SSL 连接错误导致任务失败及多个 FutureWarning

[Kamisato520]

您好，我在使用 MobileAgent-E 项目时，执行 bash scripts/run_task.sh 脚本后，任务失败并出现以下错误：

ERROR: HTTPSConnectionPool(host='dashscope.aliyuncs.com', port=443): Max retries exceeded with url: /api/v1/uploads?action=getPolicy&model=qwen-vl-plus (Caused by SSLError(SSLEOFError(8, '[SSL: UNEXPECTED_EOF_WHILE_READING] EOF occurred in violation of protocol (_ssl.c:1007)'))) 此错误表明在尝试连接到 dashscope.aliyuncs.com 时发生了 SSL 握手错误，导致连接失败。任务具体内容为：“Find me at least two survey papers on Large Language Models. Check the detailed abstract of the most cited one. And then create a new note in Notes and add the titles of the papers you found. Also include the abstract of the most cited paper.”

此外，日志中还包含多个警告和 FutureWarning，表明代码中存在一些即将弃用的功能和参数，建议在未来版本中进行更新以避免兼容性问题。以下是部分警告内容：

FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers UserWarning: Failed to load custom C++ ops. Running on CPU mode Only! modelscope - WARNING: ('PIPELINES', 'grounding-dino-task', 'Groundingdino-generation-pipe') not found in ast index file UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. modelscope - WARNING: No preprocessor field found in cfg. FutureWarning: The device argument is deprecated and will be removed in v5 of Transformers. UserWarning: torch.utils.checkpoint: the use_reentrant parameter should be passed explicitly. UserWarning: None of the inputs have requires_grad=True. Gradients will be None FutureWarning: torch.cuda.amp.autocast(args...) is deprecated. SupervisionWarnings: annotate is deprecated: BoxAnnotator is deprecated and will be removed in supervision-0.22.0. 环境信息 项目：MobileAgent-E 脚本：scripts/run_task.sh 模型：qwen-vl-plus 操作系统：Windows (MINGW64) Python 环境：mobile_agent_e 尝试的解决方法 确认网络连接正常，能够访问 dashscope.aliyuncs.com。 更新 requests 和 urllib3 库。 检查 SSL/TLS 库版本。 然而，问题依然存在。希望项目开发者能够提供关于此 SSL 错误的具体原因和解决方案，并考虑更新代码以消除 FutureWarning。 另外可以麻烦您们提供代码运行的演示视频吗？谢谢

[junyangwang0410]

您好，方便提供更详细的报错信息吗？包括错误堆栈和原始的错误输出

[Kamisato520]

好的，下面是原始的错误输出： 在执行 bash scripts/run_task.sh 脚本时，出现了以下错误输出： ERROR: HTTPSConnectionPool(host='dashscope.aliyuncs.com', port=443): Max retries exceeded with url: /api/v1/uploads?action=getPolicy&model=qwen-vl-plus (Caused by SSLError(SSLEOFError(8, '[SSL: UNEXPECTED_EOF_WHILE_READING] EOF occurred in violation of protocol (_ssl.c:1007)'))) 错误类型：SSLError(SSLEOFError) 具体原因：在与 dashscope.aliyuncs.com（端口 443）建立 HTTPS 连接时，SSL 握手过程中发生了意外的 EOF（文件结束），违反了协议规范。 影响：由于多次重试失败，最终触发了 Max retries exceeded 错误，导致任务无法完成。 任务描述： 错误发生在执行以下任务时： Failed when doing task: Find me at least two survey papers on Large Language Models. Check the detailed abstract of the most cited one. And then create a new note in Notes and add the titles of the papers you found. Also include the abstract of the most cited paper.

