Update app.py

app.py

@@ -1,35 +1,55 @@
 from transformers import AutoTokenizer, AutoModelForCausalLM
-from peft import LoraConfig, get_peft_model
+from peft import LoraConfig, get_peft_model, PeftModel, PeftConfig
 import torch
+import streamlit as st
 
+# Suppress the warning if necessary
+import warnings
+warnings.filterwarnings("ignore", category=FutureWarning)
 
-from peft import PeftModel, PeftConfig
-from transformers import AutoModelForCausalLM
-
+# Load PEFT configuration
 config = PeftConfig.from_pretrained("youssef227/llama-3-8b-Instruct-bnb-telcom-3")
-print("step 1 ")
+print("step 1")
+
+# Load the base model
 base_model = AutoModelForCausalLM.from_pretrained("unsloth/llama-3-8b-Instruct-bnb-4bit")
 print("step 2")
+
+# Apply PEFT configuration to the base model
 model = PeftModel.from_pretrained(base_model, "youssef227/llama-3-8b-Instruct-bnb-telcom-3")
 print("step 3")
 
-# Load the tokenizer and model
-# print("step 1 ")
-# tokenizer = AutoTokenizer.from_pretrained("youssef227/llama-3-8b-Instruct-bnb-telcom-3")
-# print("step 2 ")
-# model = AutoModelForCausalLM.from_pretrained("youssef227/llama-3-8b-Instruct-bnb-telcom-3")
+# Load the tokenizer
+tokenizer = AutoTokenizer.from_pretrained("unsloth/llama-3-8b-Instruct-bnb-4bit")
+
+# Define the prompt template
+alpaca_prompt = "{instruction} {input} {output}"
+
 def generator(text):
-  inputs = tokenizer(
-  [
-      alpaca_prompt.format(
-          f" {context}انت ممثل خدمة العملاء لدى شركة فودافون.و دي معلومات ممكن تفيدك", # instruction
-          text, # input
-          "", # output - leave this blank for generation!
-      )
-  ], return_tensors = "pt").to("cuda")
-  outputs = model.generate(**inputs, max_new_tokens = 64, use_cache = True)
-  return tokenizer.batch_decode(outputs)
-text = st.text_area('enter some text!')
+    # Define the context if it's used in the prompt
+    context = " "
+
+    # Prepare the inputs
+    inputs = tokenizer(
+        [
+            alpaca_prompt.format(
+                instruction="انت ممثل خدمة العملاء لدى شركة فودافون.و دي معلومات ممكن تفيدك", # instruction
+                input=text, # input
+                output="" # output - leave this blank for generation!
+            )
+        ], return_tensors="pt"
+    ).to("cuda")
+
+    # Generate the output
+    outputs = model.generate(**inputs, max_new_tokens=64, use_cache=True)
+
+    # Decode the output
+    return tokenizer.batch_decode(outputs, skip_special_tokens=True)
+
+# Streamlit application
+st.title("Text Generator")
+text = st.text_area('Enter some text!')
+
 if text:
-  out = generator(text)
-  st.json(out)
+    out = generator(text)
+    st.json(out)