diff --git a/.github/workflows/ci.yaml b/.github/workflows/ci.yaml
new file mode 100644
index 0000000..658ce2a
--- /dev/null
+++ b/.github/workflows/ci.yaml
@@ -0,0 +1,71 @@
+name: CI
+
+on:
+  pull_request:
+    branches:
+      - main
+
+jobs:
+  lint:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v2
+
+      - name: Set up Python
+        uses: actions/setup-python@v2
+        with:
+          python-version: "3.10"
+
+      - name: Install project dependencies
+        run: pip install -r requirements-ci.txt
+
+      - name: Run Ruff lint
+        run: ruff check .
+
+  format:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v2
+
+      - name: Set up Python
+        uses: actions/setup-python@v2
+        with:
+          python-version: "3.10"
+
+      - name: Install project dependencies
+        run: pip install -r requirements-ci.txt
+
+      - name: Run Ruff format
+        run: ruff format .
+
+      - name: Check for uncommitted changes (formatting)
+        run: git diff --exit-code
+
+  test:
+    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        python_version:
+          - "3.9"
+          - "3.10"
+          - "3.11"
+          - "3.12"
+
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v2
+
+      - name: Set up Python
+        uses: actions/setup-python@v2
+        with:
+          python-version: ${{ matrix.python_version }}
+
+      - name: Install project dependencies
+        run: pip install -r requirements-ci.txt
+
+      - name: Run tests
+        run: pytest
+
diff --git a/README.md b/README.md
index 33e48b6..c8b9cd3 100644
--- a/README.md
+++ b/README.md
@@ -1,7 +1,7 @@
 
 # rerankers
 
-![Python Versions](https://img.shields.io/badge/Python-3.8_3.9_3.10_3.11-blue)
+![Python Versions](https://img.shields.io/badge/Python-3.9_3.10_3.11_3.12-blue)
 [![Downloads](https://static.pepy.tech/badge/rerankers/month)](https://pepy.tech/project/rerankers)
 [![Twitter Follow](https://img.shields.io/twitter/follow/bclavie?style=social)](https://twitter.com/bclavie)
 
