Hugging Face's logo

mayflowergmbh
/
occiglot-7b-eu5-instruct-HQQ

Text Generation
Transformers
mistral
hqq
conversational
Inference Endpoints
Model card Files Community
johannhartmann commited on
Commit
120b709
•
1 Parent(s): e0862d5

Upload folder using huggingface_hub

Browse files
Files changed (8) hide show
  1. README.md +217 -0
  2. added_tokens.json +4 -0
  3. config.json +26 -0
  4. qmodel.pt +3 -0
  5. special_tokens_map.json +30 -0
  6. tokenizer.json +0 -0
  7. tokenizer.model +3 -0
  8. tokenizer_config.json +59 -0
README.md ADDED
@@ -0,0 +1,217 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - en
4
+ - es
5
+ - de
6
+ - fr
7
+ - it
8
+ license: apache-2.0
9
+ tags:
10
+ - hqq
11
+ pipeline_tag: text-generation
12
+ ---
13
+
14
+ ![image/png](https://huggingface.co/datasets/malteos/images/resolve/main/occiglot.medium.png)
15
+
16
+ # Occiglot-7B-EU5-Instruct
17
+
18
+ > A [polyglot](https://en.wikipedia.org/wiki/Multilingualism#In_individuals) language model for the [Occident](https://en.wikipedia.org/wiki/Occident).
19
+ >
20
+
21
+ **Occiglot-7B-EU5-Instruct** is a the instruct version of [occiglot-7b-eu5](https://huggingface.co/occiglot/occiglot-7b-eu5/), a generative language model with 7B parameters supporting the top-5 EU languages (English, Spanish, French, German, and Italian) and trained by the [Occiglot Research Collective](https://occiglot.github.io/occiglot/).
22
+ It was trained on 400M tokens of additional multilingual and code instructions.
23
+ Note that the model was not safety aligned and might generate problematic outputs.
24
+
25
+ This is the first release of an ongoing open research project for multilingual language models.
26
+ If you want to train a model for your own language or are working on evaluations, please contact us or join our [Discord server](https://discord.gg/wUpvYs4XvM). **We are open for collaborations!**
27
+
28
+
29
+ ### Model details
30
+
31
+ - **Instruction tuned from:** [occiglot-7b-eu5](https://huggingface.co/occiglot/occiglot-7b-eu5)
32
+ - **Model type:** Causal decoder-only transformer language model
33
+ - **Languages:** English, Spanish, French, German, Italian, and code.
34
+ - **License:** [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0.html)
35
+ - **Compute resources:** [DFKI cluster](https://www.dfki.de/en/web)
36
+ - **Contributors:** Manuel Brack, Patrick Schramowski, Pedro Ortiz, Malte Ostendorff, Fabio Barth, Georg Rehm, Kristian Kersting
37
+ - **Research labs:** [Occiglot](https://occiglot.github.io/occiglot/) with support from [SAINT](https://www.dfki.de/en/web/research/research-departments/foundations-of-systems-ai) and [SLT](https://www.dfki.de/en/web/research/research-departments/speech-and-language-technology)
38
+ - **Contact:** [Discord](https://discord.gg/wUpvYs4XvM)
39
+
40
+ ### How to use
41
+
42
+ The model was trained using the chatml instruction template. You can use the transformers chat template feature for interaction.
43
+ Since the generation relies on some randomness, we
44
+ set a seed for reproducibility:
45
+
46
+ ```python
47
+ >>> from transformers import AutoTokenizer, MistralForCausalLM, set_seed
48
+ >>> tokenizer = AutoTokenizer.from_pretrained("occiglot/occiglot-7b-eu5-instruct")
49
+ >>> model = MistralForCausalLM.from_pretrained('occiglot/occiglot-7b-eu5-instruct') # You may want to use bfloat16 and/or move to GPU here
50
+ >>> set_seed(42)
51
+ >>> messages = [
52
+ >>> {"role": "system", 'content': 'You are a helpful assistant. Please give short and concise answers.'},
53
+ >>> {"role": "user", "content": "Wer ist der deutsche Bundeskanzler?"},
54
+ >>> ]
55
+ >>> tokenized_chat = tokenizer.apply_chat_template(messages, tokenize=True, add_generation_prompt=True, return_dict=False, return_tensors='pt',)
56
+ >>> set_seed(42)
57
+ >>> outputs = model.generate(tokenized_chat.to('cuda'), max_new_tokens=200,)
58
+ >>> tokenizer.decode(out[0][len(tokenized_chat[0]):])
59
+ 'Der deutsche Bundeskanzler ist Olaf Scholz.'
60
+ ```
61
+
62
+ ## Dataset
63
+
64
+ The training data was split evenly amongst the 5 languages based on the total number of tokens. We would like to thank [Disco Research](https://huggingface.co/DiscoResearch), [Jan Philipp Harries](https://huggingface.co/jphme), and [Björn Plüster](https://huggingface.co/bjoernp) for making their dataset available to us.
65
+
66
+
67
+ **English and Code**
68
+ - [Open-Hermes-2B](https://huggingface.co/datasets/teknium/OpenHermes-2.5)
69
+
70
+ **German**
71
+ - [DiscoLM German Dataset](https://huggingface.co/DiscoResearch) includes the publicly available [germanrag](https://huggingface.co/datasets/DiscoResearch/germanrag) dataset
72
+ - [OASST-2](https://huggingface.co/datasets/OpenAssistant/oasst2) (German subset)
73
+ - [Aya-Dataset](https://huggingface.co/datasets/CohereForAI/aya_dataset) (German subset)
74
+
75
+ **Spanish**
76
+ - [Mentor-ES](https://huggingface.co/datasets/projecte-aina/MentorES)
77
+ - [Squad-es](https://huggingface.co/datasets/squad_es)
78
+ - [OASST-2](https://huggingface.co/datasets/OpenAssistant/oasst2) (Spanish subset)
79
+ - [Aya-Dataset](https://huggingface.co/datasets/CohereForAI/aya_dataset) (Spanish subset)
80
+
81
+ **French**
82
+ - [Bactrian-X](https://huggingface.co/datasets/MBZUAI/Bactrian-X) (French subset)
83
+ - [AI-Society Translated](https://huggingface.co/datasets/camel-ai/ai_society_translated) (French subset)
84
+ - [GT-Dorimiti](https://huggingface.co/datasets/Gt-Doremiti/gt-doremiti-instructions)
85
+ - [OASST-2](https://huggingface.co/datasets/OpenAssistant/oasst2) (French subset)
86
+ - [Aya-Dataset](https://huggingface.co/datasets/CohereForAI/aya_dataset) (French subset)
87
+
88
+ **Italian**
89
+ - [Quora-IT-Baize](https://huggingface.co/datasets/andreabac3/Quora-Italian-Fauno-Baize)
90
+ - [Stackoverflow-IT-Vaize](https://huggingface.co/datasets/andreabac3/StackOverflow-Italian-Fauno-Baize)
91
+ - [Camoscio](https://huggingface.co/datasets/teelinsan/camoscio_cleaned)
92
+ - [OASST-2](https://huggingface.co/datasets/OpenAssistant/oasst2) (Italian subset)
93
+ - [Aya-Dataset](https://huggingface.co/datasets/CohereForAI/aya_dataset) (Italian subset)
94
+
95
+ ## Training settings
96
+
97
+ - Full instruction fine-tuning on 8xH100.
98
+ - 0.6 - 4 training epochs (depending on dataset sampling).
99
+ - Framework: [axolotl](https://github.com/OpenAccess-AI-Collective/axolotl)
100
+ - Precision: bf16
101
+ - Optimizer: AdamW
102
+ - Global batch size: 128 (with 8192 context length)
103
+ - Cosine Annealing with Warmup
104
+
105
+
106
+ ## Tokenizer
107
+
108
+ Tokenizer is unchanged from [Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1).
109
+
110
+ ## Evaluation
111
+
112
+ Preliminary evaluation results can be found below.
113
+ Please note that the non-English results are based on partially machine-translated datasets and English prompts ([Belebele](https://huggingface.co/datasets/facebook/belebele) and [Okapi framework](https://github.com/nlp-uoregon/Okapi)) and thus should be interpreted with caution, e.g., biased towards English model performance.
114
+ Currently, we are working on more suitable benchmarks for Spanish, French, German, and Italian.
