add tokenizer files
Browse files- README.md +2 -1
- special_tokens_map.json +2 -2
- tokenizer_config.json +3 -3
README.md
CHANGED
@@ -13,7 +13,8 @@ tags:
|
|
13 |
---
|
14 |
|
15 |
|
16 |
-
#
|
|
|
17 |
|
18 |
**Model Page**: [Gemma](https://ai.google.dev/gemma/docs)
|
19 |
|
|
|
13 |
---
|
14 |
|
15 |
|
16 |
+
# SystemGemma2 9B model card
|
17 |
+
This is a version of [Gemma 2 9B](https://huggingface.co/google/gemma-2-9b-it) with system prompts enabled.
|
18 |
|
19 |
**Model Page**: [Gemma](https://ai.google.dev/gemma/docs)
|
20 |
|
special_tokens_map.json
CHANGED
@@ -1,7 +1,7 @@
|
|
1 |
{
|
2 |
"additional_special_tokens": [
|
3 |
-
|
4 |
-
|
5 |
],
|
6 |
"bos_token": {
|
7 |
"content": "<bos>",
|
|
|
1 |
{
|
2 |
"additional_special_tokens": [
|
3 |
+
"<start_of_turn>",
|
4 |
+
"<end_of_turn>"
|
5 |
],
|
6 |
"bos_token": {
|
7 |
"content": "<bos>",
|
tokenizer_config.json
CHANGED
@@ -1996,11 +1996,11 @@
|
|
1996 |
}
|
1997 |
},
|
1998 |
"additional_special_tokens": [
|
1999 |
-
|
2000 |
-
|
2001 |
],
|
2002 |
"bos_token": "<bos>",
|
2003 |
-
"chat_template": "{{ bos_token }}{% if messages[0]['role'] == 'system' %}{{
|
2004 |
"clean_up_tokenization_spaces": false,
|
2005 |
"eos_token": "<eos>",
|
2006 |
"model_max_length": 1000000000000000019884624838656,
|
|
|
1996 |
}
|
1997 |
},
|
1998 |
"additional_special_tokens": [
|
1999 |
+
"<start_of_turn>",
|
2000 |
+
"<end_of_turn>"
|
2001 |
],
|
2002 |
"bos_token": "<bos>",
|
2003 |
+
"chat_template": "{{ bos_token }}{% if messages[0]['role'] == 'system' %}{{ messages[0]['content'] + '\n' }}{% endif %}{% for message in messages %}{% if (message['role'] == 'user') != ((loop.index0 + 1) % 2 == 0) %}{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}{% endif %}{% if (message['role'] == 'assistant') %}{% set role = 'model' %}{% else %}{% set role = message['role'] %}{% endif %}{{ '<start_of_turn>' + role + '\n' + message['content'] | trim + '<end_of_turn>\n' }}{% endfor %}{% if add_generation_prompt %}{{'<start_of_turn>model\n'}}{% endif %}",
|
2004 |
"clean_up_tokenization_spaces": false,
|
2005 |
"eos_token": "<eos>",
|
2006 |
"model_max_length": 1000000000000000019884624838656,
|