AbstractPhil commited on
Commit
41c0c30
·
verified ·
1 Parent(s): 781892b

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +8 -6
app.py CHANGED
@@ -1,6 +1,5 @@
1
  # app.py
2
 
3
- from huggingface_hub import snapshot_download
4
  from transformers import AutoTokenizer, AutoModelForMaskedLM
5
  import torch
6
  import gradio as gr
@@ -9,9 +8,11 @@ from dataclasses import dataclass
9
  from pathlib import Path
10
  import spaces
11
 
 
12
  @dataclass
13
  class SymbolicConfig:
14
  repo_id: str = "AbstractPhil/bert-beatrix-2048"
 
15
  symbolic_roles: list = (
16
  "<subject>", "<subject1>", "<subject2>", "<pose>", "<emotion>",
17
  "<surface>", "<lighting>", "<material>", "<accessory>", "<footwear>",
@@ -22,14 +23,15 @@ class SymbolicConfig:
22
  )
23
 
24
  config = SymbolicConfig()
25
- model_dir = snapshot_download(repo_id=config.repo_id)
26
- tokenizer = AutoTokenizer.from_pretrained(model_dir)
27
- model = AutoModelForMaskedLM.from_pretrained(model_dir, trust_remote_code=True).eval().cuda()
28
-
 
 
29
 
30
  MASK_TOKEN = tokenizer.mask_token or "[MASK]"
31
 
32
- @spaces.GPU
33
  def mask_and_predict(text: str, selected_roles: list[str]):
34
  results = []
35
  masked_text = text
 
1
  # app.py
2
 
 
3
  from transformers import AutoTokenizer, AutoModelForMaskedLM
4
  import torch
5
  import gradio as gr
 
8
  from pathlib import Path
9
  import spaces
10
 
11
+ @spaces.GPU
12
  @dataclass
13
  class SymbolicConfig:
14
  repo_id: str = "AbstractPhil/bert-beatrix-2048"
15
+ revision: str = "main"
16
  symbolic_roles: list = (
17
  "<subject>", "<subject1>", "<subject2>", "<pose>", "<emotion>",
18
  "<surface>", "<lighting>", "<material>", "<accessory>", "<footwear>",
 
23
  )
24
 
25
  config = SymbolicConfig()
26
+ tokenizer = AutoTokenizer.from_pretrained(config.repo_id, revision=config.revision)
27
+ model = AutoModelForMaskedLM.from_pretrained(
28
+ config.repo_id,
29
+ revision=config.revision,
30
+ trust_remote_code=True
31
+ ).eval().cuda()
32
 
33
  MASK_TOKEN = tokenizer.mask_token or "[MASK]"
34
 
 
35
  def mask_and_predict(text: str, selected_roles: list[str]):
36
  results = []
37
  masked_text = text