diff --git a/examples/langchain_integration.ipynb b/examples/langchain_integration.ipynb
index 248f794..77d186c 100644
--- a/examples/langchain_integration.ipynb
+++ b/examples/langchain_integration.ipynb
@@ -262,7 +262,6 @@
    ],
    "source": [
     "from langchain.retrievers import ContextualCompressionRetriever\n",
-    "from langchain_openai import OpenAI\n",
     "\n",
     "text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)\n",
     "texts = text_splitter.split_documents(documents)\n",
@@ -274,7 +273,7 @@
     "\n",
     "\n",
     "compressed_docs = compression_retriever.get_relevant_documents(\n",
-    "     \"What did the president say about the minimum wage?\"\n",
+    "    \"What did the president say about the minimum wage?\"\n",
     ")\n",
     "\n",
     "pretty_print_docs(compressed_docs)"
diff --git a/examples/overview.ipynb b/examples/overview.ipynb
index 3e98b95..b20b35d 100644
--- a/examples/overview.ipynb
+++ b/examples/overview.ipynb
@@ -39,6 +39,7 @@
     "\n",
     "# Let's do this manually, so we don't need to install pydotenv\n",
     "import os\n",
+    "\n",
     "with open(\".env\", \"r\") as file:\n",
     "    for line in file:\n",
     "        if line.strip() and not line.startswith(\"#\"):\n",
@@ -62,7 +63,10 @@
    "outputs": [],
    "source": [
     "query = \"Gone with the wind is an absolute masterpiece\"\n",
-    "docs = [\"Gone with the wind is a masterclass in bad storytelling.\", \"Gone with the wind is an all-time classic\"]"
+    "docs = [\n",
+    "    \"Gone with the wind is a masterclass in bad storytelling.\",\n",
+    "    \"Gone with the wind is an all-time classic\",\n",
+    "]"
    ]
   },
   {
@@ -120,7 +124,7 @@
     }
    ],
    "source": [
-    "ranker = Reranker('cross-encoder')\n"
+    "ranker = Reranker(\"cross-encoder\")"
    ]
   },
   {
@@ -147,7 +151,7 @@
     }
    ],
    "source": [
-    "ranker = Reranker('cross-encoder', verbose=0)"
+    "ranker = Reranker(\"cross-encoder\", verbose=0)"
    ]
   },
   {
@@ -264,7 +268,11 @@
     }
    ],
    "source": [
-    "results = ranker.rank(query=query, docs=docs, doc_ids=[\"The Not-So Similar Document\", \"The Similar Document\"])\n",
+    "results = ranker.rank(\n",
+    "    query=query,\n",
+    "    docs=docs,\n",
+    "    doc_ids=[\"The Not-So Similar Document\", \"The Similar Document\"],\n",
+    ")\n",
     "results.top_k(1)"
    ]
   },
@@ -317,7 +325,7 @@
     }
    ],
    "source": [
-    "ranker = Reranker('cross-encoder/ms-marco-MiniLM-L-6-v2', model_type=\"cross-encoder\")"
+    "ranker = Reranker(\"cross-encoder/ms-marco-MiniLM-L-6-v2\", model_type=\"cross-encoder\")"
    ]
   },
   {
@@ -378,7 +386,7 @@
     }
    ],
    "source": [
-    "ranker = Reranker('cross-encoder', lang='fr')"
+    "ranker = Reranker(\"cross-encoder\", lang=\"fr\")"
    ]
   },
   {
@@ -409,13 +417,14 @@
     }
    ],
    "source": [
-    "ranker = Reranker('cross-encoder/ms-marco-MiniLM-L-6-v2',\n",
+    "ranker = Reranker(\n",
+    "    \"cross-encoder/ms-marco-MiniLM-L-6-v2\",\n",
     "    model_type=\"cross-encoder\",\n",
-    "    verbose = 1, # How verbose the reranker will be. Defaults to 1, setting it to 0 will suppress most messages.\n",
-    "    dtype = None, # Which dtype the model should use. If None will figure out if your platform + model combo supports fp16 and use it if so, other fp32.\n",
-    "    device = None,  # Which device the model should use. If None will figure out what the most powerful supported platform available is (cuda > mps > cpu)\n",
-    "    batch_size = 16,  # The batch size the model will use. Defaults to 16\n",
-    "    )"
+    "    verbose=1,  # How verbose the reranker will be. Defaults to 1, setting it to 0 will suppress most messages.\n",
+    "    dtype=None,  # Which dtype the model should use. If None will figure out if your platform + model combo supports fp16 and use it if so, other fp32.\n",
+    "    device=None,  # Which device the model should use. If None will figure out what the most powerful supported platform available is (cuda > mps > cpu)\n",
+    "    batch_size=16,  # The batch size the model will use. Defaults to 16\n",
+    ")"
    ]
   },
   {
@@ -463,9 +472,9 @@
    ],
    "source": [
     "# Jina\n",
-    "ranker = Reranker(\"jina\", api_key=os.environ['JINA_API_KEY'])\n",
+    "ranker = Reranker(\"jina\", api_key=os.environ[\"JINA_API_KEY\"])\n",
     "results = ranker.rank(query=query, docs=docs)\n",
-    "results.top_k(1)\n"
+    "results.top_k(1)"
    ]
   },
   {
@@ -495,7 +504,7 @@
    ],
    "source": [
     "# Cohere\n",
-    "ranker = Reranker(\"cohere\", api_key = os.environ['COHERE_API_KEY'])\n",
+    "ranker = Reranker(\"cohere\", api_key=os.environ[\"COHERE_API_KEY\"])\n",
     "results = ranker.rank(query=query, docs=docs)\n",
     "results.top_k(1)"
    ]
@@ -538,11 +547,15 @@
    ],
    "source": [
     "# Cohere\n",
-    "ranker = Reranker(\"cohere\", lang=\"en\", api_key = os.environ['COHERE_API_KEY'])\n",
-    "ranker.rank(query=\"Tell me about lord of the rings\",\n",
-    "           docs=[\"Dune is an incredibly confusing masterpiece in worldbuilding...\",\n",
-    "                 \"The silmarillion is a prequel to the Lord of The Rings...\",\n",
-    "                 \"Green Lantern uses a powerful ring to rule over his planet...\"])"
+    "ranker = Reranker(\"cohere\", lang=\"en\", api_key=os.environ[\"COHERE_API_KEY\"])\n",
+    "ranker.rank(\n",
+    "    query=\"Tell me about lord of the rings\",\n",
+    "    docs=[\n",
+    "        \"Dune is an incredibly confusing masterpiece in worldbuilding...\",\n",
+    "        \"The silmarillion is a prequel to the Lord of The Rings...\",\n",
+    "        \"Green Lantern uses a powerful ring to rule over his planet...\",\n",
+    "    ],\n",
+    ")"
    ]
   },
   {
@@ -570,7 +583,11 @@
    "source": [
     "# wrap in a try/except, as we don't have a fine-tuned model to use for this example!\n",
     "try:\n",
-    "    ranker = Reranker(\"my-finetuned-model-name\", api_provider=\"cohere\", api_key = os.environ['COHERE_API_KEY'])\n",
+    "    ranker = Reranker(\n",
+    "        \"my-finetuned-model-name\",\n",
+    "        api_provider=\"cohere\",\n",
+    "        api_key=os.environ[\"COHERE_API_KEY\"],\n",
+    "    )\n",
     "except:\n",
     "    pass"
    ]
@@ -648,7 +665,7 @@
     }
    ],
    "source": [
-    "ranker = Reranker('t5')\n",
+    "ranker = Reranker(\"t5\")\n",
     "results = ranker.rank(query=query, docs=docs)\n",
     "results.top_k(2)"
    ]
@@ -675,7 +692,7 @@
     }
    ],
    "source": [
-    "ranker = Reranker(\"unicamp-dl/ptt5-base-pt-msmarco-10k-v2\", model_type='t5', verbose=0)"
+    "ranker = Reranker(\"unicamp-dl/ptt5-base-pt-msmarco-10k-v2\", model_type=\"t5\", verbose=0)"
    ]
   },
   {
@@ -712,16 +729,17 @@
     }
    ],
    "source": [
-    "ranker = Reranker('t5',\n",
+    "ranker = Reranker(\n",
+    "    \"t5\",\n",
     "    model_type=\"t5\",\n",
