gglabs/stt-turbo-multilingual-test
Browse files- README.md +14 -24
- adapter_config.json +2 -2
- adapter_model.safetensors +1 -1
- runs/Aug16_05-02-49_ubuntu/events.out.tfevents.1755288169.ubuntu.863233.0 +3 -0
- runs/Aug16_05-03-55_ubuntu/events.out.tfevents.1755288235.ubuntu.871478.0 +3 -0
- runs/Aug16_05-05-24_ubuntu/events.out.tfevents.1755288325.ubuntu.880121.0 +3 -0
- runs/Aug16_05-08-11_ubuntu/events.out.tfevents.1755288491.ubuntu.897503.0 +3 -0
- runs/Aug16_05-10-49_ubuntu/events.out.tfevents.1755288649.ubuntu.912455.0 +3 -0
- runs/Aug16_05-13-02_ubuntu/events.out.tfevents.1755288783.ubuntu.926962.0 +3 -0
- runs/Aug16_05-15-33_ubuntu/events.out.tfevents.1755288934.ubuntu.941520.0 +3 -0
- training_args.bin +1 -1
README.md
CHANGED
@@ -9,19 +9,19 @@ tags:
|
|
9 |
metrics:
|
10 |
- wer
|
11 |
model-index:
|
12 |
-
- name: Whisper Turbo Multilingual
|
13 |
results: []
|
14 |
---
|
15 |
|
16 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
17 |
should probably proofread and complete it, then remove this comment. -->
|
18 |
|
19 |
-
# Whisper Turbo Multilingual
|
20 |
|
21 |
This model is a fine-tuned version of [openai/whisper-large-v3-turbo](https://huggingface.co/openai/whisper-large-v3-turbo) on the custom_multilingual dataset.
|
22 |
It achieves the following results on the evaluation set:
|
23 |
-
- Loss: 0.
|
24 |
-
- Wer:
|
25 |
|
26 |
## Model description
|
27 |
|
@@ -40,32 +40,22 @@ More information needed
|
|
40 |
### Training hyperparameters
|
41 |
|
42 |
The following hyperparameters were used during training:
|
43 |
-
- learning_rate: 0.
|
44 |
-
- train_batch_size:
|
45 |
-
- eval_batch_size:
|
46 |
- seed: 42
|
47 |
-
-
|
48 |
-
- total_train_batch_size: 64
|
49 |
-
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
|
50 |
- lr_scheduler_type: linear
|
51 |
-
- lr_scheduler_warmup_steps:
|
52 |
-
- training_steps:
|
53 |
- mixed_precision_training: Native AMP
|
54 |
|
55 |
### Training results
|
56 |
|
57 |
-
| Training Loss | Epoch
|
58 |
-
|
59 |
-
| 0.
|
60 |
-
| 0.
|
61 |
-
| 0.224 | 3.2913 | 1500 | 0.3503 | 17.5891 |
|
62 |
-
| 0.2132 | 4.3884 | 2000 | 0.3518 | 16.9210 |
|
63 |
-
| 0.1865 | 5.4855 | 2500 | 0.3571 | 16.2908 |
|
64 |
-
| 0.1661 | 6.5826 | 3000 | 0.3652 | 16.0491 |
|
65 |
-
| 0.1467 | 7.6796 | 3500 | 0.3692 | 16.4471 |
|
66 |
-
| 0.136 | 8.7767 | 4000 | 0.3762 | 15.9496 |
|
67 |
-
| 0.1229 | 9.8738 | 4500 | 0.3816 | 15.8453 |
|
68 |
-
| 0.1146 | 10.9709 | 5000 | 0.3860 | 15.9354 |
|
69 |
|
70 |
|
71 |
### Framework versions
|
|
|
9 |
metrics:
|
10 |
- wer
|
11 |
model-index:
|
12 |
+
- name: Whisper Turbo Multilingual
|
13 |
results: []
|
14 |
---
|
15 |
|
16 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
17 |
should probably proofread and complete it, then remove this comment. -->
|
18 |
|
19 |
+
# Whisper Turbo Multilingual
|
20 |
|
21 |
This model is a fine-tuned version of [openai/whisper-large-v3-turbo](https://huggingface.co/openai/whisper-large-v3-turbo) on the custom_multilingual dataset.
