DWA-KD: Dual-Space Weighting and Time-Warped Alignment for Cross-Tokenizer Knowledge Distillation.
Duc Trung Vu, Chi Pham Khanh, Phi Van Dat, Ngo Van Linh, Dinh Viet Sang, Trung Le
Browse the full EACL paper archive.
Duc Trung Vu, Chi Pham Khanh, Phi Van Dat, Ngo Van Linh, Dinh Viet Sang, Trung Le
Browse the full EACL paper archive.