Fastai2 and new course now released

fastai
63

Actually I don’t have a GCP setup guide yet. Help welcome on that front! :slight_smile:

4 Likes
64

Which ones are we missing? :slight_smile:

2 Likes
65

Super excited! Thanks for all the hard work! @jeremy
I assume the new course will be still on https://course.fast.ai/ as http://dev.fast.ai/ is now 404.

2 Likes
66

Course will be there, dev.fast.ai pointed to the docs which are now at docs.fast.ai. Old course will be at course19.fast.ai and old docs at http://fastai1.fast.ai/

2 Likes
67

Unable to access the video lessons. It gives error 404. Anyone else having this problem?

68

Yes, I assume it will be resolved soon. You can view them on youtube in the meantime.

69

Ok thanks

70

Please note that fastai 2 requires torch-1.6.0 and torchvision-0.7.0. The cuda drivers on the platform image are 10.1 and too old for torch-1.6.0. There are no pytorch images in the deeplearning-platform-release family with 10.2 or 11 cuda drivers. However the 10.2 drivers can be updated per @micstan

Here are the steps I followed to setup the GCP image with the new release and the book:
Follow the old GCP setup guide here: http://course19.fast.ai/start_gcp.html

Login to terminal: gcloud compute ssh --zone “us-central1-b” “jupyter@fastai-4” – -L 8080:localhost:8080

wget http://developer.download.nvidia.com/compute/cuda/10.2/Prod/local_installers/cuda_10.2.89_440.33.01_linux.run
sudo sh cuda_10.2.89_440.33.01_linux.run

cd tutorials
mv fastai fastai.old
git clone --recurse-submodules https://github.com/fastai/fastai
pip install -e “fastai[dev]”
git clone --recurse-submodules https://github.com/fastai/fastcore
cd fastcore
pip install -e “.[dev]”
cd …
git clone https://github.com/fastai/fastbook.git
cd fastbook
pip install -r requirements.txt
cd …

Launch local browser: http://localhost:8080/tree/tutorials

Verify the notebooks run

Run notebook from: http://localhost:8080/tree/tutorials/fastai/dev_nbs/course
Run notebook from: http://localhost:8080/notebooks/tutorials/fastbook/

Perhaps others will have a more elegant solution, but for something quick to get started I haven’t run into any issues running fastai v2 and notebooks on GCP this way.

Cheers and many thanks for all the stellar work on the course, book and API! Mark

6 Likes
71

Hi, I know maybe I am asking too much. Will fastai release the part 2 of 2020?

72

It has not been recorded or lectured yet, nor even announced so we’ll have to wait :slight_smile:

In the meantime part 2 is the latter half of the book for the most part

3 Likes
74

Note: If you have issues installing the new library, you may need to do pip install fastai --upgrade to snag the latest version

2 Likes
75

Hi Jeremy. Thanks for the great new course and book. But starting today I cannot access videos for part 2 of the previous courses.
I can look for the sources in github https://github.com/fastai/course-v3/tree/master/nbs/dl2,
but it’s quite unfortunate I cannot access the videos anymore.

76

@yuri_k that should have been a reply to you, sorry. ^playlist for part 2^

1 Like
77

The new FastAI course is fantastic, many thanks.

I’ll be doing something of a speedrun through it for review, wanted to post the first issue I found:

I’ll be compiling all the little issues I find and providing them to the team where appropriate. Would you recommend I start a dedicated topic for editorial review?

As I become more familiar with the infrastructure behind the course, I will provide patch submissions myself.

1 Like
78

Hi Mark - were you able to get the GPU working in your setup? I think the GPU isn’t being used because the first training takes about 10 minutes (and nvidia-smi seems to indicate that the card is idle, but it’s possible I’m misusing that tool).

79

Thank you to the wonderful fastai team. Your courses, book, and all the extra functions have been really helpful in my work. I am looking forward to where this library can go.

1 Like
80

My apologies, but fastai 2 requires torch-1.6.0 and torchvision-0.7.0. The cuda drivers on the platform image are 10.1 and too old for torch-1.6.0. There are no pytorch images in the deeplearning-platform-release family with 10.2 or 11 cuda drivers. A driver update is possible. Thanks @micstan

wget http://developer.download.nvidia.com/compute/cuda/10.2/Prod/local_installers/cuda_10.2.89_440.33.01_linux.run
sudo sh cuda_10.2.89_440.33.01_linux.run

You can then verify the gpu is all good with:

python -c ‘import torch; print(torch.__version__);print(torch.version.cuda);print(torch.cuda.is_available()); print(torch.cuda.current_device())’

2 Likes
82

hey, this one worked just fine for debian 9 on gc:

wget http://developer.download.nvidia.com/compute/cuda/10.2/Prod/local_installers/cuda_10.2.89_440.33.01_linux.run
sudo sh cuda_10.2.89_440.33.01_linux.run
83

Indeed it does! :grinning: I’ll update the instructions. Thanks, Mark

84

No idea where to put this, I can’t make new threads yet since I’m a new user, but I have a question.

How do I get the stripped notebooks? I’m running the latest version in Paperspace, and I see the following comment in the book;

Full and Stripped Notebooks: There are two folders containing different versions of the notebooks. The full folder contains the exact notebooks used to create the book you’re reading now, with all the prose and outputs. The stripped version has the same headings and code cells, but all outputs and prose have been removed.

Unfortunately, not only is there no course-v4/nbs/full or course-v4/nbs/stripped folders in the repo - which makes sense because the course is changing all the time, maybe they don’t exist anymore - but it seems that even inside of the fastbook/clean folder, the outputs are still visible.

I may be in the minority here, but when the notebooks show the output, it really take the magic of seeing everything “going” away. Devil’s advocate here, obviously doesn’t apply to people very focused on learning the content, but why run the model if I already see 16 cat images with probabilities below them.

Essentially, I don’t want to be “spoiled”, I want to actually wait the 20 minutes to train some big model and get rewarded the hard way. The only thing better than that I would say is a function definition with some comments;

# Cell
def train_new_model():
    batch_size = 12

    # 1. Make sure you load in the learner!
    # 2. Make sure you've got the data ready from "some_global"!
    # 3. Begin training and print the outputs - hint: docs.fastai.com/logging/print_stuff

I want hardmode, and since there’s nothing like the above right now, where can I find the stripped notebooks?

4 Likes