{ "id": 414918, "title": "It was written by my submission code is there a problem?", "url": "/competitions/image-matching-challenge-2023/discussion/414918", "author_name": "", "post_date": "2023-06-04T04:18:30.997530Z", "votes": null, "comment_count": 2, "views": 0, "content": "

Why did my code submit a 0.001 score and run in 3 minutes
\nIt was written by my submission is there a problem?

\n

`import gc

\n

gc.collect()
\ndatasets = []

\n

for dataset in data_dict:
\n datasets.append(dataset)

\n

for dataset in datasets:
\n print(dataset)
\n if dataset not in out_results:
\n out_results[dataset] = {}
\n for scene in data_dict[dataset]:
\n print(scene)
\n # Fail gently if the notebook has not been submitted and the test data is not populated.
\n # You may want to run this on the training data in that case?
\n img_dir = f'{src}/test/{dataset}/{scene}/images'
\n if not os.path.exists(img_dir):
\n continue
\n # Wrap the meaty part in a try-except block.
\n try:
\n out_results[dataset][scene] = {}
\n img_fnames = [f'{src}/test/{x}' for x in data_dict[dataset][scene]]
\n print (f\"Got {len(img_fnames)} images\")
\n n = 16
\n new_fnames = [img_fnames[i:i+n] for i in range(0, len(img_fnames), n)]
\n if len(new_fnames[-1]) < n:
\n new_fnames[-2:] = [new_fnames[-2] + new_fnames[-1]]
\n M_world = []
\n for index,item in enumerate(new_fnames):

\n
            img_tensor,camera = (item)\n            model = ()\n            checkpoint = torch()\n            model(checkpoint)\n            model()\n            heatmap_pred,desc_pred, = (img_tensor)\n            temp = (heatmap_pred, desc_pred,,camera)\n            M_world =M_world + temp\n         index,name  (data_dict):\n            key1 = f\n            out_results = {}\n            out_results = M_world\n            out_results = M_world\n        (f)\n        (f)\n        (out_results, data_dict)\n        ()\n        gc()\n    except:\n        pass\n
\n

\ncreate_submission(out_results, data_dict)`

", "messages": [ { "id": "2287026", "postDate": "06/04/2023 04:18:30", "content": "

Why did my code submit a 0.001 score and run in 3 minutes
\nIt was written by my submission is there a problem?

\n

`import gc

\n

gc.collect()
\ndatasets = []

\n

for dataset in data_dict:
\n datasets.append(dataset)

\n

for dataset in datasets:
\n print(dataset)
\n if dataset not in out_results:
\n out_results[dataset] = {}
\n for scene in data_dict[dataset]:
\n print(scene)
\n # Fail gently if the notebook has not been submitted and the test data is not populated.
\n # You may want to run this on the training data in that case?
\n img_dir = f'{src}/test/{dataset}/{scene}/images'
\n if not os.path.exists(img_dir):
\n continue
\n # Wrap the meaty part in a try-except block.
\n try:
\n out_results[dataset][scene] = {}
\n img_fnames = [f'{src}/test/{x}' for x in data_dict[dataset][scene]]
\n print (f\"Got {len(img_fnames)} images\")
\n n = 16
\n new_fnames = [img_fnames[i:i+n] for i in range(0, len(img_fnames), n)]
\n if len(new_fnames[-1]) < n:
\n new_fnames[-2:] = [new_fnames[-2] + new_fnames[-1]]
\n M_world = []
\n for index,item in enumerate(new_fnames):

\n
            img_tensor,camera = (item)\n            model = ()\n            checkpoint = torch()\n            model(checkpoint)\n            model()\n            heatmap_pred,desc_pred, = (img_tensor)\n            temp = (heatmap_pred, desc_pred,,camera)\n            M_world =M_world + temp\n         index,name  (data_dict):\n            key1 = f\n            out_results = {}\n            out_results = M_world\n            out_results = M_world\n        (f)\n        (f)\n        (out_results, data_dict)\n        ()\n        gc()\n    except:\n        pass\n
\n

