[WeeklyReport] gouzil 2023.11.08~2023.11.21

WeeklyReports/22_gouzil/[WeeklyReport]2023.11.08~2023.11.21.md

@@ -0,0 +1,95 @@
+### 姓名
+
+田川
+
+### 实习项目
+
+PIR 动转静组件建设与单测验证推全
+
+### 本周工作
+
+1. PIR 动转静组件单测问题排查
+
+接下来几周主要做问题排查部分
+
+相关pr: 
+
+* [#58630](https://github.com/PaddlePaddle/Paddle/pull/58630)
+* [#58686](https://github.com/PaddlePaddle/Paddle/pull/58686)
+* [#58890](https://github.com/PaddlePaddle/Paddle/pull/58890)
+* [#58936](https://github.com/PaddlePaddle/Paddle/pull/58936)
+* [#58965](https://github.com/PaddlePaddle/Paddle/pull/58965)
+* [#59007](https://github.com/PaddlePaddle/Paddle/pull/59007)
+* [#59016](https://github.com/PaddlePaddle/Paddle/pull/59016)
+
+遇到的一些问题:
+
+* `generator`无法切换问题
+我们在拿掉`with base.dygraph.guard()`时, 会遇到`name`的`generator`无法切换的问题, 报错信息如下:
+```bash
+ValueError: parameter name [Base_1_w] have be been used. In dygraph mode, the name of parameter can't be same.Please check the parameter attr value passed to self.create_parameter or constructor of dygraph Layers
+```
+
+解决方案:
+
+我们翻一下`guard`api 源码, 发现他其实在切换模式的同时调用了`framework.unique_name.guard`api
+```python
+def guard(place=None):
+    train = framework.Program()
+    startup = framework.Program()
+    tracer = Tracer()
+
+    if place is not None:
+        expected_place = _get_paddle_place(place)
+    else:
+        expected_place = framework._current_expected_place_()
+
+    with framework.program_guard(train, startup):
+        with framework.unique_name.guard():
+            with framework._dygraph_guard(tracer):
+                with framework._dygraph_place_guard(expected_place):
+                    yield
+```
+这时候我们只需要跟他做一样的操作就可以了`with unique_name.guard()`, 解决pr：[#58936](https://github.com/PaddlePaddle/Paddle/pull/58936/files#diff-97d0698d886bc904d5233448b2de3b1499e78cd1dc68d609b1a00d741ad46a90)
+
+
+* `device`切换问题
+
+在`setUp`中`paddle.set_device`会导致切换无效
+
+原因:
+
+`dygraph_tracer` 被换掉了，因为 `setUp` 后会走我们的 `to_xxx`，这个时候可能切动态图静态图，会换掉 `dygraph_tracer`，所以 `_expected_place` 就白设置了
+
+可以查看一下`python/paddle/base/framework.py`的写法
+
+```python
+def _set_dygraph_tracer_expected_place(place):
+    if global_var._dygraph_tracer_ is not None:
+        global_var._dygraph_tracer_._expected_place = place
+
+
+def _set_expected_place(place):
+    global _global_expected_place_
+    _global_expected_place_ = place
+    _set_dygraph_tracer_expected_place(place)
+```
+
+修改方案:
+
+将`paddle.set_device`移动到`test_xxx`内部
+
+pr: [#59120](https://github.com/PaddlePaddle/Paddle/pull/59120/files#diff-b71242c321499d35ff39b19d4ee34c05a1168c7eed504bf5f44c9e0fc7dbb3d6)
+
+2. 适配`OpResult.clone()`
+
+pr: [#59115](https://github.com/PaddlePaddle/Paddle/pull/59115)
+
+### 下周工作
+
+1. PIR 动转静组件单测问题排查
+2. PIR API 以及动转静最终态适配
+
+
+### 导师点评
+请联系导师填写