Dmitry Chaplinsky commited on
Commit
6a92397
1 Parent(s): 616a2eb

Updated model: 531 splits, 18.96 epochs, min_loss: 1.0162, min_ppl: 2.7628

Browse files
Files changed (2) hide show
  1. best-lm.pt +1 -1
  2. loss.txt +7 -0
best-lm.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8b54eb3ba46de29b3cfebd2f7285104a8ca2f67c98f7a39b748651a285b5808a
3
  size 22791455
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a8556f295b33ed311973a9ec4fe4adc81159dd7b7b24c9b85028580566d33583
3
  size 22791455
loss.txt CHANGED
@@ -522,3 +522,10 @@
522
  | end of split 74 / 28 | epoch 17 | time: 3256.39s | valid loss 1.0162 | valid ppl 2.7627 | learning rate 0.3125
523
  | end of split 75 / 28 | epoch 17 | time: 956.20s | valid loss 1.0162 | valid ppl 2.7627 | learning rate 0.3125
524
  | end of split 76 / 28 | epoch 17 | time: 3275.60s | valid loss 1.0162 | valid ppl 2.7627 | learning rate 0.3125
 
 
 
 
 
 
 
 
522
  | end of split 74 / 28 | epoch 17 | time: 3256.39s | valid loss 1.0162 | valid ppl 2.7627 | learning rate 0.3125
523
  | end of split 75 / 28 | epoch 17 | time: 956.20s | valid loss 1.0162 | valid ppl 2.7627 | learning rate 0.3125
524
  | end of split 76 / 28 | epoch 17 | time: 3275.60s | valid loss 1.0162 | valid ppl 2.7627 | learning rate 0.3125
525
+ | end of split 77 / 28 | epoch 17 | time: 3281.88s | valid loss 1.0162 | valid ppl 2.7626 | learning rate 0.3125
526
+ | end of split 78 / 28 | epoch 17 | time: 3282.88s | valid loss 1.0162 | valid ppl 2.7627 | learning rate 0.3125
527
+ | end of split 79 / 28 | epoch 17 | time: 3281.60s | valid loss 1.0162 | valid ppl 2.7626 | learning rate 0.3125
528
+ | end of split 80 / 28 | epoch 17 | time: 3282.62s | valid loss 1.0162 | valid ppl 2.7626 | learning rate 0.3125
529
+ | end of split 81 / 28 | epoch 17 | time: 3287.94s | valid loss 1.0162 | valid ppl 2.7627 | learning rate 0.3125
530
+ | end of split 82 / 28 | epoch 17 | time: 3278.46s | valid loss 1.0162 | valid ppl 2.7626 | learning rate 0.0781
531
+ | end of split 83 / 28 | epoch 17 | time: 3290.21s | valid loss 1.0162 | valid ppl 2.7626 | learning rate 0.0781