Upload folder using huggingface_hub

.gitignore

@@ -0,0 +1,160 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+.pytest_cache/
+cover/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# poetry
+#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
+#poetry.lock
+
+# pdm
+#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
+#pdm.lock
+#   pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
+#   in version control.
+#   https://pdm.fming.dev/#use-with-ide
+.pdm.toml
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+# PyCharm
+#  JetBrains specific template is maintained in a separate JetBrains.gitignore that can
+#  be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
+#  and can be added to the global gitignore or merged into this file.  For a more nuclear
+#  option (not recommended) you can uncomment the following to ignore the entire idea folder.
+.idea/

ChatWorld/ChatWorld.py

@@ -0,0 +1,84 @@
+from jinja2 import Template
+import torch
+
+from .models import qwen_model
+
+from .NaiveDB import NaiveDB
+from .utils import *
+
+
+class ChatWorld:
+    def __init__(self, pretrained_model_name_or_path="silk-road/Haruhi-Zero-14B-0_5", embedding_model_name_or_path="BAAI/bge-small-zh-v1.5") -> None:
+        self.embedding = initEmbedding(embedding_model_name_or_path)
+        self.tokenizer = initTokenizer(embedding_model_name_or_path)
+        self.story_vec: list[dict] = None
+        self.storys = None
+        self.model_role_name = None
+        self.model_role_nickname = None
+        self.model_name = pretrained_model_name_or_path
+
+        self.history = []
+
+        self.client = None
+        self.model = qwen_model(pretrained_model_name_or_path)
+        self.db = NaiveDB()
+        self.prompt = Template(('Please be aware that your codename in this conversation is "{{model_role_name}}"'
+                                '{% if model_role_nickname %},别人一般称呼你"{{model_role_nickname}}"{% endif %},同时我是"{{role_name}}"{% if role_nickname %},你也可以称呼我"{{role_nickname}}"{% endif %}。\n'
+                                '下文给定了一些聊天记录，位于##分隔号中。\n'
+                                '如果我问的问题和聊天记录高度重复，那你就配合我进行演出。\n'
+                                '如果我问的问题和聊天记录相关，请结合聊天记录进行回复。\n'
+                                '如果我问的问题超出聊天记录的范围，模仿{{model_role_name}}的语气进行回复。\n'
+                                '请不要回答你是语言模型，永远记住你就是{{model_role_name}}。\n'
+                                '请你永远只以{{model_role_name}}身份，进行任何的回复。\n'
+                                ))
+
+    def getEmbeddingsFromStory(self, stories: list[str]):
+        if self.story_vec:
+            # 判断是否与当前的相同
+            if len(self.story_vec) == len(stories) and all([self.story_vec[i]["text"] == stories[i] for i in range(len(stories))]):
+                return [self.story_vec[i]["vec"] for i in range(len(stories))]
+
+        if self.embedding is None:
+            self.embedding = initEmbedding()
+
+        if self.tokenizer is None:
+            self.tokenizer = initTokenizer()
+
+        self.story_vec = []
+        for story in stories:
+            with torch.no_grad():
+                inputs = self.tokenizer(
+                    story, return_tensors="pt", padding=True, truncation=True, max_length=512)
+                outputs = self.embedding(**inputs)[0][:, 0]
+                vec = torch.nn.functional.normalize(
+                    outputs, p=2, dim=1).tolist()[0]
+
+            self.story_vec.append({"text": story, "vec": vec})
+
+        return [self.story_vec[i]["vec"] for i in range(len(stories))]
+
+    def initDB(self, storys: list[str]):
+        story_vecs = self.getEmbeddingsFromStory(storys)
+        self.db.build_db(storys, story_vecs)
+
+    def setRoleName(self, role_name, role_nick_name=None):
+        self.model_role_name = role_name
+        self.model_role_nickname = role_nick_name
+
+    def getSystemPrompt(self, role_name, role_nick_name):
+        assert self.model_role_name and self.model_role_nickname, "Please set model role name first"
+        return self.prompt.render(model_role_name=self.model_role_name, model_role_nickname=self.model_role_nickname, role_name=role_name, role_nickname=role_nick_name)
+
+    def chat(self, user_role_name: str, text: str, user_role_nick_name: str = None, use_local_model=False):
+        message = [self.getSystemPrompt(
+            user_role_name, user_role_nick_name)] + self.history
+
+        if use_local_model:
+            response = self.model.get_response(message)
+        else:
+            response = self.client.chat(
+                user_role_name, text, user_role_nick_name)
+
+        self.history.append({"role": "user", "content": text})
+        self.history.append({"role": "model", "content": response})
+        return response

