· Xiaojing Yang · NLP & LLMs
LoRA for Neural Machine Translation
A practical explanation of LoRA for NMT: what it changes, why it is useful for domain adaptation, and how it differs from full fine-tuning.
Blog
Foundations and research notes on mathematics, statistics, machine learning, NLP, LLMs, model evaluation, and applied AI systems.
- · Xiaojing Yang · Research Engineering
Data Quality Diagnostics for Parallel Corpora
A practical note on diagnosing parallel corpora through alignment, duplicates, length ratio, completeness, terminology, and domain coverage.
- · Xiaojing Yang · Mathematics
Vectors, Matrices, and Embeddings
A foundation note connecting linear algebra to embeddings, similarity, neural layers, and retrieval.
- · Xiaojing Yang · Research Engineering
Hyperparameter Optimization with Grid Search and Optuna/ASHA
A practical research-engineering note on tuning LoRA rank, alpha, dropout, Pareto selection, ASHA pruning, and fANOVA analysis.
- · Xiaojing Yang · Model Evaluation
How to Evaluate Machine Translation Beyond BLEU
A practical guide to BLEU, chrF, COMET, terminology metrics, and human error analysis for domain-specific MT.
- · Xiaojing Yang · Applied ML Systems
LoRA vs Full Fine-Tuning: What Do We Actually Save?
A practical comparison of LoRA and full fine-tuning in low-resource domain MT: parameters, compute, stability, performance, and trade-offs.