我自行分析觉得有下面的问题： 网络错误：程序无法访问某个 API（可能是 dashscope.aliyuncs.com），返回了 HTTP 400 错误。 程序错误：任务失败后，抛出了 Python 异常 'NoneType' object has no attribute 'split'。 我测试了使用VPN和关闭vpn后依旧出现一样的报错。 为了方便您查看，我会把运行bash命令后的所有终端输出都放在下面：

$ bash scripts/run_task.sh

Using BACKBONE_TYPE: Gemini

INFO: Initial shortcuts: {'Tap_Type_and_Enter': {'name': 'Tap_Type_and_Enter', 'arguments': ['x', 'y', 'text'], 'description': 'Tap an input box at position (x, y), Type the "text", and then perform the Enter operation. Very useful for searching and sending messages!', 'precondition': 'There is a text input box on the screen with no previously entered content.', 'atomic_action_sequence': [{'name': 'Tap', 'arguments_map': {'x': 'x', 'y': 'y'}}, {'name': 'Type', 'arguments_map': {'text': 'text'}}, {'name': 'Enter', 'arguments_map': {}}]}} INFO: Initial tips: 0. Do not add any payment information. If you are asked to sign in, ignore it or sign in as a guest if possible. Close any pop-up windows when opening an app.

1. By default, no APPs are opened in the background.
2. Screenshots may show partial text in text boxes from your previous input; this does not count as an error.
3. When creating new Notes, you do not need to enter a title unless the user specifically requests it.

Add your custom tips here

Downloading Model from https://www.modelscope.cn to directory: C:\Users\86159.cache\modelscope\hub\models\AI-ModelScope\GroundingDINO 2025-03-22 13:30:31,995 - modelscope - INFO - Use user-specified model revision: v1.0.0 2025-03-22 13:30:32,289 - modelscope - WARNING - Use allow_remote=True. Will invoke codes from C:\Users\86159.cache\modelscope\hub\models\AI-ModelScope\GroundingDINO. Please make sure that you can trust the external codes. G:\miniconda\envs\mobile_agent_e\lib\site-packages\timm\models\layers_init_.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers warnings.warn(f"Importing from {name} is deprecated, please import via timm.layers", FutureWarning) C:\Users\86159.cache\modelscope\hub\modelscope_modules\GroundingDINO\groundingdino\models\GroundingDINO\ms_deform_attn.py:33: UserWarning: Failed to load custom C++ ops. Running on CPU mode Only! warnings.warn('Failed to load custom C++ ops. Running on CPU mode Only!') UserWarning: Failed to load custom C++ ops. Running on CPU mode Only! 2025-03-22 13:30:33,086 - modelscope - WARNING - ('PIPELINES', 'grounding-dino-task', 'Groundingdino-generation-pipe') not found in ast index file UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\native\TensorShape.cpp:3638.) final text_encoder_type: C:\Users\86159.cache\modelscope\hub\models\AI-ModelScope\GroundingDINO/bert-base-uncased 2025-03-22 13:30:34,387 - modelscope - WARNING - No preprocessor field found in cfg. 2025-03-22 13:30:34,387 - modelscope - WARNING - No val key and type key found in preprocessor domain of configuration.json file. 2025-03-22 13:30:34,387 - modelscope - WARNING - Cannot find available config to build preprocessor at mode inference, current config: {'model_dir': 'C:\Users\86159\.cache\modelscope\hub\models\AI-ModelScope\GroundingDINO'}. trying to build by task and model information. 