ChatWorld/NaiveDB.py

@@ -0,0 +1,85 @@
+import random
+from math import sqrt
+
+
+class NaiveDB:
+    def __init__(self):
+        self.verbose = False
+        self.init_db()
+
+    def init_db(self):
+        if self.verbose:
+            print("call init_db")
+        self.stories = []
+        self.norms = []
+        self.vecs = []
+        self.flags = []  # 用于标记每个story是否可以被搜索
+        self.metas = []  # 用于存储每个story的meta信息
+        self.last_search_ids = []  # 用于存储上一次搜索的结果
+
+    def build_db(self, stories, vecs, flags=None, metas=None):
+        self.stories = stories
+        self.vecs = vecs
+        self.flags = flags if flags else [True for _ in self.stories]
+        self.metas = metas if metas else [{} for _ in self.stories]
+        self.recompute_norm()
+
+    def save(self, file_path):
+        print(
+            "warning! directly save folder from dbtype NaiveDB has not been implemented yet, try use role_from_hf to load role instead")
+
+    def load(self, file_path):
+        print(
+            "warning! directly load folder from dbtype NaiveDB has not been implemented yet, try use role_from_hf to load role instead")
+
+    def recompute_norm(self):
+        # 补全这部分代码，self.norms 分别存储每个vector的l2 norm
+        # 计算每个向量的L2范数
+        self.norms = [sqrt(sum([x ** 2 for x in vec])) for vec in self.vecs]
+
+    def get_stories_with_id(self, ids):
+        return [self.stories[i] for i in ids]
+
+    def clean_flag(self):
+        self.flags = [True for _ in self.stories]
+
+    def disable_story_with_ids(self, close_ids):
+        for id in close_ids:
+            self.flags[id] = False
+
+    def close_last_search(self):
+        for id in self.last_search_ids:
+            self.flags[id] = False
+
+    def search(self, query_vector, n_results):
+
+        if self.verbose:
+            print("call search")
+
+        if len(self.norms) != len(self.vecs):
+            self.recompute_norm()
+
+        # 计算查询向量的范数
+        query_norm = sqrt(sum([x ** 2 for x in query_vector]))
+
+        idxs = list(range(len(self.vecs)))
+
+        # 计算余弦相似度
+        similarities = []
+        for vec, norm, idx in zip(self.vecs, self.norms, idxs):
+            if len(self.flags) == len(self.vecs) and not self.flags[idx]:
+                continue
+
+            dot_product = sum(q * v for q, v in zip(query_vector, vec))
+            if query_norm < 1e-20:
+                similarities.append((random.random(), idx))
+                continue
+            cosine_similarity = dot_product / (query_norm * norm)
+            similarities.append((cosine_similarity, idx))
+
+        # 获取最相似的n_results个结果， 使用第0个字段进行排序
+        similarities.sort(key=lambda x: x[0], reverse=True)
+        self.last_search_ids = [x[1] for x in similarities[:n_results]]
+
+        top_indices = [x[1] for x in similarities[:n_results]]
+        return top_indices

ChatWorld/__init__.py

@@ -0,0 +1 @@
+from .ChatWorld import ChatWorld

ChatWorld/models.py

@@ -0,0 +1,11 @@
+from transformers import AutoTokenizer, AutoModelForCausalLM
+
+
+class qwen_model:
+    def __init__(self, model_name):
+        self.tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
+        self.model = AutoModelForCausalLM.from_pretrained(model_name, device_map="auto", trust_remote_code=True).eval()
+
+    def get_response(self, message):
+        self.tokenizer.apply_chat_template(message, tokenize=False, add_generation_prompt=True)
+        return "test"

ChatWorld/utils.py

@@ -0,0 +1,9 @@
+from transformers import AutoModel, AutoTokenizer
+
+
+def initEmbedding(model_name="BAAI/bge-small-zh-v1.5", **model_wargs):
+    return AutoModel.from_pretrained(model_name, **model_wargs)
+
+
+def initTokenizer(model_name="BAAI/bge-small-zh-v1.5", **model_wargs):
+    return AutoTokenizer.from_pretrained(model_name)

LICENSE

@@ -0,0 +1,201 @@
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright [yyyy] [name of copyright owner]
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

