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 OptimizerNames.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: 100
num_epochs: 30
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Cer	Wer
0.7733	0.2717	200	0.5681	0.1786	0.8107
0.5877	0.5435	400	0.4554	0.1515	0.7426
0.5291	0.8152	600	0.4313	0.1429	0.7208
0.4806	1.0870	800	0.4011	0.1351	0.6954
0.468	1.3587	1000	0.3855	0.1341	0.6937
0.4632	1.6304	1200	0.3713	0.1274	0.6770
0.4515	1.9022	1400	0.4208	0.1348	0.6897
0.4281	2.1739	1600	0.3634	0.1266	0.6734
0.4253	2.4457	1800	0.3712	0.1299	0.6681
0.4167	2.7174	2000	0.3503	0.1229	0.6525
0.4188	2.9891	2200	0.3513	0.1238	0.6540
0.4092	3.2609	2400	0.3591	0.1237	0.6494
0.4151	3.5326	2600	0.3543	0.1224	0.6520
0.3935	3.8043	2800	0.3500	0.1230	0.6535
0.4074	4.0761	3000	0.3424	0.1215	0.6494
0.3947	4.3478	3200	0.3565	0.1211	0.6499
0.406	4.6196	3400	0.3700	0.1199	0.6532
0.413	4.8913	3600	0.3691	0.1261	0.6676
0.4426	5.1630	3800	0.3985	0.1275	0.6628
0.4658	5.4348	4000	0.4065	0.1278	0.6669
0.4736	5.7065	4200	0.4678	0.1423	0.7026
0.5586	5.9783	4400	0.4864	0.1528	0.7342
0.5504	6.25	4600	0.5632	0.1393	0.7371
0.5641	6.5217	4800	0.4616	0.1373	0.6985
0.699	6.7935	5000	0.6971	0.1770	0.7843
1.6761	7.0652	5200	1.8134	0.5793	0.9952
3.0527	7.3370	5400	2.9738	0.8306	1.0
3.091	7.6087	5600	2.9383	0.8252	0.9993
3.1507	7.8804	5800	3.0451	0.7277	0.9995
3.3528	8.1522	6000	3.0693	0.7685	0.9990
2.8426	8.4239	6200	2.4560	0.5896	0.9940
2.4325	8.6957	6400	1.9646	0.4776	0.9871
2.3467	8.9674	6600	2.0953	0.3903	0.9727
4.1072	9.2391	6800	4.2700	0.4325	0.9830
6.871	9.5109	7000	6.4158	0.4708	0.9827
9.2051	9.7826	7200	8.2100	0.5134	0.9844
10.8104	10.0543	7400	9.7932	0.5130	0.9883
13.342	10.3261	7600	11.2517	0.6396	1.0
13.5902	10.5978	7800	12.8965	0.7722	1.0
13.0844	10.8696	8000	11.8398	0.7039	1.0
12.9694	11.1413	8200	11.7697	0.7624	1.0
13.4419	11.4130	8400	11.8663	0.8062	1.0
13.0815	11.6848	8600	11.8608	0.8061	1.0
13.2691	11.9565	8800	11.8538	0.8061	1.0
13.2696	12.2283	9000	11.8520	0.8061	1.0
13.4105	12.5	9200	11.8532	0.8062	1.0
13.2602	12.7717	9400	11.8634	0.8063	1.0
13.6417	13.0435	9600	11.8639	0.8061	1.0
13.5022	13.3152	9800	11.8516	0.8062	1.0
13.432	13.5870	10000	11.8559	0.8061	1.0
13.1676	13.8587	10200	11.8649	0.8061	1.0
13.3595	14.1304	10400	11.8509	0.8061	1.0
13.3553	14.4022	10600	11.8606	0.8059	1.0
13.2227	14.6739	10800	11.8568	0.8061	1.0
13.2959	14.9457	11000	11.8527	0.8061	1.0
13.2174	15.2174	11200	11.8612	0.8061	1.0
13.5278	15.4891	11400	11.8623	0.8061	1.0