|
22 |
It achieves the following results on the evaluation set:
|
23 |
+
- Loss: 0.4791
|
24 |
+
- Wer: 18.75
|
25 |
|
26 |
## Model description
|
27 |
|
|
|
40 |
### Training hyperparameters
|
41 |
|
42 |
The following hyperparameters were used during training:
|
43 |
+
- learning_rate: 0.001
|
44 |
+
- train_batch_size: 2
|
45 |
+
- eval_batch_size: 2
|
46 |
- seed: 42
|
47 |
+
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
|
|
|
|
|
48 |
- lr_scheduler_type: linear
|
49 |
+
- lr_scheduler_warmup_steps: 5
|
50 |
+
- training_steps: 20
|
51 |
- mixed_precision_training: Native AMP
|
52 |
|
53 |
### Training results
|
54 |
|
55 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
56 |
+
|:-------------:|:-----:|:----:|:---------------:|:------:|
|
57 |
+
| 0.3929 | 1.0 | 10 | 0.7320 | 21.875 |
|
58 |
+
| 0.076 | 2.0 | 20 | 0.4791 | 18.75 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
59 |
|
60 |
|
61 |
### Framework versions
|
adapter_config.json
CHANGED
@@ -28,9 +28,9 @@
|
|
28 |
"revision": null,
|
29 |
"target_modules": [
|
30 |
"q_proj",
|
31 |
-
"
|
32 |
"v_proj",
|
33 |
-
"
|
34 |
],
|
35 |
"task_type": null,
|
36 |
"trainable_token_indices": null,
|
|
|
28 |
"revision": null,
|
29 |
"target_modules": [
|
30 |
"q_proj",
|
31 |
+
"o_proj",
|
32 |
"v_proj",
|
33 |
+
"k_proj"
|
34 |
],
|
35 |
"task_type": null,
|
36 |
"trainable_token_indices": null,
|
adapter_model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 78677456
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:227bca4c5b5125ea6832cd4d58e2ad75d97cd3d3e161a850a9c45a7de5ab59e8
|
3 |
size 78677456
|
runs/Aug16_05-02-49_ubuntu/events.out.tfevents.1755288169.ubuntu.863233.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:306154d2fea16bf6399d26b63a0f9aee1d1cc02f5ea03fc24361f2b39accb135
|
3 |
+
size 6394
|
runs/Aug16_05-03-55_ubuntu/events.out.tfevents.1755288235.ubuntu.871478.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4f2eec87cae71b9e2f2356a076752b79709d4edd4805f5f2df24ea6e13b7daee
|
3 |
+
size 6394
|
runs/Aug16_05-05-24_ubuntu/events.out.tfevents.1755288325.ubuntu.880121.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ab62afcb1338e4a7bd611e5433bb088f9b845bd142bb1bc6cce363ffdfb883de
|
3 |
+
size 6394
|
runs/Aug16_05-08-11_ubuntu/events.out.tfevents.1755288491.ubuntu.897503.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6d8bfd385acde7be3672eec940b415d0bd600302281e35c595431c07cf50befa
|
3 |
+
size 6394
|
runs/Aug16_05-10-49_ubuntu/events.out.tfevents.1755288649.ubuntu.912455.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:53e91fc2b9204c26c1341b65eac43774fcf514048409aa74db0671c2fd0ade18
|
3 |
+
size 6394
|
runs/Aug16_05-13-02_ubuntu/events.out.tfevents.1755288783.ubuntu.926962.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:252a902efa2861b602dcaefd0162e74be0d468264b1b578dedd8c5bc2826e997
|
3 |
+
size 6394
|
runs/Aug16_05-15-33_ubuntu/events.out.tfevents.1755288934.ubuntu.941520.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5e5b1ae2124631722ad8973e2e63f14135fa135ee58ebf6e42f14a445fb732f7
|
3 |
+
size 9436
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 5496
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:90e55d6a2722c603fe80e1892a5c0ad681bfeff6114717f4da5ec6d3cf336ce2
|
3 |
size 5496
|