diff --git a/Makefile b/Makefile
index 19c5155..5f46cbd 100644
--- a/Makefile
+++ b/Makefile
@@ -3,8 +3,9 @@
 
 coverage:  ## Run tests with coverage
 	python -m coverage erase
-	python -m coverage run -m pytest -ra
+	python -m coverage run -m pytest -ra -v
 	python -m coverage report -m
+	python -m coverage html
 
 deps:  ## Install dependencies
 	pip install --upgrade pip
@@ -21,7 +22,7 @@ lint:  ## Lint
 	python -m flake8 backend/src frontend/app.py tests
 
 test:  ## Run tests
-	python -m pytest -ra
+	python -m pytest -ra -v
 
 build:
 	make deps
diff --git a/backend/src/executors.py b/backend/src/executors.py
index 5db6203..84b7f59 100644
--- a/backend/src/executors.py
+++ b/backend/src/executors.py
@@ -72,7 +72,6 @@ def finetune(self, docs: DocumentArray, **kwargs) -> DocumentArray:
         train_data = []
 
         for doc in docs:
-            matches = doc.matches
             for match in doc.matches:
                 train_data.append(
                     InputExample(
diff --git a/frontend/app.py b/frontend/app.py
index a6de81b..0c483aa 100644
--- a/frontend/app.py
+++ b/frontend/app.py
@@ -1,5 +1,6 @@
-import requests
 import json
+
+import requests
 import streamlit as st
 
 
@@ -27,8 +28,8 @@ def search(abstract: str, host: str) -> dict:
 def finetune(doc: dict, match: dict, relevant: bool, host: str) -> dict:
     labeled = doc.copy()
 
-    labeled["tags"]["finetuner"] = {"label": 1 if relevant else -1}
-    match["tags"]["finetuner"] = {"label": 1 if relevant else -1}
+    labeled["tags"] = {"finetuner": {"label": 1 if relevant else -1}}
+    match["tags"] = {"finetuner": {"label": 1 if relevant else -1}}
     labeled["matches"] = [match]
     data = {"data": [labeled]}
 
@@ -58,7 +59,6 @@ def match_score(match):
 if st.button(label="Search") or query:
     if query:
         doc = search(query, host)
-
         max_score = max(match_score(match) for match in doc["matches"])
 
         for match in doc["matches"]:
@@ -78,14 +78,14 @@ def match_score(match):
                 with col1:
                     st.metric("Similarity (1 is best)", score, score_diff)
                 with col2:
-                    if st.button("✔️ Mark as relevant", id):
+                    if st.button("✔️ Mark as relevant", id + "1"):
                         with st.spinner("Finetuning..."):
                             finetune(doc, match, True, host)
                         st.success("Finetuned!")
                         st.experimental_rerun()
 
                 with col3:
-                    if st.button("✖️ Mark as irrelevant", id):
+                    if st.button("✖️ Mark as irrelevant", id + "0"):
                         with st.spinner("Finetuning..."):
                             finetune(doc, match, False, host)
                         st.success("Finetuned!")
diff --git a/tests/test_executors.py b/tests/test_executors.py
index 3f39a50..b195920 100644
--- a/tests/test_executors.py
+++ b/tests/test_executors.py
@@ -23,17 +23,14 @@ def label_doc(docs):
     d = Document(text=papers[0])
 
     matches = DocumentArray(
-        [Document(text=seq, tags={"finetuner": {"label": 1}}) for seq in papers[1:]]
+        [Document(text=seq, tags={"finetuner": {"label": 1.0}}) for seq in papers[1:]]
     )
-
     d.matches.extend(matches)
-
     return d
 
 
 def encode_sequences(docs):
     embedded_docs = SpecterExecutor().encode(docs)
-
     return embedded_docs
 
 
@@ -55,3 +52,8 @@ def test_search():
         results = flow.search(docs)
 
     return results
+
+
+def test_finetuner():
+    labelled_docs = DocumentArray(label_doc(docs))
+    SpecterExecutor().finetune(labelled_docs)