-
Notifications
You must be signed in to change notification settings - Fork 485
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
🙋 问题交流 - 图像识别不行 #1000
Comments
log:`2024-09-27 15:01:51.206 | INFO | main:http_api_thread:256 - HTTP API线程已启动! The above exception was the direct cause of the following exception: Traceback (most recent call last): The above exception was the direct cause of the following exception: Traceback (most recent call last): 2024-09-27 15:49:25.408 | ERROR | utils.common:send_request:1234 - Traceback (most recent call last): The above exception was the direct cause of the following exception: Traceback (most recent call last): The above exception was the direct cause of the following exception: urllib3.exceptions.ProxyError: ('Unable to connect to proxy', NewConnectionError('<urllib3.connection.HTTPConnection object at 0x000002263A8656F0>: Failed to establish a new connection: [WinError 10061] No connection could be made because the target machine actively refused it')) The above exception was the direct cause of the following exception: Traceback (most recent call last): During handling of the above exception, another exception occurred: Traceback (most recent call last): 2024-09-27 15:49:25.412 | ERROR | utils.common:send_request:1235 - 请求出错: HTTPConnectionPool(host='127.0.0.1', port=10809): Max retries exceeded with url: http://127.0.0.1:8082/send (Caused by ProxyError('Unable to connect to proxy', NewConnectionError('<urllib3.connection.HTTPConnection object at 0x000002263A8656F0>: Failed to establish a new connection: [WinError 10061] No connection could be made because the target machine actively refused it'))) |
谷歌代理连接失败
|
你确定你开了代理 并且端口正确吗 |
感谢您回答我的问题,我尝试了这个代理:http://guardian-antenna-finish-sublime.trycloudflare.com,但我遇到了这个问题,日志:`2024-09-27 16:23:26.108 | INFO | main:image_recognition_screenshot_and_send:6344 - 3.0后触发截图识别 The above exception was the direct cause of the following exception: Traceback (most recent call last): The above exception was the direct cause of the following exception: Traceback (most recent call last): 2024-09-27 16:24:24.221 | ERROR | utils.common:send_request:1234 - Traceback (most recent call last): 2024-09-27 16:24:24.227 | ERROR | utils.common:send_request:1235 - 请求出错: 403 Client Error: Forbidden for url: http://127.0.0.1:8082/send |
如果您可以添加 blip_image_captioning_large 或/和 OpenAI,那就太好了 |
代理不能代理 localhost,8082是程序内部HTTP服务,代理影响了数据发送 |
OpenAI 没有密钥(没经费),blip_image_captioning_large 没听说过,有空的话可以看看 |
谢谢你的回答,我不知道 Gemini 代理是如何工作的,如何在启动 (1.双击我启动程序.bat) 之前启动代理? |
如果你是在国外 可以直接访问gemini api,那就不需要启动代理。在“聊天”模式下,一键运行程序,然后再进行 摄像头识别操作 |
是的,我来自国外,可以直接访问 Gemini API,但是当我一键运行程序并尝试运行图像识别时,我遇到了第一个问题。 |
第一个问题日志说明 并没有关闭代理,导致请求发送失败,配置修改后需要保存重启 |
我应该删除 (127.0.0.1:10809) 并将其留空吗? |
log: `2024-09-27 17:16:37.333 | INFO | utils.common:capture_window_by_title:1523 - 截图已保存到:D:\AiRina\AI-Vtuber-20240913\AI-Vtuber-20240913\out\screenshots\Channel content - YouTube Studio - Opera.png The above exception was the direct cause of the following exception: Traceback (most recent call last): 2024-09-27 17:16:39.780 | ERROR | utils.common:send_request:1234 - Traceback (most recent call last): 2024-09-27 17:16:39.790 | ERROR | utils.common:send_request:1235 - 请求出错: 422 Client Error: Unprocessable Entity for url: http://127.0.0.1:8082/send` |
清空吧,不过我可能默认写死了走代理过 |
配置留空的情况下,gemini不会走代理过,可以尝试一下
|
非常感谢您的帮助,我尝试修改源代码并从 Gemini 中删除 http/https 代理,我还将模型更改为 gemini-1.5-flash,结果是 Gemini llm 运行良好,但image_recognition在 Gemini 和 Zhipu AI 上不起作用,我收到了这个日志。 `2024-09-29 02:37:27.596 | INFO | main:image_recognition_screenshot_and_send:6345 - 3.0后触发截图识别 2024-09-29 02:37:34.617 | ERROR | utils.common:send_request:1235 - 请求出错: 422 Client Error: Unprocessable Entity for url: http://127.0.0.1:8082/send |
|
这很奇怪,感觉是实例化失败了,我看看我这能不能测试一下吧 |
I looked at gemini changed a lot, and it is indeed difficult to restore and use it for a while. |
非常感谢您抽出时间关注这件事,如果有更好的选择,您可以添加,这将有很大帮助 |
zhipu ai 修改源码 模型为 glm-4v-plus 可以哦回复使用 |
我将型号更改为 glm-4v-plus,当我按下屏幕截图并发送时,我得到了此日志, `2024-09-29 17:02:27.034 | INFO | main:http_api_thread:256 - HTTP API线程已启动! 2024-09-29 17:02:43.910 | ERROR | utils.common:send_request:1234 - Traceback (most recent call last): 当我激活 (循环截图并发送) 时,我得到了这个日志, `2024-09-29 17:01:29.128 | INFO | main:http_api_thread:256 - HTTP API线程已启动! 2024-09-29 17:01:45.715 | WARNING | utils.my_handle:llm_handle:1555 - LLM没有正确返回数据,请排查配置、网络等是否正常。如果排查后都没有问题,可能是接口改动导致的兼容性问题,可以前往官方仓库提交issue,传送门:https://github.com/Ikaros-521/AI-Vtuber/issues |
我的天,还真的,奇怪了,好久没用动这块代码了,怎么变得莫名其妙了,我看看 |
#1004 |
非常感谢你,智普 AI image_recognition运行良好, 我现在正在尝试添加 (gpt-4-vision-preview),但直到现在都没有运气 也许这个链接会有所帮助: |
是的 其实我很早就想添加gpt的识别了,只不过 生活压力有点大 |
blip-image-captioning-large.zip 在过去的日子里,我试图添加(blip-image-captioning-large) 当你有时间的时候你能看看代码吗? |
Oh! This is fantastic! Thank you for your contribution to the project. I will proceed with the integration of this model in the future |
gpt.zip 当你有时间时,你可以检查代码。 |
已并入 main主线:#1011 |
非常感谢,blip-image-captioning 运行良好,你是最好的 |
操作系统
Windows
Python 版本
3.10.11
AI Vtuber 版本
20240913
描述问题
图像识别不行,Gemini和智普AI,能补充一下吗?像 open ai 和开源 blip_image_captioning_large 来进行图像识别?感谢您所做的出色工作。
(抱歉,我正在使用翻译器的语言)
复现步骤
图像识别不行,Gemini和智普AI,能补充一下吗?像 open ai 和开源 blip_image_captioning_large 来进行图像识别?感谢您所做的出色工作。
(抱歉,我正在使用翻译器的语言)
期望的结果
No response
截图或日志
No response
The text was updated successfully, but these errors were encountered: