Update stats.md
Browse files
stats.md
ADDED
|
@@ -0,0 +1,95 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language:
|
| 3 |
+
- 'no'
|
| 4 |
+
license: apache-2.0
|
| 5 |
+
base_model: NbAiLab/nb-whisper-large-v3-RC4
|
| 6 |
+
tags:
|
| 7 |
+
- audio
|
| 8 |
+
- asr
|
| 9 |
+
- automatic-speech-recognition
|
| 10 |
+
- hf-asr-leaderboard
|
| 11 |
+
model-index:
|
| 12 |
+
- name: nb-whisper-large-v0.7
|
| 13 |
+
results: []
|
| 14 |
+
---
|
| 15 |
+
|
| 16 |
+
<!-- This model card has been generated automatically according to the information Keras had access to. You should
|
| 17 |
+
probably proofread and complete it, then remove this comment. -->
|
| 18 |
+
|
| 19 |
+
# nb-whisper-large-v0.7
|
| 20 |
+
|
| 21 |
+
This model is a fine-tuned version of [NbAiLab/nb-whisper-large-v3-RC4](https://huggingface.co/NbAiLab/nb-whisper-large-v3-RC4) on the NbAiLab/ncc_speech_styling_v2 dataset.
|
| 22 |
+
It achieves the following results on the evaluation set:
|
| 23 |
+
- step: 49999
|
| 24 |
+
- validation_nst_loss: 0.4299
|
| 25 |
+
- train_loss: 0.4933
|
| 26 |
+
- validation_nst_wer: 2.1830
|
| 27 |
+
- validation_nst_cer: 0.6702
|
| 28 |
+
- validation_nst_exact_wer: 2.7220
|
| 29 |
+
- validation_nst_exact_cer: 0.7519
|
| 30 |
+
- validation_clean_stortinget_no_loss: 0.7253
|
| 31 |
+
- validation_clean_stortinget_no_wer: 8.9886
|
| 32 |
+
- validation_clean_stortinget_no_cer: 5.7594
|
| 33 |
+
- validation_clean_stortinget_no_exact_wer: 11.8515
|
| 34 |
+
- validation_clean_stortinget_no_exact_cer: 6.2132
|
| 35 |
+
|
| 36 |
+
## Model description
|
| 37 |
+
|
| 38 |
+
More information needed
|
| 39 |
+
|
| 40 |
+
## Intended uses & limitations
|
| 41 |
+
|
| 42 |
+
More information needed
|
| 43 |
+
|
| 44 |
+
## Training and evaluation data
|
| 45 |
+
|
| 46 |
+
More information needed
|
| 47 |
+
|
| 48 |
+
## Training procedure
|
| 49 |
+
|
| 50 |
+
### Training hyperparameters
|
| 51 |
+
|
| 52 |
+
The following hyperparameters were used during training:
|
| 53 |
+
- learning_rate: 7e-05
|
| 54 |
+
- lr_scheduler_type: linear
|
| 55 |
+
- per_device_train_batch_size: 8
|
| 56 |
+
- total_train_batch_size_per_node: 32
|
| 57 |
+
- total_train_batch_size: 1024
|
| 58 |
+
- total_optimization_steps: 50,000
|
| 59 |
+
- starting_optimization_step: None
|
| 60 |
+
- finishing_optimization_step: 50,000
|
| 61 |
+
- num_train_dataset_workers: 32
|
| 62 |
+
- num_hosts: 32
|
| 63 |
+
- total_num_training_examples: 51,200,000
|
| 64 |
+
- steps_per_epoch: 7798
|
| 65 |
+
- num_beams: None
|
| 66 |
+
- weight_decay: 0.01
|
| 67 |
+
- adam_beta1: 0.9
|
| 68 |
+
- adam_beta2: 0.98
|
| 69 |
+
- adam_epsilon: 1e-06
|
| 70 |
+
- dropout: True
|
| 71 |
+
- bpe_dropout_probability: 0.2
|
| 72 |
+
- activation_dropout_probability: 0.1
|
| 73 |
+
|
| 74 |
+
### Training results
|
| 75 |
+
|
| 76 |
+
| step | validation_nst_loss | train_loss | validation_nst_wer | validation_nst_cer | validation_nst_exact_wer | validation_nst_exact_cer | validation_clean_stortinget_no_loss | validation_clean_stortinget_no_wer | validation_clean_stortinget_no_cer | validation_clean_stortinget_no_exact_wer | validation_clean_stortinget_no_exact_cer |
|
| 77 |
+
|:-----:|:-------------------:|:----------:|:------------------:|:------------------:|:------------------------:|:------------------------:|:-----------------------------------:|:----------------------------------:|:----------------------------------:|:----------------------------------------:|:----------------------------------------:|
|
| 78 |
+
| 0 | 0.4271 | 0.9562 | 2.1721 | 0.6246 | 2.7056 | 0.7070 | 0.6866 | 8.5836 | 5.4517 | 11.4126 | 5.8853 |
|
| 79 |
+
| 5000 | 0.4400 | 0.5815 | 2.6621 | 0.7765 | 3.1629 | 0.8526 | 0.7085 | 9.1000 | 5.7626 | 12.1172 | 6.2354 |
|
| 80 |
+
| 10000 | 0.4377 | 0.5548 | 2.2701 | 0.6740 | 2.9016 | 0.7720 | 0.6845 | 9.2823 | 5.9461 | 12.1717 | 6.4073 |
|
| 81 |
+
| 15000 | 0.4332 | 0.5112 | 2.3246 | 0.6917 | 2.8799 | 0.7775 | 0.7101 | 9.1307 | 5.8030 | 11.9654 | 6.2408 |
|
| 82 |
+
| 20000 | 0.4345 | 0.5066 | 2.3518 | 0.7122 | 2.8962 | 0.7940 | 0.7083 | 9.0668 | 5.8133 | 11.9867 | 6.2755 |
|
| 83 |
+
| 25000 | 0.4315 | 0.4955 | 2.2266 | 0.6740 | 2.7873 | 0.7601 | 0.7034 | 9.0313 | 5.7971 | 11.9535 | 6.2588 |
|
| 84 |
+
| 30000 | 0.4332 | 0.4936 | 2.2429 | 0.6936 | 2.7764 | 0.7757 | 0.7110 | 8.9957 | 5.7534 | 11.8230 | 6.1968 |
|
| 85 |
+
| 35000 | 0.4311 | 0.4947 | 2.2102 | 0.6777 | 2.7438 | 0.7592 | 0.7138 | 9.0076 | 5.7879 | 11.8752 | 6.2463 |
|
| 86 |
+
| 40000 | 0.4305 | 0.5026 | 2.2048 | 0.6805 | 2.7492 | 0.7638 | 0.7259 | 8.9152 | 5.6809 | 11.7827 | 6.1356 |
|
| 87 |
+
| 45000 | 0.4309 | 0.4815 | 2.1612 | 0.6572 | 2.7111 | 0.7436 | 0.7293 | 9.0265 | 5.7800 | 11.9179 | 6.2404 |
|
| 88 |
+
| 50000 | 0.4299 | 0.4933 | 2.1830 | 0.6702 | 2.7220 | 0.7519 | 0.7253 | 8.9886 | 5.7594 | 11.8515 | 6.2132 |
|
| 89 |
+
|
| 90 |
+
|
| 91 |
+
### Framework versions
|
| 92 |
+
|
| 93 |
+
- Transformers 4.35.2
|
| 94 |
+
- Datasets 2.15.0
|
| 95 |
+
- Tokenizers 0.14.1
|