End of training
Browse files- README.md +117 -0
- adapter_config.json +32 -0
- adapter_model.safetensors +3 -0
- preprocessor_config.json +14 -0
- runs/Oct20_14-48-06_ce081a3621f2/events.out.tfevents.1729435695.ce081a3621f2.6025.2 +3 -0
- training_args.bin +3 -0
README.md
ADDED
@@ -0,0 +1,117 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
library_name: peft
|
3 |
+
language:
|
4 |
+
- ms
|
5 |
+
license: apache-2.0
|
6 |
+
base_model: openai/whisper-small
|
7 |
+
tags:
|
8 |
+
- generated_from_trainer
|
9 |
+
datasets:
|
10 |
+
- clt013/malay-speech-3k-rows-dataset_v2
|
11 |
+
model-index:
|
12 |
+
- name: Whisper Small FT Malay - CLT013
|
13 |
+
results: []
|
14 |
+
---
|
15 |
+
|
16 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
17 |
+
should probably proofread and complete it, then remove this comment. -->
|
18 |
+
|
19 |
+
# Whisper Small FT Malay - CLT013
|
20 |
+
|
21 |
+
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the Malay Speech 3k dataset.
|
22 |
+
It achieves the following results on the evaluation set:
|
23 |
+
- Loss: 0.8613
|
24 |
+
|
25 |
+
## Model description
|
26 |
+
|
27 |
+
More information needed
|
28 |
+
|
29 |
+
## Intended uses & limitations
|
30 |
+
|
31 |
+
More information needed
|
32 |
+
|
33 |
+
## Training and evaluation data
|
34 |
+
|
35 |
+
More information needed
|
36 |
+
|
37 |
+
## Training procedure
|
38 |
+
|
39 |
+
### Training hyperparameters
|
40 |
+
|
41 |
+
The following hyperparameters were used during training:
|
42 |
+
- learning_rate: 0.001
|
43 |
+
- train_batch_size: 8
|
44 |
+
- eval_batch_size: 8
|
45 |
+
- seed: 42
|
46 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
47 |
+
- lr_scheduler_type: linear
|
48 |
+
- lr_scheduler_warmup_steps: 100
|
49 |
+
- num_epochs: 20
|
50 |
+
- mixed_precision_training: Native AMP
|
51 |
+
|
52 |
+
### Training results
|
53 |
+
|
54 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
55 |
+
|:-------------:|:-------:|:----:|:---------------:|
|
56 |
+
| 2.1842 | 0.3731 | 100 | 0.8172 |
|
57 |
+
| 0.7488 | 0.7463 | 200 | 0.8014 |
|
58 |
+
| 0.6424 | 1.1194 | 300 | 0.8136 |
|
59 |
+
| 0.5234 | 1.4925 | 400 | 0.7511 |
|
60 |
+
| 0.4951 | 1.8657 | 500 | 0.8203 |
|
61 |
+
| 0.3835 | 2.2388 | 600 | 0.8191 |
|
62 |
+
| 0.3519 | 2.6119 | 700 | 0.8001 |
|
63 |
+
| 0.3868 | 2.9851 | 800 | 0.8011 |
|
64 |
+
| 0.2568 | 3.3582 | 900 | 0.8630 |
|
65 |
+
| 0.2781 | 3.7313 | 1000 | 0.8269 |
|
66 |
+
| 0.2535 | 4.1045 | 1100 | 0.8612 |
|
67 |
+
| 0.2105 | 4.4776 | 1200 | 0.8486 |
|
68 |
+
| 0.2104 | 4.8507 | 1300 | 0.8367 |
|
69 |
+
| 0.1726 | 5.2239 | 1400 | 0.8692 |
|
70 |
+
| 0.1672 | 5.5970 | 1500 | 0.8483 |
|
71 |
+
| 0.1641 | 5.9701 | 1600 | 0.8443 |
|
72 |
+
| 0.1186 | 6.3433 | 1700 | 0.9531 |
|
73 |
+
| 0.1261 | 6.7164 | 1800 | 0.8578 |
|
74 |
+
| 0.1211 | 7.0896 | 1900 | 0.8922 |
|
75 |
+
| 0.0962 | 7.4627 | 2000 | 0.9107 |
|
76 |
+
| 0.1188 | 7.8358 | 2100 | 0.8498 |
|
77 |
+
| 0.0847 | 8.2090 | 2200 | 0.8554 |
|
78 |
+
| 0.0802 | 8.5821 | 2300 | 0.9024 |
|
79 |
+
| 0.0805 | 8.9552 | 2400 | 0.8649 |
|
80 |
+
| 0.0559 | 9.3284 | 2500 | 0.8634 |
|
81 |
+
| 0.053 | 9.7015 | 2600 | 0.8988 |
|
82 |
+
| 0.0555 | 10.0746 | 2700 | 0.8657 |
|
83 |
+
| 0.0415 | 10.4478 | 2800 | 0.8449 |
|
84 |
+
| 0.0401 | 10.8209 | 2900 | 0.8658 |
|
85 |
+
| 0.0318 | 11.1940 | 3000 | 0.8674 |
