deploy: bda16b2

salesforce · Apr 25, 2024 · b74e528 · b74e528
1 parent 36f9d6b
commit b74e528
Show file tree

Hide file tree

Showing 2 changed files with 26 additions and 26 deletions.
diff --git a/latest/tutorials/vision/feature_visualization_tf.ipynb b/latest/tutorials/vision/feature_visualization_tf.ipynb
@@ -2,15 +2,15 @@
  "cells": [
   {
    "cell_type": "markdown",
-   "id": "2e5dbb39",
+   "id": "e2805f5e",
    "metadata": {},
    "source": [
     "### Feature visualization (Tensorflow)"
    ]
   },
   {
    "cell_type": "markdown",
-   "id": "9160f798",
+   "id": "e52bf8a4",
    "metadata": {},
    "source": [
     "This is an example of feature visualization with a Tensorflow model. The feature visualization in OmniXAI is an optimization-based method, allowing to set different objectives, e.g., layer, channel, neuron or direction. For more information, please visit https://distill.pub/2017/feature-visualization/"
@@ -19,7 +19,7 @@
   {
    "cell_type": "code",
    "execution_count": 1,
-   "id": "fe216f24",
+   "id": "55c1dae6",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -31,7 +31,7 @@
   {
    "cell_type": "code",
    "execution_count": 2,
-   "id": "452c2af5",
+   "id": "508b2b13",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -41,7 +41,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "ce291673",
+   "id": "1ebdd662",
    "metadata": {},
    "source": [
     "Here we choose the VGG16 model for demonstration (you may test other CNN models, e.g., ResNet). The target layer is the layer to analyze."
@@ -50,7 +50,7 @@
   {
    "cell_type": "code",
    "execution_count": 3,
-   "id": "64b57fd0",
+   "id": "e0632243",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -60,7 +60,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "15e3c800",
+   "id": "1f770115",
    "metadata": {},
    "source": [
     "The first example is the \"layer\" objective, where we optimize the input image such that the average output of the layer is maximized."
@@ -69,7 +69,7 @@
   {
    "cell_type": "code",
    "execution_count": 4,
-   "id": "dc721bae",
+   "id": "6430f852",
    "metadata": {},
    "outputs": [
     {
@@ -101,7 +101,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "e13cbc3a",
+   "id": "392d304b",
    "metadata": {},
    "source": [
     "The second example is the \"channel\" objective, where the input image is optimized such that the output of the specified channel is maximized."
@@ -110,7 +110,7 @@
   {
    "cell_type": "code",
    "execution_count": 5,
-   "id": "61799385",
+   "id": "bd0d943f",
    "metadata": {},
    "outputs": [
     {
@@ -142,7 +142,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "3df2eb9f",
+   "id": "93766ac9",
    "metadata": {},
    "source": [
     "We can also consider a combination of multiple objectives. The default weight for each objective is 1.0. We can set other weights as well."
@@ -151,7 +151,7 @@
   {
    "cell_type": "code",
    "execution_count": 6,
-   "id": "0242caef",
+   "id": "b718fb4d",
    "metadata": {},
    "outputs": [
     {

diff --git a/latest/tutorials/vision/feature_visualization_torch.ipynb b/latest/tutorials/vision/feature_visualization_torch.ipynb
@@ -2,15 +2,15 @@
  "cells": [
   {
    "cell_type": "markdown",
-   "id": "cbe48f6e",
+   "id": "1e7671a7",
    "metadata": {},
    "source": [
     "### Feature visualization (PyTorch)"
    ]
   },
   {
    "cell_type": "markdown",
-   "id": "c1217a2c",
+   "id": "654d78c7",
    "metadata": {},
    "source": [
     "This is an example of feature visualization with a Tensorflow model. The feature visualization in OmniXAI is an optimization-based method, allowing to set different objectives, e.g., layer, channel, neuron or direction. For more information, please visit https://distill.pub/2017/feature-visualization/"
@@ -19,7 +19,7 @@
   {
    "cell_type": "code",
    "execution_count": 1,
-   "id": "8e6f9fc6",
+   "id": "d3ae7a16",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -31,7 +31,7 @@
   {
    "cell_type": "code",
    "execution_count": 2,
-   "id": "8e9f5cdb",
+   "id": "543d5aee",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -42,7 +42,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "b29ed704",
+   "id": "99aebcd5",
    "metadata": {},
    "source": [
     "Here we choose the VGG16 model for demonstration (you may test other CNN models, e.g., ResNet). The target layer is the layer to analyze."
@@ -51,7 +51,7 @@
   {
    "cell_type": "code",
    "execution_count": 3,
-   "id": "38797d3e",
+   "id": "bad44ed4",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -62,7 +62,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "4eb35c9f",
+   "id": "8c8cba47",
    "metadata": {},
    "source": [
     "The first example is the \"layer\" objective, where we optimize the input image such that the average output of the layer is maximized."
@@ -71,7 +71,7 @@
   {
    "cell_type": "code",
    "execution_count": 4,
-   "id": "35faf532",
+   "id": "ff364761",
    "metadata": {},
    "outputs": [
     {
@@ -103,7 +103,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "6b8c3fdf",
+   "id": "a9eda779",
    "metadata": {},
    "source": [
     "The second example is the \"channel\" objective, where the input image is optimized such that the output of the specified channel is maximized."
@@ -112,7 +112,7 @@
   {
    "cell_type": "code",
    "execution_count": 5,
-   "id": "c231b61b",
+   "id": "54ba41fb",
    "metadata": {},
    "outputs": [
     {
@@ -144,7 +144,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "c3d87621",
+   "id": "47655dbd",
    "metadata": {},
    "source": [
     "We can also consider a combination of multiple objectives. The default weight for each objective is 1.0. We can set other weights as well."
@@ -153,7 +153,7 @@
   {
    "cell_type": "code",
    "execution_count": 6,
-   "id": "d1d7e16b",
+   "id": "47d62fed",
    "metadata": {
     "scrolled": true
    },
@@ -190,7 +190,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "b68212ff",
+   "id": "9232e2f7",
    "metadata": {},
    "source": [
     "Let's try another target layer and use FFT preconditioning:"
@@ -199,7 +199,7 @@
   {
    "cell_type": "code",
    "execution_count": 7,
-   "id": "6ec1c67f",
+   "id": "41bb6a0e",
    "metadata": {},
    "outputs": [
     {