| --- |
| license: apache-2.0 |
| library_name: transformers |
| datasets: |
| - kde4 |
| widget: |
| - text: Hi! How are you? |
| --- |
| |
| ## Model Summary |
|
|
| dataequity-kde4-en-de-qlora is a Transformer based language translator fine tuned using the kde dataset. The base model used is Helsinki-NLP/opus-mt-en-de |
|
|
| Our model hasn't been fine-tuned through reinforcement learning from human feedback. The intention behind crafting this open-source model is to provide the research community with a non-restricted small model to explore vital safety challenges, such as reducing toxicity, understanding societal biases, enhancing controllability, and more. |
|
|
|
|
| ### eng-spa |
|
|
| * source group: English |
| * target group: German |
|
|
| * model: transformer |
| * source language(s): en |
| * target language(s): de |
| * model: transformer |
|
|
| ### Inference Code: |
|
|
| ```python |
| from transformers import MarianMTModel, MarianTokenizer, |
| |
| hub_repo_name = 'dataequity/dataequity-kde4-en-de-qlora' |
| tokenizer = MarianTokenizer.from_pretrained(hub_repo_name) |
| finetuned_model = MarianMTModel.from_pretrained(hub_repo_name) |
| |
| questions = [ |
| "How are the first days of each season chosen?", |
| "Why are laws requiring identification for voting scrutinized by the media?", |
| "Why aren't there many new operating systems being created?" |
| ] |
| |
| translated = finetuned_model.generate(**tokenizer(questions, return_tensors="pt", padding=True)) |
| [tokenizer.decode(t, skip_special_tokens=True) for t in translated] |
| ``` |
