How can I train this model on my own dataset?

[dreamlychina]

Thank you for sharing this amazing work. I want to train this model on my own datasets, but I don't have any pretrained DDPM model. Could you please tell me how to do? Should i train DDPM model on my own datasets to get the pretrained DDPM model and then use the pretrained DDPM model to train the segmentation model?

[ellkrauze]

Hello, @dreamlychina! Did you solve this issue?

[Supzekun]

Hi！I have the same question, could you give me some advice?

[dreamlychina]

Hello, @dreamlychina! Did you solve this issue?

yes

[dreamlychina]

怎么联系你？？

[Supzekun]

怎么联系你？？

我在自己的数据集上训练，loss一直不下降，您知道这是什么原因吗？

[dreamlychina]

微信多少，我加你吧

…

---- Replied Message ---- | From | ***@***.***> | | Date | 04/03/2024 10:56 | | To | yandex-research/ddpm-segmentation ***@***.***> | | Cc | dreamlychina ***@***.***>, Mention ***@***.***> | | Subject | Re: [yandex-research/ddpm-segmentation] How can I train this model on my own dataset? (Issue #29) | 怎么联系你？？ image.png (view on web) 我在自己的数据集上训练，loss一直不下降，您知道这是什么原因吗？ — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you were mentioned.Message ID: ***@***.***>

[YonghuiTAN22]

Hello, @dreamlychina! Did you solve this issue?

yes

请问你是怎么解决的

[encoding05]

Hello, @dreamlychina! Did you solve this issue?

yes

Hi, could you please give me some advice on this question? Thanks a million!

[dreamlychina]

解决了

…

---- Replied Message ---- | From | ***@***.***> | | Date | 09/26/2024 14:39 | | To | yandex-research/ddpm-segmentation ***@***.***> | | Cc | dreamlychina ***@***.***>, Mention ***@***.***> | | Subject | Re: [yandex-research/ddpm-segmentation] How can I train this model on my own dataset? (Issue #29) | Hello, @dreamlychina! Did you solve this issue? yes Hi, could you please give me some advice on this question? Thanks a million! — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you were mentioned.Message ID: ***@***.***>

[hong-hd]

@dreamlychina 您好能请教下怎么解决的吗

[hong-hd]

解决了
…
---- Replied Message ---- | From | @.> | | Date | 09/26/2024 14:39 | | To | yandex-research/ddpm-segmentation @.> | | Cc | dreamlychina @.>, Mention @.> | | Subject | Re: [yandex-research/ddpm-segmentation] How can I train this model on my own dataset? (Issue #29) | Hello, @dreamlychina! Did you solve this issue? yes Hi, could you please give me some advice on this question? Thanks a million! — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you were mentioned.Message ID: @.***>

您好能请教下怎么解决的吗

[zxk72]

微信多少，我加你吧 微信多少，我加你吧 
…
---- Replied Message ---- | From | @.> | | Date | 04/03/2024 10:56 | | To | yandex-research/ddpm-segmentation @.> | | Cc | dreamlychina @.>, Mention @.> | | Subject | Re: [yandex-research/ddpm-segmentation] How can I train this model on my own dataset? (Issue #29) | 怎么联系你？？ image.png (view on web) 我在自己的数据集上训练，loss一直不下降，您知道这是什么原因吗？ — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you were mentioned.Message ID: @.***>
z13653725948，感谢

[zxk72]

微信多少，我加你吧 微信多少，我加你吧 
…
---- 已回复消息 ---- |来自| @。> | |日期 | 2024/04/03 10:56 | |至 | yandex-research/ddpm-segmentation _@** ._ > | |抄送 |梦幻中国_@_ . _>，提及_@。** > | |主题 |回复：[yandex-research/ddpm-segmentation] 如何在我自己的数据集上训练这个模型？ （第29期）| image.png (在网络上查看) 我在自己的数据集上训练，损失一直不上升，您知道这是什么原因吗？ — 直接回复此电子邮件，在 GitHub 上查看，或取消订阅。您收到此消息是因为您被提及。消息 ID：@ .***>
z13653725948，感谢

wxid_0qwqingx0q4h22，万分感谢

你会吗？我不会，而且你这个号我加不上

[hong-hd]

微信多少，我加你吧 微信多少，我加你吧 
…
---- 已回复消息 ---- |来自| @。> | |日期 | 2024/04/03 10:56 | |至 | yandex-research/ddpm-segmentation _@** ._ > | |抄送 |梦幻中国_@_ . _>，提及_@。** > | |主题 |回复：[yandex-research/ddpm-segmentation] 如何在我自己的数据集上训练这个模型？ （第29期）| image.png (在网络上查看) 我在自己的数据集上训练，损失一直不上升，您知道这是什么原因吗？ — 直接回复此电子邮件，在 GitHub 上查看，或取消订阅。您收到此消息是因为您被提及。消息 ID：@ .***>
z13653725948，感谢

wxid_0qwqingx0q4h22，万分感谢

你会吗？我不会，而且你这个号我加不上

可以向您请假下如何训练自己的数据集吗，这是我的邮箱[email protected]

[hong-hd]

微信多少，我加你吧 微信多少，我加你吧 
…
---- 已回复消息 ---- |来自| @。> | |日期 | 2024/04/03 10:56 | |至 | yandex-research/ddpm-segmentation _@** ._ > | |抄送 |梦幻中国_@_ . _>，提及_@。** > | |主题 |回复：[yandex-research/ddpm-segmentation] 如何在我自己的数据集上训练这个模型？ （第29期）| image.png (在网络上查看) 我在自己的数据集上训练，损失一直不上升，您知道这是什么原因吗？ — 直接回复此电子邮件，在 GitHub 上查看，或取消订阅。您收到此消息是因为您被提及。消息 ID：@ .***>
z13653725948，感谢

wxid_0qwqingx0q4h22，万分感谢

你会吗？我不会，而且你这个号我加不上

您好，可以向您请假下如何训练自己的数据集吗，这是我的邮箱[email protected]

[zxk72]

解决了
…
---- Replied Message ---- | From | @.> | | Date | 09/26/2024 14:39 | | To | yandex-research/ddpm-segmentation @.> | | Cc | dreamlychina @.>, Mention @.> | | Subject | Re: [yandex-research/ddpm-segmentation] How can I train this model on my own dataset? (Issue #29) | Hello, @dreamlychina! Did you solve this issue? yes Hi, could you please give me some advice on this question? Thanks a million! — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you were mentioned.Message ID: @.***>

@dreamlychina 你好，[email protected]这是我的邮箱，可以请教一下你是如何解决的吗？