End of training
Browse files- README.md +19 -39
- adapter_model.safetensors +1 -1
- runs/Nov02_15-35-40_a31dfab43883/events.out.tfevents.1730561792.a31dfab43883.8001.0 +3 -0
- runs/Nov02_15-37-35_a31dfab43883/events.out.tfevents.1730561865.a31dfab43883.8763.0 +3 -0
- runs/Nov02_15-38-55_a31dfab43883/events.out.tfevents.1730561950.a31dfab43883.8929.0 +3 -0
- training_args.bin +1 -1
README.md
CHANGED
@@ -1,4 +1,5 @@
|
|
1 |
---
|
|
|
2 |
language:
|
3 |
- it
|
4 |
license: apache-2.0
|
@@ -9,7 +10,6 @@ datasets:
|
|
9 |
- ASR_Synthetic_Speecht5_TTS
|
10 |
metrics:
|
11 |
- wer
|
12 |
-
library_name: peft
|
13 |
model-index:
|
14 |
- name: Whisper Medium
|
15 |
results:
|
@@ -24,7 +24,7 @@ model-index:
|
|
24 |
args: default
|
25 |
metrics:
|
26 |
- type: wer
|
27 |
-
value:
|
28 |
name: Wer
|
29 |
---
|
30 |
|
@@ -33,10 +33,10 @@ should probably proofread and complete it, then remove this comment. -->
|
|
33 |
|
34 |
# Whisper Medium
|
35 |
|
36 |
-
This model is a fine-tuned version of [
|
37 |
It achieves the following results on the evaluation set:
|
38 |
-
- Loss: 2.
|
39 |
-
- Wer:
|
40 |
|
41 |
## Model description
|
42 |
|
@@ -56,55 +56,35 @@ More information needed
|
|
56 |
|
57 |
The following hyperparameters were used during training:
|
58 |
- learning_rate: 0.001
|
59 |
-
- train_batch_size:
|
60 |
-
- eval_batch_size:
|
61 |
- seed: 42
|
|
|
|
|
62 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
63 |
- lr_scheduler_type: linear
|
64 |
- lr_scheduler_warmup_steps: 50
|
65 |
-
-
|
66 |
- mixed_precision_training: Native AMP
|
67 |
|
68 |
### Training results
|
69 |
|
70 |
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
71 |
|:-------------:|:------:|:----:|:---------------:|:--------:|
|
72 |
-
|
|
73 |
-
|
|
74 |
-
| 1.
|
75 |
-
|
|
76 |
-
|
|
77 |
-
|
|
78 |
-
|
|
79 |
-
|
|
80 |
-
| 0.4472 | 0.8789 | 450 | 2.6161 | 92.1316 |
|
81 |
-
| 0.3825 | 0.9766 | 500 | 2.5222 | 75.8226 |
|
82 |
-
| 0.2964 | 1.0742 | 550 | 2.7815 | 156.7954 |
|
83 |
-
| 0.263 | 1.1719 | 600 | 2.7807 | 101.4306 |
|
84 |
-
| 0.2474 | 1.2695 | 650 | 2.4792 | 77.9685 |
|
85 |
-
| 0.2133 | 1.3672 | 700 | 2.5918 | 84.6924 |
|
86 |
-
| 0.2148 | 1.4648 | 750 | 2.7335 | 86.4092 |
|
87 |
-
| 0.1955 | 1.5625 | 800 | 2.5843 | 79.1130 |
|
88 |
-
| 0.1943 | 1.6602 | 850 | 2.5548 | 70.6724 |
|
89 |
-
| 0.1648 | 1.7578 | 900 | 2.5600 | 84.6924 |
|
90 |
-
| 0.1411 | 1.8555 | 950 | 2.5128 | 70.8155 |
|
91 |
-
| 0.156 | 1.9531 | 1000 | 2.6710 | 76.8240 |
|
92 |
-
| 0.1215 | 2.0508 | 1050 | 2.6387 | 68.0973 |
|
93 |
-
| 0.092 | 2.1484 | 1100 | 2.5359 | 73.3906 |
|
94 |
-
| 0.0896 | 2.2461 | 1150 | 2.5957 | 75.5365 |
|
95 |
-
| 0.0861 | 2.3438 | 1200 | 2.5860 | 78.6838 |
|
96 |
-
| 0.0827 | 2.4414 | 1250 | 2.5877 | 74.5351 |
|
97 |
-
| 0.0893 | 2.5391 | 1300 | 2.5659 | 74.5351 |
|
98 |
-
| 0.0815 | 2.6367 | 1350 | 2.5564 | 74.9642 |
|
99 |
-
| 0.0599 | 2.7344 | 1400 | 2.5302 | 71.3877 |
|
100 |
-
| 0.0621 | 2.8320 | 1450 | 2.5382 | 71.8169 |
|
101 |
-
| 0.0598 | 2.9297 | 1500 | 2.5412 | 71.8169 |
|
102 |
|
103 |
|
104 |
### Framework versions
|
105 |
|
106 |
- PEFT 0.13.2
|
107 |
-
- Transformers 4.
