system
HF staff
commited on
Commit
3bae579
1 Parent(s): becbf84

Update log.txt

Browse files
Files changed (1) hide show
  1. log.txt +24 -16
log.txt CHANGED
@@ -1,26 +1,34 @@
1
- Writing logs to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-yelp_polarity-2020-06-30-16:01/log.txt.
2
  Loading nlp dataset yelp_polarity, split train.
3
  Loading nlp dataset yelp_polarity, split test.
4
  Loaded dataset. Found: 2 labels: ([0, 1])
5
  Loading transformers AutoModelForSequenceClassification: bert-base-uncased
6
  Tokenizing training data. (len: 560000)
7
  Tokenizing eval data (len: 38000)
8
- Loaded data and tokenized in 1064.7807202339172s
9
- Training model across 1 GPUs
 
 
10
  ***** Running training *****
11
  Num examples = 560000
12
- Batch size = 8
13
- Max sequence length = 512
14
- Num steps = 350000
15
  Num epochs = 5
16
  Learning rate = 5e-05
17
- Eval accuracy: 50.0%
18
- Best acc found. Saved model to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-yelp_polarity-2020-06-30-16:01/.
19
- Eval accuracy: 50.00526315789474%
20
- Best acc found. Saved model to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-yelp_polarity-2020-06-30-16:01/.
21
- Eval accuracy: 50.0%
22
- Eval accuracy: 50.0%
23
- Eval accuracy: 50.0%
24
- Saved tokenizer <textattack.models.tokenizers.auto_tokenizer.AutoTokenizer object at 0x7f6bcb56cd00> to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-yelp_polarity-2020-06-30-16:01/.
25
- Wrote README to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-yelp_polarity-2020-06-30-16:01/README.md.
26
- Wrote training args to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-yelp_polarity-2020-06-30-16:01/train_args.json.
 
 
 
 
 
 
1
+ Writing logs to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-yelp_polarity-2020-07-08-10:42/log.txt.
2
  Loading nlp dataset yelp_polarity, split train.
3
  Loading nlp dataset yelp_polarity, split test.
4
  Loaded dataset. Found: 2 labels: ([0, 1])
5
  Loading transformers AutoModelForSequenceClassification: bert-base-uncased
6
  Tokenizing training data. (len: 560000)
7
  Tokenizing eval data (len: 38000)
8
+ Loaded data and tokenized in 720.6436557769775s
9
+ Using torch.nn.DataParallel.
10
+ Training model across 4 GPUs
11
+ Wrote original training args to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-yelp_polarity-2020-07-08-10:42/train_args.json.
12
  ***** Running training *****
13
  Num examples = 560000
14
+ Batch size = 16
15
+ Max sequence length = 256
16
+ Num steps = 175000
17
  Num epochs = 5
18
  Learning rate = 5e-05
19
+ Eval accuracy: 95.95263157894736%
20
+ Best acc found. Saved model to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-yelp_polarity-2020-07-08-10:42/.
21
+ Eval accuracy: 96.59473684210526%
22
+ Best acc found. Saved model to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-yelp_polarity-2020-07-08-10:42/.
23
+ Eval accuracy: 96.69473684210527%
24
+ Best acc found. Saved model to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-yelp_polarity-2020-07-08-10:42/.
25
+ Eval accuracy: 96.91052631578947%
26
+ Best acc found. Saved model to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-yelp_polarity-2020-07-08-10:42/.
27
+ Eval accuracy: 96.99473684210527%
28
+ Best acc found. Saved model to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-yelp_polarity-2020-07-08-10:42/.
29
+ Finished training. Re-loading and evaluating model from disk.
30
+ Loading transformers AutoModelForSequenceClassification: bert-base-uncased
31
+ Eval of saved model accuracy: 96.99473684210527%
32
+ Saved tokenizer <textattack.models.tokenizers.auto_tokenizer.AutoTokenizer object at 0x7fcc548eb730> to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-yelp_polarity-2020-07-08-10:42/.
33
+ Wrote README to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-yelp_polarity-2020-07-08-10:42/README.md.
34
+ Wrote final training args to /p/qdata/jm8wx/research/text_attacks/textattack/outputs/training/bert-base-uncased-yelp_polarity-2020-07-08-10:42/train_args.json.