This model is a fine-tuned sequence-to-sequence model based on facebook/m2m100_418M for Chinese-to-English translation of Wuxia and Xianxia literature. It was developed as part of the Bachelor's Thesis (TFG):

"Enfoques de traducción automática con modelos de lenguaje en obras wuxia" (Degree in Data Science and Engineering, Universidade da Coruña)

Model Details

Base Model: facebook/m2m100_418M
Training Method: Full architectural fine-tuning
Training Dataset: HSilvosa/hsc-wuxia-100k (100,000 stratified parallel sentence pairs)
Target Domain: Wuxia/Xianxia web novels

Intended Use

This model can be used directly for Chinese-to-English translation of literary text in the Wuxia domain.

Quick Start (Inference)

python

from transformers import M2M100ForConditionalGeneration, M2M100Tokenizer

model_name = "HSilvosa/hsc-wuxia-m2m100-418m"
model = M2M100ForConditionalGeneration.from_pretrained(model_name)
tokenizer = M2M100Tokenizer.from_pretrained(model_name)

# Set source and target language
tokenizer.src_lang = "zh"
text = "一念成沧海，一念化桑田。"

encoded_zh = tokenizer(text, return_tensors="pt")
generated_tokens = model.generate(**encoded_zh, forced_bos_token_id=tokenizer.get_lang_id("en"))
print(tokenizer.batch_decode(generated_tokens, skip_special_tokens=True)[0])

Evaluation Results

The model was evaluated on the Wuxia parallel test set, comparing the base model and the fine-tuned version:

Metric	Base Model (`m2m100_418M`)	Fine-tuned Model (`hsc-wuxia-m2m100-418m`)	Delta (Improvement)
SacreBLEU	5.10	33.09	+27.99
chrF	25.19	53.24	+28.05
ROUGE-L	27.27	60.07	+32.80
METEOR	24.16	59.03	+34.87
COMET	56.17	78.52	+22.35

Training Hyperparameters

The model was fine-tuned using the following architectural training parameters:

Optimizer: AdamW
Learning Rate: 2e-5 (0.00002)
Weight Decay: 0.01
Epochs: Max 10 (with early stopping patience of 3 epochs based on validation loss)
Batch Size: 16 (per device/GPU)
Max Sequence Length: 128 tokens
Seed: 42

Project Repository

This model and the associated training/evaluation pipelines were developed in the following project repository:

GitHub Repository: HugoSilvosa/hsc-mtl-wuxia

hsilvosa/hsc-wuxia-m2m100-418m

HSC Wuxia M2M100 418M Model