moemu / muice-chatbot Goto Github PK

Linux求求

描述此Bug

清晰简洁的描述。

控制台输出

从项目开始运行时控制台的所有输出，注意删减你的个人信息

复现该Bug的操作

告诉我们你如何复现此问题

额外信息

代码输入： git clone https://github.com/Moemu/Muice-Chatbot

cd Muice-Chatbot
conda create --name Muice python=3.10.10
conda activate Muice
pip install -r requirements.txt

控制台输出：git : 无法将“git”项识别为 cmdlet、函数、脚本文件或可运行程序的名称。请检查名称的拼写，如果包括路径，请确保路径正确，
然后再试一次。
所在位置 行:1 字符: 1

• git clone https://github.com/Moemu/Muice-Chatbot

•   + CategoryInfo          : ObjectNotFound: (git:String) [], CommandNotFoundException
    + FullyQualifiedErrorId : CommandNotFoundException
  

求解Thanks!

2周前，gocq的作者Mrs4s发布1.2.0版本，同时更新文档声明不再对gocq维护；同时签名服务器作者删库、一众后端应声倒地。
可见：Mrs4s/go-cqhttp#2471
问问作者对这方面怎么考虑（高三辛苦，高一牲祝您高考顺利）

運行環境
Windows 11 Pro，conda 23.5.2

顯示卡
NVIDIA GeForce RTX 4060

過程（復現）
在安裝完所有組件後，於 Muice-Chatbot-main 啓動 Windows Terminal，依次執行 conda activate Muice 和 python main.py，會產生 AssertionError: Torch not compiled with CUDA enabled 錯誤。

檔案結構

├─Muice-Chatbot-main   <- 主路徑
│  ├─__pycache__
│  ├─llm
│  ├─model
│       ├─ chatglm2-6b
│       ├─ Muice
│  ├─qqbot
│  └─src
├─main.py  <- 主處理程式
└─...

完整輸出

Microsoft Windows [Version 10.0.22621.2134]
(c) Microsoft Corporation. All rights reserved.

C:\Users\[已略去]\Desktop\Muice\Muice-Chatbot-main>conda activate Muice

(Muice) C:\Users\[已略去]\Desktop\Muice\Muice-Chatbot-main>python main.py
Loading checkpoint shards: 100%|█████████████████████████████████████████████████████████| 7/7 [00:08<00:00,  1.17s/it]
Some weights of ChatGLMForConditionalGeneration were not initialized from the model checkpoint at model/chatglm2-6b and are newly initialized: ['transformer.prefix_encoder.embedding.weight']
You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.
Traceback (most recent call last):
  File "C:\Users\[已略去]\Desktop\Muice\Muice-Chatbot-main\main.py", line 10, in <module>
    model = Model('model/chatglm2-6b','model/Muice')
  File "C:\Users\[已略去]\Desktop\Muice\Muice-Chatbot-main\llm\chatglm2.py", line 14, in __init__
    model = AutoModel.from_pretrained(model_path, config=config, trust_remote_code=True).half().to(device)
  File "C:\Users\[已略去]\.conda\envs\Muice\lib\site-packages\transformers\modeling_utils.py", line 1900, in to
    return super().to(*args, **kwargs)
  File "C:\Users\[已略去]\.conda\envs\Muice\lib\site-packages\torch\nn\modules\module.py", line 1145, in to
    return self._apply(convert)
  File "C:\Users\[已略去]\.conda\envs\Muice\lib\site-packages\torch\nn\modules\module.py", line 797, in _apply
    module._apply(fn)
  File "C:\Users\[已略去]\.conda\envs\Muice\lib\site-packages\torch\nn\modules\module.py", line 797, in _apply
    module._apply(fn)
  File "C:\Users\[已略去]\.conda\envs\Muice\lib\site-packages\torch\nn\modules\module.py", line 797, in _apply
    module._apply(fn)
  File "C:\Users\[已略去]\.conda\envs\Muice\lib\site-packages\torch\nn\modules\module.py", line 820, in _apply
    param_applied = fn(param)
  File "C:\Users\[已略去]\.conda\envs\Muice\lib\site-packages\torch\nn\modules\module.py", line 1143, in convert
    return t.to(device, dtype if t.is_floating_point() or t.is_complex() else None, non_blocking)
  File "C:\Users\[已略去]\.conda\envs\Muice\lib\site-packages\torch\cuda\__init__.py", line 239, in _lazy_init
    raise AssertionError("Torch not compiled with CUDA enabled")
AssertionError: Torch not compiled with CUDA enabled

