# Vocabulary Trimmed [lmqg/mt5-small-koquad-qg-ae](https://huggingface.co/lmqg/mt5-small-koquad-qg-ae): `lmqg/mt5-small-koquad-qg-ae-trimmed-50000` 
This model is a trimmed version of [lmqg/mt5-small-koquad-qg-ae](https://huggingface.co/lmqg/mt5-small-koquad-qg-ae) by [`vocabtrimmer`](https://github.com/asahi417/lm-vocab-trimmer), 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-qg-ae   | lmqg/mt5-small-koquad-qg-ae-trimmed-50000   |
|:---------------------------|:------------------------------|:--------------------------------------------|
| parameter_size_full        | 300,165,504                   | 95,264,128                                  |
| parameter_size_embedding   | 256,103,424                   | 51,202,048                                  |
| vocab_size                 | 250,101                       | 50,002                                      |
| compression_rate_full      | 100.0                         | 31.74                                       |
| compression_rate_embedding | 100.0                         | 19.99                                       |


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      |               50000 |               2 |