{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "view-in-github"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/mlc-ai/notebooks/blob/main/6_Integration_with_Machine_Learning_Frameworks.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Mpn1ti5Urdsv"
      },
      "source": [
        "# Ep6: Integration with Machine Learning Frameworks"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qXysoqn-vZuF"
      },
      "source": [
        "## Install packages \n",
        "\n",
        "For the purpose of this course, we will use some ongoing development in tvm, which is an open-source machine learning compilation framework. We provide the following command to install a packaged version for mlc course."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Xe3vClsD9jlq",
        "outputId": "26a1bfbf-9182-4cfe-d507-042d11b9225a"
      },
      "outputs": [],
      "source": [
        "# !python3 -m  pip install mlc-ai-nightly -f https://mlc.ai/wheels"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "i-14C4skxIrJ"
      },
      "source": [
        "## Prelude\n",
        "\n",
        "In the past chapters, we have learned about abstractions for machine learning compilation and transformations among tensor functions.\n",
        "\n",
        "This chapter will discuss how to bring machine learning models from the existing ML framework into an MLC flow."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BBIuE2jc1DaU"
      },
      "source": [
        "## Preparations\n",
        "\n",
        "To begin with, we will import necessary dependencies.\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "id": "BVp0fHyRkYj6"
      },
      "outputs": [],
      "source": [
        "import tvm\n",
        "from tvm.ir.module import IRModule\n",
        "from tvm.script import tir as T, relax as R\n",
        "from tvm import relax\n",
        "import numpy as np\n",
        "\n",
        "# This is needed for deferring annotation parsing in TVMScript\n",
        "from __future__ import annotations "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "id": "6saIbYSCrZF7"
      },
      "outputs": [],
      "source": [
        "import torch\n",
        "import torch.nn as nn\n",
        "from torch import fx\n",
        "from torch.nn import functional as F"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8yH4IMSMvF9o"
      },
      "source": [
        "## Build an IRModule Through a Builder\n",
        "\n",
        "In the past chapters, we have been building IRModule by directly writing TVMScript. As the model gets larger, we need a programmatical way to build up an IRModule. In this section, let us review some of the tools to support that process.\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uahNKehVr2gg"
      },
      "source": [
        "### Tensor Expression for TensorIR Creation\n",
        "\n",
        "First, we review the tensor expression domain-specific language to build TensorIR functions.\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "id": "rxjwYscu4ukP"
      },
      "outputs": [],
      "source": [
        "from tvm import te"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yk63AG3t6Kdu"
      },
      "source": [
        "We begin by creating a placeholder object, which represents an input to a TensorIR function."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "id": "s_IUriaf590M"
      },
      "outputs": [],
      "source": [
        "A = te.placeholder((128, 128), name=\"A\", dtype=\"float32\")\n",
        "B = te.placeholder((128, 128), name=\"B\", dtype=\"float32\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bOfyFOkS6YVO"
      },
      "source": [
        "Each input and intermediate result here are represented as a `te.Tensor` object."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "caakIb3B59_Q",
        "outputId": "222c77b2-1d65-43ea-d079-4e67743c7d50"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "tvm.te.tensor.Tensor"
            ]
          },
          "execution_count": 6,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "type(A)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yN6-WogI6l8k"
      },
      "source": [
        "Each `te.Tensor` has a shape field and dtype field that tracks the shape\n",
        "and data type of the computation."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_5SiENj96VMa",
        "outputId": "9f77fccf-4467-4d82-cc67-e3c8afd07d35"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "[128, 128]"
            ]
          },
          "execution_count": 7,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "A.shape"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GWg6lEgR5P7i"
      },
      "source": [
        "We can describe computations through a sequence of tensor expression computation, Here `te.compute` takes the signature `te.compute(output_shape, fcompute)`. And the fcompute function describes how we want to compute the value of each element `[i, j]` for a given index.\n",
        "\n",
        "The `te_matmul` function takes in an object with type `te.Tensor`, and returns the matrix multiplication result. Note how we build up computations depending on A and B's input shape. The `te_matmul` works for A and B with different input shapes.\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "id": "zlvL1Zfkt9A4"
      },
      "outputs": [],
      "source": [
        "def te_matmul(A: te.Tensor, B: te.Tensor) -> te.Tensor:\n",
        "    assert A.shape[1] == B.shape[0]\n",
        "    n = A.shape[0]\n",
        "    m = B.shape[1]\n",
        "    k = te.reduce_axis((0, A.shape[1]), name=\"k\")\n",
        "    return te.compute((n, m), lambda i, j: te.sum(A[i, k] * B[k, j], axis=k), name=\"matmul\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dkqrFnHA7er4"
      },
      "source": [
        "We can create the result of matmul calling `te_matmul` with A and B."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "id": "ewHuGnPv7EyV"
      },
      "outputs": [],
      "source": [
        "C = te_matmul(A, B)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UJKUFKwq9D5d"
      },
      "source": [
        "To create a TensorIR function, we can call `te.create_prim_func` and pass in the input and output values."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ibJ1f-TI8k2F",
        "outputId": "e7e005fc-d6bf-49e7-8fba-45bbf6b5a719"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\u001b[38;5;30;03m# from tvm.script import tir as T\u001b[39;00m\n",
            "\u001b[38;5;129m@T\u001b[39m\u001b[38;5;129;01m.\u001b[39;00mprim_func\n",
            "\u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mfunc\u001b[39m(A: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], B: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], matmul: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m]) \u001b[38;5;129;01m-\u001b[39;00m\u001b[38;5;129;01m>\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n",
            "    \u001b[38;5;30;03m# function attr dict\u001b[39;00m\n",
            "    T\u001b[38;5;129;01m.\u001b[39;00mfunc_attr({\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mglobal_symbol\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mmain\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtir.noalias\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;28;01mTrue\u001b[39;00m})\n",
            "    \u001b[38;5;30;03m# body\u001b[39;00m\n",
            "    \u001b[38;5;30;03m# with T.block(\"root\")\u001b[39;00m\n",
            "    \u001b[38;5;28;01mfor\u001b[39;00m i0, i1, i2 \u001b[38;5;28;01min\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mgrid(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m):\n",
            "        \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mblock(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mmatmul\u001b[39m\u001b[38;5;124m\"\u001b[39m):\n",
            "            i, j, k \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00maxis\u001b[38;5;129;01m.\u001b[39;00mremap(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mSSR\u001b[39m\u001b[38;5;124m\"\u001b[39m, [i0, i1, i2])\n",
            "            T\u001b[38;5;129;01m.\u001b[39;00mreads(A[i, k], B[k, j])\n",
            "            T\u001b[38;5;129;01m.\u001b[39;00mwrites(matmul[i, j])\n",
            "            \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00minit():\n",
            "                matmul[i, j] \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mfloat32(\u001b[38;5;28m0\u001b[39m)\n",
            "            matmul[i, j] \u001b[38;5;129;01m=\u001b[39;00m matmul[i, j] \u001b[38;5;129;01m+\u001b[39;00m A[i, k] \u001b[38;5;129;01m*\u001b[39;00m B[k, j]\n",
            "\n"
          ]
        }
      ],
      "source": [
        "te.create_prim_func([A, B, C]).show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TeGKASD09NIJ"
      },
      "source": [
        "We can create a tensor expression for relu computation in a similar fashion. Here we write it in a way so that `te_relu` function can work for `te.Tensor` with any dimension and shape."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "id": "GSeib3RT778t"
      },
      "outputs": [],
      "source": [
        "def te_relu(A: te.Tensor) -> te.Tensor:\n",
        "    return te.compute(A.shape, lambda *i: te.max(A(*i), 0), name=\"relu\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CwsiEWlU91UY"
      },
      "source": [
        "Let us try out `te_relu` on two different input shapes and dimensions. First `X1` with shape `(10,)`."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PYRH0bCV8k9B",
        "outputId": "f4a27fd7-173f-438f-eb49-e83b53360ba6"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\u001b[38;5;30;03m# from tvm.script import tir as T\u001b[39;00m\n",
            "\u001b[38;5;129m@T\u001b[39m\u001b[38;5;129;01m.\u001b[39;00mprim_func\n",
