Add safetensors support (#4659)

Co-authored-by: Logan Adams <114770087+loadams@users.noreply.github.com>
Co-authored-by: Jeff Rasley <jerasley@microsoft.com>
Co-authored-by: Michael Wyatt <michaelwyatt@microsoft.com>

deepspeed/inference/v2/checkpoint/huggingface_engine.py

@@ -8,6 +8,7 @@
 import torch
 from .base_engine import CheckpointEngineBase
 from typing import Iterable, Tuple
+from functools import partial
 
 from ..logging import inference_logger
 
@@ -28,6 +29,7 @@ def __init__(self, model_name_or_path: str, auth_token: str = None) -> None:
         else:
             self.model_config.max_seq_length = self.generation_config.max_length
 
+        self._local_checkpoint_dir = None
         self._all_ckpt_paths = self._fetch_checkpoint_files()
 
     def _fetch_checkpoint_files(self):
@@ -41,29 +43,53 @@ def _fetch_checkpoint_files(self):
         # NOTE(jeff): allow_patterns here are explicitly not using safetensors or other
         # checkpoint files that may be present. Example of all files in the llama-2-7b
         # repo here: https://huggingface.co/meta-llama/Llama-2-7b-hf/tree/main
-        from huggingface_hub import snapshot_download
+        from huggingface_hub import snapshot_download, list_files_info
+
+        def model_has_safetensors(model_name_or_path: str) -> bool:
+            if os.path.isdir(model_name_or_path):
+                file_list = os.listdir(model_name_or_path)
+            else:
+                file_list = [rf.rfilename for rf in list_files_info(model_name_or_path)]
+            for f in file_list:
+                if f.endswith(".safetensors"):
+                    return True
+            return False
 
         if os.path.isdir(self.model_name_or_path):
             self._local_checkpoint_dir = self.model_name_or_path
         else:
+            # We need to download the checkpoint files from HF
+            if model_has_safetensors(self.model_name_or_path):
+                # Prioritize downloading safetensors if they are available
+                allow_patterns = ["*.safetensors", "*.json", "*.pt"]
+            else:
+                # Fallback to bin files when safetensors are not present
+                allow_patterns = ["*.bin", "*.json", "*.pt"]
             self._local_checkpoint_dir = snapshot_download(self.model_name_or_path,
-                                                           allow_patterns=[
-                                                               "*.bin",
-                                                               "*.json",
-                                                               "*.pt",
-                                                           ],
+                                                           allow_patterns=allow_patterns,
                                                            revision=None,
                                                            token=self.auth_token)
 
         assert os.path.isdir(
             self._local_checkpoint_dir
         ), f"Checkpoint dir {self._local_checkpoint_dir} is not a directory, cannot load checkpoint."
 
-        model_param_json = os.path.join(self._local_checkpoint_dir, "pytorch_model.bin.index.json")
+        # Set the appropriate file names based on whether we have safetensors or not
+        if model_has_safetensors(self._local_checkpoint_dir):
+            from safetensors.torch import load_file
+            model_param_json_fname = "model.safetensors.index.json"
+            model_file_fname = "model.safetensors"
+            self._checkpoint_load_fn = load_file
+        else:
+            model_param_json_fname = "pytorch_model.bin.index.json"
+            model_file_fname = "pytorch_model.bin"
+            self._checkpoint_load_fn = partial(torch.load, map_location="cpu")
+
+        model_param_json = os.path.join(self._local_checkpoint_dir, model_param_json_fname)
 
         if not os.path.isfile(model_param_json):
             # We don't need any json as all such HF models will have pytorch_model.bin
-            all_checkpoint_files = [os.path.join(self._local_checkpoint_dir, 'pytorch_model.bin')]
+            all_checkpoint_files = [os.path.join(self._local_checkpoint_dir, model_file_fname)]
         else:
             param_map = json.load(open(model_param_json, "r"))
 
@@ -84,7 +110,7 @@ def parameters(self) -> Iterable[Tuple[str, torch.Tensor]]:
         """
         for checkpoint in self._all_ckpt_paths:
             inference_logger().info(f"Loading checkpoint: {checkpoint}")
-            checkpoint_sd = torch.load(checkpoint, map_location='cpu')
+            checkpoint_sd = self._checkpoint_load_fn(checkpoint)
             param_keys = list(checkpoint_sd.keys())
             for param_name in param_keys:
                 param = checkpoint_sd[param_name]