Transformers
lucyknada commited on
Commit
fa5f87b
1 Parent(s): e008e1a

Upload ./README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +133 -0
README.md ADDED
@@ -0,0 +1,133 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ inference: false
3
+ library_name: transformers
4
+ language:
5
+ - en
6
+ - fr
7
+ - de
8
+ - es
9
+ - it
10
+ - pt
11
+ - ja
12
+ - ko
13
+ - zh
14
+ - ar
15
+ - el
16
+ - fa
17
+ - pl
18
+ - id
19
+ - cs
20
+ - he
21
+ - hi
22
+ - nl
23
+ - ro
24
+ - ru
25
+ - tr
26
+ - uk
27
+ - vi
28
+ license: cc-by-nc-4.0
29
+ extra_gated_prompt: "By submitting this form, you agree to the [License Agreement](https://cohere.com/c4ai-cc-by-nc-license) and acknowledge that the information you provide will be collected, used, and shared in accordance with Cohere’s [Privacy Policy]( https://cohere.com/privacy). You’ll receive email updates about C4AI and Cohere research, events, products and services. You can unsubscribe at any time."
30
+ extra_gated_fields:
31
+ Name: text
32
+ Affiliation: text
33
+ Country: country
34
+ I agree to use this model for non-commercial use ONLY: checkbox
35
+ ---
36
+ ### exl2 quant (measurement.json in main branch)
37
+ ---
38
+ ### check revisions for quants
39
+ ---
40
+
41
+
42
+ # Model Card for Aya Expanse 8B
43
+
44
+ <img src="aya-expanse-8B.png" width="650" style="margin-left:'auto' margin-right:'auto' display:'block'"/>
45
+
46
+ Aya Expanse is an open-weight research release of a model with highly advanced multilingual capabilities. It focuses on pairing a highly performant pre-trained [Command family](https://huggingface.co/CohereForAI/c4ai-command-r-plus) of models with the result of a year’s dedicated research from [Cohere For AI](https://cohere.for.ai/), including [data arbitrage](https://arxiv.org/pdf/2408.14960), [multilingual preference training](https://arxiv.org/abs/2407.02552), [safety tuning](https://arxiv.org/abs/2406.18682), and [model merging](https://arxiv.org/abs/2410.10801). The result is a powerful multilingual large language model serving 23 languages.
47
+
48
+ We cover 23 languages: Arabic, Chinese (simplified & traditional), Czech, Dutch, English, French, German, Greek, Hebrew, Hebrew, Hindi, Indonesian, Italian, Japanese, Korean, Persian, Polish, Portuguese, Romanian, Russian, Spanish, Turkish, Ukrainian, and Vietnamese
49
+
50
+ This model card corresponds to the 8-billion version of the Aya Expanse model. We also released an 32-billion version which you can find [here](https://huggingface.co/CohereForAI/aya-expanse-32B).
51
+
52
+ - Developed by: [Cohere For AI](https://cohere.for.ai/)
53
+ - Point of Contact: Cohere For AI: [cohere.for.ai](https://cohere.for.ai/)
54
+ - License: [CC-BY-NC](https://cohere.com/c4ai-cc-by-nc-license), requires also adhering to [C4AI's Acceptable Use Policy](https://docs.cohere.com/docs/c4ai-acceptable-use-policy)
55
+ - Model: Aya Expanse 8B
56
+ - Model Size: 8 billion parameters
57
+
58
+ **Try Aya Expanse**
59
+
60
+ Before downloading the weights, you can try out Aya Expanse in our hosted [Hugging Face Space](https://huggingface.co/spaces/CohereForAI/aya_expanse).
61
+
62
+
63
+ ### Usage
64
+
65
+ Please install transformers from the source repository.