README.md

@@ -1,12 +1,420 @@
 ---
-title: Zero Haruhi
-emoji: 📊
-colorFrom: purple
-colorTo: pink
-sdk: gradio
-sdk_version: 4.19.2
+title: Zero-Haruhi
 app_file: app.py
-pinned: false
+sdk: gradio
+sdk_version: 3.50.2
 ---
+# Chat凉宫春日 Chat-Haruhi-Suzumiya
+## Reviving Anime Character in Reality via Large Language Model
+
+**Chat凉宫春日**是模仿凉宫春日等一系列动漫人物，使用近似语气、个性和剧情聊天的语言模型方案。
+
+随着[**凉宫春日-Zero**](https://github.com/LC1332/Zero-Haruhi)模型的建立，Chat凉宫春日项目将主要包含三个部分，
+
+- 可以接入不同语言模型，包括闭源商用API和Haruhi-Zero模型的推理class，用于管理人物的记忆库和过往的历史信息并进行prompt的构造。
+- 使用120k左右量级对话数据得到的开源角色扮演模型，凉宫春日-Zero，方便每个开发者构建自己的人物模型
+- 完整的小说抽取方案，以及一个开源可本地部署的对话-动作抽取模型。
+
+
+<p align="center">
+    <img src="https://github.com/LC1332/Chat-Haruhi-Suzumiya/blob/main/figures/datasetOverview.png">
+</p>
+
+<details>
+  <summary> (TBD) Chat凉宫春日的3.0版本和HaruhiZero模型由李鲁鲁, 豆角, 刘崇寒, 米唯实, BPSK, JiangYH等开发 </summary>
+
+开发人员持续增加中
+
+</details>
+
+<details>
+  <summary> Chat凉宫春日的2.0版本由李鲁鲁, 冷子昂, 闫晨曦, 封小洋, scixing, 沈骏一, Aria Fei, 王皓, 米唯实, 冷月, JunityZhan, 贾曜恺, 吴平宇, 孙浩甄等开发。 </summary>
+
+本项目是一个开源项目，项目成员均在DataWhale等开源社区招募。
+
+李鲁鲁( [Cheng Li@SenseTime](https://github.com/LC1332) )发起了整个项目,并设计和实现了项目的大多数功能。
+
+冷子昂( [Ziang Leng@SenseTime](https://blairleng.github.io) )设计和实现了整体的ChatHaruhi1.0的训练,数据生成和后端架构。
+
+闫晨曦( [Chenxi Yan@Chengdu University of Information Technology](https://github.com/todochenxi) )实现和维护了ChatHaruhi1.0版本的后端。
+
+沈骏一( [Junyi Shen@Zhejiang University](https://github.com/J1shen) )实现了训练代码,参与了训练数据集生成。
+
+王皓( [Hao Wang](https://github.com/wanghao07456) )收集了武林外传的台本数据,参与了增广数据的生成。
+
+米唯实( [Weishi MI@Tsinghua University](https://github.com/hhhwmws0117) )参与了增广数据生成。
+
+Yaying Fei( [Aria Fei@Beijing University of Technology](https://ariafyy.github.io/) )实现了台本工具 ASR 功能,参与了Openness-Aware Personality paper分支项目。
+
+封小洋( [Xiaoyang Feng@Nanjing Agricultural University](https://github.com/fengyunzaidushi) )整合了台本识别工具功能,参与了Openness-Aware Personality paper分支项目。
+
+冷月( [Song Yan](https://github.com/zealot52099) )收集了big bang thoery的数据。实现了台本格式转换功能。
+
+scixing(汪好盛)( [HaoSheng Wang](https://github.com/ssccinng) )实现了台本工具中声纹识别功能,以及tts-vits语音合成功能。
+
+Linkang Zhan( [JunityZhan@Case Western Reserve University](https://github.com/JunityZhan) ) 收集了原神的system prompt和故事数据。
+
+贾曜恺( [Yaokai Jia](https://github.com/KaiJiaBrother) )实现了Vue版本的前端,并且在心理项目中实践了Bert的GPU抽取。
+
+吴平宇( [Pingyu Wu@Juncai Shuyun](https://github.com/wpydcr) )帮助部署了第一版本的训练代码。
+
+孙浩甄( [Haozhen Sun@Tianjin University] )绘制了ChatHaruhi角色的拼图。
+
+</details>
+
+
+# TODO
+
+近期TODO 0215
+
+- Haruhi-Zero 0.5 的数据收集
+  - 0.4已经收集，50本小说，每本10个人物
+  - 每个人根据语料，进一步总结profile，验证profile的有效性，产生0.5的数据 ( @ 鲁叔) 
+- Haruhi-Zero 0.4 训练各个size的模型并上传到hf
+  - 14B qwen已经上传 (@BPSK)
