This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the audiofolder dataset. It achieves the following results on the evaluation set:

Loss: nan
Wer: 1.0
Cer: 1.0

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
gradient_accumulation_steps: 2
total_train_batch_size: 16
optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 2000
num_epochs: 100
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
7.7423	0.3802	100	3.4997	1.0	1.0
3.5297	0.7605	200	3.6791	1.0	1.0
3.2609	1.1407	300	3.0158	1.0	1.0
3.3756	1.5209	400	2.9461	1.0	1.0
3.7095	1.9011	500	2.9905	1.0	1.0
3.6751	2.2814	600	2.9708	1.0	1.0
4.8025	2.6616	700	2.9062	1.0	1.0
3.5455	3.0418	800	2.9023	1.0	1.0
3.4116	3.4221	900	3.0732	1.0	1.0
3.4601	3.8023	1000	3.2853	1.0	1.0
3.1515	4.1825	1100	3.3554	1.0	1.0
3.2965	4.5627	1200	3.3260	1.0	1.0
3.4571	4.9430	1300	3.3794	1.0	1.0
3.5056	5.3232	1400	2.9404	1.0	1.0
3.5623	5.7034	1500	3.2478	1.0	1.0
4.1789	6.0837	1600	3.1654	1.0	1.0
4.2509	6.4639	1700	3.7535	1.0	1.0
5.1478	6.8441	1800	5.2032	1.0	1.0
5.2211	7.2243	1900	5.9605	1.0	1.0
6.0949	7.6046	2000	6.5013	1.0	1.0
5.7856	7.9848	2100	6.5014	1.0	1.0
6.0892	8.3650	2200	6.5020	1.0	1.0
6.0055	8.7452	2300	6.5024	1.0	1.0
6.3741	9.1255	2400	6.5014	1.0	1.0
6.1777	9.5057	2500	6.5013	1.0	1.0
6.0985	9.8859	2600	6.5011	1.0	1.0
6.039	10.2662	2700	6.5024	1.0	1.0
6.2022	10.6464	2800	6.5011	1.0	1.0
6.0846	11.0266	2900	6.5012	1.0	1.0
6.0821	11.4068	3000	6.5020	1.0	1.0
6.0395	11.7871	3100	6.5025	1.0	1.0
6.2914	12.1673	3200	6.5012	1.0	1.0
6.2113	12.5475	3300	6.5015	1.0	1.0
5.9776	12.9278	3400	6.5013	1.0	1.0
6.0498	13.3080	3500	6.5017	1.0	1.0
6.1298	13.6882	3600	6.5025	1.0	1.0
6.2922	14.0684	3700	6.5017	1.0	1.0
6.0821	14.4487	3800	6.5013	1.0	1.0
6.1075	14.8289	3900	6.5020	1.0	1.0
6.1709	15.2091	4000	6.5021	1.0	1.0
6.0219	15.5894	4100	6.5023	1.0	1.0
5.8627	15.9696	4200	6.5018	1.0	1.0
6.1203	16.3498	4300	6.5013	1.0	1.0
6.1534	16.7300	4400	6.5015	1.0	1.0
6.4003	17.1103	4500	6.5016	1.0	1.0
6.1297	17.4905	4600	6.5014	1.0	1.0
6.1184	17.8707	4700	6.5023	1.0	1.0
6.0136	18.2510	4800	6.5018	1.0	1.0
6.2258	18.6312	4900	6.5010	1.0	1.0
6.0193	19.0114	5000	6.5014	1.0	1.0
6.238	19.3916	5100	6.5017	1.0	1.0
6.0926	19.7719	5200	6.5019	1.0	1.0
6.1041	20.1521	5300	6.5018	1.0	1.0
