From 1e12944a5dfac1d0c5f38aa29b92be6d6b77bbe7 Mon Sep 17 00:00:00 2001 From: hzg0601 Date: Wed, 16 Aug 2023 22:58:59 +0800 Subject: [PATCH 01/12] =?UTF-8?q?=E6=9B=B4=E6=96=B0shutdown=5Fall.sh:?= =?UTF-8?q?=E6=9B=B4=E6=96=B0mac=E8=AE=BE=E5=A4=87=E7=9A=84=E9=A2=9D?= =?UTF-8?q?=E5=A4=96=E8=AF=B4=E6=98=8E?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- shutdown_all.sh | 1 + 1 file changed, 1 insertion(+) diff --git a/shutdown_all.sh b/shutdown_all.sh index 961260d..0218147 100644 --- a/shutdown_all.sh +++ b/shutdown_all.sh @@ -1 +1,2 @@ +# mac设备上的grep命令可能不支持grep -P选项，请使用Homebrew安装;或使用ggrep命令 ps -eo pid,user,cmd|grep -P 'server/api.py|webui.py|fastchat.serve'|grep -v grep|awk '{print $1}'|xargs kill -9 \ No newline at end of file From cb3fe84fe3f49167d8b846df4b2bdd92b52772d3 Mon Sep 17 00:00:00 2001 From: imClumsyPanda Date: Sun, 20 Aug 2023 10:41:28 +0800 Subject: [PATCH 02/12] update VERSION --- configs/__init__.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/configs/__init__.py b/configs/__init__.py index 6e0ad13..dc9dd40 100644 --- a/configs/__init__.py +++ b/configs/__init__.py @@ -1,4 +1,4 @@ from .model_config import * from .server_config import * -VERSION = "v0.2.1" +VERSION = "v0.2.2-preview" From cc0bd4efd99106a461a0074b6ab9bf7f074d2b44 Mon Sep 17 00:00:00 2001 From: hzg0601 Date: Fri, 25 Aug 2023 11:27:39 +0800 Subject: [PATCH 03/12] =?UTF-8?q?update=20server=5Fconfig.py.example:=20?= =?UTF-8?q?=E5=A2=9E=E5=8A=A0=E5=A4=9A=E5=8D=A1=E5=90=AF=E5=8A=A8=E7=9A=84?= =?UTF-8?q?=E8=AF=B4=E6=98=8E?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- configs/server_config.py.example | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/configs/server_config.py.example b/configs/server_config.py.example index 5f37779..b0f37bf 100644 --- a/configs/server_config.py.example +++ b/configs/server_config.py.example @@ -34,11 +34,11 @@ FSCHAT_MODEL_WORKERS = { "port": 20002, "device": LLM_DEVICE, # todo: 多卡加载需要配置的参数 - "gpus": None, - "numgpus": 1, + "gpus": None, # 使用的GPU，以str的格式指定，如"0,1" + "num_gpus": 1, # 使用GPU的数量 # 以下为非常用参数，可根据需要配置 - # "max_gpu_memory": "20GiB", - # "load_8bit": False, + # "max_gpu_memory": "20GiB", # 每个GPU占用的最大显存 + # "load_8bit": False, # 开启8bit量化 # "cpu_offloading": None, # "gptq_ckpt": None, # "gptq_wbits": 16, From 33ce276e3ea75a360206896e06a489d426ce6fd8 Mon Sep 17 00:00:00 2001 From: imClumsyPanda Date: Fri, 25 Aug 2023 15:18:43 +0800 Subject: [PATCH 04/12] change default search_engine option to 'duckduckgo' --- webui_pages/dialogue/dialogue.py | 16 +++++++++++----- 1 file changed, 11 insertions(+), 5 deletions(-) diff --git a/webui_pages/dialogue/dialogue.py b/webui_pages/dialogue/dialogue.py index a317aba..04ece7d 100644 --- a/webui_pages/dialogue/dialogue.py +++ b/webui_pages/dialogue/dialogue.py @@ -80,8 +80,13 @@ def dialogue_page(api: