[deploy] 0.1.6.1 support setting timeouts

Chenglong-MS · web-flow · commit 6e91164a4f49 · 2025-02-21T16:05:21.000-08:00
Dev: provide ability to set timeout options, and improved backend logging so that users can provide feedback
diff --git a/py-src/data_formulator/agents/agent_data_rec.py b/py-src/data_formulator/agents/agent_data_rec.py
@@ -181,6 +181,14 @@ def process_gpt_response(self, input_tables, messages, response):
             result['refined_goal'] = refined_goal
             candidates.append(result)
 
+        logger.info("=== Recommendation Candidates ===>")
+        for candidate in candidates:
+            for key, value in candidate.items():
+                if key in ['dialog', 'content']:
+                    logger.info(f"##{key}:\n{str(value)[:1000]}...")
+                else:
+                    logger.info(f"## {key}:\n{value}")
+
         return candidates
     
 
diff --git a/py-src/data_formulator/agents/agent_data_transform_v2.py b/py-src/data_formulator/agents/agent_data_transform_v2.py
@@ -224,16 +224,9 @@ def process_gpt_response(self, input_tables, messages, response):
 
             code_blocks = extract_code_from_gpt_response(choice.message.content + "\n", "python")
 
-            logger.info("=== Code blocks ===>")
-            logger.info(code_blocks)
-
             if len(code_blocks) > 0:
                 code_str = code_blocks[-1]
 
-                for table in input_tables:
-                    logger.info(f"Table: {table['name']}")
-                    logger.info(table['rows'])
-
                 try:
                     result = py_sandbox.run_transform_in_sandbox2020(code_str, [t['rows'] for t in input_tables])
                     result['code'] = code_str
@@ -256,8 +249,13 @@ def process_gpt_response(self, input_tables, messages, response):
             result['refined_goal'] = refined_goal
             candidates.append(result)
 
-        logger.info("=== Candidates ===>")
-        logger.info(candidates)
+        logger.info("=== Transform Candidates ===>")
+        for candidate in candidates:
+            for key, value in candidate.items():
+                if key in ['dialog', 'content']:
+                    logger.info(f"##{key}:\n{str(value)[:1000]}...")
+                else:
+                    logger.info(f"## {key}:\n{value}")
 
         return candidates
 
diff --git a/py-src/data_formulator/app.py b/py-src/data_formulator/app.py
@@ -213,8 +213,8 @@ def test_model():
         content = request.get_json()
 
         # contains endpoint, key, model, api_base, api_version
-        print("content------------------------------")
-        print(content)
+        logger.info("content------------------------------")
+        logger.info(content)
 
         client = get_client(content['model'])
 
@@ -226,8 +226,8 @@ def test_model():
                 ]
             )
 
-            print(f"model: {content['model']}")
-            print(f"welcome message: {response.choices[0].message.content}")
+            logger.info(f"model: {content['model']}")
+            logger.info(f"welcome message: {response.choices[0].message.content}")
 
             if "I can hear you." in response.choices[0].message.content:
                 result = {
@@ -236,7 +236,7 @@ def test_model():
                     "message": ""
                 }
         except Exception as e:
-            print(f"Error: {e}")
+            logger.info(f"Error: {e}")
             error_message = str(e)
             result = {
                 "model": content['model'],
@@ -250,13 +250,13 @@ def test_model():
 
 @app.route("/", defaults={"path": ""})
 def index_alt(path):
-    print(app.static_folder)
+    logger.info(app.static_folder)
     return send_from_directory(app.static_folder, "index.html")
 
 @app.errorhandler(404)
 def page_not_found(e):
     # your processing here
-    print(app.static_folder)
+    logger.info(app.static_folder)
     return send_from_directory(app.static_folder, "index.html") #'Hello 404!' #send_from_directory(app.static_folder, "index.html")
 
 ###### test functions ######
@@ -425,14 +425,21 @@ def derive_data():
         new_fields = content["new_fields"]
         instruction = content["extra_prompt"]
 
+        max_repair_attempts = content["max_repair_attempts"] if "max_repair_attempts" in content else 1
+
         if "additional_messages" in content:
             prev_messages = content["additional_messages"]
         else:
             prev_messages = []
 