-    "    verbose = 1, # How verbose the reranker will be. Defaults to 1, setting it to 0 will suppress most messages.\n",
-    "    dtype = None, # Which dtype the model should use. If None will figure out if your platform + model combo supports fp16 and use it if so, other fp32.\n",
-    "    device = None,  # Which device the model should use. If None will figure out what the most powerful supported platform available is (cuda > mps > cpu)\n",
-    "    batch_size = 16,  # The batch size the model will use. Defaults to 16\n",
-    "    token_false = \"auto\", # The output token corresponding to non-relevance.\n",
-    "    token_true = \"auto\", # The output token corresponding to relevance.\n",
-    "    return_logits = False, # Whether to return a normalised score or the raw logit for `token_true`.\n",
-    "    )\n"
+    "    verbose=1,  # How verbose the reranker will be. Defaults to 1, setting it to 0 will suppress most messages.\n",
+    "    dtype=None,  # Which dtype the model should use. If None will figure out if your platform + model combo supports fp16 and use it if so, other fp32.\n",
+    "    device=None,  # Which device the model should use. If None will figure out what the most powerful supported platform available is (cuda > mps > cpu)\n",
+    "    batch_size=16,  # The batch size the model will use. Defaults to 16\n",
+    "    token_false=\"auto\",  # The output token corresponding to non-relevance.\n",
+    "    token_true=\"auto\",  # The output token corresponding to relevance.\n",
+    "    return_logits=False,  # Whether to return a normalised score or the raw logit for `token_true`.\n",
+    ")"
    ]
   },
   {
@@ -762,7 +780,8 @@
    ],
    "source": [
     "from rerankers import Reranker\n",
-    "ranker = Reranker(\"rankgpt\", api_key = os.environ['OPENAI_API_KEY'])\n"
+    "\n",
+    "ranker = Reranker(\"rankgpt\", api_key=os.environ[\"OPENAI_API_KEY\"])"
    ]
   },
   {
@@ -810,7 +829,8 @@
    ],
    "source": [
     "from rerankers import Reranker\n",
-    "ranker = Reranker(\"rankgpt3\", api_key = os.environ['OPENAI_API_KEY'])\n",
+    "\n",
+    "ranker = Reranker(\"rankgpt3\", api_key=os.environ[\"OPENAI_API_KEY\"])\n",
     "results = ranker.rank(query=query, docs=docs)\n",
     "results.top_k(1)"
    ]
@@ -840,16 +860,21 @@
    "source": [
     "# LiteLLM uses env variables\n",
     "import os\n",
-    "os.environ['AZURE_API_KEY'] = \"\"\n",
+    "\n",
+    "os.environ[\"AZURE_API_KEY\"] = \"\"\n",
     "os.environ[\"AZURE_API_BASE\"] = \"\"\n",
     "os.environ[\"AZURE_API_VERSION\"] = \"\"\n",
     "deployment_name = \"my-azure-gpt-deployment\"\n",
     "\n",
     "# Just like Cohere's finetuned rankers above -- we try/except this as we're not actually running an Azure OpenAI model in this example!\n",
     "try:\n",
-    "    ranker = Reranker(f\"azure/{deployment_name}\", model_type=\"rankgpt\", api_key = os.environ['AZURE_API_KEY'])\n",
+    "    ranker = Reranker(\n",
+    "        f\"azure/{deployment_name}\",\n",
+    "        model_type=\"rankgpt\",\n",
+    "        api_key=os.environ[\"AZURE_API_KEY\"],\n",
+    "    )\n",
     "except:\n",
-    "    pass\n"
+    "    pass"
    ]
   },
   {
@@ -893,7 +918,8 @@
    ],
    "source": [
     "from rerankers import Reranker\n",
-    "ranker = Reranker('rankllm', api_key=os.environ['OPENAI_API_KEY'])\n",
+    "\n",
+    "ranker = Reranker(\"rankllm\", api_key=os.environ[\"OPENAI_API_KEY\"])\n",
     "ranker.rank(query=query, docs=docs)"
    ]
   },
@@ -935,7 +961,9 @@
     }
    ],
    "source": [
-    "ranker = Reranker('gpt-4-turbo', model_type='rankllm', api_key=os.environ['OPENAI_API_KEY'])\n",
+    "ranker = Reranker(\n",
+    "    \"gpt-4-turbo\", model_type=\"rankllm\", api_key=os.environ[\"OPENAI_API_KEY\"]\n",
+    ")\n",
     "ranker.rank(query=query, docs=docs)"
    ]
   },
@@ -990,7 +1018,8 @@
    ],
    "source": [
     "from rerankers import Reranker\n",
-    "ranker = Reranker('colbert')\n",
+    "\n",
+    "ranker = Reranker(\"colbert\")\n",
     "results = ranker.rank(query=query, docs=docs)\n",
     "results.top_k(1)"
    ]
@@ -1021,7 +1050,7 @@
     }
    ],
    "source": [
-    "ranker = Reranker('antoinelouis/colbertv2-camembert-L4-mmarcoFR', model_type=\"colbert\")"
+    "ranker = Reranker(\"antoinelouis/colbertv2-camembert-L4-mmarcoFR\", model_type=\"colbert\")"
    ]
   },
   {
@@ -1052,15 +1081,16 @@
     }
    ],
    "source": [
-    "ranker = Reranker('colbert',\n",
+    "ranker = Reranker(\n",
+    "    \"colbert\",\n",
     "    model_type=\"colbert\",\n",
-    "    verbose = 1, # How verbose the reranker will be. Defaults to 1, setting it to 0 will suppress most messages.\n",
-    "    dtype = None, # Which dtype the model should use. If None will figure out if your platform + model combo supports fp16 and use it if so, other fp32.\n",
-    "    device = None,  # Which device the model should use. If None will figure out what the most powerful supported platform available is (cuda > mps > cpu)\n",
-    "    batch_size = 16,  # The batch size the model will use. Defaults to 16\n",
-    "    query_token = \"[unused0]\",  # A ColBERT-specific argument. The token that your model prepends to queries.\n",
-    "    document_token = \"[unused1]\",  # A ColBERT-specific argument. The token that your model prepends to documents.\n",
-    "    )"
+    "    verbose=1,  # How verbose the reranker will be. Defaults to 1, setting it to 0 will suppress most messages.\n",
+    "    dtype=None,  # Which dtype the model should use. If None will figure out if your platform + model combo supports fp16 and use it if so, other fp32.\n",
+    "    device=None,  # Which device the model should use. If None will figure out what the most powerful supported platform available is (cuda > mps > cpu)\n",
+    "    batch_size=16,  # The batch size the model will use. Defaults to 16\n",
+    "    query_token=\"[unused0]\",  # A ColBERT-specific argument. The token that your model prepends to queries.\n",
+    "    document_token=\"[unused1]\",  # A ColBERT-specific argument. The token that your model prepends to documents.\n",
+    ")"
    ]
   },
   {
@@ -1101,7 +1131,7 @@
     }
    ],
    "source": [
-    "ranker = Reranker('flashrank') # Defaults to MiniLM-L12-v2\n",
+    "ranker = Reranker(\"flashrank\")  # Defaults to MiniLM-L12-v2\n",
     "results = ranker.rank(query=query, docs=docs)\n",
     "results.top_k(1)"
    ]
@@ -1152,9 +1182,9 @@
     }
    ],
    "source": [
-    "ranker = Reranker('ms-marco-TinyBERT-L-2-v2', model_type='flashrank')\n",
+    "ranker = Reranker(\"ms-marco-TinyBERT-L-2-v2\", model_type=\"flashrank\")\n",
     "results = ranker.rank(query=query, docs=docs)\n",
-    "results\n"
+    "results"
    ]
   },
   {
diff --git a/pyproject.toml b/pyproject.toml
index 1fab24f..a9d9eb4 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -1,4 +1,3 @@
-
 [build-system]
 requires = ["setuptools"] 
 build-backend = "setuptools.build_meta" 
@@ -10,6 +9,20 @@ packages = [
     "rerankers.integrations",
 ]
 