ApiRequest): # chunk_content = st.checkbox("关联上下文", False, disabled=True) # chunk_size = st.slider("关联长度：", 0, 500, 250, disabled=True) elif dialogue_mode == "搜索引擎问答": + search_engine_list = list(SEARCH_ENGINES.keys()) with st.expander("搜索引擎配置", True): - search_engine = st.selectbox("请选择搜索引擎", SEARCH_ENGINES.keys(), 0) + search_engine = st.selectbox( + label="请选择搜索引擎", + options=search_engine_list, + index=search_engine_list.index("duckduckgo") if "duckduckgo" in search_engine_list else 0, + ) se_top_k = st.number_input("匹配搜索结果条数：", 1, 20, 3) # Display chat messages from history on app rerun @@ -125,11 +130,12 @@ def dialogue_page(api: ApiRequest): ]) text = "" for d in api.search_engine_chat(prompt, search_engine, se_top_k): - if error_msg := check_error_msg(d): # check whether error occured + if error_msg := check_error_msg(d): # check whether error occured st.error(error_msg) - text += d["answer"] - chat_box.update_msg(text, 0) - chat_box.update_msg("\n\n".join(d["docs"]), 1, streaming=False) + else: + text += d["answer"] + chat_box.update_msg(text, 0) + chat_box.update_msg("\n\n".join(d["docs"]), 1, streaming=False) chat_box.update_msg(text, 0, streaming=False) now = datetime.now() From faba1b3877623891cfef60ac08f13780b1713586 Mon Sep 17 00:00:00 2001 From: hzg0601 Date: Fri, 25 Aug 2023 16:16:44 +0800 Subject: [PATCH 05/12] =?UTF-8?q?update=20readme.md,=20shutdown=5Fall.sh:?= =?UTF-8?q?=20=E5=9C=A8Linux=E4=B8=8A=E4=BD=BF=E7=94=A8ctrl+C=E9=80=80?= =?UTF-8?q?=E5=87=BA=E5=8F=AF=E8=83=BD=E4=BC=9A=E7=94=B1=E4=BA=8Elinux?= =?UTF-8?q?=E7=9A=84=E5=A4=9A=E8=BF=9B=E7=A8=8B=E6=9C=BA=E5=88=B6=E5=AF=BC?= =?UTF-8?q?=E8=87=B4multiprocessing=E9=81=97=E7=95=99=E5=AD=A4=E5=84=BF?= =?UTF-8?q?=E8=BF=9B=E7=A8=8B?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- README.md | 16 +++++++--------- shutdown_all.sh | 2 +- 2 files changed, 8 insertions(+), 10 deletions(-) diff --git a/README.md b/README.md index bf3a5ae..208d5de 100644 --- a/README.md +++ b/README.md @@ -208,7 +208,8 @@ embedding_model_dict = { "m3e-base": "/Users/xxx/Downloads/m3e-base", } ``` -如果你选择使用OpenAI的Embedding模型，请将模型的```key```写入`embedding_model_dict`中。使用该模型，你需要鞥能够访问OpenAI官的API，或设置代理。 + +如果你选择使用OpenAI的Embedding模型，请将模型的 ``key``写入 `embedding_model_dict`中。使用该模型，你需要鞥能够访问OpenAI官的API，或设置代理。 ### 4. 知识库初始化与迁移 @@ -219,7 +220,7 @@ embedding_model_dict = { ```shell $ python init_database.py ``` -- 如果您是第一次运行本项目，知识库尚未建立，或者配置文件中的知识库类型、嵌入模型发生变化，或者之前的向量库没有开启`normalize_L2`，需要以下命令初始化或重建知识库： +- 如果您是第一次运行本项目，知识库尚未建立，或者配置文件中的知识库类型、嵌入模型发生变化，或者之前的向量库没有开启 `normalize_L2`，需要以下命令初始化或重建知识库： ```shell $ python init_database.py --recreate-vs @@ -308,7 +309,6 @@ $ python server/llm_api_shutdown.py --serve all ![image](https://github.com/chatchat-space/Langchain-Chatchat/assets/22924096/4e056c1c-5c4b-4865-a1af-859cd58a625d) - #### 5.2 启动 API 服务 本地部署情况下，按照 [5.1 节](README.md#5.1-启动-LLM-服务)**启动 LLM 服务后**，再执行 [server/api.py](server/api.py) 脚本启动 **API** 服务； @@ -366,17 +366,13 @@ $ python startup.py -a 并可使用 `Ctrl + C` 直接关闭所有运行服务。