6.2209	20.5323	5400	6.5020	1.0	1.0
6.1275	20.9125	5500	6.5020	1.0	1.0
6.0931	21.2928	5600	6.5024	1.0	1.0
6.1613	21.6730	5700	6.5021	1.0	1.0
6.1273	22.0532	5800	6.5016	1.0	1.0
6.0681	22.4335	5900	6.5019	1.0	1.0
6.3138	22.8137	6000	6.5018	1.0	1.0
6.2613	23.1939	6100	6.5017	1.0	1.0
6.0473	23.5741	6200	6.5013	1.0	1.0
5.9002	23.9544	6300	6.5022	1.0	1.0
6.107	24.3346	6400	6.5011	1.0	1.0
6.0794	24.7148	6500	6.5009	1.0	1.0
6.2978	25.0951	6600	6.5019	1.0	1.0
5.999	25.4753	6700	6.5014	1.0	1.0
6.0059	25.8555	6800	6.5017	1.0	1.0
6.0511	26.2357	6900	6.5011	1.0	1.0
6.2068	26.6160	7000	6.5020	1.0	1.0
5.8239	26.9962	7100	6.5021	1.0	1.0
6.0688	27.3764	7200	6.5013	1.0	1.0
6.0839	27.7567	7300	6.5020	1.0	1.0
6.2684	28.1369	7400	6.5019	1.0	1.0
6.0049	28.5171	7500	6.5021	1.0	1.0
6.092	28.8973	7600	6.5015	1.0	1.0
6.2543	29.2776	7700	6.5018	1.0	1.0
6.1423	29.6578	7800	6.5017	1.0	1.0
6.1546	30.0380	7900	6.5019	1.0	1.0
6.1332	30.4183	8000	6.5022	1.0	1.0
6.0393	30.7985	8100	6.5019	1.0	1.0
6.1275	31.1787	8200	6.5015	1.0	1.0
6.1026	31.5589	8300	6.5017	1.0	1.0
6.2027	31.9392	8400	6.5017	1.0	1.0
6.182	32.3194	8500	6.5020	1.0	1.0
6.0619	32.6996	8600	6.5016	1.0	1.0
6.4214	33.0798	8700	6.5021	1.0	1.0
6.0854	33.4601	8800	6.5023	1.0	1.0
6.1748	33.8403	8900	6.5017	1.0	1.0
6.2243	34.2205	9000	6.5024	1.0	1.0
6.7435	34.6008	9100	nan	1.0	1.0
0.0	34.9810	9200	nan	1.0	1.0
0.0	35.3612	9300	nan	1.0	1.0
0.0	35.7414	9400	nan	1.0	1.0
0.0	36.1217	9500	nan	1.0	1.0
0.0	36.5019	9600	nan	1.0	1.0
0.0	36.8821	9700	nan	1.0	1.0
0.0	37.2624	9800	nan	1.0	1.0
0.0	37.6426	9900	nan	1.0	1.0
0.0	38.0228	10000	nan	1.0	1.0
0.0	38.4030	10100	nan	1.0	1.0
0.0	38.7833	10200	nan	1.0	1.0
0.0	39.1635	10300	nan	1.0	1.0
0.0	39.5437	10400	nan	1.0	1.0
0.0	39.9240	10500	nan	1.0	1.0
0.0	40.3042	10600	nan	1.0	1.0
0.0	40.6844	10700	nan	1.0	1.0
0.0	41.0646	10800	nan	1.0	1.0
0.0	41.4449	10900	nan	1.0	1.0
0.0	41.8251	11000	nan	1.0	1.0
0.0	42.2053	11100	nan	1.0	1.0
0.0	42.5856	11200	nan	1.0	1.0
0.0	42.9658	11300	nan	1.0	1.0
0.0	43.3460	11400	nan	1.0	1.0
0.0	43.7262	11500	nan	1.0	1.0
0.0	44.1065	11600	nan	1.0	1.0
0.0	44.4867	11700	nan	1.0	1.0
0.0	44.8669	11800	nan	1.0	1.0
0.0	45.2471	11900	nan	1.0	1.0
0.0	45.6274	12000	nan	1.0	1.0
0.0	46.0076	12100	nan	1.0	1.0
0.0	46.3878	12200	nan	1.0	1.0
