{ "id": 146494, "title": "[Keras, PyTorch]: All Pre-Trained Weights", "url": "/competitions/prostate-cancer-grade-assessment/discussion/146494", "author_name": "Akash Shingha", "post_date": "2020-04-27T11:21:29.021000", "votes": 4, "comment_count": 5, "views": 0, "content": "

[Pytorch]\n- https://www.kaggle.com/abhishek/pretrained-model-weights-pytorch\n- https://www.kaggle.com/hmendonca/efficientnet-pytorch

\n\n

[Keas]\n- https://www.kaggle.com/gmhost/keras-pretrain-model-weights\n- https://www.kaggle.com/ipythonx/keras-pretrained-imagenet-weights\n- https://www.kaggle.com/ipythonx/efficientnet-keras-noisystudent-weights-b0b7

", "messages": [ { "id": 823060, "postDate": "2020-04-27T11:21:29.020Z", "content": "

[Pytorch]\n- https://www.kaggle.com/abhishek/pretrained-model-weights-pytorch\n- https://www.kaggle.com/hmendonca/efficientnet-pytorch

\n\n

[Keas]\n- https://www.kaggle.com/gmhost/keras-pretrain-model-weights\n- https://www.kaggle.com/ipythonx/keras-pretrained-imagenet-weights\n- https://www.kaggle.com/ipythonx/efficientnet-keras-noisystudent-weights-b0b7

", "rawMarkdown": "[Pytorch]\n- https://www.kaggle.com/abhishek/pretrained-model-weights-pytorch\n- https://www.kaggle.com/hmendonca/efficientnet-pytorch\n\n[Keas]\n- https://www.kaggle.com/gmhost/keras-pretrain-model-weights\n- https://www.kaggle.com/ipythonx/keras-pretrained-imagenet-weights\n- https://www.kaggle.com/ipythonx/efficientnet-keras-noisystudent-weights-b0b7", "votes": 4 }, { "id": 826217, "postDate": "2020-04-29T14:08:38.630Z", "content": "

To load the weights and models with a proper string name for this weights data sets, one can follow this short notebook.\nhttps://www.kaggle.com/ipythonx/keras-all-models-weights-loader

\n\n

This mainly for keras(tf) and available weights are:\n- VGG [16, 19]\n- ResNet [18, 34, 50, 101, 152]\n- ResNeXt [50, 101]\n- SE-ResNet [18, 34, 50, 101, 152]\n- SE-ResNeXt [50, 101]\n- SE-Net [154]\n- DenseNet [121, 169, 201]\n- Inception ResNet V2\n- Inception V3\n- Xception\n- NASNet [large, mobile]\n- MobileNet\n- MobileNet v2\nBasically its a collection of weights from this repo. It's good fit for completion where we can't use internet connection.

", "rawMarkdown": "To load the weights and models with a proper string name for [this weights data sets](https://www.kaggle.com/ipythonx/keras-pretrained-imagenet-weights), one can follow this short notebook.\nhttps://www.kaggle.com/ipythonx/keras-all-models-weights-loader\n\nThis mainly for keras(tf) and available weights are:\n- VGG [16, 19]\n- ResNet [18, 34, 50, 101, 152]\n- ResNeXt [50, 101]\n- SE-ResNet [18, 34, 50, 101, 152]\n- SE-ResNeXt [50, 101]\n- SE-Net [154]\n- DenseNet [121, 169, 201]\n- Inception ResNet V2\n- Inception V3\n- Xception\n- NASNet [large, mobile]\n- MobileNet\n- MobileNet v2\nBasically its a collection of weights from [this repo](https://github.com/qubvel/classification_models). It's good fit for completion where we can't use internet connection.", "votes": 1 }, { "id": 864699, "postDate": "2020-05-28T06:27:19.207Z", "content": "

Which dataset is the model trained on? ImageNet?

", "rawMarkdown": "Which dataset is the model trained on? ImageNet?" }, { "id": 823738, "postDate": "2020-04-27T21:05:17.350Z", "content": "

Thank You!, those can be a big help considering how downloading the models with an external library which is hosting those models on github and aws, can be a pain for the people in india, china and neighboring countries, yes, those servers provide like nothing speed here until hosted in the country itself...

", "rawMarkdown": "Thank You!, those can be a big help considering how downloading the models with an external library which is hosting those models on github and aws, can be a pain for the people in india, china and neighboring countries, yes, those servers provide like nothing speed here until hosted in the country itself..." }, { "id": 823695, "postDate": "2020-04-27T20:13:33.073Z", "content": "

There aren't pretrained models to solve NLP problems?

", "rawMarkdown": "There aren't pretrained models to solve NLP problems?", "replies": [ { "id": 824064, "postDate": "2020-04-28T06:17:59.257Z", "content": "

I'm sure there are couple of these in the Data tab 😉

", "rawMarkdown": "I'm sure there are couple of these in the Data tab 😉 ", "votes": 1 } ] } ], "comments": [ { "id": 826217, "author_name": "Innat", "author_url": "", "post_date": "2020-04-29T14:08:38.630000", "content": "

To load the weights and models with a proper string name for this weights data sets, one can follow this short notebook.\nhttps://www.kaggle.com/ipythonx/keras-all-models-weights-loader

\n\n

This mainly for keras(tf) and available weights are:\n- VGG [16, 19]\n- ResNet [18, 34, 50, 101, 152]\n- ResNeXt [50, 101]\n- SE-ResNet [18, 34, 50, 101, 152]\n- SE-ResNeXt [50, 101]\n- SE-Net [154]\n- DenseNet [121, 169, 201]\n- Inception ResNet V2\n- Inception V3\n- Xception\n- NASNet [large, mobile]\n- MobileNet\n- MobileNet v2\nBasically its a collection of weights from this repo. It's good fit for completion where we can't use internet connection.

", "votes": 1, "replies": [] }, { "id": 864699, "author_name": "Karan Pathak", "author_url": "", "post_date": "2020-05-28T06:27:19.207000", "content": "

Which dataset is the model trained on? ImageNet?

", "votes": 0, "replies": [] }, { "id": 823738, "author_name": "Harshit Sheoran", "author_url": "", "post_date": "2020-04-27T21:05:17.350000", "content": "

Thank You!, those can be a big help considering how downloading the models with an external library which is hosting those models on github and aws, can be a pain for the people in india, china and neighboring countries, yes, those servers provide like nothing speed here until hosted in the country itself...

", "votes": 0, "replies": [] }, { "id": 823695, "author_name": "Hiram Coria 🧬", "author_url": "", "post_date": "2020-04-27T20:13:33.073000", "content": "

There aren't pretrained models to solve NLP problems?

", "votes": 0, "replies": [ { "id": 824064, "author_name": "mtszkw", "author_url": "", "post_date": "2020-04-28T06:17:59.257000", "content": "

I'm sure there are couple of these in the Data tab 😉

", "votes": 1, "replies": [] } ] } ], "raw_markdown_by_id": { "823060": "[Pytorch]\n- https://www.kaggle.com/abhishek/pretrained-model-weights-pytorch\n- https://www.kaggle.com/hmendonca/efficientnet-pytorch\n\n[Keas]\n- https://www.kaggle.com/gmhost/keras-pretrain-model-weights\n- https://www.kaggle.com/ipythonx/keras-pretrained-imagenet-weights\n- https://www.kaggle.com/ipythonx/efficientnet-keras-noisystudent-weights-b0b7", "826217": "To load the weights and models with a proper string name for [this weights data sets](https://www.kaggle.com/ipythonx/keras-pretrained-imagenet-weights), one can follow this short notebook.\nhttps://www.kaggle.com/ipythonx/keras-all-models-weights-loader\n\nThis mainly for keras(tf) and available weights are:\n- VGG [16, 19]\n- ResNet [18, 34, 50, 101, 152]\n- ResNeXt [50, 101]\n- SE-ResNet [18, 34, 50, 101, 152]\n- SE-ResNeXt [50, 101]\n- SE-Net [154]\n- DenseNet [121, 169, 201]\n- Inception ResNet V2\n- Inception V3\n- Xception\n- NASNet [large, mobile]\n- MobileNet\n- MobileNet v2\nBasically its a collection of weights from [this repo](https://github.com/qubvel/classification_models). It's good fit for completion where we can't use internet connection.", "864699": "Which dataset is the model trained on? ImageNet?", "823738": "Thank You!, those can be a big help considering how downloading the models with an external library which is hosting those models on github and aws, can be a pain for the people in india, china and neighboring countries, yes, those servers provide like nothing speed here until hosted in the country itself...", "823695": "There aren't pretrained models to solve NLP problems?" } }