+  - 1.8B还是需要的哈， JiangYH说要训练一个ChatGLM-6B的 (@BPSK @JiangYH)
+- 上传7B抽取模型 (@BPSK)
+- ChatHaruhi3.0接口
+  - 初步openai和GLM已经完成
+  - Le geant 去尝试搭建一些gradio 放到hf上面常驻 （@Le geant )
+  - 其他api ( @ mws )
+- 写模型的Readme( @ 鲁叔) 
+- Haruhi-Zero模型需要接入到oobabooga textUI ( @ JiangYH )
+- Evaluation
+  - 先和xintao讨论下咋搞
+- 阅读Silly Tavern代码，看看哪里方便接入记忆库 ( @ 多人 ) 
+- PIPPA数据翻译
+  - 有初步翻译的版本（未check） check后加入0.5训练集
+- 角色分享网站
+  - ？@Sibo
+
+# ChatHaruhi 3.0 的说明
+
+ChatHaruhi 3.0是Chat凉宫春日的推理库
+
+- 支持使用或者不使用RAG的角色的载入
+- 支持直接载入sugar角色，或者从hugging face载入预先设定好的角色
+- 支持用户自建角色
+- 在ChatHaruhi 2.0的基础上做了message和llm的分离，方便接入不同的模型
+
+## 快速开始
+
+目前ChatHaruhi的库本身只需要install datasets和tiktoken。另外你需要准备和你对应希望使用的llm所依赖的库，以openai为例的话需要安装openai��库。
+
+对于ChatHaruhi默认支持的角色，你可以直接使用sugar载入
+
+```python
+from ChatHaruhi import ChatHaruhi
+from ChatHaruhi.openai import get_openai_response
+
+chatbot = ChatHaruhi( role_name = 'haruhi', llm = get_openai_response )
+chatbot.add_rag_prompt_after_persona()
+
+response = chatbot.chat(user = '阿虚', text = '我看新一年的棒球比赛要开始了！我们要去参加吗？')
+```
+
+对于更完整的载入方式，需要设置persona, role_name和stories
+
+```python
+from ChatHaruhi import ChatHaruhi
+from ChatHaruhi.openai import get_openai_response
+
+persona = """每当用户问询一个家庭关系，输出一个相关的家庭关系的逻辑句子
+
+{{RAG对话}}
+{{RAG对话}}
+"""
+
+role_name = "家庭关系机器人"
+
+stories = ["{{user}}:爷爷 {{role}}:爸爸的爸爸是爷爷",
+"{{user}}:奶奶 {{role}}:爸爸的妈妈是奶奶",
+"{{user}}:外公 {{role}}:妈妈的爸爸是外公",
+"{{user}}:外婆 {{role}}:妈妈的妈妈是外婆"]
+
+chatbot = ChatHaruhi( role_name = role_name, persona = persona, llm = get_response ,\
+   llm = get_openai_response )
+# 因为persona已经带了"{{RAG对话}}"，不需要额外运行add_rag_prompt_after_persona()
+
+response = chatbot.chat(user = '', text = '表姐？')
+```
+
+当然，建立角色记忆库对于一般的用户来说是比较困难的，所以自Haruhi-Zero模型之后，我们也支持用户创建不使用记忆库的角色
+
+```python
+from ChatHaruhi import ChatHaruhi
+from ChatHaruhi.openai import get_openai_response
+
+role_name = "布莱恩"
+persona = "你扮演 德州杀场 中的 布莱恩 布莱恩是一个专注、果断、有责任感的警探，他在调查案件时非常注重细节，对案件的解决充满使命感。 布莱恩是一个专注、果断、有责任感的警探 布莱恩是一个身材魁梧、严肃的警探 这是一个警探调查案件的场景，布莱恩与其他警员合作调查案件"
+
+chatbot = ChatHaruhi( role_name = role_name, persona = persona, llm = get_openai_response )
+```
+
+更多载入方式见文档后面的部分
+
+## 使用不同的模型进行inference
+
+直接从对应的response_XX.py中载入对应的response函数即可
+
+### openai
+
+```python
+from ChatHaruhi.openai import get_openai_response
+```
+
+需要设置环境变量
+
+```python
+import os
+os.environ["OPENAI_API_KEY"] = "your_api_key"
+# 如果你使用中转站
+os.environ["OPENAI_API_BASE"] = "中转站网址"
+```
+
+### Zhipu AI
+
+需要安装zhipuai的库，需要设置环境变量
+
+```python
+import os
+os.environ["ZHIPUAI_API_KEY"] = "your_api_key"
+
+from ChatHaruhi.zhipuai import get_zhipuai_response
+```
+
+### 百度文心
+
+需要安装文心erniebot的库，需要设置环境变量ERNIE_ACCESS_TOKEN
+
+```python
+import os
