Update app.py

app.py

@@ -1,7 +1,52 @@
 import gradio as gr
+import csv
+import random
+import uuid
 from langchain.vectorstores import FAISS
 from langchain.embeddings import HuggingFaceEmbeddings
 
+USER_ID = uuid.uuid4()
+INDEXES = ["miread_large", "miread_contrastive", "scibert_contrastive"]
+MODELS = [
+    "biodatlab/MIReAD-Neuro-Large",
+    "biodatlab/MIReAD-Neuro-Contrastive",
+    "biodatlab/SciBERT-Neuro-Contrastive",
+]
+model_kwargs = {'device': 'cpu'}
+encode_kwargs = {'normalize_embeddings': False}
+faiss_embedders = [HuggingFaceEmbeddings(
+    model_name=name,
+    model_kwargs=model_kwargs,
+    encode_kwargs=encode_kwargs) for name in MODELS]
+
+vecdbs = [FAISS.load_local(index_name, faiss_embedder)
+          for index_name, faiss_embedder in zip(INDEXES, faiss_embedders)]
+
+def get_matchup():
+    choices = INDEXES
+    left, right = random.sample(choices,2)
+    return left, right
+
+def get_comp(prompt):
+    left, right = get_matchup()
+    left_output = inference(prompt,left)
+    right_output = inference(prompt,right)
+    return left_output, right_output
+
+def get_article(db_name="miread_contrastive"):
+    db = vecdbs[index_names.index(db_name)]
+    return db[0]
+
+
+
+
+
+def send_result(l_output, r_output, prompt, pick):
+    with csv.open('results.csv','a') as res_file:
+      writer = csv.writer(res_file)
+      row = [USER_ID,left,right,prompt,pick]
+      writer.writerow(row)
+
 
 def get_matches(query, db_name="miread_contrastive"):
     """
@@ -19,9 +64,7 @@ def inference(query, model="miread_contrastive"):
     """
     matches = get_matches(query, model)
     auth_counts = {}
-    j_bucket = {}
     n_table = []
-    a_table = []
     scores = [round(match[1].item(), 3) for match in matches]
     min_score = min(scores)
     max_score = max(scores)
@@ -33,18 +76,6 @@ def inference(query, model="miread_contrastive"):
         author = doc.metadata['authors'][0].title()
         date = doc.metadata.get('date', 'None')
         link = doc.metadata.get('link', 'None')
-        submitter = doc.metadata.get('submitter', 'None')
-        journal = doc.metadata['journal']
-        if (journal is None or journal.strip() == ''):
-            journal = 'None'
-        else:
-            journal = journal.strip()
-
-        # For journals
-        if journal not in j_bucket:
-            j_bucket[journal] = score
-        else:
-            j_bucket[journal] += score
 
         # For authors
         record = [i+1,
@@ -59,49 +90,12 @@ def inference(query, model="miread_contrastive"):
                 auth_counts[author] = 1
             else:
                 auth_counts[author] += 1
-
-        # For abstracts
-        record = [i+1,
-                  title,
-                  author,
-                  submitter,
-                  journal,
-                  date,
-                  link,
-                  score
-                  ]
-        a_table.append(record)
-
-    del j_bucket['None']
-    j_table = sorted([[journal, round(score, 3)] for journal,
-                     score in j_bucket.items()],
-                     key=lambda x: x[1], reverse=True)
-    j_table = [[i+1, item[0], item[1]] for i, item in enumerate(j_table)]
-    j_output = gr.Dataframe.update(value=j_table, visible=True)
     n_output = gr.Dataframe.update(value=n_table, visible=True)
-    a_output = gr.Dataframe.update(value=a_table, visible=True)
-
-    return [a_output, j_output, n_output]
-
-
-index_names = ["miread_large", "miread_contrastive", "scibert_contrastive"]
-model_names = [
-    "biodatlab/MIReAD-Neuro-Large",
-    "biodatlab/MIReAD-Neuro-Contrastive",
-    "biodatlab/SciBERT-Neuro-Contrastive",
-]
-model_kwargs = {'device': 'cpu'}
-encode_kwargs = {'normalize_embeddings': False}
-faiss_embedders = [HuggingFaceEmbeddings(
-    model_name=name,
-    model_kwargs=model_kwargs,
-    encode_kwargs=encode_kwargs) for name in model_names]
+    return n_output
 
