RuntimeError: Fail to call fit with error: 413, reason: Request Entity Too Large and text #102
Description
When trying to fit a heavily-textual dataset (melbourne-airbnb) through the API, I run into a RuntimeError. This isn't happening with a lot of other similar datasets. I didn't dive into the specific, but this dataset has A LOT of strings, and maybe some of them are very long.
import openml
from tabpfn_client import TabPFNRegressor
# naive example to replicate bug
dataset = openml.datasets.get_dataset(46665, download_data=True, download_features_meta_data=True)
x, y, _, _ = dataset.get_data(target=dataset.default_target_attribute)
x_train = x[:10000]
y_train = y[:10000]
x_test = x[10000:]
y_test = y[10000:]
model = TabPFNRegressor()
model.fit(x_train, y_train)
model.predict(x_test)
The error:
Could not download file from https://data.openml.org/datasets/0004/46665/dataset_46665.pq: Object at 'https://data.openml.org/datasets/0004/46665/dataset_46665.pq' does not exist.
Failed to download parquet, fallback on ARFF.
Welcome Back! Found existing access token, reusing it for authentication.
ERROR:tabpfn_client.client:Fail to call fit, response status: 413
Traceback (most recent call last):
File "/Users/alanarazi/Desktop/Personal/tabular/.venv/lib/python3.11/site-packages/IPython/core/interactiveshell.py", line 3579, in run_code
exec(code_obj, self.user_global_ns, self.user_ns)
File "<ipython-input-12-340f9c3d548b>", line 9, in <module>
model.fit(x_train, y_train)
File "/Users/alanarazi/Desktop/Personal/tabular/.venv/lib/python3.11/site-packages/tabpfn_client/estimator.py", line 311, in fit
self.last_train_set_uid = InferenceClient.fit(X, y, config=estimator_param)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/alanarazi/Desktop/Personal/tabular/.venv/lib/python3.11/site-packages/tabpfn_client/service_wrapper.py", line 225, in fit
return ServiceClient.fit(X, y, config=config)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/alanarazi/Desktop/Personal/tabular/.venv/lib/python3.11/site-packages/tabpfn_client/client.py", line 237, in fit
cls._validate_response(response, "fit")
File "/Users/alanarazi/Desktop/Personal/tabular/.venv/lib/python3.11/site-packages/tabpfn_client/client.py", line 477, in _validate_response
raise RuntimeError(
RuntimeError: Fail to call fit with error: 413, reason: Request Entity Too Large and text:
<html><head>
<meta http-equiv="content-type" content="text/html;charset=utf-8">
<title>413 Request Entity Too Large</title>
</head>
<body text=#000000 bgcolor=#ffffff>
<h1>Error: Request Entity Too Large</h1>
<h2>Your client issued a request that was too large.
</h2>
<h2></h2>
</body></html>