|
108 |
- Pytorch 2.2.0
|
109 |
- Datasets 3.1.0
|
110 |
- Tokenizers 0.19.1
|
|
|
1 |
---
|
2 |
+
library_name: peft
|
3 |
language:
|
4 |
- it
|
5 |
license: apache-2.0
|
|
|
10 |
- ASR_Synthetic_Speecht5_TTS
|
11 |
metrics:
|
12 |
- wer
|
|
|
13 |
model-index:
|
14 |
- name: Whisper Medium
|
15 |
results:
|
|
|
24 |
args: default
|
25 |
metrics:
|
26 |
- type: wer
|
27 |
+
value: 171.5307582260372
|
28 |
name: Wer
|
29 |
---
|
30 |
|
|
|
33 |
|
34 |
# Whisper Medium
|
35 |
|
36 |
+
This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on the ASR_Synthetic_Speecht5_TTS dataset.
|
37 |
It achieves the following results on the evaluation set:
|
38 |
+
- Loss: 2.9413
|
39 |
+
- Wer: 171.5308
|
40 |
|
41 |
## Model description
|
42 |
|
|
|
56 |
|
57 |
The following hyperparameters were used during training:
|
58 |
- learning_rate: 0.001
|
59 |
+
- train_batch_size: 4
|
60 |
+
- eval_batch_size: 4
|
61 |
- seed: 42
|
62 |
+
- gradient_accumulation_steps: 2
|
63 |
+
- total_train_batch_size: 8
|
64 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
65 |
- lr_scheduler_type: linear
|
66 |
- lr_scheduler_warmup_steps: 50
|
67 |
+
- training_steps: 200
|
68 |
- mixed_precision_training: Native AMP
|
69 |
|
70 |
### Training results
|
71 |
|
72 |
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
73 |
|:-------------:|:------:|:----:|:---------------:|:--------:|
|
74 |
+
| 6.8678 | 0.0244 | 25 | 4.4434 | 154.2203 |
|
75 |
+
| 2.6877 | 0.0489 | 50 | 3.4026 | 144.0629 |
|
76 |
+
| 1.8792 | 0.0733 | 75 | 3.2962 | 77.3963 |
|
77 |
+
| 1.5587 | 0.0978 | 100 | 3.2969 | 78.9700 |
|
78 |
+
| 1.4194 | 0.1222 | 125 | 2.9920 | 75.1073 |
|
79 |
+
| 1.2356 | 0.1467 | 150 | 2.9471 | 184.2632 |
|
80 |
+
| 1.1741 | 0.1711 | 175 | 2.9542 | 189.4134 |
|
81 |
+
| 1.0451 | 0.1956 | 200 | 2.9413 | 171.5308 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
82 |
|
83 |
|
84 |
### Framework versions
|
85 |
|
86 |
- PEFT 0.13.2
|
87 |
+
- Transformers 4.44.2
|
88 |
- Pytorch 2.2.0
|
89 |
- Datasets 3.1.0
|
90 |
- Tokenizers 0.19.1
|
adapter_model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 18915424
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:002912ddd11806afd2c9104a270119e1bfb83923eab43726cb25ce3bb21cdbec
|
3 |
size 18915424
|
runs/Nov02_15-35-40_a31dfab43883/events.out.tfevents.1730561792.a31dfab43883.8001.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f39ee68da94d4c73d847a3fbcf248278f86f5bf1ddcdb26ed3dd8c1f9e482f39
|
3 |
+
size 6733
|
runs/Nov02_15-37-35_a31dfab43883/events.out.tfevents.1730561865.a31dfab43883.8763.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e645fbdf79154587b225bbb7b43c30a0b09b9fb352f081505ce8e25ef628e0a8
|
3 |
+
size 6732
|
runs/Nov02_15-38-55_a31dfab43883/events.out.tfevents.1730561950.a31dfab43883.8929.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:55ef468e7b445fea8b16a5e3c7d920e2478ddae8968b85d9a1d5a192a663c67e
|
3 |
+
size 7458
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 5304
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:cbd761ce3ae3b878a3c0b074291501e5aafd5d9e9d57840ee210b306cbf65837
|
3 |
size 5304
|