Text Generation
Transformers
Safetensors
English
llama
climate
conversational
text-generation-inference
Inference Endpoints
dthulke commited on
Commit
d47dab5
1 Parent(s): 78ebcb4

add LICENSE and update README

Browse files
Files changed (2) hide show
  1. LICENSE.txt +118 -0
  2. README.md +22 -17
LICENSE.txt ADDED
@@ -0,0 +1,118 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ TERMS OF SERVICE - CLIMATEGPT
2
+ Non-Commercial Licensing Agreement
3
+
4
+ This Agreement ("Agreement") between you and EQTY R&D, Inc. (the "Company" or "EQTY R&D") governs the use, reproduction, distribution and modification of any ClimateGPT Materials, as set forth below. By clicking "I Accept" below or by using or distributing any portion or element of the ClimateGPT Materials, you agree to be bound by this Agreement.
5
+
6
+ 1. License Rights and Redistribution.
7
+
8
+ a. Grant of Rights. You are granted a non-exclusive, worldwide, non-transferable, and royalty-free limited license under Company’s intellectual property or other rights owned by Company embodied in the ClimateGPT Materials to use, reproduce, distribute, copy, create derivative works of, and make modifications to the ClimateGPT Materials, for non-commercial purposes only.
9
+
10
+ b. Redistribution and Use.
11
+
12
+ i. If you distribute or make the ClimateGPT Materials, or any derivative works thereof, available to a third party, you shall provide a copy of this Agreement to such third party.
13
+
14
+ ii. You must retain in all copies of the ClimateGPT Materials that you distribute the following attribution notice within a "Notice" text file distributed as a part of such copies: "ClimateGPT is licensed under the ClimateGPT Community License, Copyright © EQTY R&D, Inc. All Rights Reserved."
15
+
16
+ iii. Your use of the ClimateGPT Materials must comply with applicable laws and regulations (including trade compliance laws and regulations) and adhere to the Acceptable Use Policy set forth below.
17
+
18
+ iv. You will not use the ClimateGPT Materials or any output or results of the ClimateGPT Materials to improve any other large language model (excluding ClimateGPT or derivative works thereof).
19
+
20
+ v. You will not reverse engineer, decompile, decode, disassemble, engage in model extraction or stealing attacks, prompt injection attacks or otherwise attempt to derive or gain improper access to any software components, models, algorithms or systems of any models contained in the ClimateGPT materials, in whole or in part; and
21
+
22
+ vi. You will not utilize any output of ClimateGPT ("Output") or sell, share or disclose Output to any third parties to train an AI model or have an AI model trained (e.g., engaging in “model scraping”).
23
+
24
+ vi. You will not use any aspect of the ClimateGPT Materials (including data or models), or help anyone else do so, other than as expressly permitted under this Agreement.
25
+
26
+ c. Third-party Beneficiary. Each party acknowledges and agrees that Erasmus.AI, Inc. ("Erasmus") is a third-party beneficiary of the representations, warranties and covenants of this Agreement, and that Erasmus is otherwise an express third party beneficiary of this Agreement, entitled to enforce the terms hereof as if it were an original party hereto.
27
+
28
+ 2. Definitions. "Documentation" means the specifications, manuals and documentation accompanying ClimateGPT distributed by Company at www.eci.io and or www.huggingface.co/eci-io
29
+
30
+ "Licensee" or "you" means you, or your employer or any other person or entity (if you are entering into this Agreement on such person or entity’s behalf), of the age required under applicable laws, rules or regulations to provide legal consent and that has legal authority to bind your employer or such other person or entity if you are entering in this Agreement on their behalf.
31
+
32
+ ClimateGPT means the foundational large language models and software and algorithms, including machine-learning model code, trained model weights, inference-enabling code, training-enabling code, fine-tuning enabling code, and other elements of the foregoing distributed by EQTY R&D, Inc. at www.eci.io and or www.huggingface.co/eci-io.
