End of training
Browse files- README.md +71 -0
- generation_config.json +14 -0
README.md
ADDED
@@ -0,0 +1,71 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
language:
|
3 |
+
- ymr
|
4 |
+
license: apache-2.0
|
5 |
+
base_model: openai/whisper-small
|
6 |
+
tags:
|
7 |
+
- generated_from_trainer
|
8 |
+
metrics:
|
9 |
+
- wer
|
10 |
+
model-index:
|
11 |
+
- name: leenag/Norm_KLuke_Med
|
12 |
+
results: []
|
13 |
+
---
|
14 |
+
|
15 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
16 |
+
should probably proofread and complete it, then remove this comment. -->
|
17 |
+
|
18 |
+
# leenag/Norm_KLuke_Med
|
19 |
+
|
20 |
+
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the Spoken Bible Corpus: Malasar dataset.
|
21 |
+
It achieves the following results on the evaluation set:
|
22 |
+
- Loss: 0.5144
|
23 |
+
- Wer: 44.9541
|
24 |
+
|
25 |
+
## Model description
|
26 |
+
|
27 |
+
More information needed
|
28 |
+
|
29 |
+
## Intended uses & limitations
|
30 |
+
|
31 |
+
More information needed
|
32 |
+
|
33 |
+
## Training and evaluation data
|
34 |
+
|
35 |
+
More information needed
|
36 |
+
|
37 |
+
## Training procedure
|
38 |
+
|
39 |
+
### Training hyperparameters
|
40 |
+
|
41 |
+
The following hyperparameters were used during training:
|
42 |
+
- learning_rate: 1e-05
|
43 |
+
- train_batch_size: 32
|
44 |
+
- eval_batch_size: 16
|
45 |
+
- seed: 42
|
46 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
47 |
+
- lr_scheduler_type: linear
|
48 |
+
- lr_scheduler_warmup_steps: 500
|
49 |
+
- training_steps: 2000
|
50 |
+
- mixed_precision_training: Native AMP
|
51 |
+
|
52 |
+
### Training results
|
53 |
+
|
54 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
55 |
+
|:-------------:|:-------:|:----:|:---------------:|:-------:|
|
56 |
+
| 0.0275 | 11.3636 | 250 | 0.3935 | 48.5092 |
|
57 |
+
| 0.0091 | 22.7273 | 500 | 0.4626 | 49.0252 |
|
58 |
+
| 0.0012 | 34.0909 | 750 | 0.4798 | 47.1330 |
|
59 |
+
| 0.0002 | 45.4545 | 1000 | 0.5015 | 46.3876 |
|
60 |
+
| 0.0001 | 56.8182 | 1250 | 0.4947 | 45.5849 |
|
61 |
+
| 0.0 | 68.1818 | 1500 | 0.5088 | 45.0688 |
|
62 |
+
| 0.0 | 79.5455 | 1750 | 0.5129 | 44.9541 |
|
63 |
+
| 0.0 | 90.9091 | 2000 | 0.5144 | 44.9541 |
|
64 |
+
|
65 |
+
|
66 |
+
### Framework versions
|
67 |
+
|
68 |
+
- Transformers 4.41.2
|
69 |
+
- Pytorch 2.0.1+cu117
|
70 |
+
- Datasets 2.16.0
|
71 |
+
- Tokenizers 0.19.1
|
generation_config.json
ADDED
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"begin_suppress_tokens": [
|
3 |
+
220,
|
4 |
+
50257
|
5 |
+
],
|
6 |
+
"bos_token_id": 50257,
|
7 |
+
"decoder_start_token_id": 50258,
|
8 |
+
"eos_token_id": 50257,
|
9 |
+
"max_length": 448,
|
10 |
+
"pad_token_id": 50257,
|
11 |
+
"suppress_tokens": [],
|
12 |
+
"transformers_version": "4.41.2",
|
13 |
+
"use_cache": false
|
14 |
+
}
|