README.md

metadata

language:
  - lus
license: apache-2.0
base_model: Qwen/Qwen2-Audio-7B
tags:
  - generated_from_trainer
datasets:
  - andrewbawitlung/mizonal-v3
metrics:
  - wer
  - cer
model-index:
  - name: qwen3-asr-0.6b-mizonal3-E1-lus-v2026.06
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: MiZonal v3
          type: andrewbawitlung/mizonal-v3
          config: default
          split: test
        metrics:
          - name: Wer
            type: wer
            value: 22.1262
          - name: Cer
            type: cer
            value: 5.2278
          - name: Real Time Factor
            type: rtf
            value: 0.07

qwen3-asr-0.6b-mizonal3-E1-lus-v2026.06

This model is a fine-tuned version of Qwen/Qwen2-Audio-7B on the MiZonal v3 dataset. Note: ~1 hour of conversational speech was added to this dataset version.

It achieves the following results on the evaluation set:

• Wer: 22.1262
• Cer: 5.2278
• Real Time Factor: 0.0700

Model description

Experiment Configurations

This repository is part of a series of experiments. The different configurations are:

• E1 (Baseline): Standard training configuration.
• E2 (Noise): Training with background noise augmentation.
• E3 (Speed): Training with speed perturbation augmentation.
• E4 (SpecAug): Training with SpecAugment (time and frequency masking).
• E5 (Combined): Training with a combination of all augmentations.

All Models in this Family

Experiment	Hugging Face Repository
E1 (Baseline)	andrewbawitlung/qwen3-asr-0.6b-mizonal3-E1-lus-v2026.06
E2 (Noise)	andrewbawitlung/qwen3-asr-0.6b-mizonal3-E2-lus-v2026.06
E3 (Speed)	andrewbawitlung/qwen3-asr-0.6b-mizonal3-E3-lus-v2026.06
E4 (SpecAug)	andrewbawitlung/qwen3-asr-0.6b-mizonal3-E4-lus-v2026.06
E5 (Combined)	andrewbawitlung/qwen3-asr-0.6b-mizonal3-E5-lus-v2026.06

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• optimizer: OptimizerNames.ADAMW_TORCH_FUSED
• lr_scheduler_type: SchedulerType.LINEAR
• num_epochs: 8

Training results

step	epoch	train_loss	eval_loss	eval_wer	eval_cer	learning_rate	grad_norm
200	0.36429872495446264	0.3531	0.5678281188011169	0.42064804904154907	0.10793044953350296	1.9484200743494427e-05	11.4375
400	0.7285974499089253	0.2215	0.38337236642837524	0.3034932373260679	0.07453350296861748	1.855483271375465e-05	9.125
600	1.092896174863388	0.1323	0.33564406633377075	0.2659336382212708	0.06248232965790218	1.762546468401487e-05	8.75
800	1.4571948998178508	0.1034	0.3069060146808624	0.24978106451299018	0.06009683347469607	1.6696096654275093e-05	19.875
1000	1.8214936247723132	0.1023	0.2834382951259613	0.22983360902987254	0.053894543398360195	1.5766728624535318e-05	4.03125
1200	2.185792349726776	0.0753	0.2766484320163727	0.22370341539359737	0.051509047215154084	1.483736059479554e-05	5.625
1400	2.5500910746812386	0.0566	0.27566370368003845	0.21134572346015373	0.04871713316369805	1.3907992565055763e-05	4.71875
1600	2.9143897996357016	0.0573	0.2684209942817688	0.2139729493042717	0.0496006502685892	1.2978624535315987e-05	4.875
1800	3.278688524590164	0.037	0.28212207555770874	0.2136810353215919	0.050378145320893415	1.204925650557621e-05	3.75
2000	3.6429872495446265	0.0337	0.2878636121749878	0.20842658363335603	0.048699462821600224	1.111988847583643e-05	4.03125
2200	4.007285974499089	0.039	0.2856166362762451	0.23752067724043982	0.08289157478088775	1.0190520446096654e-05	5.4375
2400	4.371584699453552	0.0189	0.30527880787849426	0.21115111413836724	0.05044882668928471	9.261152416356878e-06	2.84375
2600	4.735883424408015	0.0178	0.3008843958377838	0.21280529337355258	0.05085524455753463	8.331784386617101e-06	2.6875
2800	5.100182149362477	0.0101	0.3245539367198944	0.21280529337355258	0.05157972858354538	7.402416356877325e-06	2.140625
3000	5.46448087431694	0.0122	0.3235035240650177	0.21407025396516494	0.05188012439920837	6.4730483271375465e-06	2.40625
3200	5.828779599271402	0.0102	0.32762861251831055	0.21173494210372676	0.05073155216284987	5.54368029739777e-06	2.1875
3400	6.193078324225866	0.0059	0.33911389112472534	0.21338912133891214	0.05253392705682782	4.614312267657993e-06	0.96875
3600	6.557377049180328	0.0074	0.341426283121109	0.2139729493042717	0.052198190556969185	3.684944237918216e-06	1.1484375
3800	6.9216757741347905	0.0084	0.3391820788383484	0.21037267685122119	0.05156205824144756	2.755576208178439e-06	2.125
4000	7.285974499089253	0.0079	0.34474754333496094	0.21309720735623236	0.051897794741306194	1.8262081784386618e-06	1.7421875
4200	7.6502732240437155	0.0082	0.34566405415534973	0.2145567772696312	0.052109838846480065	8.968401486988849e-07	3.4375