33
+
34
+ “ClimateGPT Materials” means, collectively, Company’s proprietary ClimateGPT and Documentation (and any portion thereof) made available under this Agreement.
35
+
36
+ 3. Disclaimer of Warranty. UNLESS REQUIRED BY APPLICABLE LAW, THE CLIMATEGPT MATERIALS AND ANY OUTPUT AND RESULTS THEREFROM ARE PROVIDED ON AN “AS IS” BASIS, WITHOUT WARRANTIES OF ANY KIND, EITHER EXPRESS OR IMPLIED, INCLUDING, WITHOUT LIMITATION, ANY WARRANTIES OF TITLE, NON-INFRINGEMENT, MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE. YOU ARE SOLELY RESPONSIBLE FOR DETERMINING THE APPROPRIATENESS OF USING OR REDISTRIBUTING THE CLIMATEGPT MATERIALS AND ASSUME ANY RISKS ASSOCIATED WITH YOUR USE OF THE CLIMATEGPT MATERIALS AND ANY OUTPUT AND RESULTS.
37
+
38
+ 4. Limitation of Liability. IN NO EVENT WILL COMPANY OR ITS AFFILIATES BE LIABLE UNDER ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, TORT, NEGLIGENCE, PRODUCTS LIABILITY, OR OTHERWISE, ARISING OUT OF THIS AGREEMENT, FOR ANY LOST PROFITS OR ANY INDIRECT, SPECIAL, CONSEQUENTIAL, INCIDENTAL, EXEMPLARY OR PUNITIVE DAMAGES, EVEN IF COMPANY OR ITS AFFILIATES HAVE BEEN ADVISED OF THE POSSIBILITY OF ANY OF THE FOREGOING.
39
+
40
+ 5. Intellectual Property.
41
+
42
+ a. No trademark licenses are granted under this Agreement, and in connection with the ClimateGPT Materials, neither Company nor Licensee may use any name or mark owned by or associated with the other or any of its affiliates, except as required for reasonable and customary use in describing and redistributing the ClimateGPT Materials.
43
+
44
+ b. Subject to Company’s ownership of ClimateGPT Materials and derivatives made by or for Company, with respect to any derivative works and modifications of the ClimateGPT Materials that are made by you, as between you and Company, you are and will be the owner of such derivative works and modifications.
45
+
46
+ c. If you institute litigation or other proceedings against Company or any entity (including a cross-claim or counterclaim in a lawsuit) alleging that the ClimateGPT Materials or ClimateGPT outputs or results, or any portion of any of the foregoing, constitutes an infringement of intellectual property or other rights owned or licensable by you, then any licenses granted to you under this Agreement shall terminate as of the date such litigation or claim is filed or instituted. You will indemnify and hold harmless Company from and against any claim by any third party arising out of or related to your use or distribution of the ClimateGPT Materials.
47
+
48
+ 6. Term and Termination. The term of this Agreement will commence upon your acceptance of this Agreement or access to the ClimateGPT Materials and will continue in full force and effect until terminated in accordance with the terms and conditions herein. Company may terminate this Agreement if you are in breach of any term or condition of this Agreement. Upon termination of this Agreement, you shall delete and cease use of the ClimateGPT Materials. Sections 3, 4 and 7 shall survive the termination of this Agreement.
49
+
50
+ 7. Governing Law and Jurisdiction. This Agreement will be governed and construed under the laws of the State of California without regard to choice of law principles, and the UN Convention on Contracts for the International Sale of Goods does not apply to this Agreement. The courts of California shall have exclusive jurisdiction of any dispute arising out of this Agreement.
51
+
52
+ 8. Acceptable Use Policy.
53
+
54
+ Company is committed to promoting safe and fair use of its technology, including ClimateGPT. If you access or use ClimateGPT, you agree to this Acceptable Use Policy ("Policy").
55
+
56
+ Prohibited Uses
57
+
58
+ We want everyone to use ClimateGPT safely and responsibly. You agree you will not use, or allow others to use, ClimateGPT to:
59
+
60
+ 1. Violate the law or others’ rights, including to:
61
+
62
