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

rapidfireai/utils/metric_logger.py

@@ -0,0 +1,346 @@
+"""
+Metric Logger abstraction layer for RapidFire AI.
+
+This module provides a unified interface for logging metrics to different backends
+(MLflow, TensorBoard, or both). This abstraction allows minimal changes to core ML code
+while supporting multiple tracking systems.
+"""
+
+from abc import ABC, abstractmethod
+from pathlib import Path
+from typing import Optional
+
+# Note: MLflowManager is imported lazily in MLflowMetricLogger to avoid
+# connection attempts when using tensorboard-only mode
+
+
+class MetricLogger(ABC):
+    """
+    Abstract base class for metric logging.
+
+    Provides a unified interface for logging metrics, parameters, and managing runs
+    across different tracking backends (MLflow, TensorBoard, etc.).
+    """
+
+    @abstractmethod
+    def create_run(self, run_name: str) -> str:
+        """
+        Create a new run and return run_id.
+
+        Args:
+            run_name: Name for the run
+
+        Returns:
+            Run ID string
+        """
+        pass
+
+    @abstractmethod
+    def log_param(self, run_id: str, key: str, value: str) -> None:
+        """
+        Log a parameter to a specific run.
+
+        Args:
+            run_id: Run identifier
+            key: Parameter name
+            value: Parameter value
+        """
+        pass
+
+    @abstractmethod
+    def log_metric(self, run_id: str, key: str, value: float, step: Optional[int] = None) -> None:
+        """
+        Log a metric to a specific run.
+
+        Args:
+            run_id: Run identifier
+            key: Metric name
+            value: Metric value
+            step: Optional step number for the metric
+        """
+        pass
+
+    @abstractmethod
+    def end_run(self, run_id: str) -> None:
+        """
+        End a specific run.
+
+        Args:
+            run_id: Run identifier
+        """
+        pass
+
+    @abstractmethod
+    def get_run_metrics(self, run_id: str) -> dict:
+        """
+        Get all metrics for a specific run.
+
+        Args:
+            run_id: Run identifier
+
+        Returns:
+            Dictionary of metrics
+        """
+        pass
+
+    def delete_run(self, run_id: str) -> None:
+        """
+        Delete a specific run (optional, not all backends support this).
+
+        Args:
+            run_id: Run identifier
+        """
+        pass
+
+    def clear_context(self) -> None:
+        """Clear the tracking context (optional, not all backends need this)."""
+        pass
+
+
+class MLflowMetricLogger(MetricLogger):
+    """
+    MLflow implementation of MetricLogger.
+
+    Wraps the existing MLflowManager to provide the MetricLogger interface.
+    """
+
+    def __init__(self, tracking_uri: str):
+        """
+        Initialize MLflow metric logger.
+
+        Args:
+            tracking_uri: MLflow tracking server URI
+        """
+        # Lazy import to avoid connection attempts in tensorboard-only mode
+        from rapidfireai.utils.mlflow_manager import MLflowManager
+        self.mlflow_manager = MLflowManager(tracking_uri)
+
+    def get_experiment(self, experiment_name: str) -> str:
+        """
+        Get existing experiment by name.
+
+        Args:
+            experiment_name: Name of the experiment
+
+        Returns:
+            Experiment ID
+        """
+        return self.mlflow_manager.get_experiment(experiment_name)
+
+    def create_run(self, run_name: str) -> str:
+        """Create a new MLflow run."""
+        return self.mlflow_manager.create_run(run_name)
+
+    def log_param(self, run_id: str, key: str, value: str) -> None:
+        """Log a parameter to MLflow."""
+        self.mlflow_manager.log_param(run_id, key, value)
+
+    def log_metric(self, run_id: str, key: str, value: float, step: Optional[int] = None) -> None:
+        """Log a metric to MLflow."""
+        self.mlflow_manager.log_metric(run_id, key, value, step=step)
+
+    def end_run(self, run_id: str) -> None:
+        """End an MLflow run."""
+        self.mlflow_manager.end_run(run_id)
+
+    def get_run_metrics(self, run_id: str) -> dict:
+        """Get metrics from MLflow."""
+        return self.mlflow_manager.get_run_metrics(run_id)
+
+    def delete_run(self, run_id: str) -> None:
+        """Delete an MLflow run."""
+        self.mlflow_manager.delete_run(run_id)
+
+    def clear_context(self) -> None:
+        """Clear MLflow context."""
+        self.mlflow_manager.clear_context()
+
+
+class TensorBoardMetricLogger(MetricLogger):
+    """
+    TensorBoard implementation of MetricLogger.
+
+    Uses torch.utils.tensorboard.SummaryWriter to log metrics to TensorBoard.
+    """
+
+    def __init__(self, log_dir: str):
+        """
+        Initialize TensorBoard metric logger.
+
+        Args:
+            log_dir: Directory for TensorBoard logs
+        """
+        from torch.utils.tensorboard import SummaryWriter
+
+        self.log_dir = Path(log_dir)
+        self.log_dir.mkdir(parents=True, exist_ok=True)
+        self.writers = {}  # Map run_id -> SummaryWriter
+
+    def create_run(self, run_name: str) -> str:
+        """
+        Create a new TensorBoard run.
+
+        For TensorBoard, we use run_name as the run_id and create a subdirectory
+        in the log directory.
+        """
+        from torch.utils.tensorboard import SummaryWriter
+
+        run_log_dir = self.log_dir / run_name
+        run_log_dir.mkdir(parents=True, exist_ok=True)
+
+        # Create SummaryWriter for this run
+        writer = SummaryWriter(log_dir=str(run_log_dir))
+        self.writers[run_name] = writer
+
+        return run_name
+
