README.md

---
language: ne
license: other
task_categories:
  - text-generation
tags:
  - nepali
  - devanagari
  - pretraining
  - llm
  - text-corpus
  - neBrahma
size_categories:
  - 10M<n<100M
---

# neBrahma Nepali Pretrain Corpus P2b

## Dataset Summary

The **neBrahma Nepali Pretrain Corpus P2b** is a large-scale, production-grade Nepali text
corpus assembled and certified for language model pretraining.
It contains **20,321,968 documents** and **1.845 billion tokens** of clean, verified
Devanagari Nepali text, drawn from four diverse sources and processed through an
eight-stage cleaning and quality pipeline.

This corpus serves as the training data for
[neBrahma-llm](https://github.com/ToniBirat7/neBrahma-llm) - a Nepali multimodal LLM
(text + ASR + TTS) built from scratch in C++/CUDA.
The corpus is certified **FIT TO TRAIN** with zero warnings and a 300-document human
spot-check confirming linguistic quality.

| Source | Raw Docs | Kept Docs | Keep Rate | License |
|--------|----------|-----------|-----------|---------|
| IndicCorp v2 | 21,836,879 | 17,147,383 | 78.52% | CC0 |
| FineWeb-2 | 2,381,855 | 1,583,002 | 66.46% | ODC-By |
| IRIIS Nepali | 1,878,927 | 1,563,850 | 83.23% | Research use |
| Sagarmatha ASR text | 50,790 | 27,733 | 54.60% | Author's own |
| **Total** | **26,148,451** | **20,321,968** | **77.72%** | - |

---

## EDA Visualizations

### Pipeline Overview

The corpus was constructed through an eight-stage pipeline:

1. **Fetch** - stream raw documents from source APIs (HuggingFace datasets, IRIIS, ASR text)
2. **Clean** - Unicode NFC normalization, Devanagari font conversion (Preeti/Kantipur → Unicode),
   script-ratio pre-filter
3. **Exact dedup** - SHA-1 hash deduplication within each source
4. **Language ID** - GlotLID filtering: keep `npi_Deva` (Nepali Devanagari) with score >= 0.6;
   line-level filtering for mixed-language documents
5. **Cross-source exact dedup** - SHA-1 deduplication across all sources (merged)
6. **Near-dedup** - MinHash LSH at Jaccard >= 0.8; removes near-duplicate paragraphs
7. **Quality annotation + filter** - 25+ signal gates: word count, Devanagari ratio, digit ratio,
   Latin ratio, KenLM perplexity, n-gram repetition, stopword ratio, and more
8. **Split + pack** - deterministic hash-keyed train/val/test split, then SentencePiece BPE
   tokenization into packed uint16 shards for training

![Pipeline Funnel](https://huggingface.co/datasets/tonibirat/neBrahma-Nepali-Pretrain-Corpus/resolve/main/images/pipeline_funnel.png)

### Source Composition and Keep Rates

![Source Keep Rates](https://huggingface.co/datasets/tonibirat/neBrahma-Nepali-Pretrain-Corpus/resolve/main/images/source_keep_rates.png)

### Quality Filter Gate Analysis

The chart below shows the top rejection gates (gates overlap - a document may fail multiple).

![Quality Gates](https://huggingface.co/datasets/tonibirat/neBrahma-Nepali-Pretrain-Corpus/resolve/main/images/quality_gates.png)

### Document Length Distribution

Distribution of document lengths (in characters) across the quality-filtered corpus
(sampled from 200K documents).

![Document Length Distribution](https://huggingface.co/datasets/tonibirat/neBrahma-Nepali-Pretrain-Corpus/resolve/main/images/doc_length_dist.png)

### Devanagari Ratio Distribution

All retained documents have Devanagari ratio = 1.0 (pure script, no Latin contamination).

![Devanagari Ratio](https://huggingface.co/datasets/tonibirat/neBrahma-Nepali-Pretrain-Corpus/resolve/main/images/devanagari_ratio_dist.png)

### Character Category Breakdown

![Character Categories](https://huggingface.co/datasets/tonibirat/neBrahma-Nepali-Pretrain-Corpus/resolve/main/images/char_categories.png)

### Top-100 Nepali Words by Frequency

![Top 100 Nepali Words](https://huggingface.co/datasets/tonibirat/neBrahma-Nepali-Pretrain-Corpus/resolve/main/images/word_freq_top100.png)

### Token Zipf Distribution

Vocabulary frequency distribution follows the expected Zipfian curve.
32,000-token BPE vocabulary with 97.16% coverage (31,091 of 32,000 tokens seen in corpus).

