Model save
Browse files- README.md +48 -2
- model.safetensors +1 -1
README.md
CHANGED
@@ -3,9 +3,26 @@ license: apache-2.0
|
|
3 |
base_model: openai/whisper-medium
|
4 |
tags:
|
5 |
- generated_from_trainer
|
|
|
|
|
|
|
|
|
6 |
model-index:
|
7 |
- name: whisper-medium-ach-only
|
8 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
9 |
---
|
10 |
|
11 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
@@ -14,7 +31,10 @@ should probably proofread and complete it, then remove this comment. -->
|
|
14 |
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/bakera-sunbird/huggingface/runs/rycj9ija)
|
15 |
# whisper-medium-ach-only
|
16 |
|
17 |
-
This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on
|
|
|
|
|
|
|
18 |
|
19 |
## Model description
|
20 |
|
@@ -43,6 +63,32 @@ The following hyperparameters were used during training:
|
|
43 |
- training_steps: 4000
|
44 |
- mixed_precision_training: Native AMP
|
45 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
46 |
### Framework versions
|
47 |
|
48 |
- Transformers 4.41.0.dev0
|
|
|
3 |
base_model: openai/whisper-medium
|
4 |
tags:
|
5 |
- generated_from_trainer
|
6 |
+
datasets:
|
7 |
+
- generator
|
8 |
+
metrics:
|
9 |
+
- wer
|
10 |
model-index:
|
11 |
- name: whisper-medium-ach-only
|
12 |
+
results:
|
13 |
+
- task:
|
14 |
+
name: Automatic Speech Recognition
|
15 |
+
type: automatic-speech-recognition
|
16 |
+
dataset:
|
17 |
+
name: generator
|
18 |
+
type: generator
|
19 |
+
config: default
|
20 |
+
split: train
|
21 |
+
args: default
|
22 |
+
metrics:
|
23 |
+
- name: Wer
|
24 |
+
type: wer
|
25 |
+
value: 21.152030217186024
|
26 |
---
|
27 |
|
28 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
31 |
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/bakera-sunbird/huggingface/runs/rycj9ija)
|
32 |
# whisper-medium-ach-only
|
33 |
|
34 |
+
This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on the generator dataset.
|
35 |
+
It achieves the following results on the evaluation set:
|
36 |
+
- Loss: 0.3942
|
37 |
+
- Wer: 21.1520
|
38 |
|
39 |
## Model description
|
40 |
|
|
|
63 |
- training_steps: 4000
|
64 |
- mixed_precision_training: Native AMP
|
65 |
|
66 |
+
### Training results
|
67 |
+
|
68 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
69 |
+
|:-------------:|:-------:|:----:|:---------------:|:-------:|
|
70 |
+
| 1.1662 | 0.05 | 200 | 0.7062 | 47.0255 |
|
71 |
+
| 0.6014 | 1.0248 | 400 | 0.4606 | 32.9556 |
|
72 |
+
| 0.5638 | 1.0748 | 600 | 0.4021 | 27.2899 |
|
73 |
+
| 0.3677 | 2.0495 | 800 | 0.3736 | 24.3626 |
|
74 |
+
| 0.2711 | 3.0242 | 1000 | 0.3648 | 23.5127 |
|
75 |
+
| 0.2862 | 3.0743 | 1200 | 0.3402 | 23.7016 |
|
76 |
+
| 0.2023 | 4.049 | 1400 | 0.3665 | 22.4740 |
|
77 |
+
| 0.1166 | 5.0237 | 1600 | 0.4023 | 23.6072 |
|
78 |
+
| 0.1089 | 5.0738 | 1800 | 0.3871 | 22.5685 |
|
79 |
+
| 0.0859 | 6.0485 | 2000 | 0.3837 | 25.6846 |
|
80 |
+
| 0.0557 | 7.0232 | 2200 | 0.3942 | 21.1520 |
|
81 |
+
| 0.0572 | 7.0732 | 2400 | 0.3805 | 22.0963 |
|
82 |
+
| 0.0469 | 8.048 | 2600 | 0.3995 | 23.6072 |
|
83 |
+
| 0.0308 | 9.0228 | 2800 | 0.4057 | 21.5297 |
|
84 |
+
| 0.0288 | 9.0727 | 3000 | 0.3999 | 21.1520 |
|
85 |
+
| 0.0222 | 10.0475 | 3200 | 0.4121 | 21.7186 |
|
86 |
+
| 0.0239 | 11.0222 | 3400 | 0.4162 | 21.9075 |
|
87 |
+
| 0.024 | 11.0723 | 3600 | 0.4154 | 21.9075 |
|
88 |
+
| 0.0219 | 12.047 | 3800 | 0.4186 | 21.3409 |
|
89 |
+
| 0.0133 | 13.0218 | 4000 | 0.4173 | 21.1520 |
|
90 |
+
|
91 |
+
|
92 |
### Framework versions
|
93 |
|
94 |
- Transformers 4.41.0.dev0
|
model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 3055544304
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3db6ae4274bf41b470d31097c9da1f7deb5ca4706772000fa50e779854afd169
|
3 |
size 3055544304
|