115
+
116
+ <details>
117
+ <summary>Evaluation results</summary>
118
+
119
+ ### All 5 Languages
120
+
121
+ | | avg | arc_challenge | belebele | hellaswag | mmlu | truthfulqa |
122
+ |:---------------------------|---------:|----------------:|-----------:|------------:|---------:|-------------:|
123
+ | Occiglot-7b-eu5 | 0.516895 | 0.508109 | 0.675556 | 0.718963 | 0.402064 | 0.279782 |
124
+ | Occiglot-7b-eu5-instruct | 0.537799 | 0.53632 | 0.691111 | 0.731918 | 0.405198 | 0.32445 |
125
+ | Occiglot-7b-de-en | 0.518337 | 0.496297 | 0.715111 | 0.669034 | 0.412545 | 0.298697 |
126
+ | Occiglot-7b-de-en-instruct | 0.543173 | 0.530826 | 0.745778 | 0.67676 | 0.411326 | 0.351176 |
127
+ | Occiglot-7b-it-en | 0.513221 | 0.500564 | 0.694444 | 0.668099 | 0.413528 | 0.289469 |
128
+ | Occiglot-7b-it-en-instruct | 0.53721 | 0.523128 | 0.726667 | 0.683414 | 0.414918 | 0.337927 |
129
+ | Occiglot-7b-fr-en | 0.509209 | 0.496806 | 0.691333 | 0.667475 | 0.409129 | 0.281303 |
130
+ | Occiglot-7b-fr-en-instruct | 0.52884 | 0.515613 | 0.723333 | 0.67371 | 0.413024 | 0.318521 |
131
+ | Occiglot-7b-es-en | 0.483388 | 0.482949 | 0.606889 | 0.653902 | 0.398922 | 0.274277 |
132
+ | Occiglot-7b-es-en-instruct | 0.504023 | 0.494576 | 0.65 | 0.670847 | 0.406176 | 0.298513 |
133
+ | Leo-mistral-hessianai-7b | 0.484806 | 0.462103 | 0.653556 | 0.642242 | 0.379208 | 0.28692 |
134
+ | Claire-mistral-7b-0.1 | 0.514226 | 0.502773 | 0.705111 | 0.666871 | 0.412128 | 0.284245 |
135
+ | Lince-mistral-7b-it-es | 0.543427 | 0.540222 | 0.745111 | 0.692931 | 0.426241 | 0.312629 |
136
+ | Cerbero-7b | 0.532385 | 0.513714 | 0.743111 | 0.654061 | 0.427566 | 0.323475 |
137
+ | Mistral-7b-v0.1 | 0.547111 | 0.528937 | 0.768444 | 0.682516 | 0.448253 | 0.307403 |
138
+ | Mistral-7b-instruct-v0.2 | 0.56713 | 0.547228 | 0.741111 | 0.69455 | 0.422501 | 0.430262 |
139
+
140
+
141
+ ### English
142
+
143
+ | | avg | arc_challenge | belebele | hellaswag | mmlu | truthfulqa |
144
+ |:---------------------------|---------:|----------------:|-----------:|------------:|---------:|-------------:|
145
+ | Occiglot-7b-eu5 | 0.59657 | 0.530717 | 0.726667 | 0.789882 | 0.531904 | 0.403678 |
146
+ | Occiglot-7b-eu5-instruct | 0.617905 | 0.558874 | 0.746667 | 0.799841 | 0.535109 | 0.449 |
147
+ | Leo-mistral-hessianai-7b | 0.600949 | 0.522184 | 0.736667 | 0.777833 | 0.538812 | 0.429248 |
148
+ | Mistral-7b-v0.1 | 0.668385 | 0.612628 | 0.844444 | 0.834097 | 0.624555 | 0.426201 |
149
+ | Mistral-7b-instruct-v0.2 | 0.713657 | 0.637372 | 0.824444 | 0.846345 | 0.59201 | 0.668116 |
150
+
151
+ ### German
152
+
153
+ | | avg | arc_challenge_de | belebele_de | hellaswag_de | mmlu_de | truthfulqa_de |
154
+ |:---------------------------|---------:|-------------------:|--------------:|---------------:|----------:|----------------:|
155
+ | Occiglot-7b-eu5 | 0.508311 | 0.493584 | 0.646667 | 0.666631 | 0.483406 | 0.251269 |
156
+ | Occiglot-7b-eu5-instruct | 0.531506 | 0.529512 | 0.667778 | 0.685205 | 0.488234 | 0.286802 |
157
+ | Occiglot-7b-de-en | 0.540085 | 0.50556 | 0.743333 | 0.67421 | 0.514633 | 0.26269 |
158
+ | Occiglot-7b-de-en-instruct | 0.566474 | 0.54491 | 0.772222 | 0.688407 | 0.515915 | 0.310914 |
159
+ | Leo-mistral-hessianai-7b | 0.517766 | 0.474765 | 0.691111 | 0.682109 | 0.488309 | 0.252538 |
160
+ | Mistral-7b-v0.1 | 0.527957 | 0.476476 | 0.738889 | 0.610589 | 0.529567 | 0.284264 |
161
+ | Mistral-7b-instruct-v0.2 | 0.535215 | 0.485885 | 0.688889 | 0.622438 | 0.501961 | 0.376904 |
162
+
163
+ ### Spanish
164
+
165
