Upload preprocessor with huggingface_hub
Browse files
preprocessor/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
preprocessor/tokenizer_config.yaml
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
name: bpe_tokenizer
|
| 2 |
+
config_type: preprocessor
|
| 3 |
+
truncation_strategy: no_truncation
|
| 4 |
+
padding_strategy: no_padding
|
| 5 |
+
continuing_subword_prefix: ''
|
| 6 |
+
end_of_word_suffix: ''
|
| 7 |
+
fuse_unk: false
|
| 8 |
+
train_config:
|
| 9 |
+
name: bpe_tokenizer
|
| 10 |
+
config_type: preprocessor
|
| 11 |
+
vocab_size: 30000
|
| 12 |
+
min_frequency: 2
|
| 13 |
+
limit_alphabet: 1000
|
| 14 |
+
show_progress: true
|