\ncreate_submission(out_results, data_dict)`

", "rawMarkdown": "Why did my code submit a 0.001 score and run in 3 minutes\nIt was written by my submission is there a problem?\n\n`import gc\n\n\n\ngc.collect()\ndatasets = []\n\n\nfor dataset in data_dict:\n datasets.append(dataset)\n\nfor dataset in datasets:\n print(dataset)\n if dataset not in out_results:\n out_results[dataset] = {}\n for scene in data_dict[dataset]:\n print(scene)\n # Fail gently if the notebook has not been submitted and the test data is not populated.\n # You may want to run this on the training data in that case?\n img_dir = f'{src}/test/{dataset}/{scene}/images'\n if not os.path.exists(img_dir):\n continue\n # Wrap the meaty part in a try-except block.\n try:\n out_results[dataset][scene] = {}\n img_fnames = [f'{src}/test/{x}' for x in data_dict[dataset][scene]]\n print (f\"Got {len(img_fnames)} images\")\n n = 16\n new_fnames = [img_fnames[i:i+n] for i in range(0, len(img_fnames), n)]\n if len(new_fnames[-1]) < n:\n new_fnames[-2:] = [new_fnames[-2] + new_fnames[-1]]\n M_world = []\n for index,item in enumerate(new_fnames):\n\n\n img_tensor,camera = collate_img(item)\n model = make_model()\n checkpoint = torch.load(\"/kaggle/input/params/params.pth\")\n model.load_state_dict(checkpoint)\n model.eval()\n heatmap_pred,desc_pred,img = model(img_tensor)\n temp = compute(heatmap_pred, desc_pred,img,camera)\n M_world =M_world + temp\n for index,name in enumerate(data_dict[dataset][scene]):\n key1 = f'{name}'\n out_results[dataset][scene][key1] = {}\n out_results[dataset][scene][key1][\"R\"] = M_world[index][0]\n out_results[dataset][scene][key1][\"t\"] = M_world[index][1]\n print(f'Registered: {dataset} / {scene} -> {len(out_results[dataset][scene])} images')\n print(f'Total: {dataset} / {scene} -> {len(data_dict[dataset][scene])} images')\n create_submission(out_results, data_dict)\n print(\"create_submission成功\")\n gc.collect()\n except:\n pass\n \n`\n`create_submission(out_results, data_dict)`", "votes": null }, { "id": "2287036", "postDate": "06/04/2023 04:53:56", "content": "

fyi: you may try placing all of your code in a codeblock in your question; it will make it easier for others to read (and the whitespace is important)

", "rawMarkdown": "fyi: you may try placing all of your code in a codeblock in your question; it will make it easier for others to read (and the whitespace is important)", "votes": null }, { "id": "2287074", "postDate": "06/04/2023 05:44:08", "content": "

`def compute(heatmap_pred, desc_pred,img, camera, alpha=0.2, eps=1e-6):

\n
matches,match_list = feature\nM_world = cameraworld(matches,camera,match_list)\n\n\nreturn M_world`\n
\n

`import albumentations as A
\nfrom PIL import Image

\n

transform_4 = A.Compose([
\n A.Resize(height=256, width=256, p=1.0)
\n])

\n

def collate_img(img_list):
\n newimages = []

\n
newcamera = \n index,item  (img_list):\n\n    image = cv2(item,cv2.IMREAD_COLOR)\n\n    camera_id = \n     item() != -:\n        camera_id = \n    elif item() != -:\n        camera_id = \n\n\n     = torch((image=image))\n    newimages(img)\n\n\n\n    newcamera(camera_id)\nimages = torch(torch(newimages),dtype=torch.float32)\n\nimages = images(,,,)\n\n\n\nreturn images,newcamera`\n
\n

src = '/kaggle/input/image-matching-challenge-2023'
\ndata_dict = {}\nwith open(f'{src}/sample_submission.csv', 'r') as f:\n for i, l in enumerate(f):\n # Skip header.\n if l and i > 0:\n image, dataset, scene, _, _ = l.strip().split(',')\n if dataset not in data_dict:\n data_dict[dataset] = {}\n if scene not in data_dict[dataset]:\n data_dict[dataset][scene] = []\n data_dict[dataset][scene].append(image)
\nfor dataset in data_dict:\n for scene in data_dict[dataset]:\n print(f'{dataset} / {scene} -> {len(data_dict[dataset][scene])} images')
\n`out_results = {}
\ntimings = {\"shortlisting\":[],
\n \"feature_detection\": [],
\n \"feature_matching\":[],
\n \"RANSAC\": [],
\n \"Reconstruction\": []}
\ndef arr_to_str(a):
\n return ';'.join([str(x) for x in a.reshape(-1)])
\nimport gc

