Update app.py

app.py

@@ -99,14 +99,40 @@ LANGUAGES = {
     "Western Persian": "Western Persian"
 }
 
+from huggingface_hub import snapshot_download
+from vllm import LLM, SamplingParams, BeamSearchParams
+
+save_dir = "./model_weights"
+repo_id = "ByteDance-Seed/Seed-X-PPO-7B"
+cache_dir = save_dir + "/cache"
+
+snapshot_download(
+  cache_dir=cache_dir,
+  local_dir=save_dir,
+  repo_id=repo_id,
+  local_dir_use_symlinks=False,
+  resume_download=True,
+  allow_patterns=["*.json", "*.safetensors", "*.bin", "*.py", "*.md", "*.txt"],
+)
+
+model = LLM(model=save_dir,
+            max_num_seqs=512,
+            # tensor_parallel_size=8,
+            enable_prefix_caching=True, 
+            gpu_memory_utilization=0.95)
+
+decoding_params = BeamSearchParams(beam_width=4,
+                                   max_tokens=512)
+
 @spaces.GPU(duration=120)
 def translate_text(text, source_lang, target_lang):
     if not text.strip():
         return "请输入要翻译的文本"
     
     try:
-        response = 'Test Seed-X'
-        yield response
+        response = model.generate(messages, decoding_params)
+        response = [res.outputs[0].text.strip() for res in results]
+        yield response[0]
     except Exception as e:
         yield f"翻译出错: {str(e)}"