    "renormalize_logits": True,
    "repetition_penalty": 1.2,
    "temperature": 0.1,
    "top_k": 50,
    "top_p": 1.0,
    "use_cache": True, 
  }
)

device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
generator = TextGenerationPipeline(model=model, task="text-generation", tokenizer=tokenizer, device=device)

# Generate text
prompt = "भारत की राजधानी क्या है?"
completion = generator(prompt, generation_config=generation_config)
print(completion[0]['generated_text'])

Limitations

Like almost all other language models trained on large text datasets scraped from the web, the LilMoo shows behavior that does not make it an out-of-the-box solution to many real-world applications, especially those requiring factual, reliable, and nontoxic text generation. LilMoo is subject to the following:

Hallucinations: LilMoo can produce content that can be mistaken as true facts, but are misleading or entirely false, i.e., hallucination.
Biases and Toxicity: LilMoo inherits the social and historical stereotypes from the data used to train it. Given these biases, the model can produce toxic content, i.e., harmful, offensive, or detrimental to individuals, groups, or communities.
Language Limitations: LilMoo is primarily designed to interact with Hindi. Other languages might challenge its comprehension, leading to potential misinterpretations or errors in response.
Repetition and Verbosity: LilMoo may get stuck on repetition loops (especially if the repetition penalty during generations is set to a meager value) or produce verbose responses unrelated to the prompt it was given.

Hence, even though LilMoo is released with a permissive license, we urge users to perform their risk analysis on them if they intend to use them for real-world applications.

Evaluations

The table below compares our two versions of LilMoo against other base models of similar size. The NPM (Normalized Performance Metric) is a metric designed to provide a balanced view of model performance across various tasks, accounting for the inherent difficulty of each task. It normalizes the performance of each model on a given task by comparing it to a baseline performance, which represents a the performance of a random model.

Note: In the table and plots below, benchmark scores are reported as normalized accuracy, and NPM is calculated from accuracy.

	NPM (mean)	ARC	HellaSwag	MMLU	CSQA	MILU	Global PIQA
LilMoo-v0.2	9.94	30.47	37.18	28.21	36.52	29.06	63.00
LilMoo-v0.1	8.7	30.05	37.06	26.19	37.12	28.55	65.00
Qwen3-0.6B-Base	4.08	27.48	29.82	31.05	30.96	30.93	53.00
Qwen2.5-0.5B	1.92	23.71	28.63	27.78	30.01	28.48	57.00

🏆 HellaSwag

🏆 ARC

🏆 MMLU

🏆 MILU

🏆 CSQA

🏆 Global PIQA

Aggregate NPM Across Benchmarks

Performance vs Compute

This plot compares the compute requirements (measured as C = 6 * N * D, where N is the number of parameters and D is the number of tokens processed) against the performance of each model (measured by NPM mean). It highlights the trade-offs between model size, training data, and performance. LilMoo models are compared against two baseline models: Qwen2.5-0.5B and Qwen3-0.6B-Base, which are state-of-the-art multilingual models.

Text Generation Examples

All models were prompted with the same Hindi prompts, and the generations were produced with the default parameters specified in the generation_config.json of each model. All models were set to generate up to 250 new tokens. LilMoo models generally produce more text given their custom tokenizer and vocabulary, which is more adapted to Hindi.

Model	Prompt	Generated Completion
Qwen/Qwen2.5-0.5B	भारत की राजधानी क्या है?	भारत की राजधानी क्या है? \| चिन्होराना\nभारत की राजधानी क्या है?\nभारत की राजधानी क्या है?\nभारत की राजधानी क्या है? एक राजधानी को बीच करने से चमखान लेने की तुलना में अब अधिक जानकारी हैं। आज हम एक बार जानते हैं कि राजधानी इतिहास में जब राजनीति की संस्था पर खेत थी तो एक बच्चा ने भारत के विभिन्न क्षेत .......
	भारत का राष्ट्रीय पक्षी क्या है?	भारत का राष्ट्रीय पक्षी क्या है? विशेष राष्ट्रीय पक्षी नवीन पक्षी स्वयंच कर्मचारी राष्ट्रीय राज्य अमेरिकी राष्ट्रीय द्वितीय भारतीय संघ और अमेरिकी राष्ट्रीय यूरोपीय चित्र्या विकास विवरण करने के लिए राष्ट्रीय पक्षी के रूप में संघीय अनुभव है। पक्षी के रूप में विशेष राष्ट्रीय पक्षी नवीन पक्षी स्वयंच कर्मचार ....
Qwen/Qwen3-0.6B	भारत की राजधानी क्या है?	भारत की राजधानी क्या है? जारी रूप से जानने के लिए उपयोग करें राजमार्त बाजार बाजार के बाजार के बाजार के बाजार के बाजार के बाजार के बाजार के बाजार के बाजार के बाजार के बाजार के बाजार के बाजार के बाजार के बाजार के बाजार के बाजार के बाजार के बाजार के बाजार के बाजार के बाजार के बाजार के बाजार के बाजार के ...
	भारत का राष्ट्रीय पक्षी क्या है?	भारत का राष्ट्रीय पक्षी क्या है? जारी अवधारणा अंतर्गत बर्बल प्रतिबेद्य के निर्माण निर्माण नामक चित्र जातक एक साथ करें। आप के पास एक बहुत लोटा गया ज्ञान के बारे अपनी जातक रेल एक साथ जातक रेल के निर्माण के देखा जाएं। एक कार बर्बल प्रतिबेद्य के निर्माण अपनी जातक रेल के निर्माण के निर्माण नामक चित्र जातक के रेल ......
Polygl0t/LilMoo-v0.1	भारत की राजधानी क्या है?	'भारत की राजधानी क्या है?\nउत्तर – नई दिल्ली।'
	भारत का राष्ट्रीय पक्षी क्या है?	भारत का राष्ट्रीय पक्षी कौन सा है?\nउत्तर – मोर! ...

Cite as 🤗

latex

@misc{shiza2026lilmoo,
      title={{Raising Bars, Not Parameters: LilMoo Compact Language Model for Hindi}}, 
      author={Shiza Fatimah and Aniket Sen and Sophia Falk and Florian Mai and Lucie Flek and Nicholas Kluge Corr{\^e}a},
      year={2026},
      eprint={2603.03508},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2603.03508}, 
}

Aknowlegments

Polyglot is a project funded by the Federal Ministry of Education and Research (BMBF) and the Ministry of Culture and Science of the State of North Rhine-Westphalia (MWK) as part of TRA Sustainable Futures (University of Bonn) and the Excellence Strategy of the federal and state governments.

We also gratefully acknowledge the granted access to the Marvin cluster hosted by University of Bonn along with the support provided by its High Performance Computing & Analytics Lab.

License

LilMoo is licensed under the Apache License, Version 2.0. For more details, see the LICENSE file.