Add tinyllama_awq_backup_mode_none_backend_OV test

nncf/quantization/algorithms/weight_compression/algorithm.py

@@ -277,7 +277,7 @@ def _get_bitwidth_distribution_str(
         dtype_vs_num_weights_map = {}
         ratio_defining_weight_names = set(wp.weight_name for wp in ratio_defining_params)
         for data in all_params:
-            dtype = data.compression_config.mode if data.compression_config is not None else "fp16/fp32"
+            dtype = data.compression_config.mode if data.compression_config is not None else "float"
             n_total, n_ratio_defining = dtype_vs_num_weights_map.get(dtype, ([], []))
             if data.weight_name in ratio_defining_weight_names:
                 n_ratio_defining.append(data.num_weights)
@@ -313,12 +313,8 @@ def apply(
         dataset: Optional[Dataset] = None,
     ) -> TModel:
         self._set_backend_entity(model)
-        # nodes_to_compress includes nodes from the ignored scope to be added to bitwidth_distribution_str
-        nodes_to_compress = self._get_nodes_to_compress(graph)
-
-        activations = {}
-        if dataset is not None and self._sensitivity_metric != SensitivityMetric.WEIGHT_QUANTIZATION_ERROR:
-            activations = self._get_activations(dataset, self._subset_size, nodes_to_compress, graph, model)
+        # candidates_to_compress includes nodes from the ignored scope to be added to bitwidth_distribution_str
+        candidates_to_compress = self._get_nodes_to_compress(graph)
         all_weight_params: List[WeightCompressionParameters] = []
         weight_names = set()
 
@@ -328,8 +324,8 @@ def apply(
         ignored_scope_weight_params: List[WeightCompressionParameters] = []
 
         is_last_layer_shared = False
-        n = len(nodes_to_compress)
-        for i, node in enumerate(nodes_to_compress):
+        n = len(candidates_to_compress)
+        for i, node in enumerate(candidates_to_compress):
             for weight_name, weight_port_id in self._backend_entity.get_weight_names_and_port_ids(node, graph):
                 if weight_name in weight_names:
                     if i == n - 1:
@@ -383,15 +379,23 @@ def apply(
                 all_weight_params.append(weight_params)
                 weight_names.add(weight_name)
 
+        activations = {}
+        nodes_to_compress = [node for node in candidates_to_compress if node.node_name not in ignored_names]
+        if dataset is not None and self._sensitivity_metric != SensitivityMetric.WEIGHT_QUANTIZATION_ERROR:
+            activations = self._get_activations(dataset, self._subset_size, nodes_to_compress, graph, model)
+
         ratio_defining_params = self._get_ratio_defining_params(all_weight_params, is_last_layer_shared)
         self._set_weight_compression_config(ratio_defining_params, model, graph, activations)
         nncf_logger.info(
             self._get_bitwidth_distribution_str(all_weight_params + ignored_scope_weight_params, ratio_defining_params)
         )
+        nodes_names_to_exclude = {
+            w_params.node_with_weight.node_name for w_params in all_weight_params if w_params.compression_config is None
+        }
         # Filter the weight parameters that should remain in their original floating-point precision
         all_weight_params = [w_params for w_params in all_weight_params if w_params.compression_config is not None]
-        # Remove nodes in the ignored scope from nodes_to_compress
-        nodes_to_compress = [node for node in nodes_to_compress if node.node_name not in ignored_names]
+        # Filter nodes_to_compress by excluding nodes that should remain in their original floating-point precision
+        nodes_to_compress = [node for node in nodes_to_compress if node.node_name not in nodes_names_to_exclude]
 
         if self._awq and activations is not None and self._mode != CompressWeightsMode.E2M1:
             awq_params = self._advanced_parameters.awq_params

tests/post_training/data/wc_reference_data.yaml

@@ -39,4 +39,8 @@ tinyllama_data_aware_lora_stateful_backend_OV:
 tinyllama_NF4_scale_estimation_stateful_per_channel_backend_OV:
   metric_value: 0.88663
   num_int4: 11
-  num_int8: 290
+  num_int8: 290
+tinyllama_awq_backup_mode_none_backend_OV:
+  metric_value: 0.85679
+  num_int4: 208
+  num_int8: 0

tests/post_training/model_scope.py

@@ -15,6 +15,7 @@
 import nncf
 from nncf import ModelType
 from nncf import QuantizationPreset
+from nncf.parameters import BackupMode
 from nncf.parameters import CompressWeightsMode
 from nncf.parameters import SensitivityMetric
 from nncf.quantization.advanced_parameters import AdvancedCompressionParameters
@@ -498,6 +499,21 @@
         "params": {"is_stateful": True},
         "backends": [BackendType.OV],
     },
+    {
+        "reported_name": "tinyllama_awq_backup_mode_none",
+        "model_id": "tinyllama/tinyllama-1.1b-step-50k-105b",
+        "pipeline_cls": LMWeightCompression,
+        "compression_params": {
+            "group_size": 64,
+            "ratio": 0.8,
+            "all_layers": True,
+            "backup_mode": BackupMode.NONE,
+            "mode": CompressWeightsMode.INT4_ASYM,
+            "awq": True,
+            "ignored_scope": nncf.IgnoredScope(types=["Gather"]),
+        },
+        "backends": [BackendType.OV],
+    },
 ]