![Token Zipf Distribution](https://huggingface.co/datasets/tonibirat/neBrahma-Nepali-Pretrain-Corpus/resolve/main/images/token_zipf.png)

---

## Dataset Structure

### Data Fields

| Field | Type | Description |
|-------|------|-------------|
| `id` | string | Unique document identifier (format: `{source}:{original_id}`) |
| `text` | string | Cleaned, NFC-normalized Nepali text in Devanagari script |
| `source` | string | Data source: `indiccorp`, `fineweb2`, `iriis`, or `sagarmatha` |

### Example Record

```json
{
  "id": "indiccorp:npi_Deva_2847193",
  "text": "नेपाल सरकारले नयाँ शिक्षा नीति लागू गर्ने घोषणा गरेको छ। यो नीतिले प्राथमिक विद्यालयदेखि विश्वविद्यालय तहसम्मका विद्यार्थीहरूलाई प्रभाव पार्नेछ।",
  "source": "indiccorp"
}
```

---

## Statistics

### Corpus-Level

| Metric | Value |
|--------|-------|
| Total documents | 20,321,968 |
| Raw input documents | 26,148,451 |
| Overall keep rate | 77.72% |
| Total tokens (BPE 32k) | 1,845,781,044 |
| Training tokens | 1,827,236,028 |
| Validation tokens | 9,320,924 |
| Test tokens | 9,224,092 |
| Total shards | 21 (19 train / 1 val / 1 test) |
| Vocab size | 32,000 (SentencePiece BPE) |
| Unique token types seen | 31,091 / 32,000 (97.16%) |
| Mean Devanagari ratio | 1.00 (pure script) |
| Halant density | 7.87% of Devanagari chars |
| Chandrabindu per 1K Devanagari | 4.92 |
| Anusvara per 1K Devanagari | 4.78 |
| Font conversion applied | 0% (all native Unicode) |

### Per-Source Document Counts

| Source | Kept Docs | % of Total | Est. Tokens |
|--------|-----------|------------|-------------|
| IndicCorp v2 | 17,147,383 | 84.4% | ~1.556B |
| FineWeb-2 | 1,583,002 | 7.8% | ~144M |
| IRIIS | 1,563,850 | 7.7% | ~142M |
| Sagarmatha ASR | 27,733 | 0.1% | ~2.5M |

### Quality Certification

```
Verdict   : FIT TO TRAIN
hard_ok   : true
warnings  : [] (zero)
spot_check: 300 documents - all pass
binary_checks: PASS (SHA-256 integrity, shard contract)
reversible: true (kept ∪ rejected == raw input)
```

---

## Usage

```python
from datasets import load_dataset

# Load the full corpus (~8GB parquet download)
ds = load_dataset("tonibirat/neBrahma-Nepali-Pretrain-Corpus")
print(ds)
# DatasetDict({'train': Dataset({features: ['id', 'text', 'source'], num_rows: 20321968})})

# Filter by source
indiccorp_only = ds["train"].filter(lambda x: x["source"] == "indiccorp")

# Iterate (streaming for large-scale use)
ds_stream = load_dataset("tonibirat/neBrahma-Nepali-Pretrain-Corpus", streaming=True)
for example in ds_stream["train"]:
    text = example["text"]   # Nepali Devanagari text
    src  = example["source"] # "indiccorp" | "fineweb2" | "iriis" | "sagarmatha"
    break
```

---

## Limitations

- **IRIIS crowd-source noise**: The IRIIS subset relies on crowd-sourced transcriptions which
  may contain latent inaccuracies despite passing the quality filter.
- **No demographic metadata**: No speaker, regional dialect, or domain metadata is included.
  Source provenance is available via the `source` field.
- **Sagarmatha ASR fraction is small**: The Sagarmatha ASR text subset (27,733 docs, 0.1%)
  consists of short utterance-level transcriptions and has a lower keep rate (54.6%) due to
  the minimal-sentence nature of ASR transcripts.
- **EDA sample**: Document-level EDA visualizations (length distribution, character breakdown,
  word frequency) were computed on a 200K-document sample, not the full 20.3M.
  Shard-level token statistics are exact (computed from all 21 shards).
- **Per-source license**: IndicCorp v2 is CC0; FineWeb-2 is ODC-By (attribution required);
  IRIIS is distributed for research use. Verify compliance with your use case.

---

## Citation

If you use this dataset, please cite:

```bibtex
@misc{neBrahma2026,
  title        = {neBrahma-llm: A Nepali Multimodal Language Model Trained from Scratch},
  author       = {Gautam, Toni Birat},
  year         = {2026},
  institution  = {Final Year Project},
  howpublished = {\url{https://github.com/ToniBirat7/neBrahma-llm}},
  note         = {Corpus: neBrahma Nepali Pretrain Corpus P2b --
                  20.3M docs / 1.845B tokens / 4 sources / FIT TO TRAIN}
}
```