You can deploy this model and start earning money today!

Mikile/
Bertha-translation

Name: Bertha-translation
Availability: InStock

Views

No views yet

Bertha ↔ English Translator

Bertha-Translation is a bilingual Seq2Seq neural machine translation model that translates text between English and Bertha.
It uses an Encoder–Decoder GRU architecture with attention.

Features

Translate English → Bertha
Translate Bertha → English
Bidirectional Seq2Seq model with attention
Built with PyTorch following the official Seq2Seq tutorial
Lightweight and suitable for low-latency translation

Model Details

Author: Mikiyas Zenebe
Model type: Seq2Seq (Encoder + Decoder with Attention)
Framework: PyTorch
Source / Target languages: English, Bertha
License: Apache 2.0

Usage Example

python

import torch
from huggingface_hub import hf_hub_download
from model import EncoderRNN, DecoderRNN, tensorFromSentence

# Download trained models
encoder_path = hf_hub_download("Mikile/Bertha-translation", "encoder.pth")
decoder_path = hf_hub_download("Mikile/Bertha-translation", "decoder.pth")

# Load your encoder and decoder
encoder = EncoderRNN(input_size, hidden_size)
decoder = DecoderRNN(hidden_size, output_size)
encoder.load_state_dict(torch.load(encoder_path))
decoder.load_state_dict(torch.load(decoder_path))

# Translate a sentence
sentence = "hello my friend"
input_tensor = tensorFromSentence(input_lang, sentence)
encoder_outputs, encoder_hidden = encoder(input_tensor)
decoder_outputs, _, _ = decoder(encoder_outputs, encoder_hidden)


@misc{bertha_translation,
  author = {Mikiyas Zenebe},
  title = {Bertha ↔ English Translation Model},
  year = {2025},
  howpublished = {Hugging Face Model Hub},
  url = {https://huggingface.co/Mikile/Bertha-translation}
}