Commit
•
cdaec1d
1
Parent(s):
8f6cf26
Model save
Browse files
README.md
CHANGED
@@ -2,8 +2,6 @@
|
|
2 |
license: apache-2.0
|
3 |
base_model: facebook/wav2vec2-large-xlsr-53
|
4 |
tags:
|
5 |
-
- automatic-speech-recognition
|
6 |
-
- DewiBrynJones/banc-trawsgrifiadau-bangor-clean-with-ccv
|
7 |
- generated_from_trainer
|
8 |
metrics:
|
9 |
- wer
|
@@ -17,10 +15,10 @@ should probably proofread and complete it, then remove this comment. -->
|
|
17 |
|
18 |
# wav2vec2-xlsr-53-ft-btb-ccv-cy
|
19 |
|
20 |
-
This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on
|
21 |
It achieves the following results on the evaluation set:
|
22 |
-
- Loss: 0.
|
23 |
-
- Wer: 0.
|
24 |
|
25 |
## Model description
|
26 |
|
@@ -46,39 +44,63 @@ The following hyperparameters were used during training:
|
|
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:
|
50 |
- mixed_precision_training: Native AMP
|
51 |
|
52 |
### Training results
|
53 |
|
54 |
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
55 |
|:-------------:|:------:|:----:|:---------------:|:------:|
|
56 |
-
| No log | 0.0566 | 100 | 3.
|
57 |
-
| No log | 0.1133 | 200 | 3.
|
58 |
-
| No log | 0.1699 | 300 |
|
59 |
-
| No log | 0.2265 | 400 |
|
60 |
-
| 3.
|
61 |
-
| 3.
|
62 |
-
| 3.
|
63 |
-
| 3.
|
64 |
-
| 3.
|
65 |
-
| 0.
|
66 |
-
| 0.
|
67 |
-
| 0.
|
68 |
-
| 0.
|
69 |
-
| 0.
|
70 |
-
| 0.
|
71 |
-
| 0.
|
72 |
-
| 0.
|
73 |
-
| 0.
|
74 |
-
| 0.
|
75 |
-
| 0.
|
76 |
-
| 0.
|
77 |
-
| 0.
|
78 |
-
| 0.
|
79 |
-
| 0.
|
80 |
-
| 0.
|
81 |
-
| 0.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
82 |
|
83 |
|
84 |
### Framework versions
|
|
|
2 |
license: apache-2.0
|
3 |
base_model: facebook/wav2vec2-large-xlsr-53
|
4 |
tags:
|
|
|
|
|
5 |
- generated_from_trainer
|
6 |
metrics:
|
7 |
- wer
|
|
|
15 |
|
16 |
# wav2vec2-xlsr-53-ft-btb-ccv-cy
|
17 |
|
18 |
+
This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on an unknown dataset.
|
19 |
It achieves the following results on the evaluation set:
|
20 |
+
- Loss: 0.4404
|
21 |
+
- Wer: 0.3235
|
22 |
|
23 |
## Model description
|
24 |
|
|
|
44 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
45 |
- lr_scheduler_type: linear
|
46 |
- lr_scheduler_warmup_steps: 500
|
47 |
+
- training_steps: 5000
|
48 |
- mixed_precision_training: Native AMP
|
49 |
|
50 |
### Training results
|
51 |
|
52 |
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
53 |
|:-------------:|:------:|:----:|:---------------:|:------:|
|
54 |
+
| No log | 0.0566 | 100 | 3.5966 | 1.0 |
|
55 |
+
| No log | 0.1133 | 200 | 3.5808 | 1.0 |
|
56 |
+
| No log | 0.1699 | 300 | 2.2642 | 0.9884 |
|
57 |
+
| No log | 0.2265 | 400 | 1.0231 | 0.7426 |
|
58 |
+
| 3.6826 | 0.2831 | 500 | 0.9161 | 0.6457 |
|
59 |
+
| 3.6826 | 0.3398 | 600 | 0.7942 | 0.5827 |
|
60 |
+
| 3.6826 | 0.3964 | 700 | 0.7551 | 0.5417 |
