QuanHoangNgoc/lock_s2t-small-uit-vimd_211008

s2t-small-uit-vimd-finetuned

This model is a fine-tuned version of lock_s2t-small-uit-vimd_192355 on the UIT-ViMD dataset. It achieves the following results on the evaluation set:

• Loss: 0.9924
• Wer: 0.2993

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: 3e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• 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: 1500
• training_steps: 40000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.7572	0.0490	46	1.8361	0.6165
1.7571	0.0980	92	1.8453	0.6250
1.7017	0.1470	138	1.8005	0.6551
1.6797	0.1960	184	1.7893	0.6621
1.6581	0.2449	230	1.7749	0.6178
1.6088	0.2939	276	1.7541	0.6911
1.624	0.3429	322	1.7435	0.6374
1.6397	0.3919	368	1.7343	0.6046
1.5714	0.4409	414	1.7033	0.5620
1.5615	0.4899	460	1.6827	0.6565
1.5751	0.5389	506	1.6752	0.5916
1.5767	0.5879	552	1.6542	0.5945
1.5185	0.6368	598	1.6385	0.6006
1.5641	0.6858	644	1.6269	0.5714
1.4975	0.7348	690	1.6170	0.5542
1.5288	0.7838	736	1.6039	0.5483
1.5106	0.8328	782	1.6066	0.5137
1.4704	0.8818	828	1.5884	0.5148
1.4625	0.9308	874	1.5818	0.5367
1.4827	0.9798	920	1.5753	0.5805
1.4287	1.0288	966	1.5727	0.6089
1.423	1.0777	1012	1.5546	0.5685
1.4065	1.1267	1058	1.5430	0.5897
1.4159	1.1757	1104	1.5189	0.5792
1.4114	1.2247	1150	1.5148	0.5591
1.3863	1.2737	1196	1.5100	0.5359
1.4605	1.3227	1242	1.5091	0.4837
1.4049	1.3717	1288	1.4959	0.4890
1.3761	1.4207	1334	1.4824	0.4645
1.4008	1.4696	1380	1.4540	0.4528
1.3893	1.5186	1426	1.4479	0.4949
1.3644	1.5676	1472	1.4556	0.5312
1.3774	1.6166	1518	1.4441	0.5046
1.3733	1.6656	1564	1.4311	0.4608
1.3722	1.7146	1610	1.4247	0.4935
1.3116	1.7636	1656	1.4148	0.4273
1.3391	1.8126	1702	1.4128	0.5014
1.3505	1.8616	1748	1.4087	0.5013
1.3371	1.9105	1794	1.3920	0.4507
1.3235	1.9595	1840	1.3964	0.4842
1.3062	2.0085	1886	1.3700	0.5218
1.2194	2.0575	1932	1.3757	0.4731
1.2488	2.1065	1978	1.3812	0.4668
1.2392	2.1555	2024	1.3662	0.4557
1.2676	2.2045	2070	1.3563	0.4734
1.2242	2.2535	2116	1.3378	0.4319
1.2472	2.3024	2162	1.3513	0.4442
1.2285	2.3514	2208	1.3445	0.4385
1.2297	2.4004	2254	1.3343	0.4627
1.2016	2.4494	2300	1.3360	0.4493
1.2439	2.4984	2346	1.3183	0.4480
1.2096	2.5474	2392	1.3168	0.4394
1.227	2.5964	2438	1.3125	0.4417
1.1905	2.6454	2484	1.3057	0.4047
1.1681	2.6944	2530	1.2961	0.4139
1.2209	2.7433	2576	1.2938	0.4135
1.2029	2.7923	2622	1.2849	0.4332
1.1977	2.8413	2668	1.2938	0.4267
1.1976	2.8903	2714	1.2877	0.4259
