├── requirements.txt ├── assets ├── icon.png └── mistral_pro_performance.png ├── demo ├── requirements.txt ├── README.md └── app.py ├── .gitmodules ├── scripts └── block_expansion.py ├── README.md └── LICENSE /requirements.txt: -------------------------------------------------------------------------------- 1 | fire 2 | -------------------------------------------------------------------------------- /assets/icon.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TencentARC/LLaMA-Pro/HEAD/assets/icon.png -------------------------------------------------------------------------------- /assets/mistral_pro_performance.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/TencentARC/LLaMA-Pro/HEAD/assets/mistral_pro_performance.png -------------------------------------------------------------------------------- /demo/requirements.txt: -------------------------------------------------------------------------------- 1 | accelerate 2 | bitsandbytes 3 | gradio 4 | protobuf 5 | scipy 6 | sentencepiece 7 | torch 8 | git+https://github.com/huggingface/transformers@main -------------------------------------------------------------------------------- /.gitmodules: -------------------------------------------------------------------------------- 1 | [submodule "open-instruct"] 2 | path = open-instruct 3 | url = git@github.com:hills-code/open-instruct.git 4 | [submodule "lm-evaluation-harness"] 5 | path = lm-evaluation-harness 6 | url = https://github.com/hills-code/lm-evaluation-harness.git 7 | -------------------------------------------------------------------------------- /demo/README.md: -------------------------------------------------------------------------------- 1 | # Run Gradio demo locally 2 | 3 | Here we release the code of how to run the Gradio demo locally in [app.py](./app.py) 4 | 5 | The dependency is described in [requirements.txt](./requirements.txt). 6 | 7 | You can use the following command to launch the demo in your localhost. 8 | ```bash 9 | python app.py 10 | ``` -------------------------------------------------------------------------------- /scripts/block_expansion.py: -------------------------------------------------------------------------------- 1 | 2 | import argparse 3 | from transformers import AutoModelForCausalLM 4 | import torch 5 | 6 | def main(): 7 | # Set up the argument parser 8 | parser = argparse.ArgumentParser(description="Receive deepen model's args") 9 | parser.add_argument("--model_path", default='meta-llama/Llama-2-7b-hf', type=str, help="original model path") 10 | parser.add_argument("--output_path", default='pytorch_model.bin', type=str, help="deepened model ckpt save path") 11 | parser.add_argument("--original_layers", default=32, type=int, help="original model num layers") 12 | parser.add_argument("--layers", default=40, type=int, help="deepen model num layers") 13 | 14 | # Parse the arguments 15 | args = parser.parse_args() 16 | 17 | model = AutoModelForCausalLM.from_pretrained(args.model_path, torch_dtype=torch.float16) 18 | ckpt = model.state_dict() 19 | 20 | split = int(args.original_layers / (args.layers - args.original_layers)) 21 | layer_cnt = 0 22 | 23 | output = {} 24 | for i in range(args.original_layers): 25 | for k in ckpt: 26 | if ('layers.' + str(i) + '.') in k: 27 | output[k.replace(('layers.' + str(i) + '.'), ('layers.' + str(layer_cnt) + '.'))] = ckpt[k] 28 | layer_cnt += 1 29 | if (i+1) % split == 0: 30 | for k in ckpt: 31 | if ('layers.' + str(i) + '.') in k: 32 | if 'down_proj' in k or 'o_proj' in k: 33 | output[k.replace(('layers.' + str(i) + '.'), ('layers.' + str(layer_cnt) + '.'))] = torch.zeros_like(ckpt[k]) 34 | else: 35 | output[k.replace(('layers.' + str(i) + '.'), ('layers.' + str(layer_cnt) + '.'))] = ckpt[k] 36 | 37 | 38 | layer_cnt += 1 39 | 40 | assert layer_cnt==args.layers 41 | for k in ckpt: 42 | if not 'layers' in k: 43 | output[k] = ckpt[k] 44 | 45 | torch.save(output, args.output_path) 46 | 47 | if __name__ == "__main__": 48 | main() 49 | 50 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # 图片名称 LLaMA Pro: Progressive LLaMA with Block Expansion 2 |

3 | 📃 Paper • 🤗 Demo & Model 4 |

