From 03735890f122b7d3338aa4f96f628f35c40fffed Mon Sep 17 00:00:00 2001
From: jinbridge <2635480475@qq.com>
Date: Fri, 21 Jun 2024 10:43:42 +0800
Subject: [PATCH] fix

---
 .../CPU/HF-Transformers-AutoModels/Model/glm-4v/generate.py  | 5 +++--
 .../GPU/HF-Transformers-AutoModels/Model/glm-4v/generate.py  | 5 +++--
 2 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/python/llm/example/CPU/HF-Transformers-AutoModels/Model/glm-4v/generate.py b/python/llm/example/CPU/HF-Transformers-AutoModels/Model/glm-4v/generate.py
index d4f31ad8094..384799f884a 100644
--- a/python/llm/example/CPU/HF-Transformers-AutoModels/Model/glm-4v/generate.py
+++ b/python/llm/example/CPU/HF-Transformers-AutoModels/Model/glm-4v/generate.py
@@ -40,7 +40,6 @@
     args = parser.parse_args()
     model_path = args.repo_id_or_model_path
     image_path = args.image_url_or_path
-    device = "cpu"
 
     # Load model in 4 bit,
     # which convert the relevant layers in the model into INT4 format
@@ -57,12 +56,14 @@
        image = Image.open(image_path)
     else:
        image = Image.open(requests.get(image_path, stream=True).raw)
+
+    # here the prompt tuning refers to https://huggingface.co/THUDM/glm-4v-9b/blob/main/README.md
     inputs = tokenizer.apply_chat_template([{"role": "user", "image": image, "content": query}],
                                            add_generation_prompt=True,
                                            tokenize=True,
                                            return_tensors="pt",
                                            return_dict=True)  # chat mode
-    inputs = inputs.to(device)
+    inputs = inputs.to('cpu')
     
     # Generate predicted tokens
     with torch.inference_mode():
diff --git a/python/llm/example/GPU/HF-Transformers-AutoModels/Model/glm-4v/generate.py b/python/llm/example/GPU/HF-Transformers-AutoModels/Model/glm-4v/generate.py
index 82fe1ea1dd6..6a1dd035e9e 100644
--- a/python/llm/example/GPU/HF-Transformers-AutoModels/Model/glm-4v/generate.py
+++ b/python/llm/example/GPU/HF-Transformers-AutoModels/Model/glm-4v/generate.py
@@ -40,7 +40,6 @@
     args = parser.parse_args()
     model_path = args.repo_id_or_model_path
     image_path = args.image_url_or_path
-    device = "xpu"
     
     # Load model in 4 bit,
     # which convert the relevant layers in the model into INT4 format
@@ -59,12 +58,14 @@
        image = Image.open(image_path)
     else:
        image = Image.open(requests.get(image_path, stream=True).raw)
+
+    # here the prompt tuning refers to https://huggingface.co/THUDM/glm-4v-9b/blob/main/README.md
     inputs = tokenizer.apply_chat_template([{"role": "user", "image": image, "content": query}],
                                            add_generation_prompt=True,
                                            tokenize=True,
                                            return_tensors="pt",
                                            return_dict=True)  # chat mode
-    inputs = inputs.to(device)
+    inputs = inputs.to('xpu')
 
     
     # Generate predicted tokens