glpn-nyu-finetuned-diode-230115-063851

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: 0.4360
Mae: 0.4211
Rmse: 0.6143
Abs Rel: 0.4394
Log Mae: 0.1700
Log Rmse: 0.2243
Delta1: 0.3835
Delta2: 0.6419
Delta3: 0.8181

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
1.0074	1.0	72	0.4929	0.4684	0.6424	0.5682	0.1955	0.2515	0.3151	0.5288	0.7834
0.4702	2.0	144	0.4561	0.4431	0.6292	0.4667	0.1822	0.2344	0.3377	0.6052	0.7878
0.4618	3.0	216	0.4827	0.4664	0.6351	0.5445	0.1939	0.2461	0.3163	0.5336	0.7408
0.4388	4.0	288	0.4669	0.4450	0.6251	0.5068	0.1831	0.2379	0.3465	0.5870	0.7811
0.463	5.0	360	0.4960	0.4715	0.6382	0.5942	0.1963	0.2528	0.3115	0.5361	0.7222
0.4478	6.0	432	0.4808	0.4542	0.6301	0.5439	0.1872	0.2440	0.3452	0.5647	0.7519
0.42	7.0	504	0.4645	0.4445	0.6268	0.4876	0.1822	0.2361	0.3578	0.5889	0.7741
0.3977	8.0	576	0.4767	0.4503	0.6299	0.5162	0.1857	0.2410	0.3452	0.5705	0.7726
0.4045	9.0	648	0.4747	0.4568	0.6303	0.5208	0.1885	0.2406	0.3323	0.5572	0.7497
0.392	10.0	720	0.4860	0.4571	0.6331	0.5645	0.1889	0.2475	0.3370	0.5627	0.7589
0.3749	11.0	792	0.4785	0.4502	0.6308	0.5449	0.1860	0.2446	0.3423	0.5719	0.7940
0.4292	12.0	864	0.4905	0.4574	0.6346	0.5616	0.1891	0.2483	0.3402	0.5694	0.7499
0.432	13.0	936	0.4648	0.4408	0.6229	0.4877	0.1804	0.2345	0.3607	0.5947	0.7771
0.4097	14.0	1008	0.4464	0.4303	0.6221	0.4398	0.1742	0.2285	0.3879	0.6179	0.7821
0.4212	15.0	1080	0.4773	0.4550	0.6298	0.5327	0.1874	0.2425	0.3375	0.5666	0.7588
0.3862	16.0	1152	0.4682	0.4440	0.6248	0.5171	0.1824	0.2392	0.3516	0.5906	0.7793
0.3726	17.0	1224	0.4702	0.4425	0.6243	0.5190	0.1807	0.2385	0.3591	0.5904	0.7824
0.4016	18.0	1296	0.5012	0.4789	0.6418	0.6093	0.2002	0.2561	0.3035	0.5188	0.7003
0.3772	19.0	1368	0.4935	0.4676	0.6371	0.5940	0.1947	0.2525	0.3195	0.5398	0.7340
0.3987	20.0	1440	0.4630	0.4399	0.6312	0.4934	0.1801	0.2388	0.3711	0.6044	0.7865
0.378	21.0	1512	0.4424	0.4180	0.6211	0.4329	0.1683	0.2280	0.4210	0.6415	0.8022
0.3674	22.0	1584	0.4591	0.4346	0.6272	0.5022	0.1764	0.2374	0.3819	0.6184	0.7881
0.3803	23.0	1656	0.4708	0.4483	0.6276	0.5228	0.1841	0.2404	0.3484	0.5854	0.7597
0.4082	24.0	1728	0.4753	0.4506	0.6286	0.5436	0.1854	0.2436	0.3512	0.5780	0.7593
0.3662	25.0	1800	0.4455	0.4221	0.6160	0.4622	0.1709	0.2288	0.3897	0.6406	0.8124
0.3735	26.0	1872	0.4405	0.4194	0.6219	0.4487	0.1691	0.2304	0.4091	0.6492	0.8080
0.3387	27.0	1944	0.4449	0.4235	0.6176	0.4538	0.1716	0.2282	0.3807	0.6471	0.8122
0.3826	28.0	2016	0.4521	0.4261	0.6176	0.4622	0.1716	0.2289	0.3887	0.6348	0.7957
0.358	29.0	2088	0.4299	0.4113	0.6123	0.4165	0.1643	0.2209	0.4073	0.6734	0.8179
