jainr3 commited on
Commit
d2c1bbd
1 Parent(s): 8d9c67f

Update diffusiondb-pixelart.py

Browse files
Files changed (1) hide show
  1. diffusiondb-pixelart.py +8 -95
diffusiondb-pixelart.py CHANGED
@@ -41,9 +41,7 @@ _VERSION = datasets.Version("0.9.1")
41
  # hf_hub_url() provides a more flexible way to resolve the file URLs
42
  # https://huggingface.co/datasets/jainr3/diffusiondb-pixelart/resolve/main/images/part-000001.zip
43
  _URLS = {}
44
- _URLS_LARGE = {}
45
  _PART_IDS = range(1, 2001)
46
- _PART_IDS_LARGE = range(1, 14001)
47
 
48
  for i in _PART_IDS:
49
  _URLS[i] = hf_hub_url(
@@ -52,31 +50,12 @@ for i in _PART_IDS:
52
  repo_type="dataset",
53
  )
54
 
55
- for i in _PART_IDS_LARGE:
56
- if i < 10001:
57
- _URLS_LARGE[i] = hf_hub_url(
58
- "jainr3/diffusiondb-pixelart",
59
- filename=f"diffusiondb-pixelart-large-part-1/part-{i:06}.zip",
60
- repo_type="dataset",
61
- )
62
- else:
63
- _URLS_LARGE[i] = hf_hub_url(
64
- "jainr3/diffusiondb-pixelart",
65
- filename=f"diffusiondb-pixelart-large-part-2/part-{i:06}.zip",
66
- repo_type="dataset",
67
- )
68
 
69
  # Add the metadata parquet URL as well
70
  _URLS["metadata"] = hf_hub_url(
71
  "jainr3/diffusiondb-pixelart", filename="metadata.parquet", repo_type="dataset"
72
  )
73
 
74
- _URLS_LARGE["metadata"] = hf_hub_url(
75
- "jainr3/diffusiondb-pixelart",
76
- filename="metadata-large.parquet",
77
- repo_type="dataset",
78
- )
79
-
80
  _SAMPLER_DICT = {
81
  1: "ddim",
82
  2: "plms",
@@ -93,16 +72,14 @@ _SAMPLER_DICT = {
93
  class DiffusionDBConfig(datasets.BuilderConfig):
94
  """BuilderConfig for DiffusionDB."""
95
 
96
- def __init__(self, part_ids, is_large, **kwargs):
97
  """BuilderConfig for DiffusionDB.
98
  Args:
99
  part_ids([int]): A list of part_ids.
100
- is_large(bool): If downloading data from DiffusionDB Large (14 million)
101
  **kwargs: keyword arguments forwarded to super.
102
  """
103
  super(DiffusionDBConfig, self).__init__(version=_VERSION, **kwargs)
104
  self.part_ids = part_ids
105
- self.is_large = is_large
106
 
107
 
108
  class DiffusionDB(datasets.GeneratorBasedBuilder):
@@ -114,50 +91,8 @@ class DiffusionDB(datasets.GeneratorBasedBuilder):
114
  # as the config key)
115
  for num_k in [1, 5, 10, 50, 100, 500, 1000]:
116
  for sampling in ["first", "random"]:
117
- for is_large in [False, True]:
118
- num_k_str = f"{num_k}k" if num_k < 1000 else f"{num_k // 1000}m"
119
- subset_str = "large_" if is_large else "2m_"
120
-
121
- if sampling == "random":
122
- # Name the config
123
- cur_name = subset_str + "random_" + num_k_str
124
-
125
- # Add a short description for each config
126
- cur_description = (
127
- f"Random {num_k_str} images with their prompts and parameters"
128
- )
129
-
130
- # Sample part_ids
131
- total_part_ids = _PART_IDS_LARGE if is_large else _PART_IDS
132
- part_ids = np.random.choice(
133
- total_part_ids, num_k, replace=False
134
- ).tolist()
135
- else:
136
- # Name the config
137
- cur_name = subset_str + "first_" + num_k_str
138
-
139
- # Add a short description for each config
140
- cur_description = f"The first {num_k_str} images in this dataset with their prompts and parameters"
141
-
142
- # Sample part_ids
143
- total_part_ids = _PART_IDS_LARGE if is_large else _PART_IDS
144
- part_ids = total_part_ids[1 : num_k + 1]
145
-
146
- # Create configs
147
- BUILDER_CONFIGS.append(
148
- DiffusionDBConfig(
149
- name=cur_name,
150
- part_ids=part_ids,
151
- is_large=is_large,
152
- description=cur_description,
153
- ),
154
- )
155
-
156
- # Add few more options for Large only
157
- for num_k in [5000, 10000]:
158
- for sampling in ["first", "random"]:
159
- num_k_str = f"{num_k // 1000}m"
160
- subset_str = "large_"
161
 
162
  if sampling == "random":
163
  # Name the config
@@ -169,7 +104,7 @@ class DiffusionDB(datasets.GeneratorBasedBuilder):
169
  )
170
 
171
  # Sample part_ids
172
- total_part_ids = _PART_IDS_LARGE
173
  part_ids = np.random.choice(
174
  total_part_ids, num_k, replace=False
175
  ).tolist()
@@ -181,7 +116,7 @@ class DiffusionDB(datasets.GeneratorBasedBuilder):
181
  cur_description = f"The first {num_k_str} images in this dataset with their prompts and parameters"
182
 
183
  # Sample part_ids
184
- total_part_ids = _PART_IDS_LARGE
185
  part_ids = total_part_ids[1 : num_k + 1]
186
 
