ViMATCHA

This model is a fine-tuned version of google/matcha-chartqa on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.5748

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: 5e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss
3.9077	0.1603	150	3.3649
3.1926	0.2137	200	2.7038
2.5565	0.2671	250	2.1718
2.1552	0.3205	300	1.7380
1.7834	0.3739	350	1.4287
1.428	0.4274	400	1.2302
1.3867	0.4808	450	1.1280
1.1496	0.5342	500	1.0437
1.2414	0.5876	550	0.9901
1.116	0.6410	600	0.9475
1.12	0.6944	650	0.9073
1.0132	0.7479	700	0.8859
0.9954	0.8013	750	0.8638
1.0057	0.8547	800	0.8496
0.9929	0.9081	850	0.8289
0.945	0.9615	900	0.8207
0.866	1.0150	950	0.8059
0.8901	1.0684	1000	0.7944
0.8382	1.1218	1050	0.7828
0.946	1.1752	1100	0.7748
0.9042	1.2286	1150	0.7662
0.8334	1.2821	1200	0.7549
0.8747	1.3355	1250	0.7501
0.8224	1.3889	1300	0.7424
0.7998	1.4423	1350	0.7382
0.9022	1.4957	1400	0.7341
0.8297	1.5491	1450	0.7226
0.8014	1.6026	1500	0.7165
0.8423	1.6560	1550	0.7129
0.7286	1.7094	1600	0.7063
0.7361	1.7628	1650	0.7040
0.8203	1.8162	1700	0.6982
0.8103	1.8697	1750	0.6945
0.7251	1.9231	1800	0.6926
0.7193	1.9765	1850	0.6910
0.8133	2.0299	1900	0.6843
0.7545	2.0833	1950	0.6862
0.8025	2.1368	2000	0.6768
0.7421	2.1902	2050	0.6769
0.6899	2.2436	2100	0.6744
0.7607	2.2970	2150	0.6690
0.739	2.3504	2200	0.6652
0.7095	2.4038	2250	0.6666
0.7392	2.4573	2300	0.6605
0.7307	2.5107	2350	0.6560
0.685	2.5641	2400	0.6579
0.6419	2.6175	2450	0.6499
0.6894	2.6709	2500	0.6532
0.6288	2.7244	2550	0.6482
0.7024	2.7778	2600	0.6471
0.7717	2.8312	2650	0.6475
0.7389	2.8846	2700	0.6434
0.6944	2.9380	2750	0.6406
0.6512	2.9915	2800	0.6405
0.7187	3.0449	2850	0.6410
0.6676	3.0983	2900	0.6383
0.6513	3.1517	2950	0.6359
0.5821	3.2051	3000	0.6345
0.6642	3.2585	3050	0.6338
0.6475	3.3120	3100	0.6311
0.6999	3.3654	3150	0.6281
0.696	3.4188	3200	0.6329
0.6129	3.4722	3250	0.6263
0.5601	3.5256	3300	0.6257
0.6658	3.5791	3350	0.6209
0.7212	3.6325	3400	0.6217
0.6392	3.6859	3450	0.6178
0.6877	3.7393	3500	0.6202
0.632	3.7927	3550	0.6171
0.6762	3.8462	3600	0.6175
0.6776	3.8996	3650	0.6122
0.6964	3.9530	3700	0.6120
0.6448	4.0064	3750	0.6134
0.6183	4.0598	3800	0.6125
0.6544	4.1132	3850	0.6105
0.6583	4.1667	3900	0.6102
0.6687	4.2201	3950	0.6083
0.6408	4.2735	4000	0.6078
0.6077	4.3269	4050	0.6051
0.5781	4.3803	4100	0.6078
0.645	4.4338	4150	0.6033
0.656	4.4872	4200	0.6044
0.562	4.5406	4250	0.6001
0.6568	4.5940	4300	0.6018
0.6572	4.6474	4350	0.6012
0.5749	4.7009	4400	0.6004
0.5811	4.7543	4450	0.5974
0.5765	4.8077	4500	0.6010
0.6177	4.8611	4550	0.5941
0.5371	4.9145	4600	0.5949