2025-03-22 13:30:34,387 - modelscope - WARNING - No preprocessor key ('Groundingdino', 'grounding-dino-task') found in PREPROCESSOR_MAP, skip building preprocessor. 2025-03-22 13:30:35,047 - modelscope - WARNING - Model revision not specified, use revision: v1.3.0 Downloading Model from https://www.modelscope.cn to directory: C:\Users\86159.cache\modelscope\hub\models\iic\cv_resnet18_ocr-detection-db-line-level_damo 2025-03-22 13:30:35,957 - modelscope - WARNING - Model revision not specified, use revision: v1.3.0 2025-03-22 13:30:36,267 - modelscope - INFO - initiate model from C:\Users\86159.cache\modelscope\hub\models\iic\cv_resnet18_ocr-detection-db-line-level_damo 2025-03-22 13:30:36,268 - modelscope - INFO - initiate model from location C:\Users\86159.cache\modelscope\hub\models\iic\cv_resnet18_ocr-detection-db-line-level_damo. 2025-03-22 13:30:36,276 - modelscope - INFO - initialize model from C:\Users\86159.cache\modelscope\hub\models\iic\cv_resnet18_ocr-detection-db-line-level_damo 2025-03-22 13:30:36,590 - modelscope - INFO - loading model from dir C:\Users\86159.cache\modelscope\hub\models\iic\cv_resnet18_ocr-detection-db-line-level_damo 2025-03-22 13:30:36,725 - modelscope - INFO - loading model done 2025-03-22 13:30:37,486 - modelscope - WARNING - Model revision not specified, use revision: v2.4.0 Downloading Model from https://www.modelscope.cn to directory: C:\Users\86159.cache\modelscope\hub\models\iic\cv_convnextTiny_ocr-recognition-document_damo 2025-03-22 13:30:38,277 - modelscope - WARNING - Model revision not specified, use revision: v2.4.0 2025-03-22 13:30:38,566 - modelscope - INFO - initiate model from C:\Users\86159.cache\modelscope\hub\models\iic\cv_convnextTiny_ocr-recognition-document_damo 2025-03-22 13:30:38,568 - modelscope - INFO - initiate model from location C:\Users\86159.cache\modelscope\hub\models\iic\cv_convnextTiny_ocr-recognition-document_damo. 2025-03-22 13:30:38,571 - modelscope - INFO - initialize model from C:\Users\86159.cache\modelscope\hub\models\iic\cv_convnextTiny_ocr-recognition-document_damo 2025-03-22 13:30:38,895 - modelscope - INFO - loading model from dir C:\Users\86159.cache\modelscope\hub\models\iic\cv_convnextTiny_ocr-recognition-document_damo 2025-03-22 13:30:38,944 - modelscope - INFO - loading model done INFO: Loaded perception models: - Caption model method: api | caption vlm model: qwen-vl-plus - Grounding DINO model: <GroundingDINO.ms_wrapper.GroundingdinoGenerationPipeline object at 0x000002CA6446E500> - OCR detection model: iic/cv_resnet18_ocr-detection-db-line-level_damo - OCR recognition model: iic/cv_convnextTiny_ocr-recognition-document_damo

