落叶乌龟

[论坛讨论贴](https://rustcc.cn/article?id=56332374-4b13-415d-814a-a98e291ca143) 修改方案就是在原评论的每一行添加一个">"符号。 注意❌：**功能未测试**。因为我测试时本地拉取第三方的sapper_std v0.2.1和html5ever v0.23.0编译直接失败。

BaiduNetDisk is bad for sharing file because it's speed limitation and deep bind with desktop client. It's good if you upload as a release of the repo.

feat(query): Make `BooleanQuery` supports `minimum_number_should_match`

9

see issue #2398 In this commit, a novel scorer named DisjunctionScorer is introduced, which performs the union of inverted chains with the minimal required elements. BTW, it's implemented via a...

**Is your feature request related to a problem? Please describe.** As [issue 2508](https://github.com/quickwit-oss/tantivy/issues/2508) mentioned, `QueryParser` in Tantivy cannot handle pattern like "name:*", although `ExistsQuery` are supported already. **Describe the solution...

# ChangeLog * Add entries in `LogicalLiteral` and `LogicalAst` to support `ExistsQuery` pattern.

这个 deepmatch.utils.sampledsoftmaxloss 的实现是有意为之吗？还是说仅仅是个错误

2

在调用这个 Loss 进行训练时，发现完全不收敛。翻看源码才发现其实现如下： [deepmatch/utils.py#](https://github.com/shenweichen/DeepMatch/blob/93ca4371386223e2ae0b396b8860f363bd6abf67/deepmatch/utils.py#L61C1-L63C1) ```python def sampledsoftmaxloss(y_true, y_pred): return K.mean(y_pred) ``` 如果这是个失误，我可以提一个 PR 来修复它（用 tensorflow 中带的 loss）

I connect to my Arch Linux PC which runs on a PVE platform, and it's cpu frequency works abnormally.  ```bash lebrancebw@tinker-arch ~> neofetch -` lebrancebw@tinker-arch .o+` ---------------------- `ooo/ OS:...