rapidfireai 0.10.2rc5__py3-none-any.whl → 0.11.1rc1__py3-none-any.whl

rapidfireai/experiment.py

@@ -17,9 +17,11 @@ from rapidfireai.utils.constants import MLFLOW_URL
 from rapidfireai.utils.exceptions import ExperimentException
 from rapidfireai.utils.experiment_utils import ExperimentUtils
 from rapidfireai.utils.logging import RFLogger
-from rapidfireai.utils.mlflow_manager import MLflowManager
 from rapidfireai.version import __version__
 
+# Note: MLflowManager is imported lazily in get_results() to avoid
+# connection attempts when using tensorboard-only mode
+
 
 class Experiment:
     """Class to manage the entire experiment lifecycle."""
@@ -39,6 +41,7 @@ class Experiment:
         self.experiment_id: int | None = None
         self.log_server_process: mp.Process | None = None
         self.worker_processes: list[mp.Process] = []
+        self._training_thread: Any = None  # Track background training thread (Colab only)
 
         # create db tables
         try:
@@ -88,13 +91,76 @@ class Experiment:
         seed: int = 42,
     ) -> None:
         """Run the fit"""
+
+        # Check if training is already running
+        if self._training_thread is not None and self._training_thread.is_alive():
+            print("⚠️  Training is already running in background. Please wait for it to complete.")
+            return
+
+        # Detect if running in Google Colab
         try:
-            controller = Controller(self.experiment_id, self.experiment_name)
-            controller.run_fit(param_config, create_model_fn, train_dataset, eval_dataset, num_chunks, seed)
-        except Exception as e:
-            if hasattr(self, "logger"):
-                self.logger.opt(exception=True).error(f"Error running fit: {e}")
-            raise ExperimentException(f"Error running fit: {e}, traceback: {traceback.format_exc()}") from e
+            import google.colab
+
+            in_colab = True
+        except ImportError:
+            in_colab = False
+
+        if in_colab:
+            # Run Controller in background thread to keep kernel responsive
+            import sys
+            import threading
+            from io import StringIO
+
+            from IPython.display import HTML, display
+
+            def _run_controller_background():
+                """Run controller in background thread with output suppression"""
+                # Suppress stdout to avoid print statements appearing in wrong cells
+                old_stdout = sys.stdout
+                sys.stdout = StringIO()
+
+                try:
+                    controller = Controller(self.experiment_id, self.experiment_name)
+                    controller.run_fit(param_config, create_model_fn, train_dataset, eval_dataset, num_chunks, seed)
+                except Exception as e:
+                    # Restore stdout for error logging
+                    sys.stdout = old_stdout
+                    if hasattr(self, "logger"):
+                        self.logger.opt(exception=True).error(f"Error in background training: {e}")
+                    display(HTML(f'<p style="color: red; font-weight: bold;">❌ Error in background training: {e}</p>'))
+                finally:
+                    # Restore stdout
+                    sys.stdout = old_stdout
+                    # Display completion message
+                    display(
+                        HTML(
+                            '<p style="color: blue; font-weight: bold;">🎉 Training completed! Check InteractiveController for final results.</p>'
+                        )
+                    )
+                    self._training_thread = None
+
+            self._training_thread = threading.Thread(target=_run_controller_background, daemon=True)
+            self._training_thread.start()
+
+            # Use IPython display for reliable output in Colab
+            display(
+                HTML(
+                    '<div style="padding: 10px; background-color: #d4edda; border: 1px solid #28a745; border-radius: 5px; color: #155724;">'
+                    "<b>✓ Training started in background</b><br>"
+                    "Use InteractiveController to monitor progress. The notebook kernel will remain responsive while training runs.<br>"
+                    "<small>Tip: Interact with InteractiveController periodically to keep Colab active.</small>"
+                    "</div>"
+                )
+            )
+        else:
+            # Original blocking behavior for non-Colab environments
+            try:
+                controller = Controller(self.experiment_id, self.experiment_name)
+                controller.run_fit(param_config, create_model_fn, train_dataset, eval_dataset, num_chunks, seed)
+            except Exception as e:
+                if hasattr(self, "logger"):
+                    self.logger.opt(exception=True).error(f"Error running fit: {e}")
+                raise ExperimentException(f"Error running fit: {e}, traceback: {traceback.format_exc()}") from e
 