如果一次结束不了，可以多按几次。 -可选参数包括 `-a (或--all-webui)`, `--all-api`, `--llm-api`, `-c (或--controller)`, `--openai-api`, +可选参数包括 `-a (或--all-webui)`, `--all-api`, `--llm-api`, `-c (或--controller)`, `--openai-api`, `-m (或--model-worker)`, `--api`, `--webui`，其中： - `--all-webui` 为一键启动 WebUI 所有依赖服务； - - `--all-api` 为一键启动 API 所有依赖服务； - - `--llm-api` 为一键启动 Fastchat 所有依赖的 LLM 服务； - - `--openai-api` 为仅启动 FastChat 的 controller 和 openai-api-server 服务； - - 其他为单独服务启动选项。 若想指定非默认模型，需要用 `--model-name` 选项，示例： @@ -385,7 +381,7 @@ $ python startup.py -a $ python startup.py --all-webui --model-name Qwen-7B-Chat ``` -更多信息可通过`python startup.py -h`查看。 +更多信息可通过 `python startup.py -h`查看。 **注意：** @@ -393,6 +389,8 @@ $ python startup.py --all-webui --model-name Qwen-7B-Chat **2.服务启动时间示设备不同而不同，约 3-10 分钟，如长时间没有启动请前往 `./logs`目录下监控日志，定位问题。** +**3. 在Linux上使用ctrl+C退出可能会由于linux的多进程机制导致multiprocessing遗留孤儿进程，可通过shutdown_all.sh进行退出** + ## 常见问题 参见 [常见问题](docs/FAQ.md)。 diff --git a/shutdown_all.sh b/shutdown_all.sh index 0218147..8c64806 100644 --- a/shutdown_all.sh +++ b/shutdown_all.sh @@ -1,2 +1,2 @@ # mac设备上的grep命令可能不支持grep -P选项，请使用Homebrew安装;或使用ggrep命令 -ps -eo pid,user,cmd|grep -P 'server/api.py|webui.py|fastchat.serve'|grep -v grep|awk '{print $1}'|xargs kill -9 \ No newline at end of file +ps -eo pid,user,cmd|grep -P 'server/api.py|webui.py|fastchat.serve|multiprocessing'|grep -v grep|awk '{print $1}'|xargs kill -9 \ No newline at end of file From 79d225be36d24b05c7ac4522f20e1a2c2cc64841 Mon Sep 17 00:00:00 2001 From: imClumsyPanda Date: Fri, 25 Aug 2023 16:26:51 +0800 Subject: [PATCH 06/12] update VERSION --- configs/__init__.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/configs/__init__.py b/configs/__init__.py index dc9dd40..7c105f4 100644 --- a/configs/__init__.py +++ b/configs/__init__.py @@ -1,4 +1,4 @@ from .model_config import * from .server_config import * -VERSION = "v0.2.2-preview" +VERSION = "v0.2.2" From 021677d5d66cf624cd0c8ebfe9d9a2b34c4a41a5 Mon Sep 17 00:00:00 2001 From: imClumsyPanda Date: Fri, 25 Aug 2023 18:04:06 +0800 Subject: [PATCH 07/12] Update README.md --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 208d5de..c5c2d68 100644 --- a/README.md +++ b/README.md @@ -42,7 +42,7 @@ 🚩 本项目未涉及微调、训练过程，但可利用微调或训练对本项目效果进行优化。 -🌐 [AutoDL 镜像](https://www.codewithgpu.com/i/imClumsyPanda/langchain-ChatGLM/Langchain-Chatchat) 中 `v5` 版本所使用代码已更新至本项目 `0.2.0` 版本。 +🌐 [AutoDL 镜像](https://www.codewithgpu.com/i/imClumsyPanda/langchain-ChatGLM/Langchain-Chatchat) 中 `v6` 版本所使用代码已更新至本项目 `0.2.2` 版本。 