13.225	15.7609	11600	11.8627	0.8061	1.0
13.549	16.0326	11800	11.8634	0.8061	1.0
13.2783	16.3043	12000	11.8658	0.8063	1.0
13.5079	16.5761	12200	11.8639	0.8062	1.0
13.9686	16.8478	12400	11.8615	0.8060	1.0
13.9209	17.1196	12600	11.8619	0.8061	1.0
13.6849	17.3913	12800	11.8633	0.8063	1.0
13.0396	17.6630	13000	11.8606	0.8061	1.0
13.4195	17.9348	13200	11.8664	0.8060	1.0
13.1744	18.2065	13400	11.8620	0.8062	1.0
13.4119	18.4783	13600	11.8654	0.8063	1.0
13.5461	18.75	13800	11.8569	0.8061	1.0
13.1495	19.0217	14000	11.8633	0.8063	1.0
13.4842	19.2935	14200	11.8628	0.8063	1.0
13.2245	19.5652	14400	11.8636	0.8059	1.0
13.9797	19.8370	14600	11.8648	0.8061	1.0
13.5495	20.1087	14800	11.8623	0.8060	1.0
13.6881	20.3804	15000	11.8491	0.8062	1.0
13.4038	20.6522	15200	11.8569	0.8063	1.0
13.4781	20.9239	15400	11.8518	0.8062	1.0
13.5601	21.1957	15600	11.8497	0.8062	1.0
13.1207	21.4674	15800	11.8517	0.8061	1.0
13.3136	21.7391	16000	11.8632	0.8063	1.0
13.797	22.0109	16200	11.8627	0.8062	1.0
13.2801	22.2826	16400	11.8626	0.8062	1.0
13.1204	22.5543	16600	11.8630	0.8061	1.0
13.8427	22.8261	16800	11.8625	0.8062	1.0
13.3743	23.0978	17000	11.8632	0.8064	1.0
13.1094	23.3696	17200	11.8537	0.8061	1.0
13.5959	23.6413	17400	11.8488	0.8062	1.0
13.3278	23.9130	17600	11.8533	0.8061	1.0
13.7173	24.1848	17800	11.8624	0.8062	1.0
13.5036	24.4565	18000	11.8654	0.8062	1.0
13.5662	24.7283	18200	11.8656	0.8061	1.0
13.0369	25.0	18400	11.8604	0.8061	1.0
13.4246	25.2717	18600	11.8666	0.8061	1.0
13.7556	25.5435	18800	11.8612	0.8062	1.0
13.6218	25.8152	19000	11.8637	0.8061	1.0
14.1015	26.0870	19200	11.8587	0.8063	1.0
13.3003	26.3587	19400	11.8522	0.8063	1.0
13.3842	26.6304	19600	11.8682	0.8060	1.0
13.4751	26.9022	19800	11.8490	0.8061	1.0
13.3237	27.1739	20000	11.8517	0.8060	1.0
13.7076	27.4457	20200	11.8595	0.8062	1.0
13.0914	27.7174	20400	11.8589	0.8061	1.0
13.4504	27.9891	20600	11.8649	0.8062	1.0
13.4284	28.2609	20800	11.8624	0.8060	1.0
13.7138	28.5326	21000	11.8688	0.8063	1.0
13.5526	28.8043	21200	11.8630	0.8061	1.0
13.8881	29.0761	21400	11.8604	0.8061	1.0
13.5108	29.3478	21600	11.8573	0.8062	1.0
13.3756	29.6196	21800	11.8499	0.8059	1.0
13.225	29.8913	22000	11.8616	0.8061	1.0

Framework versions

Transformers 4.52.1
Pytorch 2.9.1+cu128
Datasets 3.6.0
Tokenizers 0.21.4

ctaguchi/ssc-ady-mms-model-mix-adapt-max-longcv2

ssc-ady-mms-model-mix-adapt-max-longcv2

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

ctaguchi/
ssc-ady-mms-model-mix-adapt-max-longcv2