Perceptor ...

2025-03-22 13:30:42,433 - modelscope - WARNING - task grounding-dino-task input definition is missing FutureWarning: The device argument is deprecated and will be removed in v5 of Transformers. UserWarning: torch.utils.checkpoint: the use_reentrant parameter should be passed explicitly. In version 2.5 we will raise an exception if use_reentrant is not passed. use_reentrant=False is recommended, but if you need to preserve the current default behavior, you can pass use_reentrant=True. Refer to docs for more details on the differences between the two variants. UserWarning: None of the inputs have requires_grad=True. Gradients will be None FutureWarning: torch.cuda.amp.autocast(args...) is deprecated. Please use torch.amp.autocast('cuda', args...) instead. SupervisionWarnings: annotate is deprecated: BoxAnnotator is deprecated and will be removed in supervision-0.22.0. Use BoundingBoxAnnotator and LabelAnnotator instead 2025-03-22 13:30:47,679 - modelscope - WARNING - task grounding-dino-task output keys are missing Perception Infos: [{'text': 'text: 伞', 'coordinates': [160, 34]}, {'text': 'text: 3G', 'coordinates': [1142, 35]}, {'text': 'text: 5:30', 'coordinates': [64, 35]}, {'text': 'text: Sat, Mar 22', 'coordinates': [272, 302]}, {'text': 'text: Play Store', 'coordinates': [201, 2051]}, {'text': 'text: YouTube', 'coordinates': [1139, 2049]}, {'text': 'text: Gmail', 'coordinates': [515, 2050]}, {'text': 'text: Photos', 'coordinates': [828, 2051]}, {'text': 'text: 目', 'coordinates': [516, 2461]}, {'text': 'icon: The Google Photos app logo has four colored triangles arranged like a pinwheel, with each triangle being a different shade of green or blue. The background behind it is white.', 'coordinates': [826, 1898]}, {'text': 'icon: The Google Chrome logo, which consists of three green triangles surrounding a yellow circle with a blue dot inside it all on top of a larger dark green circle.', 'coordinates': [826, 2455]}, {'text': 'icon: The Gmail app logo on my home screen has a white background with two overlapping letters "M" colored red, yellow, blue, green respectively.', 'coordinates': [514, 1898]}, {'text': 'icon: The YouTube logo has a red play button inside a white circle with a pink border on a light blue background.', 'coordinates': [1138, 1899]}, {'text': 'icon: The app logo is a triangle with rounded corners, colored pink on top and blue at the bottom.', 'coordinates': [200, 1899]}, {'text': 'icon: The app logo is a white text bubble with three dots on it, enclosed by a blue circle.', 'coordinates': [514, 2455]}, {'text': 'icon: The icon is shaped like a speech bubble with a blue background, white border, and yellow text that says "Google."', 'coordinates': [512, 2456]}, {'text': 'icon: This is an icon.', 'coordinates': [192, 2723]}, {'text': 'icon: The icon is shaped like a microphone with rainbow colors on it, including red, blue, yellow, green, purple, orange, and pink.', 'coordinates': [670, 2724]}, {'text': 'icon: The microphone icon on the phone screen has a rounded square shape with black lines outlining it, and a red circle filled with white dots at its center.', 'coordinates': [1150, 2723]}, {'text': 'icon: The icon is a square with rounded corners, filled with white lines on a blue background to represent a memory card.', 'coordinates': [225, 34]}, {'text': 'icon: The logo for Gmail, which looks like a white capital letter "M" with blue outline and filled with rainbow colors.', 'coordinates': [516, 1899]}, {'text': 'icon: The app logo is a circle with three colored triangles inside, representing playfulness and movement.', 'coordinates': [205, 1900]}, {'text': 'icon: The icon is shaped like a gear, colored light gray with darker gray outlines.', 'coordinates': [159, 34]}, {'text': 'icon: The icon is a square with rounded corners, filled with white or light gray color.', 'coordinates': [1057, 2919]}, {'text': 'icon: This is an icon.', 'coordinates': [279, 2917]}, {'text': 'icon: The icon is circular, white with black border, and has two intersecting diagonal lines inside it.', 'coordinates': [670, 2918]}, {'text': 'icon: The time icon on the phone screen has a white background with black text that reads "5:30".', 'coordinates': [64, 36]}, {'text': 'icon: The battery icon is rectangular with rounded corners, filled with white lines to represent different levels of charge.', 'coordinates': [1246, 36]}, {'text': 'icon: The icon is shaped like a square with rounded corners, and it has white text on a blue background that says "Sat, Mar 22".', 'coordinates': [272, 301]}, {'text': 'icon: The icon is a white square with a black exclamation mark inside it, indicating urgency or importance.', 'coordinates': [1197, 36]}, {'text': 'icon: This is an icon.', 'coordinates': [1137, 34]}, {'text': 'icon: The icon is a white triangle with a red border, inside it's black text that says "3G".', 'coordinates': [1185, 35]}, {'text': 'icon: The icon is a circle with a white background, black border, and red "1" inside it.', 'coordinates': [559, 1901]}, {'text': 'icon: The Google app logo on my home screen has a blue circle with white lettering that says "Google" inside it, surrounded by green, yellow, red, and orange stripes.', 'coordinates': [825, 2457]}, {'text': 'icon: This is an icon.', 'coordinates': [1138, 1897]}]

Manager ...