            "\u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mfunc\u001b[39m(X1: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[\u001b[38;5;28m10\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], relu: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[\u001b[38;5;28m10\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m]) \u001b[38;5;129;01m-\u001b[39;00m\u001b[38;5;129;01m>\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n",
            "    \u001b[38;5;30;03m# function attr dict\u001b[39;00m\n",
            "    T\u001b[38;5;129;01m.\u001b[39;00mfunc_attr({\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mglobal_symbol\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mmain\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtir.noalias\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;28;01mTrue\u001b[39;00m})\n",
            "    \u001b[38;5;30;03m# body\u001b[39;00m\n",
            "    \u001b[38;5;30;03m# with T.block(\"root\")\u001b[39;00m\n",
            "    \u001b[38;5;28;01mfor\u001b[39;00m i0 \u001b[38;5;28;01min\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mserial(\u001b[38;5;28m10\u001b[39m):\n",
            "        \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mblock(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mrelu\u001b[39m\u001b[38;5;124m\"\u001b[39m):\n",
            "            i0_1 \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00maxis\u001b[38;5;129;01m.\u001b[39;00mspatial(\u001b[38;5;28m10\u001b[39m, i0)\n",
            "            T\u001b[38;5;129;01m.\u001b[39;00mreads(X1[i0_1])\n",
            "            T\u001b[38;5;129;01m.\u001b[39;00mwrites(relu[i0_1])\n",
            "            relu[i0_1] \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mmax(X1[i0_1], T\u001b[38;5;129;01m.\u001b[39;00mfloat32(\u001b[38;5;28m0\u001b[39m))\n",
            "\n"
          ]
        }
      ],
      "source": [
        "X1 = te.placeholder((10,), name=\"X1\", dtype=\"float32\")\n",
        "Y1 = te_relu(X1)\n",
        "te.create_prim_func([X1, Y1]).show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DDPuxjGy-AiV"
      },
      "source": [
        "Then `X2` with shape `(10, 20)`."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CXN_tu5k9o4u",
        "outputId": "16099283-d781-4635-d173-ba915cd60868"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\u001b[38;5;30;03m# from tvm.script import tir as T\u001b[39;00m\n",
            "\u001b[38;5;129m@T\u001b[39m\u001b[38;5;129;01m.\u001b[39;00mprim_func\n",
            "\u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mfunc\u001b[39m(X1: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m10\u001b[39m, \u001b[38;5;28m20\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], relu: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m10\u001b[39m, \u001b[38;5;28m20\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m]) \u001b[38;5;129;01m-\u001b[39;00m\u001b[38;5;129;01m>\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n",
            "    \u001b[38;5;30;03m# function attr dict\u001b[39;00m\n",
            "    T\u001b[38;5;129;01m.\u001b[39;00mfunc_attr({\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mglobal_symbol\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mmain\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtir.noalias\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;28;01mTrue\u001b[39;00m})\n",
            "    \u001b[38;5;30;03m# body\u001b[39;00m\n",
            "    \u001b[38;5;30;03m# with T.block(\"root\")\u001b[39;00m\n",
            "    \u001b[38;5;28;01mfor\u001b[39;00m i0, i1 \u001b[38;5;28;01min\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mgrid(\u001b[38;5;28m10\u001b[39m, \u001b[38;5;28m20\u001b[39m):\n",
            "        \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mblock(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mrelu\u001b[39m\u001b[38;5;124m\"\u001b[39m):\n",
            "            i0_1, i1_1 \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00maxis\u001b[38;5;129;01m.\u001b[39;00mremap(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mSS\u001b[39m\u001b[38;5;124m\"\u001b[39m, [i0, i1])\n",
            "            T\u001b[38;5;129;01m.\u001b[39;00mreads(X1[i0_1, i1_1])\n",
            "            T\u001b[38;5;129;01m.\u001b[39;00mwrites(relu[i0_1, i1_1])\n",
            "            relu[i0_1, i1_1] \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mmax(X1[i0_1, i1_1], T\u001b[38;5;129;01m.\u001b[39;00mfloat32(\u001b[38;5;28m0\u001b[39m))\n",
            "\n"
          ]
        }
      ],
      "source": [
        "X2 = te.placeholder((10, 20), name=\"X1\", dtype=\"float32\")\n",
        "Y2 = te_relu(X2)\n",
        "te.create_prim_func([X2, Y2]).show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1slxb7C6-FT3"
      },
      "source": [
        "One final thing that `te` API allows us to do is to compose operations and create \"fused\" operators. For example, we can take the result of matmul and apply relu again. "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {
        "id": "5h7i20Qd9pEA"
      },
      "outputs": [],
      "source": [
        "C = te_matmul(A, B)\n",
        "D = te_relu(C)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Q2vxDWDq-keD"
      },
      "source": [
        "We can create a TensorIR function by only passing the input and output values of interest, skipping intermediate values. This will cause the result of matmul being allocated as a temp space in the TensorIR function."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kYqc5Ft--jHR",
        "outputId": "c4b17b9f-d426-4372-f976-fe8e475b676d"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\u001b[38;5;30;03m# from tvm.script import tir as T\u001b[39;00m\n",
            "\u001b[38;5;129m@T\u001b[39m\u001b[38;5;129;01m.\u001b[39;00mprim_func\n",
            "\u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mfunc\u001b[39m(A: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], B: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], relu: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m]) \u001b[38;5;129;01m-\u001b[39;00m\u001b[38;5;129;01m>\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n",
            "    \u001b[38;5;30;03m# function attr dict\u001b[39;00m\n",
            "    T\u001b[38;5;129;01m.\u001b[39;00mfunc_attr({\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mglobal_symbol\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mmain\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtir.noalias\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;28;01mTrue\u001b[39;00m})\n",
            "    \u001b[38;5;30;03m# body\u001b[39;00m\n",
            "    \u001b[38;5;30;03m# with T.block(\"root\")\u001b[39;00m\n",
            "    matmul \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00malloc_buffer([\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m], dtype\u001b[38;5;129;01m=\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n",
            "    \u001b[38;5;28;01mfor\u001b[39;00m i0, i1, i2 \u001b[38;5;28;01min\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mgrid(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m):\n",
            "        \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mblock(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mmatmul\u001b[39m\u001b[38;5;124m\"\u001b[39m):\n",
            "            i, j, k \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00maxis\u001b[38;5;129;01m.\u001b[39;00mremap(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mSSR\u001b[39m\u001b[38;5;124m\"\u001b[39m, [i0, i1, i2])\n",
            "            T\u001b[38;5;129;01m.\u001b[39;00mreads(A[i, k], B[k, j])\n",
            "            T\u001b[38;5;129;01m.\u001b[39;00mwrites(matmul[i, j])\n",
            "            \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00minit():\n",
            "                matmul[i, j] \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mfloat32(\u001b[38;5;28m0\u001b[39m)\n",
            "            matmul[i, j] \u001b[38;5;129;01m=\u001b[39;00m matmul[i, j] \u001b[38;5;129;01m+\u001b[39;00m A[i, k] \u001b[38;5;129;01m*\u001b[39;00m B[k, j]\n",
            "    \u001b[38;5;28;01mfor\u001b[39;00m i0, i1 \u001b[38;5;28;01min\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mgrid(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m):\n",
            "        \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mblock(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mrelu\u001b[39m\u001b[38;5;124m\"\u001b[39m):\n",
            "            i0_1, i1_1 \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00maxis\u001b[38;5;129;01m.\u001b[39;00mremap(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mSS\u001b[39m\u001b[38;5;124m\"\u001b[39m, [i0, i1])\n",
            "            T\u001b[38;5;129;01m.\u001b[39;00mreads(matmul[i0_1, i1_1])\n",
            "            T\u001b[38;5;129;01m.\u001b[39;00mwrites(relu[i0_1, i1_1])\n",
            "            relu[i0_1, i1_1] \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mmax(matmul[i0_1, i1_1], T\u001b[38;5;129;01m.\u001b[39;00mfloat32(\u001b[38;5;28m0\u001b[39m))\n",
            "\n"
          ]
        }
      ],
      "source": [
        "te.create_prim_func([A, B, D]).show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3DVxa0rO_z26"
      },
      "source": [
        "We can also pass the intermediate result C into the argument list. In this case, the TensorIR function expects us to also pass in the buffer of C from the caller side. Normally we recommend only passing in the input/output so we can have more advanced fusion inside."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rwzFD4bt-wm3",
