load swinb dino

Author: dnth

examples/enrichment-ram-groundingdino-sam.ipynb

@@ -718,7 +718,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 11,
+   "execution_count": 32,
    "id": "5cc1e1b8-799c-461f-b08e-a95590be9a60",
    "metadata": {
     "tags": []
@@ -768,8 +768,8 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 12,
-   "id": "9fdd1d2f-4df4-422d-89a7-bbec9bf8a1dc",
+   "execution_count": 33,
+   "id": "49f8a1b8-5542-48ee-9b78-ae653640c38c",
    "metadata": {
     "tags": []
    },
@@ -794,7 +794,121 @@
        "  (2.35, 277.69, 637.63, 425.32)]}"
       ]
      },
-     "execution_count": 12,
+     "execution_count": 33,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "results"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "55d02a9a-32c1-430d-8ca8-ca10ac4b8712",
+   "metadata": {},
+   "source": [
+    "Load another SwinB variant of Grounding DINO. Weights and config can be downloaded from the [official Grounding DINO repo](https://github.com/IDEA-Research/GroundingDINO)."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 30,
+   "id": "330f87fa-fa3f-416d-a2cd-a63b52a3d8f1",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "INFO:fastdup.models.grounding_dino:Loading model checkpoint from - groundingdino_swinb_cogcoor.pth\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "final text_encoder_type: bert-base-uncased\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Some weights of the model checkpoint at bert-base-uncased were not used when initializing BertModel: ['cls.predictions.bias', 'cls.predictions.transform.dense.bias', 'cls.predictions.transform.dense.weight', 'cls.predictions.transform.LayerNorm.bias', 'cls.seq_relationship.weight', 'cls.predictions.transform.LayerNorm.weight', 'cls.seq_relationship.bias']\n",
+      "- This IS expected if you are initializing BertModel from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).\n",
+      "- This IS NOT expected if you are initializing BertModel from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).\n",
+      "INFO:fastdup.models.grounding_dino:Model loaded on device - cuda\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "final text_encoder_type: bert-base-uncased\n"
+     ]
+    }
+   ],
+   "source": [
+    "model = GroundingDINO(model_config=\"GroundingDINO_SwinB_cfg.py\", model_weights=\"groundingdino_swinb_cogcoor.pth\")\n",
+    "results = model.run_inference(image_path=\"coco_minitrain_25k/images/val2017/000000449996.jpg\",\n",
+    "                    text_prompt=\"air field . airliner . plane . airport . airport runway . airport terminal . jet . land . park . raceway . sky . tarmac . terminal\",\n",
+    "                    box_threshold=0.3,\n",
+    "                    text_threshold=0.25)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 31,
+   "id": "9fdd1d2f-4df4-422d-89a7-bbec9bf8a1dc",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "{'labels': ['sky',\n",
+       "  'plane',\n",
+       "  'plane',\n",
+       "  'plane',\n",
+       "  'plane',\n",
+       "  'jet',\n",
+       "  'jet',\n",
+       "  'jet',\n",
+       "  'plane',\n",
+       "  'jet',\n",
+       "  'plane',\n",
+       "  'tarmac'],\n",
+       " 'scores': [0.6725,\n",
+       "  0.5171,\n",
+       "  0.5012,\n",
+       "  0.5205,\n",
+       "  0.4961,\n",
+       "  0.448,\n",
+       "  0.3454,\n",
+       "  0.3646,\n",
+       "  0.4127,\n",
+       "  0.3492,\n",
+       "  0.3438,\n",
+       "  0.3614],\n",
+       " 'boxes': [(0.8, 0.64, 639.07, 243.63),\n",
+       "  (143.06, 248.17, 261.81, 296.37),\n",
+       "  (444.0, 112.47, 496.71, 129.99),\n",
+       "  (113.93, 289.86, 244.38, 340.29),\n",
+       "  (392.88, 272.05, 463.05, 294.73),\n",
+       "  (114.1, 290.14, 244.1, 340.02),\n",
+       "  (393.09, 272.23, 462.76, 294.51),\n",
+       "  (143.47, 248.59, 261.61, 296.13),\n",
+       "  (551.81, 258.11, 639.07, 297.51),\n",
+       "  (444.24, 112.87, 496.45, 129.68),\n",
+       "  (501.01, 282.23, 558.72, 296.47),\n",
+       "  (1.46, 274.81, 638.57, 426.26)]}"
+      ]
+     },
+     "execution_count": 31,
      "metadata": {},
      "output_type": "execute_result"
     }