Network Error: Request Failed Sleep 20 before retry... Network Error: Request Failed Sleep 20 before retry... Network Error: Request Failed Sleep 20 before retry... Network Error: Request Failed Sleep 20 before retry... Network Error: Request Failed Sleep 20 before retry... Network Error: [{'error': {'code': 400, 'message': 'User location is not supported for the API use.', 'status': 'FAILED_PRECONDITION'}}] Sleep 20 before retry... Failed after -1 retries... Failed when doing task: Find me at least two survery papers on Large Language Models. Check the detailed abstract of the most cited one. And then create a new note in Notes and add the titles the papers you found. Also include the abstract of the most cited paper. ERROR: 'NoneType' object has no attribute 'split'

[junyangwang0410]

User location is not supported for the API use. 推测应该是GPT-4o暂时不支持从中国IP调用，可以尝试挂🪜

[Kamisato520]

您好，我在挂载梯子并更换节点（如美国，HK，TW，新加坡等）反复尝试后出现了下面的错误： 自定义 C++ 操作加载失败 日志中多次提示“Failed to load custom C++ ops. Running on CPU mode Only!”，这表明程序未能加载预期的自定义 C++ 加速代码，导致只能以 CPU 模式运行，可能影响性能。但是我测试后发现我的cuda和安装的torch的版本是一致的，cuda也是可使用的，另外还有我的ssl的version查看： $ python -c "import torch; print(torch.version); print(torch.version.cuda)" 2.6.0+cu126 12.6 $ python -c "import ssl; print(ssl.OPENSSL_VERSION)" OpenSSL 3.0.16 11 Feb 2025 $ python -c "import torch; print(torch.cuda.is_available())" True

配置相关警告

日志显示没有找到预处理器（preprocessor）相关的字段和配置，例如“No preprocessor field found in cfg”、“No val key and type key found in preprocessor domain of configuration.json file”以及“No preprocessor key ('Groundingdino', 'grounding-dino-task') found in PREPROCESSOR_MAP, skip building preprocessor”。这可能意味着任务所需的预处理配置不完整或未正确定义。

同时，还出现了针对任务“grounding-dino-task”的输入定义和输出 key 缺失的警告：“task grounding-dino-task input definition is missing”以及“task grounding-dino-task output keys are missing”。这会导致任务在执行时无法正确解析输入和输出，可能直接影响任务执行结果。

SSL 连接错误 最后，最关键的错误出现在任务执行过程中：

ERROR: HTTPSConnectionPool(host='dashscope.aliyuncs.com', port=443): Max retries exceeded with url: /api/v1/uploads?action=getPolicy&model=qwen-vl-plus (Caused by SSLError(SSLEOFError(8, '[SSL: UNEXPECTED_EOF_WHILE_READING] EOF occurred in violation of protocol (_ssl.c:1007)'))) 这说明在与 dashscope.aliyuncs.com 建立 HTTPS 连接时，由于 SSL 协议异常（意外的 EOF），连接重试次数超限，导致任务失败。这通常与网络安全设置、SSL 协议版本或服务器端配置有关。

下面是完整的终端反馈： $ bash scripts/run_task.sh

Using BACKBONE_TYPE: Gemini

INFO: Initial shortcuts: {'Tap_Type_and_Enter': {'name': 'Tap_Type_and_Enter', 'arguments': ['x', 'y', 'text'], 'description': 'Tap an input box at position (x, y), Type the "text", and then perform the Enter operation. Very useful for searching and sending messages!', 'precondition': 'There is a text input box on the screen with no previously entered content.', 'atomic_action_sequence': [{'name': 'Tap', 'arguments_map': {'x': 'x', 'y': 'y'}}, {'name': 'Type', 'arguments_map': {'text': 'text'}}, {'name': 'Enter', 'arguments_map': {}}]}} INFO: Initial tips: 0. Do not add any payment information. If you are asked to sign in, ignore it or sign in as a guest if possible. Close any pop-up windows when opening an app.

