PremalMatalia
commited on
Commit
路
8b3a927
1
Parent(s):
5b797ee
Updated Hyperparameters
Browse files
README.md
CHANGED
@@ -22,20 +22,23 @@ Tensorflow version (GPU?): 2.5.0 (False)</br>
|
|
22 |
|
23 |
## Hyperparameters
|
24 |
```
|
|
|
|
|
|
|
|
|
|
|
25 |
batch_size=8
|
|
|
26 |
n_epochs=6
|
27 |
base_LM_model = "deepset/roberta-base-squad2"
|
28 |
-
max_seq_len=386
|
29 |
-
doc_stride=128
|
30 |
learning_rate=1.5e-5
|
31 |
adam_epsilon=1e-5
|
32 |
adam_beta1=0.95
|
33 |
adam_beta2=0.999
|
34 |
warmup_steps=100
|
35 |
weight_decay=0.01
|
36 |
-
|
37 |
-
|
38 |
-
min_null_score=7.0
|
39 |
```
|
40 |
##### There is a special threshold value CLS_threshold=-3 used to more accurately identify no answers [Logic will be available in GitHub Repo [TBD]
|
41 |
|
|
|
22 |
|
23 |
## Hyperparameters
|
24 |
```
|
25 |
+
max_seq_len=386
|
26 |
+
doc_stride=128
|
27 |
+
n_best_size=20
|
28 |
+
max_answer_length=30
|
29 |
+
min_null_score=7.0
|
30 |
batch_size=8
|
31 |
+
|
32 |
n_epochs=6
|
33 |
base_LM_model = "deepset/roberta-base-squad2"
|
|
|
|
|
34 |
learning_rate=1.5e-5
|
35 |
adam_epsilon=1e-5
|
36 |
adam_beta1=0.95
|
37 |
adam_beta2=0.999
|
38 |
warmup_steps=100
|
39 |
weight_decay=0.01
|
40 |
+
optimizer=AdamW
|
41 |
+
lr_scheduler="polynomial"
|
|
|
42 |
```
|
43 |
##### There is a special threshold value CLS_threshold=-3 used to more accurately identify no answers [Logic will be available in GitHub Repo [TBD]
|
44 |
|