Update app.py

app.py

@@ -1,9 +1,13 @@
-from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
+import os
 from flask import Flask, request, jsonify
+from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
+
+# 💡 Hugging Face cache klasörünü geçici olarak yönlendir
+os.environ["TRANSFORMERS_CACHE"] = "/tmp/hf_cache"
 
 app = Flask(__name__)
 
-# Model yükleniyor
+# Model yükle
 model_id = "memorease/fine_tuned_flan_v2"
 tokenizer = AutoTokenizer.from_pretrained(model_id)
 model = AutoModelForSeq2SeqLM.from_pretrained(model_id)
@@ -13,10 +17,8 @@ def generate_question():
     data = request.json
     memory = data.get("memory", "")
     instruction = data.get("instruction", "Generate a question based on the given memory.")
-
     prompt = instruction + " " + memory
     inputs = tokenizer(prompt, return_tensors="pt", truncation=True, padding=True)
     outputs = model.generate(**inputs, max_new_tokens=64)
-
-    question = tokenizer.decode(outputs[0], skip_special_tokens=True)
-    return jsonify({"question": question})
+    result = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    return jsonify({"question": result})