Masaki Hattori
commited on
Commit
•
8faaf99
1
Parent(s):
6473a8d
jdrt_byclass_rinnna_hubert_asr_1
Browse files- README.md +95 -0
- config.json +72 -0
- preprocessor_config.json +9 -0
- pytorch_model.bin +3 -0
- training_args.bin +3 -0
README.md
ADDED
@@ -0,0 +1,95 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
base_model: rinna/japanese-hubert-base
|
4 |
+
tags:
|
5 |
+
- generated_from_trainer
|
6 |
+
metrics:
|
7 |
+
- wer
|
8 |
+
model-index:
|
9 |
+
- name: jdrt_byclass_rinnna_hubert_asr_1
|
10 |
+
results: []
|
11 |
+
---
|
12 |
+
|
13 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
14 |
+
should probably proofread and complete it, then remove this comment. -->
|
15 |
+
|
16 |
+
# jdrt_byclass_rinnna_hubert_asr_1
|
17 |
+
|
18 |
+
This model is a fine-tuned version of [rinna/japanese-hubert-base](https://huggingface.co/rinna/japanese-hubert-base) on the None dataset.
|
19 |
+
It achieves the following results on the evaluation set:
|
20 |
+
- Loss: 0.3647
|
21 |
+
- Wer: 0.4190
|
22 |
+
- Cer: 0.2827
|
23 |
+
|
24 |
+
## Model description
|
25 |
+
|
26 |
+
More information needed
|
27 |
+
|
28 |
+
## Intended uses & limitations
|
29 |
+
|
30 |
+
More information needed
|
31 |
+
|
32 |
+
## Training and evaluation data
|
33 |
+
|
34 |
+
More information needed
|
35 |
+
|
36 |
+
## Training procedure
|
37 |
+
|
38 |
+
### Training hyperparameters
|
39 |
+
|
40 |
+
The following hyperparameters were used during training:
|
41 |
+
- learning_rate: 0.0001
|
42 |
+
- train_batch_size: 256
|
43 |
+
- eval_batch_size: 256
|
44 |
+
- seed: 42
|
45 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
46 |
+
- lr_scheduler_type: cosine
|
47 |
+
- lr_scheduler_warmup_steps: 250
|
48 |
+
- num_epochs: 50
|
49 |
+
|
50 |
+
### Training results
|
51 |
+
|
52 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|
53 |
+
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|
|
54 |
+
| 10.589 | 1.0 | 53 | 5.6588 | 0.9156 | 0.9495 |
|
55 |
+
| 5.0974 | 2.0 | 106 | 4.0914 | 0.9156 | 0.9495 |
|
56 |
+
| 3.6701 | 3.0 | 159 | 3.1732 | 0.9156 | 0.9495 |
|
57 |
+
| 2.9285 | 4.0 | 212 | 2.7238 | 0.9156 | 0.9495 |
|
58 |
+
| 2.6943 | 5.0 | 265 | 2.6600 | 0.9156 | 0.9495 |
|
59 |
+
| 2.4567 | 6.0 | 318 | 2.2231 | 0.9960 | 0.9112 |
|
60 |
+
| 2.1447 | 7.0 | 371 | 1.9716 | 0.9960 | 0.9112 |
|
61 |
+
| 1.8452 | 8.0 | 424 | 1.5058 | 0.9062 | 0.7431 |
|
62 |
+
| 1.4358 | 9.0 | 477 | 1.0988 | 0.7370 | 0.5347 |
|
63 |
+
| 1.1898 | 10.0 | 530 | 0.9512 | 0.6981 | 0.5062 |
|
64 |
+
| 1.0261 | 11.0 | 583 | 0.8354 | 0.6510 | 0.4779 |
|
65 |
+
| 0.8371 | 12.0 | 636 | 0.7158 | 0.5560 | 0.3784 |
|
66 |
+
| 0.7896 | 13.0 | 689 | 0.6381 | 0.5330 | 0.3686 |
|
67 |
+
| 0.6846 | 14.0 | 742 | 0.5720 | 0.5183 | 0.3555 |
|
68 |
+
| 0.6357 | 15.0 | 795 | 0.5879 | 0.5030 | 0.3505 |
|
69 |
+
| 0.5893 | 16.0 | 848 | 0.5501 | 0.4884 | 0.3468 |
|
70 |
+
| 0.558 | 17.0 | 901 | 0.4291 | 0.4487 | 0.3154 |
|
71 |
+
| 0.5019 | 18.0 | 954 | 0.4354 | 0.4552 | 0.3064 |
|
72 |
+
| 0.4784 | 19.0 | 1007 | 0.4199 | 0.4490 | 0.3014 |
|
73 |
+
| 0.4564 | 20.0 | 1060 | 0.4439 | 0.4508 | 0.3153 |
|
74 |
+
| 0.4291 | 21.0 | 1113 | 0.4143 | 0.4352 | 0.2845 |
|
75 |
+
| 0.4144 | 22.0 | 1166 | 0.4415 | 0.4384 | 0.2812 |
|
76 |
+
| 0.3766 | 23.0 | 1219 | 0.3706 | 0.4264 | 0.2918 |
|
77 |
+
| 0.3792 | 24.0 | 1272 | 0.3933 | 0.4377 | 0.3015 |