+ a. Engage in, promote, generate, contribute to, encourage, plan, incite, or further illegal or unlawful activity or content, such as:
63
+
64
+ i. Violence or terrorism
65
+
66
+ ii. Exploitation or harm to children, including the solicitation, creation, acquisition, or dissemination of child exploitative content or failure to report Child Sexual Abuse Material
67
+
68
+ b. Human trafficking, exploitation, and sexual violence
69
+
70
+ iii. The illegal distribution of information or materials to minors, including obscene materials, or failure to employ legally required age-gating in connection with such information or materials.
71
+
72
+ iv. Sexual solicitation
73
+
74
+ vi. Any other criminal activity
75
+
76
+ c. Engage in, promote, incite, or facilitate the harassment, abuse, threatening, or bullying of individuals or groups of individuals
77
+
78
+ d. Engage in, promote, incite, or facilitate discrimination or other unlawful or harmful conduct in the provision of employment, employment benefits, credit, housing, other economic benefits, or other essential goods and services
79
+
80
+ e. Engage in the unauthorized or unlicensed practice of any profession including, but not limited to, financial, legal, medical/health, or related professional practices
81
+
82
+ f. Collect, process, disclose, generate, or infer health, demographic, or other sensitive personal or private information about individuals without rights and consents required by applicable laws
83
+
84
+ g. Engage in or facilitate any action or generate any content that infringes, misappropriates, or otherwise violates any third-party rights, including the outputs or results of any products or services using the ClimateGPT Materials
85
+
86
+ h. Create, generate, or facilitate the creation of malicious code, malware, computer viruses or do anything else that could disable, overburden, interfere with or impair the proper working, integrity, operation or appearance of a website or computer system
87
+
88
+ 2. Engage in, promote, incite, facilitate, or assist in the planning or development of activities that present a risk of death or bodily harm to individuals, including use of ClimateGPT related to the following:
89
+
90
+ a. Military, warfare, nuclear industries or applications, espionage, use for materials or activities that are subject to the International Traffic Arms Regulations (ITAR) maintained by the United States Department of State
91
+
92
+ b. Guns and illegal weapons (including weapon development)
93
+
94
+ c. Illegal drugs and regulated/controlled substances
95
+
96
+ d. Operation of critical infrastructure, transportation technologies, or heavy machinery
97
+
98
+ e. Self-harm or harm to others, including suicide, cutting, and eating disorders
99
+
100
+ f. Any content intended to incite or promote violence, abuse, or any infliction of bodily harm to an individual
101
+
102
+ 3. Intentionally deceive or mislead others, including use of ClimateGPT related to the following:
103
+
104
+ a. Generating, promoting, or furthering fraud or the creation or promotion of disinformation
105
+
106
+ b. Generating, promoting, or furthering defamatory content, including the creation of defamatory statements, images, or other content
107
+
108
+ c. Generating, promoting, or further distributing spam
109
+
110
+ d. Impersonating another individual without consent, authorization, or legal right
111
+
112
+ e. Representing that the use of ClimateGPT or outputs are human-generated
113
+
114
+ f. Generating or facilitating false online engagement, including fake reviews and other means of fake online engagement
115
+
116
+ 4. Fail to appropriately disclose to end users any known dangers of your AI system
117
+
118
+ Please report any violation of this Policy, harmful content generated by the model, software bugs, or other problems that could lead to a violation of this Agreement to info@eci.io
README.md CHANGED
@@ -13,43 +13,39 @@ co2_eq_emissions:
13
  geographical_location: "Washington, USA"
14
  hardware_used: "8x NVIDIA H100 HBM"
15
  ---
16
- # ClimateGPT 7B FSC
17
 
