Update README.md
Browse files
README.md
CHANGED
|
@@ -57,9 +57,11 @@ Fine tuning is done using the `train` split of the GLUE MNLI dataset and the per
|
|
| 57 |
`validation_mismatched` means validation examples are not derived from the same sources as those in the training set and therefore not closely resembling any of the examples seen at training time.
|
| 58 |
|
| 59 |
Data splits for the mnli dataset are the following
|
|
|
|
| 60 |
|train |validation_matched|validation_mismatched|
|
| 61 |
|-----:|-----------------:|--------------------:|
|
| 62 |
|392702| 9815| 9832|
|
|
|
|
| 63 |
## Fine-tuning procedure
|
| 64 |
Fine tuned on a Graphcore IPU-POD64 using `popxl`.
|
| 65 |
|
|
|
|
| 57 |
`validation_mismatched` means validation examples are not derived from the same sources as those in the training set and therefore not closely resembling any of the examples seen at training time.
|
| 58 |
|
| 59 |
Data splits for the mnli dataset are the following
|
| 60 |
+
|
| 61 |
|train |validation_matched|validation_mismatched|
|
| 62 |
|-----:|-----------------:|--------------------:|
|
| 63 |
|392702| 9815| 9832|
|
| 64 |
+
|
| 65 |
## Fine-tuning procedure
|
| 66 |
Fine tuned on a Graphcore IPU-POD64 using `popxl`.
|
| 67 |
|