Model save
Browse files
README.md
ADDED
|
@@ -0,0 +1,85 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: transformers
|
| 3 |
+
license: cc-by-nc-4.0
|
| 4 |
+
base_model: facebook/mms-1b-all
|
| 5 |
+
tags:
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
metrics:
|
| 8 |
+
- wer
|
| 9 |
+
model-index:
|
| 10 |
+
- name: mms-1b-nyagen-combined-model
|
| 11 |
+
results: []
|
| 12 |
+
---
|
| 13 |
+
|
| 14 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 15 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 16 |
+
|
| 17 |
+
# mms-1b-nyagen-combined-model
|
| 18 |
+
|
| 19 |
+
This model is a fine-tuned version of [facebook/mms-1b-all](https://huggingface.co/facebook/mms-1b-all) on an unknown dataset.
|
| 20 |
+
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 0.1767
|
| 22 |
+
- Wer: 0.2447
|
| 23 |
+
|
| 24 |
+
## Model description
|
| 25 |
+
|
| 26 |
+
More information needed
|
| 27 |
+
|
| 28 |
+
## Intended uses & limitations
|
| 29 |
+
|
| 30 |
+
More information needed
|
| 31 |
+
|
| 32 |
+
## Training and evaluation data
|
| 33 |
+
|
| 34 |
+
More information needed
|
| 35 |
+
|
| 36 |
+
## Training procedure
|
| 37 |
+
|
| 38 |
+
### Training hyperparameters
|
| 39 |
+
|
| 40 |
+
The following hyperparameters were used during training:
|
| 41 |
+
- learning_rate: 0.0003
|
| 42 |
+
- train_batch_size: 4
|
| 43 |
+
- eval_batch_size: 4
|
| 44 |
+
- seed: 42
|
| 45 |
+
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
|
| 46 |
+
- lr_scheduler_type: linear
|
| 47 |
+
- lr_scheduler_warmup_steps: 100
|
| 48 |
+
- num_epochs: 30.0
|
| 49 |
+
- mixed_precision_training: Native AMP
|
| 50 |
+
|
| 51 |
+
### Training results
|
| 52 |
+
|
| 53 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 54 |
+
|:-------------:|:------:|:----:|:---------------:|:------:|
|
| 55 |
+
| 6.9978 | 0.1364 | 100 | 0.6384 | 0.5015 |
|
| 56 |
+
| 0.482 | 0.2729 | 200 | 0.2777 | 0.3713 |
|
| 57 |
+
| 0.3907 | 0.4093 | 300 | 0.2484 | 0.3481 |
|
| 58 |
+
| 0.3782 | 0.5457 | 400 | 0.2290 | 0.3232 |
|
| 59 |
+
| 0.3316 | 0.6821 | 500 | 0.2222 | 0.3148 |
|
| 60 |
+
| 0.3158 | 0.8186 | 600 | 0.2127 | 0.3042 |
|
| 61 |
+
| 0.3199 | 0.9550 | 700 | 0.2106 | 0.2932 |
|
| 62 |
+
| 0.3223 | 1.0914 | 800 | 0.2013 | 0.2826 |
|
| 63 |
+
| 0.3075 | 1.2278 | 900 | 0.1975 | 0.2709 |
|
| 64 |
+
| 0.3015 | 1.3643 | 1000 | 0.1942 | 0.2762 |
|
| 65 |
+
| 0.3049 | 1.5007 | 1100 | 0.1895 | 0.2729 |
|
| 66 |
+
| 0.3029 | 1.6371 | 1200 | 0.1888 | 0.2718 |
|
| 67 |
+
| 0.2626 | 1.7735 | 1300 | 0.1866 | 0.2683 |
|
| 68 |
+
| 0.2803 | 1.9100 | 1400 | 0.1830 | 0.2615 |
|
| 69 |
+
| 0.2725 | 2.0464 | 1500 | 0.1814 | 0.2626 |
|
| 70 |
+
| 0.2732 | 2.1828 | 1600 | 0.1783 | 0.2641 |
|
| 71 |
+
| 0.249 | 2.3192 | 1700 | 0.1828 | 0.2560 |
|
| 72 |
+
| 0.2423 | 2.4557 | 1800 | 0.1762 | 0.2480 |
|
| 73 |
+
| 0.2668 | 2.5921 | 1900 | 0.1732 | 0.2458 |
|
| 74 |
+
| 0.2653 | 2.7285 | 2000 | 0.1727 | 0.2460 |
|
| 75 |
+
| 0.2614 | 2.8649 | 2100 | 0.1749 | 0.2533 |
|
| 76 |
+
| 0.2474 | 3.0014 | 2200 | 0.1733 | 0.2438 |
|
| 77 |
+
| 0.2317 | 3.1378 | 2300 | 0.1767 | 0.2447 |
|
| 78 |
+
|
| 79 |
+
|
| 80 |
+
### Framework versions
|
| 81 |
+
|
| 82 |
+
- Transformers 4.48.0.dev0
|
| 83 |
+
- Pytorch 2.5.1+cu124
|
| 84 |
+
- Datasets 3.2.0
|
| 85 |
+
- Tokenizers 0.21.0
|