I save model from pickle.
Load model :
with open(AICROWD_ASSETS_DIR + ‘/models.pickle’, ‘rb’) as f:
models = pickle.load(f)
Submit error “No such file or directory: ‘assets/models.pickle’”.
How to specify the path correctly?
1 Like
Hello @BanKhv
TLDR; Can you submit your solution again without removing the headings provided in the starter notebook?
For the recent submissions that failed with docker build error, it is due to removing some of the blocks/headings from the notebook. We rely on the headings in the notebook to detect which part of the notebook should be run.
In this case, the submissions were failing due to missing Define preprocessing code 💻 heading after Define preprocessing code 💻 .
The following should be the structure of your notebook
# Setup AIcrowd Utilities
...
# Install packages
...
# Define preprocessing code
...
# Training phase
...
# Prediction phase
...
# Submit to AIcrowd
During the package installation phase, we run the notebook cells between # Install packages and # Define preprocessing code. During the generate predictions phase, we ignore the notebook cells between (# Install packages, # Define preprocessing code) and (# Training phase, # Prediction phase).
Please let me know if anything is unclear.
3 Likes
Thanks. It worked.
Can you reset the count of erroneous submissions?
Hi,
Just to be sure I correctly understand: we don’t have access to the test dataset (the one used for Leaderboard, that contains 1,473 rows), even its features, right? Our model will be loaded and applied on it?
Thank you.
Hello @demarsylvain
Yes, you are right. The participants will not have access to the test data. The leaderboard right now shows the scores on 60% of the test data. There will be a new leaderboard released at the end of the challenge with scores on 100% of the test data.
1 Like