Datasets:
patrickvonplaten
commited on
Commit
•
9deacee
1
Parent(s):
3b46101
Update README.md
Browse files
README.md
CHANGED
@@ -63,6 +63,11 @@ As can be seen in this experiments, training the model for less than 2 epochs gi
|
|
63 |
|---|---|
|
64 |
| 25.27 | 25.21 |
|
65 |
|
66 |
-
as can be seen [here](
|
67 |
|
68 |
-
|
|
|
|
|
|
|
|
|
|
|
|
63 |
|---|---|
|
64 |
| 25.27 | 25.21 |
|
65 |
|
66 |
+
as can be seen [here](https://huggingface.co/patrickvonplaten/ami-wav2vec2-large-lv60).
|
67 |
|
68 |
+
The results are in-line with results of published papers:
|
69 |
+
|
70 |
+
- [*Hybrid acoustic models for distant and multichannel large vocabulary speech recognition*](https://www.researchgate.net/publication/258075865_Hybrid_acoustic_models_for_distant_and_multichannel_large_vocabulary_speech_recognition)
|
71 |
+
- [Multi-Span Acoustic Modelling using Raw Waveform Signals](https://arxiv.org/abs/1906.11047)
|
72 |
+
|
73 |
+
You can run [run.sh](https://huggingface.co/patrickvonplaten/ami-wav2vec2-large-lv60/blob/main/run.sh) to reproduce the result.
|