zapqqqwe

Results 6 comments of zapqqqwe

Problem when executing Bard().get_answer(...)

Response Error: b')]}\'\n\n38\n[["wrb.fr",null,null,null,null,[8]]]\n54\n[["di",75],["af.httprm",75,"6198447553110271097",8]]\n25\n[["e",4,null,null,129]]\n'. [] i use the code. bard = Bard(token=token,proxies={'http':'http://127.0.0.1:7890', 'https':'http://127.0.0.1:7890'}, timeout=10) however i still meet the error,what should i do?

Problem when executing Bard().get_answer(...)

i know .thank you.   ... ***@***.***   ------------------ 原始邮件 ------------------ 发件人: "Daniel ***@***.***>; 发送时间: 2023年5月24日(星期三) 晚上10:37 收件人: ***@***.***>; 抄送: ***@***.***>; ***@***.***>; 主题: Re: [dsdanielpark/Bard-API] Problem when executing Bard().get_answer(...) (Issue #1) chavarera...

Problem when executing Bard().get_answer(...)

have no idea.google limit the times to use . ... ***@***.***   ------------------ 原始邮件 ------------------ 发件人: "dsdanielpark/Bard-API" ***@***.***>; 发送时间: 2023年6月4日(星期天) 凌晨3:56 ***@***.***>; ***@***.******@***.***>; 主题: Re: [dsdanielpark/Bard-API] Problem when executing Bard().get_answer(...) (Issue #1) This error...

error:RuntimeError: Error(s) in loading state_dict for CLIPVisionModel: size mismatch for vision_model.embeddings.class_embedding: copying a param with shape torch.Size([1024]) from checkpoint, the shape in current model is torch.Size([768]).

huggface里面的LanguageBind/LanguageBind_Video_merge 和 LanguageBind/LanguageBind_Image

不用flash attention推理结果异常

+1 同样的问题

Open AI接口Qwen3Vl做多帧图像视频理解

我是用的下面的方案解决的，应该也是可以的： selected_frames = b64_frames # 3. 构造 messages：每帧一个 image_url，最后加上 text messages = [ { "role": "user", "content": [ *map( lambda x: { "type": "image_url", "image_url": { "url": f"data:image/jpg;base64,{x}", # 或者...