0.0	46.7681	12300	nan	1.0	1.0
0.0	47.1483	12400	nan	1.0	1.0
0.0	47.5285	12500	nan	1.0	1.0
0.0	47.9087	12600	nan	1.0	1.0
0.0	48.2890	12700	nan	1.0	1.0
0.0	48.6692	12800	nan	1.0	1.0
0.0	49.0494	12900	nan	1.0	1.0
0.0	49.4297	13000	nan	1.0	1.0
0.0	49.8099	13100	nan	1.0	1.0
0.0	50.1901	13200	nan	1.0	1.0
0.0	50.5703	13300	nan	1.0	1.0
0.0	50.9506	13400	nan	1.0	1.0
0.0	51.3308	13500	nan	1.0	1.0
0.0	51.7110	13600	nan	1.0	1.0
0.0	52.0913	13700	nan	1.0	1.0
0.0	52.4715	13800	nan	1.0	1.0
0.0	52.8517	13900	nan	1.0	1.0
0.0	53.2319	14000	nan	1.0	1.0
0.0	53.6122	14100	nan	1.0	1.0
0.0	53.9924	14200	nan	1.0	1.0
0.0	54.3726	14300	nan	1.0	1.0
0.0	54.7529	14400	nan	1.0	1.0
0.0	55.1331	14500	nan	1.0	1.0
0.0	55.5133	14600	nan	1.0	1.0
0.0	55.8935	14700	nan	1.0	1.0
0.0	56.2738	14800	nan	1.0	1.0
0.0	56.6540	14900	nan	1.0	1.0
0.0	57.0342	15000	nan	1.0	1.0
0.0	57.4144	15100	nan	1.0	1.0
0.0	57.7947	15200	nan	1.0	1.0
0.0	58.1749	15300	nan	1.0	1.0
0.0	58.5551	15400	nan	1.0	1.0
0.0	58.9354	15500	nan	1.0	1.0
0.0	59.3156	15600	nan	1.0	1.0
0.0	59.6958	15700	nan	1.0	1.0
0.0	60.0760	15800	nan	1.0	1.0
0.0	60.4563	15900	nan	1.0	1.0
0.0	60.8365	16000	nan	1.0	1.0
0.0	61.2167	16100	nan	1.0	1.0
0.0	61.5970	16200	nan	1.0	1.0
0.0	61.9772	16300	nan	1.0	1.0
0.0	62.3574	16400	nan	1.0	1.0
0.0	62.7376	16500	nan	1.0	1.0
0.0	63.1179	16600	nan	1.0	1.0
0.0	63.4981	16700	nan	1.0	1.0
0.0	63.8783	16800	nan	1.0	1.0
0.0	64.2586	16900	nan	1.0	1.0
0.0	64.6388	17000	nan	1.0	1.0
0.0	65.0190	17100	nan	1.0	1.0
0.0	65.3992	17200	nan	1.0	1.0
0.0	65.7795	17300	nan	1.0	1.0
0.0	66.1597	17400	nan	1.0	1.0
0.0	66.5399	17500	nan	1.0	1.0
0.0	66.9202	17600	nan	1.0	1.0
0.0	67.3004	17700	nan	1.0	1.0
0.0	67.6806	17800	nan	1.0	1.0
0.0	68.0608	17900	nan	1.0	1.0
0.0	68.4411	18000	nan	1.0	1.0
0.0	68.8213	18100	nan	1.0	1.0
0.0	69.2015	18200	nan	1.0	1.0
0.0	69.5817	18300	nan	1.0	1.0
0.0	69.9620	18400	nan	1.0	1.0
0.0	70.3422	18500	nan	1.0	1.0
0.0	70.7224	18600	nan	1.0	1.0
0.0	71.1027	18700	nan	1.0	1.0
0.0	71.4829	18800	nan	1.0	1.0
0.0	71.8631	18900	nan	1.0	1.0
0.0	72.2433	19000	nan	1.0	1.0
0.0	72.6236	19100	nan	1.0	1.0
0.0	73.0038	19200	nan	1.0	1.0
0.0	73.3840	19300	nan	1.0	1.0
0.0	73.7643	19400	nan	1.0	1.0
0.0	74.1445	19500	nan	1.0	1.0
0.0	74.5247	19600	nan	1.0	1.0
