End of training
Browse files
README.md
ADDED
@@ -0,0 +1,94 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
base_model: openai/whisper-large-v3
|
3 |
+
datasets:
|
4 |
+
- mozilla-foundation/common_voice_11_0
|
5 |
+
language:
|
6 |
+
- th
|
7 |
+
library_name: peft
|
8 |
+
license: apache-2.0
|
9 |
+
metrics:
|
10 |
+
- wer
|
11 |
+
- bleu
|
12 |
+
tags:
|
13 |
+
- generated_from_trainer
|
14 |
+
model-index:
|
15 |
+
- name: Whisper Large V3 Thai Lora - Magi Boss
|
16 |
+
results:
|
17 |
+
- task:
|
18 |
+
type: automatic-speech-recognition
|
19 |
+
name: Automatic Speech Recognition
|
20 |
+
dataset:
|
21 |
+
name: Common Voice 11.0
|
22 |
+
type: mozilla-foundation/common_voice_11_0
|
23 |
+
config: th
|
24 |
+
split: None
|
25 |
+
args: 'config: th, split: validation'
|
26 |
+
metrics:
|
27 |
+
- type: wer
|
28 |
+
value: 0.8100890207715133
|
29 |
+
name: Wer
|
30 |
+
- type: bleu
|
31 |
+
value: 8.739138222980559
|
32 |
+
name: Bleu
|
33 |
+
---
|
34 |
+
|
35 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
36 |
+
should probably proofread and complete it, then remove this comment. -->
|
37 |
+
|
38 |
+
# Whisper Large V3 Thai Lora - Magi Boss
|
39 |
+
|
40 |
+
This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) on the Common Voice 11.0 dataset.
|
41 |
+
It achieves the following results on the evaluation set:
|
42 |
+
- Loss: 0.1894
|
43 |
+
- Wer: 0.8101
|
44 |
+
- Ter: {'score': 81.00890207715133, 'num_edits': 546, 'ref_length': 674.0}
|
45 |
+
- Chrf: {'score': 87.48106146298329, 'char_order': 6, 'word_order': 0, 'beta': 2}
|
46 |
+
- Cer: 0.1041
|
47 |
+
- Bleu: 8.7391
|
48 |
+
- Average Suber: 0.8189
|
49 |
+
|
50 |
+
## Model description
|
51 |
+
|
52 |
+
More information needed
|
53 |
+
|
54 |
+
## Intended uses & limitations
|
55 |
+
|
56 |
+
More information needed
|
57 |
+
|
58 |
+
## Training and evaluation data
|
59 |
+
|
60 |
+
More information needed
|
61 |
+
|
62 |
+
## Training procedure
|
63 |
+
|
64 |
+
### Training hyperparameters
|
65 |
+
|
66 |
+
The following hyperparameters were used during training:
|
67 |
+
- learning_rate: 1e-05
|
68 |
+
- train_batch_size: 32
|
69 |
+
- eval_batch_size: 8
|
70 |
+
- seed: 42
|
71 |
+
- gradient_accumulation_steps: 4
|
72 |
+
- total_train_batch_size: 128
|
73 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
74 |
+
- lr_scheduler_type: linear
|
75 |
+
- lr_scheduler_warmup_steps: 25
|
76 |
+
- num_epochs: 1
|
77 |
+
- mixed_precision_training: Native AMP
|
78 |
+
|
79 |
+
### Training results
|
80 |
+
|
81 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer | Ter | Chrf | Cer | Bleu | Average Suber |
|
82 |
+
|:-------------:|:-----:|:----:|:---------------:|:------:|:-------------------------------------------------------------------:|:-------------------------------------------------------------------------:|:------:|:------:|:-------------:|
|
83 |
+
| 0.3053 | 0.32 | 50 | 0.2052 | 0.8457 | {'score': 84.56973293768546, 'num_edits': 570, 'ref_length': 674.0} | {'score': 86.69717825036054, 'char_order': 6, 'word_order': 0, 'beta': 2} | 0.1115 | 7.8703 | 0.8640 |
|
84 |
+
| 0.3752 | 0.64 | 100 | 0.1937 | 0.8323 | {'score': 83.23442136498517, 'num_edits': 561, 'ref_length': 674.0} | {'score': 86.98005540895491, 'char_order': 6, 'word_order': 0, 'beta': 2} | 0.1087 | 8.1510 | 0.8469 |
|
85 |
+
| 0.2794 | 0.96 | 150 | 0.1894 | 0.8101 | {'score': 81.00890207715133, 'num_edits': 546, 'ref_length': 674.0} | {'score': 87.48106146298329, 'char_order': 6, 'word_order': 0, 'beta': 2} | 0.1041 | 8.7391 | 0.8189 |
|
86 |
+
|
87 |
+
|
88 |
+
### Framework versions
|
89 |
+
|
90 |
+
- PEFT 0.12.1.dev0
|
91 |
+
- Transformers 4.45.0.dev0
|
92 |
+
- Pytorch 2.1.2
|
93 |
+
- Datasets 2.20.0
|
94 |
+
- Tokenizers 0.19.1
|
runs/Aug06_21-16-30_0992a6553fa1/events.out.tfevents.1722978995.0992a6553fa1.1460.0
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ca63f4408854e960bb0bcecb3c0c7e1ee89b328b73e6257b4563d437f7669e7c
|
3 |
+
size 40440
|