Update app.py

app.py

@@ -1,30 +1,4 @@
-import random
-import torch
-import gradio as gr
-from gradio_rangeslider import RangeSlider
-import pandas as pd
-from utils import create_vocab, setup_seed
-from dataset_mlm import  get_paded_token_idx_gen, add_tokens_to_vocab
-import time 
-seed = random.randint(0,100000)
-
-setup_seed(seed)
-device = torch.device("cpu")
-vocab_mlm = create_vocab()
-vocab_mlm = add_tokens_to_vocab(vocab_mlm)
-save_path = 'mlm-model-27.pt'  
-train_seqs = pd.read_csv('C0_seq.csv')  
-train_seq = train_seqs['Seq'].tolist()
-model = torch.load(save_path)
-model = model.to(device)
-
-def temperature_sampling(logits, temperature):
-    logits = logits / temperature
-    probabilities = torch.softmax(logits, dim=-1)
-    sampled_token = torch.multinomial(probabilities, 1)
-    return sampled_token
-
-def CTXGen(τ, g_num, length_range):
+def CTXGen(τ, g_num, length_range, progress=gr.Progress()):
     start, end = length_range
     X1 = "X"
     X2 = "X"
@@ -114,7 +88,8 @@ def CTXGen(τ, g_num, length_range):
                     out = pd.DataFrame({'Generated_seq': generated_seqs_FINAL, 'Subtype': cls_pos_all, 'Subtype_probability': cls_probability_all, 'Potency': act_pos_all, 'Potency_probability': act_probability_all, 'random_seed': seed})
                     out.to_csv("output.csv", index=False, encoding='utf-8-sig')
                     count += 1
-    return 'output.csv'
+                    progress(count / gen_num, desc="Generating sequences...")
+    return 'output.csv', f"Generated {count} sequences."
 
 iface = gr.Interface(
     fn=CTXGen,
@@ -123,6 +98,9 @@ iface = gr.Interface(
         gr.Dropdown(choices=[1,10,100], label="Number of generations"),
         RangeSlider(minimum=8, maximum=50, step=1, value=(12, 17), label="Length range")
     ],
-    outputs=["file"]
+    outputs=[
+        gr.File(label="Download generated sequences"),
+        gr.Textbox(label="Progress")
+    ]
 )
 iface.launch()