🐳 [Docker 镜像](registry.cn-beijing.aliyuncs.com/chatchat/chatchat:0.2.0) From 2e930cf43c49fb3287155c31736b63d6f2768221 Mon Sep 17 00:00:00 2001 From: imClumsyPanda Date: Fri, 25 Aug 2023 22:14:56 +0800 Subject: [PATCH 08/12] update requirements.txt --- requirements.txt | 3 +++ 1 file changed, 3 insertions(+) diff --git a/requirements.txt b/requirements.txt index 93908dd..171e2a7 100644 --- a/requirements.txt +++ b/requirements.txt @@ -14,6 +14,8 @@ python-magic-bin; sys_platform == 'win32' SQLAlchemy==2.0.19 faiss-cpu nltk +accelerate +spacy # uncomment libs if you want to use corresponding vector store # pymilvus==2.1.3 # requires milvus==2.1.3 @@ -28,3 +30,4 @@ streamlit-antd-components>=0.1.11 streamlit-chatbox>=1.1.6 streamlit-aggrid>=0.3.4.post3 httpx~=0.24.1 +watchdog \ No newline at end of file From ab5d1df2a34d0ebbae6f901b0705bcfd09021f36 Mon Sep 17 00:00:00 2001 From: imClumsyPanda Date: Fri, 25 Aug 2023 22:16:49 +0800 Subject: [PATCH 09/12] update requirements --- requirements_api.txt | 2 ++ requirements_webui.txt | 3 ++- 2 files changed, 4 insertions(+), 1 deletion(-) diff --git a/requirements_api.txt b/requirements_api.txt index f567f9f..58dbc0c 100644 --- a/requirements_api.txt +++ b/requirements_api.txt @@ -14,6 +14,8 @@ python-magic-bin; sys_platform == 'win32' SQLAlchemy==2.0.19 faiss-cpu nltk +accelerate +spacy # uncomment libs if you want to use corresponding vector store # pymilvus==2.1.3 # requires milvus==2.1.3 diff --git a/requirements_webui.txt b/requirements_webui.txt index a832550..1645fe1 100644 --- a/requirements_webui.txt +++ b/requirements_webui.txt @@ -6,4 +6,5 @@ streamlit-antd-components>=0.1.11 streamlit-chatbox>=1.1.6 streamlit-aggrid>=0.3.4.post3 httpx~=0.24.1 -nltk \ No newline at end of file +nltk +watchdog \ No newline at end of file From 32bf508745254e74c4b845a926cba38079c04e16 Mon Sep 17 00:00:00 2001 From: imClumsyPanda Date: Fri, 25 Aug 2023 22:53:49 +0800 Subject: [PATCH 10/12] update requirements --- requirements.txt | 2 +- requirements_webui.txt | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/requirements.txt b/requirements.txt index 171e2a7..a72c760 100644 --- a/requirements.txt +++ b/requirements.txt @@ -24,7 +24,7 @@ spacy numpy~=1.24.4 pandas~=2.0.3 -streamlit>=1.25.0 +streamlit>=1.26.0 streamlit-option-menu>=0.3.6 streamlit-antd-components>=0.1.11 streamlit-chatbox>=1.1.6 diff --git a/requirements_webui.txt b/requirements_webui.txt index 1645fe1..da66c30 100644 --- a/requirements_webui.txt +++ b/requirements_webui.txt @@ -1,6 +1,6 @@ numpy~=1.24.4 pandas~=2.0.3 -streamlit>=1.25.0 +streamlit>=1.26.0 streamlit-option-menu>=0.3.6 streamlit-antd-components>=0.1.11 streamlit-chatbox>=1.1.6 From fea7e8ddf172eb5ad66213608f05c981f0e333b1 Mon Sep 17 00:00:00 2001 From: imClumsyPanda Date: Sat, 26 Aug 2023 11:10:45 +0800 Subject: [PATCH 11/12] update VERSION --- configs/__init__.