        "outputId": "359f4aa0-5f08-4ee6-94ac-33b5e4e17898"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\u001b[38;5;30;03m# from tvm.script import tir as T\u001b[39;00m\n",
            "\u001b[38;5;129m@T\u001b[39m\u001b[38;5;129;01m.\u001b[39;00mprim_func\n",
            "\u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mfunc\u001b[39m(A: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], B: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], matmul: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], relu: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m]) \u001b[38;5;129;01m-\u001b[39;00m\u001b[38;5;129;01m>\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n",
            "    \u001b[38;5;30;03m# function attr dict\u001b[39;00m\n",
            "    T\u001b[38;5;129;01m.\u001b[39;00mfunc_attr({\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mglobal_symbol\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mmain\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtir.noalias\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;28;01mTrue\u001b[39;00m})\n",
            "    \u001b[38;5;30;03m# body\u001b[39;00m\n",
            "    \u001b[38;5;30;03m# with T.block(\"root\")\u001b[39;00m\n",
            "    \u001b[38;5;28;01mfor\u001b[39;00m i0, i1, i2 \u001b[38;5;28;01min\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mgrid(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m):\n",
            "        \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mblock(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mmatmul\u001b[39m\u001b[38;5;124m\"\u001b[39m):\n",
            "            i, j, k \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00maxis\u001b[38;5;129;01m.\u001b[39;00mremap(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mSSR\u001b[39m\u001b[38;5;124m\"\u001b[39m, [i0, i1, i2])\n",
            "            T\u001b[38;5;129;01m.\u001b[39;00mreads(A[i, k], B[k, j])\n",
            "            T\u001b[38;5;129;01m.\u001b[39;00mwrites(matmul[i, j])\n",
            "            \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00minit():\n",
            "                matmul[i, j] \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mfloat32(\u001b[38;5;28m0\u001b[39m)\n",
            "            matmul[i, j] \u001b[38;5;129;01m=\u001b[39;00m matmul[i, j] \u001b[38;5;129;01m+\u001b[39;00m A[i, k] \u001b[38;5;129;01m*\u001b[39;00m B[k, j]\n",
            "    \u001b[38;5;28;01mfor\u001b[39;00m i0, i1 \u001b[38;5;28;01min\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mgrid(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m):\n",
            "        \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mblock(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mrelu\u001b[39m\u001b[38;5;124m\"\u001b[39m):\n",
            "            i0_1, i1_1 \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00maxis\u001b[38;5;129;01m.\u001b[39;00mremap(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mSS\u001b[39m\u001b[38;5;124m\"\u001b[39m, [i0, i1])\n",
            "            T\u001b[38;5;129;01m.\u001b[39;00mreads(matmul[i0_1, i1_1])\n",
            "            T\u001b[38;5;129;01m.\u001b[39;00mwrites(relu[i0_1, i1_1])\n",
            "            relu[i0_1, i1_1] \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mmax(matmul[i0_1, i1_1], T\u001b[38;5;129;01m.\u001b[39;00mfloat32(\u001b[38;5;28m0\u001b[39m))\n",
            "\n"
          ]
        }
      ],
      "source": [
        "te.create_prim_func([A, B, C, D]).show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oOTcJQDB7EeJ"
      },
      "source": [
        "### Use BlockBuilder to Create an IRModule"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Vm71Tp87Qz8M"
      },
      "source": [
        "So far, we have created a single TensorIR function. In order to build end-to-end model execution, we also need to be able to connect multiple TensorIR functions through a computational graph.\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SnaE0Szgu5cq"
      },
      "source": [
        "Let us first create a block builder, which helps us incrementally build a `relax.Function`."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {
        "id": "0xwqttTtAESi"
      },
      "outputs": [],
      "source": [
        "A = relax.Var(\"A\", (128, 128), relax.DynTensorType(2, \"float32\"))\n",
        "B = relax.Var(\"B\", (128, 128), relax.DynTensorType(2, \"float32\"))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IlXZjFsT02-Y"
      },
      "source": [
        "We construct the relax function by creating a block builder and then a sequence of primitive tensor operations."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Hz4UojV8vT8Y",
        "outputId": "e7cb31e3-d7f3-4cba-9042-f1ea5a140d36"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\u001b[38;5;129m@tvm\u001b[39m\u001b[38;5;129;01m.\u001b[39;00mscript\u001b[38;5;129;01m.\u001b[39;00mir_module\n",
            "\u001b[38;5;28;01mclass\u001b[39;00m \u001b[38;5;21;01mModule\u001b[39;00m:\n",
            "    \u001b[38;5;129m@T\u001b[39m\u001b[38;5;129;01m.\u001b[39;00mprim_func\n",
            "    \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mte_matmul\u001b[39m(rxplaceholder: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], rxplaceholder_1: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], matmul: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m]) \u001b[38;5;129;01m-\u001b[39;00m\u001b[38;5;129;01m>\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n",
            "        \u001b[38;5;30;03m# function attr dict\u001b[39;00m\n",
            "        T\u001b[38;5;129;01m.\u001b[39;00mfunc_attr({\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mglobal_symbol\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mte_matmul\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtir.noalias\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;28;01mTrue\u001b[39;00m})\n",
            "        \u001b[38;5;30;03m# body\u001b[39;00m\n",
            "        \u001b[38;5;30;03m# with T.block(\"root\")\u001b[39;00m\n",
            "        \u001b[38;5;28;01mfor\u001b[39;00m i0, i1, i2 \u001b[38;5;28;01min\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mgrid(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m):\n",
            "            \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mblock(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mmatmul\u001b[39m\u001b[38;5;124m\"\u001b[39m):\n",
            "                i, j, k \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00maxis\u001b[38;5;129;01m.\u001b[39;00mremap(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mSSR\u001b[39m\u001b[38;5;124m\"\u001b[39m, [i0, i1, i2])\n",
            "                T\u001b[38;5;129;01m.\u001b[39;00mreads(rxplaceholder[i, k], rxplaceholder_1[k, j])\n",
            "                T\u001b[38;5;129;01m.\u001b[39;00mwrites(matmul[i, j])\n",
            "                \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00minit():\n",
            "                    matmul[i, j] \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mfloat32(\u001b[38;5;28m0\u001b[39m)\n",
            "                matmul[i, j] \u001b[38;5;129;01m=\u001b[39;00m matmul[i, j] \u001b[38;5;129;01m+\u001b[39;00m rxplaceholder[i, k] \u001b[38;5;129;01m*\u001b[39;00m rxplaceholder_1[k, j]\n",
            "    \n",
            "    \u001b[38;5;129m@T\u001b[39m\u001b[38;5;129;01m.\u001b[39;00mprim_func\n",
            "    \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mte_relu\u001b[39m(rxplaceholder: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], relu: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m]) \u001b[38;5;129;01m-\u001b[39;00m\u001b[38;5;129;01m>\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n",
            "        \u001b[38;5;30;03m# function attr dict\u001b[39;00m\n",
            "        T\u001b[38;5;129;01m.\u001b[39;00mfunc_attr({\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mglobal_symbol\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mte_relu\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtir.noalias\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;28;01mTrue\u001b[39;00m})\n",
            "        \u001b[38;5;30;03m# body\u001b[39;00m\n",
            "        \u001b[38;5;30;03m# with T.block(\"root\")\u001b[39;00m\n",
            "        \u001b[38;5;28;01mfor\u001b[39;00m i0, i1 \u001b[38;5;28;01min\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mgrid(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m):\n",
            "            \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mblock(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mrelu\u001b[39m\u001b[38;5;124m\"\u001b[39m):\n",
            "                i0_1, i1_1 \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00maxis\u001b[38;5;129;01m.\u001b[39;00mremap(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mSS\u001b[39m\u001b[38;5;124m\"\u001b[39m, [i0, i1])\n",
            "                T\u001b[38;5;129;01m.\u001b[39;00mreads(rxplaceholder[i0_1, i1_1])\n",
            "                T\u001b[38;5;129;01m.\u001b[39;00mwrites(relu[i0_1, i1_1])\n",
            "                relu[i0_1, i1_1] \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mmax(rxplaceholder[i0_1, i1_1], T\u001b[38;5;129;01m.\u001b[39;00mfloat32(\u001b[38;5;28m0\u001b[39m))\n",
            "    \n",
            "    \u001b[38;5;129m@R\u001b[39m\u001b[38;5;129;01m.\u001b[39;00mfunction\n",
            "    \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mmain\u001b[39m(A: Tensor((\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m), B: Tensor((\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m)) \u001b[38;5;129;01m-\u001b[39;00m\u001b[38;5;129;01m>\u001b[39;00m Tensor(\u001b[38;5;28;01mNone\u001b[39;00m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m, ndim \u001b[38;5;129;01m=\u001b[39;00m \u001b[38;5;28m2\u001b[39m):\n",
