File size: 4,104 Bytes
080af4f | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 | ---
base_model: Qwen/Qwen2.5-0.5B-Instruct
library_name: peft
pipeline_tag: text-generation
language:
- en
tags:
- lora
- peft
- nigeria
- nigerian-english
- nigerian-pidgin
- customer-service
- scam-safety
- business-writing
license: apache-2.0
---
# GaiaLab Naija Assistant v0.5
GaiaLab Naija Assistant v0.5 is an experimental LoRA adapter for `Qwen/Qwen2.5-0.5B-Instruct`.
This release adds a reproducible dataset workflow for CSV ingestion, JSONL generation, validation, duplicate checking, statistics, and CPU-compatible LoRA training.
## Model Details
| Field | Value |
|---|---|
| Version | v0.5 |
| Base model | `Qwen/Qwen2.5-0.5B-Instruct` |
| Fine-tuning method | LoRA / PEFT |
| Model type | Causal language model adapter |
| Training examples | 47 |
| Dataset health score | 95/100 |
| Developer | Oluwafemi Idiakhoa |
| Project | GaiaLab AI |
## Training Categories
| Category | Examples |
|---|---:|
| Safety and scams | 13 |
| Professional boundaries | 12 |
| Customer service | 10 |
| Nigerian English | 10 |
| Business writing | 1 |
| Nigerian Pidgin | 1 |
| **Total** | **47** |
## Risk-Level Distribution
| Risk level | Examples |
|---|---:|
| High | 19 |
| Medium | 7 |
| Low | 21 |
## Dataset Validation
The v0.5 pipeline reported:
- Valid JSONL
- Required fields present
- Correct system, user, and assistant role order
- Zero duplicate IDs
- Zero duplicate prompts
- Zero missing prompts
- Zero missing responses
- Dataset health score of 95/100
## Evaluation Status
A formal side-by-side benchmark comparing v0.4 and v0.5 has not yet been published. This model card does not claim that v0.5 outperforms v0.4.
## Intended Uses
- Research and education
- Nigerian customer-service prototypes
- Professional message drafting
- Scam-awareness demonstrations
- Nigerian English experimentation
- Basic Nigerian Pidgin experimentation
- CPU-friendly LoRA research
## Limitations
- The training dataset contains only 47 examples
- Business writing and Pidgin each contain only one example
- The dataset is unevenly distributed
- The adapter may overfit specific wording
- Cultural coverage is narrow
- The model may hallucinate
- Human review is required for important outputs
## Installation
```bash
pip install torch transformers peft
```
## Usage
```python
import torch
from peft import PeftModel
from transformers import AutoModelForCausalLM, AutoTokenizer
base_model_id = "Qwen/Qwen2.5-0.5B-Instruct"
adapter_id = "mgbam/gaialab-naija-adapter-v0.5"
tokenizer = AutoTokenizer.from_pretrained(
base_model_id,
trust_remote_code=True,
)
base_model = AutoModelForCausalLM.from_pretrained(
base_model_id,
torch_dtype=torch.float32,
trust_remote_code=True,
)
model = PeftModel.from_pretrained(base_model, adapter_id)
messages = [
{
"role": "system",
"content": (
"You are GaiaLab Naija Assistant. Be helpful, concise, "
"culturally aware, truthful, and safe."
),
},
{
"role": "user",
"content": "Write a polite payment reminder for a customer.",
},
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
)
inputs = tokenizer(text, return_tensors="pt")
with torch.no_grad():
output = model.generate(
**inputs,
max_new_tokens=120,
do_sample=False,
)
new_tokens = output[0][inputs["input_ids"].shape[1]:]
print(tokenizer.decode(new_tokens, skip_special_tokens=True))
```
## Responsible Use
Do not use this model as the sole authority for medical, legal, financial, emergency, employment, identity-verification, or other high-impact decisions.
Never provide passwords, PINs, one-time passwords, bank verification codes, private keys, or other sensitive credentials to the model.
## Author
Developed by **Oluwafemi Idiakhoa** under the **GaiaLab AI** initiative.
## Project Links
- GitHub: https://github.com/oluwafemidiakhoa/gaialab-naija-assistant
- Model: https://huggingface.co/mgbam/gaialab-naija-adapter-v0.5
- GaiaLab AI: https://www.gailabai.com
|