+[tool.ruff]
+exclude = [
+    "examples/*"
+]
+
+[tool.mypy]
+python_version = "3.12"
+disallow_untyped_defs = true
+namespace_packages = true
+
+[[tool.mypy.overrides]]
+module = ["requests"]
+ignore_missing_imports = true
+
 [project]
 name = "rerankers" 
 
@@ -73,7 +86,7 @@ rankllm = [
   "nmslib-metabrainz; python_version >= '3.10'",
   "rank-llm; python_version >= '3.10'"
 ]
-dev = ["ruff", "isort", "pytest", "ipyprogress", "ipython", "ranx", "ir_datasets", "srsly"]
+dev = ["ruff", "pytest", "ipyprogress", "ipython", "ranx", "ir_datasets", "srsly"]
 
 [project.urls]
-"Homepage" = "https://github.com/answerdotai/rerankers"
\ No newline at end of file
+"Homepage" = "https://github.com/answerdotai/rerankers"
diff --git a/requirements-ci.txt b/requirements-ci.txt
new file mode 100644
index 0000000..23dbb10
--- /dev/null
+++ b/requirements-ci.txt
@@ -0,0 +1,4 @@
+ruff
+pytest
+
+.[transformers]
diff --git a/rerankers/models/__init__.py b/rerankers/models/__init__.py
index cd0439d..5842c0e 100644
--- a/rerankers/models/__init__.py
+++ b/rerankers/models/__init__.py
@@ -1,4 +1,6 @@
-AVAILABLE_RANKERS = {}
+from typing import Any
+
+AVAILABLE_RANKERS: dict[str, Any] = {}
 
 try:
     from rerankers.models.transformer_ranker import TransformerRanker
diff --git a/rerankers/models/api_rankers.py b/rerankers/models/api_rankers.py
index bf3a667..495b30d 100644
--- a/rerankers/models/api_rankers.py
+++ b/rerankers/models/api_rankers.py
@@ -16,26 +16,28 @@
     "mixedbread.ai": "https://api.mixedbread.ai/v1/reranking",
 }
 
-DOCUMENT_KEY_MAPPING = {
-    "mixedbread.ai": "input",
-    "text-embeddings-inference":"texts"
-}
+DOCUMENT_KEY_MAPPING = {"mixedbread.ai": "input", "text-embeddings-inference": "texts"}
 RETURN_DOCUMENTS_KEY_MAPPING = {
-    "mixedbread.ai":"return_input",
-    "text-embeddings-inference":"return_text"
+    "mixedbread.ai": "return_input",
+    "text-embeddings-inference": "return_text",
 }
 RESULTS_KEY_MAPPING = {
     "voyage": "data",
     "mixedbread.ai": "data",
-    "text-embeddings-inference": None
-}
-SCORE_KEY_MAPPING = {
-    "mixedbread.ai": "score",
-    "text-embeddings-inference":"score"
+    "text-embeddings-inference": None,
 }
+SCORE_KEY_MAPPING = {"mixedbread.ai": "score", "text-embeddings-inference": "score"}
+
 
 class APIRanker(BaseRanker):
-    def __init__(self, model: str, api_key: str, api_provider: str, verbose: int = 1, url: str = None):
+    def __init__(
+        self,
+        model: str,
+        api_key: str,
+        api_provider: str,
+        verbose: int = 1,
+        url: str = None,
+    ):
         self.api_key = api_key
         self.model = model
         self.api_provider = api_provider.lower()
@@ -48,7 +50,6 @@ def __init__(self, model: str, api_key: str, api_provider: str, verbose: int = 1
         }
         self.url = url if url else URLS[self.api_provider]
 
-
     def _get_document_text(self, r: dict) -> str:
         if self.api_provider == "voyage":
             return r["document"]
@@ -60,14 +61,16 @@ def _get_document_text(self, r: dict) -> str:
             return r["document"]["text"]
 
     def _get_score(self, r: dict) -> float:
-        score_key = SCORE_KEY_MAPPING.get(self.api_provider,"relevance_score")
+        score_key = SCORE_KEY_MAPPING.get(self.api_provider, "relevance_score")
         return r[score_key]
 
     def _parse_response(
-        self, response: dict,  docs: List[Document],
+        self,
+        response: dict,
+        docs: List[Document],
     ) -> RankedResults:
         ranked_docs = []
-        results_key = RESULTS_KEY_MAPPING.get(self.api_provider,"results")
+        results_key = RESULTS_KEY_MAPPING.get(self.api_provider, "results")
         print(response)
 
         for i, r in enumerate(response[results_key] if results_key else response):
@@ -95,13 +98,14 @@ def rank(
         results = self._parse_response(response.json(), docs)
         return RankedResults(results=results, query=query, has_scores=True)
 