+    def log_param(self, run_id: str, key: str, value: str) -> None:
+        """
+        Log a parameter to TensorBoard.
+
+        TensorBoard doesn't have native parameter logging, so we log as text.
+        """
+        if run_id not in self.writers:
+            self.create_run(run_id)
+
+        writer = self.writers[run_id]
+        writer.add_text(f"params/{key}", str(value), global_step=0)
+        writer.flush()
+
+    def log_metric(self, run_id: str, key: str, value: float, step: Optional[int] = None) -> None:
+        """
+        Log a metric to TensorBoard.
+
+        Args:
+            run_id: Run identifier
+            key: Metric name
+            value: Metric value
+            step: Step number (required for TensorBoard time series)
+        """
+        if run_id not in self.writers:
+            self.create_run(run_id)
+
+        writer = self.writers[run_id]
+        # Use step=0 if not provided (fallback)
+        writer.add_scalar(key, value, global_step=step if step is not None else 0)
+        # Flush immediately to ensure real-time updates
+        writer.flush()
+
+    def end_run(self, run_id: str) -> None:
+        """End a TensorBoard run by closing the writer."""
+        if run_id in self.writers:
+            self.writers[run_id].close()
+            del self.writers[run_id]
+
+    def get_run_metrics(self, run_id: str) -> dict:
+        """
+        Get metrics from TensorBoard.
+
+        Note: TensorBoard doesn't provide easy API access to logged metrics.
+        This returns an empty dict. For viewing metrics, use TensorBoard UI.
+        """
+        return {}
+
+    def __del__(self):
+        """Clean up all writers on deletion."""
+        for writer in self.writers.values():
+            writer.close()
+
+
+class DualMetricLogger(MetricLogger):
+    """
+    Dual implementation that logs to both MLflow and TensorBoard.
+
+    This allows users to benefit from both tracking systems simultaneously:
+    - MLflow for experiment comparison and model registry
+    - TensorBoard for real-time training visualization (especially useful in Colab)
+    """
+
+    def __init__(self, mlflow_tracking_uri: str, tensorboard_log_dir: str):
+        """
+        Initialize dual metric logger.
+
+        Args:
+            mlflow_tracking_uri: MLflow tracking server URI
+            tensorboard_log_dir: Directory for TensorBoard logs
+        """
+        self.mlflow_logger = MLflowMetricLogger(mlflow_tracking_uri)
+        self.tensorboard_logger = TensorBoardMetricLogger(tensorboard_log_dir)
+
+    def get_experiment(self, experiment_name: str) -> str:
+        """Get experiment from MLflow (TensorBoard doesn't have experiments)."""
+        return self.mlflow_logger.get_experiment(experiment_name)
+
+    def create_run(self, run_name: str) -> str:
+        """Create run in both MLflow and TensorBoard."""
+        mlflow_run_id = self.mlflow_logger.create_run(run_name)
+        self.tensorboard_logger.create_run(run_name)
+        # Return MLflow run_id as the canonical ID
+        return mlflow_run_id
+
+    def log_param(self, run_id: str, key: str, value: str) -> None:
+        """Log parameter to both backends."""
+        self.mlflow_logger.log_param(run_id, key, value)
+        self.tensorboard_logger.log_param(run_id, key, value)
+
+    def log_metric(self, run_id: str, key: str, value: float, step: Optional[int] = None) -> None:
+        """Log metric to both backends."""
+        self.mlflow_logger.log_metric(run_id, key, value, step=step)
+        self.tensorboard_logger.log_metric(run_id, key, value, step=step)
+
+    def end_run(self, run_id: str) -> None:
+        """End run in both backends."""
+        self.mlflow_logger.end_run(run_id)
+        self.tensorboard_logger.end_run(run_id)
+
+    def get_run_metrics(self, run_id: str) -> dict:
+        """Get metrics from MLflow (primary source)."""
+        return self.mlflow_logger.get_run_metrics(run_id)
+
+    def delete_run(self, run_id: str) -> None:
+        """Delete run from MLflow (TensorBoard logs remain on disk)."""
+        self.mlflow_logger.delete_run(run_id)
+
+    def clear_context(self) -> None:
+        """Clear context in both backends."""
+        self.mlflow_logger.clear_context()
+
+
+def create_metric_logger(
+    backend: str,
+    mlflow_tracking_uri: Optional[str] = None,
+    tensorboard_log_dir: Optional[str] = None,
+) -> MetricLogger:
+    """
+    Factory function to create the appropriate metric logger.
+
+    Args:
+        backend: Tracking backend to use ('mlflow', 'tensorboard', or 'both')
+        mlflow_tracking_uri: MLflow tracking server URI (required if backend includes MLflow)
+        tensorboard_log_dir: TensorBoard log directory (required if backend includes TensorBoard)
+
+    Returns:
+        MetricLogger instance
+
+    Raises:
+        ValueError: If backend is invalid or required parameters are missing
+    """
+    backend = backend.lower()
+
+    if backend == "mlflow":
+        if not mlflow_tracking_uri:
+            raise ValueError("mlflow_tracking_uri required for MLflow backend")
+        return MLflowMetricLogger(mlflow_tracking_uri)
+
+    elif backend == "tensorboard":
+        if not tensorboard_log_dir:
+            raise ValueError("tensorboard_log_dir required for TensorBoard backend")
+        return TensorBoardMetricLogger(tensorboard_log_dir)
+
+    elif backend == "both":
+        if not mlflow_tracking_uri or not tensorboard_log_dir:
+            raise ValueError("Both mlflow_tracking_uri and tensorboard_log_dir required for dual backend")
+        return DualMetricLogger(mlflow_tracking_uri, tensorboard_log_dir)
+
+    else:
+        raise ValueError(f"Invalid backend: {backend}. Must be 'mlflow', 'tensorboard', or 'both'")

