[Tool]: Update tools/convert2llama.py to support `safetensors` format (#730)

tools/convert2llama.py

@@ -9,6 +9,28 @@ from tqdm import tqdm
 from transformers import AutoConfig, LlamaConfig, LlamaTokenizer
 
 
+def weight_load(fp, **kwargs):
+    """Load weights from a file."""
+    is_safetensors = kwargs.pop('is_safetensors', False)
+
+    if is_safetensors:
+        try:
+            from safetensors import safe_open
+        except ImportError:
+            raise ImportError(
+                'Before loading ckpts in the `safetensors` format, '
+                'please install the `safetensors` package first.')
+
+        model = safe_open(fp, framework='pt')
+        state_dict = {}
+        for k in model.keys():
+            state_dict[k] = model.get_tensor(k)
+        return state_dict
+
+    else:
+        return torch.load(fp, **kwargs)
+
+
 def save_conifg(config, tgt):
     config_dict = config.to_dict()
     unnecessary_keys = [
@@ -41,19 +63,29 @@ def convert(src, tgt):
         // config.num_key_value_heads
 
     # load index json file
-    index_file = os.path.join(src, 'pytorch_model.bin.index.json')
-    if os.path.exists(index_file):
-        with open(index_file) as fp:
+    index_file = 'pytorch_model.bin.index.json'
+    if os.path.exists(os.path.join(src, index_file)):
+        with open(os.path.join(src, index_file)) as fp:
             index_dict = json.load(fp)
             index_dict['weight_map'] = {}
     else:
-        index_dict = None
+        index_file = 'model.safetensors.index.json'
+        if os.path.exists(os.path.join(src, index_file)):
+            with open(os.path.join(src, index_file)) as fp:
+                index_dict = json.load(fp)
+                index_dict['weight_map'] = {}
+        else:
+            index_dict = None
 
     os.makedirs(tgt, exist_ok=True)
     for filename in tqdm(os.listdir(src)):
-        if not filename.endswith('.bin'):
+        if not any(filename.endswith(ext) for ext in ('.bin', '.safetensors')):
             continue
-        states = torch.load(os.path.join(src, filename))
+
+        print(f'Loading {os.path.join(src, filename)}...', flush=True)
+        states = weight_load(os.path.join(src, filename),
+                             is_safetensors=filename.endswith('.safetensors'))
+
         llama_states = {}
         for k, v in states.copy().items():
             if 'wqkv' in k:
@@ -104,15 +136,15 @@ def convert(src, tgt):
         if index_dict is not None:
             for k in llama_states:
                 index_dict['weight_map'][k] = filename
-        print(f"Saving to {os.path.join(tgt, filename)}...", flush=True)
+
+        print(f'Saving to {os.path.join(tgt, filename)}...', flush=True)
         torch.save(llama_states, os.path.join(tgt, filename))
         del states
 
-    print('Saving config and tokenizer...')
+    print('Saving config and tokenizer...', flush=True)
     # index.json
     if index_dict is not None:
-        with open(os.path.join(tgt, 'pytorch_model.bin.index.json'),
-                  'w') as fp:
+        with open(os.path.join(tgt, index_file), 'w') as fp:
             json.dump(index_dict, fp, indent=2)
     # tokenizer
     tokenizer = LlamaTokenizer.from_pretrained(src)
@@ -120,7 +152,8 @@ def convert(src, tgt):
     tokenizer.save_pretrained(tgt)
     # config
     save_conifg(config, tgt)
-    print('Done!')
+
+    print('Done!', flush=True)
 
 
 def parse_args():