Vocabulary Trimmed google/mt5-small: vocabtrimmer/mt5-small-trimmed-en-90000

This model is a trimmed version of google/mt5-small by vocabtrimmer, a tool for trimming vocabulary of language models to compress the model size. Following table shows a summary of the trimming process.

google/mt5-small vocabtrimmer/mt5-small-trimmed-en-90000
parameter_size_full 300,176,768 136,223,104
parameter_size_embedding 256,114,688 92,161,024
vocab_size 250,112 90,001
compression_rate_full 100.0 45.38
compression_rate_embedding 100.0 35.98

Following table shows the parameter used to trim vocabulary.

language dataset dataset_column dataset_name dataset_split target_vocab_size min_frequency
en vocabtrimmer/mc4_validation text en validation 90000 2