smolvlm-mmocr-sft

This model is a fine-tuned version of HuggingFaceTB/SmolVLM-Instruct on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.1884

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: 0.0003
train_batch_size: 1
eval_batch_size: 1
seed: 42
gradient_accumulation_steps: 8
total_train_batch_size: 8
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: 200
num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss
0.7936	0.0138	20	0.8587
0.7632	0.0275	40	0.8587
0.7891	0.0413	60	0.8584
0.7804	0.0551	80	0.8581
0.802	0.0689	100	0.8572
0.7793	0.0826	120	0.8562
0.7857	0.0964	140	0.8546
0.7788	0.1102	160	0.8524
0.7979	0.1240	180	0.8498
0.78	0.1377	200	0.8459
0.7597	0.1515	220	0.8410
0.7325	0.1653	240	0.8366
0.761	0.1791	260	0.8316
0.7802	0.1928	280	0.8265
0.7239	0.2066	300	0.8217
0.7196	0.2204	320	0.8176
0.7355	0.2342	340	0.8137
0.72	0.2479	360	0.8097
0.7521	0.2617	380	0.8061
0.725	0.2755	400	0.8030
0.7314	0.2893	420	0.8001
0.7148	0.3030	440	0.7972
0.7272	0.3168	460	0.7945
0.7194	0.3306	480	0.7919
0.7303	0.3444	500	0.7893
0.7205	0.3581	520	0.7868
0.7239	0.3719	540	0.7846
0.7029	0.3857	560	0.7824
0.695	0.3994	580	0.7804
0.7207	0.4132	600	0.7785
0.7406	0.4270	620	0.7763
0.7032	0.4408	640	0.7747
0.7011	0.4545	660	0.7726
0.6904	0.4683	680	0.7710
0.6949	0.4821	700	0.7694
0.7226	0.4959	720	0.7676
0.6762	0.5096	740	0.7661
0.7739	0.5234	760	0.7646
0.7166	0.5372	780	0.7633
0.6984	0.5510	800	0.7616
0.6933	0.5647	820	0.7603
0.679	0.5785	840	0.7592
0.7128	0.5923	860	0.7578
0.6924	0.6061	880	0.7567
0.6899	0.6198	900	0.7553
0.6965	0.6336	920	0.7542
0.6746	0.6474	940	0.7531
0.6708	0.6612	960	0.7518
0.6746	0.6749	980	0.7506
0.6747	0.6887	1000	0.7497
0.6814	0.7025	1020	0.7486
0.6758	0.7163	1040	0.7475
0.6752	0.7300	1060	0.7465
0.7218	0.7438	1080	0.7454
0.6733	0.7576	1100	0.7443
0.685	0.7713	1120	0.7435
0.6592	0.7851	1140	0.7427
0.6827	0.7989	1160	0.7417
0.732	0.8127	1180	0.7410
0.6803	0.8264	1200	0.7401
0.6643	0.8402	1220	0.7392
0.6805	0.8540	1240	0.7385
0.7031	0.8678	1260	0.7377
0.6857	0.8815	1280	0.7371
0.6663	0.8953	1300	0.7364
0.6788	0.9091	1320	0.7354
0.7035	0.9229	1340	0.7347
0.6669	0.9366	1360	0.7343
0.6869	0.9504	1380	0.7333
0.6996	0.9642	1400	0.7326
0.6985	0.9780	1420	0.7320
0.6678	0.9917	1440	0.7312
0.6306	1.0055	1460	0.7307
0.6634	1.0193	1480	0.7300
0.6708	1.0331	1500	0.7293
0.6596	1.0468	1520	0.7290
0.6837	1.0606	1540	0.7282
0.684	1.0744	1560	0.7276
0.6889	1.0882	1580	0.7269
0.6758	1.1019	1600	0.7265
0.6513	1.1157	1620	0.7260
0.6555	1.1295	1640	0.7255