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/configs/__init__.py b/configs/__init__.py index 7c105f4..b42f0b5 100644 --- a/configs/__init__.py +++ b/configs/__init__.py @@ -1,4 +1,4 @@ from .model_config import * from .server_config import * -VERSION = "v0.2.2" +VERSION = "v0.2.3-preview" From f95d41ef471707ddc0f0a2430c1cbc9faf2dfdcc Mon Sep 17 00:00:00 2001 From: imClumsyPanda Date: Sat, 26 Aug 2023 11:45:01 +0800 Subject: [PATCH 12/12] =?UTF-8?q?[BUG]=20=E4=BF=AE=E5=A4=8Dcsv=E6=96=87?= =?UTF-8?q?=E4=BB=B6=E8=AF=BB=E5=8F=96=E5=90=8E=EF=BC=8C=E5=8D=95=E8=A1=8C?= =?UTF-8?q?=E6=95=B0=E6=8D=AE=E8=A2=AB=E5=88=86=E6=88=90=E5=A4=9A=E6=AE=B5?= =?UTF-8?q?=E3=80=82?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- server/knowledge_base/utils.py | 43 ++++++++++++++++++---------------- 1 file changed, 23 insertions(+), 20 deletions(-) diff --git a/server/knowledge_base/utils.py b/server/knowledge_base/utils.py index da53049..34f2083 100644 --- a/server/knowledge_base/utils.py +++ b/server/knowledge_base/utils.py @@ -104,32 +104,35 @@ class KnowledgeFile: else: loader = DocumentLoader(self.filepath) - try: - if self.text_splitter_name is None: + if self.ext in ".csv": + docs = loader.load() + else: + try: + if self.text_splitter_name is None: + text_splitter_module = importlib.import_module('langchain.text_splitter') + TextSplitter = getattr(text_splitter_module, "SpacyTextSplitter") + text_splitter = TextSplitter( + pipeline="zh_core_web_sm", + chunk_size=CHUNK_SIZE, + chunk_overlap=OVERLAP_SIZE, + ) + self.text_splitter_name = "SpacyTextSplitter" + else: + text_splitter_module = importlib.import_module('langchain.text_splitter') + TextSplitter = getattr(text_splitter_module, self.text_splitter_name) + text_splitter = TextSplitter( + chunk_size=CHUNK_SIZE, + chunk_overlap=OVERLAP_SIZE) + except Exception as e: + print(e) text_splitter_module = importlib.import_module('langchain.text_splitter') - TextSplitter = getattr(text_splitter_module, "SpacyTextSplitter") + TextSplitter = getattr(text_splitter_module, "RecursiveCharacterTextSplitter") text_splitter = TextSplitter( - pipeline="zh_core_web_sm", chunk_size=CHUNK_SIZE, chunk_overlap=OVERLAP_SIZE, ) - self.text_splitter_name = "SpacyTextSplitter" - else: - text_splitter_module = importlib.import_module('langchain.text_splitter') - TextSplitter = getattr(text_splitter_module, self.text_splitter_name) - text_splitter = TextSplitter( - chunk_size=CHUNK_SIZE, - chunk_overlap=OVERLAP_SIZE) - except Exception as e: - print(e) - text_splitter_module = importlib.import_module('langchain.text_splitter') - TextSplitter = getattr(text_splitter_module, "RecursiveCharacterTextSplitter") - text_splitter = TextSplitter( - chunk_size=CHUNK_SIZE, - chunk_overlap=OVERLAP_SIZE, - ) - docs = loader.load_and_split(text_splitter) + docs = loader.load_and_split(text_splitter) print(docs[0]) if using_zh_title_enhance: docs = zh_title_enhance(docs)