            "        \u001b[38;5;30;03m# block 0\u001b[39;00m\n",
            "        \u001b[38;5;28;01mwith\u001b[39;00m R\u001b[38;5;129;01m.\u001b[39;00mdataflow():\n",
            "            lv \u001b[38;5;129;01m=\u001b[39;00m R\u001b[38;5;129;01m.\u001b[39;00mcall_tir(te_matmul, (A, B), (\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m), dtype\u001b[38;5;129;01m=\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n",
            "            lv1 \u001b[38;5;129;01m=\u001b[39;00m R\u001b[38;5;129;01m.\u001b[39;00mcall_tir(te_relu, (lv,), (\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m), dtype\u001b[38;5;129;01m=\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n",
            "            gv: Tensor((\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m) \u001b[38;5;129;01m=\u001b[39;00m lv1\n",
            "            R\u001b[38;5;129;01m.\u001b[39;00moutput(gv)\n",
            "        \u001b[38;5;28;01mreturn\u001b[39;00m gv\n",
            "    \n",
            "\n"
          ]
        }
      ],
      "source": [
        "bb = relax.BlockBuilder()\n",
        "\n",
        "with bb.function(\"main\"):\n",
        "    with bb.dataflow():\n",
        "        C = bb.emit_te(te_matmul, A, B)\n",
        "        D = bb.emit_te(te_relu, C)\n",
        "        R = bb.emit_output(D)\n",
        "    bb.emit_func_output(R, params=[A, B])\n",
        "\n",
        "MyModule = bb.get()\n",
        "MyModule.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2yBnEX5C2lG5"
      },
      "source": [
        "### Deep Dive into Block Builder APIs\n",
        "\n",
        "Now let us do a deep dive into each block builder API. It is helpful to put the block builder code and the resulting module side by side."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nzJgNT4kyh3i"
      },
      "source": [
        "![Screen Shot 2022-07-29 at 11.16.24 PM.png](data:image/png;base64,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)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "erv3pnzy25-7"
      },
      "source": [
        "The block builder comes with scopes that correspond to the scopes in the relax function. For example, `bb.dataflow()` creates a dataflow\n",
        "block where all the block builder calls inside the scope belonging to the dataflow scope. \n",
        "\n",
        "```python\n",
        "with bb.function(\"main\"):\n",
        "    with bb.dataflow():\n",
        "        # every emit call generates a variable inside a dataflow block.\n",
        "```"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LA-0uerSAnUC"
      },
      "source": [
        "Each intermediate result is a `relax.Var` corresponding to a variable that stores the result of the computation. `DataflowVar` indicates that the var is an intermediate step inside a dataflow block (computational graph)."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "H8l9fbJQAlY_",
        "outputId": "2eca5a8b-7290-447c-854d-99f99d8d5ff6"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "tvm.relax.expr.DataflowVar"
            ]
          },
          "execution_count": 19,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "type(C)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 20,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ufvwcuD5ytYk",
        "outputId": "fd34141a-0d32-4a80-8e1f-77e890a55c1b"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "True"
            ]
          },
          "execution_count": 20,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "isinstance(C, relax.Var)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "f61-biEx82nr"
      },
      "source": [
        "Each line in the relax function is generated by an `emit_te` call. For example, \n",
        "\n",
        "```python\n",
        "lv = R.call_tir(te_matmul, (A, B), (128, 128), dtype=\"float32\")\n",
        "```\n",
        "is generated by \n",
        "```python\n",
        "C = bb.emit_te(te_matmul, A, B).\n",
        "```\n",
        "Under the hood, the bb.emit_te does the following things:\n",
        "- Create an input `te.placeholder` for A and B\n",
        "- Run them through `te_matmul` function.\n",
        "- Call into `te.create_prim_func` to create a TensorIR function.\n",
        "- Generate a call into the function via `call_tir`.\n",
        "\n",
        "We can find that the result is a computational graph with two intermediate values, with one node corresponding to the te_matmul operation and another one corresponding to `te_relu`."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hWd1GbGHBZlt"
      },
      "source": [
        "We can create output variable of each dataflow block through `bb.emit_output`.\n",
        "\n",
        "```python\n",
        "with bb.dataflow():\n",
        "    ...\n",
        "    R = bb.emit_output(D)\n",
        "```\n",
        "The above code marks that D is a variable that can be referred to outside of the dataflow block.\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FrqQCHu3BrGC"
      },
      "source": [
        "Finally, the function output is marked by `bb.emit_func_output`. We can only call `emit_func_output` once in each function scope.\n",
        "\n",
        "Notably, we can specify the list of parameters of the function in the output emission stage. Doing so helps us in cases where we collect the list of parameters on the fly.\n",
        "\n",
        "```python\n",
        "with bb.function(\"main\"):\n",
        "    ...\n",
        "    # specify parameters in the end\n",
        "    bb.emit_func_output(R, params=[A, B])\n",
        "```\n",
        "Alternatively, we can specify the list of parameters at the beginning of the function scope.\n",
        "```python\n",
        "# specify parameters in the beginning.\n",
        "with bb.function(\"main\", params=[A, B]):\n",
        "    ...\n",
        "    bb.emit_func_output(R)\n",
        "```\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "a-c8N-7sEa1M"
      },
      "source": [
        "## Import Model From PyTorch"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PUcCRU2IQPm-"
      },
      "source": [
        "Now that we have learned the tools to construct an IRModule programmatically. Let us use them to bring a model from PyTorch into the IRModule format.\n",
        "\n",
        "Most machine learning framework comes with computational graph abstractions, where each node corresponds to an operation, and the edges correspond to the dependency among them. We will take a PyTorch model, obtain a computational graph in PyTorch's native format, and translate that into IRModule.\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HJFs5zDkGAOU"
      },
      "source": [
        "Let us begin by defining a model in PyTorch. To keep the example consistent, we will use matmul relu example."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {
        "id": "F6gzwCFfF6V9"
      },
      "outputs": [],
      "source": [
        "class MyModel(nn.Module):\n",
        "    def __init__(self):\n",
        "        super(MyModel, self).__init__()\n",
        "        self.weight = nn.Parameter(torch.randn(128, 128))\n",
        "\n",
        "    def forward(self, x):\n",
        "        x = torch.matmul(x, self.weight)\n",
        "        x = torch.relu(x)\n",
        "        return x"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PoX9dyqVG0ZB"
      },
      "source": [
        "### Create TorchFX GraphModule\n",
        "\n",
        "We use TorchFX to trace a graph from the PyTorch module."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 22,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4vGt8XleGH5q",
        "outputId": "1ffd6f09-38c4-4902-b4bf-76ccf99d6ad4"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "torch.fx.graph_module.GraphModule.__new__.<locals>.GraphModuleImpl"
            ]
          },
          "execution_count": 22,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "model = MyModel()\n",
        "fx_module = fx.symbolic_trace(model)\n",
        "type(fx_module)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eYJUSv2qHm5x"
      },
      "source": [
        "The `fx_module` contains a simple computation graph view that can be printed as tabular data. Our goal is to translate this graph into an IRModule."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 23,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oxvS3jjmGztF",
        "outputId": "970a7536-c672-48fa-9b8e-fadeffb8edf1"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "opcode         name    target                                                     args         kwargs\n",
            "-------------  ------  ---------------------------------------------------------  -----------  --------\n",
            "placeholder    x       x                                                          ()           {}\n",
            "get_attr       weight  weight                                                     ()           {}\n",
            "call_function  matmul  <built-in method matmul of type object at 0x7ff306e899a0>  (x, weight)  {}\n",
            "call_function  relu    <built-in method relu of type object at 0x7ff306e899a0>    (matmul,)    {}\n",
            "output         output  output                                                     (relu,)      {}\n"
          ]
        }
      ],
      "source": [
        "fx_module.graph.print_tabular()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RjVbP-mmIdUs"
      },
      "source": [
        "### Create Map Function\n",
        "\n",
        "Let us define the overall high-level translation logic. The main flow is as follows:\n",
