adjust generation time

app.py

@@ -13,9 +13,16 @@ import psutil
 def print_system_resources():
     memory = psutil.virtual_memory()
     cpu_percent = psutil.cpu_percent(interval=1)
-    print(f"Total physical memory: {memory.total/1e9:.2f} GB")
+    # Get container memory limit (for Docker)
+    try:
+        with open('/sys/fs/cgroup/memory/memory.limit_in_bytes', 'r') as f:
+            mem_limit = int(f.read().strip()) / 1e9  # Convert to GB
+    except:
+        mem_limit = memory.total / 1e9  # Fallback to psutil
+    print(f"Total physical memory (psutil): {memory.total/1e9:.2f} GB")
+    print(f"Container memory limit: {mem_limit:.2f} GB")
     print(f"CPU usage: {cpu_percent}%")
-    print(f"Memory usage: {memory.percent}% ({memory.used/1e9:.2f}/{memory.total/1e9:.2f} GB)")
+    print(f"Memory usage: {memory.percent}% ({memory.used/1e9:.2f}/{mem_limit:.2f} GB)")
     print(f"Active processes: {len(psutil.pids())}")
 
 # Print Gradio version for debugging
@@ -51,9 +58,17 @@ print(f"Memory allocated: {torch.cuda.memory_allocated(device)/1e9:.2f} GB" if t
 print_system_resources()
 
 def clean_text(text):
-    """Clean generated text by removing non-alphabetic characters and extra spaces."""
+    """Clean generated text by removing non-alphabetic characters and incomplete sentences."""
     text = re.sub(r'[^\w\s.,!?àáâãèéêìíòóôõùúýăđĩũơưạảấầẩẫậắằẳẵặẹẻẽếềểễệỉịọỏốồổỗộớờởỡợụủứừửữựỳỵỷỹ]', '', text)
     text = re.sub(r'\s+', ' ', text).strip()
+    # Remove incomplete sentence (ends without punctuation)
+    if not re.search(r'[.!?]$', text):
+        last_period = text.rfind('.')
+        last_exclam = text.rfind('!')
+        last_question = text.rfind('?')
+        last_punct = max(last_period, last_exclam, last_question)
+        if last_punct != -1:
+            text = text[:last_punct + 1]
     return text
 
 def generate_text(prompt, max_length=50, temperature=0.9):
@@ -73,10 +88,10 @@ def generate_text(prompt, max_length=50, temperature=0.9):
         outputs = model.generate(
             input_ids=inputs["input_ids"],
             attention_mask=inputs["attention_mask"],
-            max_new_tokens=30,  # Increase for more content
+            max_new_tokens=20,  # Reduce for speed
             min_length=10,
-            do_sample=True,  # Enable sampling for diversity
-            top_k=50,
+            do_sample=True,
+            top_k=40,  # Reduce for speed
             top_p=0.9,
             temperature=temperature,
             no_repeat_ngram_size=2,
@@ -110,4 +125,4 @@ demo = gr.Interface(
 )
 
 if __name__ == "__main__":
-    demo.launch(server_name="0.0.0.0", server_port=7860)  # Remove queue=False
+    demo.launch(server_name="0.0.0.0", server_port=7860)