rapidfireai/utils/mlflow_manager.py

@@ -104,7 +104,6 @@ class MLflowManager:
             current_run = mlflow.active_run()
             if current_run:
                 run_id = current_run.info.run_id
-                print(f"[DEBUG] Active run ID: {run_id}")
 
                 # Try to end the run properly using the client first
                 try:

rapidfireai/utils/ping.py

@@ -1,9 +1,10 @@
 #!/usr/bin/env python
-import socket
 import argparse
+import socket
+
 
 def ping_server(server: str, port: int, timeout=3):
-    """ping server:port """
+    """ping server:port"""
     try:
         socket.setdefaulttimeout(timeout)
         s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
@@ -14,6 +15,7 @@ def ping_server(server: str, port: int, timeout=3):
         s.close()
         return True
 
+
 if __name__ == "__main__":
     parser = argparse.ArgumentParser(description="Ping a server port")
     parser.add_argument("server", type=str, help="Server to ping")

rapidfireai/utils/worker_manager.py

@@ -84,7 +84,14 @@ class WorkerManager:
         while not self.shutdown_event.is_set():
             try:
                 # Check if parent process is still alive
-                os.getpgid(self.parent_pid)
+                try:
+                    os.getpgid(self.parent_pid)
+                except PermissionError:
+                    # Fallback for restricted environments (e.g., Colab)
+                    # Use psutil to check if parent process exists
+                    import psutil
+                    if not psutil.pid_exists(self.parent_pid):
+                        raise ProcessLookupError("Parent process no longer exists")
                 time.sleep(self.parent_check_interval)
             except ProcessLookupError:
                 self.logger.debug(f"Parent process {self.parent_pid} died, shutting down workers...")
@@ -101,11 +108,14 @@ class WorkerManager:
         """
         self.logger.debug(f"Creating {self.num_workers} worker processes...")
 