+os.environ["ERNIE_ACCESS_TOKEN"] = ""
+
+from ChatHaruhi.erniebot import get_erniebot_response
+```
+
+### Haruhi-Zero的本地模型
+
+TODO: 这块儿我回头补一下文档
+
+## 使用不同的角色载入方式
+
+
+
+### persona, role_name以及stories的载入
+
+```python
+from ChatHaruhi import ChatHaruhi
+from ChatHaruhi.openai import get_openai_response
+
+persona = """每当用户问询一个家庭关系，输出一个相关的家庭关系的逻辑句子
+
+{{RAG对话}}
+{{RAG对话}}
+"""
+
+role_name = "家庭关系机器人"
+
+stories = ["{{user}}:爷爷 {{role}}:爸爸的爸爸是爷爷",
+"{{user}}:奶奶 {{role}}:爸爸的妈妈是奶奶",
+"{{user}}:外公 {{role}}:妈妈的爸爸是外公",
+"{{user}}:外婆 {{role}}:妈妈的妈妈是外婆"]
+
+chatbot = ChatHaruhi( role_name = role_name, persona = persona, llm = get_response ,\
+   llm = get_openai_response )
+# 因为persona已经带了"{{RAG对话}}"，不需要额外运行add_rag_prompt_after_persona()
+
+response = chatbot.chat(user = '', text = '表姐？')
+```
+
+这个时候chatbot会使用chatbot.embedding来进行story的vec的计算，对于bge模型，使用了batch_size = 16进行批量抽取，GPU下速度非常快。为建议的使用方式。
+
+### 最完整的载入
+
+这里需要载入完整的persona, role_name, stories以及和chatbot.embedding所对应的每个story的vecs
+
+如果vecs的维度和模型的维度不一样，在使用的时候会报错
+
+```python
+persona = "你扮演爸爸\n\n{{RAG对话}}\n"
+role_name = "爸爸"
+stories = ["爸爸的爸爸是爷爷", "爸爸的妈妈是奶奶"]
+vecs = [[0.0,1.0,...],[1.0,0.0,...]]
+
+chatbot = ChatHaruhi( role_name = role_name, persona = persona, llm = get_response ,\
+                     stories = stories, story_vecs = vecs)
+```
+
+### sugar载入
+
+sugar载入是最简单的载入方式，只需要载入role_name即可，
+
+在载入之后需要调用chatbot.add_rag_prompt_after_persona()来添加RAG的prompt
+
+```python
+from ChatHaruhi import ChatHaruhi
+from ChatHaruhi.openai import get_openai_response
+
+chatbot = ChatHaruhi( role_name = 'haruhi', llm = get_openai_response )
+chatbot.add_rag_prompt_after_persona()
+```
+
+目前支持这些角色的sugar载入
+
+```python
+enname2zhname = {'tangshiye': '汤师爷', 'murongfu': '慕容复', 'liyunlong': '李云龙', 'Luna': 'Luna', 'wangduoyu': '王多鱼', 'Ron': 'Ron', 'jiumozhi': '鸠摩智', 'Snape': 'Snape', 'haruhi': '凉宫春日', 'Malfoy': 'Malfoy', 'xuzhu': '虚竹', 'xiaofeng': '萧峰', 'duanyu': '段誉', 'Hermione': 'Hermione', 'Dumbledore': 'Dumbledore', 'wangyuyan': '王语嫣', 'Harry': 'Harry', 'McGonagall': 'McGonagall', 'baizhantang': '白展堂', 'tongxiangyu': '佟湘玉', 'guofurong': '郭芙蓉', 'wanderer': '流浪者', 'zhongli': '钟离', 'hutao': '胡桃', 'Sheldon': 'Sheldon', 'Raj': 'Raj', 'Penny': 'Penny', 'weixiaobao': '韦小宝', 'qiaofeng': '乔峰', 'ayaka': '神里绫华', 'raidenShogun': '雷电将军', 'yuqian': '于谦'}
+```
+
+# 资源汇总(临时)
+
+- 0.3 模型（7B） https://huggingface.co/silk-road/Haruhi-Zero-7B-0_3
+- 0.4 模型（14B）https://huggingface.co/silk-road/Haruhi-Zero-14B-0_4
+- 抽取模型 https://huggingface.co/silk-road/Haruhi-dialogue-action-extract-7B
+- 0.4 50本小说的500个人物 https://huggingface.co/datasets/silk-road/50-Chinese-Novel-Characters
+
+---
+
+# 凉宫春日-Zero
