featrixsphere 0.2.1233__py3-none-any.whl → 0.2.1237__py3-none-any.whl

featrixsphere/__init__.py

@@ -38,7 +38,7 @@ Example:
     ...                                labels=['Experiment A', 'Experiment B'])
 """
 
-__version__ = "0.2.1233"
+__version__ = "0.2.1237"
 __author__ = "Featrix"
 __email__ = "support@featrix.com"
 __license__ = "MIT"

featrixsphere/client.py

@@ -4194,93 +4194,126 @@ class FeatrixSphereClient:
         print(f"Training predictor on foundation model {foundation_model_id}...")
         print(f"  Target: {target_column} ({target_column_type})")
         
-        # Validate that only one data source is provided
-        if input_filename and df is not None:
-            raise ValueError("Provide either input_filename or df, not both")
-        
-        # If DataFrame provided, save to temp file and upload it
-        temp_file = None
-        if df is not None:
-            import pandas as pd
-            import tempfile
-            import os
-            
-            if not isinstance(df, pd.DataFrame):
-                raise ValueError("df must be a pandas DataFrame")
-            
+        # Get the compute cluster from the foundation model session
+        # This ensures we upload files to the same node where the foundation model lives
+        foundation_session = self.get_session_status(foundation_model_id)
+        foundation_compute_cluster = self.get_last_server_metadata()
+        foundation_compute_cluster = foundation_compute_cluster.get('compute_cluster') if foundation_compute_cluster else None
+        
+        # Temporarily set compute cluster for file uploads if we found one
+        original_compute_cluster = self.compute_cluster
+        original_headers = self.session.headers.copy()
+        if foundation_compute_cluster:
+            self.set_compute_cluster(foundation_compute_cluster)
             if verbose:
-                print(f"📊 Using provided DataFrame ({len(df)} rows, {len(df.columns)} columns)")
-            
-            # Create temporary CSV file
-            temp_file = tempfile.NamedTemporaryFile(mode='w', suffix='.csv', delete=False)
-            temp_file_path = temp_file.name
-            temp_file.close()
-            
-            # Save DataFrame to temp file
-            df.to_csv(temp_file_path, index=False)
-            
-            if verbose:
-                print(f"📁 Saved to temporary file: {os.path.basename(temp_file_path)}")
-                print(f"📤 Uploading file to server...")
-            
-            # Upload the file
-            uploaded_filename = self.upload_file(temp_file_path)
-            input_filename = uploaded_filename
+                print(f"  Using compute cluster: {foundation_compute_cluster}")
+        
+        try:
+            # Validate that only one data source is provided
+            if input_filename and df is not None:
+                raise ValueError("Provide either input_filename or df, not both")
             
-            if verbose:
-                print(f"✅ File uploaded: {input_filename}")
+            # Prepare multipart form data (like train_single_predictor_with_file does)
+            files = None
+            data = {
+                "foundation_model_id": foundation_model_id,
+                "target_column": target_column,
+                "target_column_type": target_column_type,
+                "epochs": str(epochs),
+                "optimize_for": optimize_for,
+            }
             
-            # Clean up temp file
-            try:
-                os.unlink(temp_file_path)
-            except Exception:
-                pass  # Ignore cleanup errors
-        
-        data = {
-            "foundation_model_id": foundation_model_id,
-            "target_column": target_column,
-            "target_column_type": target_column_type,
-            "epochs": epochs,
-            "optimize_for": optimize_for,
-        }
-        
-        if input_filename:
-            # If absolute path provided, upload the file first
-            from pathlib import Path
-            input_path = Path(input_filename)
-            if input_path.is_absolute():
-                # Upload the file first, then use the uploaded filename
-                if not input_path.exists():
-                    raise FileNotFoundError(f"Input file not found: {input_filename}")
+            # Handle file upload - send file directly in multipart form
+            if df is not None:
+                import pandas as pd
+                import tempfile
+                import os
+                
+                if not isinstance(df, pd.DataFrame):
+                    raise ValueError("df must be a pandas DataFrame")
                 
                 if verbose:
-                    print(f"📤 Uploading file from absolute path: {input_filename}")
+                    print(f"📊 Using provided DataFrame ({len(df)} rows, {len(df.columns)} columns)")
+                
+                # Create temporary CSV file
+                temp_file = tempfile.NamedTemporaryFile(mode='w', suffix='.csv', delete=False)
+                temp_file_path = temp_file.name
+                temp_file.close()
                 
-                # Upload the file
-                uploaded_filename = self.upload_file(str(input_path))
+                # Save DataFrame to temp file
+                df.to_csv(temp_file_path, index=False)
                 
                 if verbose:
-                    print(f"✅ File uploaded as: {uploaded_filename}")
+                    print(f"📁 Saved to temporary file: {os.path.basename(temp_file_path)}")
+                    print(f"📤 Uploading file directly with training request...")
                 
