Training in progress, step 1250
Browse files
README.md
CHANGED
@@ -2,12 +2,12 @@
|
|
2 |
language:
|
3 |
- ko
|
4 |
license: apache-2.0
|
|
|
5 |
tags:
|
6 |
- hf-asr-leaderboard
|
7 |
- generated_from_trainer
|
8 |
datasets:
|
9 |
- hyojin99/EBRC
|
10 |
-
base_model: openai/whisper-base
|
11 |
model-index:
|
12 |
- name: ft_model
|
13 |
results: []
|
@@ -19,9 +19,6 @@ should probably proofread and complete it, then remove this comment. -->
|
|
19 |
# ft_model
|
20 |
|
21 |
This model is a fine-tuned version of [openai/whisper-base](https://huggingface.co/openai/whisper-base) on the EBRC dataset.
|
22 |
-
It achieves the following results on the evaluation set:
|
23 |
-
- Loss: 0.4127
|
24 |
-
- Cer: 18.2381
|
25 |
|
26 |
## Model description
|
27 |
|
@@ -40,34 +37,16 @@ More information needed
|
|
40 |
### Training hyperparameters
|
41 |
|
42 |
The following hyperparameters were used during training:
|
43 |
-
- learning_rate:
|
44 |
-
- train_batch_size:
|
45 |
-
- eval_batch_size:
|
46 |
- seed: 42
|
47 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
48 |
- lr_scheduler_type: linear
|
49 |
- lr_scheduler_warmup_steps: 50
|
50 |
-
- training_steps:
|
51 |
- mixed_precision_training: Native AMP
|
52 |
|
53 |
-
### Training results
|
54 |
-
|
55 |
-
| Training Loss | Epoch | Step | Validation Loss | Cer |
|
56 |
-
|:-------------:|:-----:|:-----:|:---------------:|:-------:|
|
57 |
-
| 0.5845 | 0.4 | 1000 | 0.5512 | 34.8150 |
|
58 |
-
| 0.4598 | 0.8 | 2000 | 0.4840 | 22.6069 |
|
59 |
-
| 0.3077 | 1.2 | 3000 | 0.4570 | 20.7128 |
|
60 |
-
| 0.3212 | 1.6 | 4000 | 0.4381 | 21.4198 |
|
61 |
-
| 0.3027 | 2.0 | 5000 | 0.4181 | 20.1164 |
|
62 |
-
| 0.219 | 2.4 | 6000 | 0.4180 | 19.6479 |
|
63 |
-
| 0.2373 | 2.8 | 7000 | 0.4089 | 18.6477 |
|
64 |
-
| 0.1342 | 3.2 | 8000 | 0.4127 | 18.3603 |
|
65 |
-
| 0.1601 | 3.6 | 9000 | 0.4104 | 18.4824 |
|
66 |
-
| 0.1489 | 4.0 | 10000 | 0.4084 | 18.0628 |
|
67 |
-
| 0.1308 | 4.4 | 11000 | 0.4134 | 18.2856 |
|
68 |
-
| 0.114 | 4.8 | 12000 | 0.4127 | 18.2381 |
|
69 |
-
|
70 |
-
|
71 |
### Framework versions
|
72 |
|
73 |
- Transformers 4.40.0.dev0
|
|
|
2 |
language:
|
3 |
- ko
|
4 |
license: apache-2.0
|
5 |
+
base_model: openai/whisper-base
|
6 |
tags:
|
7 |
- hf-asr-leaderboard
|
8 |
- generated_from_trainer
|
9 |
datasets:
|
10 |
- hyojin99/EBRC
|
|
|
11 |
model-index:
|
12 |
- name: ft_model
|
13 |
results: []
|
|
|
19 |
# ft_model
|
20 |
|
21 |
This model is a fine-tuned version of [openai/whisper-base](https://huggingface.co/openai/whisper-base) on the EBRC dataset.
|
|
|
|
|
|
|
22 |
|
23 |
## Model description
|
24 |
|
|
|
37 |
### Training hyperparameters
|
38 |
|
39 |
The following hyperparameters were used during training:
|
40 |
+
- learning_rate: 5e-05
|
41 |
+
- train_batch_size: 32
|
42 |
+
- eval_batch_size: 16
|
43 |
- seed: 42
|
44 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
45 |
- lr_scheduler_type: linear
|
46 |
- lr_scheduler_warmup_steps: 50
|
47 |
+
- training_steps: 7500
|
48 |
- mixed_precision_training: Native AMP
|
49 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
50 |
### Framework versions
|
51 |
|
52 |
- Transformers 4.40.0.dev0
|
model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 290403936
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:07769cc94b659062fca0a00d6cf27d708a1c14a98d4757b1daa4c591cc4593aa
|
3 |
size 290403936
|
runs/Mar22_10-52-59_dde68c4478c7/events.out.tfevents.1711104798.dde68c4478c7.2036.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:890814c7df51cdcca35fc494b15df8c7c62a81f54e7a4f54b74c4f45261a0d33
|
3 |
+
size 18895
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 5048
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:65071e29eb227a50e8c2a1c31c0df61c6f6b15737fc2e61d8e0ccbe4e4e0b843
|
3 |
size 5048
|