update README

README.md

@@ -1,40 +1,40 @@
 ---
 language:
-  - hi
   - en
+  - sw
 tags:
 - translation
 license: cc-by-4.0
 ---
 
-## HPLT MT release v1.0
+### HPLT MT release v1.0
 
-This repository contains the translation model for hi-en trained with OPUS and HPLT data. For usage instructions, evaluation scripts, and inference scripts, please refer to the [HPLT-MT-Models v1.0](https://github.com/hplt-project/HPLT-MT-Models/tree/main/v1.0) GitHub repository.
+This repository contains the translation model for en-sw trained with HPLT data only. For usage instructions, evaluation scripts, and inference scripts, please refer to the [HPLT-MT-Models v1.0](https://github.com/hplt-project/HPLT-MT-Models/tree/main/v1.0) GitHub repository.
 
 ### Model Info
 
-* Source language: hi
-* Target language: en
-* Dataset: OPUS and HPLT data
+* Source language: English
+* Target language: Swahili
+* Data: HPLT data only
 * Model architecture: Transformer-base
 * Tokenizer: SentencePiece (Unigram)
-* Cleaning: We used OpusCleaner with a set of basic rules. Details can be found in the filter files in [Github](https://github.com/hplt-project/HPLT-MT-Models/tree/main/v1.0/data/en-hi/raw/v2)
+* Cleaning: We used OpusCleaner with a set of basic rules. Details can be found in the filter files in [Github](https://github.com/hplt-project/HPLT-MT-Models/tree/main/v1.0/data/en-sw/raw/v0)
 
 You can also read our deliverable report [here](https://hplt-project.org/HPLT_D5_1___Translation_models_for_select_language_pairs.pdf) for more details.
 
 ### Usage
-
+**Note** that for quality considerations, we recommend using [HPLT/translate-en-sw-v1.0-hplt_opus](https://huggingface.co/HPLT/translate-en-sw-v1.0-hplt_opus) instead of this model.
 
 The model has been trained with Marian. To run inference, refer to the [Inference/Decoding/Translation](https://github.com/hplt-project/HPLT-MT-Models/tree/main/v1.0#inferencedecodingtranslation) section of our GitHub repository.
 
 The model can be used with the Hugging Face framework if the weights are converted to the Hugging Face format. We might provide this in the future; contributions are also welcome.
 
-### Benchmarks
+## Benchmarks
 
 | testset                                | BLEU | chrF++ | COMET22 |
 | -------------------------------------- | ---- | ----- | ----- |
-| flores200     | 38.2 | 62.1  | 0.8838  |
-| ntrex | 32.6   | 57.9  | 0.862  |
+| flores200     | 28.4 | 54.6  | 0.7743  |
+| ntrex | 30.5   | 55.2  | 0.7572  |
 
 ### Acknowledgements