        "- Create a `node_map` that maps `fx.Node` to the corresponding `relax.Var` that represents the translated node in IRModule.\n",
        "- Iterate over the nodes in the fx graph in topological order.\n",
        "- Compute the mapped output of the node given the mapped inputs."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 24,
      "metadata": {
        "id": "U6t1bQ7kOmLs"
      },
      "outputs": [],
      "source": [
        "def map_param(param: nn.Parameter):\n",
        "    ndim = len(param.data.shape)\n",
        "    return relax.const(\n",
        "        param.data.cpu().numpy(), relax.DynTensorType(ndim, \"float32\")\n",
        "    )\n",
        "\n",
        "def fetch_attr(fx_mod, target: str):\n",
        "    \"\"\"Helper function to fetch an attr\"\"\"\n",
        "    target_atoms = target.split('.')\n",
        "    attr_itr = fx_mod\n",
        "    for i, atom in enumerate(target_atoms):\n",
        "        if not hasattr(attr_itr, atom):\n",
        "            raise RuntimeError(f\"Node referenced nonexistant target {'.'.join(target_atoms[:i])}\")\n",
        "        attr_itr = getattr(attr_itr, atom)\n",
        "    return attr_itr\n",
        "\n",
        "def from_fx(fx_mod, input_shapes, call_function_map, call_module_map):\n",
        "    input_index = 0\n",
        "    node_map = {}\n",
        "    named_modules = dict(fx_mod.named_modules())\n",
        "\n",
        "    bb = relax.BlockBuilder()\n",
        "\n",
        "    fn_inputs = []\n",
        "    fn_output = None\n",
        "    with bb.function(\"main\"):\n",
        "        with bb.dataflow():\n",
        "            for node in fx_mod.graph.nodes:\n",
        "                if node.op == \"placeholder\":\n",
        "                    # create input placeholder\n",
        "                    shape = input_shapes[input_index]\n",
        "                    input_index += 1 \n",
        "                    input_var = relax.Var(\n",
        "                        node.target, shape, relax.DynTensorType(len(shape), \"float32\")\n",
        "                    )\n",
        "                    fn_inputs.append(input_var)\n",
        "                    node_map[node] = input_var\n",
        "                elif node.op == \"get_attr\":\n",
        "                    node_map[node] = map_param(fetch_attr(fx_mod, node.target))\n",
        "                elif node.op == \"call_function\":\n",
        "                    node_map[node] = call_function_map[node.target](bb, node_map, node)\n",
        "                elif node.op == \"call_module\":\n",
        "                    named_module = named_modules[node.target]\n",
        "                    node_map[node] = call_module_map[type(named_module)](bb, node_map, node, named_module)\n",
        "                elif node.op == \"output\":\n",
        "                    output = node_map[node.args[0]]\n",
        "                    assert fn_output is None\n",
        "                    fn_output = bb.emit_output(output)\n",
        "        # output and finalize the function\n",
        "        bb.emit_func_output(output, fn_inputs)\n",
        "    return bb.get()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HBlaVkysfrCu"
      },
      "source": [
        "We did not define the function map in the `from_fx` function. We will supply the translation rule of each torch function via a map. Specifically, the following code block shows how we can do that through the `emit_te` API."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 25,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CKqkz9jrfrbh",
        "outputId": "cbd008cf-6862-48fe-db45-59931c0fd68c"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\u001b[38;5;129m@tvm\u001b[39m\u001b[38;5;129;01m.\u001b[39;00mscript\u001b[38;5;129;01m.\u001b[39;00mir_module\n",
            "\u001b[38;5;28;01mclass\u001b[39;00m \u001b[38;5;21;01mModule\u001b[39;00m:\n",
            "    \u001b[38;5;129m@T\u001b[39m\u001b[38;5;129;01m.\u001b[39;00mprim_func\n",
            "    \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mte_matmul\u001b[39m(rxplaceholder: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], rxplaceholder_1: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], matmul: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m]) \u001b[38;5;129;01m-\u001b[39;00m\u001b[38;5;129;01m>\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n",
            "        \u001b[38;5;30;03m# function attr dict\u001b[39;00m\n",
            "        T\u001b[38;5;129;01m.\u001b[39;00mfunc_attr({\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mglobal_symbol\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mte_matmul\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtir.noalias\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;28;01mTrue\u001b[39;00m})\n",
            "        \u001b[38;5;30;03m# body\u001b[39;00m\n",
            "        \u001b[38;5;30;03m# with T.block(\"root\")\u001b[39;00m\n",
            "        \u001b[38;5;28;01mfor\u001b[39;00m i0, i1, i2 \u001b[38;5;28;01min\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mgrid(\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m128\u001b[39m):\n",
            "            \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mblock(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mmatmul\u001b[39m\u001b[38;5;124m\"\u001b[39m):\n",
            "                i, j, k \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00maxis\u001b[38;5;129;01m.\u001b[39;00mremap(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mSSR\u001b[39m\u001b[38;5;124m\"\u001b[39m, [i0, i1, i2])\n",
            "                T\u001b[38;5;129;01m.\u001b[39;00mreads(rxplaceholder[i, k], rxplaceholder_1[k, j])\n",
            "                T\u001b[38;5;129;01m.\u001b[39;00mwrites(matmul[i, j])\n",
            "                \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00minit():\n",
            "                    matmul[i, j] \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mfloat32(\u001b[38;5;28m0\u001b[39m)\n",
            "                matmul[i, j] \u001b[38;5;129;01m=\u001b[39;00m matmul[i, j] \u001b[38;5;129;01m+\u001b[39;00m rxplaceholder[i, k] \u001b[38;5;129;01m*\u001b[39;00m rxplaceholder_1[k, j]\n",
            "    \n",
            "    \u001b[38;5;129m@T\u001b[39m\u001b[38;5;129;01m.\u001b[39;00mprim_func\n",
            "    \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mte_relu\u001b[39m(rxplaceholder: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], relu: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m]) \u001b[38;5;129;01m-\u001b[39;00m\u001b[38;5;129;01m>\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n",
            "        \u001b[38;5;30;03m# function attr dict\u001b[39;00m\n",
            "        T\u001b[38;5;129;01m.\u001b[39;00mfunc_attr({\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mglobal_symbol\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mte_relu\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtir.noalias\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;28;01mTrue\u001b[39;00m})\n",
            "        \u001b[38;5;30;03m# body\u001b[39;00m\n",
            "        \u001b[38;5;30;03m# with T.block(\"root\")\u001b[39;00m\n",
            "        \u001b[38;5;28;01mfor\u001b[39;00m i0, i1 \u001b[38;5;28;01min\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mgrid(\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m):\n",
            "            \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mblock(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mrelu\u001b[39m\u001b[38;5;124m\"\u001b[39m):\n",
            "                i0_1, i1_1 \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00maxis\u001b[38;5;129;01m.\u001b[39;00mremap(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mSS\u001b[39m\u001b[38;5;124m\"\u001b[39m, [i0, i1])\n",
            "                T\u001b[38;5;129;01m.\u001b[39;00mreads(rxplaceholder[i0_1, i1_1])\n",
            "                T\u001b[38;5;129;01m.\u001b[39;00mwrites(relu[i0_1, i1_1])\n",
            "                relu[i0_1, i1_1] \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mmax(rxplaceholder[i0_1, i1_1], T\u001b[38;5;129;01m.\u001b[39;00mfloat32(\u001b[38;5;28m0\u001b[39m))\n",
            "    \n",
            "    \u001b[38;5;129m@R\u001b[39m\u001b[38;5;129;01m.\u001b[39;00mfunction\n",
            "    \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mmain\u001b[39m(x: Tensor((\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m)) \u001b[38;5;129;01m-\u001b[39;00m\u001b[38;5;129;01m>\u001b[39;00m Tensor(\u001b[38;5;28;01mNone\u001b[39;00m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m, ndim \u001b[38;5;129;01m=\u001b[39;00m \u001b[38;5;28m2\u001b[39m):\n",
            "        \u001b[38;5;30;03m# block 0\u001b[39;00m\n",
            "        \u001b[38;5;28;01mwith\u001b[39;00m R\u001b[38;5;129;01m.\u001b[39;00mdataflow():\n",
            "            lv \u001b[38;5;129;01m=\u001b[39;00m R\u001b[38;5;129;01m.\u001b[39;00mcall_tir(te_matmul, (x, meta[relay\u001b[38;5;129;01m.\u001b[39;00mConstant][\u001b[38;5;28m0\u001b[39m]), (\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m), dtype\u001b[38;5;129;01m=\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n",
            "            lv1 \u001b[38;5;129;01m=\u001b[39;00m R\u001b[38;5;129;01m.\u001b[39;00mcall_tir(te_relu, (lv,), (\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m), dtype\u001b[38;5;129;01m=\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n",
            "            gv: Tensor((\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m) \u001b[38;5;129;01m=\u001b[39;00m lv1\n",
