teknium commited on
Commit
40d4c37
1 Parent(s): 95074f9

Upload 2 files

Browse files
Files changed (2) hide show
  1. tokenizer.json +3 -3
  2. tokenizer_config.json +3 -5
tokenizer.json CHANGED
@@ -356,16 +356,16 @@
356
  },
357
  {
358
  "id": 128039,
359
- "content": "<|reserved_special_token_31|>",
360
  "single_word": false,
361
  "lstrip": false,
362
  "rstrip": false,
363
  "normalized": false,
364
- "special": true
365
  },
366
  {
367
  "id": 128040,
368
- "content": "<|reserved_special_token_32|>",
369
  "single_word": false,
370
  "lstrip": false,
371
  "rstrip": false,
 
356
  },
357
  {
358
  "id": 128039,
359
+ "content": "<|im_start|>",
360
  "single_word": false,
361
  "lstrip": false,
362
  "rstrip": false,
363
  "normalized": false,
364
+ "special": false
365
  },
366
  {
367
  "id": 128040,
368
+ "content": "<|im_end|>",
369
  "single_word": false,
370
  "lstrip": false,
371
  "rstrip": false,
tokenizer_config.json CHANGED
@@ -313,15 +313,15 @@
313
  "special": false
314
  },
315
  "128039": {
316
- "content": "<|reserved_special_token_31|>",
317
  "lstrip": false,
318
  "normalized": false,
319
  "rstrip": false,
320
  "single_word": false,
321
- "special": true
322
  },
323
  "128040": {
324
- "content": "<|reserved_special_token_32|>",
325
  "lstrip": false,
326
  "normalized": false,
327
  "rstrip": false,
@@ -2050,10 +2050,8 @@
2050
  }
2051
  },
2052
  "bos_token": "<|begin_of_text|>",
2053
- "chat_template": "{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}",
2054
  "clean_up_tokenization_spaces": true,
2055
  "eos_token": "<|end_of_text|>",
2056
- "pad_token": "<|end_of_text|>",
2057
  "model_input_names": [
2058
  "input_ids",
2059
  "attention_mask"
 
313
  "special": false
314
  },
315
  "128039": {
316
+ "content": "<|im_start|>",
317
  "lstrip": false,
318
  "normalized": false,
319
  "rstrip": false,
320
  "single_word": false,
321
+ "special": false
322
  },
323
  "128040": {
324
+ "content": "<|im_end|>",
325
  "lstrip": false,
326
  "normalized": false,
327
  "rstrip": false,
 
2050
  }
2051
  },
2052
  "bos_token": "<|begin_of_text|>",
 
2053
  "clean_up_tokenization_spaces": true,
2054
  "eos_token": "<|end_of_text|>",
 
2055
  "model_input_names": [
2056
  "input_ids",
2057
  "attention_mask"