187
  # Create configs
@@ -189,26 +124,16 @@ class DiffusionDB(datasets.GeneratorBasedBuilder):
189
  DiffusionDBConfig(
190
  name=cur_name,
191
  part_ids=part_ids,
192
- is_large=True,
193
  description=cur_description,
194
  ),
195
  )
196
 
197
- # Need to manually add all (2m) and all (large)
 
198
  BUILDER_CONFIGS.append(
199
  DiffusionDBConfig(
200
  name="2m_all",
201
  part_ids=_PART_IDS,
202
- is_large=False,
203
- description="All images with their prompts and parameters",
204
- ),
205
- )
206
-
207
- BUILDER_CONFIGS.append(
208
- DiffusionDBConfig(
209
- name="large_all",
210
- part_ids=_PART_IDS_LARGE,
211
- is_large=True,
212
  description="All images with their prompts and parameters",
213
  ),
214
  )
@@ -218,19 +143,10 @@ class DiffusionDB(datasets.GeneratorBasedBuilder):
218
  DiffusionDBConfig(
219
  name="2m_text_only",
220
  part_ids=[],
221
- is_large=False,
222
  description="Only include all prompts and parameters (no image)",
223
  ),
224
  )
225
 
226
- BUILDER_CONFIGS.append(
227
- DiffusionDBConfig(
228
- name="large_text_only",
229
- part_ids=[],
230
- is_large=True,
231
- description="Only include all prompts and parameters (no image)",
232
- ),
233
- )
234
 
235
  # Default to only load 1k random images
236
  DEFAULT_CONFIG_NAME = "2m_random_1k"
@@ -300,10 +216,7 @@ class DiffusionDB(datasets.GeneratorBasedBuilder):
300
  json_paths = []
301
 
302
  # Resolve the urls
303
- if self.config.is_large:
304
- urls = _URLS_LARGE
305
- else:
306
- urls = _URLS
307
 
308
  for cur_part_id in self.config.part_ids:
309
  cur_url = urls[cur_part_id]
 
41
  # hf_hub_url() provides a more flexible way to resolve the file URLs
42
  # https://huggingface.co/datasets/jainr3/diffusiondb-pixelart/resolve/main/images/part-000001.zip
43
  _URLS = {}
 
44
  _PART_IDS = range(1, 2001)
 
45
 
46
  for i in _PART_IDS:
47
  _URLS[i] = hf_hub_url(
 
50
  repo_type="dataset",
51
  )
52
 
 
 
 
 
 
 
 
 
 
 
 
 
 
53
 
54
  # Add the metadata parquet URL as well
55
  _URLS["metadata"] = hf_hub_url(
56
  "jainr3/diffusiondb-pixelart", filename="metadata.parquet", repo_type="dataset"
57
  )
58
 
 
 
 
 
 
 
59
  _SAMPLER_DICT = {
60
  1: "ddim",
61
  2: "plms",
 
72
  class DiffusionDBConfig(datasets.BuilderConfig):
73
  """BuilderConfig for DiffusionDB."""
74
 
75
+ def __init__(self, part_ids, **kwargs):
76
  """BuilderConfig for DiffusionDB.
77
  Args:
78
  part_ids([int]): A list of part_ids.
 
79
  **kwargs: keyword arguments forwarded to super.
80
  """
81
  super(DiffusionDBConfig, self).__init__(version=_VERSION, **kwargs)
82
  self.part_ids = part_ids
 
83
 
84
 
85
  class DiffusionDB(datasets.GeneratorBasedBuilder):
 
91
  # as the config key)
92
  for num_k in [1, 5, 10, 50, 100, 500, 1000]:
93
  for sampling in ["first", "random"]:
94
+ num_k_str = f"{num_k}k" if num_k < 1000 else f"{num_k // 1000}m"
95
+ subset_str = "2m_"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
96
 
97
  if sampling == "random":
98
  # Name the config
 
104
  )
105
 
106
  # Sample part_ids
107
+ total_part_ids = _PART_IDS
108
  part_ids = np.random.choice(
109
  total_part_ids, num_k, replace=False
110
  ).tolist()
 
116
  cur_description = f"The first {num_k_str} images in this dataset with their prompts and parameters"
117
 
118
  # Sample part_ids
119
+ total_part_ids = _PART_IDS
120
  part_ids = total_part_ids[1 : num_k + 1]
121
 
122
  # Create configs
 
124
  DiffusionDBConfig(
125
  name=cur_name,
126
  part_ids=part_ids,
 
127
  description=cur_description,
128
  ),
129
  )
130
 
131
+
132
+ # Need to manually add all (2m)
133
  BUILDER_CONFIGS.append(
134
  DiffusionDBConfig(
135
  name="2m_all",
136
  part_ids=_PART_IDS,
 
 
 
 
 
 
 
 
 
 
137
  description="All images with their prompts and parameters",
138
  ),
139
  )
 
143
  DiffusionDBConfig(
144
  name="2m_text_only",
145
  part_ids=[],
 
146
  description="Only include all prompts and parameters (no image)",
147
  ),
148
  )
149
 
 
 
 
 
 
 
 
 
150
 
151
  # Default to only load 1k random images
152
  DEFAULT_CONFIG_NAME = "2m_random_1k"
 
216
  json_paths = []
217
 
218
  # Resolve the urls
219
+ urls = _URLS
 
 
 
220
 
221
  for cur_part_id in self.config.part_ids:
222
  cur_url = urls[cur_part_id]