teacher

This model is a fine-tuned version of microsoft/resnet-18 on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 1.6827
Accuracy: 0.5862
Precision: 0.6144
Recall: 0.5862
F1: 0.5791

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: 8
eval_batch_size: 8
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: cosine
lr_scheduler_warmup_ratio: 0.1
num_epochs: 50
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
5.543	0.012	150	5.5415	0.0052	0.0020	0.0052	0.0017
5.5236	0.024	300	5.5266	0.0054	0.0015	0.0054	0.0017
5.4413	0.036	450	5.5019	0.0064	0.0017	0.0064	0.0020
5.4126	0.048	600	5.4706	0.0069	0.0021	0.0069	0.0025
5.4049	0.06	750	5.4372	0.0085	0.0029	0.0085	0.0034
5.3589	0.072	900	5.4040	0.0087	0.0033	0.0087	0.0033
5.3584	0.084	1050	5.3520	0.0124	0.0046	0.0124	0.0052
5.3175	0.096	1200	5.3003	0.0144	0.0078	0.0144	0.0072
5.2656	0.108	1350	5.2471	0.0176	0.0100	0.0176	0.0092
5.2616	0.12	1500	5.1812	0.0257	0.0196	0.0257	0.0153
5.1724	0.132	1650	5.1143	0.0304	0.0286	0.0304	0.0192
5.1141	0.144	1800	5.0395	0.0395	0.0324	0.0395	0.0268
5.0431	0.156	1950	4.9608	0.0511	0.0448	0.0511	0.0354
4.9545	0.168	2100	4.8770	0.0647	0.0610	0.0647	0.0453
4.8579	0.18	2250	4.7872	0.0844	0.0794	0.0844	0.0620
4.8546	0.192	2400	4.6793	0.106	0.1085	0.106	0.0832
4.7785	0.204	2550	4.5924	0.1216	0.1237	0.1216	0.0949
4.7172	0.216	2700	4.4810	0.1424	0.1535	0.1424	0.1136
4.6015	0.228	2850	4.3700	0.1607	0.1727	0.1607	0.1287
4.4687	0.24	3000	4.2548	0.184	0.2157	0.184	0.1492
4.4424	0.252	3150	4.1712	0.2061	0.2314	0.2061	0.1701
4.3885	0.264	3300	4.0614	0.217	0.2609	0.217	0.1809
4.2685	0.276	3450	3.9597	0.2417	0.2793	0.2417	0.2038
4.23	0.288	3600	3.8267	0.2582	0.3022	0.2582	0.2209
4.1418	0.3	3750	3.7437	0.2719	0.3104	0.2719	0.2359
4.1243	0.312	3900	3.6720	0.2876	0.3513	0.2876	0.2525
3.9686	0.324	4050	3.5553	0.2943	0.3786	0.2943	0.2647
3.8717	0.336	4200	3.4799	0.3088	0.3827	0.3088	0.2769
3.8616	0.348	4350	3.3777	0.3295	0.3700	0.3295	0.2886
3.7505	0.36	4500	3.2887	0.3454	0.4185	0.3454	0.3092
3.7404	0.372	4650	3.2421	0.353	0.4278	0.353	0.3174
3.5638	0.384	4800	3.1349	0.3557	0.4332	0.3557	0.3199
3.6686	0.396	4950	3.0576	0.3658	0.4488	0.3658	0.3330
3.447	0.408	5100	3.0050	0.3821	0.4593	0.3821	0.3521
3.4422	0.42	5250	2.9250	0.3857	0.4522	0.3857	0.3518
3.4536	0.432	5400	2.8535	0.3969	0.4663	0.3969	0.3646
3.2877	0.444	5550	2.8128	0.411	0.4663	0.411	0.3802
3.2011	0.456	5700	2.7647	0.4265	0.4791	0.4265	0.3961
3.22	0.468	5850	2.6465	0.4333	0.5000	0.4333	0.4073