|
86 |
+
| 0.0245 | 11.5672 | 3100 | 0.8491 |
|
87 |
+
| 0.032 | 11.9403 | 3200 | 0.8694 |
|
88 |
+
| 0.0186 | 12.3134 | 3300 | 0.8620 |
|
89 |
+
| 0.0179 | 12.6866 | 3400 | 0.8555 |
|
90 |
+
| 0.015 | 13.0597 | 3500 | 0.8730 |
|
91 |
+
| 0.0176 | 13.4328 | 3600 | 0.8458 |
|
92 |
+
| 0.0155 | 13.8060 | 3700 | 0.8454 |
|
93 |
+
| 0.0121 | 14.1791 | 3800 | 0.8533 |
|
94 |
+
| 0.0139 | 14.5522 | 3900 | 0.8604 |
|
95 |
+
| 0.009 | 14.9254 | 4000 | 0.8676 |
|
96 |
+
| 0.0095 | 15.2985 | 4100 | 0.8649 |
|
97 |
+
| 0.0059 | 15.6716 | 4200 | 0.8728 |
|
98 |
+
| 0.0065 | 16.0448 | 4300 | 0.8570 |
|
99 |
+
| 0.0049 | 16.4179 | 4400 | 0.8521 |
|
100 |
+
| 0.0042 | 16.7910 | 4500 | 0.8600 |
|
101 |
+
| 0.0051 | 17.1642 | 4600 | 0.8741 |
|
102 |
+
| 0.0037 | 17.5373 | 4700 | 0.8666 |
|
103 |
+
| 0.0037 | 17.9104 | 4800 | 0.8691 |
|
104 |
+
| 0.0029 | 18.2836 | 4900 | 0.8619 |
|
105 |
+
| 0.0023 | 18.6567 | 5000 | 0.8603 |
|
106 |
+
| 0.0019 | 19.0299 | 5100 | 0.8629 |
|
107 |
+
| 0.0018 | 19.4030 | 5200 | 0.8608 |
|
108 |
+
| 0.0018 | 19.7761 | 5300 | 0.8613 |
|
109 |
+
|
110 |
+
|
111 |
+
### Framework versions
|
112 |
+
|
113 |
+
- PEFT 0.13.2
|
114 |
+
- Transformers 4.44.2
|
115 |
+
- Pytorch 2.4.1+cu121
|
116 |
+
- Datasets 3.0.1
|
117 |
+
- Tokenizers 0.19.1
|
adapter_config.json
ADDED
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"alpha_pattern": {},
|
3 |
+
"auto_mapping": {
|
4 |
+
"base_model_class": "WhisperForConditionalGeneration",
|
5 |
+
"parent_library": "transformers.models.whisper.modeling_whisper"
|
6 |
+
},
|
7 |
+
"base_model_name_or_path": "openai/whisper-small",
|
8 |
+
"bias": "none",
|
9 |
+
"fan_in_fan_out": false,
|
10 |
+
"inference_mode": true,
|
11 |
+
"init_lora_weights": true,
|
12 |
+
"layer_replication": null,
|
13 |
+
"layers_pattern": null,
|
14 |
+
"layers_to_transform": null,
|
15 |
+
"loftq_config": {},
|
16 |
+
"lora_alpha": 64,
|
17 |
+
"lora_dropout": 0.05,
|
18 |
+
"megatron_config": null,
|
19 |
+
"megatron_core": "megatron.core",
|
20 |
+
"modules_to_save": null,
|
21 |
+
"peft_type": "LORA",
|
22 |
+
"r": 32,
|
23 |
+
"rank_pattern": {},
|
24 |
+
"revision": null,
|
25 |
+
"target_modules": [
|
26 |
+
"q_proj",
|
27 |
+
"v_proj"
|
28 |
+
],
|
29 |
+
"task_type": null,
|
30 |
+
"use_dora": false,
|
31 |
+
"use_rslora": false
|
32 |
+
}
|
adapter_model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8317af7735099706c1fd3aed94e4e97c30cb6415de20c888484a43ea35c25d62
|
3 |
+
size 14176064
|
preprocessor_config.json
ADDED
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"chunk_length": 30,
|
3 |
+
"feature_extractor_type": "WhisperFeatureExtractor",
|
4 |
+
"feature_size": 80,
|
5 |
+
"hop_length": 160,
|
6 |
+
"n_fft": 400,
|
7 |
+
"n_samples": 480000,
|
8 |
+
"nb_max_frames": 3000,
|
9 |
+
"padding_side": "right",
|
10 |
+
"padding_value": 0.0,
|
11 |
+
"processor_class": "WhisperProcessor",
|
12 |
+
"return_attention_mask": false,
|
13 |
+
"sampling_rate": 16000
|
14 |
+
}
|
runs/Oct20_14-48-06_ce081a3621f2/events.out.tfevents.1729435695.ce081a3621f2.6025.2
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:2beb706c91d41d41d5693b2da82dffcb62a5d10148b020abbd74b26725c4da94
|
3 |
+
size 33121
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:19ddd4550daf6437bb68092373b28d269a6724f254d4b3455a4167eb878a7a33
|
3 |
+
size 5368
|