{
 "cells": [
  {
   "cell_type": "markdown",
   "source": [
    "# Parameter\n",
    "\n",
    "`Ascend` `GPU` `CPU` `入门`\n",
    "\n",
    "[![在线运行](https://gitee.com/mindspore/docs/raw/r1.6/resource/_static/logo_modelarts.png)](https://authoring-modelarts-cnnorth4.huaweicloud.com/console/lab?share-url-b64=aHR0cHM6Ly9vYnMuZHVhbHN0YWNrLmNuLW5vcnRoLTQubXlodWF3ZWljbG91ZC5jb20vbWluZHNwb3JlLXdlYnNpdGUvbm90ZWJvb2svbWFzdGVyL3Byb2dyYW1taW5nX2d1aWRlL3poX2NuL21pbmRzcG9yZV9wYXJhbWV0ZXJfaW50cm9kdWN0aW9uLmlweW5i&imageid=65f636a0-56cf-49df-b941-7d2a07ba8c8c) [![下载Notebook](https://gitee.com/mindspore/docs/raw/r1.6/resource/_static/logo_notebook.png)](https://obs.dualstack.cn-north-4.myhuaweicloud.com/mindspore-website/notebook/r1.6/programming_guide/zh_cn/mindspore_parameter_introduction.ipynb) [![下载样例代码](https://gitee.com/mindspore/docs/raw/r1.6/resource/_static/logo_download_code.png)](https://obs.dualstack.cn-north-4.myhuaweicloud.com/mindspore-website/notebook/r1.6/programming_guide/zh_cn/mindspore_parameter_introduction.py) [![查看源文件](https://gitee.com/mindspore/docs/raw/r1.6/resource/_static/logo_source.png)](https://gitee.com/mindspore/docs/blob/r1.6/docs/mindspore/programming_guide/source_zh_cn/parameter_introduction.ipynb)"
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "`Parameter`是变量张量，代表在训练网络时，需要被更新的参数，一般包括权重`weight`和偏置`bias`。\n",
    "\n",
    "例如，对于卷积算子`nn.Conv2D`， `weight`是`[out_channel, in_channel, kernel_size, kernel_size]`大小的张量， `bias`是`[out_channel]`大小的向量。而对于全连接操作`nn.Dense`，`weight`是`[out_channel, in_channel]`大小的张量， `bias`是`[out_channel]`大小的向量。\n",
    "\n",
    "```python\n",
    "import numpy as np\n",
    "from mindspore import Tensor\n",
    "from mindspore import nn\n",
    "\n",
    "conv_net = nn.Conv2d(in_channels=120, out_channels=240, kernel_size=4, has_bias=True, weight_init='normal')\n",
    "dense_net = nn.Dense(in_channels=3, out_channels=4, weight_init='normal', bias_init='zeros', has_bias=True)\n",
    "\n",
    "print(\"conv_net, weight shape:\", conv_net.weight.shape, \", bias shape:\", conv_net.bias.shape)\n",
    "print(\"dense_net, weight_shape：\", dense_net.weight.shape, \", bias shape:\", dense_net.bias.shape)\n",
    "```\n",
    "\n",
    "```python\n",
    "conv_net, weight shape: (240, 120, 4, 4) , bias shape:  (240,)\n",
    "dense_net, weight_shape: (4, 3) , bias shape:  (4,)\n",
    "```\n",
    "\n",
    "`nn`下的算子接口中一般集成了权重和偏置，用户可以指定初始化方式来初始化这些`Parameter`。\n",
    "\n",
    "在网络训练过程中，优化器根据反向传播计算出的梯度不断更新`Parameter`的值，最优的`Parameter`最终会被保存下来供推理使用。\n",
    "\n",
    "`Parameter`不支持稀疏矩阵，如[mindspore.RowTensor](https://www.mindspore.cn/docs/api/zh-CN/r1.6/api_python/mindspore/mindspore.RowTensor.html)、[mindspore.SparseTensor](https://www.mindspore.cn/docs/api/zh-CN/r1.6/api_python/mindspore/mindspore.SparseTensor.html)。其初始化和更新的详细介绍可参考[参数初始化](https://www.mindspore.cn/docs/programming_guide/zh-CN/r1.6/initializer.html)和[参数更新](https://www.mindspore.cn/docs/programming_guide/zh-CN/r1.6/parameter.html)。"
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "\n"
   ],
   "metadata": {}
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "MindSpore",
   "language": "python",
   "name": "mindspore"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}