AlanRobotics
commited on
Commit
•
962a5b4
1
Parent(s):
455d6c5
Upload tokenizer
Browse files- tokenizer.json +1 -0
- tokenizer_config.json +1 -0
tokenizer.json
CHANGED
@@ -382,6 +382,7 @@
|
|
382 |
"end_of_word_suffix": "",
|
383 |
"fuse_unk": false,
|
384 |
"byte_fallback": false,
|
|
|
385 |
"vocab": {
|
386 |
"!": 0,
|
387 |
"\"": 1,
|
|
|
382 |
"end_of_word_suffix": "",
|
383 |
"fuse_unk": false,
|
384 |
"byte_fallback": false,
|
385 |
+
"ignore_merges": false,
|
386 |
"vocab": {
|
387 |
"!": 0,
|
388 |
"\"": 1,
|
tokenizer_config.json
CHANGED
@@ -318,6 +318,7 @@
|
|
318 |
"clean_up_tokenization_spaces": true,
|
319 |
"eos_token": "<|endoftext|>",
|
320 |
"model_max_length": 2048,
|
|
|
321 |
"tokenizer_class": "CodeGenTokenizer",
|
322 |
"unk_token": "<|endoftext|>"
|
323 |
}
|
|
|
318 |
"clean_up_tokenization_spaces": true,
|
319 |
"eos_token": "<|endoftext|>",
|
320 |
"model_max_length": 2048,
|
321 |
+
"return_token_type_ids": false,
|
322 |
"tokenizer_class": "CodeGenTokenizer",
|
323 |
"unk_token": "<|endoftext|>"
|
324 |
}
|