\n

gc.collect()
\ndatasets = []

\n

for dataset in data_dict:
\n datasets.append(dataset)

\n

for dataset in datasets:
\n print(dataset)
\n if dataset not in out_results:
\n out_results[dataset] = {}
\n for scene in data_dict[dataset]:
\n print(scene)
\n # Fail gently if the notebook has not been submitted and the test data is not populated.
\n # You may want to run this on the training data in that case?
\n img_dir = f'{src}/test/{dataset}/{scene}/images'
\n if not os.path.exists(img_dir):
\n continue
\n # Wrap the meaty part in a try-except block.
\n try:
\n out_results[dataset][scene] = {}
\n img_fnames = [f'{src}/test/{x}' for x in data_dict[dataset][scene]]
\n print (f\"Got {len(img_fnames)} images\")
\n n = 16
\n new_fnames = [img_fnames[i:i+n] for i in range(0, len(img_fnames), n)]
\n if len(new_fnames[-1]) < n:
\n new_fnames[-2:] = [new_fnames[-2] + new_fnames[-1]]
\n M_world = []
\n for index,item in enumerate(new_fnames):

\n
            img_tensor,camera = (item)\n            model = ()\n            checkpoint = torch()\n            model(checkpoint)\n            model()\n            heatmap_pred,desc_pred, = (img_tensor)\n            temp = (heatmap_pred, desc_pred,,camera)\n            M_world =M_world + temp\n         index,name  (data_dict):\n            key1 = f\n            out_results = {}\n            out_results = M_world\n            out_results = M_world\n        (f)\n        (f)\n        (out_results, data_dict)\n        ()\n        gc()\n    except:\n        pass\n\n\n        `\n
\n

create_submission(out_results, data_dict)
\nThis is all my code except the model

", "rawMarkdown": "`def compute(heatmap_pred, desc_pred,img, camera, alpha=0.2, eps=1e-6):\n \n \n matches,match_list = feature_matching(heatmap_pred,desc_pred,img)\n M_world = cameraworld(matches,camera,match_list)\n \n \n return M_world`\n`import albumentations as A\nfrom PIL import Image\n\n\ntransform_4 = A.Compose([\n A.Resize(height=256, width=256, p=1.0)\n])\n\ndef collate_img(img_list):\n newimages = []\n \n newcamera = []\n for index,item in enumerate(img_list):\n \n image = cv2.imread(item,cv2.IMREAD_COLOR)\n \n camera_id = 2\n if item[0].find(\"dioscuri\") != -1:\n camera_id = 54\n elif item[0].find(\"kyiv-puppet-theater\") != -1:\n camera_id = 2\n \n \n img = torch.from_numpy(transform_4(image=image)['image'])\n newimages.append(img)\n \n \n \n newcamera.append(camera_id)\n images = torch.tensor(torch.stack(newimages),dtype=torch.float32)\n print(\"images:\",images.shape)\n images = images.permute(0,3,1,2)\n print(\"images:\",images.shape)\n \n \n return images,newcamera`\n`src = '/kaggle/input/image-matching-challenge-2023'`\n`data_dict = {}\nwith open(f'{src}/sample_submission.csv', 'r') as f:\n for i, l in enumerate(f):\n # Skip header.\n if l and i > 0:\n image, dataset, scene, _, _ = l.strip().split(',')\n if dataset not in data_dict:\n data_dict[dataset] = {}\n if scene not in data_dict[dataset]:\n data_dict[dataset][scene] = []\n data_dict[dataset][scene].append(image)`\n`for dataset in data_dict:\n for scene in data_dict[dataset]:\n print(f'{dataset} / {scene} -> {len(data_dict[dataset][scene])} images')`\n`out_results = {}\ntimings = {\"shortlisting\":[],\n \"feature_detection\": [],\n \"feature_matching\":[],\n \"RANSAC\": [],\n \"Reconstruction\": []}\ndef arr_to_str(a):\n return ';'.join([str(x) for x in a.reshape(-1)])\nimport gc\n\n\n\ngc.collect()\ndatasets = []\n\n\nfor dataset in data_dict:\n datasets.append(dataset)\n\nfor dataset in datasets:\n print(dataset)\n if dataset not in out_results:\n out_results[dataset] = {}\n for scene in data_dict[dataset]:\n print(scene)\n # Fail gently if the notebook has not been submitted and the test data is not populated.\n # You may want to run this on the training data in that case?\n img_dir = f'{src}/test/{dataset}/{scene}/images'\n if not os.path.exists(img_dir):\n continue\n # Wrap the meaty part in a try-except block.\n try:\n out_results[dataset][scene] = {}\n img_fnames = [f'{src}/test/{x}' for x in data_dict[dataset][scene]]\n print (f\"Got {len(img_fnames)} images\")\n n = 16\n new_fnames = [img_fnames[i:i+n] for i in range(0, len(img_fnames), n)]\n if len(new_fnames[-1]) < n:\n new_fnames[-2:] = [new_fnames[-2] + new_fnames[-1]]\n M_world = []\n for index,item in enumerate(new_fnames):\n\n\n img_tensor,camera = collate_img(item)\n model = make_model()\n checkpoint = torch.load(\"/kaggle/input/params/DCNV3_4_params.pth\")\n model.load_state_dict(checkpoint)\n model.eval()\n heatmap_pred,desc_pred,img = model(img_tensor)\n temp = compute(heatmap_pred, desc_pred,img,camera)\n M_world =M_world + temp\n for index,name in enumerate(data_dict[dataset][scene]):\n key1 = f'{name}'\n out_results[dataset][scene][key1] = {}\n out_results[dataset][scene][key1][\"R\"] = M_world[index][0]\n out_results[dataset][scene][key1][\"t\"] = M_world[index][1]\n print(f'Registered: {dataset} / {scene} -> {len(out_results[dataset][scene])} images')\n print(f'Total: {dataset} / {scene} -> {len(data_dict[dataset][scene])} images')\n create_submission(out_results, data_dict)\n print(\"create_submission成功\")\n gc.collect()\n except:\n pass\n \n\n `\n`create_submission(out_results, data_dict)`\nThis is all my code except the model", "votes": null } ], "comments": [ { "id": 2287036, "author_name": "joshuabruning", "author_url": "", "post_date": "06/04/2023 04:53:56", "content": "

