search

ZQPei / deep_sort_pytorch

MOT using deepsort and yolov3 with pytorch
MIT License
2.76k stars 721 forks source link

deepsort模型ckpt.t7怎么重新训练？ #192

Open deeplearningers opened 3 years ago

deeplearningers commented 3 years ago

我要训练自己的数据，ckpt.t7需要重新训练吗？怎么训练啊？ i want to train my own datasets,does ckpt.t7 model need to retrain again? how to train it?

Sarouch commented 3 years ago

Hello, Same issue here, anyone has the answer please ? Thank you.

yuninn commented 3 years ago

I ran into the same problem. Do I need to retrain the weights? If I need to retrain, how do I label the data?

JAYCHOU2020 commented 2 years ago

I ran into the same problem. Do I need to retrain the weights? If I need to retrain, how do I label the data?

help i also want to know!!!!!!!!thanks

yuninn commented 2 years ago

You can refer to this example，https://blog.csdn.net/qq_35975447/article/details/114412978

------------------ 原始邮件 ------------------ 发件人: @.>; 发送时间: 2021年7月28日(星期三) 中午11:37 收件人: @.>; 抄送: @.>; @.>; 主题: Re: [ZQPei/deep_sort_pytorch] deepsort模型ckpt.t7怎么重新训练？ (#192)

I ran into the same problem. Do I need to retrain the weights? If I need to retrain, how do I label the data?

help i also want to know!!!!!!!!thanks

— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.

Sarouch commented 2 years ago

You can refer to this example，https://blog.csdn.net/qq_35975447/article/details/114412978 … ------------------ 原始邮件 ------------------ 发件人: @.>; 发送时间: 2021年7月28日(星期三) 中午11:37 收件人: @.>; 抄送: @.>; @.>; 主题: Re: [ZQPei/deep_sort_pytorch] deepsort模型ckpt.t7怎么重新训练？ (#192) I ran into the same problem. Do I need to retrain the weights? If I need to retrain, how do I label the data? help i also want to know!!!!!!!!thanks — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.

Hello @yuninn , unfortunatelly I can't access to the hole web site, may be because I live in France, could you please copy the tutorial to me and send it to gymnouem@gmail.com ? image

yuninn commented 2 years ago

My email was rejected by your mailbox

------------------ 原始邮件 ------------------ 发件人: @.>; 发送时间: 2021年8月6日(星期五) 凌晨3:02 收件人: @.>; 抄送: @.>; @.>; 主题: Re: [ZQPei/deep_sort_pytorch] deepsort模型ckpt.t7怎么重新训练？ (#192)

You can refer to this example，https://blog.csdn.net/qq_35975447/article/details/114412978 … ------------------ 原始邮件 ------------------ 发件人: @.>; 发送时间: 2021年7月28日(星期三) 中午11:37 收件人: @.>; 抄送: @.>; @.>; 主题: Re: [ZQPei/deep_sort_pytorch] deepsort模型ckpt.t7怎么重新训练？ (#192) I ran into the same problem. Do I need to retrain the weights? If I need to retrain, how do I label the data? help i also want to know!!!!!!!!thanks — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.

Hello @yuninn , unfortunatelly I can't access to the hole web site, may be because I live in France, could you please copy the tutorial to me and send it to @.*** ?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android.

JAYCHOU2020 commented 2 years ago

My email was rejected by your mailbox … ------------------ 原始邮件 ------------------ 发件人: @.>; 发送时间: 2021年8月6日(星期五) 凌晨3:02 收件人: @.>; 抄送: @.>; @.>; 主题: Re: [ZQPei/deep_sort_pytorch] deepsort模型ckpt.t7怎么重新训练？ (#192) You can refer to this example，https://blog.csdn.net/qq_35975447/article/details/114412978 … ------------------ 原始邮件 ------------------ 发件人: @.>; 发送时间: 2021年7月28日(星期三) 中午11:37 收件人: @.>; 抄送: @.>; @.>; 主题: Re: [ZQPei/deep_sort_pytorch] deepsort模型ckpt.t7怎么重新训练？ (#192) I ran into the same problem. Do I need to retrain the weights? If I need to retrain, how do I label the data? help i also want to know!!!!!!!!thanks — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe. Hello @yuninn , unfortunatelly I can't access to the hole web site, may be because I live in France, could you please copy the tutorial to me and send it to @.*** ? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. when i train deepsort maybe my target is too small, how to fix it please?thanks 微信图片_20210806161644

