Updates README.md
Browse filesSigned-off-by: Giovani <giovanitavares@outlook.com>
README.md
CHANGED
@@ -95,12 +95,16 @@ $10k$ sentence pairs equally distributed between *ptbr* and *ptpt* subsets.
|
|
95 |
### Fine-Tuning Procedure
|
96 |
|
97 |
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
|
98 |
-
The fine-tuning procedure can be summarized in three major subsequent tasks:
|
99 |
-
i
|
100 |
-
|
101 |
-
|
102 |
-
|
103 |
-
|
|
|
|
|
|
|
|
|
104 |
|
105 |
#### Training Hyperparameters
|
106 |
|
|
|
95 |
### Fine-Tuning Procedure
|
96 |
|
97 |
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
|
98 |
+
The model fine-tuning procedure can be summarized in three major subsequent tasks:
|
99 |
+
<ol type="i">
|
100 |
+
<li>**[Data Processing](#data-processing):**</li> [ASSIN](https://huggingface.co/datasets/assin)'s *validation* and *train* splits were loaded from the **Hugging Face Hub** and processed afterwards;
|
101 |
+
<li>**Hyperparameter Tuning:**</li> [XLM-RoBERTa-base](https://huggingface.co/xlm-roberta-base)'s hyperparameters were chosen with the help of the [Weights & Biases] API to track the results and upload the fine-tuned models;
|
102 |
+
<li>**Final Model Loading and Testing:**</li>
|
103 |
+
using the *cross-tests* approach described in the [this section](#evaluation), the models' performance were measured using different datasets and metrics.
|
104 |
+
</ol>
|
105 |
+
#### Data Processing [optional]
|
106 |
+
##### Class Label Column Renaming
|
107 |
+
The **Hugging Face**'s ```transformers``` module's ```DataCollator``` used by its ```Trainer``` requires that the ```class label``` column of the collated dataset to be called ```label```. [ASSIN](https://huggingface.co/datasets/assin)'s class label column for each hypothesis/premise pair is called ```entailment_judgement```. Therefore, as the first step of the data preprocessing pipeline the column ```entailment_judgement``` was renamed to ```label``` so that the **Hugging Face**'s ```transformers``` module's ```Trainer``` could be used.
|
108 |
|
109 |
#### Training Hyperparameters
|
110 |
|