-        print("spec------------------------------")
-        print(new_fields)
-        print(instruction)
+        logger.info("== input tables ===>")
+        for table in input_tables:
+            logger.info(f"===> Table: {table['name']} (first 5 rows)")
+            logger.info(table['rows'][:5])
+
+        logger.info("== user spec ===")
+        logger.info(new_fields)
+        logger.info(instruction)
 
         mode = "transform"
         if len(new_fields) == 0:
@@ -447,7 +454,7 @@ def derive_data():
             results = agent.run(input_tables, instruction, [field['name'] for field in new_fields], prev_messages)
 
         repair_attempts = 0
-        while results[0]['status'] == 'error' and repair_attempts == 0: # only try once
+        while results[0]['status'] == 'error' and repair_attempts < max_repair_attempts: # try up to n times
             error_message = results[0]['content']
             new_instruction = f"We run into the following problem executing the code, please fix it:\n\n{error_message}\n\nPlease think step by step, reflect why the error happens and fix the code so that no more errors would occur."
 
@@ -482,16 +489,23 @@ def refine_data():
         output_fields = content["output_fields"]
         dialog = content["dialog"]
         new_instruction = content["new_instruction"]
+        max_repair_attempts = content["max_repair_attempts"] if "max_repair_attempts" in content else 1
+        
+        logger.info("== input tables ===>")
+        for table in input_tables:
+            logger.info(f"===> Table: {table['name']} (first 5 rows)")
+            logger.info(table['rows'][:5])
         
-        print("previous dialog")
-        print(dialog)
+        logger.info("== user spec ===>")
+        logger.info(output_fields)
+        logger.info(new_instruction)
 
         # always resort to the data transform agent       
         agent = DataTransformationAgentV2(client=client)
         results = agent.followup(input_tables, dialog, [field['name'] for field in output_fields], new_instruction)
 
         repair_attempts = 0
-        while results[0]['status'] == 'error' and repair_attempts == 0: # only try once
+        while results[0]['status'] == 'error' and repair_attempts < max_repair_attempts: # only try once
             error_message = results[0]['content']
             new_instruction = f"We run into the following problem executing the code, please fix it:\n\n{error_message}\n\nPlease think step by step, reflect why the error happens and fix the code so that no more errors would occur."
             prev_dialog = results[0]['dialog']
diff --git a/pyproject.toml b/pyproject.toml
@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "data_formulator"
-version = "0.1.6"
+version = "0.1.6.1"
 
 requires-python = ">=3.9"
 authors = [
diff --git a/src/app/App.tsx b/src/app/App.tsx
@@ -35,6 +35,7 @@ import {
     ToggleButton,
     Menu,
     MenuItem,
+    TextField,
 } from '@mui/material';
 
 
@@ -46,7 +47,7 @@ import { DataFormulatorFC } from '../views/DataFormulator';
 
 import GridViewIcon from '@mui/icons-material/GridView';
 import ViewSidebarIcon from '@mui/icons-material/ViewSidebar';
-
+import SettingsIcon from '@mui/icons-material/Settings';
 import {
     createBrowserRouter,
     RouterProvider,
@@ -295,10 +296,122 @@ const ResetDialog: React.FC = () => {
     );
 };
 
