Control trace cache warnings (#7039)

Make trace cache warnings configurable, and disabled by default. 

Fix #6985, #4081, #5033, #5006, #5662

---------

Signed-off-by: Olatunji Ruwase <olruwase@microsoft.com>

deepspeed/runtime/engine.py

@@ -983,6 +983,9 @@ class DeepSpeedEngine(Module):
     def zeropp_loco_param(self):
         return self._config.zero_config.zeropp_loco_param
 
+    def zero_log_trace_cache_warnings(self):
+        return self._config.zero_config.log_trace_cache_warnings
+
     def dump_state(self):
         return self._config.dump_state
 
@@ -1692,6 +1695,7 @@ class DeepSpeedEngine(Module):
                     zero_quantized_weights=self.zero_quantized_weights(),
                     zero_quantized_nontrainable_weights=self.zero_quantized_nontrainable_weights(),
                     zero_module_granularity_threshold=self.zero_module_granularity_threshold(),
+                    log_trace_cache_warnings=self.zero_log_trace_cache_warnings(),
                 )
             else:
                 log_dist(
@@ -1740,6 +1744,7 @@ class DeepSpeedEngine(Module):
                     zero_quantized_nontrainable_weights=self.zero_quantized_nontrainable_weights(),
                     zero_module_granularity_threshold=self.zero_module_granularity_threshold(),
                     zeropp_loco_param=self.zeropp_loco_param(),
+                    log_trace_cache_warnings=self.zero_log_trace_cache_warnings(),
                 )
 
         else:

deepspeed/runtime/zero/config.py

@@ -45,6 +45,7 @@ ZeRO optimization should be enabled as:
     "memory_efficient_linear": [true|false],
     "override_module_apply": [true|false],
     "zeropp_loco_param": {...},
+    "log_trace_cache_warnings" : [true|false],
     }
 }
 """
@@ -340,6 +341,11 @@ class DeepSpeedZeroConfig(DeepSpeedConfigModel):
     Override nn.Module apply function, for Stage 3.
     """
 
+    log_trace_cache_warnings: bool = False
+    """
+    Whether to log warnings from trace cache, such as invalidation events.
+    """
+
     # Validators
     @model_validator(mode="after")
     def overlap_comm_valid(self):

deepspeed/runtime/zero/parameter_offload.py

@@ -103,6 +103,7 @@ class DeepSpeedZeRoOffload(object):
         zero_quantized_weights=False,
         zero_quantized_nontrainable_weights=False,
         zero_module_granularity_threshold=0,
+        log_trace_cache_warnings=False,
     ):
 
         see_memory_usage("DeepSpeedZeRoOffload initialize [begin]", force=True)
@@ -118,6 +119,7 @@ class DeepSpeedZeRoOffload(object):
         self.zero_param_parallel_group = zero_param_parallel_group
         self.zero_quantized_weights = zero_quantized_weights
         self.zero_quantized_nontrainable_weights = zero_quantized_nontrainable_weights
+        self.log_trace_cache_warnings = log_trace_cache_warnings
 
         if offload_param_config is not None and offload_param_config.device != OffloadDeviceEnum.none:
             self.offload_device = offload_param_config.device
@@ -165,7 +167,9 @@ class DeepSpeedZeRoOffload(object):
             timers=self.timers,
             zero_quantized_weights=self.zero_quantized_weights,
             zero_quantized_nontrainable_weights=self.zero_quantized_nontrainable_weights,
-            fast_sharding_for_leaf_module=self.fast_sharding_for_leaf_module)
+            fast_sharding_for_leaf_module=self.fast_sharding_for_leaf_module,
+            log_trace_cache_warnings=self.log_trace_cache_warnings,
+        )
 
         self.forward_hooks = []
         self.backward_hooks = []

deepspeed/runtime/zero/partitioned_param_coordinator.py

@@ -76,7 +76,8 @@ class PartitionedParameterCoordinator:
         param: Parameter
         step_id_last_used_at: int
 
-    def __init__(self,
+    def __init__(
+        self,
         prefetch_bucket_sz: int,
         max_reuse_distance_in_numel: int,
         max_available_parameters_in_numel: int,
@@ -86,7 +87,9 @@ class PartitionedParameterCoordinator:
         timers=None,
         zero_quantized_weights=False,
         zero_quantized_nontrainable_weights=False,
-                 fast_sharding_for_leaf_module=False) -> None:
+        fast_sharding_for_leaf_module=False,
+        log_trace_cache_warnings=False,
+    ) -> None:
         # mapping of param -> handle for each param that is currently in flight
         self.__inflight_param_registry = inflight_param_registry
         # keeps track of the number of submodules invoked so far.
@@ -129,6 +132,9 @@ class PartitionedParameterCoordinator:
         self.__max_ongoing_fetch_events: int = 2
         self.__profiler = PartitionedParameterProfiler(timers if ENABLE_PROFILER else None)
 
