Edit model card

Model description:

Model: pgajo/mbert-xlwa-en-it

Dataset: TASTEset

Unshuffled ratio: ['0']

Shuffled ratio: ['1']

Best exact match epoch: 6

Best exact match: 84.89

Best epoch: 6

Drop duplicates: ['1']

Max epochs = 10

Optimizer lr = 3e-05

Optimizer eps = 1e-08

Batch size = 32

Dataset path = pgajo/EW-TT-PE_U0_S1_Tingredient_DROP1_mbert

Results

epoch train_loss train_f1 train_exact dev_loss dev_f1 dev_exact test_loss test_f1 test_exact
1 1.29 64.97 47.04 0.56 81.58 71.7 0 0 0
2 0.44 86.17 76.17 0.48 85.76 78.57 0 0 0
3 0.2 94.29 89.39 0.51 88.31 81.87 0 0 0
4 0.11 96.45 93.66 0.49 88.36 82.69 0 0 0
5 0.08 97.25 95.25 0.56 88.42 82.42 0 0 0
6 0.05 98.35 96.97 0.55 89.65 84.89 0 0 0
7 0.04 99.06 98.14 0.56 88.35 83.79 0 0 0
8 0.02 99.37 99.04 0.63 88.79 84.07 0 0 0
9 0.02 99.31 98.9 0.63 89.55 84.62 0 0 0
Downloads last month
2
Safetensors
Model size
177M params
Tensor type
F32