update

app.py

@@ -2,10 +2,10 @@ import gradio as gr
 from transformers import AutoTokenizer, pipeline
 import torch
 
-tokenizer1 = AutoTokenizer.from_pretrained("notexist/ttt")
-tdk1 = pipeline('text-generation', model='notexist/ttt', tokenizer=tokenizer1)
-tokenizer2 = AutoTokenizer.from_pretrained("notexist/ttt2")
-tdk2 = pipeline('text-generation', model='notexist/ttt2', tokenizer=tokenizer2)
+tokenizer1 = AutoTokenizer.from_pretrained("notexist/ttt2")
+tdk1 = pipeline('text-generation', model='notexist/ttt2', tokenizer=tokenizer1)
+tokenizer2 = AutoTokenizer.from_pretrained("notexist/ttte")
+tdk2 = pipeline('text-generation', model='notexist/ttte', tokenizer=tokenizer2)
 
 def predict(name, sl, topk, topp):
     if name == "":
@@ -17,7 +17,8 @@ def predict(name, sl, topk, topp):
             num_return_sequences=1,
             repetition_penalty=sl
         )[0]["generated_text"]
-        x2 = tdk1(f"<|endoftext|>",
+        new_name = x1[len(f"<|endoftext|>"):x1.index("\n\n")]
+        x2 = tdk2(f"<|endoftext|>{new_name}\n\n",
             do_sample=True, 
             max_length=64, 
             top_k=topk, 
@@ -26,7 +27,7 @@ def predict(name, sl, topk, topp):
             repetition_penalty=sl
         )[0]["generated_text"]
 
-        return x1[len(f"<|endoftext|>"):]+"\n\n"+x2[len(f"<|endoftext|>"):]
+        return x1[len(f"<|endoftext|>"):]+"\n\n"+x2[len(f"<|endoftext|>{new_name}\n\n"):]+"\n\n"+new_name
     else:
         x1 = tdk1(f"<|endoftext|>{name}\n\n",
             do_sample=True,