osanseviero HF staff commited on
Commit
f479f1f
1 Parent(s): 70c0533

Update pipeline.py

Browse files
Files changed (1) hide show
  1. pipeline.py +0 -4
pipeline.py CHANGED
@@ -8,15 +8,12 @@ from fastai.learner import load_learner
8
 
9
  from helpers import is_cat
10
 
11
- def is_cat(x): return x[0].isupper()
12
-
13
  class PreTrainedPipeline():
14
  def __init__(self, path=""):
15
  # IMPLEMENT_THIS
16
  # Preload all the elements you are going to need at inference.
17
  # For instance your model, processors, tokenizer that might be needed.
18
  # This function is only called once, so do all the heavy processing I/O here"""
19
- def is_cat(x): return x[0].isupper()
20
  self.model = load_learner(os.path.join(path, "model.pkl"))
21
  with open(os.path.join(path, "config.json")) as config:
22
  config = json.load(config)
@@ -32,7 +29,6 @@ class PreTrainedPipeline():
32
  A :obj:`list`:. The list contains items that are dicts should be liked {"label": "XXX", "score": 0.82}
33
  It is preferred if the returned list is in decreasing `score` order
34
  """
35
- def is_cat(x): return x[0].isupper()
36
  # IMPLEMENT_THIS
37
  # FastAI expects a np array, not a PIL Image.
38
  _, _, preds = self.model.predict(np.array(inputs))
 
8
 
9
  from helpers import is_cat
10
 
 
 
11
  class PreTrainedPipeline():
12
  def __init__(self, path=""):
13
  # IMPLEMENT_THIS
14
  # Preload all the elements you are going to need at inference.
15
  # For instance your model, processors, tokenizer that might be needed.
16
  # This function is only called once, so do all the heavy processing I/O here"""
 
17
  self.model = load_learner(os.path.join(path, "model.pkl"))
18
  with open(os.path.join(path, "config.json")) as config:
19
  config = json.load(config)
 
29
  A :obj:`list`:. The list contains items that are dicts should be liked {"label": "XXX", "score": 0.82}
30
  It is preferred if the returned list is in decreasing `score` order
31
  """
 
32
  # IMPLEMENT_THIS
33
  # FastAI expects a np array, not a PIL Image.
34
  _, _, preds = self.model.predict(np.array(inputs))