metadata

license: apache-2.0
base_model: openai/whisper-medium
tags:
  - generated_from_trainer
model-index:
  - name: MAScIR_elderly_whisper-medium-LoRA-data-augmented
    results: []

MAScIR_elderly_whisper-medium-LoRA-data-augmented

This model is a fine-tuned version of openai/whisper-medium on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.0358

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: 0.001
train_batch_size: 8
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 200
num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss
0.8119	0.09	100	0.2422
0.7907	0.19	200	0.2357
0.6762	0.28	300	0.2311
0.7081	0.38	400	0.2256
0.5623	0.47	500	0.1946
0.569	0.57	600	0.1697
9.0833	0.66	700	8.1242
6.1681	0.76	800	5.9288
5.5565	0.85	900	4.9360
2.0714	0.95	1000	0.2584
0.6051	1.04	1100	0.2062
0.485	1.14	1200	0.1824
0.637	1.23	1300	0.1522
0.5521	1.33	1400	0.1371
0.3999	1.42	1500	0.1331
0.4788	1.52	1600	0.1344
0.3738	1.61	1700	0.0952
0.3046	1.71	1800	0.0871
0.4335	1.8	1900	0.0770
0.3876	1.9	2000	0.0654
0.4226	1.99	2100	0.0638
0.2651	2.09	2200	0.0612
0.2075	2.18	2300	0.0541
0.2464	2.28	2400	0.0473
0.1797	2.37	2500	0.0482
0.2393	2.47	2600	0.0428
0.1764	2.56	2700	0.0396
0.1398	2.66	2800	0.0390
0.1855	2.75	2900	0.0382
0.232	2.85	3000	0.0369
0.2	2.94	3100	0.0358

Framework versions

Transformers 4.34.0.dev0
Pytorch 2.0.1+cu118
Datasets 2.14.5
Tokenizers 0.13.3