-        # Create new process group
-        os.setpgrp()
-        self.process_group_id = os.getpgrp()
-
-        self.logger.debug(f"Starting worker processes in process group {self.process_group_id}")
+        # Create new process group (may not be permitted in restricted environments like Colab)
+        try:
+            os.setpgrp()
+            self.process_group_id = os.getpgrp()
+            self.logger.debug(f"Starting worker processes in process group {self.process_group_id}")
+        except PermissionError:
+            self.logger.debug("Cannot create process group (restricted environment) - will use individual process termination")
+            self.process_group_id = None
 
         worker_ids = []
 

rapidfireai/version.py

@@ -2,5 +2,5 @@
 Version information for RapidFire AI
 """
 
-__version__ = "0.10.2rc5"
-__version_info__ = (0, 10, "2rc5") 
+__version__ = "0.11.1rc1"
+__version_info__ = (0, 11, "1rc1") 

{rapidfireai-0.10.2rc5.dist-info → rapidfireai-0.11.1rc1.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: rapidfireai
-Version: 0.10.2rc5
+Version: 0.11.1rc1
 Summary: RapidFire AI: Rapid Experimentation Engine for Customizing LLMs
 Author-email: "RapidFire AI Inc." <support@rapidfire.ai>
 License: Apache-2.0
@@ -22,6 +22,7 @@ Requires-Python: >=3.12
 Description-Content-Type: text/markdown
 License-File: LICENSE
 Requires-Dist: pandas>=2.3.1
+Requires-Dist: psutil>=5.9.0
 Requires-Dist: torch>=2.8.0
 Requires-Dist: transformers>=4.55.2
 Requires-Dist: peft>=0.17.0
@@ -40,11 +41,13 @@ Requires-Dist: requests>=2.32.5
 Requires-Dist: loguru>=0.7.3
 Requires-Dist: ipython>=8.37.0
 Requires-Dist: jupyter>=1.1.1
-Requires-Dist: ipywidgets>=8.1.7
+Requires-Dist: ipywidgets<9.0.0,>=7.3.4
 Requires-Dist: uv>=0.8.14
+Requires-Dist: tensorboard>=2.11.0
 Provides-Extra: dev
 Requires-Dist: pytest>=6.0; extra == "dev"
 Requires-Dist: pytest-cov>=2.0; extra == "dev"
+Requires-Dist: pytest-mock>=3.10.0; extra == "dev"
 Requires-Dist: black>=21.0; extra == "dev"
 Requires-Dist: flake8>=3.8; extra == "dev"
 Requires-Dist: mypy>=0.800; extra == "dev"
@@ -147,7 +150,7 @@ If you encounter port conflicts, you can kill existing processes:
 
 ```bash
 lsof -t -i:5002 | xargs kill -9  # mlflow
-lsof -t -i:8080 | xargs kill -9  # dispatcher
+lsof -t -i:8081 | xargs kill -9  # dispatcher
 lsof -t -i:3000 | xargs kill -9  # frontend server
 ```
 
@@ -292,7 +295,7 @@ chmod +x ./rapidfireai/start_dev.sh
 # VSCode can port-forward localhost:3000 where the rf-frontend server will be running
 