+const ConfigDialog: React.FC = () => {
+    const [open, setOpen] = useState(false);
+    const dispatch = useDispatch();
+    const config = useSelector((state: DataFormulatorState) => state.config);
+
+    const [formulateTimeoutSeconds, setFormulateTimeoutSeconds] = useState(config.formulateTimeoutSeconds);
+    const [maxRepairAttempts, setMaxRepairAttempts] = useState(config.maxRepairAttempts);
+
+    // Add check for changes
+    const hasChanges = formulateTimeoutSeconds !== config.formulateTimeoutSeconds || 
+                      maxRepairAttempts !== config.maxRepairAttempts;
+
+    return (
+        <>
+            <Button variant="text" sx={{textTransform: 'none'}} onClick={() => setOpen(true)} startIcon={<SettingsIcon />}>
+                 <Box component="span" sx={{lineHeight: 1.2, display: 'flex', flexDirection: 'column', alignItems: 'left'}}>
+                    <Box component="span" sx={{py: 0, my: 0, fontSize: '10px', mr: 'auto'}}>timeout={config.formulateTimeoutSeconds}s</Box>
+                    <Box component="span" sx={{py: 0, my: 0, fontSize: '10px', mr: 'auto'}}>max_repair={config.maxRepairAttempts}</Box>
+                </Box>
+            </Button>
+            <Dialog onClose={() => setOpen(false)} open={open}>
+                <DialogTitle>Data Formulator Configuration</DialogTitle>
+                <DialogContent>
+                    <DialogContentText>
+                        <Box sx={{ 
+                            display: 'flex', 
+                            flexDirection: 'column', 
+                            gap: 3,
+                            my: 2,
+                            maxWidth: 400
+                        }}>
+                            <Box sx={{ display: 'flex', alignItems: 'center', gap: 2 }}>
+                                <Box sx={{ flex: 1 }}>
+                                    <TextField
+                                        label="formulate timeout (seconds)"
+                                        type="number"
+                                        variant="outlined"
+                                        value={formulateTimeoutSeconds}
+                                        onChange={(e) => {
+                                            const value = parseInt(e.target.value);
+                                            setFormulateTimeoutSeconds(value);
+                                        }}
+                                        inputProps={{
+                                            min: 0,
+                                            max: 3600,
+                                        }}
+                                        error={formulateTimeoutSeconds <= 0 || formulateTimeoutSeconds > 3600}
+                                        helperText={formulateTimeoutSeconds <= 0 || formulateTimeoutSeconds > 3600 ? 
+                                            "Value must be between 1 and 3600 seconds" : ""}
+                                        fullWidth
+                                    />
+                                    <Typography variant="caption" color="text.secondary" sx={{ mt: 1, display: 'block' }}>
+                                        Maximum time allowed for the formulation process before timing out. 
+                                    </Typography>
+                                    <Typography variant="caption" color="text.secondary" sx={{ mt: 1, display: 'block' }}>
+                                        Smaller values (&lt;30s) make the model fails fast thus providing a smoother UI experience. Increase this value for slow models.
+                                    </Typography>
+                                </Box>
+                            </Box>
+                            <Box sx={{ display: 'flex', alignItems: 'center', gap: 2 }}>
+                                <Box sx={{ flex: 1 }}>
+                                    <TextField
+                                        label="max repair attempts"
+                                        type="number"
+                                        variant="outlined"
+                                        value={maxRepairAttempts}
+                                        onChange={(e) => {
+                                            const value = parseInt(e.target.value);
+                                            setMaxRepairAttempts(value);
+                                        }}
+                                        fullWidth
+                                        inputProps={{
+                                            min: 1,
+                                            max: 5,
+                                        }}
+                                        error={maxRepairAttempts <= 0 || maxRepairAttempts > 5}
+                                        helperText={maxRepairAttempts <= 0 || maxRepairAttempts > 5 ? 
+                                            "Value must be between 1 and 5" : ""}
+                                    />
+                                    <Typography variant="caption" color="text.secondary" sx={{ mt: 1, display: 'block' }}>
+                                        Maximum number of times the LLM will attempt to repair code if generated code fails to execute (recommended = 1).
+                                    </Typography>
+                                    <Typography variant="caption" color="text.secondary" sx={{ mt: 1, display: 'block' }}>
+                                        Higher values might slightly increase the chance of success but can crash the backend. Repair time is as part of the formulate timeout.
+                                    </Typography>
+                                </Box>
+                            </Box>
+                        </Box>
+                    </DialogContentText>
+                </DialogContent>
+                <DialogActions sx={{'.MuiButton-root': {textTransform: 'none'}}}>
+                    <Button sx={{marginRight: 'auto'}} onClick={() => {
+                        setFormulateTimeoutSeconds(30);
+                        setMaxRepairAttempts(1);
+                    }}>Reset to default</Button>
+                    <Button onClick={() => setOpen(false)}>Cancel</Button>
+                    <Button 
+                        variant={hasChanges ? "contained" : "text"}
+                        disabled={!hasChanges || isNaN(maxRepairAttempts) || maxRepairAttempts <= 0 || maxRepairAttempts > 5 || isNaN(formulateTimeoutSeconds) || formulateTimeoutSeconds <= 0 || formulateTimeoutSeconds > 3600}
+                        onClick={() => {
+                            dispatch(dfActions.setConfig({formulateTimeoutSeconds, maxRepairAttempts}));
+                            setOpen(false);
+                        }}
+                    >
+                        Apply
+                    </Button>
+                </DialogActions>
+            </Dialog>
+        </>
+    );  
+}
+
 export const AppFC: FC<AppFCProps> = function AppFC(appProps) {
 
     const visViewMode = useSelector((state: DataFormulatorState) => state.visViewMode);
-    const betaMode = useSelector((state: DataFormulatorState) => state.betaMode);
+    const config = useSelector((state: DataFormulatorState) => state.config);
     const tables = useSelector((state: DataFormulatorState) => state.tables);
 
