Update README.md
Browse files
README.md
CHANGED
@@ -11,13 +11,7 @@ probably proofread and complete it, then remove this comment. -->
|
|
11 |
|
12 |
# xtremedistil-l6-h256-uncased-nli
|
13 |
|
14 |
-
|
15 |
-
It achieves the following results on the evaluation set:
|
16 |
-
- Train Loss: 0.1618
|
17 |
-
- Train Accuracy: 0.9368
|
18 |
-
- Validation Loss: 0.2226
|
19 |
-
- Validation Accuracy: 0.9102
|
20 |
-
- Epoch: 2
|
21 |
|
22 |
## Model description
|
23 |
|
@@ -41,12 +35,11 @@ The following hyperparameters were used during training:
|
|
41 |
|
42 |
### Training results
|
43 |
|
44 |
-
|
45 |
-
|
46 |
-
|
47 |
-
|
48 |
-
|
49 |
-
|
50 |
|
51 |
### Framework versions
|
52 |
|
|
|
11 |
|
12 |
# xtremedistil-l6-h256-uncased-nli
|
13 |
|
14 |
+
The model base is xtremedistil-l6-h256-uncased. It has been fine-tuned on: snli, wanli, mnli, anli, scitail
|
|
|
|
|
|
|
|
|
|
|
|
|
15 |
|
16 |
## Model description
|
17 |
|
|
|
35 |
|
36 |
### Training results
|
37 |
|
38 |
+
It achieved the following accuracy during training (on validation sets):
|
39 |
+
SNLI: 87,90%
|
40 |
+
MNLI: 82.27%
|
41 |
+
ANLI_r3: 44.83%
|
42 |
+
scitail: 91.02%
|
|
|
43 |
|
44 |
### Framework versions
|
45 |
|