Modify example from fp32 to fp16 (#10528)

* Modify example from fp32 to fp16

* Remove Falcon from fp16 example for now

* Remove MPT from fp16 example

python/llm/example/GPU/HF-Transformers-AutoModels/Model/baichuan2/generate.py

@@ -51,7 +51,7 @@ if __name__ == '__main__':
                                                  load_in_4bit=True,
                                                  trust_remote_code=True,
                                                  use_cache=True)
-    model = model.to('xpu')
+    model = model.half().to('xpu')
 
     # Load tokenizer
     tokenizer = AutoTokenizer.from_pretrained(model_path,

python/llm/example/GPU/HF-Transformers-AutoModels/Model/chatglm2/generate.py

@@ -48,7 +48,7 @@ if __name__ == '__main__':
                                       optimize_model=True,
                                       trust_remote_code=True,
                                       use_cache=True)
-    model = model.to('xpu')
+    model = model.half().to('xpu')
 
     # Load tokenizer
     tokenizer = AutoTokenizer.from_pretrained(model_path,

python/llm/example/GPU/HF-Transformers-AutoModels/Model/chatglm3/generate.py

@@ -48,7 +48,7 @@ if __name__ == '__main__':
                                       optimize_model=True,
                                       trust_remote_code=True,
                                       use_cache=True)
-    model = model.to('xpu')
+    model = model.half().to('xpu')
 
     # Load tokenizer
     tokenizer = AutoTokenizer.from_pretrained(model_path,

python/llm/example/GPU/HF-Transformers-AutoModels/Model/llama2/generate.py

@@ -61,7 +61,7 @@ if __name__ == '__main__':
                                                  optimize_model=True,
                                                  trust_remote_code=True,
                                                  use_cache=True)
-    model = model.to('xpu')
+    model = model.half().to('xpu')
 
     # Load tokenizer
     tokenizer = LlamaTokenizer.from_pretrained(model_path, trust_remote_code=True)

python/llm/example/GPU/HF-Transformers-AutoModels/Model/mistral/generate.py

@@ -47,7 +47,7 @@ if __name__ == '__main__':
                                                  optimize_model=True,
                                                  trust_remote_code=True,
                                                  use_cache=True)
-    model = model.to('xpu')
+    model = model.half().to('xpu')
 
     # Load tokenizer
     tokenizer = AutoTokenizer.from_pretrained(model_path, trust_remote_code=True)

python/llm/example/GPU/HF-Transformers-AutoModels/Model/qwen/generate.py

@@ -54,7 +54,7 @@ if __name__ == '__main__':
                                                  optimize_model=True,
                                                  trust_remote_code=True,
                                                  use_cache=True)
-    model = model.to('xpu')
+    model = model.half().to('xpu')
 
     # Load tokenizer
     tokenizer = AutoTokenizer.from_pretrained(model_path,

python/llm/example/GPU/HF-Transformers-AutoModels/Model/qwen1.5/generate.py

@@ -44,7 +44,7 @@ if __name__ == '__main__':
     model = AutoModelForCausalLM.from_pretrained(model_path,
                                                  load_in_4bit=True,
                                                  trust_remote_code=True)
-    model = model.to("xpu")
+    model = model.half().to("xpu")
 
     # Load tokenizer
     tokenizer = AutoTokenizer.from_pretrained(model_path,