18
  <blockquote style="padding: 10px; margin: 0 0 10px; border-left: 5px solid #ddd;">
19
- ⚠️ This is a research experiment to explore training from scratch on climate related data. If you're just interested in using the model, we recommend to use the Llama 2 based [ClimateGPT 7B](https://huggingface.co/eci-io/climategpt-7b).
20
  </blockquote>
21
 
22
- ClimateGPT is an ensemble of AI models designed to augment human decisions on the fast-moving field of climate change.
23
- ClimateGPT 7B FSB (from scratch climate) is a 7 billion transformer decoder model that was pre-trained for 319.5B tokens including a collection of 4.2B tokens from curated climate documents.
24
  The model is further instruction fine-tuned on a dataset of instruction-completion pairs manually collected by AppTek in cooperation with climate scientists.
25
- [ClimateGPT 7B](https://huggingface.co/eci-io/climategpt-7b) outperforms Llama 2 70B Chat on our climate-specific benchmarks.
26
  The model is designed to be used together with retrieval augmentation to extend the knowledge, and increase the factuality of the model and with cascaded machine translation to increase the language coverage.
27
 
28
- <blockquote style="padding: 10px; margin: 0 0 10px; border-left: 5px solid #ddd;">
29
- A paper describing our approach will be released soon.
30
- </blockquote>
31
-
32
  ## Model Details
33
  - **Trained by:** [AppTek](https://apptek.com)
34
  - **Powered by:** [Erasmus AI](https://erasmus.ai)
35
  - **Verified by:** [EQTYLab](https://eqtylab.io)
36
  - **Model type:** decoder-only Transformer
37
  - **Language(s) (NLP):** English
38
- - **License:** TO BE ADDED
39
- - **Continued pre-trained from:** Llama 2 7B
40
  - **Context length:** 4K tokens
41
  - **Input:** Text-only data
42
  - **Output:** Model generates text only
43
- - **Paper:** The paper will be released soon.
44
  - **Website:** [eci.io](https://eci.io)
45
 
46
  ## Uses
47
- - This is an experimental model and it is only intended to be used to reproduce our results and for LLM research. For any other use-case, we recommend to use [ClimateGPT 7B](https://huggingface.co/eci-io/climategpt-7b), [13B](https://huggingface.co/eci-io/climategpt-13b) or [70B](https://huggingface.co/eci-io/climategpt-70b)
48
  - **Despite the efforts from the development team to eliminate them, as every other chat-capable LLMs, this model may generate biased, offensive or inaccurate responses.**
49
 
50
  ## Downstream Use
51
 
52
- ClimateGPT 7B FSC is an instruction-tuned model that can be directly used for climate-specific question-answering applications.
53
  It was trained to perform well with retrieval augmentation and supports up to 5 references in context.
54
 
55
  The model was trained using ChatML so the following format should be followed when prompting, including the `<|im_start|>`, `<|im_end|>` tags, `system`, `user`, `context` and `assistant` identifiers and `[[0]]`, `[[1]]]` etc. tokens to indicate references.
@@ -69,12 +65,12 @@ The model was trained using ChatML so the following format should be followed wh
69
  """
70
 
71
  ## Training
72
- - Details on the pre-training data with climate focus are given in our paper.
73
  - For instruction fine-tuning, about 272K instruction-completion pairs (both in the climate domain but also general domain) are used.
74
 
75
  ## Evaluation
76
 
77
- Detailed evaluation results are presented on our model card website: [eci.io/model-card](https://eci.io/model-card)
78
 
79
  ## Environmental Impact
80
  - **Hardware Type:** 8x NVIDIA H100 HBM
@@ -86,4 +82,13 @@ Detailed evaluation results are presented on our model card website: [eci.io/mod
86
  - **Carbon Emitted:** 262.8kg CO2eq
87
 
88
  ## Citation
89
- **BibTeX:** Paper will be released soon.
 
 
 
 
 
 
 
 
 
 
13
  geographical_location: "Washington, USA"
14
  hardware_used: "8x NVIDIA H100 HBM"
15
  ---
16
+ # ClimateGPT-7B-FSC
17
 
18
  <blockquote style="padding: 10px; margin: 0 0 10px; border-left: 5px solid #ddd;">
19
+ ⚠️ This is a research experiment to explore training from scratch on climate related data. If you're just interested in using the model, we recommend to use the Llama-2-based [ClimateGPT-7B](https://huggingface.co/eci-io/climategpt-7b).
20
  </blockquote>
21
 
