Upload tokenizer_config.json with huggingface_hub
Browse files- tokenizer_config.json +21 -0
tokenizer_config.json
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": "<|bos|>",
|
| 5 |
+
"eos_token": "<|eos|>",
|
| 6 |
+
"errors": "replace",
|
| 7 |
+
"extra_special_tokens": [
|
| 8 |
+
"<|user_start|>",
|
| 9 |
+
"<|user_end|>",
|
| 10 |
+
"<|assistant_start|>",
|
| 11 |
+
"<|assistant_end|>",
|
| 12 |
+
"<|system_start|>",
|
| 13 |
+
"<|system_end|>"
|
| 14 |
+
],
|
| 15 |
+
"is_local": false,
|
| 16 |
+
"model_max_length": 1024,
|
| 17 |
+
"pad_token": null,
|
| 18 |
+
"tokenizer_class": "GPT2Tokenizer",
|
| 19 |
+
"unk_token": "<|endoftext|>",
|
| 20 |
+
"chat_template": "{% for message in messages %}{% if message['role'] == 'system' %}<|system_start|>{{ message['content'] }}<|system_end|>{% elif message['role'] == 'user' %}<|user_start|>{{ message['content'] }}<|user_end|>{% elif message['role'] == 'assistant' %}<|assistant_start|>{{ message['content'] }}<|assistant_end|>{% endif %}{% endfor %}{% if add_generation_prompt %}<|assistant_start|>{% endif %}"
|
| 21 |
+
}
|