adrianSauer
commited on
Commit
•
907af20
1
Parent(s):
5601071
End of training
Browse files- README.md +17 -15
- generation_config.json +1 -1
README.md
CHANGED
@@ -1,7 +1,8 @@
|
|
1 |
---
|
2 |
language:
|
3 |
- gn
|
4 |
-
|
|
|
5 |
tags:
|
6 |
- generated_from_trainer
|
7 |
datasets:
|
@@ -23,7 +24,7 @@ model-index:
|
|
23 |
metrics:
|
24 |
- name: Wer
|
25 |
type: wer
|
26 |
-
value: 58.
|
27 |
---
|
28 |
|
29 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
@@ -31,10 +32,10 @@ should probably proofread and complete it, then remove this comment. -->
|
|
31 |
|
32 |
# Common Voice 16 - Guarani
|
33 |
|
34 |
-
This model is a fine-tuned version of [openai/whisper-base
|
35 |
It achieves the following results on the evaluation set:
|
36 |
-
- Loss: 0.
|
37 |
-
- Wer: 58.
|
38 |
|
39 |
## Model description
|
40 |
|
@@ -59,23 +60,24 @@ The following hyperparameters were used during training:
|
|
59 |
- seed: 42
|
60 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
61 |
- lr_scheduler_type: constant_with_warmup
|
|
|
62 |
- training_steps: 500
|
63 |
- mixed_precision_training: Native AMP
|
64 |
|
65 |
### Training results
|
66 |
|
67 |
-
| Training Loss | Epoch
|
68 |
-
|
69 |
-
|
|
70 |
-
| 0.
|
71 |
-
| 0.
|
72 |
-
| 0.
|
73 |
-
| 0.
|
74 |
|
75 |
|
76 |
### Framework versions
|
77 |
|
78 |
-
- Transformers 4.
|
79 |
- Pytorch 2.2.1+cu121
|
80 |
-
- Datasets 2.
|
81 |
-
- Tokenizers 0.
|
|
|
1 |
---
|
2 |
language:
|
3 |
- gn
|
4 |
+
license: apache-2.0
|
5 |
+
base_model: openai/whisper-base
|
6 |
tags:
|
7 |
- generated_from_trainer
|
8 |
datasets:
|
|
|
24 |
metrics:
|
25 |
- name: Wer
|
26 |
type: wer
|
27 |
+
value: 58.90378171373863
|
28 |
---
|
29 |
|
30 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
32 |
|
33 |
# Common Voice 16 - Guarani
|
34 |
|
35 |
+
This model is a fine-tuned version of [openai/whisper-base](https://huggingface.co/openai/whisper-base) on the Common Voice 16 dataset.
|
36 |
It achieves the following results on the evaluation set:
|
37 |
+
- Loss: 0.5927
|
38 |
+
- Wer: 58.9038
|
39 |
|
40 |
## Model description
|
41 |
|
|
|
60 |
- seed: 42
|
61 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
62 |
- lr_scheduler_type: constant_with_warmup
|
63 |
+
- lr_scheduler_warmup_steps: 50
|
64 |
- training_steps: 500
|
65 |
- mixed_precision_training: Native AMP
|
66 |
|
67 |
### Training results
|
68 |
|
69 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
70 |
+
|:-------------:|:------:|:----:|:---------------:|:-------:|
|
71 |
+
| 2.3403 | 1.0101 | 100 | 0.9377 | 81.1872 |
|
72 |
+
| 0.5917 | 2.0202 | 200 | 0.6872 | 69.3633 |
|
73 |
+
| 0.3348 | 3.0303 | 300 | 0.6128 | 65.4380 |
|
74 |
+
| 0.2026 | 4.0404 | 400 | 0.5885 | 61.0340 |
|
75 |
+
| 0.1237 | 5.0505 | 500 | 0.5927 | 58.9038 |
|
76 |
|
77 |
|
78 |
### Framework versions
|
79 |
|
80 |
+
- Transformers 4.40.0
|
81 |
- Pytorch 2.2.1+cu121
|
82 |
+
- Datasets 2.19.0
|
83 |
+
- Tokenizers 0.19.1
|
generation_config.json
CHANGED
@@ -252,5 +252,5 @@
|
|
252 |
"transcribe": 50359,
|
253 |
"translate": 50358
|
254 |
},
|
255 |
-
"transformers_version": "4.
|
256 |
}
|
|
|
252 |
"transcribe": 50359,
|
253 |
"translate": 50358
|
254 |
},
|
255 |
+
"transformers_version": "4.40.0"
|
256 |
}
|