fixed the query_engine tests

scripts/backup.jac

@@ -5,24 +5,23 @@ import:py zipfile;
 import:py datetime;
 
 glob current_time = datetime.datetime.now().strftime('%Y-%m-%d_%H-%M-%S');
+
 glob GDRIVE_FOLDER_ID = os.environ['GDRIVE_FOLDER_ID'];
 
-can zipdir(folders:list[str], out:str) {
+can zipdir(folders: list[str], out: str) {
     with zipfile.ZipFile(out, 'w', zipfile.ZIP_DEFLATED) as zipf {
         for folder in folders {
             for (root, _, files) in os.walk(folder) {
                 for file in files {
-                    zipf.write(os.path.join(root, file), 
-                        os.path.relpath(os.path.join(root, file), 
-                        os.path.join(folder, '..')));
+                    zipf.write(os.path.join(root, file), os.path.relpath(os.path.join(root, file), os.path.join(folder, '..')));
                 }
             }
         }
     }
 }
 
 can upload_file_to_drive(file: str, folder_id: str) {
-    file_drive = drive.CreateFile({'parents': [{'id': folder_id}]});
+    file_drive = drive.CreateFile({'parents':[{'id':folder_id }] });
     file_drive.SetContentFile(file);
     file_drive.Upload();
 }
@@ -31,8 +30,7 @@ with entry {
     gauth = GoogleAuth();
     gauth.LocalWebserverAuth();
     drive = GoogleDrive(gauth);
-
     zipdir([".human_eval_config", "results", "data"], f"{current_time}.zip");
     upload_file_to_drive(f"{current_time}.zip", GDRIVE_FOLDER_ID);
     os.remove(f"{current_time}.zip");
-}
+}

scripts/runs2data.jac

@@ -3,46 +3,39 @@ import:py os;
 import:py argparse;
 
 can convert_run(run: str, prompt_disc: str) {
-    responses_files = [f for f in os.listdir(os.path.join("runs", run)) if f.endswith(".json")];
-    data = {
-        "run": run,
-        "prompt_disc": prompt_disc,
-        "outputs": {}
-    };
-
+    responses_files = [f  for f in os.listdir(os.path.join("runs", run)) if f.endswith(".json")];
+    data = {"run":run, "prompt_disc":prompt_disc, "outputs":{} };
+
     for responses_file in responses_files {
         with open(os.path.join("runs", run, responses_file), "r") as f {
             model_data = json.load(f);
             model_name = model_data["model_name"];
             prompt = model_data["full_prompt"];
-            responses = [r["response"] for r in model_data["outputs"]];
+            responses = [r["response"]  for r in model_data["outputs"]];
             data["outputs"][model_name] = responses;
             data["prompt"] = prompt if "prompt" not in data else data["prompt"];
         }
     }
-
     with open(os.path.join("data", f"{run}_responses"), "w") as f {
         json.dump(data, f, indent=4);
     }
 }
 
-
 with entry {
     parser = argparse.ArgumentParser(description="Convert the runs to data");
     parser.add_argument("-r", "--run_name", default="all", help="The run to convert");
     args = parser.parse_args();
-
     os.makedirs("data", exist_ok=True);
-    runs = [f for f in os.listdir("runs") if os.path.isdir(os.path.join("runs", f))];
+    runs = [f  for f in os.listdir("runs") if os.path.isdir(os.path.join("runs", f))];
     if args.run == "all" {
         for run in runs {
             prompt_disc_md = input(f"Provide the relative path to the prompt description markdown (.md) for run");
             prompt_disc = open(prompt_disc_md, "r").read();
             convert_run(run, prompt_disc);
         }
-    } else { 
+    } else {
         prompt_disc_md = input(f"Provide the relative path to the prompt description markdown (.md) for run");
         prompt_disc = open(prompt_disc_md, "r").read();
         convert_run(args.run, prompt_disc);
     }
-}
+}

src/app.jac

@@ -1,5 +1,4 @@
 import:py streamlit as st;
-
 import:jac from components.utils, login;
 import:jac from components.generator, generator;
 import:jac from components.setup, setup;
@@ -10,20 +9,34 @@ import:jac from components.about, about;
 import:jac from components.theme, footer;
 
 can main {
-    if "admin_privileges" not in st.session_state {st.session_state.admin_privileges = False;} 
+    if "admin_privileges" not in st.session_state {
+        st.session_state.admin_privileges = False;
+    }
     (human_eval_tab, admin_tab) = st.tabs(["Human Evaluation", "Admin Panel"]);
-    with human_eval_tab {human_eval();}
+    with human_eval_tab {
+        human_eval();
+    }
     with admin_tab {
-        if not st.session_state.admin_privileges {login();}
-        else {
-            (dashboard_tab, generator_tab, evaluation_tab, setup_tab, about_tab) = st.tabs(["Dashboard",
-                                        "Response Generator", "Auto Evaluator", "Human Eval Setup", "About"]);
-            with dashboard_tab {dashboard();}
-            with generator_tab {generator();}
-            with evaluation_tab { auto_eval();}
-            with setup_tab {setup();}
-            with about_tab {about();}
+        if not st.session_state.admin_privileges {
+            login();
+        } else {
+            (dashboard_tab, generator_tab, evaluation_tab, setup_tab, about_tab) = st.tabs(["Dashboard", "Response Generator", "Auto Evaluator", "Human Eval Setup", "About"]);
+            with dashboard_tab {
+                dashboard();
+            }
+            with generator_tab {
+                generator();
+            }
+            with evaluation_tab {
+                auto_eval();
+            }
+            with setup_tab {
+                setup();
+            }
+            with about_tab {
+                about();
+            }
         }
     }
     footer();
-}
+}

src/components/about.jac

@@ -9,4 +9,4 @@ can about {
     st.markdown(citation);
     st.header("Contacts");
     st.markdown(socials_mk);
-}
+}

src/components/auto_eval.jac

@@ -1,12 +1,17 @@
 import:py streamlit as st;
-
 import:jac from emb_sim_scorer, emb_sim_scorer;
 import:jac from model_ranking, model_win_percentage_table;
 import:jac from llm_as_evaluator, llm_as_evaluator;
 
 can auto_eval {
-    (gpt_evaluator, sim_scorer_tab,model_ranking) = st.tabs(["LLM as Evaluator", "Similarity Scorer", "Model Ranking"]);
-    with sim_scorer_tab {emb_sim_scorer();}
-    with model_ranking {model_win_percentage_table();}
-    with gpt_evaluator {llm_as_evaluator();}
-}
+    (gpt_evaluator, sim_scorer_tab, model_ranking) = st.tabs(["LLM as Evaluator", "Similarity Scorer", "Model Ranking"]);
+    with sim_scorer_tab {
+        emb_sim_scorer();
+    }
+    with model_ranking {
+        model_win_percentage_table();
+    }
+    with gpt_evaluator {
+        llm_as_evaluator();
+    }
+}