Ai-Ya-Ya

This error occurs when trying to load any model. `predictor.load` will work fine, but any other methods called to predictor, such as `predictor.predict()` will not.

Here is a more verbose stack trace. ``` --------------------------------------------------------------------------- EOFError Traceback (most recent call last) Cell In[12], line 1 ----> 1 predictor.predict(train_data, known_covariates=test_data[['income', 'weekend', 'holiday']], model='WeightedEnsemble') File /fs/cml-projects/teamgahsp/miniforge3/envs/new_env/lib/python3.11/site-packages/autogluon/timeseries/predictor.py:845, in TimeSeriesPredictor.predict(self,...

I added a line `print(fin)` right before `object = pickle.load(fin)`. It produced the following output: ``` ``` The error seems to be in the last line, which for some reason...

Doing a few more runs, it's clear that every other path has `./` included except for the path to `cached_predictions.pkl`.

Deleting `cached_predictions.pkl` fixed the error. For completeness, here is the training log. ``` Beginning AutoGluon training... AutoGluon will save models to './models/moco_total_crimes_year/' =================== System Info =================== AutoGluon Version: 1.1.0 Python...

Can confirm, still having this issue despite proper MF properties (EBE 0.10.1+1.21, MF 5.19.0+mc1.21). F3+T reloads correct textures though

It does seem to work with `dynamic_stacking=True`. Thanks. So, since you added the milestone, is a fix in progress for 1.4?

Nevermind. Halloy was hiding in the very far corner of my screen. Still, I do not believe it's expected that Halloy open in such a place.

I'm accessing it via `my-ip:port listed in config.toml`. I don't believe I'm using nginx, and I don't have it setup; I was only following the online documentation. When following the...