(Muice) C:\Users\[已略去]\Desktop\Muice\Muice-Chatbot-main>

登录pygocq后(使用安卓手表协议)可以正确收到消息 但是qq没有回复 同时cou和内存负载并不高 疑似模型未正常运行 请问是否需要提供其他log文件等

⌈AI 女孩子⌋的設想固然美好，或許還帶有幾分浪漫，但人工智慧系統系統是程式和算法的集合，它們是為了執行特定任務而設計的。它們沒有自我意識、情感或自主性。在技術層面，將其視為機器或工具更為準確。雖然賦予其人格化的稱謂可以提升用戶體驗（説是⌈顯著提升⌋也不爲過），但身爲開發者必須考慮到一種⌈用戶沈迷其中，而視 AI 爲情感依賴⌋的極端可能。AI畢竟不能取代人類在情感依賴方面的體驗，因此希望您可以在項目的説明文檔中闡明這一點，和/或在模型中加入類似的限制（可參考 ChatGPT 對於此類問題的回答）。

在執行 python main.py 後，pycqbot 正常載入，但會一直卡在 device lock is disable. http api may fail 處。如下所示：

PS C:\Users\[已略去]\Desktop\Muice> conda activate Muice 
PS C:\Users\[已略去]\Desktop\Muice> python main.py 
Loading checkpoint shards: 100%|█████████████████████████████████████████████████████████| 7/7 [00:08<00:00,  1.25s/it] 
Some weights of ChatGLMForConditionalGeneration were not initialized from the model checkpoint at model/chatglm2-6b and are newly initialized: ['transformer.prefix_encoder.embedding.weight'] 
You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference. 
[2023-09-09 02:49:43,201][MainThread/INFO] PyCqBot: 创建定时任务 CreateANewTopic 
--- Logging error --- 
Traceback (most recent call last): 
  File "C:\Users\[已略去]\AppData\Local\Programs\Python\Python310\lib\logging\init.py", line 1103, in emit 
    stream.write(msg + self.terminator) 
UnicodeEncodeError: 'cp950' codec can't encode character '\u521b' in position 56: illegal multibyte sequence 
Call stack: 
  File "C:\Users\[已略去]\Desktop\Muice\main.py", line 16, in <module> 
    qqbot(muice, Trust_QQ_list=configs['Trust_QQ_list'], AutoCreateTopic=configs['AutoCreateTopic']) 
  File "C:\Users\[已略去]\Desktop\Muice\qqbot.py", line 53, in qqbot 
    bot.timing(CreateANewChat,"CreateANewTopic",{"timeSleep":60}) 
  File "C:\Users\[已略去]\AppData\Local\Programs\Python\Python310\lib\site-packages\pycqBot\cqHttpApi.py", line 638, in timing 
    logging.info("创建定时任务 %s " % timing_name) 
Message: '创建定时任务 CreateANewTopic '
Arguments: ()

#################################################################
██████╗ ██╗   ██╗ ██████╗ ██████╗ ██████╗  ██████╗ ████████╗
██╔══██╗╚██╗ ██╔╝██╔════╝██╔═══██╗██╔══██╗██╔═══██╗╚══██╔══╝
██████╔╝ ╚████╔╝ ██║     ██║   ██║██████╔╝██║   ██║   ██║
██╔═══╝   ╚██╔╝  ██║     ██║▄▄ ██║██╔══██╗██║   ██║   ██║
██║        ██║   ╚██████╗╚██████╔╝██████╔╝╚██████╔╝   ██║
╚═╝        ╚═╝    ╚═════╝ ╚══▀▀═╝ ╚═════╝  ╚═════╝    ╚═╝
                                            v0.5.1.1  BY FengLiu
#################################################################

[2023-09-09 12:00:17,557][go-cqhttp/WARNING] PyCqBot: go-cqhttp 警告 Protocol -> device lock is disable. http api may fail.
|