5 | 6 | ## News 7 | * [2024/01/06] We open source the [LLaMA-Pro repository](https://github.com/TencentARC/LLaMA-Pro) and [Demo & Model](https://huggingface.co/TencentARC/LLaMA-Pro-8B). 8 | * [2024/01/07] Add how to run gradio demo locally in [demo](./demo/app.py) 9 | * [2024/01/18] Add the training code in [open-instruct](https://github.com/hills-code/open-instruct/tree/llama-pro). 10 | * [2024/02/23] We release the [Mistral-Pro-8B-v0.1](https://huggingface.co/TencentARC/Mistral_Pro_8B_v0.1) with superior performance on a range of benchmarks. It enhances the code and math performance of Mistral and matches the performance of the recently dominant model, [Gemma](https://huggingface.co/google/gemma-7b). 11 | ![assets/mistral_pro_performance.png](assets/mistral_pro_performance.png) 12 | * [2024/02/23] We release the evaluation code of [Mistral-Pro-8B-v0.1](https://huggingface.co/TencentARC/Mistral_Pro_8B_v0.1) in [lm-evaluation-harness](https://github.com/hills-code/lm-evaluation-harness). 13 | * [2024/02/23] We release [MetaMath-Mistral-Pro](https://huggingface.co/TencentARC/MetaMath-Mistral-Pro) that surpasses previous MetaMath series 7B models at both GSM8k and MATH. The evaluation is following [the official MetaMath repo](https://github.com/meta-math/MetaMath). 14 | * [2024/05/08] Add the pre-train example script for cosmopedia in [open-instruct](https://github.com/hills-code/open-instruct/tree/llama-pro). 15 | * [2024/05/16] [LLaMA Pro](https://arxiv.org/abs/2401.02415) has been accepted to the main conference of ACL 2024! 16 | 17 | 18 | 🔥 Comprehensive Results 19 | 20 | | Model | GSM8k Pass@1 | MATH Pass@1 | 21 | |---------------------|--------------|-------------| 22 | | WizardMath-7B | 54.9 | 10.7 | 23 | | LLaMA-2-70B | 56.8 | 13.5 | 24 | | WizardMath-13B | 63.9 | 14.0 | 25 | | MetaMath-7B | 66.5 | 19.8 | 26 | | MetaMath-13B | 72.3 | 22.4 | 27 | | MetaMath-Mistral-7B | 77.7 | 28.2 | 28 | | MetaMath-Llemma-7B | 69.2 | 30.0 | 29 | | 🔥 **MetaMath-Mistral-Pro** | **78.4** | **30.3** | 30 | 31 | ## Acknowledgement 32 | The code of instruction tuning is based on the official implementation of [open-instruct](https://github.com/allenai/open-instruct). 33 | 34 | Thanks [huggingface](https://huggingface.co/TencentARC/LLaMA-Pro-8B) & [wisemodel](https://wisemodel.cn/models/TencentARC/LLaMA-Pro-8B) for hosting our checkpoint. 35 | 36 | ## Citation 37 | The code and model in this repository is mostly developed for or derived from the paper below. Please cite it if you find the repository helpful. 38 | ``` 39 | @article{wu2024llama, 40 | title={Llama pro: Progressive llama with block expansion}, 41 | author={Wu, Chengyue and Gan, Yukang and Ge, Yixiao and Lu, Zeyu and Wang, Jiahao and Feng, Ye and Luo, Ping and Shan, Ying}, 42 | journal={arXiv preprint arXiv:2401.02415}, 43 | year={2024} 44 | } 45 | ``` 46 | -------------------------------------------------------------------------------- /demo/app.py: -------------------------------------------------------------------------------- 1 | import gradio as gr 2 | import torch 3 | import sys 4 | import html 5 | from transformers import AutoTokenizer, AutoModelForCausalLM, TextIteratorStreamer 6 | from threading import Thread 7 | 8 | model_name_or_path = 'TencentARC/LLaMA-Pro-8B-Instruct' 9 | 10 | tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, use_fast=False) 11 | model = AutoModelForCausalLM.from_pretrained(model_name_or_path) 12 | 13 | model.half().cuda() 14 | 15 | def convert_message(message): 16 | message_text = "" 17 | if message["content"] is None and message["role"] == "assistant": 18 | message_text += "<|assistant|>\n" # final msg 19 | elif message["role"] == "system": 20 | message_text += "<|system|>\n" + message["content"].strip() + "\n" 21 | elif message["role"] == "user": 22 | message_text += "<|user|>\n" + message["content"].strip() + "\n" 23 | elif message["role"] == "assistant": 24 | message_text += "<|assistant|>\n" + message["content"].strip() + "\n" 25 | else: 26 | raise ValueError("Invalid role: {}".format(message["role"])) 27 | # gradio cleaning - it converts stuff to html entities 28 | # we would need special handling for where we want to keep the html... 29 | message_text = html.unescape(message_text) 30 | # it also converts newlines to
, undo this. 31 | message_text = message_text.replace("
