Downloads · 30 days
4
4% of all-time downloads
Debbyjaye001/MarketFit-Africa-Llama4Scout
MarketFit-Africa-Llama4Scout is a text generation model from Debbyjaye001. Use it when you need the model to write or continue text. It is set up for peft. The card lists the license as llama4.
MarketFit-Africa-Llama4Scout is a PEFT LoRA adapter built on Llama 4 Scout for evaluating and improving marketing communication tailored to African audiences.
Downloads · 30 days
4
4% of all-time downloads
All-time downloads
110
Public
Repo size
921 MB
Likes
0
Public
Click a slice to open those files.
.safetensors893 MB · 97%
From the Hugging Face model README
MarketFit-Africa-Llama4Scout is a PEFT LoRA adapter built on Llama 4 Scout for evaluating and improving marketing communication tailored to African audiences.
Rather than simply rewriting copy, the model analyzes marketing messages through the lens of trust, cultural context, audience alignment, platform behavior, conversion psychology, and decision friction before recommending strategic improvements.
It is designed for founders, marketers, agencies, startups, and businesses that want marketing communication with stronger local relevance and higher conversion potential.
This model was developed as part of the Adaption AutoScientist Challenge.
| Item | Value |
|---|---|
| Model Name | MarketFit-Africa-Llama4Scout |
| Base Model | togethercomputer/Llama-4-Scout-17B-16E-Instruct_bnb_4bit |
| Architecture | PEFT LoRA Adapter |
| Primary Language | English |
| Region | Africa (Nigeria & Ghana emphasis) |
| Developer | John Bolaji Deborah |
This model is designed to:
Evaluate the following marketing copy for an African audience.
"We are the #1 fintech solution guaranteed to change your life instantly."
Trust Score: Moderate
Credibility Issues:
- Uses exaggerated claims.
- "Guaranteed" reduces credibility.
Audience Alignment:
- Needs stronger localization.
Recommendations:
- Add proof or social validation.
- Use more realistic language.
- Focus on measurable customer benefits.
Improved Version:
"Join thousands of Africans using our fintech platform to simplify everyday payments with speed, security, and convenience."
The adapter was trained on a custom dataset of marketing evaluation examples.
Each training sample contains:
The dataset emphasizes:
| Metric | Value |
|---|---|
| Dataset Size | 815 prompt-completion pairs |
| Average Prompt Length | 30 words |
| Average Completion Length | 152.88 words |
| Adaptive Data Quality Improvement | 23.8% |
| Grade Improvement | B → A |
| Percentile Improvement | 16.7 → 57.7 |
| AutoScientist Win Rate | 74% |
| Marketing Category Win Rate | 76% |
Compared with the base model, the adapter demonstrates improvements in:
This model is intended only for marketing evaluation and communication improvement.
It should not be used for:
Human review is recommended before using outputs in production.
If you use this model, please cite:
MarketFit-Africa-Llama4Scout
Developed by John Bolaji Deborah
Adaption AutoScientist Challenge
Built using:
Special thanks to the Adaption Team for providing the Adaption platform used to train this model.