| Model | \nAUC | \n
|---|---|
SEResNeXt50 | \n0.8668 | \n
SEResNet152 | \n0.857 | \n
SEResNet101 | \n0.854 | \n
SEResNet50 | \n0.859 | \n
| Model | \nAUC | \n
|---|---|
SEResNeXt50 | \n0.8668 | \n
SEResNet152 | \n0.857 | \n
SEResNet101 | \n0.854 | \n
SEResNet50 | \n0.859 | \n
PS: I'll also be sharing pretrained weights for reproducibility.
", "rawMarkdown": "PS: I'll also be sharing pretrained weights for reproducibility.", "votes": 1 }, { "id": 1406822, "postDate": "2021-08-01T08:19:36.010Z", "content": "Do you have any plans to share your pipeline ?
", "rawMarkdown": "Do you have any plans to share your pipeline ?", "votes": 1, "replies": [ { "id": 1407712, "postDate": "2021-08-02T05:26:56.867Z", "content": "My pipeline was similar to @miklgr500's Tensorflow TPU Kernel. I also used Mixup, but with a custom LR Scheduler and a \"Dense\" Output Classification Head.
", "rawMarkdown": "My pipeline was similar to @miklgr500's Tensorflow TPU Kernel. I also used Mixup, but with a custom LR Scheduler and a \"Dense\" Output Classification Head. " }, { "id": 1407719, "postDate": "2021-08-02T05:30:22.207Z", "content": "Thanks for the insight
", "rawMarkdown": "Thanks for the insight ", "votes": 1 } ] }, { "id": 1404987, "postDate": "2021-07-30T12:09:19.460Z", "content": "If you used TensorFlow can you share from where you got these models ?
\nEdit - It looks like he probably used this repo https://github.com/qubvel/classification_models
Yes, I used classification_models
Hey All,
\nThank you all for taking part in our competition. The participation has been overwhelmingly positive. We are currently conducting a survey to gauge the demographic and outreach achieved. Kindly spare 2min and fill in this survey https://forms.gle/QP9L16niPexozyhu5.
\nThank you all,
\nRegards,
\nChris
Just Uploaded the Dataset containing the pre-trained weights here
", "rawMarkdown": "Just Uploaded the Dataset containing the pre-trained weights [here](https://www.kaggle.com/sauravmaheshkar/g2netfinetunedclassifiermodels)" }, { "id": 1405348, "postDate": "2021-07-30T18:30:28.193Z", "rawMarkdown": "", "votes": 1, "isDeleted": true, "replies": [ { "id": 1406692, "postDate": "2021-08-01T05:59:45.447Z", "content": "I used 256 images generated using CQT
", "rawMarkdown": "I used 256 images generated using CQT", "votes": 1 } ] }, { "id": 1405325, "postDate": "2021-07-30T17:55:52.463Z", "content": "Thanks for sharing @sauravmaheshkar !
", "rawMarkdown": "Thanks for sharing @sauravmaheshkar !", "votes": 1 } ], "comments": [ { "id": 1407714, "author_name": "Saurav Maheshkar ☕️", "author_url": "", "post_date": "2021-08-02T05:27:22.720000", "content": "PS: I'll also be sharing pretrained weights for reproducibility.
", "votes": 1, "replies": [] }, { "id": 1406822, "author_name": "Mithil Salunkhe", "author_url": "", "post_date": "2021-08-01T08:19:36.010000", "content": "Do you have any plans to share your pipeline ?
", "votes": 1, "replies": [ { "id": 1407712, "author_name": "Saurav Maheshkar ☕️", "author_url": "", "post_date": "2021-08-02T05:26:56.867000", "content": "My pipeline was similar to @miklgr500's Tensorflow TPU Kernel. I also used Mixup, but with a custom LR Scheduler and a \"Dense\" Output Classification Head.
", "votes": 0, "replies": [] }, { "id": 1407719, "author_name": "Mithil Salunkhe", "author_url": "", "post_date": "2021-08-02T05:30:22.207000", "content": "Thanks for the insight
", "votes": 1, "replies": [] } ] }, { "id": 1404987, "author_name": "Mithil Salunkhe", "author_url": "", "post_date": "2021-07-30T12:09:19.460000", "content": "If you used TensorFlow can you share from where you got these models ?
\nEdit - It looks like he probably used this repo https://github.com/qubvel/classification_models
Yes, I used classification_models
Hey All,
\nThank you all for taking part in our competition. The participation has been overwhelmingly positive. We are currently conducting a survey to gauge the demographic and outreach achieved. Kindly spare 2min and fill in this survey https://forms.gle/QP9L16niPexozyhu5.
\nThank you all,
\nRegards,
\nChris
Just Uploaded the Dataset containing the pre-trained weights here
", "votes": 0, "replies": [] }, { "id": 1405348, "author_name": "", "author_url": "", "post_date": "2021-07-30T18:30:28.193000", "content": "", "votes": 1, "replies": [ { "id": 1406692, "author_name": "Saurav Maheshkar ☕️", "author_url": "", "post_date": "2021-08-01T05:59:45.447000", "content": "I used 256 images generated using CQT
", "votes": 1, "replies": [] } ] }, { "id": 1405325, "author_name": "Old Monk", "author_url": "", "post_date": "2021-07-30T17:55:52.463000", "content": "Thanks for sharing @sauravmaheshkar !
", "votes": 1, "replies": [] } ], "raw_markdown_by_id": { "1404665": "- 4 fold CV\n- 20 epochs\n\n---\n\n| **Model** | **AUC** |\n| --- | --- |\n| **`SEResNeXt50`** | 0.8668 |\n| **`SEResNet152`**| 0.857 |\n| **`SEResNet101`** | 0.854 |\n| **`SEResNet50`** | 0.859 |", "1407714": "PS: I'll also be sharing pretrained weights for reproducibility.", "1406822": "Do you have any plans to share your pipeline ?", "1404987": "If you used TensorFlow can you share from where you got these models ?\nEdit - It looks like he probably used this repo https://github.com/qubvel/classification_models", "1561212": "Hey All,\n\nThank you all for taking part in our competition. The participation has been overwhelmingly positive. We are currently conducting a survey to gauge the demographic and outreach achieved. Kindly spare 2min and fill in this survey https://forms.gle/QP9L16niPexozyhu5.\n\nThank you all,\n\nRegards,\nChris", "1474340": "Just Uploaded the Dataset containing the pre-trained weights [here](https://www.kaggle.com/sauravmaheshkar/g2netfinetunedclassifiermodels)", "1405348": "", "1405325": "Thanks for sharing @sauravmaheshkar !" } }