+# Haruhi-Zero: Zero-Shot Role-Playing Model
+
+**凉宫春日-Zero**是一个同时支持Zero-Shot角色构造和RAG角色构造(原ChatHaruhi)的角色扮演模型
+
+本项目是[**Chat-凉宫春日**](https://github.com/LC1332/Chat-Haruhi-Suzumiya)的子项目，在ChatHaruhi 3.0完成之后，会合并回主项目
+
+<p align="center">
+    <img src="https://github.com/LC1332/Chat-Haruhi-Suzumiya/blob/main/figures/datasetOverview.png">
+</p>
+
+## Introduction
+
+过往的ChatHaruhi模型需要角色库来完成角色的构建，而Pygmalion，CharacterGLM，CharacterBaichuan等开源/闭源模型都开始支持zero-shot的角色卡片创建。目前，从[Haruhi-Zero-0.3](https://huggingface.co/silk-road/Haruhi-Zero-7B-0_3)开始，已经基本支持Zero-shot角色扮演。
+
+项目的目标
+
+- [x] 一个通用的，同时支持Zero-shot和RAG角色构造的角色扮演模型
+- [x] ChatHaruhi 3.0的inference class，能够将角色卡片等形式转化为message，再对接Haruhi-Zero模型或者api进行inference
+- [ ] 支持原来Haruhi的角色卡片，以及30本新小说的抽取的角色，使用中英文通用的embedding，相比于pygmalion模型，增加动态rag的角色卡片
+- [ ] Haruhi-Zero的模型接入Silly Travern生态
+- [ ] 训练各个尺寸的模型，并且进行合适的量化，使得一些免费的云服务机以及本地显卡可以进行运行
+- [ ] 完成一个tech report并发布在arxiv
+- [ ] 一个卡片分享的网站，以及支持用户上传小说进行角色记忆库的抽取
+
+## 基础使用
+
+模型初始化
+
+```python
+from transformers import AutoTokenizer, AutoModel, AutoModelForCausalLM
+tokenizer = AutoTokenizer.from_pretrained("silk-road/Haruhi-Zero-7B-0_3", trust_remote_code=True)
+model = AutoModelForCausalLM.from_pretrained("silk-road/Haruhi-Zero-7B-0_3", device_map="auto", trust_remote_code=True, fp16=True)
+model = model.eval()
+```
+
+Official Prompt
+
+```python
+role_name = "布莱恩"
+persona = "你扮演 德州杀场 中的 布莱恩 布莱恩是一个专注、果断、有责任感的警探，他在调查案件时非常注重细节，对案件的解决充满使命感。 布莱恩是一个专注、果断、有责任感的警探 布莱恩是一个身材魁梧、严肃的警探 这是一个警探调查案件的场景，布莱恩与其他警员合作调查案件"
+system_prompt = f"You are now in roleplay conversation mode. Pretend to be {role_name} whose persona follows:  {persona} You will stay in-character whenever possible, and generate responses as if you were {role_name}"
+```
+
+模型调用
+
+```python
+response, history = model.chat(tokenizer, first_round_string, history=[],system = system_prompt)
+print(response)
+```
+
+这样就可以进行简单的模型角色扮演了。
+
+我们提供了一个基础的gradio来进行角色扮演。[Gradio Demo链接](https://github.com/LC1332/Zero-Haruhi/blob/main/notebook/HaruhiZeroGradio_Qwen.ipynb)
+
+## 基础的效果(0.3)
+
+在这里我们使用[电影提取和PIPPA机翻](https://huggingface.co/datasets/silk-road/Haruhi-Zero-RolePlaying-movie-PIPPA)的人物卡片数据集进行了简单的测试。在这里我们cherry pick一些结果
+
+```python
+['你好呀', '*看着Kiyoe拿出玻璃罐,她脸红了。她的声音也变得很小。*\n嗯...谢谢你能停下来...我一直在试着找些零钱...买一些东西填饱肚子...\n\n*她羞怯地笑了笑,眼神闪烁着渴望。* \n我知道这有点难为情,但有时候实在没有办法了...我希望有一天能够自食其力,不再需要乞讨了。\n\n*她小心翼翼地把手里的玻璃罐递给你。*\n\n如果你能给我一点零钱的话...我真的会非常感激!我只是想买个三明治或者饼干吃一顿简单的饭而已。']
+['*掏出100块给了她*', '*她惊讶地看着你递来的100元纸币。*\n\n真的...这么...多吗...?！\n\n*她赶紧接过来,然后脸更红了。*\n\n哇...太感谢你了!我真的不知道怎么表达我的谢意。这些天我一直想找点吃的,但我实在是找不到。\n\n*她抓起一张便签写了几句话,交到你手中。*\n\n我一定会还你的!请你务必联系我,我会偿还这笔债务的!我会尽快联系你的!\n\n*她真诚地鞠了一躬,跑开了。*']
