Update README.md

README.md

@@ -30,32 +30,25 @@ base_model:
 
 ## TL;DR & Quick results
 
-Try it on [Space demo](https://huggingface.co/spaces/XSkills/nllb-turkmen-english) Article with full technical journey is available [Medium](). 
-
-### Test Results
-
-| Direction | BLEU ↑ | chrF ↑ | TER ↓ | Test pairs |
-|-----------|-------:|-------:|------:|-----------:|
-| **tk → en** | **26.07** | 52.97 | 68.39 | 50 |
-| **en → tk** | **8.13**  | 39.39 | 87.30 | 50 |
+Try it on [Space demo](https://huggingface.co/spaces/XSkills/nllb-turkmen-english) Article with full technical journey is available [Medium](https://medium.com/@meinnps/fine-tuning-nllb-200-with-lora-on-a-650-sentence-turkmen-english-corpus-082f68bdec71). 
 
 ### Model Comparison (Fine-tuned vs Original)
 
-### tk → en (Turkmen to English)
+#### English to Turkmen
 
-| Metric | Fine-tuned | Original |
-|--------|-----------|----------|
-| BLEU   | 26.07     | 26.48    |
-| chrF   | 52.97     | 52.91    |
-| TER    | 68.39     | 69.70    |
+| Metric                    | Fine-tuned | Original | Difference |
+|---------------------------|-----------:|---------:|-----------:|
+| **BLEU**                  |       8.24 |     8.12 |      +0.12 |
+| **chrF**                  |      39.55 |    39.46 |      +0.09 |
+| **TER (lower is better)** |      87.20 |    87.30 |      -0.10 |
 
-#### tk → en (Turkmen to English)
+#### Turkmen to English
 
-| Metric | Fine-tuned | Original |
-|--------|-----------|----------|
-| BLEU   | 26.07     | 26.48    |
-| chrF   | 52.97     | 52.91    |
-| TER    | 68.39     | 69.70    |
+| Metric                    | Fine-tuned | Original | Difference |
+|---------------------------|-----------:|---------:|-----------:|
+| **BLEU**                  |      25.88 |    26.48 |      -0.60 |
+| **chrF**                  |      52.71 |    52.91 |      -0.20 |
+| **TER (lower is better)** |      67.70 |    69.70 |      -2.00 |
 
 *Scores computed with sacre BLEU 2.5, chrF, TER on the official `test` split.  
 A separate spreadsheet with **human adequacy/fluency ratings** is available in the article.*
@@ -180,6 +173,35 @@ A manual review on 50 random test sentences showed:
 - Gender/ politeness nuances not guaranteed.
 - CC-BY-NC licence forbids commercial use; respect Meta’s original terms.
 
+## How to Contribute
+
+We welcome contributions to improve Turkmen-English translation capabilities! Here's how you can help:
+
+### Data Contributions
+- **Read Dataset Contribution**: You can find the instructions for contributing to the dataset at [Dataset Readme](https://huggingface.co/datasets/XSkills/turkmen_english_s500/blob/main/README.md)
+
+### Code Contributions
+- **Hyperparameter experiments**: Try different LoRA configurations and document your results
+- **Evaluation**: Help with human evaluation of translation quality and fluency
+- **Bug fixes**: Report issues or submit fixes for the model implementation
+
+### Use Cases & Documentation
+- **Example applications**: Share how you're using the model for research or projects
+- **Domain-specific guides**: Create guides for using the model in specific domains
+- **Translation examples**: Share interesting or challenging translation examples
+
+### Getting Started
+1. Fork the repository
+2. Make your changes
+3. Submit a pull request with clear documentation of your contribution
+4. For data contributions, contact the maintainer directly
+
+All contributors will be acknowledged in the model documentation. Contact [meinnps@gmail.com](mailto:meinnps@gmail.com) with any questions or to discuss potential contributions.
+
+---
+
+*Note: This model is licensed under CC-BY-NC-4.0, so all contributions must be compatible with non-commercial use only.*
+
 ## Citation
 ```bibtex
 @misc{durdyyev2025turkmenNLLBLoRA,