-                data["input_filename"] = uploaded_filename
-            else:
-                # Relative filename - assume it's already on the server
-                data["input_filename"] = input_filename
-        if name:
-            data["name"] = name
-        if session_name_prefix:
-            data["session_name_prefix"] = session_name_prefix
-        if rare_label_value:
-            data["rare_label_value"] = rare_label_value
-        if class_imbalance:
-            data["class_imbalance"] = class_imbalance
-        if webhooks:
-            data["webhooks"] = webhooks
-        
-        response_data = self._post_json("/compute/train_on_foundational_model", data)
-        
-        new_session_id = response_data.get('session_id')
-        print(f"✅ Predictor training session created: {new_session_id}")
+                # Send file in multipart form
+                files = {'file': (os.path.basename(temp_file_path), open(temp_file_path, 'rb'), 'text/csv')}
+                
+            elif input_filename:
+                # If absolute path provided, send file directly
+                from pathlib import Path
+                input_path = Path(input_filename)
+                if input_path.is_absolute():
+                    if not input_path.exists():
+                        raise FileNotFoundError(f"Input file not found: {input_filename}")
+                    
+                    if verbose:
+                        print(f"📤 Sending file directly from absolute path: {input_filename}")
+                    
+                    # Send file in multipart form
+                    files = {'file': (input_path.name, open(input_path, 'rb'), 'text/csv' if input_path.suffix == '.csv' else 'application/gzip')}
+                else:
+                    # Relative filename - assume it's already on the server
+                    data["input_filename"] = input_filename
+            
+            if name:
+                data["name"] = name
+            if session_name_prefix:
+                data["session_name_prefix"] = session_name_prefix
+            if rare_label_value:
+                data["rare_label_value"] = rare_label_value
+            if class_imbalance:
+                import json
+                data["class_imbalance"] = json.dumps(class_imbalance)
+            if webhooks:
+                import json
+                data["webhooks"] = json.dumps(webhooks)
+            
+            # Send request with file if provided
+            try:
+                if files:
+                    response = self._make_request("POST", "/compute/train_on_foundational_model", files=files, data=data)
+                else:
+                    response = self._make_request("POST", "/compute/train_on_foundational_model", json=data)
+                response_data = response.json()
+            finally:
+                # Close file handles
+                if files and 'file' in files:
+                    files['file'][1].close()
+                    # Clean up temp file if we created one
+                    if df is not None and temp_file_path:
+                        try:
+                            os.unlink(temp_file_path)
+                        except Exception:
+                            pass
+            
+            new_session_id = response_data.get('session_id')
+            print(f"✅ Predictor training session created: {new_session_id}")
+            
+            # Restore original compute cluster setting
+            if original_compute_cluster != self.compute_cluster:
+                if original_compute_cluster:
+                    self.set_compute_cluster(original_compute_cluster)
+                else:
+                    self.session.headers = original_headers
+        finally:
+            # Ensure we restore headers even if there's an error
+            if original_compute_cluster != self.compute_cluster:
+                if original_compute_cluster:
+                    self.set_compute_cluster(original_compute_cluster)
+                else:
+                    self.session.headers = original_headers
         
         if verbose:
             print(f"⏳ Waiting for training to complete...")

{featrixsphere-0.2.1233.dist-info → featrixsphere-0.2.1237.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: featrixsphere
-Version: 0.2.1233
+Version: 0.2.1237
 Summary: Transform any CSV into a production-ready ML model in minutes, not months.
 Home-page: https://github.com/Featrix/sphere
 Author: Featrix

featrixsphere-0.2.1237.dist-info/RECORD

@@ -0,0 +1,9 @@
+featrixsphere/__init__.py,sha256=B4omX54pm9GR-ascatiSaHJPIOziXObJlkcE1U5Lte0,1888
+featrixsphere/cli.py,sha256=AW9O3vCvCNJ2UxVGN66eRmeN7XLSiHJlvK6JLZ9UJXc,13358
+featrixsphere/client.py,sha256=5BgOKD0lC1Tsnkj1NsKkx_QMdvQJM3Qn8F6DfQjlCQ8,384584
+featrixsphere/test_client.py,sha256=4SiRbib0ms3poK0UpnUv4G0HFQSzidF3Iswo_J2cjLk,11981
+featrixsphere-0.2.1237.dist-info/METADATA,sha256=UOQ8XXBXkiZzIvATWl8X1xIkcGM8soxDYfs9IzY5u7k,16232
+featrixsphere-0.2.1237.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+featrixsphere-0.2.1237.dist-info/entry_points.txt,sha256=QreJeYfD_VWvbEqPmMXZ3pqqlFlJ1qZb-NtqnyhEldc,51
+featrixsphere-0.2.1237.dist-info/top_level.txt,sha256=AyN4wjfzlD0hWnDieuEHX0KckphIk_aC73XCG4df5uU,14
+featrixsphere-0.2.1237.dist-info/RECORD,,

featrixsphere-0.2.1233.dist-info/RECORD

@@ -1,9 +0,0 @@
-featrixsphere/__init__.py,sha256=Iwvv27mlpZjhl7DSUg6kz9Qa24K1S-o5aa-QqwgwcFA,1888
-featrixsphere/cli.py,sha256=AW9O3vCvCNJ2UxVGN66eRmeN7XLSiHJlvK6JLZ9UJXc,13358
-featrixsphere/client.py,sha256=xjYFie7KxepDRp8j8MA0divHyhC_t7fLnjgQUClV0_c,382285
-featrixsphere/test_client.py,sha256=4SiRbib0ms3poK0UpnUv4G0HFQSzidF3Iswo_J2cjLk,11981
-featrixsphere-0.2.1233.dist-info/METADATA,sha256=lQDkfyCT4NMnCnob9Vhgv77YfKrH96XNSoaP0IFpsZM,16232
-featrixsphere-0.2.1233.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-featrixsphere-0.2.1233.dist-info/entry_points.txt,sha256=QreJeYfD_VWvbEqPmMXZ3pqqlFlJ1qZb-NtqnyhEldc,51
-featrixsphere-0.2.1233.dist-info/top_level.txt,sha256=AyN4wjfzlD0hWnDieuEHX0KckphIk_aC73XCG4df5uU,14
-featrixsphere-0.2.1233.dist-info/RECORD,,