Teamim-AllNusah-whisper-tiny_Random-True_Mid

cantillation/Teamim-AllNusah-whisper-tiny_Random-True_Mid

Teamim-AllNusah-whisper-tiny_Random-True_Mid is a automatic speech recognition model from cantillation. Use it when you need speech turned into text. It is set up for transformers. The card lists the license as apache-2.0.

should probably proofread and complete it, then remove this comment. --

Downloads · 30 days

25% of all-time downloads

All-time downloads

Public

Parameters

37.8M

1.5 GB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors151 MB · 98%

At a glance

Task: Automatic Speech Recognition
Library: transformers
License: apache-2.0
Model type: whisper
Access: Public
Created: Mar 21, 2024
Updated: Mar 21, 2024
SHA: 275b13c1

Base models

openai/whisper-tiny

Model card

From the Hugging Face model README

he-cantillation

This model is a fine-tuned version of openai/whisper-tiny on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.2139
Wer: 17.8049
Avg Precision Exact: 0.8322
Avg Recall Exact: 0.8338
Avg F1 Exact: 0.8325
Avg Precision Letter Shift: 0.8586
Avg Recall Letter Shift: 0.8603
Avg F1 Letter Shift: 0.8589
Avg Precision Word Level: 0.8646
Avg Recall Word Level: 0.8661
Avg F1 Word Level: 0.8648
Avg Precision Word Shift: 0.9560
Avg Recall Word Shift: 0.9594
Avg F1 Word Shift: 0.9570
Precision Median Exact: 0.9231
Recall Median Exact: 0.9231
F1 Median Exact: 0.9231
Precision Max Exact: 1.0
Recall Max Exact: 1.0
F1 Max Exact: 1.0
Precision Min Exact: 0.0
Recall Min Exact: 0.0
F1 Min Exact: 0.0
Precision Min Letter Shift: 0.0
Recall Min Letter Shift: 0.0
F1 Min Letter Shift: 0.0
Precision Min Word Level: 0.0
Recall Min Word Level: 0.0
F1 Min Word Level: 0.0
Precision Min Word Shift: 0.1111
Recall Min Word Shift: 0.1111
F1 Min Word Shift: 0.125

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 1e-05
train_batch_size: 8
eval_batch_size: 32
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 1000
training_steps: 100000
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Avg Precision Exact	Avg Recall Exact	Avg F1 Exact	Avg Precision Letter Shift	Avg Recall Letter Shift	Avg F1 Letter Shift	Avg Precision Word Level	Avg Recall Word Level	Avg F1 Word Level	Avg Precision Word Shift	Avg Recall Word Shift	Avg F1 Word Shift	Precision Median Exact	Recall Median Exact	F1 Median Exact	Precision Max Exact	Recall Max Exact	F1 Max Exact	Precision Min Word Shift	Recall Min Word Shift	F1 Min Word Shift
No log	0.0	1	9.8845	100.0	0.0005	0.0042	0.0008	0.0151	0.0160	0.0155	0.0037	0.0339	0.0066	0.0689	0.0732	0.0706	0.0	0.0	0.0	0.1429	1.0	0.25	0.0	0.0	0.0
0.0658	0.8	10000	0.1829	26.4745	0.7434	0.7474	0.7447	0.7770	0.7813	0.7784	0.7850	0.7890	0.7862	0.9183	0.9221	0.9192	0.8516	0.8571	0.8571	1.0	1.0	1.0	0.0833	0.0909	0.0870
0.03	1.6	20000	0.1693	21.5854	0.7959	0.7942	0.7945	0.8256	0.8239	0.8241	0.8319	0.8300	0.8302	0.9415	0.9414	0.9406	0.9091	0.9	0.9	1.0	1.0	1.0	0.1111	0.1111	0.125
0.0279	2.4	30000	0.1784	20.1072	0.8107	0.8146	0.8121	0.8398	0.8438	0.8412	0.8459	0.8503	0.8474	0.9442	0.9508	0.9467	0.9091	0.9091	0.9091	1.0	1.0	1.0	0.0	0.0	0.0
0.0175	3.2	40000	0.1883	19.3459	0.8145	0.8149	0.8141	0.8411	0.8415	0.8407	0.8465	0.8472	0.8463	0.9480	0.9503	0.9485	0.9167	0.9167	0.9167	1.0	1.0	1.0	0.0909	0.1111	0.1000
0.0046	4.0	50000	0.1997	19.2535	0.8253	0.8230	0.8236	0.8527	0.8504	0.8510	0.8592	0.8567	0.8574	0.9523	0.9517	0.9513	0.9167	0.9167	0.9167	1.0	1.0	1.0	0.1429	0.125	0.1333
0.0072	4.8	60000	0.2034	18.3407	0.8247	0.8272	0.8254	0.8527	0.8553	0.8535	0.8595	0.8619	0.8602	0.9551	0.9586	0.9562	0.9167	0.9167	0.9167	1.0	1.0	1.0	0.0	0.0	0.0
0.0041	5.6	70000	0.2095	18.2262	0.8263	0.8263	0.8258	0.8534	0.8535	0.8529	0.8589	0.8591	0.8584	0.9554	0.9578	0.9560	0.9231	0.9167	0.9231	1.0	1.0	1.0	0.0	0.0	0.0
0.0011	6.4	80000	0.2145	18.1744	0.8287	0.8292	0.8284	0.8550	0.8555	0.8547	0.8602	0.8606	0.8598	0.9545	0.9570	0.9551	0.9167	0.9231	0.9231	1.0	1.0	1.0	0.0909	0.1111	0.1000
0.002	7.2	90000	0.2144	17.8640	0.8323	0.8332	0.8323	0.8587	0.8598	0.8587	0.8646	0.8655	0.8646	0.9559	0.9587	0.9566	0.9231	0.9231	0.9231	1.0	1.0	1.0	0.0909	0.1111	0.1000
0.0018	8.0	100000	0.2139	17.8049	0.8322	0.8338	0.8325	0.8586	0.8603	0.8589	0.8646	0.8661	0.8648	0.9560	0.9594	0.9570	0.9231	0.9231	0.9231	1.0	1.0	1.0	0.1111	0.1111	0.125

Framework versions

Transformers 4.39.0.dev0
Pytorch 2.2.1+cu121
Datasets 2.16.1
Tokenizers 0.15.0

Task: Automatic Speech Recognition
Library: transformers
Type: whisper
License: apache-2.0
Languages: he
Created: Mar 21, 2024
Updated: Mar 21, 2024