 # for port clash issues -
-lsof -t -i:8080 | xargs kill -9 # dispatcher
+lsof -t -i:8081 | xargs kill -9 # dispatcher
 lsof -t -i:5002 | xargs kill -9 # mlflow
 lsof -t -i:3000 | xargs kill -9 # frontend
 ```

{rapidfireai-0.10.2rc5.dist-info → rapidfireai-0.11.1rc1.dist-info}/RECORD

@@ -1,31 +1,31 @@
 rapidfireai/__init__.py,sha256=mSV8CiaJ9LwjCpMdHSBd9bM-JBijDx-lc8hGny1KEsQ,368
-rapidfireai/cli.py,sha256=j9gjZKw4LqOI5WwSYF4PlKjuN_pyLDXLxvB3fFjhgfI,16048
-rapidfireai/experiment.py,sha256=YvUAzwM3o-gEYECDOmDhbB9szKthfMAw6aiAvJ91gYA,6962
-rapidfireai/start.sh,sha256=o0Mp2EMrZ3lHkTdyCfOqg4JvatN220Kq2VuZNfhFFGg,21268
-rapidfireai/version.py,sha256=c5X_WW-3KMxmWVkp6_-9D0uoIDWyvqJKCuGcRxTcLg8,107
+rapidfireai/cli.py,sha256=EzIqiIVTvXHZA0MP51G3SNbm8fq4hg--lsOg3vlb71U,16766
+rapidfireai/experiment.py,sha256=HTZCDD3VfH-DP7nFmD6xGHy3SpttZ6Wd3-tAZGr0HUU,10612
+rapidfireai/start.sh,sha256=HGl-J-CpkZuwFl5iVMzPMR_Zf0OGJqPfJFmPKigu_80,24384
+rapidfireai/version.py,sha256=r80_4K2SERzBy0aLKlQ2IBxufsANR-PhibCNUprhmqM,107
 rapidfireai/automl/__init__.py,sha256=QnzWa33i9aMp1NatoQYJFPrGZchtTUAPkgSOyyDXbSU,501
 rapidfireai/automl/base.py,sha256=pF6NQMr8DeEFm4PBbmbUbNAtP0S-yDfeUnKMqz2D9Zk,1947
 rapidfireai/automl/datatypes.py,sha256=rbocXidGekpeukKQuMSZLFK6h6h4PIo1Fvre2FWmhqU,1470
-rapidfireai/automl/grid_search.py,sha256=0Drz4xKvNxYKYIjiYSTzHIkgYFhgHW30UjGp3TrORJs,6158
-rapidfireai/automl/model_config.py,sha256=7GwE5fIL2ZDdDXTimvAOnTlUEhu6qQqw4PI1peoBZo0,3776
-rapidfireai/automl/random_search.py,sha256=UBEE-T9a0DUT4a76xHzrNe82qk3LW6ej2jtFDmTPLoU,5689
+rapidfireai/automl/grid_search.py,sha256=S_lxFuSt0MZcztSwzUMg_lDGmneybXZ8VqoUzdGIlfw,6127
+rapidfireai/automl/model_config.py,sha256=eqRbUxcKj-O_Cxk3A1sW4pFMJQgjs_75m81EtJKArQM,3665
+rapidfireai/automl/random_search.py,sha256=VzDHFM-hJ3KBnLuSz6ExOofeEOMJbHkTgMYvXi2dNHg,5324
 rapidfireai/backend/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 rapidfireai/backend/chunks.py,sha256=83n_jz08lxEHi0FZ_xqV-ZQxgzzp3XQdCyW1zvW8Pb0,6981
-rapidfireai/backend/controller.py,sha256=7datz74r1bBs2kL1fRrmVQIi-vIebTiJTpk4MEcZMIs,34002
+rapidfireai/backend/controller.py,sha256=jha417YwPBbzVuSSCDW_s5Wh6NawUCpPxi4k8Q8N7P8,32817
 rapidfireai/backend/scheduler.py,sha256=D_C6XXZ6D7UT8-N5XskeagG2e7Il8DnN3s788y9vc1Y,6026
-rapidfireai/backend/worker.py,sha256=EFoAdWoiwOG1BvG-tZqOypMaml8SFX4p9qxMOvvDen8,12662
+rapidfireai/backend/worker.py,sha256=sCWKB0NrarfIFJVNaNaIbkxrf5_1it9HxbBRAWmQq_0,13106
 rapidfireai/db/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 rapidfireai/db/db_interface.py,sha256=891uxZP00hVnqd8jVf_jL84Ui85BO06JFdYVLzqxaPU,5237
-rapidfireai/db/rf_db.py,sha256=0LmHHaDqZvzjS3k5QpKlFHiJWnCNOIuZwFdmsBHVxR8,26492
