hdallatorre commited on
Commit
58e9a17
·
1 Parent(s): 7cace1f

feat: revert to working version

Browse files
nucleotide_transformer_downstream_tasks_multilabel.py CHANGED
@@ -44,17 +44,19 @@ _LICENSE = "https://github.com/instadeepai/nucleotide-transformer/LICENSE.md"
44
  # The toy_classification and toy_regression are two manually created configurations
45
  # with 5 samples in both the train and test fasta files. It is notably used in order to
46
  # test the scripts.
47
- _TASKS_DTYPE = [
48
- ("deepstarr", "float32"),
49
- ("toy_classification", "int32"),
50
- ("toy_regression", "float32"),
51
  ]
52
 
53
 
54
  class NucleotideTransformerDownstreamTasksConfig(datasets.BuilderConfig):
55
  """BuilderConfig for The Nucleotide Transformer downstream taks dataset."""
56
 
57
- def __init__(self, *args, task: str, dtype: str = "int32", **kwargs):
 
 
58
  """BuilderConfig downstream tasks dataset.
59
  Args:
60
  task (:obj:`str`): Task name.
@@ -66,6 +68,7 @@ class NucleotideTransformerDownstreamTasksConfig(datasets.BuilderConfig):
66
  **kwargs,
67
  )
68
  self.task = task
 
69
  self.dtype = dtype
70
 
71
 
@@ -73,18 +76,24 @@ class NucleotideTransformerDownstreamTasks(datasets.GeneratorBasedBuilder):
73
  VERSION = datasets.Version("1.1.0")
74
  BUILDER_CONFIG_CLASS = NucleotideTransformerDownstreamTasksConfig
75
  BUILDER_CONFIGS = [
76
- NucleotideTransformerDownstreamTasksConfig(task=task, dtype=dtype)
77
- for (task, dtype) in _TASKS_DTYPE
 
 
78
  ]
79
  DEFAULT_CONFIG_NAME = "deepstarr"
80
 
81
  def _info(self):
82
-
83
- features = {
84
  "sequence": datasets.Value("string"),
85
  "name": datasets.Value("string"),
86
- # "labels": datasets.Sequence(self.config.dtype),
87
  }
 
 
 
 
 
 
88
 
89
  return datasets.DatasetInfo(
90
  # This is the description that will appear on the datasets page.
@@ -126,10 +135,15 @@ class NucleotideTransformerDownstreamTasks(datasets.GeneratorBasedBuilder):
126
  sequence, name = str(record.seq), str(record.name)
127
  labels = [float(label) for label in name.split("|")[1:]]
128
 
129
- # yield example
130
- yield key, {
131
  "sequence": sequence,
132
  "name": name,
133
- # "labels": labels,
134
  }
 
 
 
 
 
 
 
135
  key += 1
 
44
  # The toy_classification and toy_regression are two manually created configurations
45
  # with 5 samples in both the train and test fasta files. It is notably used in order to
46
  # test the scripts.
47
+ _TASKS_NUM_LABELS_DTYPE = [
48
+ ("deepstarr", 6, "float32"),
49
+ ("toy_classification", 2, "int32"),
50
+ ("toy_regression", 2, "float32"),
51
  ]
52
 
53
 
54
  class NucleotideTransformerDownstreamTasksConfig(datasets.BuilderConfig):
55
  """BuilderConfig for The Nucleotide Transformer downstream taks dataset."""
56
 
57
+ def __init__(
58
+ self, *args, task: str, num_labels=int, dtype: str = "int32", **kwargs
59
+ ):
60
  """BuilderConfig downstream tasks dataset.
61
  Args:
62
  task (:obj:`str`): Task name.
 
68
  **kwargs,
69
  )
70
  self.task = task
71
+ self.num_labels = num_labels
72
  self.dtype = dtype
73
 
74
 
 
76
  VERSION = datasets.Version("1.1.0")
77
  BUILDER_CONFIG_CLASS = NucleotideTransformerDownstreamTasksConfig
78
  BUILDER_CONFIGS = [
79
+ NucleotideTransformerDownstreamTasksConfig(
80
+ task=task, num_labels=num_labels, dtype=dtype
81
+ )
82
+ for (task, num_labels, dtype) in _TASKS_NUM_LABELS_DTYPE
83
  ]
84
  DEFAULT_CONFIG_NAME = "deepstarr"
85
 
86
  def _info(self):
87
+ features_dict = {
 
88
  "sequence": datasets.Value("string"),
89
  "name": datasets.Value("string"),
 
90
  }
91
+ labels_dict = {
92
+ f"label_{i}": datasets.Value(self.config.dtype)
93
+ for i in range(self.config.num_labels)
94
+ }
95
+ features_dict.update(labels_dict)
96
+ features = datasets.Features(features_dict)
97
 
98
  return datasets.DatasetInfo(
99
  # This is the description that will appear on the datasets page.
 
135
  sequence, name = str(record.seq), str(record.name)
136
  labels = [float(label) for label in name.split("|")[1:]]
137
 
138
+ sequence_name_dict = {
 
139
  "sequence": sequence,
140
  "name": name,
 
141
  }
142
+
143
+ labels_dict = {
144
+ f"label_{i}": labels[i] for i in range(self.config.num_labels)
145
+ }
146
+ sequence_name_dict.update(labels_dict)
147
+ # yield example
148
+ yield key, sequence_name_dict
149
  key += 1