{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"#| default_exp app"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/home/long_ng/anaconda3/envs/pytorch/lib/python3.9/site-packages/scipy/__init__.py:146: UserWarning: A NumPy version >=1.16.5 and <1.23.0 is required for this version of SciPy (detected version 1.23.1\n",
" warnings.warn(f\"A NumPy version >={np_minversion} and <{np_maxversion}\"\n"
]
}
],
"source": [
"#|export\n",
"from fastai.vision.all import *\n",
"from huggingface_hub import push_to_hub_fastai, from_pretrained_fastai\n",
"import gradio as gr"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
"PILImage mode=RGB size=192x188"
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"im = PILImage.create('IP1.jpg')\n",
"im.thumbnail((192,192))\n",
"im"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [],
"source": [
"#|export\n",
"learner = load_learner('export.pkl')"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n"
],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"CPU times: user 799 ms, sys: 0 ns, total: 799 ms\n",
"Wall time: 94.7 ms\n"
]
},
{
"data": {
"text/plain": [
"('Noise', TensorBase(2), TensorBase([2.0159e-12, 1.9818e-06, 1.0000e+00]))"
]
},
"execution_count": 8,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"%time learner.predict(im)"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [],
"source": [
"#|export\n",
"categories = ('In-phase', 'Invert-Phase', 'Noise')\n",
"\n",
"def classify_image(img):\n",
" pred, idx, probs = learner.predict(img)\n",
" return dict(zip(categories, map(float, probs)))"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n"
],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"{'In-phase': 0.9999990463256836,\n",
" 'Invert-Phase': 9.950312005457818e-07,\n",
" 'Noise': 1.0946109574305751e-09}"
]
},
"execution_count": 12,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"classify_image(im)"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/home/long_ng/anaconda3/envs/pytorch/lib/python3.9/site-packages/gradio/inputs.py:256: UserWarning: Usage of gradio.inputs is deprecated, and will not be supported in the future, please import your component from gradio.components\n",
" warnings.warn(\n",
"/home/long_ng/anaconda3/envs/pytorch/lib/python3.9/site-packages/gradio/deprecation.py:40: UserWarning: `optional` parameter is deprecated, and it has no effect\n",
" warnings.warn(value)\n",
"/home/long_ng/anaconda3/envs/pytorch/lib/python3.9/site-packages/gradio/outputs.py:196: UserWarning: Usage of gradio.outputs is deprecated, and will not be supported in the future, please import your components from gradio.components\n",
" warnings.warn(\n",
"/home/long_ng/anaconda3/envs/pytorch/lib/python3.9/site-packages/gradio/deprecation.py:40: UserWarning: The 'type' parameter has been deprecated. Use the Number component instead.\n",
" warnings.warn(value)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Running on local URL: http://127.0.0.1:7860/\n",
"\n",
"To create a public link, set `share=True` in `launch()`.\n"
]
},
{
"data": {
"text/plain": [
"(, 'http://127.0.0.1:7860/', None)"
]
},
"execution_count": 13,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"text/html": [
"\n",
"\n"
],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"\n",
"\n"
],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"\n",
"\n"
],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"\n",
"\n"
],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"#|export\n",
"image = gr.inputs.Image(shape=(192,192))\n",
"label = gr.outputs.Label()\n",
"examples = ['IVP1.jpg','IVP2.jpg', 'IP1.jpg', 'IP2.jpg','Noise1.jpg', 'Noise2.jpg']\n",
"\n",
"intf = gr.Interface(fn=classify_image, inputs=image, outputs=label, examples=examples)\n",
"intf.launch(inline=False)"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3.9.12 ('pytorch')",
"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": "3.9.12"
},
"orig_nbformat": 4,
"vscode": {
"interpreter": {
"hash": "116405ed73c8aa81284e79f7dbed5bb2183b3f2b3168f79d61a0664a2160e2ed"
}
}
},
"nbformat": 4,
"nbformat_minor": 2
}