ecmwf · anaprietonem · Nov 4, 2024 · Aug 20, 2024 · Aug 20, 2024 · Aug 20, 2024
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -21,6 +21,8 @@ Keep it human-readable, your future self will thank you!
    - Fix that applies the metric_ranges in the post-processed variable space [#116](https://github.com/ecmwf/anemoi-training/pull/116)
 - Sub-hour datasets [#63](https://github.com/ecmwf/anemoi-training/pull/63)
 - Add synchronisation workflow [#92](https://github.com/ecmwf/anemoi-training/pull/92)
+- Feat: Anemoi Profiler compatible with mlflow and using Pytorch (Kineto) Profiler for memory report [38](https://github.com/ecmwf/anemoi-training/pull/38/)
+
 
 ### Changed
 - Modified training configuration to support max_steps and tied lr iterations to max_steps by default [#67](https://github.com/ecmwf/anemoi-training/pull/67)
@@ -32,6 +34,8 @@ Keep it human-readable, your future self will thank you!
 
 - Lock python version <3.13 [#107](https://github.com/ecmwf/anemoi-training/pull/107)
 
+
+
 ## [0.2.1 - Bugfix: resuming mlflow runs](https://github.com/ecmwf/anemoi-training/compare/0.2.0...0.2.1) - 2024-10-24
 
 ### Added
@@ -82,6 +86,7 @@ Keep it human-readable, your future self will thank you!
 - Feature: `AnemoiMlflowClient`, an mlflow client with authentication support [#86](https://github.com/ecmwf/anemoi-training/pull/86)
 - Long Rollout Plots
 
+
 ### Fixed
 
 - Fix `TypeError` raised when trying to JSON serialise `datetime.timedelta` object - [#43](https://github.com/ecmwf/anemoi-training/pull/43)

diff --git a/docs/images/profiler/anemoi_profiler_architecture.png b/docs/images/profiler/anemoi_profiler_architecture.png
diff --git a/docs/images/profiler/anemoi_profiler_benchmark_profiler.png b/docs/images/profiler/anemoi_profiler_benchmark_profiler.png
diff --git a/docs/images/profiler/anemoi_profiler_config.png b/docs/images/profiler/anemoi_profiler_config.png
diff --git a/docs/images/profiler/anemoi_profiler_high_level.png b/docs/images/profiler/anemoi_profiler_high_level.png
diff --git a/docs/images/profiler/anemoi_profiler_mlflow_integration.png b/docs/images/profiler/anemoi_profiler_mlflow_integration.png
diff --git a/docs/images/profiler/anemoi_profiler_mlflow_integration_2.png b/docs/images/profiler/anemoi_profiler_mlflow_integration_2.png
diff --git a/docs/images/profiler/anemoi_profiler_mlflow_integration_3.png b/docs/images/profiler/anemoi_profiler_mlflow_integration_3.png
diff --git a/docs/images/profiler/anemoi_profiler_speed_report.png b/docs/images/profiler/anemoi_profiler_speed_report.png
diff --git a/docs/images/profiler/anemoi_profiler_speedreport_diagram.png b/docs/images/profiler/anemoi_profiler_speedreport_diagram.png
diff --git a/docs/images/profiler/anemoi_profiler_training_rates.png b/docs/images/profiler/anemoi_profiler_training_rates.png
diff --git a/docs/images/profiler/anemoi_profiler_validation_rates.png b/docs/images/profiler/anemoi_profiler_validation_rates.png
diff --git a/docs/images/profiler/example_memory_report.png b/docs/images/profiler/example_memory_report.png
diff --git a/docs/images/profiler/example_memory_timeline.png b/docs/images/profiler/example_memory_timeline.png
diff --git a/docs/images/profiler/example_model_summary.png b/docs/images/profiler/example_model_summary.png
diff --git a/docs/images/profiler/example_model_summary_2.png b/docs/images/profiler/example_model_summary_2.png
diff --git a/docs/images/profiler/example_system_report.png b/docs/images/profiler/example_system_report.png
diff --git a/docs/images/profiler/example_time_report.png b/docs/images/profiler/example_time_report.png
diff --git a/docs/images/profiler/idle_time_breakdown.png b/docs/images/profiler/idle_time_breakdown.png
diff --git a/docs/images/profiler/kernel_breakdown_dfs.png b/docs/images/profiler/kernel_breakdown_dfs.png
diff --git a/docs/images/profiler/kernel_breakdown_plots.png b/docs/images/profiler/kernel_breakdown_plots.png
diff --git a/docs/images/profiler/memory_snapshot_diagram.png b/docs/images/profiler/memory_snapshot_diagram.png
diff --git a/docs/images/profiler/memory_snapshot_output.png b/docs/images/profiler/memory_snapshot_output.png
diff --git a/docs/images/profiler/temporal_breakdown.png b/docs/images/profiler/temporal_breakdown.png
diff --git a/docs/index.rst b/docs/index.rst
@@ -43,6 +43,7 @@ This package provides the *Anemoi* training functionality.
    user-guide/training
    user-guide/models
    user-guide/tracking
+   user-guide/benchmarking
    user-guide/distributed
    user-guide/debugging
 

diff --git a/docs/overview.rst b/docs/overview.rst
@@ -91,6 +91,18 @@ and resolve issues during the training process, including:
 -  Debug configurations for quick error identification
 -  Guidance on isolating and addressing common problems
 
+8. Benchmarking and HPC Profiling
+=================================
+
+Anemoi Training offers tools and configurations to support benchmarking
+and High-Performance Computing (HPC) profiling, allowing users to
+optimize training performance. This includes:
+
+-  Benchmarking configurations for evaluating training efficiency across
+   different hardware setups.
+-  Profiling tools for monitoring resource utilization (CPU, GPU,
+   memory) and identifying performance bottlenecks.
+
 **************************
  Components and Structure
 **************************