", "\n") 32 | return message_text 33 | 34 | def convert_history(chat_history, max_input_length=1024): 35 | history_text = "" 36 | idx = len(chat_history) - 1 37 | # add messages in reverse order until we hit max_input_length 38 | while len(tokenizer(history_text).input_ids) < max_input_length and idx >= 0: 39 | user_message, chatbot_message = chat_history[idx] 40 | user_message = convert_message({"role": "user", "content": user_message}) 41 | chatbot_message = convert_message({"role": "assistant", "content": chatbot_message}) 42 | history_text = user_message + chatbot_message + history_text 43 | idx = idx - 1 44 | # if nothing was added, add <|assistant|> to start generation. 45 | if history_text == "": 46 | history_text = "<|assistant|>\n" 47 | return history_text 48 | 49 | @torch.inference_mode() 50 | def instruct(instruction, max_token_output=1024): 51 | input_text = instruction 52 | streamer = TextIteratorStreamer(tokenizer, skip_prompt=True) 53 | input_ids = tokenizer(input_text, return_tensors='pt', truncation=False) 54 | input_ids["input_ids"] = input_ids["input_ids"].cuda() 55 | input_ids["attention_mask"] = input_ids["attention_mask"].cuda() 56 | generation_kwargs = dict(input_ids, streamer=streamer, max_new_tokens=max_token_output, do_sample=False) 57 | thread = Thread(target=model.generate, kwargs=generation_kwargs) 58 | thread.start() 59 | return streamer 60 | 61 | 62 | with gr.Blocks() as demo: 63 | # chatbot-style model 64 | with gr.Tab("Chatbot"): 65 | chatbot = gr.Chatbot([], elem_id="chatbot") 66 | msg = gr.Textbox() 67 | clear = gr.Button("Clear") 68 | # fn to add user message to history 69 | def user(user_message, history): 70 | return "", history + [[user_message, None]] 71 | 72 | def bot(history): 73 | prompt = convert_history(history) 74 | streaming_out = instruct(prompt) 75 | history[-1][1] = "" 76 | for new_token in streaming_out: 77 | history[-1][1] += new_token 78 | yield history 79 | 80 | msg.submit(user, [msg, chatbot], [msg, chatbot], queue=False).then( 81 | bot, chatbot, chatbot 82 | ) 83 | 84 | clear.click(lambda: None, None, chatbot, queue=False) 85 | 86 | if __name__ == "__main__": 87 | demo.queue().launch(share=True) -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | Apache License 2 | Version 2.0, January 2004 3 | http://www.apache.org/licenses/ 4 | 5 | TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION 6 | 7 | 1. Definitions. 8 | 9 | "License" shall mean the terms and conditions for use, reproduction, 10 | and distribution as defined by Sections 1 through 9 of this document. 11 | 12 | "Licensor" shall mean the copyright owner or entity authorized by 13 | the copyright owner that is granting the License. 14 | 15 | "Legal Entity" shall mean the union of the acting entity and all 16 | other entities that control, are controlled by, or are under common 17 | control with that entity. For the purposes of this definition, 18 | "control" means (i) the power, direct or indirect, to cause the 19 | direction or management of such entity, whether by contract or 20 | otherwise, or (ii) ownership of fifty percent (50%) or more of the 21 | outstanding shares, or (iii) beneficial ownership of such entity. 22 | 23 | "You" (or "Your") shall mean an individual or Legal Entity 24 | exercising permissions granted by this License. 25 | 26 | "Source" form shall mean the preferred form for making modifications, 27 | including but not limited to software source code, documentation 28 | source, and configuration files. 29 | 30 | "Object" form shall mean any form resulting from mechanical 31 | transformation or translation of a Source form, including but 32 | not limited to compiled object code, generated documentation, 33 | and conversions to other media types. 34 | 35 | "Work" shall mean the work of authorship, whether in Source or 36 | Object form, made available under the License, as indicated by a 37 | copyright notice that is included in or attached to the work 38 | (an example is provided in the Appendix below). 39 | 40 | "Derivative Works" shall mean any work, whether in Source or Object 41 | form, that is based on (or derived from) the Work and for which the 42 | editorial revisions, annotations, elaborations, or other modifications 43 | represent, as a whole, an original work of authorship. For the purposes 44 | of this License, Derivative Works shall not include works that remain 45 | separable from, or merely link (or bind by name) to the interfaces of, 46 | the Work and Derivative Works thereof. 