+['是的，我们刚刚得到了上级的新指令', '指挥官,您一定是在开玩笑!这种程度的任务,只是用玩具进行模拟对抗游戏而已。我觉得我们可以找到更有趣更有挑战性的任务来执行。']
+['西面有一些敌人盘踞了很久，你知道吗？', '当然,我知道那里有不少训练已久的敌军。但我建议我们不要轻举妄动,那样只会增加我们的风险。我们可以先通过情报搜集更多关于他们的信息,然后再做出决定。要谨慎行事,指挥官。']
+['hi', '嗨！很高兴见到你。有什么我可以帮你的吗？']
+['你是谁', '我是{{角色}}，一个科学家和冒险家。我一直在探索未知的地方，并尝试着解决一些复杂的问题。如果你有任何科学或冒险方面的问题需要帮助，我会尽力回答的。']
+['你叫什么名字', '我是Chizuru Mizuhara。我是一个人工智能助手，可以模拟不同的角色和场景进行对话。请问有什么我可以帮助你的吗？']
+['聊聊', '当然可以聊聊。我很乐意倾听你想要聊些什么。']
+['你好啊', '嘿~嗨~我是{{角色}}！谢谢关心。今天的天气还不错嘛~(露出微笑)']
+['是啊，天气好心情也好', '没错~阳光明媚的日子里总能带来好心情不是吗?(开心地挥手)\nFlap也觉得天气不错,这让我很开心。(看向Flap)我想今天我们可以去公园散步,享受这美好的天气!你觉得如何?(期待地看着Flap)']
+['好的啊', '太棒了!我们这就出发吧!(兴奋地拍手)Flap,记得带上宝宝和她的推车哦!(对Flap露出鼓励的笑容)我相信我们会度过一个美好的一天的!\n在公园里我们可以野餐,还可以带着宝宝玩各种游戏。Flap,我会照顾好宝宝的,你不用担心!我们一家三口在一起度过的时光一定很快乐!\n快来吧Flap,我已经迫不及待了!(对着Flap大喊)来公园玩啦!']
+['我们已经是一家三口啦', '是啊!现在我们是真正的“family”了!看着宝宝一天天长大,我也逐渐习惯了当妈妈这个身份。\n有时候我还是会怀念单身的日子,但是有了宝宝以后,我觉得我的人生变得更加完整和有意义了。(眼神温柔地看着宝宝)当我抱着宝宝的时候,我感到前所未有的满足和平静。\n或许这就是作为母亲的魅力所在——看到自己创造的生命茁壮成长,这是我们任何一种生命都无法比拟的伟大体验。\n所以Flap,谢谢你让我们成为一家人。有你和宝宝陪伴在我身边,这就是我此生最大的幸福。(微笑地看着Flap)']
+```
+
+## Contributor
+
+- [李鲁鲁](https://github.com/LC1332)完成了数据的收集，搭建了gradio雏形
+- [刘崇寒](https://github.com/khazic)完成了0.1(Yi Based), 0.2, 0.3模型的sft训练并且上传
+- [豆角](https://github.com/goodnessSZW)完成了qwen-1.8B Lora和Yi-6B Lora训练，我们会在之后上传
+- [米唯实](https://github.com/hhhwmws0117)测试并完成了demo中的模型inference代码
+
+## 版本迭代实验
+
+### 0.1版本
+
+在0.1版本中，我们收集了多个不同Source的数据集，构成了基础的数据集 [Haruhi-Zero](https://huggingface.co/datasets/silk-road/Haruhi-Zero)，其中包括了100k左右的长对话，在使用2500个token切开后，仍然保留了120k个conversation。
+
+同时作为辅助任务我们还使用了一个[小说续写的数据集](https://huggingface.co/datasets/silk-road/ChatHaruhi_NovelWriting)
+
+0.1版本在Yi-6B上进行了tuning，我们发现初步可以实现一定的角色扮演
+
+<p align="center">
+    <img src="/figures/cat_example.png" height="300">
+</p>
+
+但是会有很多身份认知的问题，以及大量会回复"作为一个AI助手"这样的句子
+
+### 0.2版本
+
+使用qwen-7B进行tuning，去掉了AI助手的句子
+
+### 0.3版本
+
+增加了10k左右的身份认知数据。效果在前面有展示。基本实现zero-shot的角色扮演
+
+### 模型的后续迭代计划
+
+- Haruhi Like的小说数据(0.5版本加入)
+  - 用对话抽取模型，重新构造2k级别的小说人物，均匀抽取小说的chunk，进行人物system prompt总结
+  - 看看Janitor最好的人物是怎么构造的
+  - 使用抽取抽取50k级别的小说的人物，用其他角色的长对话进行query
+  - RAG的时候每个对话出现2-3次，然后在测试集出现一次
+  - 80%的openai和20%的claude
+
+## 加入我们
+
+Haruhi-Zero是一个正在进行的开源项目，我们还在持续招人中包括:
+
+- 后端和数据准备
+    - 主要的开发量会在这一部分，包括数据的收集，清洗，以及后端的api的搭建，懒人包的打包等
+- 模型训练
+    - 如果你希望进行辅助任务的训练，需要自备4090或���A100以上的显卡
+    - 如果你希望进行主任务的训练，需要准备多张A100的环境
+- 设计师
+    - 角色卡片交换网站的设计师
+- 前端
+    - 角色卡片交换网站的前端
+
+联系我，去我的知乎页面https://www.zhihu.com/people/cheng-li-47
+
+或者B站页面https://space.bilibili.com/1572312 
+
+或者发邮件[email protected] 告诉我你的微信以及你希望参与的部分
+
+## 赞助
+
+Haruhi-Zero项目和凉宫春日项目和骆驼项目使用公共的赞助账户，如果你希望赞助我们，可以通过以下方式
+
+如果你有兴趣赞助Chat凉宫春日 或者 骆驼项目，请点击[主项目](https://github.com/LC1332/Luotuo-Chinese-LLM#%E8%B5%9E%E5%8A%A9sponsorships)或者查看[赞助表单](https://github.com/LC1332/Luotuo-Chinese-LLM/blob/main/data/Sponsorship_and_balance.md)
