Update app.py

app.py

@@ -1,13 +1,26 @@
 import gradio as gr
-from transformers import GPT2LMHeadModel, GPT2Tokenizer
+from transformers import AutoTokenizer, AutoModelForCausalLM
+import torch
 
-model_name = "speakleash/Bielik-1.5B-v3"
-tokenizer = GPT2Tokenizer.from_pretrained(model_name)
-model = GPT2LMHeadModel.from_pretrained(model_name)
+# Nazwa modelu na Hugging Face
+model_name = "speakleash/Bielik-1.5B-v3.0-Instruct-FP8-Dynamic"
 
-def generate_text(prompt):
-    input_ids = tokenizer.encode(prompt, return_tensors="pt")
-    output = model.generate(input_ids, max_length=100, num_return_sequences=1)
+# Ładowanie tokenizera i modelu
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype=torch.float32)
+
+# Funkcja generująca odpowiedź
+def generate(prompt):
+    input_ids = tokenizer(prompt, return_tensors="pt").input_ids
+    with torch.no_grad():
+        output = model.generate(
+            input_ids=input_ids,
+            max_new_tokens=150,
+            temperature=0.7,
+            top_p=0.95,
+            do_sample=True
+        )
     return tokenizer.decode(output[0], skip_special_tokens=True)
 
-gr.Interface(fn=generate_text, inputs="text", outputs="text", title="Polish GPT-2 Demo").launch()
+# Interfejs Gradio
+gr.Interface