Update README.md
Browse files
README.md
CHANGED
@@ -36,6 +36,12 @@ model-index:
|
|
36 |
|
37 |
Fine-tuned [facebook/wav2vec2-base-fr-voxpopuli-v2](https://huggingface.co/facebook/wav2vec2-base-fr-voxpopuli-v2) for **French speech-to-phoneme** using the train and validation splits of [Common Voice v13](https://huggingface.co/datasets/mozilla-foundation/common_voice_13_0).
|
38 |
|
|
|
|
|
|
|
|
|
|
|
|
|
39 |
- The model has been trained for 14 epochs on 4x2080 Ti GPUs using a ddp strategy and gradient-accumulation procedure (256 audios per update, corresponding roughly to 25 minutes of speech per update -> 2k updates per epoch)
|
40 |
- Learning rate schedule : Double Tri-state schedule
|
41 |
- Warmup from 1e-5 for 7% of total updates
|
@@ -47,5 +53,3 @@ Fine-tuned [facebook/wav2vec2-base-fr-voxpopuli-v2](https://huggingface.co/faceb
|
|
47 |
|
48 |
- The set of hyperparameters used for training are those detailed in Annex B and Table 6 of [wav2vec2 paper](https://arxiv.org/pdf/2006.11477.pdf).
|
49 |
|
50 |
-
When using this model, make sure that your speech input is **sampled at 16kHz**.
|
51 |
-
|
|
|
36 |
|
37 |
Fine-tuned [facebook/wav2vec2-base-fr-voxpopuli-v2](https://huggingface.co/facebook/wav2vec2-base-fr-voxpopuli-v2) for **French speech-to-phoneme** using the train and validation splits of [Common Voice v13](https://huggingface.co/datasets/mozilla-foundation/common_voice_13_0).
|
38 |
|
39 |
+
## Samplerate of audio
|
40 |
+
|
41 |
+
When using this model, make sure that your speech input is **sampled at 16kHz**.
|
42 |
+
|
43 |
+
## Training procedure details
|
44 |
+
|
45 |
- The model has been trained for 14 epochs on 4x2080 Ti GPUs using a ddp strategy and gradient-accumulation procedure (256 audios per update, corresponding roughly to 25 minutes of speech per update -> 2k updates per epoch)
|
46 |
- Learning rate schedule : Double Tri-state schedule
|
47 |
- Warmup from 1e-5 for 7% of total updates
|
|
|
53 |
|
54 |
- The set of hyperparameters used for training are those detailed in Annex B and Table 6 of [wav2vec2 paper](https://arxiv.org/pdf/2006.11477.pdf).
|
55 |
|
|
|
|