Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
tags:
|
| 3 |
+
- question-answer
|
| 4 |
+
language:
|
| 5 |
+
- it
|
| 6 |
+
model-index:
|
| 7 |
+
- name: it5-squadv1-it
|
| 8 |
+
results: []
|
| 9 |
+
datasets:
|
| 10 |
+
- z-uo/squad-it
|
| 11 |
+
---
|
| 12 |
+
|
| 13 |
+
# Question and Answer with Italian T5
|
| 14 |
+
|
| 15 |
+
This model is a fine-tuned version of [gsarti/it5-base](https://huggingface.co/gsarti/it5-base) on an unknown dataset.
|
| 16 |
+
It achieves the following results/params on the train set:
|
| 17 |
+
- epoch: 2.0
|
| 18 |
+
- train_loss: 0.0042
|
| 19 |
+
- train_samples: 87599
|
| 20 |
+
|
| 21 |
+
# Train the model
|
| 22 |
+
To train the model use [this repo](https://gitlab.com/nicolalandro/qandatrain), inside you find the requirements.txt and the src to create train.
|