0.0	74.9049	19700	nan	1.0	1.0
0.0	75.2852	19800	nan	1.0	1.0
0.0	75.6654	19900	nan	1.0	1.0
0.0	76.0456	20000	nan	1.0	1.0
0.0	76.4259	20100	nan	1.0	1.0
0.0	76.8061	20200	nan	1.0	1.0
0.0	77.1863	20300	nan	1.0	1.0
0.0	77.5665	20400	nan	1.0	1.0
0.0	77.9468	20500	nan	1.0	1.0
0.0	78.3270	20600	nan	1.0	1.0
0.0	78.7072	20700	nan	1.0	1.0
0.0	79.0875	20800	nan	1.0	1.0
0.0	79.4677	20900	nan	1.0	1.0
0.0	79.8479	21000	nan	1.0	1.0
0.0	80.2281	21100	nan	1.0	1.0
0.0	80.6084	21200	nan	1.0	1.0
0.0	80.9886	21300	nan	1.0	1.0
0.0	81.3688	21400	nan	1.0	1.0
0.0	81.7490	21500	nan	1.0	1.0
0.0	82.1293	21600	nan	1.0	1.0
0.0	82.5095	21700	nan	1.0	1.0
0.0	82.8897	21800	nan	1.0	1.0
0.0	83.2700	21900	nan	1.0	1.0
0.0	83.6502	22000	nan	1.0	1.0
0.0	84.0304	22100	nan	1.0	1.0
0.0	84.4106	22200	nan	1.0	1.0
0.0	84.7909	22300	nan	1.0	1.0
0.0	85.1711	22400	nan	1.0	1.0
0.0	85.5513	22500	nan	1.0	1.0
0.0	85.9316	22600	nan	1.0	1.0
0.0	86.3118	22700	nan	1.0	1.0
0.0	86.6920	22800	nan	1.0	1.0
0.0	87.0722	22900	nan	1.0	1.0
0.0	87.4525	23000	nan	1.0	1.0
0.0	87.8327	23100	nan	1.0	1.0
0.0	88.2129	23200	nan	1.0	1.0
0.0	88.5932	23300	nan	1.0	1.0
0.0	88.9734	23400	nan	1.0	1.0
0.0	89.3536	23500	nan	1.0	1.0
0.0	89.7338	23600	nan	1.0	1.0
0.0	90.1141	23700	nan	1.0	1.0
0.0	90.4943	23800	nan	1.0	1.0
0.0	90.8745	23900	nan	1.0	1.0
0.0	91.2548	24000	nan	1.0	1.0
0.0	91.6350	24100	nan	1.0	1.0
0.0	92.0152	24200	nan	1.0	1.0
0.0	92.3954	24300	nan	1.0	1.0
0.0	92.7757	24400	nan	1.0	1.0
0.0	93.1559	24500	nan	1.0	1.0
0.0	93.5361	24600	nan	1.0	1.0
0.0	93.9163	24700	nan	1.0	1.0
0.0	94.2966	24800	nan	1.0	1.0
0.0	94.6768	24900	nan	1.0	1.0
0.0	95.0570	25000	nan	1.0	1.0
0.0	95.4373	25100	nan	1.0	1.0
0.0	95.8175	25200	nan	1.0	1.0
0.0	96.1977	25300	nan	1.0	1.0
0.0	96.5779	25400	nan	1.0	1.0
0.0	96.9582	25500	nan	1.0	1.0
0.0	97.3384	25600	nan	1.0	1.0
0.0	97.7186	25700	nan	1.0	1.0
0.0	98.0989	25800	nan	1.0	1.0
0.0	98.4791	25900	nan	1.0	1.0
0.0	98.8593	26000	nan	1.0	1.0
0.0	99.2395	26100	nan	1.0	1.0
0.0	99.6198	26200	nan	1.0	1.0
0.0	100.0	26300	nan	1.0	1.0

Framework versions

Transformers 4.57.3
Pytorch 2.6.0+cu124
Datasets 3.2.0
Tokenizers 0.22.1

aconeil/w2v2-lmk_full_auto

w2v2-lmk_full_auto

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

aconeil/
w2v2-lmk_full_auto