+ | | avg | arc_challenge_es | belebele_es | hellaswag_es | mmlu_es | truthfulqa_es |
166
+ |:---------------------------|---------:|-------------------:|--------------:|---------------:|----------:|----------------:|
167
+ | Occiglot-7b-eu5 | 0.533194 | 0.508547 | 0.676667 | 0.725411 | 0.499325 | 0.25602 |
168
+ | Occiglot-7b-eu5-instruct | 0.548155 | 0.535043 | 0.68 | 0.737039 | 0.503525 | 0.285171 |
169
+ | Occiglot-7b-es-en | 0.527264 | 0.529915 | 0.627778 | 0.72253 | 0.512749 | 0.243346 |
170
+ | Occiglot-7b-es-en-instruct | 0.5396 | 0.545299 | 0.636667 | 0.734372 | 0.524374 | 0.257288 |
171
+ | Lince-mistral-7b-it-es | 0.547212 | 0.52906 | 0.721111 | 0.687967 | 0.512749 | 0.285171 |
172
+ | Mistral-7b-v0.1 | 0.554817 | 0.528205 | 0.747778 | 0.672712 | 0.544023 | 0.281369 |
173
+ | Mistral-7b-instruct-v0.2 | 0.568575 | 0.54188 | 0.73 | 0.685406 | 0.511699 | 0.373891 |
174
+
175
+ ### French
176
+
177
+ | | avg | arc_challenge_fr | belebele_fr | hellaswag_fr | mmlu_fr | truthfulqa_fr |
178
+ |:---------------------------|---------:|-------------------:|--------------:|---------------:|----------:|----------------:|
179
+ | Occiglot-7b-eu5 | 0.525017 | 0.506416 | 0.675556 | 0.712358 | 0.495684 | 0.23507 |
180
+ | Occiglot-7b-eu5-instruct | 0.554216 | 0.541488 | 0.7 | 0.724245 | 0.499122 | 0.306226 |
181
+ | Occiglot-7b-fr-en | 0.542903 | 0.532934 | 0.706667 | 0.718891 | 0.51333 | 0.242694 |
182
+ | Occiglot-7b-fr-en-instruct | 0.567079 | 0.542344 | 0.752222 | 0.72553 | 0.52051 | 0.29479 |
183
+ | Claire-mistral-7b-0.1 | 0.515127 | 0.486741 | 0.694444 | 0.642964 | 0.479566 | 0.271919 |
184
+ | Cerbero-7b | 0.526044 | 0.462789 | 0.735556 | 0.624438 | 0.516462 | 0.290978 |
185
+ | Mistral-7b-v0.1 | 0.558129 | 0.525235 | 0.776667 | 0.66481 | 0.543121 | 0.280813 |
186
+ | Mistral-7b-instruct-v0.2 | 0.575821 | 0.551754 | 0.758889 | 0.67916 | 0.506837 | 0.382465 |
187
+
188
+ ### Italian
189
+
190
+ | | avg | arc_challenge_it | belebele_it | hellaswag_it | mmlu_it | truthfulqa_it |
191
+ |:---------------------------|---------:|-------------------:|--------------:|---------------:|----------:|----------------:|
192
+ | Occiglot-7b-eu5 | 0.421382 | 0.501283 | 0.652222 | 0.700533 | 0 | 0.252874 |
193
+ | Occiglot-7b-eu5-instruct | 0.437214 | 0.516681 | 0.661111 | 0.71326 | 0 | 0.295019 |
194
+ | Occiglot-7b-it-en | 0.432667 | 0.536356 | 0.684444 | 0.694768 | 0 | 0.247765 |
195
+ | Occiglot-7b-it-en-instruct | 0.456261 | 0.545766 | 0.717778 | 0.713804 | 0 | 0.303959 |
196
+ | Cerbero-7b | 0.434939 | 0.522669 | 0.717778 | 0.631567 | 0 | 0.302682 |
197
+ | Mistral-7b-v0.1 | 0.426264 | 0.502139 | 0.734444 | 0.630371 | 0 | 0.264368 |
198
+ | Mistral-7b-instruct-v0.2 | 0.442383 | 0.519247 | 0.703333 | 0.6394 | 0 | 0.349936 |
199
+
200
+
201
+ </details>
202
+
203
+ ## Acknowledgements
204
+
205
+ The pre-trained model training was supported by a compute grant at the [42 supercomputer](https://hessian.ai/) which is a central component in the development of [hessian AI](https://hessian.ai/), the [AI Innovation Lab](https://hessian.ai/infrastructure/ai-innovationlab/) (funded by the [Hessian Ministry of Higher Education, Research and the Art (HMWK)](https://wissenschaft.hessen.de) & the [Hessian Ministry of the Interior, for Security and Homeland Security (HMinD)](https://innen.hessen.de)) and the [AI Service Centers](https://hessian.ai/infrastructure/ai-service-centre/) (funded by the [German Federal Ministry for Economic Affairs and Climate Action (BMWK)](https://www.bmwk.de/Navigation/EN/Home/home.html)).
