Model Overview

baban/QwenTranslate_English_Telugu is a specialized machine translation model, fine-tuned from the Qwen/Qwen2.5-3B-Instruct base model. With 3.1 billion parameters, it focuses on translating text from English to Telugu. The model was trained using a learning rate of 5e-05 over 3 epochs, with a total training batch size of 1024 across 8 GPUs.

Key Capabilities

• English to Telugu Translation: Optimized for high-quality translation between these two languages.
• Fine-tuned Performance: Achieved a loss of 0.2652 on its evaluation set, indicating effective learning for the translation task.
• Qwen2.5-3B-Instruct Base: Benefits from the robust architecture and capabilities of the Qwen2.5-3B-Instruct model.

Training Details

The model underwent training with specific hyperparameters including an AdamW optimizer, inverse square root learning rate scheduler, and a distributed training setup. It processed approximately 4.9 billion input tokens during its training phase.

Good For

• Applications requiring dedicated English-Telugu machine translation.
• Developers seeking a specialized model for language conversion in this pair.