0.66	1.1433	1660	0.7248
0.6808	1.1570	1680	0.7244
0.6482	1.1708	1700	0.7239
0.6662	1.1846	1720	0.7236
0.6438	1.1983	1740	0.7230
0.6369	1.2121	1760	0.7226
0.6516	1.2259	1780	0.7223
0.6547	1.2397	1800	0.7217
0.6489	1.2534	1820	0.7212
0.6729	1.2672	1840	0.7206
0.6717	1.2810	1860	0.7202
0.6622	1.2948	1880	0.7198
0.6587	1.3085	1900	0.7192
0.6796	1.3223	1920	0.7190
0.6571	1.3361	1940	0.7185
0.6237	1.3499	1960	0.7182
0.6473	1.3636	1980	0.7177
0.6528	1.3774	2000	0.7172
0.6795	1.3912	2020	0.7169
0.6397	1.4050	2040	0.7164
0.6471	1.4187	2060	0.7162
0.6247	1.4325	2080	0.7157
0.6623	1.4463	2100	0.7154
0.6656	1.4601	2120	0.7149
0.6573	1.4738	2140	0.7146
0.6317	1.4876	2160	0.7144
0.6455	1.5014	2180	0.7141
0.6426	1.5152	2200	0.7136
0.6472	1.5289	2220	0.7133
0.6447	1.5427	2240	0.7129
0.6618	1.5565	2260	0.7127
0.6706	1.5702	2280	0.7121
0.6581	1.5840	2300	0.7120
0.6337	1.5978	2320	0.7117
0.6526	1.6116	2340	0.7115
0.6379	1.6253	2360	0.7113
0.6366	1.6391	2380	0.7110
0.659	1.6529	2400	0.7107
0.6685	1.6667	2420	0.7103
0.6317	1.6804	2440	0.7100
0.6611	1.6942	2460	0.7098
0.6431	1.7080	2480	0.7094
0.6249	1.7218	2500	0.7091
0.6502	1.7355	2520	0.7088
0.6506	1.7493	2540	0.7086
0.6707	1.7631	2560	0.7083
0.6399	1.7769	2580	0.7081
0.6189	1.7906	2600	0.7079
0.6167	1.8044	2620	0.7078
0.6469	1.8182	2640	0.7075
0.6611	1.8320	2660	0.7073
0.6446	1.8457	2680	0.7071
0.6374	1.8595	2700	0.7068
0.6394	1.8733	2720	0.7066
0.6195	1.8871	2740	0.7063
0.6255	1.9008	2760	0.7060
0.6346	1.9146	2780	0.7059
0.6375	1.9284	2800	0.7058
0.6254	1.9421	2820	0.7056
0.6203	1.9559	2840	0.7056
0.6619	1.9697	2860	0.7039
0.6151	1.9835	2880	0.6930
0.6233	1.9972	2900	0.6823
0.5892	2.0110	2920	0.6747
0.6042	2.0248	2940	0.6678
0.6045	2.0386	2960	0.6627
0.5495	2.0523	2980	0.6552
0.579	2.0661	3000	0.6479
0.5868	2.0799	3020	0.6444
0.564	2.0937	3040	0.6419
0.5657	2.1074	3060	0.6381
0.6204	2.1212	3080	0.6348
0.5565	2.1350	3100	0.6314
0.5645	2.1488	3120	0.6253
0.5375	2.1625	3140	0.6246
0.5386	2.1763	3160	0.6194
0.5427	2.1901	3180	0.6194
0.556	2.2039	3200	0.6147
0.5428	2.2176	3220	0.6134
0.5598	2.2314	3240	0.6102
0.5304	2.2452	3260	0.6081
0.5201	2.2590	3280	0.6084
0.5168	2.2727	3300	0.6081
0.5309	2.2865	3320	0.6060
0.5051	2.3003	3340	0.6055
0.516	2.3140	3360	0.6026
0.5164	2.3278	3380	0.6016
0.5445	2.3416	3400	0.5978
0.5139	2.3554	3420	0.5984
0.5178	2.3691	3440	0.5968
0.5028	2.3829	3460	0.5974
0.5499	2.3967	3480	0.5940
0.493	2.4105	3500	0.5956
0.5218	2.4242	3520	0.6022
0.5468	2.4380	3540	0.5990
0.5282	2.4518	3560	0.5985