-vecdbs = [FAISS.load_local(index_name, faiss_embedder)
-          for index_name, faiss_embedder in zip(index_names, faiss_embedders)]
 
 with gr.Blocks(theme=gr.themes.Soft()) as demo:
-    gr.Markdown("# NBDT Recommendation Engine for Editors")
+    gr.Markdown("# NBDT Recommendation Engine Arena")
     gr.Markdown("NBDT Recommendation Engine for Editors is a tool for neuroscience authors/abstracts/journalsrecommendation built for NBDT journal editors. \
     It aims to help an editor to find similar reviewers, abstracts, and journals to a given submitted abstract.\
     To find a recommendation, paste a `title[SEP]abstract` or `abstract` in the text box below and click on the appropriate \"Find Matches\" button.\
@@ -109,56 +103,44 @@ with gr.Blocks(theme=gr.themes.Soft()) as demo:
     The data in our current demo includes authors associated with the NBDT Journal. We will update the data monthly for an up-to-date publications.")
 
     abst = gr.Textbox(label="Abstract", lines=10)
-
-    action_btn1 = gr.Button(value="Find Matches with MIReAD-Neuro-Large")
-    action_btn2 = gr.Button(value="Find Matches with MIReAD-Neuro-Contrastive")
-    action_btn3 = gr.Button(
-        value="Find Matches with SciBERT-Neuro-Contrastive")
-
-    with gr.Tab("Authors"):
-        n_output = gr.Dataframe(
+    models = gr.State(value=get_matchup())
+    prompt = gr.State(value=get_prompt())
+    action_btn = gr.Button(value="Get comparison")
+    with gr.Row().style(equal_height=True):
+      with gr.Column(scale=1):
+        l_output = gr.Dataframe(
             headers=['No.', 'Score', 'Name', 'Title', 'Link', 'Date'],
             datatype=['number', 'number', 'str', 'str', 'str', 'str'],
             col_count=(6, "fixed"),
             wrap=True,
-            visible=False
-        )
-    with gr.Tab("Abstracts"):
-        a_output = gr.Dataframe(
-            headers=['No.', 'Title', 'Author', 'Corresponding Author',
-                     'Journal', 'Date', 'Link', 'Score'],
-            datatype=['number', 'str', 'str', 'str',
-                      'str', 'str', 'str', 'number'],
-            col_count=(8, "fixed"),
-            wrap=True,
-            visible=False
-        )
-    with gr.Tab("Journals"):
-        j_output = gr.Dataframe(
-            headers=['No.', 'Name', 'Score'],
-            datatype=['number', 'str', 'number'],
-            col_count=(3, "fixed"),
+            visible=True,
+            label='Model A',
+            show_label = True
+            scale=1
+            )
+        l_btn = gr.Button(value="Model A is better",scale=1)
+      with gr.Column(scale=1):
+        r_output = gr.Dataframe(
+            headers=['No.', 'Score', 'Name', 'Title', 'Link', 'Date'],
+            datatype=['number', 'number', 'str', 'str', 'str', 'str'],
+            col_count=(6, "fixed"),
             wrap=True,
-            visible=False
-        )
-
-    action_btn1.click(fn=lambda x: inference(x, index_names[0]),
-                      inputs=[
-        abst,
-    ],
-        outputs=[a_output, j_output, n_output],
-        api_name="neurojane")
-    action_btn2.click(fn=lambda x: inference(x, index_names[1]),
-                      inputs=[
-        abst,
-    ],
-        outputs=[a_output, j_output, n_output],
-        api_name="neurojane")
-    action_btn3.click(fn=lambda x: inference(x, index_names[2]),
-                      inputs=[
-        abst,
-    ],
-        outputs=[a_output, j_output, n_output],
-        api_name="neurojane")
-
-demo.launch(debug=True)
+            visible=True,
+            label='Model B',
+            show_label = True
+            scale=1
+            )
+        r_btn = gr.Button(value="Model B is better",scale=1)
+
+    action_btn.click(fn=get_comp,
+        inputs=[prompt,],
+        outputs=[l_output, r_output],
+        api_name="arena")
+    l_btn.click(fn=lambda x,y,z: send_result(x,y,z,'left'),
+                inputs=[l_output,r_output,prompt],
+                api_name="feedleft")
+    l_btn.click(fn=lambda x,y,z: send_result(x,y,z,'right'),
+                inputs=[l_output,r_output,prompt],
+                api_name="feedright")
+
+demo.launch(debug=True)