File size: 1,637 Bytes
20464d7 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 |
# Vocabulary Trimmed [lmqg/mt5-base-esquad-ae](https://huggingface.co/lmqg/mt5-base-esquad-ae): `lmqg/mt5-base-esquad-ae-trimmed-50000`
This model is a trimmed version of [lmqg/mt5-base-esquad-ae](https://huggingface.co/lmqg/mt5-base-esquad-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-base-esquad-ae | lmqg/mt5-base-esquad-ae-trimmed-50000 |
|:---------------------------|:--------------------------|:----------------------------------------|
| parameter_size_full | 582,384,384 | 275,032,320 |
| parameter_size_embedding | 384,155,136 | 76,803,072 |
| vocab_size | 250,101 | 50,002 |
| compression_rate_full | 100.0 | 47.23 |
| 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 |
|:-----------|:----------------------------|:-----------------|:---------------|:----------------|--------------------:|----------------:|
| es | vocabtrimmer/mc4_validation | text | es | validation | 50000 | 2 | |