Fix load validation

CHANGELOG.md

@@ -5,6 +5,11 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## 2.7.1
+### Fixed
+- In 2.7.0 a bug was introduced, failing the loading of models. Here this is fixed. 
+- Added a check that a model file exists, for clearer error handling. 
+
 ## 2.7.0
 ### Added
 - The training pair sampling for both ionmodes is now balanced over the different ionmode pairs.

ms2deepscore/SettingsMS2Deepscore.py

@@ -93,8 +93,6 @@ class SettingsMS2Deepscore:
             If True the model will do separate pair sampling for training for each ionmode.
             This gives better balance over the ionmodes. Initial results showed a decrease in pos-pos prediction
             accuracy. Which you can find in the notebook model_benchmarking/Compare balanced cross ion moe sampling.ipynb
-        additional_metadata:
-            Additional metadata that should be used in training the model. e.g. precursor_mz
         dropout_rate:
             The dropout rate that should be used during training
         learning_rate:
@@ -272,8 +270,6 @@ def __init__(self, validate_settings=True, **settings):
             np.random.seed(self.random_seed)
 
         if self.spectrum_file_path is not None:
-            if not os.path.isfile(self.spectrum_file_path):
-                raise ValueError("The spectrum file specified is not an existing file")
             root_dir = os.path.dirname(self.spectrum_file_path)
             spectrum_file_name = os.path.basename(self.spectrum_file_path)
 
@@ -304,7 +300,10 @@ def validate_settings(self):
         validate_bin_order(self.same_prob_bins)
         if self.balanced_sampling_across_ionmodes and self.ionisation_mode != "both":
             raise ValueError("Balanced sampling across ionmodes only works if you train on both ionmodes")
-
+        if self.spectrum_file_path is not None:
+            if not os.path.isfile(self.spectrum_file_path):
+                raise ValueError("The spectrum file specified is not an existing file")
+
     def create_model_directory_name(self):
         """Creates a directory name using metadata, it will contain the metadata, the binned spectra and final model"""
         binning_file_label = ""

ms2deepscore/__version__.py

@@ -1 +1 @@
-__version__ = '2.7.0'
+__version__ = '2.7.1'

ms2deepscore/models/load_model.py

@@ -1,3 +1,4 @@
+import os
 from typing import Union, Dict, Optional, Any, Callable
 import json
 from pathlib import Path
@@ -130,6 +131,8 @@ def _load_model_generic(
       3) Instantiate model via model_factory(settings), then load state_dict.
       4) If safe path fails and allow_legacy=True, attempt unsafe legacy path.
     """
+    if not os.path.isfile(filename):
+        raise ValueError(f"The file given does not exist: {filename}")
     # --- preferred safe path
     try:
         ckpt = _load_ckpt_safe(filename)

tests/test_load_model.py

@@ -10,7 +10,7 @@
 
 def _dummy_siamese_model():
     """Helper function to create a dummy SiameseSpectralModel."""
-    settings = SettingsMS2Deepscore(base_dims=(100, 100), embedding_dim=10)
+    settings = SettingsMS2Deepscore(base_dims=(100, 100), embedding_dim=10, spectrum_file_path="./nonexisting_path.mgf", validate_settings=False)
     model = SiameseSpectralModel(settings=settings)
     return model