            "            R\u001b[38;5;129;01m.\u001b[39;00moutput(gv)\n",
            "        \u001b[38;5;28;01mreturn\u001b[39;00m lv1\n",
            "    \n",
            "\n"
          ]
        }
      ],
      "source": [
        "def map_matmul(bb, node_map, node: fx.Node):\n",
        "    A = node_map[node.args[0]]\n",
        "    B = node_map[node.args[1]]\n",
        "    return bb.emit_te(te_matmul, A, B)\n",
        "\n",
        "def map_relu(bb, node_map, node: fx.Node):\n",
        "    A = node_map[node.args[0]]\n",
        "    return bb.emit_te(te_relu, A)\n",
        "\n",
        "MyModule = from_fx(\n",
        "    fx_module, \n",
        "    input_shapes = [(1, 128)], \n",
        "    call_function_map = {\n",
        "      torch.matmul: map_matmul,\n",
        "      torch.relu: map_relu, \n",
        "    },\n",
        "    call_module_map={},\n",
        ")\n",
        "\n",
        "MyModule.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6OBLhELUgpBT"
      },
      "source": [
        "## Coming back to FashionMNIST Example"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 26,
      "metadata": {
        "id": "o1JM175Ri_hq"
      },
      "outputs": [],
      "source": [
        "import torch\n",
        "import torchvision\n",
        "\n",
        "test_data = torchvision.datasets.FashionMNIST(\n",
        "    root=\"data\",\n",
        "    train=False,\n",
        "    download=True,\n",
        "    transform=torchvision.transforms.ToTensor()\n",
        ")\n",
        "test_loader = torch.utils.data.DataLoader(test_data, batch_size=1, shuffle=True)\n",
        "class_names = ['T-shirt/top', 'Trouser', 'Pullover', 'Dress', 'Coat',\n",
        "               'Sandal', 'Shirt', 'Sneaker', 'Bag', 'Ankle boot']\n",
        "\n",
        "img, label = next(iter(test_loader))\n",
        "img = img.reshape(1, 28, 28).numpy()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 27,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 287
        },
        "id": "VOoQL7r7i_qf",
        "outputId": "3685c1c1-2121-47e1-c297-40624727763f"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Class: Bag\n"
          ]
        }
      ],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "plt.figure()\n",
        "plt.imshow(img[0])\n",
        "plt.colorbar()\n",
        "plt.grid(False)\n",
        "plt.show()\n",
        "\n",
        "print(\"Class:\", class_names[label[0]])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 28,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_-xEQ84Fi_yR",
        "outputId": "3ecdeebe-f28a-406e-921d-83c3a7072559"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "File ‘fasionmnist_mlp_params.pkl’ already there; not retrieving.\n",
            "\n"
          ]
        }
      ],
      "source": [
        "# Hide outputs\n",
        "!wget -nc https://github.com/mlc-ai/web-data/raw/main/models/fasionmnist_mlp_params.pkl"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mBwreM4qjn7x"
      },
      "source": [
        "![image.png](data:image/png;base64,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)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Vqh3yklvjquJ"
      },
      "source": [
        "The above is our model of interest, we can build the PyTorch model as follows"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 29,
      "metadata": {
        "id": "OqbgbCtEgfoV"
      },
      "outputs": [],
      "source": [
        "class MLP(nn.Module):\n",
        "    def __init__(self):\n",
        "        super(MLP, self).__init__()\n",
        "        self.linear0 = nn.Linear(784, 128, bias=True)\n",
        "        self.relu = nn.ReLU()\n",
        "        self.linear1 = nn.Linear(128, 10, bias=True)\n",
        "\n",
        "    def forward(self, x):\n",
        "        x = self.linear0(x)\n",
        "        x = self.relu(x)\n",
        "        x = self.linear1(x)\n",
        "        return x"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 30,
      "metadata": {
        "id": "usBESp0ojUdc"
      },
      "outputs": [],
      "source": [
        "import pickle as pkl\n",
        "mlp_model = MLP()\n",
        "\n",
        "mlp_params = pkl.load(open(\"fasionmnist_mlp_params.pkl\", \"rb\"))\n",
        "mlp_model.linear0.weight.data = torch.from_numpy(mlp_params[\"w0\"])\n",
        "mlp_model.linear0.bias.data = torch.from_numpy(mlp_params[\"b0\"])\n",
        "mlp_model.linear1.weight.data = torch.from_numpy(mlp_params[\"w1\"])\n",
        "mlp_model.linear1.bias.data = torch.from_numpy(mlp_params[\"b1\"])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 31,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "MDKt0OAWk3Jr",
        "outputId": "8686b68c-bc0e-4eb9-8916-7d96b86c3ed4"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Torch Prediction: Bag\n"
          ]
        }
      ],
      "source": [
        "torch_res = mlp_model(torch.from_numpy(img.reshape(1, 784)))\n",
        "\n",
        "pred_kind = np.argmax(torch_res.detach().numpy(), axis=1)\n",
        "print(\"Torch Prediction:\", class_names[pred_kind[0]])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6BEuhKdzqlUW"
      },
      "source": [
        "Let us try to translate from fx by defining mapping functions for the corresponding `nn.Module`. Here we are reusing pre-defined TE libraries from TVM `topi` instead of defining our own tensor expression.\n",
        "\n",
        "- `topi.nn.dense(x, w)` performs transposed matrix multiplication `x @ w.T`\n",
        "- `topi.add` performs broadcast add.\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 32,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0YmBlh3ClbNX",
        "outputId": "a498fa9b-6ea7-4c2f-8a91-146a30cf5940"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\u001b[38;5;129m@tvm\u001b[39m\u001b[38;5;129;01m.\u001b[39;00mscript\u001b[38;5;129;01m.\u001b[39;00mir_module\n",
            "\u001b[38;5;28;01mclass\u001b[39;00m \u001b[38;5;21;01mModule\u001b[39;00m:\n",
            "    \u001b[38;5;129m@T\u001b[39m\u001b[38;5;129;01m.\u001b[39;00mprim_func\n",
            "    \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mdense1\u001b[39m(rxplaceholder: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], rxplaceholder_1: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m10\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], T_matmul_NT: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m10\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m]) \u001b[38;5;129;01m-\u001b[39;00m\u001b[38;5;129;01m>\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n",
            "        \u001b[38;5;30;03m# function attr dict\u001b[39;00m\n",
            "        T\u001b[38;5;129;01m.\u001b[39;00mfunc_attr({\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mglobal_symbol\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mdense1\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtir.noalias\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;28;01mTrue\u001b[39;00m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mlayout_free_buffers\u001b[39m\u001b[38;5;124m\"\u001b[39m: [\u001b[38;5;28m1\u001b[39m]})\n",
            "        \u001b[38;5;30;03m# body\u001b[39;00m\n",
            "        \u001b[38;5;30;03m# with T.block(\"root\")\u001b[39;00m\n",
            "        \u001b[38;5;28;01mfor\u001b[39;00m i0, i1, i2 \u001b[38;5;28;01min\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mgrid(\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m10\u001b[39m, \u001b[38;5;28m128\u001b[39m):\n",
            "            \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mblock(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mT_matmul_NT\u001b[39m\u001b[38;5;124m\"\u001b[39m):\n",
            "                i, j, k \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00maxis\u001b[38;5;129;01m.\u001b[39;00mremap(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mSSR\u001b[39m\u001b[38;5;124m\"\u001b[39m, [i0, i1, i2])\n",
            "                T\u001b[38;5;129;01m.\u001b[39;00mreads(rxplaceholder[i, k], rxplaceholder_1[j, k])\n",
            "                T\u001b[38;5;129;01m.\u001b[39;00mwrites(T_matmul_NT[i, j])\n",
            "                \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00minit():\n",
            "                    T_matmul_NT[i, j] \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mfloat32(\u001b[38;5;28m0\u001b[39m)\n",
            "                T_matmul_NT[i, j] \u001b[38;5;129;01m=\u001b[39;00m T_matmul_NT[i, j] \u001b[38;5;129;01m+\u001b[39;00m rxplaceholder[i, k] \u001b[38;5;129;01m*\u001b[39;00m rxplaceholder_1[j, k]\n",
            "    \n",
            "    \u001b[38;5;129m@T\u001b[39m\u001b[38;5;129;01m.\u001b[39;00mprim_func\n",
            "    \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mte_relu\u001b[39m(rxplaceholder: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], relu: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m]) \u001b[38;5;129;01m-\u001b[39;00m\u001b[38;5;129;01m>\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n",
            "        \u001b[38;5;30;03m# function attr dict\u001b[39;00m\n",
            "        T\u001b[38;5;129;01m.\u001b[39;00mfunc_attr({\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mglobal_symbol\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mte_relu\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtir.noalias\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;28;01mTrue\u001b[39;00m})\n",
            "        \u001b[38;5;30;03m# body\u001b[39;00m\n",
            "        \u001b[38;5;30;03m# with T.block(\"root\")\u001b[39;00m\n",
            "        \u001b[38;5;28;01mfor\u001b[39;00m i0, i1 \u001b[38;5;28;01min\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mgrid(\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m):\n",