1.1939	2.9393	2760	1.2771	0.4133
1.2151	2.9883	2806	1.2670	0.4197
1.1395	3.0373	2852	1.2865	0.4354
1.0845	3.0863	2898	1.2641	0.4316
1.0471	3.1353	2944	1.2609	0.4099
1.1189	3.1842	2990	1.2681	0.4265
1.1358	3.2332	3036	1.2564	0.3983
1.1279	3.2822	3082	1.2441	0.4326
1.157	3.3312	3128	1.2515	0.3814
1.0963	3.3802	3174	1.2468	0.4001
1.0981	3.4292	3220	1.2414	0.4206
1.1165	3.4782	3266	1.2393	0.3809
1.0594	3.5272	3312	1.2229	0.3851
1.0668	3.5761	3358	1.2264	0.3760
1.0991	3.6251	3404	1.2175	0.3931
1.1016	3.6741	3450	1.2042	0.3946
1.068	3.7231	3496	1.2146	0.3618
1.0918	3.7721	3542	1.1883	0.3902
1.077	3.8211	3588	1.2077	0.3696
1.0912	3.8701	3634	1.1838	0.3970
1.0739	3.9191	3680	1.1903	0.3668
1.1475	3.9681	3726	1.1983	0.3841
1.0658	4.0170	3772	1.1937	0.3725
1.0177	4.0660	3818	1.1894	0.3790
0.9442	4.1150	3864	1.2047	0.3572
1.0337	4.1640	3910	1.1978	0.3561
1.0644	4.2130	3956	1.2001	0.3586
0.9971	4.2620	4002	1.1912	0.3502
1.0139	4.3110	4048	1.1946	0.3671
1.007	4.3600	4094	1.1893	0.3618
1.03	4.4089	4140	1.1738	0.3644
1.0011	4.4579	4186	1.1887	0.4018
0.9874	4.5069	4232	1.1751	0.3741
1.0079	4.5559	4278	1.1790	0.3867
1.0059	4.6049	4324	1.1649	0.3768
1.0051	4.6539	4370	1.1648	0.3918
1.0397	4.7029	4416	1.1517	0.3698
1.0216	4.7519	4462	1.1600	0.3585
0.9799	4.8009	4508	1.1628	0.3538
1.0347	4.8498	4554	1.1560	0.3562
0.9958	4.8988	4600	1.1504	0.3481
1.019	4.9478	4646	1.1421	0.3637
1.0019	4.9968	4692	1.1248	0.3765
0.9473	5.0458	4738	1.1379	0.3521
0.9194	5.0948	4784	1.1369	0.3481
0.928	5.1438	4830	1.1418	0.3495
0.9204	5.1928	4876	1.1392	0.3977
0.9382	5.2417	4922	1.1333	0.3462
0.9572	5.2907	4968	1.1376	0.3471
0.939	5.3397	5014	1.1436	0.3446
0.9695	5.3887	5060	1.1225	0.3422
0.9303	5.4377	5106	1.1261	0.3562
0.9572	5.4867	5152	1.1320	0.3543
0.9614	5.5357	5198	1.1267	0.3564
0.9328	5.5847	5244	1.1318	0.3872
0.9005	5.6337	5290	1.1323	0.3379
0.9758	5.6826	5336	1.1254	0.3390
0.9411	5.7316	5382	1.1224	0.3919
0.9359	5.7806	5428	1.1112	0.3617
0.9433	5.8296	5474	1.0961	0.3325
0.9454	5.8786	5520	1.0988	0.3443
0.923	5.9276	5566	1.1049	0.3267
0.9525	5.9766	5612	1.0973	0.3682
0.9083	6.0256	5658	1.0923	0.3368
0.912	6.0745	5704	1.1067	0.3322
0.858	6.1235	5750	1.1026	0.3382
0.9162	6.1725	5796	1.0954	0.3486
0.8776	6.2215	5842	1.0930	0.3620
0.8828	6.2705	5888	1.0965	0.3266
0.8723	6.3195	5934	1.0964	0.3175
0.8524	6.3685	5980	1.0978	0.3379
0.8791	6.4175	6026	1.1007	0.3640
0.8962	6.4665	6072	1.0807	0.3363
0.9189	6.5154	6118	1.0934	0.3363