|
61 |
+
| 3.6826 | 0.4530 | 800 | 0.7243 | 0.5352 |
|
62 |
+
| 3.6826 | 0.5096 | 900 | 0.6665 | 0.4985 |
|
63 |
+
| 0.4714 | 0.5663 | 1000 | 0.6403 | 0.4863 |
|
64 |
+
| 0.4714 | 0.6229 | 1100 | 0.6220 | 0.4727 |
|
65 |
+
| 0.4714 | 0.6795 | 1200 | 0.6100 | 0.4642 |
|
66 |
+
| 0.4714 | 0.7361 | 1300 | 0.5863 | 0.4442 |
|
67 |
+
| 0.4714 | 0.7928 | 1400 | 0.5668 | 0.4384 |
|
68 |
+
| 0.3661 | 0.8494 | 1500 | 0.5753 | 0.4421 |
|
69 |
+
| 0.3661 | 0.9060 | 1600 | 0.5641 | 0.4311 |
|
70 |
+
| 0.3661 | 0.9626 | 1700 | 0.5466 | 0.4153 |
|
71 |
+
| 0.3661 | 1.0193 | 1800 | 0.5304 | 0.4023 |
|
72 |
+
| 0.3661 | 1.0759 | 1900 | 0.5280 | 0.3938 |
|
73 |
+
| 0.3069 | 1.1325 | 2000 | 0.5220 | 0.3902 |
|
74 |
+
| 0.3069 | 1.1891 | 2100 | 0.5064 | 0.3929 |
|
75 |
+
| 0.3069 | 1.2458 | 2200 | 0.5033 | 0.3823 |
|
76 |
+
| 0.3069 | 1.3024 | 2300 | 0.4920 | 0.3773 |
|
77 |
+
| 0.3069 | 1.3590 | 2400 | 0.4955 | 0.3737 |
|
78 |
+
| 0.264 | 1.4156 | 2500 | 0.4955 | 0.3739 |
|
79 |
+
| 0.264 | 1.4723 | 2600 | 0.4846 | 0.3651 |
|
80 |
+
| 0.264 | 1.5289 | 2700 | 0.4761 | 0.3645 |
|
81 |
+
| 0.264 | 1.5855 | 2800 | 0.4672 | 0.3627 |
|
82 |
+
| 0.264 | 1.6421 | 2900 | 0.4645 | 0.3640 |
|
83 |
+
| 0.2498 | 1.6988 | 3000 | 0.4652 | 0.3589 |
|
84 |
+
| 0.2498 | 1.7554 | 3100 | 0.4596 | 0.3534 |
|
85 |
+
| 0.2498 | 1.8120 | 3200 | 0.4505 | 0.3510 |
|
86 |
+
| 0.2498 | 1.8686 | 3300 | 0.4490 | 0.3522 |
|
87 |
+
| 0.2498 | 1.9253 | 3400 | 0.4460 | 0.3456 |
|
88 |
+
| 0.2304 | 1.9819 | 3500 | 0.4440 | 0.3405 |
|
89 |
+
| 0.2304 | 2.0385 | 3600 | 0.4493 | 0.3425 |
|
90 |
+
| 0.2304 | 2.0951 | 3700 | 0.4386 | 0.3374 |
|
91 |
+
| 0.2304 | 2.1518 | 3800 | 0.4424 | 0.3381 |
|
92 |
+
| 0.2304 | 2.2084 | 3900 | 0.4385 | 0.3373 |
|
93 |
+
| 0.1947 | 2.2650 | 4000 | 0.4394 | 0.3355 |
|
94 |
+
| 0.1947 | 2.3216 | 4100 | 0.4354 | 0.3324 |
|
95 |
+
| 0.1947 | 2.3783 | 4200 | 0.4345 | 0.3325 |
|
96 |
+
| 0.1947 | 2.4349 | 4300 | 0.4343 | 0.3331 |
|
97 |
+
| 0.1947 | 2.4915 | 4400 | 0.4517 | 0.3349 |
|
98 |
+
| 0.1996 | 2.5481 | 4500 | 0.4571 | 0.3328 |
|
99 |
+
| 0.1996 | 2.6048 | 4600 | 0.4423 | 0.3255 |
|
100 |
+
| 0.1996 | 2.6614 | 4700 | 0.4405 | 0.3254 |
|
101 |
+
| 0.1996 | 2.7180 | 4800 | 0.4437 | 0.3238 |
|
102 |
+
| 0.1996 | 2.7746 | 4900 | 0.4401 | 0.3232 |
|
103 |
+
| 0.2024 | 2.8313 | 5000 | 0.4404 | 0.3235 |
|
104 |
|
105 |
|
106 |
### Framework versions
|
model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 1261996080
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:542c2e46a436d07be34394a9c80a389da699d88cfa58989e55c02333f0eaeec1
|
3 |
size 1261996080
|
runs/May27_13-44-18_5cce808c63ba/events.out.tfevents.1716813896.5cce808c63ba.274.0
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:f20588c57a1bf32da734a8aea6474ba1999264657a36cbe8ad6b35f2d2717e91
|
3 |
+
size 24795
|