glpn-nyu-finetuned-diode-221229-103851

This model is a fine-tuned version of vinvino02/glpn-nyu on the diode-subset dataset. It achieves the following results on the evaluation set:

Loss: 0.4367
Mae: 0.4201
Rmse: 0.6202
Abs Rel: 0.4454
Log Mae: 0.1684
Log Rmse: 0.2265
Delta1: 0.3989
Delta2: 0.6516
Delta3: 0.8162

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: 24
eval_batch_size: 48
seed: 2022
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.15
num_epochs: 75
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Mae	Rmse	Abs Rel	Log Mae	Log Rmse	Delta1	Delta2	Delta3
1.0076	1.0	72	0.4910	0.4686	0.6460	0.5610	0.1948	0.2509	0.3172	0.5344	0.7863
0.4695	2.0	144	0.4518	0.4410	0.6304	0.4547	0.1805	0.2328	0.3427	0.6159	0.7933
0.4629	3.0	216	0.4766	0.4636	0.6340	0.5289	0.1921	0.2433	0.3189	0.5391	0.7521
0.4366	4.0	288	0.4581	0.4417	0.6300	0.4828	0.1801	0.2349	0.3591	0.6006	0.7918
0.4643	5.0	360	0.4687	0.4514	0.6290	0.5164	0.1856	0.2394	0.3340	0.5755	0.7834
0.4594	6.0	432	0.4515	0.4269	0.6214	0.4603	0.1727	0.2302	0.3894	0.6253	0.8033
0.4169	7.0	504	0.4669	0.4451	0.6271	0.4934	0.1830	0.2374	0.3538	0.5812	0.7772
0.4023	8.0	576	0.4734	0.4533	0.6353	0.5114	0.1869	0.2416	0.3403	0.5711	0.7772
0.4094	9.0	648	0.4752	0.4491	0.6329	0.5141	0.1843	0.2420	0.3621	0.5831	0.7587
0.3848	10.0	720	0.4515	0.4309	0.6246	0.4686	0.1750	0.2323	0.3837	0.6285	0.7958
0.365	11.0	792	0.4613	0.4541	0.6318	0.4734	0.1865	0.2348	0.3399	0.5601	0.7451
0.4244	12.0	864	0.4875	0.4616	0.6342	0.5606	0.1908	0.2475	0.3382	0.5512	0.7328
0.4397	13.0	936	0.4286	0.4183	0.6224	0.3908	0.1672	0.2211	0.4020	0.6556	0.8030
0.408	14.0	1008	0.4501	0.4377	0.6278	0.4376	0.1774	0.2289	0.3785	0.5970	0.7752
0.4233	15.0	1080	0.4330	0.4207	0.6252	0.3906	0.1680	0.2223	0.4184	0.6421	0.7951
0.3984	16.0	1152	0.4731	0.4550	0.6295	0.5166	0.1870	0.2394	0.3386	0.5589	0.7554
0.3706	17.0	1224	0.4291	0.4145	0.6170	0.3908	0.1652	0.2189	0.4184	0.6484	0.8062
0.424	18.0	1296	0.4910	0.4692	0.6388	0.5791	0.1944	0.2506	0.3195	0.5378	0.7355
0.3814	19.0	1368	0.4982	0.4770	0.6415	0.6067	0.1988	0.2552	0.3065	0.5249	0.7135
0.4224	20.0	1440	0.4222	0.4093	0.6149	0.3911	0.1621	0.2182	0.4257	0.6726	0.8145
0.3773	21.0	1512	0.4464	0.4213	0.6223	0.4333	0.1696	0.2283	0.4087	0.6412	0.7998