fyi: you may try placing all of your code in a codeblock in your question; it will make it easier for others to read (and the whitespace is important)

", "votes": null, "replies": [] }, { "id": 2287074, "author_name": "lanshank", "author_url": "", "post_date": "06/04/2023 05:44:08", "content": "

`def compute(heatmap_pred, desc_pred,img, camera, alpha=0.2, eps=1e-6):

\n
matches,match_list = feature\nM_world = cameraworld(matches,camera,match_list)\n\n\nreturn M_world`\n
\n

`import albumentations as A
\nfrom PIL import Image

\n

transform_4 = A.Compose([
\n A.Resize(height=256, width=256, p=1.0)
\n])

\n

def collate_img(img_list):
\n newimages = []

\n
newcamera = \n index,item  (img_list):\n\n    image = cv2(item,cv2.IMREAD_COLOR)\n\n    camera_id = \n     item() != -:\n        camera_id = \n    elif item() != -:\n        camera_id = \n\n\n     = torch((image=image))\n    newimages(img)\n\n\n\n    newcamera(camera_id)\nimages = torch(torch(newimages),dtype=torch.float32)\n\nimages = images(,,,)\n\n\n\nreturn images,newcamera`\n
\n

src = '/kaggle/input/image-matching-challenge-2023'
\ndata_dict = {}\nwith open(f'{src}/sample_submission.csv', 'r') as f:\n for i, l in enumerate(f):\n # Skip header.\n if l and i > 0:\n image, dataset, scene, _, _ = l.strip().split(',')\n if dataset not in data_dict:\n data_dict[dataset] = {}\n if scene not in data_dict[dataset]:\n data_dict[dataset][scene] = []\n data_dict[dataset][scene].append(image)
\nfor dataset in data_dict:\n for scene in data_dict[dataset]:\n print(f'{dataset} / {scene} -> {len(data_dict[dataset][scene])} images')
\n`out_results = {}
\ntimings = {\"shortlisting\":[],
\n \"feature_detection\": [],
\n \"feature_matching\":[],
\n \"RANSAC\": [],
\n \"Reconstruction\": []}
\ndef arr_to_str(a):
\n return ';'.join([str(x) for x in a.reshape(-1)])
\nimport gc