0.521	2.4656	3580	0.5965
0.5267	2.4793	3600	0.5952
0.4896	2.4931	3620	0.5923
0.5165	2.5069	3640	0.5877
0.4976	2.5207	3660	0.5880
0.5296	2.5344	3680	0.5865
0.506	2.5482	3700	0.5853
0.4869	2.5620	3720	0.5822
0.5062	2.5758	3740	0.5800
0.5116	2.5895	3760	0.5818
0.4781	2.6033	3780	0.5800
0.4819	2.6171	3800	0.5784
0.4937	2.6309	3820	0.5768
0.4934	2.6446	3840	0.5747
0.4932	2.6584	3860	0.5729
0.4938	2.6722	3880	0.5728
0.4741	2.6860	3900	0.5709
0.5275	2.6997	3920	0.5691
0.4808	2.7135	3940	0.5667
0.5362	2.7273	3960	0.5669
0.4926	2.7410	3980	0.5656
0.452	2.7548	4000	0.5679
0.482	2.7686	4020	0.5662
0.5015	2.7824	4040	0.5646
0.4782	2.7961	4060	0.5644
0.4462	2.8099	4080	0.5668
0.5052	2.8237	4100	0.5630
0.4967	2.8375	4120	0.5625
0.4944	2.8512	4140	0.5599
0.4818	2.8650	4160	0.5635
0.4883	2.8788	4180	0.5629
0.4817	2.8926	4200	0.5605
0.4229	2.9063	4220	0.5576
0.466	2.9201	4240	0.5557
0.4666	2.9339	4260	0.5596
0.4579	2.9477	4280	0.5565
0.4947	2.9614	4300	0.5535
0.4747	2.9752	4320	0.5531
0.4776	2.9890	4340	0.5535
0.493	3.0028	4360	0.5543
0.4521	3.0165	4380	0.5535
0.4488	3.0303	4400	0.5515
0.4858	3.0441	4420	0.5528
0.4496	3.0579	4440	0.5518
0.4564	3.0716	4460	0.5516
0.4418	3.0854	4480	0.5488
0.4803	3.0992	4500	0.5477
0.4678	3.1129	4520	0.5590
0.495	3.1267	4540	0.5565
0.4729	3.1405	4560	0.5506
0.491	3.1543	4580	0.5578
0.4929	3.1680	4600	0.5468
0.4558	3.1818	4620	0.5410
0.4504	3.1956	4640	0.5394
0.4641	3.2094	4660	0.5370
0.4694	3.2231	4680	0.5399
0.4549	3.2369	4700	0.5296
0.4759	3.2507	4720	0.5266
0.4405	3.2645	4740	0.5258
0.4444	3.2782	4760	0.5155
0.4494	3.2920	4780	0.5159
0.4451	3.3058	4800	0.5025
0.4292	3.3196	4820	0.4966
0.4197	3.3333	4840	0.4877
0.454	3.3471	4860	0.4852
0.3973	3.3609	4880	0.4778
0.3518	3.3747	4900	0.4709
0.4021	3.3884	4920	0.4593
0.4024	3.4022	4940	0.4510
0.3711	3.4160	4960	0.4521
0.3724	3.4298	4980	0.4366
0.3733	3.4435	5000	0.4260
0.3816	3.4573	5020	0.4199
0.3673	3.4711	5040	0.4169
0.3428	3.4848	5060	0.4063
0.3369	3.4986	5080	0.3998
0.3553	3.5124	5100	0.3898
0.3304	3.5262	5120	0.3827
0.3403	3.5399	5140	0.3773
0.3	3.5537	5160	0.3737
0.3441	3.5675	5180	0.3787
0.3022	3.5813	5200	0.3602
0.3205	3.5950	5220	0.3591
0.304	3.6088	5240	0.3527
0.3291	3.6226	5260	0.3457
0.2545	3.6364	5280	0.3405
0.2878	3.6501	5300	0.3324
0.2974	3.6639	5320	0.3316
0.278	3.6777	5340	0.3256
0.3123	3.6915	5360	0.3239
0.2838	3.7052	5380	0.3164
0.2876	3.7190	5400	0.3143
0.2974	3.7328	5420	0.3113
0.2508	3.7466	5440	0.3087
0.2793	3.7603	5460	0.3043
0.2858	3.7741	5480	0.2988
0.2761	3.7879	5500	0.2918
