|
171 | 171 | " untar=True, cache_dir='.',\n", |
172 | 172 | " cache_subdir='')\n", |
173 | 173 | "\n", |
174 | | - "dataset_dir = os.path.join(os.path.dirname(dataset), 'aclImdb')" |
| 174 | + "dataset_dir = os.path.join(os.path.dirname(dataset), 'aclImdb_v1')" |
175 | 175 | ] |
176 | 176 | }, |
177 | 177 | { |
|
193 | 193 | }, |
194 | 194 | "outputs": [], |
195 | 195 | "source": [ |
196 | | - "train_dir = os.path.join(dataset_dir, 'train')\n", |
| 196 | + "train_dir = os.path.join(dataset_dir, 'aclImdb', 'train')\n", |
| 197 | + "test_dir = os.path.join(dataset_dir, 'aclImdb', 'test')\n", |
197 | 198 | "os.listdir(train_dir)" |
198 | 199 | ] |
199 | 200 | }, |
|
214 | 215 | }, |
215 | 216 | "outputs": [], |
216 | 217 | "source": [ |
217 | | - "sample_file = os.path.join(train_dir, 'pos/1181_9.txt')\n", |
| 218 | + "sample_file = os.path.join(train_dir, 'pos', '1181_9.txt')\n", |
218 | 219 | "with open(sample_file) as f:\n", |
219 | 220 | " print(f.read())" |
220 | 221 | ] |
|
286 | 287 | "seed = 42\n", |
287 | 288 | "\n", |
288 | 289 | "raw_train_ds = tf.keras.utils.text_dataset_from_directory(\n", |
289 | | - " 'aclImdb/train',\n", |
| 290 | + " train_dir,\n", |
290 | 291 | " batch_size=batch_size,\n", |
291 | 292 | " validation_split=0.2,\n", |
292 | 293 | " subset='training',\n", |
|
366 | 367 | "outputs": [], |
367 | 368 | "source": [ |
368 | 369 | "raw_val_ds = tf.keras.utils.text_dataset_from_directory(\n", |
369 | | - " 'aclImdb/train',\n", |
| 370 | + " train_dir,\n", |
370 | 371 | " batch_size=batch_size,\n", |
371 | 372 | " validation_split=0.2,\n", |
372 | 373 | " subset='validation',\n", |
|
382 | 383 | "outputs": [], |
383 | 384 | "source": [ |
384 | 385 | "raw_test_ds = tf.keras.utils.text_dataset_from_directory(\n", |
385 | | - " 'aclImdb/test',\n", |
| 386 | + " test_dir,\n", |
386 | 387 | " batch_size=batch_size)" |
387 | 388 | ] |
388 | 389 | }, |
|
0 commit comments