Haritz Puerto
commited on
Commit
•
b9d4392
1
Parent(s):
e7b628d
Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language:
|
| 3 |
+
- en
|
| 4 |
+
tags:
|
| 5 |
+
- QA
|
| 6 |
+
- Question Answering
|
| 7 |
+
- SQuAD
|
| 8 |
+
license: "mit"
|
| 9 |
+
datasets:
|
| 10 |
+
- squad
|
| 11 |
+
metrics:
|
| 12 |
+
- squad_metric
|
| 13 |
+
---
|
| 14 |
+
|
| 15 |
+
microsoft/xtremedistil-l6-h256-uncased fined-tuned on SQuAD (https://huggingface.co/datasets/squad)
|
| 16 |
+
|
| 17 |
+
Hyperparameters:
|
| 18 |
+
- epochs: 1
|
| 19 |
+
- lr: 1e-5
|
| 20 |
+
- train batch sie: 16
|
| 21 |
+
- optimizer: adamW
|
| 22 |
+
- lr_scheduler: linear
|
| 23 |
+
- num warming steps: 0
|
| 24 |
+
- max_length: 512
|
| 25 |
+
|
| 26 |
+
Results on the dev set:
|
| 27 |
+
- 'exact_match': 62.66792809839168
|
| 28 |
+
- 'f1': 74.99490608582015
|