-rapidfireai/db/tables.sql,sha256=Qj1WPA2W44jKITBCbox8W7lE-wNhk3k7xr1DyFGAfEc,1939
-rapidfireai/dispatcher/dispatcher.py,sha256=uCpHFOoG2-I1ZRYiNJcZsXcZOqmB2rdppAy-sf2yF70,16026
-rapidfireai/dispatcher/gunicorn.conf.py,sha256=4DHeHZe-dg2IlM4YzIf7j3v3GfB0_I2e7jy1llOtqw0,770
+rapidfireai/db/rf_db.py,sha256=0GptQv0PlS13dUEaxcBfV78xo1XGr8cyxPCSFnFcVTE,26499
+rapidfireai/db/tables.sql,sha256=QRZs9iu4fkEfM9J--Xrdu-ar1zZOUdWmPWrUYL1o9GY,1930
+rapidfireai/dispatcher/dispatcher.py,sha256=kAUZw9eK_c9sfp9LHjbCUVDWO6QWBWjyP887PklqC04,16158
+rapidfireai/dispatcher/gunicorn.conf.py,sha256=CfS1lPfDK5ggl63c5qYtydfNC3TNJ0_irEI0qCL37d0,837
 rapidfireai/frontend/package.json,sha256=sVJmBQ0-1b7RW_h_jD-kKWL3QdhlTmdx12f80nxNd8w,8778
-rapidfireai/frontend/proxy_middleware.py,sha256=FkdCYOXRczUqT1nq0RomlETEc38X3WEYtz8eI_7Lt7U,8537
+rapidfireai/frontend/proxy_middleware.py,sha256=FLOmJbzzEq19HzuJvBmneCI4MKNUWGtNEYZAfCpSCBs,8537
 rapidfireai/frontend/server.py,sha256=uTCzigwy2v14X1g-PHs9PsJ_LlJFmd2HLTi6A0MFfL0,984
-rapidfireai/frontend/build/asset-manifest.json,sha256=oGSMA5Cd9XkO3PaUi_KgEjy7kqK9fzUkTLplzY7qJj0,19906
+rapidfireai/frontend/build/asset-manifest.json,sha256=hUsuVYGmwulDfi8HQx019EokmC87CL_DF3Rm2SgNG2k,19906
 rapidfireai/frontend/build/favicon.ico,sha256=t1wT9kqUZ_H6n5WgN9FTZBoN1Fb_gBBBBb1GJm0j1tg,15406
-rapidfireai/frontend/build/index.html,sha256=gmlC3Qeli0BVxPwMn-qkhJcqVX0amtV2V2eIAWDoRls,707
+rapidfireai/frontend/build/index.html,sha256=u5CofDvF2BkqqnjSJ4VADwpwpve-X9XRs8-e1bQa4WE,707
 rapidfireai/frontend/build/manifest.json,sha256=ShxPBqSqfpMC7XKD10XThODp-eVtCalJ2SuJbLxN6wQ,317
 rapidfireai/frontend/build/lib/notebook-trace-renderer/index.html,sha256=TExHVdUiSmR8VgCyjlxkW4OHk2ZW6XjOMPqSP2ABEcM,205
 rapidfireai/frontend/build/lib/notebook-trace-renderer/js/ml-model-trace-renderer.5490ebc325fe0f300ad9.js,sha256=WmABEAnu39aZ3UPTdoYYkgscNfhidEBbEd6-Q1TF_xY,5096
@@ -230,9 +230,9 @@ rapidfireai/frontend/build/static/js/9831.43bfa1a6.chunk.js,sha256=FgAwC9cUR31aY
 rapidfireai/frontend/build/static/js/9831.43bfa1a6.chunk.js.map,sha256=zqQ06JHJuOM7xYXWVBbvT5vnPNhY5X-h5B7zHysx3lw,11180
 rapidfireai/frontend/build/static/js/experimentPage.230db0b3.chunk.js,sha256=QsBAHGwXMz1fDE0nwtviNM2l6gaiL2ZL3K6kNYDzqos,9322
 rapidfireai/frontend/build/static/js/experimentPage.230db0b3.chunk.js.map,sha256=ZMZJzwksHJ5epKDudSY7UesbhRmg_-QVVzRykc4vt30,40878
-rapidfireai/frontend/build/static/js/main.1bf27639.js,sha256=DX1XbKdgsq0jI4lOzK8AJ6_wSdDrlQbXNlzE3z0X7I0,2605534
-rapidfireai/frontend/build/static/js/main.1bf27639.js.LICENSE.txt,sha256=0aWPNna3cImqZ2xi_GQ1LK0CTZvzZ2zFrD_mwHYMmko,6269
-rapidfireai/frontend/build/static/js/main.1bf27639.js.map,sha256=ucZe7RREf1rLQ4ITuwQuhHo96IkTU66IU8Pr_gDhhEU,11862186