47 | 48 | "Contribution" shall mean any work of authorship, including 49 | the original version of the Work and any modifications or additions 50 | to that Work or Derivative Works thereof, that is intentionally 51 | submitted to Licensor for inclusion in the Work by the copyright owner 52 | or by an individual or Legal Entity authorized to submit on behalf of 53 | the copyright owner. For the purposes of this definition, "submitted" 54 | means any form of electronic, verbal, or written communication sent 55 | to the Licensor or its representatives, including but not limited to 56 | communication on electronic mailing lists, source code control systems, 57 | and issue tracking systems that are managed by, or on behalf of, the 58 | Licensor for the purpose of discussing and improving the Work, but 59 | excluding communication that is conspicuously marked or otherwise 60 | designated in writing by the copyright owner as "Not a Contribution." 61 | 62 | "Contributor" shall mean Licensor and any individual or Legal Entity 63 | on behalf of whom a Contribution has been received by Licensor and 64 | subsequently incorporated within the Work. 65 | 66 | 2. Grant of Copyright License. Subject to the terms and conditions of 67 | this License, each Contributor hereby grants to You a perpetual, 68 | worldwide, non-exclusive, no-charge, royalty-free, irrevocable 69 | copyright license to reproduce, prepare Derivative Works of, 70 | publicly display, publicly perform, sublicense, and distribute the 71 | Work and such Derivative Works in Source or Object form. 72 | 73 | 3. Grant of Patent License. Subject to the terms and conditions of 74 | this License, each Contributor hereby grants to You a perpetual, 75 | worldwide, non-exclusive, no-charge, royalty-free, irrevocable 76 | (except as stated in this section) patent license to make, have made, 77 | use, offer to sell, sell, import, and otherwise transfer the Work, 78 | where such license applies only to those patent claims licensable 79 | by such Contributor that are necessarily infringed by their 80 | Contribution(s) alone or by combination of their Contribution(s) 81 | with the Work to which such Contribution(s) was submitted. If You 82 | institute patent litigation against any entity (including a 83 | cross-claim or counterclaim in a lawsuit) alleging that the Work 84 | or a Contribution incorporated within the Work constitutes direct 85 | or contributory patent infringement, then any patent licenses 86 | granted to You under this License for that Work shall terminate 87 | as of the date such litigation is filed. 88 | 89 | 4. Redistribution. You may reproduce and distribute copies of the 90 | Work or Derivative Works thereof in any medium, with or without 91 | modifications, and in Source or Object form, provided that You 92 | meet the following conditions: 93 | 94 | (a) You must give any other recipients of the Work or 95 | Derivative Works a copy of this License; and 96 | 97 | (b) You must cause any modified files to carry prominent notices 98 | stating that You changed the files; and 99 | 100 | (c) You must retain, in the Source form of any Derivative Works 101 | that You distribute, all copyright, patent, trademark, and 102 | attribution notices from the Source form of the Work, 103 | excluding those notices that do not pertain to any part of 104 | the Derivative Works; and 105 | 106 | (d) If the Work includes a "NOTICE" text file as part of its 107 | distribution, then any Derivative Works that You distribute must 108 | include a readable copy of the attribution notices contained 109 | within such NOTICE file, excluding those notices that do not 110 | pertain to any part of the Derivative Works, in at least one 111 | of the following places: within a NOTICE text file distributed 112 | as part of the Derivative Works; within the Source form or 113 | documentation, if provided along with the Derivative Works; or, 114 | within a display generated by the Derivative Works, if and 115 | wherever such third-party notices normally appear. The contents 116 | of the NOTICE file are for informational purposes only and 117 | do not modify the License. You may add Your own attribution 118 | notices within Derivative Works that You distribute, alongside 119 | or as an addendum to the NOTICE text from the Work, provided 120 | that such additional attribution notices cannot be construed 121 | as modifying the License. 122 | 123 | You may add Your own copyright statement to Your modifications and 124 | may provide additional or different license terms and conditions 125 | for use, reproduction, or distribution of Your modifications, or 126 | for any such Derivative Works as a whole, provided Your use, 127 | reproduction, and distribution of the Work otherwise complies with 128 | the conditions stated in this License. 