Check point

.gitignore

@@ -0,0 +1,4 @@
+.ipynb_checkpoints
+__pycache__
+.pyc
+.DS_Store

dataset.py

@@ -0,0 +1,31 @@
+from torch.utils.data import Dataset
+from PIL import Image
+import os
+import re
+import torch
+
+class AgeDataset(Dataset):
+    def __init__(self, target_dir, transform=None):
+        input_file_format = ('png', 'jpg', 'jpeg')
+
+        self.target_dir = target_dir
+        self.paths = [f for f in os.listdir(self.target_dir) if f.lower().endswith(input_file_format)]
+        self.transform = transform
+        
+    def load_img(self, idx):
+        img_path = os.path.join(self.target_dir, self.paths[idx])
+        target_value = re.search(r"(\d+)_", self.paths[idx])
+        target_value = int(target_value.group(1))
+
+        return Image.open(img_path).convert("RGB"), torch.tensor(target_value, dtype=torch.float).unsqueeze(dim=0)
+        
+    def __len__(self):
+        return len(self.paths)
+        
+    def __getitem__(self, idx):
+        img, target_value = self.load_img(idx)
+            
+        if self.transform:
+            img = self.transform(img)
+        
+        return img, target_value

pretrained_weight/vit_medium_patch16_clip_224.tinyclip_yfcc15m(trainable 0.00) (eval Score 0.9067, Loss 29.465482).pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:395d85bd8ff805c07120fd64407dbc47e1352d3ddc6876a63b309ce9072a0100
+size 154448566

test_gradio.ipynb

@@ -0,0 +1,175 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "9782f4d6-2dc8-44a7-bb91-6636a71415e8",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import gradio as gr\n",
+    "import time\n",
+    "from PIL import Image\n",
+    "import torch\n",
+    "from torch import nn\n",
+    "import timm\n",
+    "from custom_torch_module import setup_utils"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "7cfc4e19-e2c5-41d0-b105-9e4a661001c4",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "title = \"Age Prediction model\"\n",
+    "description = \"ViT(medium clip) based model. transfer trained with custom dataset\"\n",
+    "article = \"Through bunch of fine tuning and experiments. REMEMBER! This model can be wrong.\""
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "504cd630-6e4e-43b8-b444-d49775bbcf5d",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<All keys matched successfully>"
+      ]
+     },
+     "execution_count": 3,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "MODEL_NAME = \"vit_medium_patch16_clip_224.tinyclip_yfcc15m\"\n",
+    "FILE_NAME = \"pretrained_weight/vit_medium_patch16_clip_224.tinyclip_yfcc15m(trainable 0.00) (eval Score 0.9067, Loss 29.465482).pth\"\n",
+    "DEVICE = \"cpu\"\n",
+    "\n",
+    "torch.set_default_device(DEVICE)\n",
+    "\n",
+    "model = timm.create_model(MODEL_NAME, pretrained=True, num_classes=0, drop_rate=0.7)\n",
+    "\n",
+    "model_classifier = nn.Sequential(nn.Linear(512, 512),\n",
+    "                                nn.BatchNorm1d(512),\n",
+    "                                nn.GELU(),\n",
+    "                                nn.Linear(512, 1))\n",
+    "\n",
+    "model = nn.Sequential(model, model_classifier)\n",
+    "\n",
+    "test_transform = setup_utils.build_transform(img_size=224, is_data_aug=False)\n",
+    "\n",
+    "model.load_state_dict(state_dict=torch.load(FILE_NAME, weights_only=True))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "id": "1dba4088-961a-4f14-8d1e-2f9044a652e3",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def predict(img):\n",
+    "    start_time = time.time()\n",
+    "    model.eval()\n",
+    "    with torch.inference_mode():\n",
+    "        img = test_transform(img).unsqueeze(dim=0).to(DEVICE)\n",
+    "        pred_age = model(img).item()\n",
+    "    \n",
+    "    end_time = time.time()\n",
+    "    \n",
+    "    elapsed_time = end_time - start_time\n",
+    "    fps = 1 / elapsed_time\n",
+    "    return pred_age, fps\n",
+    "\n",
+    "# img = Image.open(img_path[0]).convert(\"RGB\")\n",
+    "# pred_label_and_probs, elapsed_time = predict(img)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "63313e6f-f86f-4d42-a8e1-986fa394e51a",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Running on local URL:  http://127.0.0.1:7860\n",
+      "Running on public URL: https://3b12d13cbee982d501.gradio.live\n",
+      "\n",
+      "This share link expires in 72 hours. For free permanent hosting and GPU upgrades, run `gradio deploy` from Terminal to deploy to Spaces (https://huggingface.co/spaces)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div><iframe src=\"https://3b12d13cbee982d501.gradio.live\" width=\"100%\" height=\"500\" allow=\"autoplay; camera; microphone; clipboard-read; clipboard-write;\" frameborder=\"0\" allowfullscreen></iframe></div>"
+      ],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "# Create the Gradio demo\n",
+    "demo = gr.Interface(fn=predict, \n",
+    "                    inputs=gr.Image(type=\"pil\"),\n",
+    "                    outputs=[gr.Number(label=\"Age Prediction\"),\n",
+    "                             gr.Number(label=\"Prediction speed (fps)\")], \n",
+    "                    title=title,\n",
+    "                    description=description,\n",
+    "                    article=article)\n",
+    "\n",
+    "# Launch the demo!\n",
+    "demo.launch(debug=True, # print errors locally?\n",
+    "            share=True) # generate a publically shareable URL?"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "9a17b040-b662-45ca-92e7-517c52bc5950",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "f58e0cc1-22b1-4dcc-935a-4d9c98961c9e",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "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": "3.12.3"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}