-
     def _format_payload(self, query: str, docs: List[str]) -> str:
         top_key = (
             "top_n" if self.api_provider not in ["voyage", "mixedbread.ai"] else "top_k"
         )
-        documents_key = DOCUMENT_KEY_MAPPING.get(self.api_provider,"documents")
-        return_documents_key = RETURN_DOCUMENTS_KEY_MAPPING.get(self.api_provider,"return_documents")
+        documents_key = DOCUMENT_KEY_MAPPING.get(self.api_provider, "documents")
+        return_documents_key = RETURN_DOCUMENTS_KEY_MAPPING.get(
+            self.api_provider, "return_documents"
+        )
 
         payload = {
             "model": self.model,
diff --git a/rerankers/models/colbert_ranker.py b/rerankers/models/colbert_ranker.py
index 57f02d9..025f1c2 100644
--- a/rerankers/models/colbert_ranker.py
+++ b/rerankers/models/colbert_ranker.py
@@ -3,7 +3,7 @@
 
 import torch
 import torch.nn as nn
-from transformers import BertPreTrainedModel, BertModel, AutoModel, AutoTokenizer
+from transformers import BertPreTrainedModel, BertModel, AutoTokenizer
 from typing import List, Optional, Union
 from math import ceil
 
diff --git a/rerankers/models/ranker.py b/rerankers/models/ranker.py
index 6c8f45a..95f53a4 100644
--- a/rerankers/models/ranker.py
+++ b/rerankers/models/ranker.py
@@ -33,10 +33,10 @@ async def rank_async(
         docs: List[str],
         doc_ids: Optional[Union[List[str], str]] = None,
     ) -> RankedResults:
-
-
         loop = get_event_loop()
-        return await loop.run_in_executor(None, partial(self.rank, query, docs, doc_ids))
+        return await loop.run_in_executor(
+            None, partial(self.rank, query, docs, doc_ids)
+        )
 
     def as_langchain_compressor(self, k: int = 10):
         try:
diff --git a/rerankers/models/t5ranker.py b/rerankers/models/t5ranker.py
index 30a4e18..5be6c45 100644
--- a/rerankers/models/t5ranker.py
+++ b/rerankers/models/t5ranker.py
@@ -14,8 +14,6 @@
 from rerankers.documents import Document
 
 
-import torch
-
 from rerankers.results import RankedResults, Result
 from rerankers.utils import (
     vprint,
@@ -89,7 +87,7 @@ def __init__(
         token_false: str = "auto",
         token_true: str = "auto",
         return_logits: bool = False,
-        inputs_template: str = "Query: {query} Document: {text} Relevant:"
+        inputs_template: str = "Query: {query} Document: {text} Relevant:",
     ):
         """
         Implementation of the key functions from https://github.com/unicamp-dl/InRanker/blob/main/inranker/rankers.py
@@ -197,8 +195,7 @@ def _get_scores(
             total=ceil(len(docs) / batch_size),
         ):
             queries_documents = [
-                self.inputs_template.format(query=query, text=text)
-                for text in batch
+                self.inputs_template.format(query=query, text=text) for text in batch
             ]
             tokenized = self.tokenizer(
                 queries_documents,
diff --git a/rerankers/reranker.py b/rerankers/reranker.py
index 5c7599b..4e68f46 100644
--- a/rerankers/reranker.py
+++ b/rerankers/reranker.py
@@ -52,7 +52,9 @@
 PROVIDERS = ["cohere", "jina", "voyage", "mixedbread.ai", "text-embeddings-inference"]
 