0.2378	3.8017	5520	0.2905
0.2419	3.8154	5540	0.2908
0.2414	3.8292	5560	0.2874
0.2702	3.8430	5580	0.2871
0.2875	3.8567	5600	0.2836
0.2457	3.8705	5620	0.2810
0.2574	3.8843	5640	0.2779
0.2391	3.8981	5660	0.2777
0.2426	3.9118	5680	0.2750
0.2459	3.9256	5700	0.2735
0.2283	3.9394	5720	0.2703
0.2269	3.9532	5740	0.2662
0.2051	3.9669	5760	0.2645
0.2235	3.9807	5780	0.2612
0.2038	3.9945	5800	0.2594
0.2268	4.0083	5820	0.2593
0.2068	4.0220	5840	0.2538
0.245	4.0358	5860	0.2508
0.2426	4.0496	5880	0.2520
0.1992	4.0634	5900	0.2506
0.2809	4.0771	5920	0.2482
0.195	4.0909	5940	0.2422
0.2125	4.1047	5960	0.2429
0.2376	4.1185	5980	0.2428
0.2237	4.1322	6000	0.2406
0.2138	4.1460	6020	0.2395
0.2001	4.1598	6040	0.2371
0.2051	4.1736	6060	0.2351
0.2127	4.1873	6080	0.2328
0.173	4.2011	6100	0.2335
0.1769	4.2149	6120	0.2344
0.1615	4.2287	6140	0.2298
0.1935	4.2424	6160	0.2286
0.1954	4.2562	6180	0.2290
0.208	4.2700	6200	0.2267
0.1896	4.2837	6220	0.2232
0.2094	4.2975	6240	0.2206
0.1854	4.3113	6260	0.2212
0.1948	4.3251	6280	0.2196
0.1667	4.3388	6300	0.2194
0.1926	4.3526	6320	0.2168
0.1657	4.3664	6340	0.2158
0.1802	4.3802	6360	0.2140
0.1564	4.3939	6380	0.2164
0.1864	4.4077	6400	0.2145
0.187	4.4215	6420	0.2145
0.1868	4.4353	6440	0.2130
0.189	4.4490	6460	0.2107
0.1808	4.4628	6480	0.2102
0.1828	4.4766	6500	0.2079
0.1771	4.4904	6520	0.2081
0.1856	4.5041	6540	0.2061
0.1685	4.5179	6560	0.2045
0.1567	4.5317	6580	0.2059
0.1913	4.5455	6600	0.2051
0.1937	4.5592	6620	0.2031
0.1823	4.5730	6640	0.2024
0.1613	4.5868	6660	0.2021
0.1837	4.6006	6680	0.2012
0.1419	4.6143	6700	0.2012
0.1769	4.6281	6720	0.1997
0.1683	4.6419	6740	0.1977
0.1614	4.6556	6760	0.1986
0.1686	4.6694	6780	0.1990
0.1851	4.6832	6800	0.1976
0.1529	4.6970	6820	0.1978
0.1746	4.7107	6840	0.2065
0.1474	4.7245	6860	0.1999
0.1415	4.7383	6880	0.1980
0.1709	4.7521	6900	0.1965
0.1673	4.7658	6920	0.1958
0.1732	4.7796	6940	0.1953
0.1424	4.7934	6960	0.1947
0.1271	4.8072	6980	0.1940
0.1893	4.8209	7000	0.1936
0.1696	4.8347	7020	0.1917
0.1644	4.8485	7040	0.1916
0.1509	4.8623	7060	0.1912
0.1507	4.8760	7080	0.1912
0.1471	4.8898	7100	0.1900
0.1554	4.9036	7120	0.1895
0.1547	4.9174	7140	0.1892
0.1787	4.9311	7160	0.1888
0.1436	4.9449	7180	0.1889
0.1522	4.9587	7200	0.1886
0.1657	4.9725	7220	0.1886
0.1716	4.9862	7240	0.1885
0.1889	5.0	7260	0.1884

Framework versions

PEFT 0.14.0
Transformers 4.49.0
Pytorch 2.5.1+cu124
Datasets 3.3.2
Tokenizers 0.21.1

chuuhtetnaing/smolvlm-mmocr-sft

smolvlm-mmocr-sft

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Seed this model