Update README.md
Browse files
README.md
CHANGED
|
@@ -67,7 +67,7 @@ tensorboard --logdir callback_path
|
|
| 67 |
|
| 68 |
| Model | Gigaspeech_test<br>(Greedy / n-gram Beam Search) | VLSP2023_pb_test<br>(Greedy / n-gram Beam Search) | VLSP2023_pr_test<br>(Greedy / n-gram Beam Search) |
|
| 69 |
|:--------------------------------------|:------------------------------------------------:|:-------------------------------------------------:|:-------------------------------------------------:|
|
| 70 |
-
| **EC-Small-CTC** | **19.61 / 17.
|
| 71 |
|
| 72 |
|
| 73 |
In the competition organized by VLSP, I used the Efficient Conformer Large architecture with approximately 127 million parameters. You can find the detailed results in the technical report below:
|
|
|
|
| 67 |
|
| 68 |
| Model | Gigaspeech_test<br>(Greedy / n-gram Beam Search) | VLSP2023_pb_test<br>(Greedy / n-gram Beam Search) | VLSP2023_pr_test<br>(Greedy / n-gram Beam Search) |
|
| 69 |
|:--------------------------------------|:------------------------------------------------:|:-------------------------------------------------:|:-------------------------------------------------:|
|
| 70 |
+
| **EC-Small-CTC** | **19.61 / 17.47** | **23.06 / 20.83** | **23.17 / 21.15** |
|
| 71 |
|
| 72 |
|
| 73 |
In the competition organized by VLSP, I used the Efficient Conformer Large architecture with approximately 127 million parameters. You can find the detailed results in the technical report below:
|