     def get_results(self) -> pd.DataFrame:
         """
@@ -102,6 +168,19 @@ class Experiment:
         """
         try:
             runs_info_df = self.experiment_utils.get_runs_info()
+
+            # Check if there are any mlflow_run_ids before importing MLflow
+            has_mlflow_runs = (
+                runs_info_df.get("mlflow_run_id") is not None and runs_info_df["mlflow_run_id"].notna().any()
+            )
+
+            if not has_mlflow_runs:
+                # No MLflow runs to fetch, return empty DataFrame
+                return pd.DataFrame(columns=["run_id", "step"])
+
+            # Lazy import - only import when we actually have MLflow runs to fetch
+            from rapidfireai.utils.mlflow_manager import MLflowManager
+
             mlflow_manager = MLflowManager(MLFLOW_URL)
 
             metrics_data = []

rapidfireai/frontend/build/asset-manifest.json

@@ -1,7 +1,7 @@
 {
   "files": {
     "main.css": "/static-files/static/css/main.702595df.css",
-    "main.js": "/static-files/static/js/main.1bf27639.js",
+    "main.js": "/static-files/static/js/main.58393d31.js",
     "ml-model-trace-renderer.js": "/static-files/lib/notebook-trace-renderer/js/ml-model-trace-renderer.5490ebc325fe0f300ad9.js",
     "static/js/6019.9025341e.chunk.js": "/static-files/static/js/6019.9025341e.chunk.js",
     "static/js/6336.8153bc1c.chunk.js": "/static-files/static/js/6336.8153bc1c.chunk.js",
@@ -120,7 +120,7 @@
     "static/media/chart-line.svg": "/static-files/static/media/chart-line.0adaa2036bb4eb5956db6d0c7e925a3d.svg",
     "lib/notebook-trace-renderer/index.html": "/static-files/lib/notebook-trace-renderer/index.html",
     "main.702595df.css.map": "/static-files/static/css/main.702595df.css.map",
-    "main.1bf27639.js.map": "/static-files/static/js/main.1bf27639.js.map",
+    "main.58393d31.js.map": "/static-files/static/js/main.58393d31.js.map",
     "ml-model-trace-renderer.js.map": "/static-files/lib/notebook-trace-renderer/js/ml-model-trace-renderer.5490ebc325fe0f300ad9.js.map",
     "6336.8153bc1c.chunk.js.map": "/static-files/static/js/6336.8153bc1c.chunk.js.map",
     "9478.cbf55ef3.chunk.js.map": "/static-files/static/js/9478.cbf55ef3.chunk.js.map",
@@ -216,6 +216,6 @@
   },
   "entrypoints": [
     "static/css/main.702595df.css",
-    "static/js/main.1bf27639.js"
+    "static/js/main.58393d31.js"
   ]
 }

rapidfireai/frontend/build/index.html

@@ -1 +1 @@
-<!doctype html><html lang="en"><head><meta charset="utf-8"/><meta name="viewport" content="width=device-width,initial-scale=1,shrink-to-fit=no"/><link rel="shortcut icon" href="./static-files/favicon.ico"/><meta name="theme-color" content="#000000"/><link rel="manifest" href="./static-files/manifest.json" crossorigin="use-credentials"/><title>RapidFire AI</title><script defer="defer" src="static-files/static/js/main.1bf27639.js"></script><link href="static-files/static/css/main.702595df.css" rel="stylesheet"></head><body><noscript>You need to enable JavaScript to run this app.</noscript><div id="root" class="mlflow-ui-container"></div><div id="modal" class="mlflow-ui-container"></div></body></html>
+<!doctype html><html lang="en"><head><meta charset="utf-8"/><meta name="viewport" content="width=device-width,initial-scale=1,shrink-to-fit=no"/><link rel="shortcut icon" href="./static-files/favicon.ico"/><meta name="theme-color" content="#000000"/><link rel="manifest" href="./static-files/manifest.json" crossorigin="use-credentials"/><title>RapidFire AI</title><script defer="defer" src="static-files/static/js/main.58393d31.js"></script><link href="static-files/static/css/main.702595df.css" rel="stylesheet"></head><body><noscript>You need to enable JavaScript to run this app.</noscript><div id="root" class="mlflow-ui-container"></div><div id="modal" class="mlflow-ui-container"></div></body></html>