            "            \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mblock(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mrelu\u001b[39m\u001b[38;5;124m\"\u001b[39m):\n",
            "                i0_1, i1_1 \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00maxis\u001b[38;5;129;01m.\u001b[39;00mremap(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mSS\u001b[39m\u001b[38;5;124m\"\u001b[39m, [i0, i1])\n",
            "                T\u001b[38;5;129;01m.\u001b[39;00mreads(rxplaceholder[i0_1, i1_1])\n",
            "                T\u001b[38;5;129;01m.\u001b[39;00mwrites(relu[i0_1, i1_1])\n",
            "                relu[i0_1, i1_1] \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mmax(rxplaceholder[i0_1, i1_1], T\u001b[38;5;129;01m.\u001b[39;00mfloat32(\u001b[38;5;28m0\u001b[39m))\n",
            "    \n",
            "    \u001b[38;5;129m@T\u001b[39m\u001b[38;5;129;01m.\u001b[39;00mprim_func\n",
            "    \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21madd\u001b[39m(rxplaceholder: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], rxplaceholder_1: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[\u001b[38;5;28m128\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], T_add: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m]) \u001b[38;5;129;01m-\u001b[39;00m\u001b[38;5;129;01m>\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n",
            "        \u001b[38;5;30;03m# function attr dict\u001b[39;00m\n",
            "        T\u001b[38;5;129;01m.\u001b[39;00mfunc_attr({\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mglobal_symbol\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124madd\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtir.noalias\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;28;01mTrue\u001b[39;00m})\n",
            "        \u001b[38;5;30;03m# body\u001b[39;00m\n",
            "        \u001b[38;5;30;03m# with T.block(\"root\")\u001b[39;00m\n",
            "        \u001b[38;5;28;01mfor\u001b[39;00m i0, i1 \u001b[38;5;28;01min\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mgrid(\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m):\n",
            "            \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mblock(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mT_add\u001b[39m\u001b[38;5;124m\"\u001b[39m):\n",
            "                ax0, ax1 \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00maxis\u001b[38;5;129;01m.\u001b[39;00mremap(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mSS\u001b[39m\u001b[38;5;124m\"\u001b[39m, [i0, i1])\n",
            "                T\u001b[38;5;129;01m.\u001b[39;00mreads(rxplaceholder[ax0, ax1], rxplaceholder_1[ax1])\n",
            "                T\u001b[38;5;129;01m.\u001b[39;00mwrites(T_add[ax0, ax1])\n",
            "                T_add[ax0, ax1] \u001b[38;5;129;01m=\u001b[39;00m rxplaceholder[ax0, ax1] \u001b[38;5;129;01m+\u001b[39;00m rxplaceholder_1[ax1]\n",
            "    \n",
            "    \u001b[38;5;129m@T\u001b[39m\u001b[38;5;129;01m.\u001b[39;00mprim_func\n",
            "    \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21madd1\u001b[39m(rxplaceholder: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m10\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], rxplaceholder_1: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[\u001b[38;5;28m10\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], T_add: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m10\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m]) \u001b[38;5;129;01m-\u001b[39;00m\u001b[38;5;129;01m>\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n",
            "        \u001b[38;5;30;03m# function attr dict\u001b[39;00m\n",
            "        T\u001b[38;5;129;01m.\u001b[39;00mfunc_attr({\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mglobal_symbol\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124madd1\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtir.noalias\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;28;01mTrue\u001b[39;00m})\n",
            "        \u001b[38;5;30;03m# body\u001b[39;00m\n",
            "        \u001b[38;5;30;03m# with T.block(\"root\")\u001b[39;00m\n",
            "        \u001b[38;5;28;01mfor\u001b[39;00m i0, i1 \u001b[38;5;28;01min\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mgrid(\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m10\u001b[39m):\n",
            "            \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mblock(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mT_add\u001b[39m\u001b[38;5;124m\"\u001b[39m):\n",
            "                ax0, ax1 \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00maxis\u001b[38;5;129;01m.\u001b[39;00mremap(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mSS\u001b[39m\u001b[38;5;124m\"\u001b[39m, [i0, i1])\n",
            "                T\u001b[38;5;129;01m.\u001b[39;00mreads(rxplaceholder[ax0, ax1], rxplaceholder_1[ax1])\n",
            "                T\u001b[38;5;129;01m.\u001b[39;00mwrites(T_add[ax0, ax1])\n",
            "                T_add[ax0, ax1] \u001b[38;5;129;01m=\u001b[39;00m rxplaceholder[ax0, ax1] \u001b[38;5;129;01m+\u001b[39;00m rxplaceholder_1[ax1]\n",
            "    \n",
            "    \u001b[38;5;129m@T\u001b[39m\u001b[38;5;129;01m.\u001b[39;00mprim_func\n",
            "    \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mdense\u001b[39m(rxplaceholder: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m784\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], rxplaceholder_1: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m784\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m], T_matmul_NT: T\u001b[38;5;129;01m.\u001b[39;00mBuffer[(\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m]) \u001b[38;5;129;01m-\u001b[39;00m\u001b[38;5;129;01m>\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n",
            "        \u001b[38;5;30;03m# function attr dict\u001b[39;00m\n",
            "        T\u001b[38;5;129;01m.\u001b[39;00mfunc_attr({\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mglobal_symbol\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mdense\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtir.noalias\u001b[39m\u001b[38;5;124m\"\u001b[39m: \u001b[38;5;28;01mTrue\u001b[39;00m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mlayout_free_buffers\u001b[39m\u001b[38;5;124m\"\u001b[39m: [\u001b[38;5;28m1\u001b[39m]})\n",
            "        \u001b[38;5;30;03m# body\u001b[39;00m\n",
            "        \u001b[38;5;30;03m# with T.block(\"root\")\u001b[39;00m\n",
            "        \u001b[38;5;28;01mfor\u001b[39;00m i0, i1, i2 \u001b[38;5;28;01min\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mgrid(\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m, \u001b[38;5;28m784\u001b[39m):\n",
            "            \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mblock(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mT_matmul_NT\u001b[39m\u001b[38;5;124m\"\u001b[39m):\n",
            "                i, j, k \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00maxis\u001b[38;5;129;01m.\u001b[39;00mremap(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mSSR\u001b[39m\u001b[38;5;124m\"\u001b[39m, [i0, i1, i2])\n",
            "                T\u001b[38;5;129;01m.\u001b[39;00mreads(rxplaceholder[i, k], rxplaceholder_1[j, k])\n",
            "                T\u001b[38;5;129;01m.\u001b[39;00mwrites(T_matmul_NT[i, j])\n",
            "                \u001b[38;5;28;01mwith\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00minit():\n",
            "                    T_matmul_NT[i, j] \u001b[38;5;129;01m=\u001b[39;00m T\u001b[38;5;129;01m.\u001b[39;00mfloat32(\u001b[38;5;28m0\u001b[39m)\n",
            "                T_matmul_NT[i, j] \u001b[38;5;129;01m=\u001b[39;00m T_matmul_NT[i, j] \u001b[38;5;129;01m+\u001b[39;00m rxplaceholder[i, k] \u001b[38;5;129;01m*\u001b[39;00m rxplaceholder_1[j, k]\n",
            "    \n",
            "    \u001b[38;5;129m@R\u001b[39m\u001b[38;5;129;01m.\u001b[39;00mfunction\n",
            "    \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mmain\u001b[39m(x: Tensor((\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m784\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m)) \u001b[38;5;129;01m-\u001b[39;00m\u001b[38;5;129;01m>\u001b[39;00m Tensor(\u001b[38;5;28;01mNone\u001b[39;00m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m, ndim \u001b[38;5;129;01m=\u001b[39;00m \u001b[38;5;28m2\u001b[39m):\n",
            "        \u001b[38;5;30;03m# block 0\u001b[39;00m\n",
            "        \u001b[38;5;28;01mwith\u001b[39;00m R\u001b[38;5;129;01m.\u001b[39;00mdataflow():\n",
            "            lv \u001b[38;5;129;01m=\u001b[39;00m R\u001b[38;5;129;01m.\u001b[39;00mcall_tir(dense, (x, meta[relay\u001b[38;5;129;01m.\u001b[39;00mConstant][\u001b[38;5;28m0\u001b[39m]), (\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m), dtype\u001b[38;5;129;01m=\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n",
            "            lv1 \u001b[38;5;129;01m=\u001b[39;00m R\u001b[38;5;129;01m.\u001b[39;00mcall_tir(add, (lv, meta[relay\u001b[38;5;129;01m.\u001b[39;00mConstant][\u001b[38;5;28m1\u001b[39m]), (\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m), dtype\u001b[38;5;129;01m=\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n",
            "            lv2 \u001b[38;5;129;01m=\u001b[39;00m R\u001b[38;5;129;01m.\u001b[39;00mcall_tir(te_relu, (lv1,), (\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m), dtype\u001b[38;5;129;01m=\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n",
