Update app.py

app.py

@@ -1,59 +1,35 @@
-
-import json
-import subprocess
-from threading import Thread
-
 import torch
-import spaces
 import gradio as gr
 from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig, TextIteratorStreamer
+from threading import Thread
 
-#subprocess.run('pip install flash-attn --no-build-isolation', env={'FLASH_ATTENTION_SKIP_CUDA_BUILD': "TRUE"}, shell=True)
-
+# تنظیمات مدل
 MODEL_ID = "deepseek-ai/DeepSeek-R1-Distill-Qwen-14B"
 CHAT_TEMPLATE = "Auto"
 MODEL_NAME = MODEL_ID.split("/")[-1]
 CONTEXT_LENGTH = 16000
 
-# Estableciendo valores directamente para las variables
-COLOR = "blue"  # Color predeterminado de la interfaz
-EMOJI = "🤖"  # Emoji predeterminado para el modelo
-DESCRIPTION = f"This is the {MODEL_NAME} model designed for testing thinking for general AI tasks."  # Descripción predeterminada
 
-latex_delimiters_set = [{
-        "left": "\\(",
-        "right": "\\)",
-        "display": False 
-    }, {
-        "left": "\\begin{equation}",
-        "right": "\\end{equation}",
-        "display": True 
-    }, {
-        "left": "\\begin{align}",
-        "right": "\\end{align}",
-        "display": True
-    }, {
-        "left": "\\begin{alignat}",
-        "right": "\\end{alignat}",
-        "display": True
-    }, {
-        "left": "\\begin{gather}",
-        "right": "\\end{gather}",
-        "display": True
-    }, {
-        "left": "\\begin{CD}",
-        "right": "\\end{CD}",
-        "display": True
-    }, {
-        "left": "\\[",
-        "right": "\\]",
-        "display": True
-    }]
+COLOR = "blue"
+EMOJI = "🤖"
+DESCRIPTION = f"This is the {MODEL_NAME} model designed for testing thinking for general AI tasks."
+
+
+device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
+quantization_config = BitsAndBytesConfig(
+    load_in_4bit=True,
+    bnb_4bit_compute_dtype=torch.bfloat16
+)
+tokenizer = AutoTokenizer.from_pretrained(MODEL_ID)
+model = AutoModelForCausalLM.from_pretrained(
+    MODEL_ID,
+    device_map="auto",
+    quantization_config=quantization_config,
+).to(device)
 
 
-@spaces.GPU()
 def predict(message, history, system_prompt, temperature, max_new_tokens, top_k, repetition_penalty, top_p):
-    # Format history with a given chat template
+   
     if CHAT_TEMPLATE == "Auto":
         stop_tokens = [tokenizer.eos_token_id]
         instruction = system_prompt + "\n\n"
@@ -74,18 +50,18 @@ def predict(message, history, system_prompt, temperature, max_new_tokens, top_k,
         instruction += f' {message} [/INST]'
     else:
         raise Exception("Incorrect chat template, select 'Auto', 'ChatML' or 'Mistral Instruct'")
-    print(instruction)
+    
     
     streamer = TextIteratorStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True)
     enc = tokenizer(instruction, return_tensors="pt", padding=True, truncation=True)
-    input_ids, attention_mask = enc.input_ids, enc.attention_mask
+    input_ids, attention_mask = enc.input_ids.to(device), enc.attention_mask.to(device)
 
     if input_ids.shape[1] > CONTEXT_LENGTH:
         input_ids = input_ids[:, -CONTEXT_LENGTH:]
         attention_mask = attention_mask[:, -CONTEXT_LENGTH:]
 
     generate_kwargs = dict(
-        input_ids=input_ids.to(device),
+        input_ids=input_ids,
         attention_mask=attention_mask,
         streamer=streamer,
         do_sample=True,
@@ -105,28 +81,10 @@ def predict(message, history, system_prompt, temperature, max_new_tokens, top_k,
         yield "".join(outputs)
 
 
-# Load model
-device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
-quantization_config = BitsAndBytesConfig(
-    load_in_4bit=True,
-    bnb_4bit_compute_dtype=torch.bfloat16
-)
-tokenizer = AutoTokenizer.from_pretrained(MODEL_ID)
-model = AutoModelForCausalLM.from_pretrained(
-    MODEL_ID,
-    device_map="auto",
-    quantization_config=quantization_config,
-    #attn_implementation="flash_attention_2",
-)
-
-# Create Gradio interface
 gr.ChatInterface(
     predict,
     title=EMOJI + " " + MODEL_NAME,
     description=DESCRIPTION,
-    
-
-     
     additional_inputs_accordion=gr.Accordion(label="⚙️ Parameters", open=False),
     additional_inputs=[
         gr.Textbox("You are a code assistant.", label="System prompt"),
@@ -137,5 +95,4 @@ gr.ChatInterface(
         gr.Slider(0, 1, 0.95, label="Top P sampling"),
     ],
     theme=gr.themes.Soft(primary_hue=COLOR),
-).queue().launch()
-
+).queue().launch()