66
+
67
+ ```python
68
+ # pip install 'git+https://github.com/huggingface/transformers.git'
69
+ from transformers import AutoTokenizer, AutoModelForCausalLM
70
+
71
+ model_id = "CohereForAI/aya-expanse-8b"
72
+ tokenizer = AutoTokenizer.from_pretrained(model_id)
73
+ model = AutoModelForCausalLM.from_pretrained(model_id)
74
+
75
+ # Format the message with the chat template
76
+ messages = [{"role": "user", "content": "Anneme onu ne kadar sevdiğimi anlatan bir mektup yaz"}]
77
+ input_ids = tokenizer.apply_chat_template(messages, tokenize=True, add_generation_prompt=True, return_tensors="pt")
78
+ ## <BOS_TOKEN><|START_OF_TURN_TOKEN|><|USER_TOKEN|>Anneme onu ne kadar sevdiğimi anlatan bir mektup yaz<|END_OF_TURN_TOKEN|><|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>
79
+
80
+ gen_tokens = model.generate(
81
+ input_ids,
82
+ max_new_tokens=100,
83
+ do_sample=True,
84
+ temperature=0.3,
85
+ )
86
+
87
+ gen_text = tokenizer.decode(gen_tokens[0])
88
+ print(gen_text)
89
+ ```
90
+
91
+ ### Example Notebooks
92
+
93
+ **Fine-Tuning**:
94
+ - [This notebook](https://colab.research.google.com/drive/1ryPYXzqb7oIn2fchMLdCNSIH5KfyEtv4) showcases a detailed use of fine-tuning Aya Expanse on more languages.
95
+
96
+ **Example Use cases**:
97
+
98
+ The following notebooks contributed by *Cohere For AI Community* members show how Aya Expanse can be used for different use cases:
99
+
100
+ - [Mulitlingual Writing Assistant](https://colab.research.google.com/drive/1SRLWQ0HdYN_NbRMVVUHTDXb-LSMZWF60#scrollTo=qBK1H7WO9UHG)
101
+ - [AyaMCooking](https://colab.research.google.com/drive/1-cnn4LXYoZ4ARBpnsjQM3sU7egOL_fLB?usp=sharing#scrollTo=ukHwdlrgXSdI)
102
+ - [Multilingual Question-Answering System](https://colab.research.google.com/drive/1bbB8hzyzCJbfMVjsZPeh4yNEALJFGNQy?usp=sharing)
103
+
104
+
105
+ ## Model Details
106
+
107
+ **Input**: Models input text only.
108
+
109
+ **Output**: Models generate text only.
110
+
111
+ **Model Architecture**: Aya Expanse 8B is an auto-regressive language model that uses an optimized transformer architecture. Post-training includes supervised finetuning, preference training, and model merging.
112
+
113
+ **Languages covered**: The model is particularly optimized for multilinguality and supports the following languages: Arabic, Chinese (simplified & traditional), Czech, Dutch, English, French, German, Greek, Hebrew, Hindi, Indonesian, Italian, Japanese, Korean, Persian, Polish, Portuguese, Romanian, Russian, Spanish, Turkish, Ukrainian, and Vietnamese
114
+
115
+ **Context length**: 8K
116
+
117
+ ### Evaluation
118
+
119
+ <img src="winrates_marenahard.png" width="650" style="margin-left:'auto' margin-right:'auto' display:'block'"/>
120
+ <img src="winrates_by_lang.png" width="650" style="margin-left:'auto' margin-right:'auto' display:'block'"/>
121
+ <img src="winrates_step_by_step.png" width="650" style="margin-left:'auto' margin-right:'auto' display:'block'"/>
122
+
123
+ ### Model Card Contact
124
+
125
+ For errors or additional questions about details in this model card, contact info@for.ai.
126
+
127
+ ### Terms of Use
128
+
129
+ We hope that the release of this model will make community-based research efforts more accessible, by releasing the weights of a highly performant multilingual model to researchers all over the world. This model is governed by a [CC-BY-NC](https://cohere.com/c4ai-cc-by-nc-license) License with an acceptable use addendum, and also requires adhering to [C4AI's Acceptable Use Policy](https://docs.cohere.com/docs/c4ai-acceptable-use-policy).
130
+
131
+ ### Try the model today
132
+
133
+ You can try Aya Expanse in the Cohere [playground](https://dashboard.cohere.com/playground/chat) here. You can also use it in our dedicated Hugging Face Space [here](https://huggingface.co/spaces/CohereForAI/aya_expanse).