You can deploy this model and start earning money today!

Allen1984/
whisper-large-v2-zh-common-voice-finetuned-20240917

Name: whisper-large-v2-zh-common-voice-finetuned-20240917
Availability: InStock

peft

safetensors

generated_from_trainer

common_voice_16_1

openai/whisper-large-v2

adapter

apache-2.0

Views

No views yet

whisper-large-v2-zh-common-voice-finetuned-20240917

This model is a fine-tuned version of openai/whisper-large-v2 on the common_voice_16_1 dataset. It achieves the following results on the evaluation set:

Loss: 0.1829

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: 4
eval_batch_size: 4
seed: 42
distributed_type: multi-GPU
num_devices: 4
total_train_batch_size: 16
total_eval_batch_size: 16
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.2
num_epochs: 10
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
7.0069	0.1	25	6.8203
4.2746	0.2	50	2.4824
2.0669	0.3	75	1.4531
1.6991	0.4	100	4.1953
0.5368	0.5	125	0.1449
0.1844	0.6	150	0.1372
0.1682	0.7	175	0.1311
0.2004	0.8	200	0.1311
0.1654	0.9	225	0.1307
0.1582	1.0	250	0.1360
0.1475	1.1	275	0.1539
0.1475	1.2	300	0.1467
0.1454	1.3	325	0.1586
0.154	1.4	350	0.1545
0.1546	1.5	375	0.1754
0.1457	1.6	400	0.1825
0.2113	1.7	425	0.1768
0.2046	1.8	450	0.1901
0.2163	1.9	475	0.1906
0.2854	2.0	500	0.2280
0.1895	2.1	525	0.2496
0.2169	2.2	550	0.2576
0.1841	2.3	575	0.2651
0.2053	2.4	600	0.2410
0.2036	2.5	625	0.2295
0.209	2.6	650	0.2581
0.2163	2.7	675	0.2257
0.2219	2.8	700	0.2314
0.194	2.9	725	0.2312
0.2058	3.0	750	0.2233
0.1472	3.1	775	0.2517
0.1386	3.2	800	0.2334
0.1569	3.3	825	0.2239
0.1401	3.4	850	0.2363
0.1489	3.5	875	0.2180
0.1522	3.6	900	0.2148
0.1408	3.7	925	0.2192
0.1735	3.8	950	0.2477
0.1373	3.9	975	0.2094
0.1844	4.0	1000	0.2207
0.0817	4.1	1025	0.2272
0.0992	4.2	1050	0.2222
0.1156	4.3	1075	0.2241
0.1006	4.4	1100	0.2566
0.1107	4.5	1125	0.2178
0.1111	4.6	1150	0.2145
0.1056	4.7	1175	0.2206
0.0943	4.8	1200	0.2144
0.1069	4.9	1225	0.2042
0.1112	5.0	1250	0.2153
0.0548	5.1	1275	0.2151
0.0685	5.2	1300	0.2163
0.0702	5.3	1325	0.2155
0.0687	5.4	1350	0.2175
0.0697	5.5	1375	0.2135
0.0619	5.6	1400	0.2083
0.0678	5.7	1425	0.2052
0.0627	5.8	1450	0.2061
0.0704	5.9	1475	0.2129
0.0756	6.0	1500	0.1948
0.035	6.1	1525	0.2025
0.0299	6.2	1550	0.2063
0.0238	6.3	1575	0.2068
0.0407	6.4	1600	0.2032
0.0366	6.5	1625	0.2004
0.0273	6.6	1650	0.1947
0.0312	6.7	1675	0.2030
0.0276	6.8	1700	0.1907
0.0347	6.9	1725	0.1943
0.0265	7.0	1750	0.1940
0.0133	7.1	1775	0.1954
0.0169	7.2	1800	0.1906
0.0119	7.3	1825	0.1893
0.014	7.4	1850	0.1897
0.0117	7.5	1875	0.1912
0.0108	7.6	1900	0.1917
0.0161	7.7	1925	0.1890
0.0218	7.8	1950	0.1896
0.0168	7.9	1975	0.1925
0.0144	8.0	2000	0.1852
0.0083	8.1	2025	0.1884
0.0101	8.2	2050	0.1870
0.0074	8.3	2075	0.1816
0.0057	8.4	2100	0.1843
0.011	8.5	2125	0.1829
0.007	8.6	2150	0.1847
0.0091	8.7	2175	0.1882
0.0086	8.8	2200	0.1852
0.0057	8.9	2225	0.1825
0.0049	9.0	2250	0.1831
0.0028	9.1	2275	0.1831
0.0025	9.2	2300	0.1837
0.003	9.3	2325	0.1844
0.0036	9.4	2350	0.1843
0.0048	9.5	2375	0.1853
0.0039	9.6	2400	0.1833
0.0024	9.7	2425	0.1835
0.0042	9.8	2450	0.1835
0.0077	9.9	2475	0.1831
0.0026	10.0	2500	0.1829

Framework versions

PEFT 0.11.1
Transformers 4.42.3
Pytorch 2.3.1+cu121
Datasets 2.20.0
Tokenizers 0.19.1

Allen1984/whisper-large-v2-zh-common-voice-finetuned-20240917

whisper-large-v2-zh-common-voice-finetuned-20240917

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Allen1984/
whisper-large-v2-zh-common-voice-finetuned-20240917