但若直接由 Powershell 啓動 go-cqhttp，則可以正常登入，如下所示：

[2023-09-09 12:05:11] [INFO]: 当前版本:v1.1.0
[2023-09-09 12:05:11] [INFO]: 将使用 device.json 内的设备信息运行Bot.
[2023-09-09 12:05:11] [INFO]: 使用服务器 http://127.0.0.1:13579 进行数据包签名
[2023-09-09 12:05:11] [INFO]: Bot将在5秒后登录并开始信息处理, 按 Ctrl+C 取消.
[2023-09-09 12:05:16] [INFO]: 开始尝试登录并同步消息...
[2023-09-09 12:05:16] [INFO]: 使用协议: iPad 8.9.33.614
[2023-09-09 12:05:16] [INFO]: Protocol -> connect to server: [::ffff:14.22.9.53]:443
[2023-09-09 12:05:17] [WARNING]: Protocol -> device lock is disable. http api may fail.
[2023-09-09 12:05:19] [INFO]: 正在检查协议更新...
[2023-09-09 12:05:19] [INFO]: 登录成功 欢迎使用: [QQ昵稱]
[2023-09-09 12:05:19] [INFO]: 开始加载好友列表...
[2023-09-09 12:05:19] [INFO]: 共加载 [已略去] 个好友.
[2023-09-09 12:05:19] [INFO]: 开始加载群列表...
[2023-09-09 12:05:19] [INFO]: 共加载 [已略去] 个群.
[2023-09-09 12:05:19] [INFO]: 资源初始化完成, 开始处理信息.
[2023-09-09 12:05:19] [INFO]: アトリは、高性能ですから!
[2023-09-09 12:05:19] [INFO]: 正在检查更新.
[2023-09-09 12:05:19] [INFO]: 检查更新完成. 当前已运行最新版本.
[2023-09-09 12:05:19] [INFO]: 开始诊断网络情况
[2023-09-09 12:05:21] [INFO]: 网络诊断完成. 未发现问题

描述此Bug

Muice-Chatbot根目录：

model文件夹下的chatglm2-6b-int4文件夹内部：

model文件夹下的Muice文件夹内部：

main.py文件:

控制台输出

若将chatglm2-6b-int4文件夹放在ChatGLM2-6B文件夹内，可正常运行int4模型，因此并非是模型出错。

复现该Bug的操作

从GitHub上下载Muice-Chatbot源文件并不小心覆盖已修改的main.py文件。

额外信息

之前修改main.py文件后，运行成功了一次，但因go-cqhttp无法连接而失败。后因失误，把原main.py文件覆盖了修改后的main.py文件，且忘记怎么输入的，导致无法再次运行。尝试绝对路径但依然加载模型失败。希望up有空时能帮忙检查下问题所在。
另外因为腾讯围堵，go-cq应该是用不了了，希望能尝试下对接ntqq的方法。
up辛苦了，祝你高考顺利考上理想的大学！

重新编写的main.py并提供onebot服务不过目前不支持使用指令以及主动对话,未来应该会调试的

你好，能否共享一下训练数据呢

描述此Bug

原版ChatGLM2可以使用，但使用这个微调模型会报错

控制台输出

Traceback (most recent call last):
File "D:\chatglm\web_demo.py", line 6, in
tokenizer = AutoTokenizer.from_pretrained("D:\chatglm\models\Muice", trust_remote_code=True)
File "D:\Programs Files\Python\Python311\Lib\site-packages\transformers\models\auto\tokenization_auto.py", line 678, in from_pretrained
return tokenizer_class.from_pretrained(pretrained_model_name_or_path, *inputs, **kwargs)
File "D:\Programs Files\Python\Python311\Lib\site-packages\transformers\tokenization_utils_base.py", line 1825, in from_pretrained
return cls._from_pretrained(
File "D:\Programs Files\Python\Python311\Lib\site-packages\transformers\tokenization_utils_base.py", line 1988, in _from_pretrained
tokenizer = cls(*init_inputs, **init_kwargs)
File "C:\Users\ZhaoYunFeng/.cache\huggingface\modules\transformers_modules\Muice\tokenization_chatglm.py", line 69, in init
super().init(padding_side=padding_side, clean_up_tokenization_spaces=False, **kwargs)
TypeError: transformers.tokenization_utils.PreTrainedTokenizer.init() got multiple values for keyword argument 'clean_up_tokenization_spaces'

复现该Bug的操作

tokenizer = AutoTokenizer.from_pretrained("D:\chatglm\models\Muice", trust_remote_code=True)
model = AutoModel.from_pretrained("D:\chatglm\models\Muice", trust_remote_code=True).cuda()

