---
language:
- pcm
- en
tags:
- transformers
- encoder-decoder
- xlm-roberta
- afriberta
- pidgin
- nigerian-pidgin
- nlp
library_name: transformers
---

# Pidgin14 Encoder (AfriBERTa-based)

## Overview

This repository hosts the **encoder-side tokenizer** for `pidgin14`, an encoder-decoder sequence-to-sequence system for Nigerian Pidgin English ("Naija") built by [Ephraim](https://huggingface.co/Ephraimmm) at Analytics Intelligence.

`pidgin14` is composed of two halves published as separate repositories:
- **Encoder** (this repo) — based on AfriBERTa, reads source text and produces contextual representations.
- **Decoder** — [`Ephraimmm/pidgin14-decoder`](https://huggingface.co/Ephraimmm/pidgin14-decoder), based on GPT-2-medium, consumes the encoder's representations via cross-attention and generates output text.

The two halves are combined and trained together as a single `EncoderDecoderModel`, whose full weights are published at [`Ephraimmm/pidgin14`](https://huggingface.co/Ephraimmm/pidgin14). The architecture facts below are taken directly from that combined model's `config.json` (`encoder` sub-config), since this component repository itself contains only tokenizer files (`tokenizer.json`, `tokenizer_config.json`, `special_tokens_map.json`, `sentencepiece.bpe.model`) and not a standalone `config.json` or weight file.

## Architecture Details

From the `encoder` sub-configuration of the combined `Ephraimmm/pidgin14` model:

| Field | Value |
|---|---|
| Base model | `castorini/afriberta_small` |
| Model type | `xlm-roberta` (architecture class `XLMRobertaForMaskedLM`, used as the encoder half of an `EncoderDecoderModel`) |
| Hidden size | 768 |
| Hidden layers | 4 |
| Attention heads | 6 |
| Intermediate (feed-forward) size | 3072 |
| Max position embeddings | 514 |
| Hidden activation | GELU |
| Vocabulary size | 70,006 |

Tokenizer shipped in **this** repository:
- Tokenizer class: `XLMRobertaTokenizer`
- Underlying algorithm: SentencePiece **Unigram** model (`sentencepiece.bpe.model`)
- Vocabulary size: 70,006 tokens
- Special tokens: `<s>` (bos/cls), `<pad>`, `</s>` (eos/sep), `<unk>`, `<mask>`

## Training Details

- Fine-tuned from: `castorini/afriberta_small`, used as the encoder half of the `pidgin14` `EncoderDecoderModel`.
- Framework: Hugging Face `transformers` (the combined model's config records `transformers_version: 4.44.2`).
- Stored precision: `float32` (per the combined model's config).
- No `trainer_state.json`, training-step/epoch counts, optimizer settings, or training-dataset identifiers are published in this repository or in the combined `Ephraimmm/pidgin14` repository. These details are therefore omitted rather than estimated.

## Intended Use

- Encoding Nigerian Pidgin English and/or English text as the first stage of the `pidgin14` sequence-to-sequence pipeline (e.g. translation, paraphrasing, conversational response generation).
- Research and experimentation on low-resource West African language NLP.
- Must be paired with the [`pidgin14-decoder`](https://huggingface.co/Ephraimmm/pidgin14-decoder) tokenizer and the trained weights in [`Ephraimmm/pidgin14`](https://huggingface.co/Ephraimmm/pidgin14) to produce output.

## How to Use

```python
from transformers import AutoTokenizer, EncoderDecoderModel

# Tokenizers for each half of the system
encoder_tokenizer = AutoTokenizer.from_pretrained("Ephraimmm/pidgin14-encoder")
decoder_tokenizer = AutoTokenizer.from_pretrained("Ephraimmm/pidgin14-decoder")

# The trained combined encoder-decoder weights
model = EncoderDecoderModel.from_pretrained("Ephraimmm/pidgin14")

text = "How you dey?"
inputs = encoder_tokenizer(text, return_tensors="pt")

output_ids = model.generate(
    **inputs,
    decoder_start_token_id=decoder_tokenizer.bos_token_id,
    max_length=50,
)
print(decoder_tokenizer.decode(output_ids[0], skip_special_tokens=True))
```

## Limitations

- This repository provides the **tokenizer only** for the encoder half of `pidgin14`; it is not a usable standalone model and contains no weight file or `config.json` of its own.
- Must be paired with [`Ephraimmm/pidgin14-decoder`](https://huggingface.co/Ephraimmm/pidgin14-decoder) and the weights in [`Ephraimmm/pidgin14`](https://huggingface.co/Ephraimmm/pidgin14) to perform any task.
- Nigerian Pidgin English is a low-resource language with substantial dialectal and orthographic variation; a fixed AfriBERTa-derived vocabulary may not fully capture all spelling variants encountered in real usage.
- No evaluation metrics, benchmark results, or training-dataset documentation are published for this model. Outputs should be independently validated before any production use.
- License terms are not specified in the repository; users should contact the author before commercial reuse.

## Author

Developed by [Ephraimmm](https://huggingface.co/Ephraimmm)