 
-def _get_api_provider(model_name: str, model_type: Optional[str] = None) -> str:
+def _get_api_provider(
+    model_name: str, model_type: Optional[str] = None
+) -> Optional[str]:
     if model_type in PROVIDERS or any(provider in model_name for provider in PROVIDERS):
         return model_type or next(
             (provider for provider in PROVIDERS if provider in model_name), None
@@ -137,7 +139,7 @@ def _get_defaults(
     model_type: Optional[str] = None,
     lang: str = "en",
     verbose: int = 1,
-) -> str:
+) -> tuple[Optional[str], Optional[str]]:
     if model_name in DEFAULTS.keys():
         print(f"Loading default {model_name} model for language {lang}")
         try:
diff --git a/tests/consistency_notebooks/test_colbert.ipynb b/tests/consistency_notebooks/test_colbert.ipynb
index 229b6b4..46be64d 100644
--- a/tests/consistency_notebooks/test_colbert.ipynb
+++ b/tests/consistency_notebooks/test_colbert.ipynb
@@ -58,8 +58,8 @@
    "source": [
     "import srsly\n",
     "\n",
-    "corpus = [x for x in srsly.read_jsonl('./data/scifact/corpus.jsonl')]\n",
-    "queries = [x for x in srsly.read_jsonl('./data/scifact/queries.jsonl')]\n",
+    "corpus = [x for x in srsly.read_jsonl(\"./data/scifact/corpus.jsonl\")]\n",
+    "queries = [x for x in srsly.read_jsonl(\"./data/scifact/queries.jsonl\")]\n",
     "\n",
     "corpus[0]"
    ]
@@ -79,7 +79,7 @@
     }
    ],
    "source": [
-    "ranker = Reranker('colbert', device='cuda', verbose=0)"
+    "ranker = Reranker(\"colbert\", device=\"cuda\", verbose=0)"
    ]
   },
   {
@@ -88,8 +88,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "top100 = srsly.read_json('./data/scifact/scifact_top_100.json')\n",
-    "\n"
+    "top100 = srsly.read_json(\"./data/scifact/scifact_top_100.json\")"
    ]
   },
   {
@@ -98,7 +97,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "corpus_map = {x['_id']: f\"{x['title']} {x['text']}\" for x in corpus}"
+    "corpus_map = {x[\"_id\"]: f\"{x['title']} {x['text']}\" for x in corpus}"
    ]
   },
   {
@@ -115,15 +114,16 @@
     }
    ],
    "source": [
-    "qrels_dict = dict(qrels)\n",
-    "queries = [q for q in queries if q['_id'] in qrels_dict]\n",
     "from tqdm import tqdm\n",
     "\n",
+    "qrels_dict = dict(qrels)\n",
+    "queries = [q for q in queries if q[\"_id\"] in qrels_dict]\n",
+    "\n",
     "scores = {}\n",
     "for q in tqdm(queries):\n",
-    "    doc_ids = top100[q['_id']]\n",
+    "    doc_ids = top100[q[\"_id\"]]\n",
     "    docs = [corpus_map[x] for x in doc_ids]\n",
-    "    scores[q['_id']] = ranker.rank(q['text'], docs, doc_ids=doc_ids)\n"
+    "    scores[q[\"_id\"]] = ranker.rank(q[\"text\"], docs, doc_ids=doc_ids)"
    ]
   },
   {
@@ -154,12 +154,15 @@
    ],
    "source": [
     "from ranx import evaluate\n",
-    "evaluation_score = evaluate(qrels, run, 'ndcg@10')\n",
-    "litterature_result = 0.693 # From ColBERTv2 Paper https://arxiv.org/abs/2112.01488\n",
+    "\n",
+    "evaluation_score = evaluate(qrels, run, \"ndcg@10\")\n",
+    "litterature_result = 0.693  # From ColBERTv2 Paper https://arxiv.org/abs/2112.01488\n",
     "if abs(evaluation_score - litterature_result) > 0.01:\n",
-    "    print(f\"Score {evaluation_score:0.3f} differs by more than 0.01 from the the reported score.\")\n",
+    "    print(\n",
+    "        f\"Score {evaluation_score:0.3f} differs by more than 0.01 from the the reported score.\"\n",
+    "    )\n",
     "else:\n",
-    "    print(f\"Score is within 0.01 NDCG@10 of the reported score!\")\n"
+    "    print(\"Score is within 0.01 NDCG@10 of the reported score!\")"
    ]
   }
  ],
diff --git a/tests/consistency_notebooks/test_crossenc.ipynb b/tests/consistency_notebooks/test_crossenc.ipynb
index 5c0f604..b0fac80 100644
--- a/tests/consistency_notebooks/test_crossenc.ipynb
+++ b/tests/consistency_notebooks/test_crossenc.ipynb
@@ -6,7 +6,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "from ranx import Qrels, Run\n"
+    "from ranx import Qrels, Run"
    ]
   },
   {
@@ -63,7 +63,8 @@
    ],
    "source": [
     "from rerankers import Reranker\n",
-    "ranker = Reranker('castorini/monot5-base-msmarco-10k', device='cuda', batch_size=128)"
+    "\n",
+    "ranker = Reranker(\"castorini/monot5-base-msmarco-10k\", device=\"cuda\", batch_size=128)"
    ]
   },
   {
@@ -88,8 +89,8 @@
    "source": [
     "import srsly\n",
     "\n",
-    "corpus = [x for x in srsly.read_jsonl('./data/scifact/corpus.jsonl')]\n",
-    "queries = [x for x in srsly.read_jsonl('./data/scifact/queries.jsonl')]\n",
+    "corpus = [x for x in srsly.read_jsonl(\"./data/scifact/corpus.jsonl\")]\n",
+    "queries = [x for x in srsly.read_jsonl(\"./data/scifact/queries.jsonl\")]\n",
     "\n",
     "corpus[0]"
    ]
@@ -122,7 +123,7 @@
     }
    ],
    "source": [
-    "ranker = Reranker('mixedbread-ai/mxbai-rerank-base-v1', device='cuda')"
+    "ranker = Reranker(\"mixedbread-ai/mxbai-rerank-base-v1\", device=\"cuda\")"
    ]
   },
   {
@@ -131,7 +132,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "top100 = srsly.read_json('data/scifact/scifact_top_100.json')\n"
+    "top100 = srsly.read_json(\"data/scifact/scifact_top_100.json\")"
    ]
   },
   {
@@ -140,7 +141,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "corpus_map = {x['_id']: f\"{x['title']} {x['text']}\" for x in corpus}"
+    "corpus_map = {x[\"_id\"]: f\"{x['title']} {x['text']}\" for x in corpus}"
    ]
   },
   {
@@ -157,15 +158,16 @@
     }
    ],
    "source": [
-    "qrels_dict = dict(qrels)\n",
-    "queries = [q for q in queries if q['_id'] in qrels_dict]\n",
     "from tqdm import tqdm\n",
     "\n",
+    "qrels_dict = dict(qrels)\n",
+    "queries = [q for q in queries if q[\"_id\"] in qrels_dict]\n",
+    "\n",
     "scores = {}\n",
     "for q in tqdm(queries):\n",
-    "    doc_ids = top100[q['_id']]\n",
+    "    doc_ids = top100[q[\"_id\"]]\n",
     "    docs = [corpus_map[x] for x in doc_ids]\n",
-    "    scores[q['_id']] = ranker.rank(q['text'], docs, doc_ids=doc_ids)\n"
+    "    scores[q[\"_id\"]] = ranker.rank(q[\"text\"], docs, doc_ids=doc_ids)"
    ]
   },
   {
@@ -177,7 +179,7 @@
     "scores_dict = {}\n",
     "for q_id, ranked_results in scores.items():\n",
     "    top_10_results = ranked_results.top_k(10)\n",
-    "    scores_dict[q_id] = {result.doc_id: result.score for result in top_10_results}\n"
+    "    scores_dict[q_id] = {result.doc_id: result.score for result in top_10_results}"
    ]
   },
   {
@@ -224,12 +226,15 @@
    ],
    "source": [
     "from ranx import evaluate\n",
-    "evaluation_score = evaluate(qrels, run, 'ndcg@10')\n",
-    "litterature_result = 0.724 # from MXBAI https://docs.google.com/spreadsheets/d/15ELkSMFv-oHa5TRiIjDvhIstH9dlc3pnZeO-iGz4Ld4/edit#gid=0\n",
+    "\n",
+    "evaluation_score = evaluate(qrels, run, \"ndcg@10\")\n",
+    "litterature_result = 0.724  # from MXBAI https://docs.google.com/spreadsheets/d/15ELkSMFv-oHa5TRiIjDvhIstH9dlc3pnZeO-iGz4Ld4/edit#gid=0\n",
     "if abs(evaluation_score - litterature_result) > 0.01:\n",