额外信息

国内大陆无法正常访问huggingface.co难以部署,是否可以将文件全部下载后手动部署启动(如果可以完成一键部署就更好啦 (大佬辛苦了!) 另外请问大佬是否有支持ntqq消息或者nonebot插件的想法呢

(ノಠ益ಠ)ノ彡┻━┻
（╯°□°）╯︵( .o.)
(ノ°益°)ノ彡
(ノಥ,_｣ಥ)ノ彡┻━┻
┻━┻ ︵ヽ(Д´)ﾉ︵ ┻━┻ (╯°□°）╯︵ ┻━┻ ლ(ಠ益ಠლ) (ノಠ ∩ಠ)ノ彡( o°o) (╯ರ ~ ರ）╯︵ ┻━┻ (╯°Д°）╯︵/(.□ . ) (╯°□°）╯︵ ┻━┻ (ノ｀⌒´)ノ┫：・’.：：・┻┻：・’.：：・ (╯°□°）╯︵ ┬─┻ ┻━┻ ︵ヽ(Д´)ﾉ︵ ┻━┻

这样难以跟踪bug问题

适当的上下文

无

出现问题的Prompt

你复活啦

出现问题的Answer

啊......没有吧，好像真的especially especially since she's an AI

期望的Answer

(至少是中文)

额外信息

使用2.2的glm2-6b

非常好工作！爱来自瓷器！（雾

Hello，在b站刷到了视频，然后有一点想法（

感觉16G显存需求对于大多数人还是有点过高了
不知道您有没有考虑过模型量化之类的事情，量化到4bit的话，显存需求对比于fp16的模型权重会有一半的下降
期待您的新版本！

输入代码：git clone https://github.com/Moemu/Muice-Chatbot
cd Muice-Chatbot
conda create --name Muice python=3.10.10
conda activate Muice
pip install -r requirements.txt
控制台：Downloading sentencepiece-0.1.99-cp310-cp310-win_amd64.whl (977 kB)
ERROR: Exception:
Traceback (most recent call last):
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_vendor\urllib3\response.py", line 438, in _error_catcher
yield
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_vendor\urllib3\response.py", line 561, in read
data = self._fp_read(amt) if not fp_closed else b""
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_vendor\urllib3\response.py", line 527, in _fp_read
return self._fp.read(amt) if amt is not None else self._fp.read()
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_vendor\cachecontrol\filewrapper.py", line 90, in read
data = self.__fp.read(amt)
File "C:\Users\HP.conda\envs\Muice\lib\http\client.py", line 465, in read
s = self.fp.read(amt)
File "C:\Users\HP.conda\envs\Muice\lib\socket.py", line 705, in readinto
return self._sock.recv_into(b)
File "C:\Users\HP.conda\envs\Muice\lib\ssl.py", line 1274, in recv_into
return self.read(nbytes, buffer)
File "C:\Users\HP.conda\envs\Muice\lib\ssl.py", line 1130, in read
return self._sslobj.read(len, buffer)
TimeoutError: The read operation timed out

