File size: 1,696 Bytes
5470819
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
# Vocabulary Trimmed [lmqg/mt5-small-jaquad-qa](https://huggingface.co/lmqg/mt5-small-jaquad-qa): `vocabtrimmer/mt5-small-jaquad-qa-trimmed-ja` 
This model is a trimmed version of [lmqg/mt5-small-jaquad-qa](https://huggingface.co/lmqg/mt5-small-jaquad-qa) 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-jaquad-qa   | vocabtrimmer/mt5-small-jaquad-qa-trimmed-ja   |
|:---------------------------|:---------------------------|:----------------------------------------------|
| parameter_size_full        | 300,165,504                | 172,987,776                                   |
| parameter_size_embedding   | 256,103,424                | 128,925,696                                   |
| vocab_size                 | 250,101                    | 125,904                                       |
| compression_rate_full      | 100.0                      | 57.63                                         |
| compression_rate_embedding | 100.0                      | 50.34                                         |


Following table shows the parameter used to trim vocabulary.

 | language   | dataset                     | dataset_column   | dataset_name   | dataset_split   | target_vocab_size   |   min_frequency |
|:-----------|:----------------------------|:-----------------|:---------------|:----------------|:--------------------|----------------:|
| ja         | vocabtrimmer/mc4_validation | text             | ja             | validation      |                     |               2 |