0.3529	22.0	1584	0.4541	0.4341	0.6237	0.4770	0.1756	0.2320	0.3749	0.6176	0.7872
0.3726	23.0	1656	0.4439	0.4229	0.6216	0.4461	0.1702	0.2285	0.3984	0.6456	0.8022
0.4254	24.0	1728	0.4898	0.4680	0.6389	0.5684	0.1936	0.2490	0.3196	0.5408	0.7341
0.3814	25.0	1800	0.4462	0.4292	0.6276	0.4380	0.1732	0.2294	0.3901	0.6324	0.7938
0.352	26.0	1872	0.4534	0.4316	0.6257	0.4686	0.1747	0.2336	0.3875	0.6286	0.7918
0.3372	27.0	1944	0.4472	0.4275	0.6198	0.4542	0.1733	0.2286	0.3788	0.6311	0.8053
0.372	28.0	2016	0.4750	0.4540	0.6298	0.5259	0.1867	0.2413	0.3413	0.5671	0.7596
0.3603	29.0	2088	0.4366	0.4209	0.6243	0.4156	0.1685	0.2248	0.3979	0.6576	0.8041
0.4185	30.0	2160	0.4559	0.4359	0.6256	0.4713	0.1769	0.2324	0.3805	0.6140	0.7716
0.3697	31.0	2232	0.4553	0.4375	0.6254	0.4788	0.1780	0.2334	0.3643	0.6158	0.7801
0.355	32.0	2304	0.4293	0.4128	0.6199	0.4038	0.1646	0.2222	0.4249	0.6666	0.8078
0.3927	33.0	2376	0.4597	0.4398	0.6317	0.4655	0.1799	0.2359	0.3758	0.6075	0.7719
0.3399	34.0	2448	0.4503	0.4333	0.6228	0.4604	0.1754	0.2300	0.3806	0.6181	0.7830
0.3398	35.0	2520	0.4467	0.4270	0.6227	0.4498	0.1726	0.2291	0.3898	0.6346	0.7976
0.3443	36.0	2592	0.4491	0.4309	0.6217	0.4545	0.1751	0.2293	0.3701	0.6267	0.7970
0.3423	37.0	2664	0.4592	0.4377	0.6212	0.4751	0.1792	0.2324	0.3582	0.6021	0.7870
0.384	38.0	2736	0.4587	0.4401	0.6294	0.4862	0.1791	0.2348	0.3573	0.6123	0.7955
0.3577	39.0	2808	0.4588	0.4441	0.6263	0.4830	0.1808	0.2335	0.3516	0.5952	0.7732
0.3839	40.0	2880	0.4611	0.4399	0.6249	0.5006	0.1792	0.2358	0.3614	0.6049	0.7866
0.3545	41.0	2952	0.4412	0.4242	0.6182	0.4461	0.1709	0.2264	0.3914	0.6353	0.8051
0.346	42.0	3024	0.4428	0.4273	0.6233	0.4429	0.1721	0.2275	0.3870	0.6369	0.8006
0.3516	43.0	3096	0.4380	0.4221	0.6177	0.4352	0.1692	0.2241	0.3934	0.6416	0.8063
0.3178	44.0	3168	0.4502	0.4312	0.6221	0.4730	0.1750	0.2311	0.3688	0.6241	0.8114
0.3327	45.0	3240	0.4446	0.4273	0.6233	0.4516	0.1729	0.2293	0.3796	0.6380	0.8100
0.3519	46.0	3312	0.4438	0.4229	0.6175	0.4492	0.1702	0.2261	0.3927	0.6393	0.8035
0.369	47.0	3384	0.4488	0.4270	0.6280	0.4590	0.1720	0.2320	0.3958	0.6465	0.7991
0.3331	48.0	3456	0.4545	0.4332	0.6221	0.4858	0.1755	0.2324	0.3696	0.6198	0.7990
0.3618	49.0	3528	0.4449	0.4250	0.6233	0.4549	0.1710	0.2295	0.3990	0.6408	0.8032
0.3315	50.0	3600	0.4489	0.4311	0.6236	0.4613	0.1739	0.2298	0.3842	0.6253	0.7941