129 | 130 | 5. Submission of Contributions. Unless You explicitly state otherwise, 131 | any Contribution intentionally submitted for inclusion in the Work 132 | by You to the Licensor shall be under the terms and conditions of 133 | this License, without any additional terms or conditions. 134 | Notwithstanding the above, nothing herein shall supersede or modify 135 | the terms of any separate license agreement you may have executed 136 | with Licensor regarding such Contributions. 137 | 138 | 6. Trademarks. This License does not grant permission to use the trade 139 | names, trademarks, service marks, or product names of the Licensor, 140 | except as required for reasonable and customary use in describing the 141 | origin of the Work and reproducing the content of the NOTICE file. 142 | 143 | 7. Disclaimer of Warranty. Unless required by applicable law or 144 | agreed to in writing, Licensor provides the Work (and each 145 | Contributor provides its Contributions) on an "AS IS" BASIS, 146 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or 147 | implied, including, without limitation, any warranties or conditions 148 | of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A 149 | PARTICULAR PURPOSE. You are solely responsible for determining the 150 | appropriateness of using or redistributing the Work and assume any 151 | risks associated with Your exercise of permissions under this License. 152 | 153 | 8. Limitation of Liability. In no event and under no legal theory, 154 | whether in tort (including negligence), contract, or otherwise, 155 | unless required by applicable law (such as deliberate and grossly 156 | negligent acts) or agreed to in writing, shall any Contributor be 157 | liable to You for damages, including any direct, indirect, special, 158 | incidental, or consequential damages of any character arising as a 159 | result of this License or out of the use or inability to use the 160 | Work (including but not limited to damages for loss of goodwill, 161 | work stoppage, computer failure or malfunction, or any and all 162 | other commercial damages or losses), even if such Contributor 163 | has been advised of the possibility of such damages. 164 | 165 | 9. Accepting Warranty or Additional Liability. While redistributing 166 | the Work or Derivative Works thereof, You may choose to offer, 167 | and charge a fee for, acceptance of support, warranty, indemnity, 168 | or other liability obligations and/or rights consistent with this 169 | License. However, in accepting such obligations, You may act only 170 | on Your own behalf and on Your sole responsibility, not on behalf 171 | of any other Contributor, and only if You agree to indemnify, 172 | defend, and hold each Contributor harmless for any liability 173 | incurred by, or claims asserted against, such Contributor by reason 174 | of your accepting any such warranty or additional liability. 175 | 176 | END OF TERMS AND CONDITIONS 177 | 178 | APPENDIX: How to apply the Apache License to your work. 179 | 180 | To apply the Apache License to your work, attach the following 181 | boilerplate notice, with the fields enclosed by brackets "[]" 182 | replaced with your own identifying information. (Don't include 183 | the brackets!) The text should be enclosed in the appropriate 184 | comment syntax for the file format. We also recommend that a 185 | file or class name and description of purpose be included on the 186 | same "printed page" as the copyright notice for easier 187 | identification within third-party archives. 188 | 189 | Copyright [yyyy] [name of copyright owner] 190 | 191 | Licensed under the Apache License, Version 2.0 (the "License"); 192 | you may not use this file except in compliance with the License. 193 | You may obtain a copy of the License at 194 | 195 | http://www.apache.org/licenses/LICENSE-2.0 196 | 197 | Unless required by applicable law or agreed to in writing, software 198 | distributed under the License is distributed on an "AS IS" BASIS, 199 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 200 | See the License for the specific language governing permissions and 201 | limitations under the License. 202 | --------------------------------------------------------------------------------