|
78 |
+
| 0.3759 | 25.0 | 1325 | 0.3708 | 0.4231 | 0.3023 |
|
79 |
+
| 0.337 | 26.0 | 1378 | 0.3762 | 0.4250 | 0.2942 |
|
80 |
+
| 0.3282 | 27.0 | 1431 | 0.3595 | 0.4253 | 0.2937 |
|
81 |
+
| 0.3174 | 28.0 | 1484 | 0.3998 | 0.4269 | 0.2898 |
|
82 |
+
| 0.3156 | 29.0 | 1537 | 0.4056 | 0.4268 | 0.3093 |
|
83 |
+
| 0.2921 | 30.0 | 1590 | 0.3694 | 0.4274 | 0.3041 |
|
84 |
+
| 0.2929 | 31.0 | 1643 | 0.3917 | 0.4228 | 0.2881 |
|
85 |
+
| 0.2686 | 32.0 | 1696 | 0.3880 | 0.4276 | 0.2969 |
|
86 |
+
| 0.2776 | 33.0 | 1749 | 0.4038 | 0.4273 | 0.2963 |
|
87 |
+
| 0.2619 | 34.0 | 1802 | 0.3647 | 0.4190 | 0.2827 |
|
88 |
+
|
89 |
+
|
90 |
+
### Framework versions
|
91 |
+
|
92 |
+
- Transformers 4.34.0.dev0
|
93 |
+
- Pytorch 2.0.1+cu118
|
94 |
+
- Datasets 2.14.5
|
95 |
+
- Tokenizers 0.13.3
|
config.json
ADDED
@@ -0,0 +1,72 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "rinna/japanese-hubert-base",
|
3 |
+
"activation_dropout": 0.1,
|
4 |
+
"apply_spec_augment": true,
|
5 |
+
"architectures": [
|
6 |
+
"HubertForCTC"
|
7 |
+
],
|
8 |
+
"attention_dropout": 0.1,
|
9 |
+
"bos_token_id": 1,
|
10 |
+
"classifier_proj_size": 256,
|
11 |
+
"conv_bias": false,
|
12 |
+
"conv_dim": [
|
13 |
+
512,
|
14 |
+
512,
|
15 |
+
512,
|
16 |
+
512,
|
17 |
+
512,
|
18 |
+
512,
|
19 |
+
512
|
20 |
+
],
|
21 |
+
"conv_kernel": [
|
22 |
+
10,
|
23 |
+
3,
|
24 |
+
3,
|
25 |
+
3,
|
26 |
+
3,
|
27 |
+
2,
|
28 |
+
2
|
29 |
+
],
|
30 |
+
"conv_stride": [
|
31 |
+
5,
|
32 |
+
2,
|
33 |
+
2,
|
34 |
+
2,
|
35 |
+
2,
|
36 |
+
2,
|
37 |
+
2
|
38 |
+
],
|
39 |
+
"ctc_loss_reduction": "mean",
|
40 |
+
"ctc_zero_infinity": false,
|
41 |
+
"do_stable_layer_norm": false,
|
42 |
+
"eos_token_id": 2,
|
43 |
+
"feat_extract_activation": "gelu",
|
44 |
+
"feat_extract_norm": "group",
|
45 |
+
"feat_proj_dropout": 0.1,
|
46 |
+
"feat_proj_layer_norm": true,
|
47 |
+
"final_dropout": 0.1,
|
48 |
+
"hidden_act": "gelu",
|
49 |
+
"hidden_dropout": 0,
|
50 |
+
"hidden_size": 768,
|
51 |
+
"initializer_range": 0.02,
|
52 |
+
"intermediate_size": 3072,
|
53 |
+
"layer_norm_eps": 1e-05,
|
54 |
+
"layerdrop": 0.1,
|
55 |
+
"mask_feature_length": 10,
|
56 |
+
"mask_feature_min_masks": 0,
|
57 |
+
"mask_feature_prob": 0.0,
|
58 |
+
"mask_time_length": 10,
|
59 |
+
"mask_time_min_masks": 2,
|
60 |
+
"mask_time_prob": 0.05,
|
61 |
+
"model_type": "hubert",
|
62 |
+
"num_attention_heads": 12,
|
63 |
+
"num_conv_pos_embedding_groups": 16,
|
64 |
+
"num_conv_pos_embeddings": 128,
|
65 |
+
"num_feat_extract_layers": 7,
|
66 |
+
"num_hidden_layers": 12,
|
67 |
+
"pad_token_id": 28,
|
68 |
+
"torch_dtype": "float32",
|
69 |
+
"transformers_version": "4.34.0.dev0",
|
70 |
+
"use_weighted_layer_sum": false,
|
71 |
+
"vocab_size": 29
|
72 |
+
}
|
preprocessor_config.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"do_normalize": true,
|
3 |
+
"feature_extractor_type": "Wav2Vec2FeatureExtractor",
|
4 |
+
"feature_size": 1,
|
5 |
+
"padding_side": "right",
|
6 |
+
"padding_value": 0.0,
|
7 |
+
"return_attention_mask": true,
|
8 |
+
"sampling_rate": 16000
|
9 |
+
}
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1fbad066105339ed50405ab1b41a4056f95e7c65ecd3981306547fdc2c132bb5
|
3 |
+
size 377649057
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:effc66fb5f8dea434e199ed16151b4fa0912bbf14cb4fd4021a07b451e468e03
|
3 |
+
size 4027
|