afriquellama_8b-lora-r64-eng-amh

afriquellama_8b-lora-r64-eng-amh

This is a LoRA adapter for the AfriScience-MT project, enabling efficient scientific machine translation for African languages.

Adapter Description

Why LoRA?

LoRA (Low-Rank Adaptation) enables efficient fine-tuning by training only a small number of additional parameters. This adapter adds only ~32.0M parameters to the base model while achieving strong translation performance.

Usage

Quick Start

from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig
from peft import PeftModel
import torch

# Configure 4-bit quantization (recommended for memory efficiency)
bnb_config = BitsAndBytesConfig(
    load_in_4bit=True,
    bnb_4bit_compute_dtype=torch.bfloat16,
    bnb_4bit_quant_type="nf4",
    bnb_4bit_use_double_quant=True,
)

# Load base model
base_model = AutoModelForCausalLM.from_pretrained(
    "afriquellama_8b",
    quantization_config=bnb_config,
    device_map="auto",
    torch_dtype=torch.bfloat16,
)
tokenizer = AutoTokenizer.from_pretrained("afriquellama_8b")

# Load LoRA adapter
adapter_name = "dsfsi/afriquellama_8b-lora-r64-eng-amh"
model = PeftModel.from_pretrained(base_model, adapter_name)
model.eval()

# Prepare translation prompt
source_text = "Climate change significantly impacts agricultural productivity in sub-Saharan Africa."
instruction = "Translate the following English scientific text to Amharic."

# Format prompt
prompt = f"""### Instruction:
{instruction}

### Input:
{source_text}

### Response:
"""

# Generate translation
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
with torch.no_grad():
    outputs = model.generate(
        **inputs,
        max_new_tokens=256,
        num_beams=5,
        early_stopping=True,
        pad_token_id=tokenizer.pad_token_id,
    )

# Decode only the generated part
generated = outputs[0][inputs["input_ids"].shape[1]:]
translation = tokenizer.decode(generated, skip_special_tokens=True)
print(translation)

Without Quantization (Full Precision)

# For GPUs with sufficient memory (>24GB for larger models)
base_model = AutoModelForCausalLM.from_pretrained(
    "afriquellama_8b",
    device_map="auto",
    torch_dtype=torch.bfloat16,
)
model = PeftModel.from_pretrained(base_model, "dsfsi/afriquellama_8b-lora-r64-eng-amh")

Hardware Requirements

Reproducibility

To reproduce this adapter:

# Clone the AfriScience-MT repository
git clone https://github.com/afriscience-mt/afriscience-mt.git
cd afriscience-mt

# Install dependencies
pip install -r requirements.txt

# Run LoRA training
python -m afriscience_mt.scripts.run_lora_training \
    --data_dir ./data \
    --source_lang eng \
    --target_lang amh \
    --model_name afriquellama_8b \
    --model_type llama \
    --lora_rank 64 \
    --output_dir ./output \
    --num_epochs 3 \
    --batch_size 4 \
    --load_in_4bit

Limitations

• Domain Specificity: Optimized for scientific/academic texts; may underperform on casual or colloquial language.
• Language Direction: Only supports English → Amharic translation.
• Base Model Required: Must be used with the afriquellama_8b base model.
• Context Length: Maximum context is model-dependent; longer texts should be chunked.

Citation

If you use this model, please cite the AfriScience-MT paper (arXiv:2605.29741):

@article{abdulmumin2026afriscience,
  title   = {AfriScience-MT: Towards Decolonizing Science in Africa through Text Translation},
  author  = {Abdulmumin, Idris and Gwadabe, Tajuddeen and Muhammad, Shamsuddeen Hassan and Adelani, David Ifeoluwa and Khalo, Nomonde and Ahmad, Ibrahim Said and Modupe, Abiodun and Mumm, Anina and Biyela, Sibusiso and Rabie, Michelle and Havemann, Johanna and Rei, Marek and Abbott, Jade and Marivate, Vukosi},
  journal = {arXiv preprint arXiv:2605.29741},
  year    = {2026},
  url     = {https://arxiv.org/abs/2605.29741}
}

License

This adapter is released under the Apache 2.0 License.

Acknowledgments

• Base model: afriquellama_8b
• LoRA implementation: PEFT
• Evaluation: SSA-COMET for African language assessment