refactor: added new profiling setup to the profiling tutorial's config

Author: le1nux

tutorials/profiling/configs/distributed_8B_model_profiling.yaml

@@ -6,7 +6,26 @@ settings:
   benchmark:
     sequence_length: 4096
     vocab_size: 50304
-    batch_size: 2
+    batch_size: 1
+  paths: 
+    experiment_root_path: ${modalities_env:config_folder_path}
+
+profiler:
+  component_key: steppable_profiler
+  variant_key: kernel_tracing
+  config:
+    num_wait_steps: 5
+    num_warmup_steps: 5
+    num_active_steps: 3
+    profiler_activities: [CPU, CUDA]
+    record_shapes: false
+    profile_memory: false
+    with_stack: false
+    with_flops: false
+    with_modules: false
+    tracked_ranks: [0, 1]
+    output_folder_path: ${settings.paths.experiment_root_path}/kernel_traces
+    
 
 steppable_component:
   component_key: steppable_component