+        # Whether to log trace cache warnings, e.g. invalidation events
+        self.__log_trace_cache_warnings = log_trace_cache_warnings
+
         # whether to enable fast fetch for the z3 leaf module.
         # this will improve fetch speed but will not break down leaf module parameters to alleviate memory pressure.
         self.fast_sharding_for_leaf_module = fast_sharding_for_leaf_module
@@ -177,7 +183,7 @@ class PartitionedParameterCoordinator:
                 print_rank_0(
                     f"Invalidate trace cache @ step {self.__step_id} and module {sub_module.ds_id}: "
                     f"cache has only {len(self.__submodule_order)} modules",
-                    force=True)
+                    force=self.__log_trace_cache_warnings)
                 self._invalidate_trace()
                 return
 
@@ -186,7 +192,7 @@ class PartitionedParameterCoordinator:
                 print_rank_0(
                     f"Invalidate trace cache @ step {self.__step_id}: "
                     f"expected module {expected_module_id}, but got module {sub_module.ds_id}",
-                    force=True)
+                    force=self.__log_trace_cache_warnings)
                 self._invalidate_trace()
 
     @compiler.disable

deepspeed/runtime/zero/stage3.py

@@ -160,6 +160,7 @@ class DeepSpeedZeroOptimizer_Stage3(ZeROOptimizer):
         zero_quantized_nontrainable_weights=False,
         zero_module_granularity_threshold=0,
         zeropp_loco_param=None,
+        log_trace_cache_warnings=False,
     ):
         see_memory_usage("Stage 3 initialize beginning", force=True)
 
@@ -231,7 +232,9 @@ class DeepSpeedZeroOptimizer_Stage3(ZeROOptimizer):
             zero_param_parallel_group=zero_param_parallel_group,
             zero_quantized_weights=zero_quantized_weights,
             zero_quantized_nontrainable_weights=zero_quantized_nontrainable_weights,
-            zero_module_granularity_threshold=zero_module_granularity_threshold)
+            zero_module_granularity_threshold=zero_module_granularity_threshold,
+            log_trace_cache_warnings=log_trace_cache_warnings,
+        )
 
         self.persistent_parameters = self.parameter_offload.persistent_parameters
         self._configure_offloading(offload_optimizer_config, offload_param_config)
@@ -465,6 +468,7 @@ class DeepSpeedZeroOptimizer_Stage3(ZeROOptimizer):
         zero_quantized_weights,
         zero_quantized_nontrainable_weights,
         zero_module_granularity_threshold,
+        log_trace_cache_warnings,
     ):
         return DeepSpeedZeRoOffload(module=module,
                                     timers=timers,
@@ -481,7 +485,8 @@ class DeepSpeedZeroOptimizer_Stage3(ZeROOptimizer):
                                     zero_param_parallel_group=zero_param_parallel_group,
                                     zero_quantized_weights=zero_quantized_weights,
                                     zero_quantized_nontrainable_weights=zero_quantized_nontrainable_weights,
-                                    zero_module_granularity_threshold=zero_module_granularity_threshold)
+                                    zero_module_granularity_threshold=zero_module_granularity_threshold,
+                                    log_trace_cache_warnings=log_trace_cache_warnings)
 
     def _get_trainable_parameter_groups(self):
         param_groups = []

docs/_pages/config-json.md

@@ -371,11 +371,12 @@ Enabling and configuring ZeRO memory optimizations
     "sub_group_size" : 1e12,
     "elastic_checkpoint" : [true|false],
     "stage3_gather_16bit_weights_on_model_save": [true|false],
-    "ignore_unused_parameters": [true|false]
-    "round_robin_gradients": [true|false]
-    "zero_hpz_partition_size": 1
-    "zero_quantized_weights": [true|false]
-    "zero_quantized_gradients": [true|false]
+    "ignore_unused_parameters": [true|false],
+    "round_robin_gradients": [true|false],
+    "zero_hpz_partition_size": 1,
+    "zero_quantized_weights": [true|false],
+    "zero_quantized_gradients": [true|false],
+    "log_trace_cache_warnings": [true|false],
     }
 ```
 
@@ -512,6 +513,12 @@ Enabling and configuring ZeRO memory optimizations
 | ----------------------------------------------------------------------------------------------------------------------------------- | ------- |
 |Boolean indicating whether to enable communication efficient quantized gradients of ZeRO++. | `False`   |
 
+<i>**log_trace_cache_warnings**</i>: [boolean]
+
+| Description                                                                                                         | Default |
+| ------------------------------------------------------------------------------------------------------------------- | ------- |
+| Log warnings from trace cache optimization of parameter sharding, such as cache invalidation events. | `False`  |
+
 ***cpu_offload***: [boolean]
 
 **Deprecated:** **cpu_offload** is deprecated and will be removed in future, please use `offload_optimizer` instead.