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_internal\cli\base_command.py", line 180, in exc_logging_wrapper
status = run_func(*args)
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_internal\cli\req_command.py", line 248, in wrapper
return func(self, options, args)
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_internal\commands\install.py", line 377, in run
requirement_set = resolver.resolve(
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_internal\resolution\resolvelib\resolver.py", line 92, in resolve
result = self._result = resolver.resolve(
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_vendor\resolvelib\resolvers.py", line 546, in resolve
state = resolution.resolve(requirements, max_rounds=max_rounds)
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_vendor\resolvelib\resolvers.py", line 397, in resolve
self._add_to_criteria(self.state.criteria, r, parent=None)
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_vendor\resolvelib\resolvers.py", line 173, in _add_to_criteria
if not criterion.candidates:
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_vendor\resolvelib\structs.py", line 156, in bool
return bool(self._sequence)
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_internal\resolution\resolvelib\found_candidates.py", line 155, in bool
return any(self)
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_internal\resolution\resolvelib\found_candidates.py", line 143, in
return (c for c in iterator if id(c) not in self._incompatible_ids)
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_internal\resolution\resolvelib\found_candidates.py", line 47, in _iter_built
candidate = func()
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_internal\resolution\resolvelib\factory.py", line 206, in _make_candidate_from_link
self._link_candidate_cache[link] = LinkCandidate(
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_internal\resolution\resolvelib\candidates.py", line 293, in init
super().init(
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_internal\resolution\resolvelib\candidates.py", line 156, in init
self.dist = self._prepare()
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_internal\resolution\resolvelib\candidates.py", line 225, in _prepare
dist = self._prepare_distribution()
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_internal\resolution\resolvelib\candidates.py", line 304, in _prepare_distribution
return preparer.prepare_linked_requirement(self._ireq, parallel_builds=True)
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_internal\operations\prepare.py", line 538, in prepare_linked_requirement
return self._prepare_linked_requirement(req, parallel_builds)
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_internal\operations\prepare.py", line 609, in _prepare_linked_requirement
local_file = unpack_url(
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_internal\operations\prepare.py", line 166, in unpack_url
file = get_http_url(
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_internal\operations\prepare.py", line 107, in get_http_url
from_path, content_type = download(link, temp_dir.path)
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_internal\network\download.py", line 147, in call
for chunk in chunks:
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_internal\cli\progress_bars.py", line 53, in _rich_progress_bar
for chunk in iterable:
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_internal\network\utils.py", line 63, in response_chunks
for chunk in response.raw.stream(
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_vendor\urllib3\response.py", line 622, in stream
data = self.read(amt=amt, decode_content=decode_content)
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_vendor\urllib3\response.py", line 560, in read
with self._error_catcher():
File "C:\Users\HP.conda\envs\Muice\lib\contextlib.py", line 153, in exit
self.gen.throw(typ, value, traceback)
File "C:\Users\HP.conda\envs\Muice\lib\site-packages\pip_vendor\urllib3\response.py", line 443, in _error_catcher
raise ReadTimeoutError(self._pool, None, "Read timed out.")
pip._vendor.urllib3.exceptions.ReadTimeoutError: HTTPSConnectionPool(host='files.pythonhosted.org', port=443): Read timed out.

描述此Bug

运行时会在调用go-cqhttp后卡死在检查更新失败

控制台输出

PS C:\Gu\word\Muice-Chatbot> python main.py
在09.10更新中, 将使用(QQ号).json的方式来存储聊天记录, 对于重新拉取的源码， 可能会出现记忆缺失的情况, 对此, 请手动重命名memory下的chat_memory.json文件, 以便恢复记忆
Loading checkpoint shards: 100%|█████████████████████████████████████████████████████████| 7/7 [00:11<00:00, 1.61s/it]
Some weights of ChatGLMForConditionalGeneration were not initialized from the model checkpoint at model/chatglm2-6b and are newly initialized: ['transformer.prefix_encoder.embedding.weight']
You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.
[2023-09-17 12:34:18,746][MainThread/INFO] PyCqBot: 创建定时任务 CreateANewTopic

#################################################################
██████╗ ██╗ ██╗ ██████╗ ██████╗ ██████╗ ██████╗ ████████╗
██╔══██╗╚██╗ ██╔╝██╔════╝██╔═══██╗██╔══██╗██╔═══██╗╚══██╔══╝
██████╔╝ ╚████╔╝ ██║ ██║ ██║██████╔╝██║ ██║ ██║
██╔═══╝ ╚██╔╝ ██║ ██║▄▄ ██║██╔══██╗██║ ██║ ██║
██║ ██║ ╚██████╗╚██████╔╝██████╔╝╚██████╔╝ ██║
╚═╝ ╚═╝ ╚═════╝ ╚══▀▀═╝ ╚═════╝ ╚═════╝ ╚═╝
v0.5.1.1 BY FengLiu
#################################################################

[2023-09-17 12:34:19,184][go-cqhttp/WARNING] PyCqBot: go-cqhttp 警告 自动注 册实例已关闭，请配置 sign-server 端自动注册实例以保持正常签名
[2023-09-17 12:34:25,023][go-cqhttp/WARNING] PyCqBot: go-cqhttp 警告 Protocol -> device lock is disabled. HTTP API may fail.
[2023-09-17 12:34:27,818][go-cqhttp/WARNING] PyCqBot: go-cqhttp 警告 检查更 新失败！

复现该Bug的操作

仅需要按照正常流程部署于Win10

额外信息

使用22H1版本Win10，最新版cq-http Rlease,Cuda cuDNN均使用可在本电脑部署的最高版本，即12.1，qsign使用https://github.com/rhwong/unidbg-fetch-qsign-onekey/tree/main项目的一键部署，QQ版本8.9.63
另
感谢您的解答

[enhancement]作者大大可以加一个调用API吗？就是使用第三方代理（不用自己家电脑）