206
+ The curation of the training data is partially funded by the [German Federal Ministry for Economic Affairs and Climate Action (BMWK)](https://www.bmwk.de/Navigation/EN/Home/home.html)
207
+ through the project [OpenGPT-X](https://opengpt-x.de/en/) (project no. 68GX21007D).
208
+
209
+
210
+ ## License
211
+
212
+ [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0.html)
213
+
214
+ ## See also
215
+
216
+ - https://huggingface.co/collections/occiglot/occiglot-eu5-7b-v01-65dbed502a6348b052695e01
217
+
added_tokens.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "<|im_end|>": 32001,
3
+ "<|im_start|>": 32000
4
+ }
config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "./mayflowergmbh/occiglot-7b-eu5-instruct-hf",
3
+ "architectures": [
4
+ "MistralForCausalLM"
5
+ ],
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 1,
8
+ "eos_token_id": 2,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 4096,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 14336,
13
+ "max_position_embeddings": 32768,
14
+ "model_type": "mistral",
15
+ "num_attention_heads": 32,
16
+ "num_hidden_layers": 32,
17
+ "num_key_value_heads": 8,
18
+ "rms_norm_eps": 1e-05,
19
+ "rope_theta": 10000.0,
20
+ "sliding_window": 4096,
21
+ "tie_word_embeddings": false,
22
+ "torch_dtype": "float16",
23
+ "transformers_version": "4.40.0.dev0",
24
+ "use_cache": false,
25
+ "vocab_size": 32002
26
+ }
qmodel.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6843b6234b2abb668d3251e9a1744c07c5db6b8ae990039cee1f5e51286bf689
3
+ size 2597218112
special_tokens_map.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "</s>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "unk_token": {
24
+ "content": "<unk>",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ }
30
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dadfd56d766715c61d2ef780a525ab43b8e6da4de6865bda3d95fdef5e134055
3
+ size 493443
tokenizer_config.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "added_tokens_decoder": {
5
+ "0": {
6
+ "content": "<unk>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "1": {
14
+ "content": "<s>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "2": {
22
+ "content": "</s>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ },
29
+ "32000": {
30
+ "content": "<|im_start|>",
31
+ "lstrip": false,
32
+ "normalized": false,
33
+ "rstrip": false,
34
+ "single_word": false,
35
+ "special": false
36
+ },
37
+ "32001": {
38
+ "content": "<|im_end|>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false,
43
+ "special": false
44
+ }
45
+ },
46
+ "additional_special_tokens": [],
47
+ "bos_token": "<s>",
48
+ "chat_template": "{{'<s>'}}{% if messages[0]['role'] == 'system' %}{% set loop_messages = messages[1:] %}{% set system_message = messages[0]['content'] %}{% else %}{% set loop_messages = messages %}{% set system_message = 'You are a helpful assistant. Please give a long and detailed answer.' %}{% endif %}{% if not add_generation_prompt is defined %}{% set add_generation_prompt = false %}{% endif %}{% for message in loop_messages %}{% if loop.index0 == 0 %}{{'<|im_start|>system\n' + system_message + '<|im_end|>\n'}}{% endif %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}",
49
+ "clean_up_tokenization_spaces": false,
50
+ "eos_token": "</s>",
51
+ "legacy": true,
52
+ "model_max_length": 1000000000000000019884624838656,
53
+ "pad_token": "</s>",
54
+ "sp_model_kwargs": {},
55
+ "spaces_between_special_tokens": false,
56
+ "tokenizer_class": "LlamaTokenizer",
57
+ "unk_token": "<unk>",
58
+ "use_default_system_prompt": false
59
+ }