1. By default, no APPs are opened in the background.
2. Screenshots may show partial text in text boxes from your previous input; this does not count as an error.
3. When creating new Notes, you do not need to enter a title unless the user specifically requests it.

Add your custom tips here

Downloading Model from https://www.modelscope.cn to directory: C:\Users\86159.cache\modelscope\hub\models\AI-ModelScope\GroundingDINO 2025-03-22 19:37:16,776 - modelscope - INFO - Use user-specified model revision: v1.0.0 2025-03-22 19:37:17,238 - modelscope - WARNING - Use allow_remote=True. Will invoke codes from C:\Users\86159.cache\modelscope\hub\models\AI-ModelScope\GroundingDINO. Please make sure that you can trust the external codes. G:\miniconda\envs\mobile_agent_e\lib\site-packages\timm\models\layers_init_.py:48: FutureWarning: Importing from timm.models.layers is deprecated, please import via timm.layers warnings.warn(f"Importing from {name} is deprecated, please import via timm.layers", FutureWarning) C:\Users\86159.cache\modelscope\hub\modelscope_modules\GroundingDINO\groundingdino\models\GroundingDINO\ms_deform_attn.py:33: UserWarning: Failed to load custom C++ ops. Running on CPU mode Only! warnings.warn('Failed to load custom C++ ops. Running on CPU mode Only!') UserWarning: Failed to load custom C++ ops. Running on CPU mode Only! 2025-03-22 19:37:17,940 - modelscope - WARNING - ('PIPELINES', 'grounding-dino-task', 'Groundingdino-generation-pipe') not found in ast index file UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at C:\actions-runner_work\pytorch\pytorch\pytorch\aten\src\ATen\native\TensorShape.cpp:3638.) final text_encoder_type: C:\Users\86159.cache\modelscope\hub\models\AI-ModelScope\GroundingDINO/bert-base-uncased 2025-03-22 19:37:18,892 - modelscope - WARNING - No preprocessor field found in cfg. 2025-03-22 19:37:18,892 - modelscope - WARNING - No val key and type key found in preprocessor domain of configuration.json file. 2025-03-22 19:37:18,892 - modelscope - WARNING - Cannot find available config to build preprocessor at mode inference, current config: {'model_dir': 'C:\Users\86159\.cache\modelscope\hub\models\AI-ModelScope\GroundingDINO'}. trying to build by task and model information. 2025-03-22 19:37:18,892 - modelscope - WARNING - No preprocessor key ('Groundingdino', 'grounding-dino-task') found in PREPROCESSOR_MAP, skip building preprocessor. 2025-03-22 19:37:19,649 - modelscope - WARNING - Model revision not specified, use revision: v1.3.0 Downloading Model from https://www.modelscope.cn to directory: C:\Users\86159.cache\modelscope\hub\models\iic\cv_resnet18_ocr-detection-db-line-level_damo 2025-03-22 19:37:20,479 - modelscope - WARNING - Model revision not specified, use revision: v1.3.0 2025-03-22 19:37:20,755 - modelscope - INFO - initiate model from C:\Users\86159.cache\modelscope\hub\models\iic\cv_resnet18_ocr-detection-db-line-level_damo 2025-03-22 19:37:20,755 - modelscope - INFO - initiate model from location C:\Users\86159.cache\modelscope\hub\models\iic\cv_resnet18_ocr-detection-db-line-level_damo. 2025-03-22 19:37:20,758 - modelscope - INFO - initialize model from C:\Users\86159.cache\modelscope\hub\models\iic\cv_resnet18_ocr-detection-db-line-level_damo 2025-03-22 19:37:20,897 - modelscope - INFO - loading model from dir C:\Users\86159.cache\modelscope\hub\models\iic\cv_resnet18_ocr-detection-db-line-level_damo 2025-03-22 19:37:21,029 - modelscope - INFO - loading model done 2025-03-22 19:37:21,793 - modelscope - WARNING - Model revision not specified, use revision: v2.4.0 Downloading Model from https://www.modelscope.cn to directory: C:\Users\86159.cache\modelscope\hub\models\iic\cv_convnextTiny_ocr-recognition-document_damo 2025-03-22 19:37:22,750 - modelscope - WARNING - Model revision not specified, use revision: v2.4.0 2025-03-22 19:37:23,094 - modelscope - INFO - initiate model from C:\Users\86159.cache\modelscope\hub\models\iic\cv_convnextTiny_ocr-recognition-document_damo 2025-03-22 19:37:23,094 - modelscope - INFO - initiate model from location C:\Users\86159.cache\modelscope\hub\models\iic\cv_convnextTiny_ocr-recognition-document_damo. 2025-03-22 19:37:23,097 - modelscope - INFO - initialize model from C:\Users\86159.cache\modelscope\hub\models\iic\cv_convnextTiny_ocr-recognition-document_damo 2025-03-22 19:37:23,276 - modelscope - INFO - loading model from dir C:\Users\86159.cache\modelscope\hub\models\iic\cv_convnextTiny_ocr-recognition-document_damo 2025-03-22 19:37:23,313 - modelscope - INFO - loading model done INFO: Loaded perception models: - Caption model method: api | caption vlm model: qwen-vl-plus - Grounding DINO model: <GroundingDINO.ms_wrapper.GroundingdinoGenerationPipeline object at 0x000001A773C7E500> - OCR detection model: iic/cv_resnet18_ocr-detection-db-line-level_damo - OCR recognition model: iic/cv_convnextTiny_ocr-recognition-document_damo

