feat: add model.json to package data

.gitignore

@@ -182,3 +182,6 @@ lightning_logs/
 
 # Test resources
 tests/instanovo_test_resources*
+
+# Notebook output
+notebooks/predictions_kbs.csv

MANIFEST.in

@@ -1 +1,2 @@
 recursive-include instanovo/configs *.yaml
+include instanovo/models.json

instanovo/__init__.py

@@ -1,3 +1,3 @@
 from __future__ import annotations
 
-__version__ = "1.0.1"
+__version__ = "1.0.2"

instanovo/transformer/model.py

@@ -16,6 +16,7 @@
 from omegaconf import DictConfig
 from torch import nn
 from torch import Tensor
+from importlib import resources
 
 from instanovo.constants import MAX_SEQUENCE_LENGTH
 from instanovo.transformer.layers import ConvPeakEmbedding
@@ -33,7 +34,6 @@
 from instanovo.types import SpectrumMask
 from instanovo.utils import ResidueSet
 
-MODELS_PATH = Path(__file__).parent.parent / "models.json"
 MODEL_TYPE = "transformer"
 
 
@@ -126,7 +126,7 @@ def _get_causal_mask(seq_len: int, return_float: bool = False) -> PeptideMask:
     def get_pretrained() -> list[str]:
         """Get a list of pretrained model ids."""
         # Load the models.json file
-        with open(MODELS_PATH, "r") as f:
+        with resources.open_text("instanovo", "models.json") as f:
             models_config = json.load(f)
 
         if MODEL_TYPE not in models_config:
@@ -177,7 +177,7 @@ def from_pretrained(cls, model_id: str) -> Tuple["InstaNovo", "DictConfig"]:
                 raise FileNotFoundError(f"No file found at path: {model_id}")
 
         # Load the models.json file
-        with open(MODELS_PATH, "r") as f:
+        with resources.open_text("instanovo", "models.json") as f:
             models_config = json.load(f)
 
         # Find the model in the config

notebooks/getting_started_with_instanovo.ipynb

@@ -85,7 +85,7 @@
    },
    "outputs": [],
    "source": [
-    "!pip install instanovo"
+    "!pip install instanovo \"torch<2.5\""
    ]
   },
   {
@@ -110,7 +110,9 @@
     "import numpy as np\n",
     "import pandas as pd\n",
     "import urllib.request\n",
-    "import zipfile"
+    "import zipfile\n",
+    "\n",
+    "os.environ[\"PROTOCOL_BUFFERS_PYTHON_IMPLEMENTATION\"] = \"python\""
    ]
   },
   {
@@ -698,7 +700,8 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "!pip uninstall -y instanovo && pip install instanovo==0.1.7"
+    "!pip uninstall -y instanovo && pip install \"instanovo==0.1.7\"\n",
+    "!pip install 'protobuf<3.20' tensorboard --force-reinstall"
    ]
   },
   {
@@ -711,7 +714,8 @@
     "from datasets import load_dataset\n",
     "\n",
     "# Only evaluate on a subset of the data for demo\n",
-    "dataset = load_dataset(\"InstaDeepAI/ms_ninespecies_benchmark\", split=\"test[:1%]\")"
+    "dataset = load_dataset(\"InstaDeepAI/ms_ninespecies_benchmark\", split=\"test[:1%]\")\n",
+    "dataset"
    ]
   },
   {
@@ -922,7 +926,7 @@
    "provenance": []
   },
   "kernelspec": {
-   "display_name": "Python 3 (ipykernel)",
+   "display_name": "instanovo",
    "language": "python",
    "name": "python3"
   },
@@ -936,7 +940,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.11.9"
+   "version": "3.10.14"
   }
  },
  "nbformat": 4,