Ben Rood
Ben Rood
> GGUF/exl2 + Qwen + multigpu should have been fixed in latest dev, please test if it works for you. Yes, it works now. Thank you for your great work.
> Exl2 is currently broken in the rc_054 branch. Please read the PR description at #481 When will exl2 be support again? or Should I stick to 0.5.3 until the...
> Hi folks, there are 2-3 options for this you might want to look into: > > * Reduce the `max_num_seqs` (set it to around 1) > * Reduce the...
> @bash99 nice catch! Since we have been working on providing the new chunker within [DS4SD/docling-core#68](https://github.com/DS4SD/docling-core/pull/68), and to speed things up, I added your fix directly there, giving author credit...
> It is a bug and the auther dont fix it，so bad！！！！ and looking my test: [#593](https://github.com/VikParuchuri/marker/issues/593) Does my workaround fix your issue?
And Please don't use fake data for sparse/tense vector. Just use some small embedding model to encode data.
我有点怀疑是连接到了国外的deepseek api接入点导致的，不知道cline能不能让我们之间配置deepseek baseurl的地址。
But when I try got last msg from chat session client.get_previous_messages(bot='claude-3.5-sonnet', chatId=chatId, count=1) the result msg is {'author': 'claude_3_igloo', 'text': '', 'messageId': xxxxxxx, 'contentType': 'text_markdown'} so the author is still...
> 好的，后面会添加一个参数来控制出错后的行为 或者来两个参数 1. 重试次数 2. 允许错误次数，将接口出错的回答直接当答错/无回答处理。
> I was able to narrow down the behavior. If I set `force_full_page_ocr=True` EasyOCR is extracting the text correctly (i.e., performing OCR). However, since this is a PDF that is...