Commit
·
c5c1c24
1
Parent(s):
9587eef
update model card README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,90 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
tags:
|
| 4 |
+
- speech-recognition
|
| 5 |
+
- librispeech_asr
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
model-index:
|
| 8 |
+
- name: hubert-librispeech-clean-100h-demo-dist
|
| 9 |
+
results: []
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 13 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 14 |
+
|
| 15 |
+
# hubert-librispeech-clean-100h-demo-dist
|
| 16 |
+
|
| 17 |
+
This model is a fine-tuned version of [facebook/hubert-large-ll60k](https://huggingface.co/facebook/hubert-large-ll60k) on the LIBRISPEECH_ASR - CLEAN dataset.
|
| 18 |
+
It achieves the following results on the evaluation set:
|
| 19 |
+
- Loss: 0.0984
|
| 20 |
+
- Wer: 0.0883
|
| 21 |
+
|
| 22 |
+
## Model description
|
| 23 |
+
|
| 24 |
+
More information needed
|
| 25 |
+
|
| 26 |
+
## Intended uses & limitations
|
| 27 |
+
|
| 28 |
+
More information needed
|
| 29 |
+
|
| 30 |
+
## Training and evaluation data
|
| 31 |
+
|
| 32 |
+
More information needed
|
| 33 |
+
|
| 34 |
+
## Training procedure
|
| 35 |
+
|
| 36 |
+
### Training hyperparameters
|
| 37 |
+
|
| 38 |
+
The following hyperparameters were used during training:
|
| 39 |
+
- learning_rate: 0.0003
|
| 40 |
+
- train_batch_size: 4
|
| 41 |
+
- eval_batch_size: 8
|
| 42 |
+
- seed: 42
|
| 43 |
+
- distributed_type: multi-GPU
|
| 44 |
+
- num_devices: 8
|
| 45 |
+
- total_train_batch_size: 32
|
| 46 |
+
- total_eval_batch_size: 64
|
| 47 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 48 |
+
- lr_scheduler_type: linear
|
| 49 |
+
- lr_scheduler_warmup_steps: 500
|
| 50 |
+
- num_epochs: 3.0
|
| 51 |
+
- mixed_precision_training: Native AMP
|
| 52 |
+
|
| 53 |
+
### Training results
|
| 54 |
+
|
| 55 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 56 |
+
|:-------------:|:-----:|:----:|:---------------:|:------:|
|
| 57 |
+
| 2.9031 | 0.11 | 100 | 2.9220 | 1.0 |
|
| 58 |
+
| 2.6437 | 0.22 | 200 | 2.6268 | 1.0 |
|
| 59 |
+
| 0.3934 | 0.34 | 300 | 0.4860 | 0.4182 |
|
| 60 |
+
| 0.3531 | 0.45 | 400 | 0.3088 | 0.2894 |
|
| 61 |
+
| 0.2255 | 0.56 | 500 | 0.2568 | 0.2426 |
|
| 62 |
+
| 0.3379 | 0.67 | 600 | 0.2073 | 0.2011 |
|
| 63 |
+
| 0.2419 | 0.78 | 700 | 0.1849 | 0.1838 |
|
| 64 |
+
| 0.2128 | 0.9 | 800 | 0.1662 | 0.1690 |
|
| 65 |
+
| 0.1341 | 1.01 | 900 | 0.1600 | 0.1541 |
|
| 66 |
+
| 0.0946 | 1.12 | 1000 | 0.1431 | 0.1404 |
|
| 67 |
+
| 0.1643 | 1.23 | 1100 | 0.1373 | 0.1304 |
|
| 68 |
+
| 0.0663 | 1.35 | 1200 | 0.1293 | 0.1307 |
|
| 69 |
+
| 0.162 | 1.46 | 1300 | 0.1247 | 0.1266 |
|
| 70 |
+
| 0.1433 | 1.57 | 1400 | 0.1246 | 0.1262 |
|
| 71 |
+
| 0.1581 | 1.68 | 1500 | 0.1219 | 0.1154 |
|
| 72 |
+
| 0.1036 | 1.79 | 1600 | 0.1127 | 0.1081 |
|
| 73 |
+
| 0.1352 | 1.91 | 1700 | 0.1087 | 0.1040 |
|
| 74 |
+
| 0.0471 | 2.02 | 1800 | 0.1085 | 0.1005 |
|
| 75 |
+
| 0.0945 | 2.13 | 1900 | 0.1066 | 0.0973 |
|
| 76 |
+
| 0.0843 | 2.24 | 2000 | 0.1102 | 0.0964 |
|
| 77 |
+
| 0.0774 | 2.35 | 2100 | 0.1079 | 0.0940 |
|
| 78 |
+
| 0.0952 | 2.47 | 2200 | 0.1056 | 0.0927 |
|
| 79 |
+
| 0.0635 | 2.58 | 2300 | 0.1026 | 0.0920 |
|
| 80 |
+
| 0.0665 | 2.69 | 2400 | 0.1012 | 0.0905 |
|
| 81 |
+
| 0.034 | 2.8 | 2500 | 0.1009 | 0.0900 |
|
| 82 |
+
| 0.0251 | 2.91 | 2600 | 0.0993 | 0.0883 |
|
| 83 |
+
|
| 84 |
+
|
| 85 |
+
### Framework versions
|
| 86 |
+
|
| 87 |
+
- Transformers 4.11.0.dev0
|
| 88 |
+
- Pytorch 1.9.0+cu111
|
| 89 |
+
- Datasets 1.12.1
|
| 90 |
+
- Tokenizers 0.10.3
|