2 llm

app.py

@@ -4,7 +4,7 @@ import torch
 import numpy as np
 from transformers import pipeline
 
-name_list = ['microsoft/biogpt', 'google/flan-ul2', 'facebook/galactica-1.3b']
+name_list = ['microsoft/biogpt', 'google/flan-ul2']
 
 examples = [['COVID-19 is'],['A 65-year-old female patient with a past medical history of']] 
 
@@ -12,20 +12,17 @@ print(f"Is CUDA available: {torch.cuda.is_available()}")
 print(f"CUDA device: {torch.cuda.get_device_name(torch.cuda.current_device())}")
 
 pipe_biogpt = pipeline("text-generation", model="microsoft/BioGPT-Large", device="cuda:0", model_kwargs={"torch_dtype":torch.bfloat16})
-pipe_flan_t5_xxl = pipeline("text-generation", model="google/flan-t5-xxl", device="cuda:0", model_kwargs={"torch_dtype":torch.bfloat16})
-pipe_gpt_2 = pipeline("text-generation", model="gpt2", device="cuda:0", model_kwargs={"torch_dtype":torch.bfloat16})
+pipe_flan_ul2 = pipeline("text-generation", model="google/flan-ul2", device="cuda:0", model_kwargs={"torch_dtype":torch.bfloat16})
 
 title = "LLM vs LLM!"
 description = "**Disclaimer:** this demo was made for research purposes only."
 
 def inference(text):
   output_biogpt = pipe_biogpt(text, max_length=100)[0]["generated_text"]
-  output_flan_t5_xxl = pipe_flan_t5_xxl(text, max_length=100)[0]["generated_text"]
-  output_gpt_2 = pipe_gpt_2(text, max_length=100)[0]["generated_text"]
+  output_flan_ul2 = pipe_flan_ul2(text, max_length=100)[0]["generated_text"]
   return [
       output_biogpt, 
-      output_flan_t5_xxl,
-      output_gpt_2
+      output_flan_ul2
   ]
 
 io = gr.Interface(
@@ -33,8 +30,7 @@ io = gr.Interface(
   gr.Textbox(lines=3),
   outputs=[
     gr.Textbox(lines=3, label="BioGPT-Large"),
-    gr.Textbox(lines=3, label="Flan T5 XXL"),
-    gr.Textbox(lines=3, label="GPT-2"),
+    gr.Textbox(lines=3, label="Flan UL2"),
   ],
   title=title,
   description=description,