0.8487	6.5644	6164	1.0807	0.3247
0.8904	6.6134	6210	1.0922	0.3398
0.8679	6.6624	6256	1.0910	0.3491
0.89	6.7114	6302	1.0819	0.3277
0.9157	6.7604	6348	1.0806	0.3457
0.8636	6.8094	6394	1.0757	0.3245
0.8912	6.8584	6440	1.0726	0.3460
0.9072	6.9073	6486	1.0707	0.3315
0.8774	6.9563	6532	1.0611	0.3221
0.8869	7.0053	6578	1.0692	0.3204
0.8122	7.0543	6624	1.0877	0.3116
0.8716	7.1033	6670	1.0745	0.3306
0.8196	7.1523	6716	1.0662	0.3261
0.8396	7.2013	6762	1.0625	0.3172
0.8319	7.2503	6808	1.0560	0.3226
0.848	7.2993	6854	1.0851	0.3277
0.8479	7.3482	6900	1.0611	0.3226
0.8153	7.3972	6946	1.0742	0.3350
0.8269	7.4462	6992	1.0741	0.3334
0.8385	7.4952	7038	1.0639	0.3465
0.8096	7.5442	7084	1.0757	0.3494
0.8544	7.5932	7130	1.0519	0.3507
0.8415	7.6422	7176	1.0547	0.3524
0.8645	7.6912	7222	1.0560	0.3416
0.8594	7.7401	7268	1.0589	0.3264
0.8092	7.7891	7314	1.0535	0.3441
0.8595	7.8381	7360	1.0498	0.3379
0.8432	7.8871	7406	1.0523	0.3252
0.8735	7.9361	7452	1.0557	0.3749
0.8588	7.9851	7498	1.0536	0.3186
0.8199	8.0341	7544	1.0510	0.3318
0.7679	8.0831	7590	1.0604	0.3446
0.789	8.1321	7636	1.0530	0.3444
0.7756	8.1810	7682	1.0377	0.3570
0.7923	8.2300	7728	1.0643	0.3747
0.7933	8.2790	7774	1.0546	0.3181
0.7804	8.3280	7820	1.0455	0.3275
0.7988	8.3770	7866	1.0362	0.3358
0.8083	8.4260	7912	1.0341	0.3301
0.7837	8.4750	7958	1.0330	0.3460
0.8011	8.5240	8004	1.0370	0.3537
0.8241	8.5729	8050	1.0383	0.3272
0.7776	8.6219	8096	1.0266	0.3468
0.7931	8.6709	8142	1.0358	0.3177
0.8014	8.7199	8188	1.0329	0.3178
0.814	8.7689	8234	1.0409	0.3838
0.8312	8.8179	8280	1.0403	0.3232
0.8477	8.8669	8326	1.0332	0.3193
0.816	8.9159	8372	1.0259	0.3216
0.8083	8.9649	8418	1.0208	0.3099
0.8057	9.0138	8464	1.0238	0.3218
0.7335	9.0628	8510	1.0403	0.3240
0.7546	9.1118	8556	1.0388	0.3309
0.7287	9.1608	8602	1.0240	0.3197
0.7685	9.2098	8648	1.0375	0.3317
0.7734	9.2588	8694	1.0431	0.3336
0.8028	9.3078	8740	1.0522	0.3272
0.7345	9.3568	8786	1.0398	0.3146
0.7659	9.4058	8832	1.0259	0.3548
0.7735	9.4547	8878	1.0353	0.3199
0.7758	9.5037	8924	1.0259	0.3487
0.7496	9.5527	8970	1.0229	0.3287
0.7764	9.6017	9016	1.0332	0.3502
0.7826	9.6507	9062	1.0285	0.3231
0.7731	9.6997	9108	1.0202	0.3481
0.7714	9.7487	9154	1.0272	0.3344
0.7781	9.7977	9200	1.0170	0.3204
0.8076	9.8466	9246	1.0104	0.3266
0.7609	9.8956	9292	1.0145	0.3127
0.7595	9.9446	9338	1.0116	0.3118
0.7975	9.9936	9384	1.0146	0.3099
0.7265	10.0426	9430	1.0183	0.3097
0.702	10.0916	9476	1.0155	0.3114
