commit files to HF hub
Browse files
README.md
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# Vocabulary Trimmed [facebook/mbart-large-cc25](https://huggingface.co/facebook/mbart-large-cc25): `vocabtrimmer/mbart-large-cc25-trimmed-ja`
|
| 2 |
+
This model is a trimmed version of [facebook/mbart-large-cc25](https://huggingface.co/facebook/mbart-large-cc25) by [`vocabtrimmer`](https://github.com/asahi417/lm-vocab-trimmer), a tool for trimming vocabulary of language models to compress the model size.
|
| 3 |
+
Following table shows a summary of the trimming process.
|
| 4 |
+
|
| 5 |
+
| | facebook/mbart-large-cc25 | vocabtrimmer/mbart-large-cc25-trimmed-ja |
|
| 6 |
+
|:---------------------------|:----------------------------|:-------------------------------------------|
|
| 7 |
+
| parameter_size_full | 610,851,840 | 434,447,360 |
|
| 8 |
+
| parameter_size_embedding | 512,055,296 | 159,246,336 |
|
| 9 |
+
| vocab_size | 250,027 | 77,757 |
|
| 10 |
+
| compression_rate_full | 100.0 | 71.12 |
|
| 11 |
+
| compression_rate_embedding | 100.0 | 31.1 |
|
| 12 |
+
|
| 13 |
+
|
| 14 |
+
Following table shows the parameter used to trim vocabulary.
|
| 15 |
+
|
| 16 |
+
| language | dataset | dataset_column | dataset_name | dataset_split | target_vocab_size | min_frequency |
|
| 17 |
+
|:-----------|:----------------------------|:-----------------|:---------------|:----------------|:--------------------|----------------:|
|
| 18 |
+
| ja | vocabtrimmer/mc4_validation | text | ja | validation | | 2 |
|