Hello. Has anyone successfully installed and used the L5Kit on their local machine. I was able to install it with PyTorch 1.5 but getting errors while importing. Any suggestions.
", "messages": [ { "id": "1001135", "postDate": "09/07/2020 04:59:06", "content": "Hello. Has anyone successfully installed and used the L5Kit on their local machine. I was able to install it with PyTorch 1.5 but getting errors while importing. Any suggestions.
", "rawMarkdown": "Hello. Has anyone successfully installed and used the L5Kit on their local machine. I was able to install it with PyTorch 1.5 but getting errors while importing. Any suggestions.", "votes": null }, { "id": "1001138", "postDate": "09/07/2020 05:03:19", "content": "Do u have a screen shot of ur error? I can install l5kit and use that to run experiments.
", "rawMarkdown": "Do u have a screen shot of ur error? I can install l5kit and use that to run experiments.", "votes": null }, { "id": "1001150", "postDate": "09/07/2020 05:19:19", "content": "Ok sure. This is the import code and the error.
\nCODE:
\n`from typing import Dict
from tempfile import gettempdir
\nimport matplotlib.pyplot as plt
\nimport numpy as np
\nimport torch
\nfrom torch import nn, optim
\nfrom torch.utils.data import DataLoader
\nfrom torchvision.models.resnet import resnet18
\nfrom tqdm import tqdm
from l5kit.configs import load_config_data
\nfrom l5kit.data import LocalDataManager, ChunkedDataset
\nfrom l5kit.dataset import AgentDataset, EgoDataset
\nfrom l5kit.rasterization import build_rasterizer
\nfrom l5kit.evaluation import write_pred_csv, compute_metrics_csv, read_gt_csv, create_chopped_dataset
\nfrom l5kit.evaluation.chop_dataset import MIN_FUTURE_STEPS
\nfrom l5kit.evaluation.metrics import neg_multi_log_likelihood, time_displace
\nfrom l5kit.geometry import transform_points
\nfrom l5kit.visualization import PREDICTED_POINTS_COLOR, TARGET_POINTS_COLOR, draw_trajectory
\nfrom prettytable import PrettyTable
\nfrom pathlib import Path
import os`
\nERROR
\n`ValueError Traceback (most recent call last)
\n in
\n 12 from l5kit.configs import load_config_data
\n 13 from l5kit.data import LocalDataManager, ChunkedDataset
\n---> 14 from l5kit.dataset import AgentDataset, EgoDataset
\n 15 from l5kit.rasterization import build_rasterizer
\n 16 from l5kit.evaluation import write_pred_csv, compute_metrics_csv, read_gt_csv, create_chopped_dataset
d:\\installed_softwares\\miniconda\\envs\\t151\\lib\\site-packages\\l5kit\\dataset__init__.py in
\n----> 1 from .agent import AgentDataset
\n 2 from .ego import EgoDataset
\n 3 from .select_agents import select_agents
\n 4
\n 5 all = [\"EgoDataset\", \"AgentDataset\", \"select_agents\"]
d:\\installed_softwares\\miniconda\\envs\\t151\\lib\\site-packages\\l5kit\\dataset\\agent.py in
\n 10 from ..rasterization import Rasterizer
\n 11 from .ego import EgoDataset
\n---> 12 from .select_agents import TH_DISTANCE_AV, TH_EXTENT_RATIO, TH_YAW_DEGREE, select_agents
\n 13
\n 14 # WARNING: changing these values impact the number of instances selected for both train and inference!
d:\\installed_softwares\\miniconda\\envs\\t151\\lib\\site-packages\\l5kit\\dataset\\select_agents.py in
\n 19 from l5kit.data.filter import _get_label_filter # TODO expose this without digging
\n 20
\n---> 21 multiprocessing.set_start_method(\"fork\", force=True) # this fix loop in python 3.8 on MacOS
\n 22 os.environ[\"BLOSC_NOLOCK\"] = \"1\" # this is required for multiprocessing
\n 23
d:\\installed_softwares\\miniconda\\envs\\t151\\lib\\multiprocessing\\context.py in set_start_method(self, method, force)
\n 244 self._actual_context = None
\n 245 return
\n--> 246 self._actual_context = self.get_context(method)
\n 247
\n 248 def get_start_method(self, allow_none=False):
d:\\installed_softwares\\miniconda\\envs\\t151\\lib\\multiprocessing\\context.py in get_context(self, method)
\n 236 return self._actual_context
\n 237 else:
\n--> 238 return super().get_context(method)
\n 239
\n 240 def set_start_method(self, method, force=False):
d:\\installed_softwares\\miniconda\\envs\\t151\\lib\\multiprocessing\\context.py in get_context(self, method)
\n 190 ctx = _concrete_contexts[method]
\n 191 except KeyError:
\n--> 192 raise ValueError('cannot find context for %r' % method) from None
\n 193 ctx._check_available()
\n 194 return ctx
ValueError: cannot find context for 'fork'`
", "rawMarkdown": "Ok sure. This is the import code and the error.\nCODE:\n`from typing import Dict\n\nfrom tempfile import gettempdir\nimport matplotlib.pyplot as plt\nimport numpy as np\nimport torch\nfrom torch import nn, optim\nfrom torch.utils.data import DataLoader\nfrom torchvision.models.resnet import resnet18\nfrom tqdm import tqdm\n\nfrom l5kit.configs import load_config_data\nfrom l5kit.data import LocalDataManager, ChunkedDataset\nfrom l5kit.dataset import AgentDataset, EgoDataset\nfrom l5kit.rasterization import build_rasterizer\nfrom l5kit.evaluation import write_pred_csv, compute_metrics_csv, read_gt_csv, create_chopped_dataset\nfrom l5kit.evaluation.chop_dataset import MIN_FUTURE_STEPS\nfrom l5kit.evaluation.metrics import neg_multi_log_likelihood, time_displace\nfrom l5kit.geometry import transform_points\nfrom l5kit.visualization import PREDICTED_POINTS_COLOR, TARGET_POINTS_COLOR, draw_trajectory\nfrom prettytable import PrettyTable\nfrom pathlib import Path\n\nimport os`\n\nERROR\n`ValueError Traceback (most recent call last)\n@sovitrath I assume you are running widows. There is a discussion multiprocessing thread where people have discussed the same issue. Partial solution is mentioned there. You can follow that. What you need to do is, comment out the 'fork' multiprocessing set for MAC users in select_agent.py and you need to follow multi processing guideline from python/pytorch which talks about how to handle multiprocessing issue in windows. And another thing from my side is don't use Jupyter. This is the only solution I know. If you find any other alternative. Come back and comment.
Thanks for the reply. Will try out what you said. And the thing about not using Jupyter. Is it because the error is Jupyter specific. Because I don't usually use Jypyter. I always execute python scripts. But since there were so many problems regarding L5Kit in this competition, therefore I started with some minimal Jupyter code. And since I ran into error I did not try python scripts. Have been running on Kaggle notebooks. Will try with python scripts on my local system. Thanks, again.
", "rawMarkdown": "Thanks for the reply. Will try out what you said. And the thing about not using Jupyter. Is it because the error is Jupyter specific. Because I don't usually use Jypyter. I always execute python scripts. But since there were so many problems regarding L5Kit in this competition, therefore I started with some minimal Jupyter code. And since I ran into error I did not try python scripts. Have been running on Kaggle notebooks. Will try with python scripts on my local system. Thanks, again.", "votes": null }, { "id": "1001875", "postDate": "09/07/2020 16:31:39", "content": "hey, Just one question. How are you storing 22 GB of lift data locally when you are running your scripts?
", "rawMarkdown": "hey, Just one question. How are you storing 22 GB of lift data locally when you are running your scripts?", "votes": null }, { "id": "1002238", "postDate": "09/08/2020 00:40:26", "content": "It will be better if you can be a bit clearer in your question. Are you asking, how do I download them, or how do I load the dataset to memory?
", "rawMarkdown": "It will be better if you can be a bit clearer in your question. Are you asking, how do I download them, or how do I load the dataset to memory?", "votes": null }, { "id": "1002245", "postDate": "09/08/2020 00:53:13", "content": "sorry.I am asking , how do you download them? are you using GCS ?
", "rawMarkdown": "sorry.I am asking , how do you download them? are you using GCS ?", "votes": null }, { "id": "1002314", "postDate": "09/08/2020 03:40:35", "content": "For downloading the dataset, you can simply click the download button the Data section of the competition page or even use the API to download the dataset.
Do u have a screen shot of ur error? I can install l5kit and use that to run experiments.
", "votes": null, "replies": [ { "id": 1001150, "author_name": "sovitrath", "author_url": "", "post_date": "09/07/2020 05:19:19", "content": "Ok sure. This is the import code and the error.
\nCODE:
\n`from typing import Dict
from tempfile import gettempdir
\nimport matplotlib.pyplot as plt
\nimport numpy as np
\nimport torch
\nfrom torch import nn, optim
\nfrom torch.utils.data import DataLoader
\nfrom torchvision.models.resnet import resnet18
\nfrom tqdm import tqdm
from l5kit.configs import load_config_data
\nfrom l5kit.data import LocalDataManager, ChunkedDataset
\nfrom l5kit.dataset import AgentDataset, EgoDataset
\nfrom l5kit.rasterization import build_rasterizer
\nfrom l5kit.evaluation import write_pred_csv, compute_metrics_csv, read_gt_csv, create_chopped_dataset
\nfrom l5kit.evaluation.chop_dataset import MIN_FUTURE_STEPS
\nfrom l5kit.evaluation.metrics import neg_multi_log_likelihood, time_displace
\nfrom l5kit.geometry import transform_points
\nfrom l5kit.visualization import PREDICTED_POINTS_COLOR, TARGET_POINTS_COLOR, draw_trajectory
\nfrom prettytable import PrettyTable
\nfrom pathlib import Path
import os`
\nERROR
\n`ValueError Traceback (most recent call last)
\n in
\n 12 from l5kit.configs import load_config_data
\n 13 from l5kit.data import LocalDataManager, ChunkedDataset
\n---> 14 from l5kit.dataset import AgentDataset, EgoDataset
\n 15 from l5kit.rasterization import build_rasterizer
\n 16 from l5kit.evaluation import write_pred_csv, compute_metrics_csv, read_gt_csv, create_chopped_dataset
d:\\installed_softwares\\miniconda\\envs\\t151\\lib\\site-packages\\l5kit\\dataset__init__.py in
\n----> 1 from .agent import AgentDataset
\n 2 from .ego import EgoDataset
\n 3 from .select_agents import select_agents
\n 4
\n 5 all = [\"EgoDataset\", \"AgentDataset\", \"select_agents\"]
d:\\installed_softwares\\miniconda\\envs\\t151\\lib\\site-packages\\l5kit\\dataset\\agent.py in
\n 10 from ..rasterization import Rasterizer
\n 11 from .ego import EgoDataset
\n---> 12 from .select_agents import TH_DISTANCE_AV, TH_EXTENT_RATIO, TH_YAW_DEGREE, select_agents
\n 13
\n 14 # WARNING: changing these values impact the number of instances selected for both train and inference!
d:\\installed_softwares\\miniconda\\envs\\t151\\lib\\site-packages\\l5kit\\dataset\\select_agents.py in
\n 19 from l5kit.data.filter import _get_label_filter # TODO expose this without digging
\n 20
\n---> 21 multiprocessing.set_start_method(\"fork\", force=True) # this fix loop in python 3.8 on MacOS
\n 22 os.environ[\"BLOSC_NOLOCK\"] = \"1\" # this is required for multiprocessing
\n 23
d:\\installed_softwares\\miniconda\\envs\\t151\\lib\\multiprocessing\\context.py in set_start_method(self, method, force)
\n 244 self._actual_context = None
\n 245 return
\n--> 246 self._actual_context = self.get_context(method)
\n 247
\n 248 def get_start_method(self, allow_none=False):
d:\\installed_softwares\\miniconda\\envs\\t151\\lib\\multiprocessing\\context.py in get_context(self, method)
\n 236 return self._actual_context
\n 237 else:
\n--> 238 return super().get_context(method)
\n 239
\n 240 def set_start_method(self, method, force=False):
d:\\installed_softwares\\miniconda\\envs\\t151\\lib\\multiprocessing\\context.py in get_context(self, method)
\n 190 ctx = _concrete_contexts[method]
\n 191 except KeyError:
\n--> 192 raise ValueError('cannot find context for %r' % method) from None
\n 193 ctx._check_available()
\n 194 return ctx
ValueError: cannot find context for 'fork'`
", "votes": null, "replies": [] } ] }, { "id": 1001227, "author_name": "deepakrajpurushothaman", "author_url": "", "post_date": "09/07/2020 06:30:22", "content": "@sovitrath I assume you are running widows. There is a discussion multiprocessing thread where people have discussed the same issue. Partial solution is mentioned there. You can follow that. What you need to do is, comment out the 'fork' multiprocessing set for MAC users in select_agent.py and you need to follow multi processing guideline from python/pytorch which talks about how to handle multiprocessing issue in windows. And another thing from my side is don't use Jupyter. This is the only solution I know. If you find any other alternative. Come back and comment.
Thanks for the reply. Will try out what you said. And the thing about not using Jupyter. Is it because the error is Jupyter specific. Because I don't usually use Jypyter. I always execute python scripts. But since there were so many problems regarding L5Kit in this competition, therefore I started with some minimal Jupyter code. And since I ran into error I did not try python scripts. Have been running on Kaggle notebooks. Will try with python scripts on my local system. Thanks, again.
", "votes": null, "replies": [] }, { "id": 1001875, "author_name": "sujaydkhandekar", "author_url": "", "post_date": "09/07/2020 16:31:39", "content": "hey, Just one question. How are you storing 22 GB of lift data locally when you are running your scripts?
", "votes": null, "replies": [] }, { "id": 1002238, "author_name": "sovitrath", "author_url": "", "post_date": "09/08/2020 00:40:26", "content": "It will be better if you can be a bit clearer in your question. Are you asking, how do I download them, or how do I load the dataset to memory?
", "votes": null, "replies": [] }, { "id": 1002245, "author_name": "sujaydkhandekar", "author_url": "", "post_date": "09/08/2020 00:53:13", "content": "sorry.I am asking , how do you download them? are you using GCS ?
", "votes": null, "replies": [] }, { "id": 1002314, "author_name": "sovitrath", "author_url": "", "post_date": "09/08/2020 03:40:35", "content": "For downloading the dataset, you can simply click the download button the Data section of the competition page or even use the API to download the dataset.