# Google Colab Ready Quick Inference
# 1. Install dependencies:
!pip install huggingface_hub omnilingual-asr torchaudio

# 2. Download the model checkpoint directly from Hugging Face:
import torch
from huggingface_hub import hf_hub_download

model_path = hf_hub_download(
    repo_id="andrewbawitlung/omni-asr-ctc-300m-v2-mizonal3-E3-lus-v2026.06",
    filename="sdp_00.pt"
)

# 3. Load the base model pipeline and inject fine-tuned weights:
from omnilingual_asr.models.inference.pipeline import ASRInferencePipeline

print("Loading base model pipeline...")
pipeline = ASRInferencePipeline(model_card="omniASR_CTC_300M_v2")

print("Loading fine-tuned checkpoint weights...")
checkpoint = torch.load(model_path, map_location="cpu")
state_dict = checkpoint['model'] if 'model' in checkpoint else checkpoint
pipeline.model.load_state_dict(state_dict, strict=False)

# 4. Run transcription (make sure audio files are present in the specified paths):
audio_files = ["/content/segment_003.wav", "/content/segment_108.wav"]
lang = ["lus_Latn", "lus_Latn"]
transcriptions = pipeline.transcribe(audio_files, lang=lang, batch_size=2)

print("Transcription 0:", transcriptions[0])
print("Transcription 1:", transcriptions[1])

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/omni-asr-ctc-300m-v2-mizonal3-E1-lus-v2026.06
E2 (Noise)	andrewbawitlung/omni-asr-ctc-300m-v2-mizonal3-E2-lus-v2026.06
E3 (Speed)	andrewbawitlung/omni-asr-ctc-300m-v2-mizonal3-E3-lus-v2026.06
E4 (SpecAug)	andrewbawitlung/omni-asr-ctc-300m-v2-mizonal3-E4-lus-v2026.06
E5 (Combined)	andrewbawitlung/omni-asr-ctc-300m-v2-mizonal3-E5-lus-v2026.06

Training hyperparameters

Training hyperparameters not available in bin format.

Training results

step	epoch	train_loss	eval_loss	eval_wer	eval_cer	grad_norm
100	3.0000	65.5965				281.3520
200	5.0000	39.0984				88.8331
300	8.0000	32.0597				98.2777
400	10.0000	23.7244				59.9031
500	13.0000	20.8239				79.3340
600	15.0000	16.3057				47.8548
700	18.0000	14.4927				53.2894
800	20.0000	11.2795	12.5893	20.0156	4.0160	38.7695
900	23.0000	9.8886				35.1124
1000	25.0000	8.1665				28.8110
1100	28.0000	7.1241				27.0514
1200	30.0000	5.5548				25.8380
1300	33.0000	5.0678				26.2448
1400	35.0000	4.3349				24.4729
1500	38.0000	3.7787				24.0198
1600	40.0000	3.4933	15.9423	16.8980	3.2639	18.6042
1700	43.0000	2.7611				18.6376
1800	45.0000	2.3575				17.8721
1900	48.0000	2.4540				18.4342
2000	50.0000	2.2154				15.9295
2100	53.0000	1.8771				18.9938
2200	55.0000	1.5710				12.7145
2300	58.0000	1.5092				15.3270
2400	60.0000	1.4374	22.2271	15.9402	3.1113	13.3519
2500	63.0000	1.2930				13.1228
2600	65.0000	1.2291				12.2031
2700	68.0000	1.2222				12.3354
2800	70.0000	1.2905				11.6585
2900	73.0000	0.9952				10.9727
3000	75.0000	1.0097				10.5416
3100	78.0000	0.9109				17.1557
3200	80.0000	0.7360	24.9567	15.5493	2.9907	9.6445
3300	83.0000	0.6933				10.1479
3400	85.0000	0.6904				9.1531
3500	88.0000	0.6576				9.4965
3600	90.0000	0.5875				7.7222
3700	93.0000	0.7039				9.1735
3800	95.0000	0.5678				13.2920
3900	98.0000	0.5372				8.1872
4000	100.0000	0.4710	26.3384	14.5426	2.8151	6.2237
4100	103.0000	0.5047				8.6809
4200	105.0000	0.3869				5.7722
4300	108.0000	0.3877				6.3638
4400	110.0000	0.4078				6.3570
4500	113.0000	0.6695				10.3843
4600	115.0000	0.3720				6.5772
4700	118.0000	0.4218				7.0805
4800	120.0000	0.3339	27.5587	14.5035	2.8044	5.6652
4900	123.0000	0.3387				6.7879
5000	125.0000	0.3084				6.1141
5100	128.0000	0.3111				7.2847
5200	130.0000	0.3574				5.8786
5300	133.0000	0.3028				14.1547
5400	135.0000	0.2235				4.3351
5500	138.0000	0.3968				5.9151
5600	140.0000	0.2775	27.9162	14.2494	2.7743	4.8575
5700	143.0000	0.2755				4.6821
5800	145.0000	0.3046				5.0620
5900	148.0000	0.2812				4.9860
6000	150.0000	0.2982				5.3985
6100	153.0000	0.2727				5.1145
6200	155.0000	0.2642				4.7418
6300	158.0000	0.3387				5.6535
6400	160.0000	0.2759	28.4943	14.1028	2.7282	5.1401