22
+ ClimateGPT is a family of AI models designed to synthesize interdisciplinary research on climate change.
23
+ ClimateGPT-7B-FSB (from scratch climate) is a 7 billion parameter transformer decoder model that was pre-trained for 319.5B tokens including a collection of 4.2B tokens from curated climate documents.
24
  The model is further instruction fine-tuned on a dataset of instruction-completion pairs manually collected by AppTek in cooperation with climate scientists.
25
+ [ClimateGPT-7B](https://huggingface.co/eci-io/climategpt-7b) outperforms Llama-2-70B Chat on our climate-specific benchmarks.
26
  The model is designed to be used together with retrieval augmentation to extend the knowledge, and increase the factuality of the model and with cascaded machine translation to increase the language coverage.
27
 
 
 
 
 
28
  ## Model Details
29
  - **Trained by:** [AppTek](https://apptek.com)
30
  - **Powered by:** [Erasmus AI](https://erasmus.ai)
31
  - **Verified by:** [EQTYLab](https://eqtylab.io)
32
  - **Model type:** decoder-only Transformer
33
  - **Language(s) (NLP):** English
34
+ - **License:** ClimateGPT Community License
35
+ - **Continued pre-trained from:** Llama-2-7B
36
  - **Context length:** 4K tokens
37
  - **Input:** Text-only data
38
  - **Output:** Model generates text only
39
+ - **Paper:** [Download](https://shareddatastgacct.blob.core.windows.net/shared-data/climategpt-v1-publication.pdf)
40
  - **Website:** [eci.io](https://eci.io)
41
 
42
  ## Uses
43
+ - This is an experimental model and it is only intended to be used to reproduce our results and for LLM research. For any other use-case, we recommend to use [ClimateGPT-7B](https://huggingface.co/eci-io/climategpt-7b), [13B](https://huggingface.co/eci-io/climategpt-13b) or [70B](https://huggingface.co/eci-io/climategpt-70b)
44
  - **Despite the efforts from the development team to eliminate them, as every other chat-capable LLMs, this model may generate biased, offensive or inaccurate responses.**
45
 
46
  ## Downstream Use
47
 
48
+ ClimateGPT-7B-FSC is an instruction-tuned model that can be directly used for climate-specific question-answering applications.
49
  It was trained to perform well with retrieval augmentation and supports up to 5 references in context.
50
 
51
  The model was trained using ChatML so the following format should be followed when prompting, including the `<|im_start|>`, `<|im_end|>` tags, `system`, `user`, `context` and `assistant` identifiers and `[[0]]`, `[[1]]]` etc. tokens to indicate references.
 
65
  """
66
 
67
  ## Training
68
+ - Details on the climate-specific pre-training data are given in our paper.
69
  - For instruction fine-tuning, about 272K instruction-completion pairs (both in the climate domain but also general domain) are used.
70
 
71
  ## Evaluation
72
 
73
+ Detailed evaluation results are presented in our [paper](https://shareddatastgacct.blob.core.windows.net/shared-data/climategpt-v1-publication.pdf) on our model card website: [eci.io/model-card](https://eci.io/model-card)
74
 
75
  ## Environmental Impact
76
  - **Hardware Type:** 8x NVIDIA H100 HBM
 
82
  - **Carbon Emitted:** 262.8kg CO2eq
83
 
84
  ## Citation
85
+
86
+ If you find ClimateGPT is useful in your work, please cite it with:
87
+
88
+ ```
89
+ @misc{thulke2024climategpt,
90
+ title={ClimateGPT: Towards AI Synthesizing Interdisciplinary Research on Climate Change},
91
+ author={David Thulke and Yingbo Gao and Petrus Pelser and Rein Brune and Rricha Jalota and Floris Fok and Michael Ramos and Ian van Wyk and Abdallah Nasir and Hayden Goldstein and Taylor Tragemann and Katie Nguyen and Ariana Fowler and Andrew Stanco and Jon Gabriel and Jordan Taylor and Dean Moro and Evgenii Tsymbalov and Juliette de Waal and Evgeny Matusov and Mudar Yaghi and Mohammad Shihadah and Hermann Ney and Christian Dugast and Jonathan Dotan and Daniel Erasmus},
92
+ year={2024},
93
+ }
94
+ ```