jingwangsg

Results 9 comments of jingwangsg

> same problem It's similar to #4094 > > 1. I modify the `num_experts_per_tok` to 8 in `config.json` of mixtral-8x7b and everything works well. So I think the problem is...

@favourhong Hi, 这款主题简直太棒了,作为科研人做一些简洁的ppt不再需要纠结排版。 这边在obsidian中的preview的页面基本正常,只是这边在obsidian会出现导出为pdf样式全部丢失的情况。不知道大佬有没有什么想法可以解决

> I think @K-Nick is suggesting this: Clicking on a node in the mindmap moves the text editor to that part of the markdown document. Exactly. Thank you for your...

> Hello! > > Thanks for the request. > > I'm yet to test this viewer (and [term-image](https://github.com/AnonymouX47/term-image), the library it depends on) on [xterm-addon-image](https://github.com/jerch/xterm-addon-image) which the VS Code terminal...

@LuigiNixy Hi, can you still use pytubefix to massively download videos now?

[This line](https://github.com/huggingface/transformers/blob/9f06fb05059a973048f5865e7e385c9db5d6daa4/src/transformers/models/t5/modeling_t5.py#L779) is the root cause for fp32 conversion. And it looks like it's only affecting fp16, due to the [internal logic](https://github.com/huggingface/transformers/blob/9f06fb05059a973048f5865e7e385c9db5d6daa4/src/transformers/modeling_utils.py#L4056-L4059) to handle `_keep_in_fp32_modules` A quick fix is to...

Got it — so each “value” will be recovered and placed on the CPU. coredumpy just tracks these values, rather than the full context (e.g., the original device where each...

点击到整个library文件夹时会严重卡顿(可能是库里的文献太多) 可能的两个解决思路 1. 仅渲染存在连接的节点 2. 当在library文件夹时仅进行以特定paper为中心的局部子图展示

> @K-Nick > > > 仅渲染存在连接的节点 > > 我觉得渲染不会占用太多资源,卡顿可能来自所有条目读取分析。 > > > 当在library文件夹时仅进行以特定paper为中心的局部子图展示 > > 在没有优化关系图谱之前,可以选择性打开它(我平时都不开,都快把它忘了) 等一波优化👍🏻(虽然我平常也不开)