Use "nyu-mll/glue" instead of "glue" for encoder datasets to fix 404 error

Signed-off-by: Jeremy Berchtold <jberchtold@nvidia.com>

Author: jberchtold-nvidia

examples/jax/encoder/test_model_parallel_encoder.py

@@ -219,11 +219,11 @@ def get_datasets(max_seq_len):
     vocab = {}
     word_id = 0
 
-    train_ds = load_dataset("glue", "cola", split="train")
+    train_ds = load_dataset("nyu-mll/glue", "cola", split="train")
     train_ds.set_format(type="np")
     train_ds, vocab, word_id = data_preprocess(train_ds, vocab, word_id, max_seq_len)
 
-    test_ds = load_dataset("glue", "cola", split="validation")
+    test_ds = load_dataset("nyu-mll/glue", "cola", split="validation")
     test_ds.set_format(type="np")
     test_ds, vocab, word_id = data_preprocess(test_ds, vocab, word_id, max_seq_len)
     return train_ds, test_ds, word_id

examples/jax/encoder/test_multigpu_encoder.py

@@ -197,11 +197,11 @@ def get_datasets(max_seq_len):
     vocab = {}
     word_id = 0
 
-    train_ds = load_dataset("glue", "cola", split="train")
+    train_ds = load_dataset("nyu-mll/glue", "cola", split="train")
     train_ds.set_format(type="np")
     train_ds, vocab, word_id = data_preprocess(train_ds, vocab, word_id, max_seq_len)
 
-    test_ds = load_dataset("glue", "cola", split="validation")
+    test_ds = load_dataset("nyu-mll/glue", "cola", split="validation")
     test_ds.set_format(type="np")
     test_ds, vocab, word_id = data_preprocess(test_ds, vocab, word_id, max_seq_len)
     return train_ds, test_ds, word_id

examples/jax/encoder/test_multiprocessing_encoder.py

@@ -307,11 +307,11 @@ def get_datasets(max_seq_len):
     vocab = {}
     word_id = 0
 
-    train_ds = load_dataset("glue", "cola", split="train")
+    train_ds = load_dataset("nyu-mll/glue", "cola", split="train")
     train_ds.set_format(type="np")
     train_ds, vocab, word_id = data_preprocess(train_ds, vocab, word_id, max_seq_len)
 
-    test_ds = load_dataset("glue", "cola", split="validation")
+    test_ds = load_dataset("nyu-mll/glue", "cola", split="validation")
     test_ds.set_format(type="np")
     test_ds, vocab, word_id = data_preprocess(test_ds, vocab, word_id, max_seq_len)
     return train_ds, test_ds, word_id

examples/jax/encoder/test_single_gpu_encoder.py

@@ -195,11 +195,11 @@ def get_datasets(max_seq_len):
     vocab = {}
     word_id = 0
 
-    train_ds = load_dataset("glue", "cola", split="train")
+    train_ds = load_dataset("nyu-mll/glue", "cola", split="train")
     train_ds.set_format(type="np")
     train_ds, vocab, word_id = data_preprocess(train_ds, vocab, word_id, max_seq_len)
 
-    test_ds = load_dataset("glue", "cola", split="validation")
+    test_ds = load_dataset("nyu-mll/glue", "cola", split="validation")
     test_ds.set_format(type="np")
     test_ds, vocab, word_id = data_preprocess(test_ds, vocab, word_id, max_seq_len)
     return train_ds, test_ds, word_id