Upload tokenizer_config.json

tokenizer_config.json

@@ -17,7 +17,7 @@
  "special": true
  },
  "128002": {
- "content": "<|reserved_special_token_0|>",
+ "content": "<|im_start|>",
  "lstrip": false,
  "normalized": false,
  "rstrip": false,
@@ -25,7 +25,7 @@
  "special": true
  },
  "128003": {
- "content": "<|reserved_special_token_1|>",
+ "content": "<|im_end|>",
  "lstrip": false,
  "normalized": false,
  "rstrip": false,
@@ -33,7 +33,7 @@
  "special": true
  },
  "128004": {
- "content": "<|reserved_special_token_2|>",
+ "content": "<tool_call>",
  "lstrip": false,
  "normalized": false,
  "rstrip": false,
@@ -41,7 +41,7 @@
  "special": true
  },
  "128005": {
- "content": "<|reserved_special_token_3|>",
+ "content": "<tool_response>",
  "lstrip": false,
  "normalized": false,
  "rstrip": false,
@@ -65,7 +65,7 @@
  "special": true
  },
  "128008": {
- "content": "<|reserved_special_token_4|>",
+ "content": "<tools>",
  "lstrip": false,
  "normalized": false,
  "rstrip": false,
@@ -81,7 +81,7 @@
  "special": true
  },
  "128010": {
- "content": "<|reserved_special_token_5|>",
+ "content": "</tools>",
  "lstrip": false,
  "normalized": false,
  "rstrip": false,
@@ -89,7 +89,7 @@
  "special": true
  },
  "128011": {
- "content": "<|reserved_special_token_6|>",
+ "content": "</tool_call>",
  "lstrip": false,
  "normalized": false,
  "rstrip": false,
@@ -97,7 +97,7 @@
  "special": true
  },
  "128012": {
- "content": "<|reserved_special_token_7|>",
+ "content": "</tool_response>",
  "lstrip": false,
  "normalized": false,
  "rstrip": false,
@@ -2050,12 +2050,14 @@
  }
  },
  "bos_token": "<|begin_of_text|>",
+ "chat_template": "{{bos_token}}{% for message in messages %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}",
  "clean_up_tokenization_spaces": true,
- "eos_token": "<|end_of_text|>",
+ "eos_token": "<|im_end|>",
  "model_input_names": [
  "input_ids",
  "attention_mask"
  ],
  "model_max_length": 1000000000000000019884624838656,
+ "pad_token": "<|end_of_text|>",
  "tokenizer_class": "PreTrainedTokenizerFast"
 }