\n

gc.collect()
\ndatasets = []

\n

for dataset in data_dict:
\n datasets.append(dataset)

\n

for dataset in datasets:
\n print(dataset)
\n if dataset not in out_results:
\n out_results[dataset] = {}
\n for scene in data_dict[dataset]:
\n print(scene)
\n # Fail gently if the notebook has not been submitted and the test data is not populated.
\n # You may want to run this on the training data in that case?
\n img_dir = f'{src}/test/{dataset}/{scene}/images'
\n if not os.path.exists(img_dir):
\n continue
\n # Wrap the meaty part in a try-except block.
\n try:
\n out_results[dataset][scene] = {}
\n img_fnames = [f'{src}/test/{x}' for x in data_dict[dataset][scene]]
\n print (f\"Got {len(img_fnames)} images\")
\n n = 16
\n new_fnames = [img_fnames[i:i+n] for i in range(0, len(img_fnames), n)]
\n if len(new_fnames[-1]) < n:
\n new_fnames[-2:] = [new_fnames[-2] + new_fnames[-1]]
\n M_world = []
\n for index,item in enumerate(new_fnames):

\n
            img_tensor,camera = (item)\n            model = ()\n            checkpoint = torch()\n            model(checkpoint)\n            model()\n            heatmap_pred,desc_pred, = (img_tensor)\n            temp = (heatmap_pred, desc_pred,,camera)\n            M_world =M_world + temp\n         index,name  (data_dict):\n            key1 = f\n            out_results = {}\n            out_results = M_world\n            out_results = M_world\n        (f)\n        (f)\n        (out_results, data_dict)\n        ()\n        gc()\n    except:\n        pass\n\n\n        `\n
\n

create_submission(out_results, data_dict)
\nThis is all my code except the model

", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "2287026": "Why did my code submit a 0.001 score and run in 3 minutes\nIt was written by my submission is there a problem?\n\n`import gc\n\n\n\ngc.collect()\ndatasets = []\n\n\nfor dataset in data_dict:\n datasets.append(dataset)\n\nfor dataset in datasets:\n print(dataset)\n if dataset not in out_results:\n out_results[dataset] = {}\n for scene in data_dict[dataset]:\n print(scene)\n # Fail gently if the notebook has not been submitted and the test data is not populated.\n # You may want to run this on the training data in that case?\n img_dir = f'{src}/test/{dataset}/{scene}/images'\n if not os.path.exists(img_dir):\n continue\n # Wrap the meaty part in a try-except block.\n try:\n out_results[dataset][scene] = {}\n img_fnames = [f'{src}/test/{x}' for x in data_dict[dataset][scene]]\n print (f\"Got {len(img_fnames)} images\")\n n = 16\n new_fnames = [img_fnames[i:i+n] for i in range(0, len(img_fnames), n)]\n if len(new_fnames[-1]) < n:\n new_fnames[-2:] = [new_fnames[-2] + new_fnames[-1]]\n M_world = []\n for index,item in enumerate(new_fnames):\n\n\n img_tensor,camera = collate_img(item)\n model = make_model()\n checkpoint = torch.load(\"/kaggle/input/params/params.pth\")\n model.load_state_dict(checkpoint)\n model.eval()\n heatmap_pred,desc_pred,img = model(img_tensor)\n temp = compute(heatmap_pred, desc_pred,img,camera)\n M_world =M_world + temp\n for index,name in enumerate(data_dict[dataset][scene]):\n key1 = f'{name}'\n out_results[dataset][scene][key1] = {}\n out_results[dataset][scene][key1][\"R\"] = M_world[index][0]\n out_results[dataset][scene][key1][\"t\"] = M_world[index][1]\n print(f'Registered: {dataset} / {scene} -> {len(out_results[dataset][scene])} images')\n print(f'Total: {dataset} / {scene} -> {len(data_dict[dataset][scene])} images')\n create_submission(out_results, data_dict)\n print(\"create_submission成功\")\n gc.collect()\n except:\n pass\n \n`\n`create_submission(out_results, data_dict)`", "2287036": "fyi: you may try placing all of your code in a codeblock in your question; it will make it easier for others to read (and the whitespace is important)", "2287074": "`def compute(heatmap_pred, desc_pred,img, camera, alpha=0.2, eps=1e-6):\n \n \n matches,match_list = feature_matching(heatmap_pred,desc_pred,img)\n M_world = cameraworld(matches,camera,match_list)\n \n \n return M_world`\n`import albumentations as A\nfrom PIL import Image\n\n\ntransform_4 = A.Compose([\n A.Resize(height=256, width=256, p=1.0)\n])\n\ndef collate_img(img_list):\n newimages = []\n \n newcamera = []\n for index,item in enumerate(img_list):\n \n image = cv2.imread(item,cv2.IMREAD_COLOR)\n \n camera_id = 2\n if item[0].find(\"dioscuri\") != -1:\n camera_id = 54\n elif item[0].find(\"kyiv-puppet-theater\") != -1:\n camera_id = 2\n \n \n img = torch.from_numpy(transform_4(image=image)['image'])\n newimages.append(img)\n \n \n \n newcamera.append(camera_id)\n images = torch.tensor(torch.stack(newimages),dtype=torch.float32)\n print(\"images:\",images.shape)\n images = images.permute(0,3,1,2)\n print(\"images:\",images.shape)\n \n \n return images,newcamera`\n`src = '/kaggle/input/image-matching-challenge-2023'`\n`data_dict = {}\nwith open(f'{src}/sample_submission.csv', 'r') as f:\n for i, l in enumerate(f):\n # Skip header.\n if l and i > 0:\n image, dataset, scene, _, _ = l.strip().split(',')\n if dataset not in data_dict:\n data_dict[dataset] = {}\n if scene not in data_dict[dataset]:\n data_dict[dataset][scene] = []\n data_dict[dataset][scene].append(image)`\n`for dataset in data_dict:\n for scene in data_dict[dataset]:\n print(f'{dataset} / {scene} -> {len(data_dict[dataset][scene])} images')`\n`out_results = {}\ntimings = {\"shortlisting\":[],\n \"feature_detection\": [],\n \"feature_matching\":[],\n \"RANSAC\": [],\n \"Reconstruction\": []}\ndef arr_to_str(a):\n return ';'.join([str(x) for x in a.reshape(-1)])\nimport gc\n\n\n\ngc.collect()\ndatasets = []\n\n\nfor dataset in data_dict:\n datasets.append(dataset)\n\nfor dataset in datasets:\n print(dataset)\n if dataset not in out_results:\n out_results[dataset] = {}\n for scene in data_dict[dataset]:\n print(scene)\n # Fail gently if the notebook has not been submitted and the test data is not populated.\n # You may want to run this on the training data in that case?\n img_dir = f'{src}/test/{dataset}/{scene}/images'\n if not os.path.exists(img_dir):\n continue\n # Wrap the meaty part in a try-except block.\n try:\n out_results[dataset][scene] = {}\n img_fnames = [f'{src}/test/{x}' for x in data_dict[dataset][scene]]\n print (f\"Got {len(img_fnames)} images\")\n n = 16\n new_fnames = [img_fnames[i:i+n] for i in range(0, len(img_fnames), n)]\n if len(new_fnames[-1]) < n:\n new_fnames[-2:] = [new_fnames[-2] + new_fnames[-1]]\n M_world = []\n for index,item in enumerate(new_fnames):\n\n\n img_tensor,camera = collate_img(item)\n model = make_model()\n checkpoint = torch.load(\"/kaggle/input/params/DCNV3_4_params.pth\")\n model.load_state_dict(checkpoint)\n model.eval()\n heatmap_pred,desc_pred,img = model(img_tensor)\n temp = compute(heatmap_pred, desc_pred,img,camera)\n M_world =M_world + temp\n for index,name in enumerate(data_dict[dataset][scene]):\n key1 = f'{name}'\n out_results[dataset][scene][key1] = {}\n out_results[dataset][scene][key1][\"R\"] = M_world[index][0]\n out_results[dataset][scene][key1][\"t\"] = M_world[index][1]\n print(f'Registered: {dataset} / {scene} -> {len(out_results[dataset][scene])} images')\n print(f'Total: {dataset} / {scene} -> {len(data_dict[dataset][scene])} images')\n create_submission(out_results, data_dict)\n print(\"create_submission成功\")\n gc.collect()\n except:\n pass\n \n\n `\n`create_submission(out_results, data_dict)`\nThis is all my code except the model" }, "source": "meta" }