-    "    print(f\"Score {evaluation_score:0.3f} differs by more than 0.01 from the the reported score.\")\n",
+    "    print(\n",
+    "        f\"Score {evaluation_score:0.3f} differs by more than 0.01 from the the reported score.\"\n",
+    "    )\n",
     "else:\n",
-    "    print(f\"Score is within 0.01NDCG@10 of the reported score!\")\n"
+    "    print(\"Score is within 0.01NDCG@10 of the reported score!\")"
    ]
   }
  ],
diff --git a/tests/consistency_notebooks/test_inranker.ipynb b/tests/consistency_notebooks/test_inranker.ipynb
index 119bdb8..466f2f8 100644
--- a/tests/consistency_notebooks/test_inranker.ipynb
+++ b/tests/consistency_notebooks/test_inranker.ipynb
@@ -6,8 +6,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "from ranx import Qrels, Run\n",
-    "\n"
+    "from ranx import Qrels, Run"
    ]
   },
   {
@@ -59,8 +58,8 @@
    "source": [
     "import srsly\n",
     "\n",
-    "corpus = [x for x in srsly.read_jsonl('./data/scifact/corpus.jsonl')]\n",
-    "queries = [x for x in srsly.read_jsonl('./data/scifact/queries.jsonl')]\n",
+    "corpus = [x for x in srsly.read_jsonl(\"./data/scifact/corpus.jsonl\")]\n",
+    "queries = [x for x in srsly.read_jsonl(\"./data/scifact/queries.jsonl\")]\n",
     "\n",
     "corpus[0]"
    ]
@@ -87,7 +86,7 @@
     }
    ],
    "source": [
-    "ranker = Reranker('unicamp-dl/InRanker-base', device='cuda', batch_size=32, verbose=0)"
+    "ranker = Reranker(\"unicamp-dl/InRanker-base\", device=\"cuda\", batch_size=32, verbose=0)"
    ]
   },
   {
@@ -96,8 +95,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "top100 = srsly.read_json('data/scifact/scifact_top_100.json')\n",
-    "\n"
+    "top100 = srsly.read_json(\"data/scifact/scifact_top_100.json\")"
    ]
   },
   {
@@ -106,7 +104,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "corpus_map = {x['_id']: f\"{x['title']} {x['text']}\" for x in corpus}"
+    "corpus_map = {x[\"_id\"]: f\"{x['title']} {x['text']}\" for x in corpus}"
    ]
   },
   {
@@ -123,15 +121,16 @@
     }
    ],
    "source": [
-    "qrels_dict = dict(qrels)\n",
-    "queries = [q for q in queries if q['_id'] in qrels_dict]\n",
     "from tqdm import tqdm\n",
     "\n",
+    "qrels_dict = dict(qrels)\n",
+    "queries = [q for q in queries if q[\"_id\"] in qrels_dict]\n",
+    "\n",
     "scores = {}\n",
     "for q in tqdm(queries):\n",
-    "    doc_ids = top100[q['_id']]\n",
+    "    doc_ids = top100[q[\"_id\"]]\n",
     "    docs = [corpus_map[x] for x in doc_ids]\n",
-    "    scores[q['_id']] = ranker.rank(q['text'], docs, doc_ids=doc_ids)\n"
+    "    scores[q[\"_id\"]] = ranker.rank(q[\"text\"], docs, doc_ids=doc_ids)"
    ]
   },
   {
@@ -162,12 +161,15 @@
    ],
    "source": [
     "from ranx import evaluate\n",
-    "evaluation_score = evaluate(qrels, run, 'ndcg@10')\n",
-    "litterature_result = 0.7618 # From InRanker Paper https://arxiv.org/pdf/2401.06910.pdf\n",
+    "\n",
+    "evaluation_score = evaluate(qrels, run, \"ndcg@10\")\n",
+    "litterature_result = 0.7618  # From InRanker Paper https://arxiv.org/pdf/2401.06910.pdf\n",
     "if abs(evaluation_score - litterature_result) > 0.01:\n",
-    "    print(f\"Score {evaluation_score:0.3f} differs by more than 0.01 from the the reported score.\")\n",
+    "    print(\n",
+    "        f\"Score {evaluation_score:0.3f} differs by more than 0.01 from the the reported score.\"\n",
+    "    )\n",
     "else:\n",
-    "    print(f\"Score is within 0.01 NDCG@10 of the reported score!\")\n"
+    "    print(\"Score is within 0.01 NDCG@10 of the reported score!\")"
    ]
   }
  ],
diff --git a/tests/consistency_notebooks/test_t5.ipynb b/tests/consistency_notebooks/test_t5.ipynb
index 421f037..351ab0c 100644
--- a/tests/consistency_notebooks/test_t5.ipynb
+++ b/tests/consistency_notebooks/test_t5.ipynb
@@ -58,8 +58,8 @@
    "source": [
     "import srsly\n",
     "\n",
-    "corpus = [x for x in srsly.read_jsonl('./data/scifact/corpus.jsonl')]\n",
-    "queries = [x for x in srsly.read_jsonl('./data/scifact/queries.jsonl')]\n",
+    "corpus = [x for x in srsly.read_jsonl(\"./data/scifact/corpus.jsonl\")]\n",
+    "queries = [x for x in srsly.read_jsonl(\"./data/scifact/queries.jsonl\")]\n",
     "\n",
     "corpus[0]"
    ]
@@ -86,7 +86,9 @@
     }
    ],
    "source": [
-    "ranker = Reranker('castorini/monot5-base-msmarco-10k', device='cuda', batch_size=128, verbose=0)"
+    "ranker = Reranker(\n",
+    "    \"castorini/monot5-base-msmarco-10k\", device=\"cuda\", batch_size=128, verbose=0\n",
+    ")"
    ]
   },
   {
@@ -95,8 +97,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "top100 = srsly.read_json('data/scifact/scifact_top_100.json')\n",
-    "\n"
+    "top100 = srsly.read_json(\"data/scifact/scifact_top_100.json\")"
    ]
   },
   {
@@ -105,7 +106,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "corpus_map = {x['_id']: f\"{x['title']} {x['text']}\" for x in corpus}"
+    "corpus_map = {x[\"_id\"]: f\"{x['title']} {x['text']}\" for x in corpus}"
    ]
   },
   {
@@ -122,15 +123,16 @@
     }
    ],
    "source": [
-    "qrels_dict = dict(qrels)\n",
-    "queries = [q for q in queries if q['_id'] in qrels_dict]\n",
     "from tqdm import tqdm\n",
     "\n",
+    "qrels_dict = dict(qrels)\n",
+    "queries = [q for q in queries if q[\"_id\"] in qrels_dict]\n",
+    "\n",
     "scores = {}\n",
     "for q in tqdm(queries):\n",
-    "    doc_ids = top100[q['_id']]\n",
+    "    doc_ids = top100[q[\"_id\"]]\n",
     "    docs = [corpus_map[x] for x in doc_ids]\n",
-    "    scores[q['_id']] = ranker.rank(q['text'], docs, doc_ids=doc_ids)\n"
+    "    scores[q[\"_id\"]] = ranker.rank(q[\"text\"], docs, doc_ids=doc_ids)"
    ]
   },
   {
@@ -161,12 +163,15 @@
    ],
    "source": [
     "from ranx import evaluate\n",
-    "evaluation_score = evaluate(qrels, run, 'ndcg@10')\n",
-    "litterature_result = 0.734 # From RankGPT Paper https://arxiv.org/pdf/2304.09542.pdf\n",
+    "\n",
+    "evaluation_score = evaluate(qrels, run, \"ndcg@10\")\n",
+    "litterature_result = 0.734  # From RankGPT Paper https://arxiv.org/pdf/2304.09542.pdf\n",
     "if abs(evaluation_score - litterature_result) > 0.01:\n",
-    "    print(f\"Score {evaluation_score:0.3f} differs by more than 0.01 from the the reported score.\")\n",
+    "    print(\n",
+    "        f\"Score {evaluation_score:0.3f} differs by more than 0.01 from the the reported score.\"\n",
+    "    )\n",
     "else:\n",
-    "    print(f\"Score is within 0.01 NDCG@10 of the reported score!\")\n"
+    "    print(\"Score is within 0.01 NDCG@10 of the reported score!\")"
    ]
   }
  ],
diff --git a/tests/test_crossenc.py b/tests/test_crossenc.py
index 9637708..e149d11 100644
--- a/tests/test_crossenc.py
+++ b/tests/test_crossenc.py
@@ -1,10 +1,9 @@
 from unittest.mock import patch
-import torch
-from rerankers import Reranker
 from rerankers.models.transformer_ranker import TransformerRanker
 from rerankers.results import Result, RankedResults
 from rerankers.documents import Document
 