     // if the user has logged in
@@ -410,7 +523,7 @@ export const AppFC: FC<AppFCProps> = function AppFC(appProps) {
 
     let appBar = [
         <AppBar className="app-bar" position="static" key="app-bar-main">
-            <Toolbar variant="dense" sx={{ backgroundColor: betaMode ? 'lavender' : '' }}>
+            <Toolbar variant="dense">
                 <Button href={"/"} sx={{
                     display: "flex", flexDirection: "row", textTransform: "none",
                     backgroundColor: 'transparent',
@@ -420,7 +533,7 @@ export const AppFC: FC<AppFCProps> = function AppFC(appProps) {
                 }} color="inherit">
                     <Box component="img" sx={{ height: 32, marginRight: "12px" }} alt="" src={dfLogo} />
                     <Typography variant="h6" noWrap component="h1" sx={{ fontWeight: 300, display: { xs: 'none', sm: 'block' } }}>
-                        {toolName} {betaMode ? "β" : ""} {process.env.NODE_ENV == "development" ? "" : ""}
+                        {toolName} {process.env.NODE_ENV == "development" ? "" : ""}
                     </Typography>
                 </Button>
                 <Box sx={{ flexGrow: 1, textAlign: 'center', display: 'flex', justifyContent: 'center' }} >
@@ -432,6 +545,8 @@ export const AppFC: FC<AppFCProps> = function AppFC(appProps) {
                         about
                     </Button>
                     <Divider orientation="vertical" variant="middle" flexItem /> */}
+                    <ConfigDialog />
+                    <Divider orientation="vertical" variant="middle" flexItem />
                     <ModelSelectionButton />
                     <Divider orientation="vertical" variant="middle" flexItem />
                     <Typography sx={{ display: 'flex', fontSize: 14, alignItems: 'center', gap: 1 }}>
diff --git a/src/app/dfSlice.tsx b/src/app/dfSlice.tsx
@@ -65,7 +65,10 @@ export interface DataFormulatorState {
 
     chartSynthesisInProgress: string[];
 
-    betaMode: boolean;
+    config: {
+        formulateTimeoutSeconds: number;
+        maxRepairAttempts: number;
+    }
 }
 
 // Define the initial state using that type
@@ -98,7 +101,10 @@ const initialState: DataFormulatorState = {
 
     chartSynthesisInProgress: [],
 
-    betaMode: false,
+    config: {
+        formulateTimeoutSeconds: 30,
+        maxRepairAttempts: 1,
+    }
 }
 
 let getUnrefedDerivedTableIds = (state: DataFormulatorState) => {
@@ -247,7 +253,11 @@ export const dataFormulatorSlice = createSlice({
 
             state.chartSynthesisInProgress = [];
 
-            state.betaMode = false;
+            // avoid resetting config
+            // state.config = {
+            //     formulateTimeoutSeconds: 30,
+            //     repairAttempts: 1,
+            // }
         },
         loadState: (state, action: PayloadAction<any>) => {
 
@@ -274,10 +284,10 @@ export const dataFormulatorSlice = createSlice({
 
             state.chartSynthesisInProgress = [];
 
-            state.betaMode = savedState.betaMode;
+            state.config = savedState.config;
         },
-        toggleBetaMode: (state, action: PayloadAction<boolean>) => {
-            state.betaMode = action.payload;
+        setConfig: (state, action: PayloadAction<{formulateTimeoutSeconds: number, maxRepairAttempts: number}>) => {
+            state.config = action.payload;
         },
         selectModel: (state, action: PayloadAction<string | undefined>) => {
             state.selectedModelId = action.payload;
diff --git a/src/views/EncodingShelfCard.tsx b/src/views/EncodingShelfCard.tsx
diff --git a/src/views/EncodingShelfThread.tsx b/src/views/EncodingShelfThread.tsx
diff --git a/src/views/ModelSelectionDialog.tsx b/src/views/ModelSelectionDialog.tsx
