carolinetfls/plant-seedlings-model-ConvNet-final

plant-seedlings-model-ConvNet-all-train

This model is a fine-tuned version of facebook/convnext-tiny-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.2653
• Accuracy: 0.9392

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: 0.0002
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.2307	0.25	100	0.4912	0.8729
0.0652	0.49	200	0.3280	0.9085
0.1854	0.74	300	0.4850	0.8711
0.1831	0.98	400	0.3827	0.8938
0.1636	1.23	500	0.4071	0.9012
0.0868	1.47	600	0.3980	0.8999
0.2298	1.72	700	0.4855	0.8846
0.2291	1.97	800	0.4019	0.8883
0.2698	2.21	900	0.3855	0.8944
0.0923	2.46	1000	0.3690	0.8938
0.1396	2.7	1100	0.4715	0.8760
0.174	2.95	1200	0.3710	0.9006
0.1009	3.19	1300	0.3481	0.9030
0.1162	3.44	1400	0.3502	0.9153
0.1737	3.69	1500	0.4034	0.8999
0.2478	3.93	1600	0.4053	0.8913
0.1471	4.18	1700	0.3555	0.9036
0.1873	4.42	1800	0.3769	0.9122
0.0615	4.67	1900	0.4147	0.8987
0.1718	4.91	2000	0.2779	0.9214
0.1012	5.16	2100	0.3239	0.9159
0.0967	5.41	2200	0.3290	0.9079
0.0873	5.65	2300	0.4057	0.9055
0.0567	5.9	2400	0.3821	0.9018
0.1356	6.14	2500	0.4183	0.8944
0.168	6.39	2600	0.3755	0.9067
0.1592	6.63	2700	0.3413	0.9079
0.1239	6.88	2800	0.3299	0.9091
0.0382	7.13	2900	0.3391	0.9165
0.1167	7.37	3000	0.4274	0.8987
0.109	7.62	3100	0.3952	0.9018
0.0591	7.86	3200	0.4043	0.9122
0.1407	8.11	3300	0.3325	0.9134
0.054	8.35	3400	0.3333	0.9177
0.0633	8.6	3500	0.3275	0.9208
0.1038	8.85	3600	0.3982	0.9042
0.0435	9.09	3700	0.3656	0.9190
0.1549	9.34	3800	0.3367	0.9190
0.2299	9.58	3900	0.3872	0.9134
0.0375	9.83	4000	0.3206	0.9245
0.0204	10.07	4100	0.3133	0.9263
0.1208	10.32	4200	0.3373	0.9196
0.0617	10.57	4300	0.3045	0.9220
0.1426	10.81	4400	0.2972	0.9294
0.0351	11.06	4500	0.3409	0.9147
0.0311	11.3	4600	0.3003	0.9233
0.1255	11.55	4700	0.3447	0.9282
0.0569	11.79	4800	0.2703	0.9331
0.0918	12.04	4900	0.3170	0.9245
0.0656	12.29	5000	0.3223	0.9190
0.0971	12.53	5100	0.3209	0.9196
0.0742	12.78	5200	0.3030	0.9282
0.0662	13.02	5300	0.2780	0.9319
0.0453	13.27	5400	0.3360	0.9227
0.0869	13.51	5500	0.2417	0.9343
0.1786	13.76	5600	0.3078	0.9263
0.1563	14.0	5700	0.3046	0.9312
0.0584	14.25	5800	0.3011	0.9288
0.0783	14.5	5900	0.2705	0.9288
0.0486	14.74	6000	0.2583	0.9288
0.094	14.99	6100	0.2854	0.9282
0.0852	15.23	6200	0.2693	0.9325
0.0665	15.48	6300	0.2754	0.9282
0.0948	15.72	6400	0.2598	0.9349
0.0368	15.97	6500	0.2875	0.9355
0.0031	16.22	6600	0.2679	0.9325
0.0796	16.46	6700	0.2642	0.9300
0.0903	16.71	6800	0.2977	0.9269
0.0952	16.95	6900	0.2615	0.9337
0.1344	17.2	7000	0.2948	0.9251
0.0854	17.44	7100	0.2748	0.9368
0.0891	17.69	7200	0.2386	0.9325
0.1202	17.94	7300	0.2509	0.9355
0.0832	18.18	7400	0.2406	0.9398
0.0949	18.43	7500	0.2356	0.9386
0.0404	18.67	7600	0.2415	0.9386
0.1008	18.92	7700	0.2582	0.9355
0.092	19.16	7800	0.2724	0.9325
0.0993	19.41	7900	0.2655	0.9325
0.0593	19.66	8000	0.2423	0.9386
0.1011	19.9	8100	0.2653	0.9392

Framework versions

• Transformers 4.28.1
• Pytorch 2.0.0+cu118
• Datasets 2.11.0
• Tokenizers 0.13.3