+rapidfireai/frontend/build/static/js/main.58393d31.js,sha256=hk0_IVrF9Qj0q3Vr5UzRPjo7216vbD3AaF4GhH6DoHM,2605534
+rapidfireai/frontend/build/static/js/main.58393d31.js.LICENSE.txt,sha256=0aWPNna3cImqZ2xi_GQ1LK0CTZvzZ2zFrD_mwHYMmko,6269
+rapidfireai/frontend/build/static/js/main.58393d31.js.map,sha256=_oau_iXX3pIJsL0LDLgePZs2jD1atbIV3Tg7YYC_nlM,11862186
 rapidfireai/frontend/build/static/media/404-overflow.fad9a31861b0afba6f921ebb8e769688.svg,sha256=brmh2WLNo6Kp2QZKU5MezMUUB9sScckdhTTkee8FxcU,4637
 rapidfireai/frontend/build/static/media/RapidFire_Square_Bug.27ceb48296314a4bc0d4.png,sha256=2Lyp_ZlHCfOnzC2KoOvV9ahZPLad_XzWb96F1BvWn5I,20682
 rapidfireai/frontend/build/static/media/chart-bar.0fd4a63680fba840a7b69fbf07969f79.svg,sha256=TyChv8IKe0v04YBeS7qz-IFOGjn0YJhjdflz3FxdwRc,498
@@ -257,31 +257,34 @@ rapidfireai/frontend/build/static/media/registered-model-grey-ok.8274b58d39504c8
 rapidfireai/frontend/build/static/media/versions-empty.3e2d8bb8c5b8ddd0a19a34890b36ad41.svg,sha256=qknDzhv6z7j1AONJhHgV5Xsi84J3iNw2gLeE2v1oghA,23148
 rapidfireai/frontend/build/static/media/warning.290a3b14118933547965e91ea61c5a61.svg,sha256=aR7OrJItpEixvhf_sl1vuTsEPZLeizUm2xR12dKxfrI,646
 rapidfireai/ml/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-rapidfireai/ml/callbacks.py,sha256=vi7YUStCNkBoPSYGEuwVkoeRpsEuZ_sSIBi1jfrZZ0E,9733
+rapidfireai/ml/callbacks.py,sha256=RKm_XGOqHw0ShhPIJKstfseaFeK_QRJEWHxITnfkpUg,11236
 rapidfireai/ml/checkpoint_utils.py,sha256=T6BzkvQmL77tvOiif-06oPyM-ikEDsj8ezbNELSzjM0,23422
-rapidfireai/ml/trainer.py,sha256=TPmuHGxTa3DtYML_LQXjzPGYsJwFFL5oy2BZTKwEpQE,15520
+rapidfireai/ml/trainer.py,sha256=PYb4SPgDs5LuYO93upBRSCllELQZSMjOWgrBlqmzYiQ,14985
 rapidfireai/utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 rapidfireai/utils/automl_utils.py,sha256=4IeGZyYRxSdoKk1dBcTI5_JRms70TyiWL9F6Gta31BI,2004
-rapidfireai/utils/constants.py,sha256=-p4FFjlfAieyZDOqE6zokGCmiCuRb1mJYb1aQcwab7c,3332
+rapidfireai/utils/constants.py,sha256=lMQraeikDhlQBjsAp3T_00moxe9wCOZG5m_wqp6sisE,4088
 rapidfireai/utils/datapaths.py,sha256=PKgZu_qWx2z6QBIfmzmjY0lWG79GaU6W3577_34yX10,2554
 rapidfireai/utils/exceptions.py,sha256=RA6kMSV3nCz3oE-yhuNLDEneDqTUrZC6N0AkSRBdAlg,2002
-rapidfireai/utils/experiment_utils.py,sha256=7ow1RGk4dnXOKVnkjcHNSYGjLLlVgPlrvnjt_hq_0Ik,14688
-rapidfireai/utils/logging.py,sha256=X6hLKk4alVUhPqs4CdBmPj4ppSOkQ0WoyczNzCWs02E,3050
-rapidfireai/utils/mlflow_manager.py,sha256=iGuA5ubmhTjhxtZrLCsStpCHBAidnnvONb5LVWZv-RE,5046
-rapidfireai/utils/ping.py,sha256=d8d5Ykx-Tn0HRFeo3xzxwc__KMn2t9FvEd7ur9YLts8,976
+rapidfireai/utils/experiment_utils.py,sha256=ZbKb2argSlCpFpqVoFO5_KwMK9ql1lVme3QZamDPePM,17440