            "            lv3 \u001b[38;5;129;01m=\u001b[39;00m R\u001b[38;5;129;01m.\u001b[39;00mcall_tir(dense1, (lv2, meta[relay\u001b[38;5;129;01m.\u001b[39;00mConstant][\u001b[38;5;28m2\u001b[39m]), (\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m10\u001b[39m), dtype\u001b[38;5;129;01m=\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n",
            "            lv4 \u001b[38;5;129;01m=\u001b[39;00m R\u001b[38;5;129;01m.\u001b[39;00mcall_tir(add1, (lv3, meta[relay\u001b[38;5;129;01m.\u001b[39;00mConstant][\u001b[38;5;28m3\u001b[39m]), (\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m10\u001b[39m), dtype\u001b[38;5;129;01m=\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n",
            "            gv: Tensor((\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m10\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m) \u001b[38;5;129;01m=\u001b[39;00m lv4\n",
            "            R\u001b[38;5;129;01m.\u001b[39;00moutput(gv)\n",
            "        \u001b[38;5;28;01mreturn\u001b[39;00m lv4\n",
            "    \n",
            "\n"
          ]
        }
      ],
      "source": [
        "from tvm import topi\n",
        "\n",
        "def map_nn_linear(bb, node_map, node, nn_mod):\n",
        "    x = node_map[node.args[0]]\n",
        "    w = map_param(nn_mod.weight)\n",
        "    if nn_mod.bias is not None:\n",
        "        b = map_param(nn_mod.bias)\n",
        "    y = bb.emit_te(topi.nn.dense, x, w)\n",
        "    return bb.emit_te(topi.add, y, b)\n",
        "\n",
        "def map_nn_relu(bb, node_map, node, nn_mod):\n",
        "    return map_relu(bb, node_map, node)\n",
        "\n",
        "\n",
        "MLPModule = from_fx(\n",
        "    fx.symbolic_trace(mlp_model), \n",
        "    input_shapes = [(1, 784)], \n",
        "    call_function_map={\n",
        "    },\n",
        "    call_module_map={\n",
        "        torch.nn.Linear: map_nn_linear,\n",
        "        torch.nn.ReLU: map_nn_relu,\n",
        "    },\n",
        ")\n",
        "\n",
        "MLPModule.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 33,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "WAg9RZV2lbbK",
        "outputId": "476cd64c-6dca-4e3f-e9da-94cf81eb02be"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "MLPModule Prediction: Bag\n"
          ]
        }
      ],
      "source": [
        "ex = relax.vm.build(MLPModule, target=\"llvm\")\n",
        "vm = relax.VirtualMachine(ex, tvm.cpu())\n",
        "data_nd = tvm.nd.array(img.reshape(1, 784))\n",
        "\n",
        "nd_res = vm[\"main\"](data_nd)\n",
        "\n",
        "pred_kind = np.argmax(nd_res.numpy(), axis=1)\n",
        "print(\"MLPModule Prediction:\", class_names[pred_kind[0]])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3BSOnEbUgu60"
      },
      "source": [
        "## Remark: Translating into High-level Operators\n",
        "\n",
        "In most machine learning frameworks, it is sometimes helpful to first translate into high-level built-in primitive operators. The following code block gives an example to do that.\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 34,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YOo0H-5mhOGl",
        "outputId": "e3bca65f-6504-41c2-b41a-4120bf4e5ab9"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\u001b[38;5;129m@tvm\u001b[39m\u001b[38;5;129;01m.\u001b[39;00mscript\u001b[38;5;129;01m.\u001b[39;00mir_module\n",
            "\u001b[38;5;28;01mclass\u001b[39;00m \u001b[38;5;21;01mModule\u001b[39;00m:\n",
            "    \u001b[38;5;129m@R\u001b[39m\u001b[38;5;129;01m.\u001b[39;00mfunction\n",
            "    \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mmain\u001b[39m(x: Tensor((\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m784\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m)) \u001b[38;5;129;01m-\u001b[39;00m\u001b[38;5;129;01m>\u001b[39;00m Tensor(\u001b[38;5;28;01mNone\u001b[39;00m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m, ndim \u001b[38;5;129;01m=\u001b[39;00m \u001b[38;5;28m2\u001b[39m):\n",
            "        \u001b[38;5;30;03m# block 0\u001b[39;00m\n",
            "        \u001b[38;5;28;01mwith\u001b[39;00m R\u001b[38;5;129;01m.\u001b[39;00mdataflow():\n",
            "            lv: Tensor((\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m) \u001b[38;5;129;01m=\u001b[39;00m relax\u001b[38;5;129;01m.\u001b[39;00mnn\u001b[38;5;129;01m.\u001b[39;00mdense(x, meta[relay\u001b[38;5;129;01m.\u001b[39;00mConstant][\u001b[38;5;28m0\u001b[39m])\n",
            "            lv1: Tensor((\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m) \u001b[38;5;129;01m=\u001b[39;00m relax\u001b[38;5;129;01m.\u001b[39;00madd(lv, meta[relay\u001b[38;5;129;01m.\u001b[39;00mConstant][\u001b[38;5;28m1\u001b[39m])\n",
            "            lv2: Tensor((\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m128\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m) \u001b[38;5;129;01m=\u001b[39;00m relax\u001b[38;5;129;01m.\u001b[39;00mnn\u001b[38;5;129;01m.\u001b[39;00mrelu(lv1)\n",
            "            lv3: Tensor((\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m10\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m) \u001b[38;5;129;01m=\u001b[39;00m relax\u001b[38;5;129;01m.\u001b[39;00mnn\u001b[38;5;129;01m.\u001b[39;00mdense(lv2, meta[relay\u001b[38;5;129;01m.\u001b[39;00mConstant][\u001b[38;5;28m2\u001b[39m])\n",
            "            lv4: Tensor((\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m10\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m) \u001b[38;5;129;01m=\u001b[39;00m relax\u001b[38;5;129;01m.\u001b[39;00madd(lv3, meta[relay\u001b[38;5;129;01m.\u001b[39;00mConstant][\u001b[38;5;28m3\u001b[39m])\n",
            "            gv: Tensor((\u001b[38;5;28m1\u001b[39m, \u001b[38;5;28m10\u001b[39m), \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mfloat32\u001b[39m\u001b[38;5;124m\"\u001b[39m) \u001b[38;5;129;01m=\u001b[39;00m lv4\n",
            "            R\u001b[38;5;129;01m.\u001b[39;00moutput(gv)\n",
            "        \u001b[38;5;28;01mreturn\u001b[39;00m lv4\n",
            "    \n",
            "\n"
          ]
        }
      ],
      "source": [
        "def map_nn_relu_op(bb, node_map, node, nn_mod):\n",
        "    A = node_map[node.args[0]]\n",
        "    return bb.emit(relax.op.relu(A))\n",
        "\n",
        "def map_nn_linear_op(bb, node_map, node, nn_mod):\n",
        "    x = node_map[node.args[0]]\n",
        "    w = map_param(nn_mod.weight)\n",
        "    if nn_mod.bias is not None:\n",
        "        b = map_param(nn_mod.bias)\n",
        "    y = bb.emit(relax.op.dense(x, w))\n",
        "    return bb.emit(relax.op.add(y, b))\n",
        "\n",
        "MLPModuleHighLevel = from_fx(\n",
        "    fx.symbolic_trace(mlp_model), \n",
        "    input_shapes = [(1, 784)], \n",
        "    call_function_map={\n",
        "    },\n",
        "    call_module_map={\n",
        "        torch.nn.Linear: map_nn_linear_op,\n",
        "        torch.nn.ReLU: map_nn_relu_op,\n",
        "    },\n",
        ")\n",
        "\n",
        "MLPModuleHighLevel.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eQvtGR0rqci1"
      },
      "source": [
        "After we get the model into IRModule with those built-in operator calls.\n",
        "These built-in operators are **higher-level abstraction** than the TensorIR functions. There can be different opportunities to further translate these primitive operators into either library or TensorIR functions.\n",
        "\n",
        "In most cases, it can be helpful to translate into high-level builtins when they are available. However, there are many cases where we cannot find the corresponding high-level built-in or when we want to specify the TensorIR function directly. In those cases, we can customize the translation logic or transformation to generate `call_tir` or call into the library functions. Usually, we can get the best result by combining the high-level op, TensorIR, and library abstractions. We will discuss the tradeoffs in the follow-up lectures."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bMHXI_ccFS-z"
      },
      "source": [
        "## Discussions\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VkB5oHttOY0U"
      },
      "source": [
        "In this chapter, we focus on the **develop** part of the MLC flow. We studied different ways to get models from machine learning frameworks onto the IRModule. We also briefly touched upon the high-level primitive operators.\n",
        "\n",
        "Once we get the model into the IRModule, we can introduce more kinds of transformations on primitive functions and computational graph functions. A good MLC process composes these transformations together to form an end deployment form.\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "y2KrBILMsNGf"
      },
      "source": [
        "![image.png](data:image/png;base64,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)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pZGWq5EjJ0BA"
      },
      "source": [
        "## Summary\n",
        "\n",
        "- Tensor expression API allows us to create a primitive TensorIR function.\n",
        "- BlockBuilder API creates IRModule through `emit_te` and other functions.\n",
        "- Integrate with existing machine learning frameworks by transforming models into an IRModule."
      ]
    }
  ],
  "metadata": {
    "colab": {
      "collapsed_sections": [],
      "include_colab_link": true,
      "name": "6 Integration with Machine Learning Frameworks.ipynb",
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3.10.4 ('mlc': conda)",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.10.4"
    },
    "vscode": {
      "interpreter": {
        "hash": "d8a760899c905ec5a15e0d212432af25d7f0b614c7ae634224dffa77837bb03c"
      }
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}