Comment
Accuracy over SST2
Accuracy over SST2
Losses
Losses
Runs over SST2
10
Name
10 visualized
Tags
Runtime
Sweep
best_acc
method
loss_type
alpha
acc_and_f1
corr
f1
mcc
pearson
spearmanr
distill
sst2
42m 4s
-
0.84404
cut
distill
0.5
-
-
-
-
-
-
distill
sst2
1h 12m 45s
-
0.85665
cut
distill
0.2
-
-
-
-
-
-
attention_loss
sst2
1h 21m 23s
-
0.90711
cut
attention
0.2
-
-
-
-
-
-
prune
sst2
1h 49m 23s
-
0.8383
prune
distill
0.5
-
-
-
-
-
-
attention_loss
prune
sst2
1h 59m 51s
-
0.84862
prune
attention
0.5
-
-
-
-
-
-
distill
sst2
1h 13m 11s
-
0.8555
cut
distill
0.5
-
-
-
-
-
-
attention_loss
distill
sst2
1h 20m 16s
-
0.91055
cut
attention
0.5
-
-
-
-
-
-
distill
hidden
sst2
1h 14m 1s
-
0.85206
cut
distill
0.8
-
-
-
-
-
-
attention_loss
distill
hidden
sst2
1h 21m 33s
-
0.9094
cut
attention
0.8
-
-
-
-
-
-
baseline
sst2
1h 25m 39s
-
0.93234
cut
attention
0.8
-
-
-
-
-
-
1-10
of 10
Add a comment
Created with ❤️ on Weights & Biases.
https://wandb.ai/geraltofrivia783/lilbert/reports/SST2-Distillations--VmlldzoxOTcxNw