+rapidfireai/utils/interactive_controller.py,sha256=lCpjkDWZ9ixkf8TEu1FTsjIa4T1fRUDomfM7wJgaSnk,17862
+rapidfireai/utils/logging.py,sha256=jq87rfMd0aQMnfh6BvXKTn5WtoZorizS_83uk16lYBo,3026
+rapidfireai/utils/metric_logger.py,sha256=J8PxEWwaRSG5x07e8tz-mqYYRQchNQSTW9lm7UTOYFU,11078
+rapidfireai/utils/mlflow_manager.py,sha256=ocUCLMkA3LDaHc0809DqUmJK8hAmoPFo3qmRQyvmyQk,4988
+rapidfireai/utils/ping.py,sha256=9HJmiIBOc5HtgxSrEWEkbLj08nXHa14Smf-91BmYNNQ,977
 rapidfireai/utils/serialize.py,sha256=_A9egs2uhlYNGT3Ntv2fzH7rwp6I-GGVoS4ViY3sufU,401
 rapidfireai/utils/shm_manager.py,sha256=MlK-lKR-GjQ7G161U65922fh19YqfZmDabfi5P_kCpw,23323
 rapidfireai/utils/trainer_config.py,sha256=xrEcx-jW3NnSb6qCTTQT7OhLO9R9hWvNIHd2AsymQ-g,629
-rapidfireai/utils/worker_manager.py,sha256=LsXnXC2yDwnIp7tm1shpI6DMpif6XGtZ-4kDoo302tk,7971
-rapidfireai-0.10.2rc5.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+rapidfireai/utils/worker_manager.py,sha256=VomHVgn7GgXghX1pVaRdI9HVEnJjbnhpyJT0-IKBUt8,8643
+rapidfireai-0.11.1rc1.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+tutorial_notebooks/rf-colab-tensorboard-tutorial.ipynb,sha256=u82tZDxuJ__ij8lMquIwRSckibTDd1p-3dWS8eWAYwQ,15893
 tutorial_notebooks/rf-tutorial-dpo-alignment-lite.ipynb,sha256=ZEoP6hmBuwWPE5BgJXUAwBgfaAs3-s0Jj3hbcCl6buo,15082
 tutorial_notebooks/rf-tutorial-dpo-alignment.ipynb,sha256=EmHZkmsVUqgonlcNgJfYri4-MK0K_Ad7lkkLQDTGRjs,15698
 tutorial_notebooks/rf-tutorial-grpo-mathreasoning-lite.ipynb,sha256=BXBQTzAdHmugUA0a5_eLK5AWitQaM_fOdfH6V-DnrOY,13111
 tutorial_notebooks/rf-tutorial-grpo-mathreasoning.ipynb,sha256=vL8gdGXqDYl1iGAi2wbPnqU7WcS5pzQG9KEiP4dAWm4,14255
 tutorial_notebooks/rf-tutorial-sft-chatqa-lite.ipynb,sha256=A5nLbJyKDa1cPIBZbZsqHH2MS15Eb9m57zV22486av8,12235
 tutorial_notebooks/rf-tutorial-sft-chatqa.ipynb,sha256=qGMwOZM9VYJVLSz02hl5pfUcwdMNhLLcypBogJ7G3Bo,12140
-rapidfireai-0.10.2rc5.dist-info/METADATA,sha256=ln_7P0MRzkzHUf1p0Is3KMoCjiPHxmUhz4N9WiS9dx0,12489
-rapidfireai-0.10.2rc5.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-rapidfireai-0.10.2rc5.dist-info/entry_points.txt,sha256=tuZF1oC4KyQ9H767o83S8Y-ZiGvw_PVADPL1vRykY3g,53
-rapidfireai-0.10.2rc5.dist-info/top_level.txt,sha256=A28FddyVhe1LHCbvbigLRtmEWKHGVgOVKH1_FfbUQ2U,12
-rapidfireai-0.10.2rc5.dist-info/RECORD,,
+rapidfireai-0.11.1rc1.dist-info/METADATA,sha256=635WC_wVxFUH7seXaYpQQ4Lpu0S_uRapOAan8iq61H0,12611
+rapidfireai-0.11.1rc1.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+rapidfireai-0.11.1rc1.dist-info/entry_points.txt,sha256=tuZF1oC4KyQ9H767o83S8Y-ZiGvw_PVADPL1vRykY3g,53
+rapidfireai-0.11.1rc1.dist-info/top_level.txt,sha256=A28FddyVhe1LHCbvbigLRtmEWKHGVgOVKH1_FfbUQ2U,12
+rapidfireai-0.11.1rc1.dist-info/RECORD,,