+
 @patch("rerankers.models.transformer_ranker.TransformerRanker.rank")
 def test_transformer_ranker_rank(mock_rank):
     query = "Gone with the wind is an absolute masterpiece"
@@ -15,12 +14,17 @@ def test_transformer_ranker_rank(mock_rank):
     expected_results = RankedResults(
         results=[
             Result(
-                document=Document(id=1, text="Gone with the wind is an all-time classic"),
+                document=Document(
+                    doc_id=1, text="Gone with the wind is an all-time classic"
+                ),
                 score=1.6181640625,
                 rank=1,
             ),
             Result(
-                document=Document(id=0, text="Gone with the wind is a masterclass in bad storytelling."),
+                document=Document(
+                    doc_id=0,
+                    text="Gone with the wind is a masterclass in bad storytelling.",
+                ),
                 score=0.88427734375,
                 rank=2,
             ),
diff --git a/tests/test_results.py b/tests/test_results.py
index 2fbeb22..f60be44 100644
--- a/tests/test_results.py
+++ b/tests/test_results.py
@@ -6,8 +6,8 @@
 def test_ranked_results_functions():
     results = RankedResults(
         results=[
-            Result(document=Document(id=0, text="Doc 0"), score=0.9, rank=2),
-            Result(document=Document(id=1, text="Doc 1"), score=0.95, rank=1),
+            Result(document=Document(doc_id=0, text="Doc 0"), score=0.9, rank=2),
+            Result(document=Document(doc_id=1, text="Doc 1"), score=0.95, rank=1),
         ],
         query="Test Query",
         has_scores=True,
@@ -20,7 +20,7 @@ def test_ranked_results_functions():
 
 
 def test_result_attributes():
-    result = Result(document=Document(id=1, text="Doc 1"), score=0.95, rank=1)
+    result = Result(document=Document(doc_id=1, text="Doc 1"), score=0.95, rank=1)
     assert result.doc_id == 1
     assert result.text == "Doc 1"
     assert result.score == 0.95