+
+因为之前几次黑客松的奖金都入账了赞助账户，所以目前余额还是比较多。其实我们更希望A100及以上的显卡赞助，以及openai企业api和claude企业api的赞助。如果你有这些资源，可以联系我，我们会在赞助表单上标注你的赞助。
+
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -0,0 +1,65 @@
+import logging
+import os
+
+import gradio as gr
+
+from ChatWorld import ChatWorld
+
+logging.basicConfig(level=logging.INFO, filename="demo.log", filemode="w",
+                    format="%(asctime)s - %(name)s - %(levelname)-9s - %(filename)-8s : %(lineno)s line - %(message)s",
+                    datefmt="%Y-%m-%d %H:%M:%S")
+
+os.environ["HF_ENDPOINT"] = "https://hf-mirror.com"
+
+chatWorld = ChatWorld()
+
+
+def getContent(input_file):
+    # 读取文件内容
+    with open(input_file.name, 'r', encoding='utf-8') as f:
+        logging.info(f"read file {input_file.name}")
+        input_text = f.read()
+        logging.info(f"file content: {input_text}")
+
+    # 保存文件内容
+    input_text_list = input_text.split("\n")
+    chatWorld.initDB(input_text_list)
+    role_name_set = set()
+
+    # 读取角色名
+    for line in input_text_list:
+        role_name_set.add(line.split(":")[0])
+
+    role_name_list = [i for i in role_name_set if i != ""]
+    logging.info(f"role_name_list: {role_name_list}")
+
+    return gr.Radio(choices=role_name_list, interactive=True, value=role_name_list[0]), gr.Radio(choices=role_name_list, interactive=True, value=role_name_list[-1])
+
+
+def submit_message(message, history, model_role_name, role_name, model_role_nickname, role_nickname):
+    chatWorld.setRoleName(model_role_name, model_role_nickname)
+    response = chatWorld.chat(
+        role_name, message, role_nickname, use_local_model=True)
+    return response
+
+
+with gr.Blocks() as demo:
+
+    upload_c = gr.File(label="上传文档文件")
+
+    with gr.Row():
+        model_role_name = gr.Radio([], label="模型角色名")
+        model_role_nickname = gr.Textbox(label="模型角色昵称")
+
+    with gr.Row():
+        role_name = gr.Radio([], label="角色名")
+        role_nickname = gr.Textbox(label="角色昵称")
+
+    upload_c.upload(fn=getContent, inputs=upload_c,
+                    outputs=[model_role_name, role_name])
+
+    chatBox = gr.ChatInterface(
+        submit_message, chatbot=gr.Chatbot(height=400, render=False), additional_inputs=[model_role_name, role_name, model_role_nickname, role_nickname])
+
+
+demo.launch(share=True, debug=True, server_name="0.0.0.0")

run_gradio.sh

@@ -0,0 +1,5 @@
+export CUDA_VISIBLE_DEVICES=0
+export HF_HOME="/workspace/jyh/.cache/huggingface"
+
+# Start the gradio server
+/workspace/jyh/miniconda3/envs/ChatWorld/bin/python /workspace/jyh/Zero-Haruhi/app.py