0.3367	51.0	3672	0.4406	0.4211	0.6202	0.4442	0.1695	0.2278	0.4020	0.6512	0.8083
0.3239	52.0	3744	0.4256	0.4076	0.6159	0.4108	0.1620	0.2220	0.4333	0.6807	0.8152
0.3228	53.0	3816	0.4434	0.4205	0.6171	0.4564	0.1694	0.2283	0.4045	0.6400	0.8092
0.3176	54.0	3888	0.4378	0.4196	0.6165	0.4396	0.1687	0.2255	0.4028	0.6421	0.8102
0.3333	55.0	3960	0.4302	0.4156	0.6183	0.4203	0.1661	0.2226	0.4074	0.6593	0.8127
0.3234	56.0	4032	0.4450	0.4283	0.6219	0.4572	0.1729	0.2290	0.3892	0.6275	0.8010
0.3153	57.0	4104	0.4414	0.4236	0.6201	0.4511	0.1706	0.2277	0.3922	0.6391	0.8092
0.3326	58.0	4176	0.4325	0.4177	0.6177	0.4258	0.1672	0.2234	0.4113	0.6507	0.8085
0.3261	59.0	4248	0.4439	0.4256	0.6185	0.4576	0.1720	0.2284	0.3875	0.6331	0.8081
0.3117	60.0	4320	0.4346	0.4181	0.6159	0.4364	0.1678	0.2248	0.4067	0.6473	0.8107
0.3196	61.0	4392	0.4386	0.4204	0.6177	0.4461	0.1688	0.2261	0.4004	0.6430	0.8132
0.3016	62.0	4464	0.4376	0.4188	0.6152	0.4493	0.1686	0.2265	0.3981	0.6490	0.8156
0.3084	63.0	4536	0.4346	0.4170	0.6167	0.4360	0.1673	0.2249	0.4015	0.6585	0.8154
0.3086	64.0	4608	0.4401	0.4233	0.6186	0.4518	0.1703	0.2265	0.3858	0.6396	0.8172
0.3119	65.0	4680	0.4382	0.4207	0.6182	0.4450	0.1692	0.2262	0.3927	0.6488	0.8141
0.3025	66.0	4752	0.4436	0.4257	0.6206	0.4646	0.1717	0.2292	0.3834	0.6357	0.8131
0.3307	67.0	4824	0.4358	0.4188	0.6178	0.4423	0.1681	0.2258	0.4003	0.6525	0.8165
0.3173	68.0	4896	0.4368	0.4190	0.6175	0.4436	0.1680	0.2256	0.4016	0.6503	0.8156
0.3172	69.0	4968	0.4346	0.4172	0.6212	0.4380	0.1669	0.2257	0.4078	0.6604	0.8159
0.3128	70.0	5040	0.4376	0.4196	0.6182	0.4470	0.1685	0.2265	0.3961	0.6508	0.8147
0.3273	71.0	5112	0.4376	0.4217	0.6225	0.4446	0.1690	0.2265	0.3954	0.6486	0.8147
0.3234	72.0	5184	0.4379	0.4219	0.6191	0.4491	0.1696	0.2269	0.3917	0.6458	0.8150
0.3087	73.0	5256	0.4364	0.4198	0.6187	0.4460	0.1685	0.2264	0.3985	0.6498	0.8160
0.3167	74.0	5328	0.4371	0.4200	0.6190	0.4465	0.1686	0.2266	0.3981	0.6509	0.8155
0.3151	75.0	5400	0.4367	0.4201	0.6202	0.4454	0.1684	0.2265	0.3989	0.6516	0.8162

Framework versions

Transformers 4.24.0
Pytorch 1.12.1+cu116
Datasets 2.8.0
Tokenizers 0.13.2

glpn-nyu-finetuned-diode-221229-103851

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

NSDT 3DConvert

UnrealSynth

DreamTexture.js