JAYCHOU2020 commented 2 years ago

My email was rejected by your mailbox … ------------------ 原始邮件 ------------------ 发件人: @.>; 发送时间: 2021年8月6日(星期五) 凌晨3:02 收件人: @.>; 抄送: @.>; @.>; 主题: Re: [ZQPei/deep_sort_pytorch] deepsort模型ckpt.t7怎么重新训练？ (#192) You can refer to this example，https://blog.csdn.net/qq_35975447/article/details/114412978 … ------------------ 原始邮件 ------------------ 发件人: @.>; 发送时间: 2021年7月28日(星期三) 中午11:37 收件人: @.>; 抄送: @.>; @.>; 主题: Re: [ZQPei/deep_sort_pytorch] deepsort模型ckpt.t7怎么重新训练？ (#192) I ran into the same problem. Do I need to retrain the weights? If I need to retrain, how do I label the data? help i also want to know!!!!!!!!thanks — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe. Hello @yuninn , unfortunatelly I can't access to the hole web site, may be because I live in France, could you please copy the tutorial to me and send it to @.*** ? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android.

when i modify torchvision.transforms.RandomCrop((128,64)) to torchvision.transforms.RandomCrop((30,20)) i meet this mistake 微信图片_20210806162042

yuninn commented 2 years ago

transform_train = torchvision.transforms.Compose([     torchvision.transforms.Resize((128, 64)),     torchvision.transforms.RandomCrop((128, 64), padding=4),     torchvision.transforms.RandomHorizontalFlip(),     torchvision.transforms.ToTensor(),     torchvision.transforms.Normalize(         [0.485, 0.456, 0.406], [0.229, 0.224, 0.225]) ])

train.py  resize your data to(128,64)

------------------ 原始邮件 ------------------ 发件人: @.>; 发送时间: 2021年8月6日(星期五) 下午4:21 收件人: @.>; 抄送: @.>; @.>; 主题: Re: [ZQPei/deep_sort_pytorch] deepsort模型ckpt.t7怎么重新训练？ (#192)

My email was rejected by your mailbox … ------------------ 原始邮件 ------------------ 发件人: @.>; 发送时间: 2021年8月6日(星期五) 凌晨3:02 收件人: @.>; 抄送: @.>; @.>; 主题: Re: [ZQPei/deep_sort_pytorch] deepsort模型ckpt.t7怎么重新训练？ (#192) You can refer to this example，https://blog.csdn.net/qq_35975447/article/details/114412978 … ------------------ 原始邮件 ------------------ 发件人: @.>; 发送时间: 2021年7月28日(星期三) 中午11:37 收件人: @.>; 抄送: @.>; @.>; 主题: Re: [ZQPei/deep_sort_pytorch] deepsort模型ckpt.t7怎么重新训练？ (#192) I ran into the same problem. Do I need to retrain the weights? If I need to retrain, how do I label the data? help i also want to know!!!!!!!!thanks — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe. Hello @yuninn , unfortunatelly I can't access to the hole web site, may be because I live in France, could you please copy the tutorial to me and send it to @.*** ? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android.

when i modify torchvision.transforms.RandomCrop((128,64)) to torchvision.transforms.RandomCrop((30,20)) i meet this mistake

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android.

Sarouch commented 2 years ago

@yuninn Try this please : sarah.chouchene@univ-lorraine.fr Thank you

JAYCHOU2020 commented 2 years ago

transform_train = torchvision.transforms.Compose([     torchvision.transforms.Resize((128, 64)),     torchvision.transforms.RandomCrop((128, 64), padding=4),     torchvision.transforms.RandomHorizontalFlip(),     torchvision.transforms.ToTensor(),     torchvision.transforms.Normalize(         [0.485, 0.456, 0.406], [0.229, 0.224, 0.225]) ]) train.py  resize your data to(128,64) … ------------------ 原始邮件 ------------------ 发件人: @.>; 发送时间: 2021年8月6日(星期五) 下午4:21 收件人: @.>; 抄送: @.>; @.>; 主题: Re: [ZQPei/deep_sort_pytorch] deepsort模型ckpt.t7怎么重新训练？ (#192) My email was rejected by your mailbox … ------------------ 原始邮件 ------------------ 发件人: @.>; 发送时间: 2021年8月6日(星期五) 凌晨3:02 收件人: @.>; 抄送: @.>; @.>; 主题: Re: [ZQPei/deep_sort_pytorch] deepsort模型ckpt.t7怎么重新训练？ (#192) You can refer to this example，https://blog.csdn.net/qq_35975447/article/details/114412978 … ------------------ 原始邮件 ------------------ 发件人: @.>; 发送时间: 2021年7月28日(星期三) 中午11:37 收件人: @.>; 抄送: @.>; @.>; 主题: Re: [ZQPei/deep_sort_pytorch] deepsort模型ckpt.t7怎么重新训练？ (#192) I ran into the same problem. Do I need to retrain the weights? If I need to retrain, how do I label the data? help i also want to know!!!!!!!!thanks — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe. Hello @yuninn , unfortunatelly I can't access to the hole web site, may be because I live in France, could you please copy the tutorial to me and send it to @.*** ? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. when i modify torchvision.transforms.RandomCrop((128,64)) to torchvision.transforms.RandomCrop((30,20)) i meet this mistake — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android.

thank u very much !!! I'll try it