0.7254	10.1406	9522	1.0299	0.3419
0.7453	10.1896	9568	1.0093	0.3127
0.7232	10.2386	9614	1.0165	0.3242
0.7516	10.2875	9660	1.0097	0.3395
0.7404	10.3365	9706	1.0181	0.3535
0.7419	10.3855	9752	1.0072	0.3199
0.7193	10.4345	9798	1.0088	0.3116
0.6984	10.4835	9844	1.0169	0.3216
0.7382	10.5325	9890	1.0142	0.3462
0.7495	10.5815	9936	1.0101	0.3336
0.7388	10.6305	9982	1.0111	0.3389
0.7348	10.6794	10028	1.0037	0.3397
0.7586	10.7284	10074	0.9927	0.3521
0.7304	10.7774	10120	1.0082	0.3298
0.7667	10.8264	10166	1.0086	0.3279
0.7357	10.8754	10212	0.9974	0.3196
0.793	10.9244	10258	1.0059	0.3545
0.7668	10.9734	10304	0.9908	0.3186
0.7007	11.0224	10350	1.0085	0.3162
0.7216	11.0714	10396	0.9879	0.3159
0.6698	11.1203	10442	0.9882	0.3197
0.7225	11.1693	10488	1.0060	0.3102
0.709	11.2183	10534	1.0000	0.2944
0.6797	11.2673	10580	1.0035	0.3127
0.7062	11.3163	10626	0.9957	0.3146
0.7043	11.3653	10672	0.9986	0.3097
0.6759	11.4143	10718	1.0029	0.3201
0.7107	11.4633	10764	0.9810	0.3609
0.7013	11.5122	10810	0.9798	0.3280
0.6903	11.5612	10856	0.9827	0.3207
0.6967	11.6102	10902	0.9906	0.3105
0.7206	11.6592	10948	0.9862	0.3127
0.7435	11.7082	10994	0.9853	0.3078
0.7311	11.7572	11040	1.0085	0.3309
0.7318	11.8062	11086	1.0002	0.3172
0.7397	11.8552	11132	0.9791	0.3169
0.7179	11.9042	11178	0.9846	0.3145
0.7143	11.9531	11224	0.9977	0.3033
0.7496	12.0021	11270	0.9967	0.3285
0.6547	12.0511	11316	0.9954	0.3065
0.6897	12.1001	11362	1.0002	0.3076
0.6642	12.1491	11408	0.9978	0.3142
0.661	12.1981	11454	0.9927	0.3057
0.6781	12.2471	11500	0.9992	0.3162
0.6856	12.2961	11546	1.0077	0.3151
0.6776	12.3450	11592	0.9895	0.3197
0.6921	12.3940	11638	1.0124	0.3204
0.6982	12.4430	11684	1.0040	0.3103
0.6885	12.4920	11730	0.9951	0.3113
0.679	12.5410	11776	1.0012	0.3089
0.6974	12.5900	11822	0.9946	0.3202
0.6862	12.6390	11868	0.9875	0.3076
0.6929	12.6880	11914	0.9867	0.3099
0.6526	12.7370	11960	0.9897	0.3078
0.7107	12.7859	12006	0.9849	0.3228
0.6697	12.8349	12052	0.9829	0.3430
0.6986	12.8839	12098	0.9949	0.3373
0.7056	12.9329	12144	0.9660	0.3389
0.722	12.9819	12190	0.9513	0.3239
0.6482	13.0309	12236	0.9834	0.3559
0.651	13.0799	12282	0.9834	0.3317
0.612	13.1289	12328	0.9858	0.3354
0.6633	13.1778	12374	0.9837	0.3266
0.6286	13.2268	12420	0.9792	0.3303
0.6242	13.2758	12466	0.9844	0.3293
0.663	13.3248	12512	0.9814	0.3285
0.6688	13.3738	12558	0.9924	0.2993

Framework versions

• Transformers 4.49.0
• Pytorch 2.6.0+cu124
• Datasets 3.4.0
• Tokenizers 0.21.0