Bug Fix: recipe stages were not being concatenated

src/llmcompressor/transformers/sparsification/sparse_model.py

@@ -14,10 +14,14 @@
 from torch.nn import Module
 from transformers import AutoModelForCausalLM, PreTrainedModel
 
+from llmcompressor.pytorch.model_load.helpers import initialize_recipe
 from llmcompressor.transformers.sparsification.compressed_tensors_utils import (
     modify_save_pretrained,
 )
-from llmcompressor.transformers.utils.helpers import download_model_directory
+from llmcompressor.transformers.utils.helpers import (
+    download_model_directory,
+    resolve_recipe,
+)
 
 __all__ = ["SparseAutoModel", "SparseAutoModelForCausalLM", "get_shared_tokenizer_src"]
 
@@ -142,6 +146,10 @@ def skip(*args, **kwargs):
                 compressor.decompress(
                     model_path=pretrained_model_name_or_path, model=model
                 )
+        recipe = resolve_recipe(recipe=recipe, model_path=pretrained_model_name_or_path)
+
+        if recipe:
+            initialize_recipe(model=model, recipe_path=recipe)
 
         return model
 

tests/testing_utils.py

@@ -104,6 +104,7 @@ def _parse_configs_dir(current_config_dir):
                 logging.info(
                     f"Skipping testing model: {file} for cadence: {expected_cadence}"
                 )
+
     if isinstance(configs_directory, list):
         for config in configs_directory:
             _parse_configs_dir(config)