0.5863	4.9679	4650	0.5930
0.6274	5.0214	4700	0.5996
0.592	5.0748	4750	0.5972
0.6286	5.1282	4800	0.5936
0.6056	5.1816	4850	0.5954
0.5697	5.2350	4900	0.5943
0.5562	5.2885	4950	0.5923
0.5173	5.3419	5000	0.5944
0.6037	5.3953	5050	0.5943
0.6125	5.4487	5100	0.5921
0.5729	5.5021	5150	0.5920
0.561	5.5556	5200	0.5908
0.6244	5.6090	5250	0.5877
0.5338	5.6624	5300	0.5876
0.57	5.7158	5350	0.5912
0.555	5.7692	5400	0.5863
0.5914	5.8226	5450	0.5855
0.5785	5.8761	5500	0.5826
0.6014	5.9295	5550	0.5830
0.6192	5.9829	5600	0.5834
0.6025	6.0363	5650	0.5854
0.5327	6.0897	5700	0.5861
0.5958	6.1432	5750	0.5856
0.5319	6.1966	5800	0.5852
0.5722	6.25	5850	0.5848
0.585	6.3034	5900	0.5854
0.5549	6.3568	5950	0.5829
0.5559	6.4103	6000	0.5816
0.5539	6.4637	6050	0.5820
0.5847	6.5171	6100	0.5797
0.5964	6.5705	6150	0.5803
0.5613	6.6239	6200	0.5808
0.5194	6.6774	6250	0.5816
0.5664	6.7308	6300	0.5808
0.5971	6.7842	6350	0.5783
0.4937	6.8376	6400	0.5773
0.577	6.8910	6450	0.5779
0.5422	6.9444	6500	0.5799
0.4881	6.9979	6550	0.5789
0.5806	7.0513	6600	0.5802
0.5029	7.1047	6650	0.5798
0.5436	7.1581	6700	0.5782
0.5298	7.2115	6750	0.5783
0.5947	7.2650	6800	0.5777
0.4975	7.3184	6850	0.5762
0.5652	7.3718	6900	0.5761
0.5214	7.4252	6950	0.5793
0.535	7.4786	7000	0.5791
0.5035	7.5321	7050	0.5774
0.5004	7.5855	7100	0.5769
0.6187	7.6389	7150	0.5778
0.5042	7.6923	7200	0.5770
0.5116	7.7457	7250	0.5760
0.5178	7.7991	7300	0.5744
0.5154	7.8526	7350	0.5748
0.5641	7.9060	7400	0.5764
0.536	7.9594	7450	0.5762
0.5423	8.0128	7500	0.5743
0.5402	8.0662	7550	0.5755
0.5559	8.1197	7600	0.5755
0.4998	8.1731	7650	0.5769
0.5895	8.2265	7700	0.5777
0.5252	8.2799	7750	0.5786
0.4802	8.3333	7800	0.5774
0.5285	8.3868	7850	0.5742
0.4931	8.4402	7900	0.5740
0.5126	8.4936	7950	0.5761
0.5376	8.5470	8000	0.5754
0.4825	8.6004	8050	0.5761
0.509	8.6538	8100	0.5762
0.4823	8.7073	8150	0.5747
0.5379	8.7607	8200	0.5733
0.5283	8.8141	8250	0.5740
0.4662	8.8675	8300	0.5743
0.5325	8.9209	8350	0.5727
0.5628	8.9744	8400	0.5727
0.4885	9.0278	8450	0.5731
0.5187	9.0812	8500	0.5761
0.5286	9.1346	8550	0.5752
0.4738	9.1880	8600	0.5744
0.4931	9.2415	8650	0.5733
0.5403	9.2949	8700	0.5751
0.4927	9.3483	8750	0.5755
0.5415	9.4017	8800	0.5743
0.5627	9.4551	8850	0.5746
0.468	9.5085	8900	0.5743
0.5002	9.5620	8950	0.5751
0.4891	9.6154	9000	0.5740
0.5456	9.6688	9050	0.5747
0.5054	9.7222	9100	0.5747
0.5453	9.7756	9150	0.5741
0.477	9.8291	9200	0.5745
0.4798	9.8825	9250	0.5747
0.5003	9.9359	9300	0.5748
0.4785	9.9893	9350	0.5748

Framework versions

• Transformers 4.40.1
• Pytorch 2.2.1+cu121
• Datasets 2.19.0
• Tokenizers 0.19.1