Perceptor ...

2025-03-22 19:37:27,674 - modelscope - WARNING - task grounding-dino-task input definition is missing FutureWarning: The device argument is deprecated and will be removed in v5 of Transformers. UserWarning: torch.utils.checkpoint: the use_reentrant parameter should be passed explicitly. In version 2.5 we will raise an exception if use_reentrant is not passed. use_reentrant=False is recommended, but if you need to preserve the current default behavior, you can pass use_reentrant=True. Refer to docs for more details on the differences between the two variants. UserWarning: None of the inputs have requires_grad=True. Gradients will be None FutureWarning: torch.cuda.amp.autocast(args...) is deprecated. Please use torch.amp.autocast('cuda', args...) instead. SupervisionWarnings: annotate is deprecated: BoxAnnotator is deprecated and will be removed in supervision-0.22.0. Use BoundingBoxAnnotator and LabelAnnotator instead 2025-03-22 19:37:31,462 - modelscope - WARNING - task grounding-dino-task output keys are missing Failed when doing task: Find me at least two survery papers on Large Language Models. Check the detailed abstract of the most cited one. And then create a new note in Notes and add the titles the papers you found. Also include the abstract of the most cited paper. ERROR: HTTPSConnectionPool(host='dashscope.aliyuncs.com', port=443): Max retries exceeded with url: /api/v1/uploads?action=getPolicy&model=qwen-vl-plus (Caused by SSLError(SSLEOFError(8, '[SSL: UNEXPECTED_EOF_WHILE_READING] EOF occurred in violation of protocol (_ssl.c:1007)')))

[junyangwang0410]

感谢反馈，我想您的情况可能遇到了如下情况： 使用🪜后的IP或原本的IP会导致GPT和Qwen只能调用其中一个而无法调用另一个。 这种情况下，建议自行部署qwen模型，它的功能较为单一，仅为图标描述工具，可以尝试qwen-2.5-2b模型

[Kamisato520]

好的，我会尝试自行部署，感谢您的回复。另外可以透露一些v3的信息吗？比如多久上线开源？能不能全部支持qwen系列的模型？可以考虑有专门的软件优化吗？谢谢