3.2558	0.48	6000	2.5661	0.4436	0.4919	0.4436	0.4161
2.9677	0.492	6150	2.5566	0.4511	0.5021	0.4511	0.4204
3.0638	0.504	6300	2.4858	0.4521	0.5161	0.4521	0.4230
3.1151	0.516	6450	2.4440	0.4558	0.5238	0.4558	0.4254
2.9704	0.528	6600	2.4065	0.4714	0.5249	0.4714	0.4480
2.8427	0.54	6750	2.3941	0.4683	0.5263	0.4683	0.4421
2.8918	0.552	6900	2.3228	0.4779	0.5206	0.4779	0.4520
2.7433	0.564	7050	2.3036	0.4848	0.5276	0.4848	0.4563
2.688	0.576	7200	2.2720	0.4883	0.5575	0.4883	0.4587
2.6728	0.588	7350	2.2135	0.4919	0.5487	0.4919	0.4630
2.7158	0.6	7500	2.2095	0.494	0.5379	0.494	0.4699
2.6621	0.612	7650	2.1368	0.5081	0.5597	0.5081	0.4851
2.62	0.624	7800	2.1086	0.519	0.5616	0.519	0.4987
2.6065	0.636	7950	2.1081	0.5134	0.5478	0.5134	0.4913
2.6039	0.648	8100	2.0900	0.5124	0.5489	0.5124	0.4938
2.5656	0.66	8250	2.0528	0.5253	0.5762	0.5253	0.5071
2.6685	0.672	8400	2.0159	0.526	0.5640	0.526	0.5079
2.4459	0.684	8550	1.9924	0.5322	0.5745	0.5322	0.5141
2.6334	0.696	8700	1.9847	0.5326	0.5739	0.5326	0.5156
2.352	0.708	8850	1.9473	0.5334	0.5859	0.5334	0.5149
2.5241	0.72	9000	1.9282	0.545	0.5959	0.545	0.5306
2.4239	0.732	9150	1.8883	0.5514	0.5842	0.5514	0.5335
2.4066	0.744	9300	1.8828	0.5503	0.5874	0.5503	0.5340
2.3955	0.756	9450	1.8484	0.5518	0.5850	0.5518	0.5393
2.3239	0.768	9600	1.8370	0.5574	0.5869	0.5574	0.5445
2.2789	0.78	9750	1.8495	0.5513	0.5884	0.5513	0.5368
2.361	0.792	9900	1.8085	0.5644	0.6008	0.5644	0.5511
2.2039	0.804	10050	1.7917	0.5696	0.6035	0.5696	0.5587
2.197	0.816	10200	1.7957	0.5623	0.5963	0.5623	0.5475
2.35	0.828	10350	1.7508	0.5686	0.5958	0.5686	0.5545
2.2012	0.84	10500	1.7497	0.5766	0.6105	0.5766	0.5662
2.1991	0.852	10650	1.7474	0.577	0.6104	0.577	0.5642
2.1711	0.864	10800	1.7279	0.567	0.5997	0.567	0.5534
2.2713	0.876	10950	1.7623	0.5669	0.6008	0.5669	0.5563
2.1433	0.888	11100	1.7210	0.5757	0.6002	0.5757	0.5613
2.0696	0.9	11250	1.7007	0.5726	0.5952	0.5726	0.5582
2.0547	0.912	11400	1.7004	0.5802	0.6078	0.5802	0.5697
1.9903	0.924	11550	1.6861	0.5799	0.6183	0.5799	0.5667
1.9698	0.936	11700	1.6907	0.5817	0.6174	0.5817	0.5704
2.1005	0.948	11850	1.6509	0.5877	0.6184	0.5877	0.5788
2.1018	0.96	12000	1.6890	0.5789	0.6095	0.5789	0.5683
2.0815	0.972	12150	1.6827	0.5862	0.6144	0.5862	0.5791

Framework versions

Transformers 4.57.3
Pytorch 2.9.1+cu128
Datasets 4.4.2
Tokenizers 0.22.1

thethinkmachine/teacher-resnet18-tiny-imagenet

teacher

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions