Daniel Bourke comments

Results 107 comments of


                                            Daniel Bourke

Notebook 05: load_weights results in "Incompatible tensor with shape (1280, 10)..."

Update: looks like TensorFlow 2.9.0 is still the most stable here, see: https://github.com/mrdbourke/tensorflow-deep-learning/issues/553 TL;DR tried `tf-nightly(2.14.0-dev20230520)` and it still broke.

Notebook 05: load_weights results in "Incompatible tensor with shape (1280, 10)..."

Hi all, After much troubleshooting, I've found the best fix for `tf.keras.applications.EfficientNetB0` problems is to simply upgrade to [`tf.keras.applications.efficientnet_v2.EfficientNetV2B0`](https://www.tensorflow.org/api_docs/python/tf/keras/applications/efficientnet_v2/EfficientNetV2B0). You can see a full write-up of the fix here: https://github.com/mrdbourke/tensorflow-deep-learning/discussions/575

Notebook 05: load_weights results in "Incompatible tensor with shape (1280, 10)..."

@talha-0 Great catch! Thank you for the update!

Notebook 05: load_weights results in "Incompatible tensor with shape (1280, 10)..."

> > Hi all, > > After much troubleshooting, I've found the best fix for `tf.keras.applications.EfficientNetB0` problems is to simply upgrade to [`tf.keras.applications.efficientnet_v2.EfficientNetV2B0`](https://www.tensorflow.org/api_docs/python/tf/keras/applications/efficientnet_v2/EfficientNetV2B0). > > You can see a full...

removed a deprecated method and corrected a typo in notebook 00

Hey @mazarrazi , Thank you for this! I appreciate the typo fix and am happy to merge that. However, as for the `torch.Tensor.type`, I'm not sure that's deprecated? (I may...

Put section number for Section 5 "Equation 2: Multihead Self-Attention (MSA block)"

Thank you @ideaguy3d !

Issue in video 118: RuntimeError: Inplace update to inference tensor outside InferenceMode is not allowed.

Hi @HrushikeshChaphalkar and @jyli559 , I get this quite often too. It happens most when I forget to make sure everything in the `with torch.inference_mode():` context manager has the right...

ModuleNotFoundError: No module named 'going_modular'

Hi @NaiyerRafique , This is strange. I just re-ran [Notebook 08 - PyTorch Paper Replicating](https://github.com/mrdbourke/pytorch-deep-learning/blob/main/08_pytorch_paper_replicating.ipynb) in Google Colab and found it worked. I was able to download the `going_modular` scripts...

deffrent y_pred calculating in leasson 02(pytorch cllassification) and extras in pytorch learning.why?!

Hi @jenabesaman , It will depend on what kind of data you're using for `torch.softmax` or `torch.sigmoid`. Do you have a more full example of your code so I can...

Add a note to README to start locally

Maybe fix this too? ```python File ~/code/pytorch-deep-learning/env/lib/python3.11/site-packages/torchvision/transforms/functional.py:5 3 import sys 4 import math ----> 5 from PIL import Image, ImageOps, ImageEnhance, PILLOW_VERSION 6 try: 7 import accimage ImportError: cannot import...