Vocabulary Trimmed lmqg/mt5-small-koquad-qa: vocabtrimmer/mt5-small-koquad-qa-trimmed-ko-5000
This model is a trimmed version of lmqg/mt5-small-koquad-qa by vocabtrimmer, a tool for trimming vocabulary of language models to compress the model size.
Following table shows a summary of the trimming process.
| lmqg/mt5-small-koquad-qa | vocabtrimmer/mt5-small-koquad-qa-trimmed-ko-5000 | |
|---|---|---|
| parameter_size_full | 300,165,504 | 49,184,128 |
| parameter_size_embedding | 256,103,424 | 5,122,048 |
| vocab_size | 250,101 | 5,002 |
| compression_rate_full | 100.0 | 16.39 |
| compression_rate_embedding | 100.0 | 2.0 |
Following table shows the parameter used to trim vocabulary.
| language | dataset | dataset_column | dataset_name | dataset_split | target_vocab_size | min_frequency |
|---|---|---|---|---|---|---|
| ko | vocabtrimmer/mc4_validation | text | ko | validation | 5000 | 2 |