0.3466	30.0	2160	0.4357	0.4154	0.6172	0.4177	0.1666	0.2237	0.4067	0.6619	0.8109
0.3698	31.0	2232	0.4735	0.4469	0.6256	0.5423	0.1842	0.2425	0.3421	0.5840	0.7896
0.3578	32.0	2304	0.4405	0.4156	0.6126	0.4429	0.1674	0.2253	0.4016	0.6521	0.8146
0.3908	33.0	2376	0.4829	0.4584	0.6315	0.5698	0.1895	0.2472	0.3317	0.5601	0.7479
0.3398	34.0	2448	0.4451	0.4253	0.6187	0.4517	0.1720	0.2283	0.3869	0.6347	0.8013
0.3368	35.0	2520	0.4491	0.4259	0.6186	0.4619	0.1725	0.2299	0.3774	0.6392	0.8056
0.3786	36.0	2592	0.4419	0.4254	0.6150	0.4497	0.1726	0.2262	0.3677	0.6346	0.8181
0.3373	37.0	2664	0.4562	0.4365	0.6224	0.4909	0.1780	0.2346	0.3690	0.6071	0.7911
0.3628	38.0	2736	0.4643	0.4433	0.6244	0.5107	0.1822	0.2378	0.3437	0.5898	0.7946
0.3746	39.0	2808	0.4746	0.4525	0.6278	0.5310	0.1865	0.2417	0.3388	0.5716	0.7541
0.3994	40.0	2880	0.4740	0.4498	0.6280	0.5399	0.1857	0.2431	0.3415	0.5791	0.7742
0.3583	41.0	2952	0.4500	0.4260	0.6197	0.4717	0.1731	0.2318	0.3885	0.6316	0.8052
0.369	42.0	3024	0.4369	0.4176	0.6181	0.4334	0.1681	0.2261	0.4051	0.6604	0.8066
0.35	43.0	3096	0.4514	0.4297	0.6182	0.4802	0.1753	0.2321	0.3702	0.6155	0.8117
0.3249	44.0	3168	0.4382	0.4209	0.6180	0.4332	0.1698	0.2256	0.3981	0.6443	0.8054
0.3329	45.0	3240	0.4558	0.4380	0.6222	0.4840	0.1789	0.2335	0.3578	0.5989	0.7958
0.3553	46.0	3312	0.4420	0.4173	0.6150	0.4520	0.1679	0.2274	0.4029	0.6572	0.8098
0.3671	47.0	3384	0.4479	0.4294	0.6174	0.4734	0.1750	0.2304	0.3595	0.6255	0.8145
0.3244	48.0	3456	0.4542	0.4369	0.6189	0.4872	0.1786	0.2325	0.3520	0.6026	0.8070
0.3803	49.0	3528	0.4447	0.4256	0.6174	0.4635	0.1721	0.2291	0.3850	0.6347	0.8041
0.332	50.0	3600	0.4434	0.4279	0.6167	0.4573	0.1735	0.2276	0.3689	0.6301	0.8082
0.3249	51.0	3672	0.4379	0.4242	0.6170	0.4448	0.1716	0.2260	0.3783	0.6400	0.8117
0.3257	52.0	3744	0.4277	0.4151	0.6169	0.4110	0.1664	0.2215	0.4075	0.6594	0.8110
0.3256	53.0	3816	0.4493	0.4317	0.6189	0.4776	0.1755	0.2309	0.3654	0.6152	0.8077
0.3164	54.0	3888	0.4503	0.4303	0.6173	0.4821	0.1750	0.2313	0.3687	0.6181	0.8128
0.3276	55.0	3960	0.4503	0.4322	0.6187	0.4765	0.1763	0.2311	0.3641	0.6098	0.8127
0.3207	56.0	4032	0.4524	0.4320	0.6199	0.4807	0.1759	0.2324	0.3622	0.6234	0.8072
0.3204	57.0	4104	0.4425	0.4238	0.6149	0.4532	0.1715	0.2266	0.3800	0.6413	0.8086
0.3282	58.0	4176	0.4440	0.4267	0.6162	0.4592	0.1731	0.2278	0.3777	0.6260	0.8088
0.3232	59.0	4248	0.4439	0.4298	0.6165	0.4603	0.1748	0.2278	0.3621	0.6190	0.8141
0.307	60.0	4320	0.4452	0.4275	0.6165	0.4623	0.1737	0.2286	0.3741	0.6235	0.8105
0.3142	61.0	4392	0.4432	0.4270	0.6159	0.4578	0.1732	0.2275	0.3763	0.6236	0.8133
0.3062	62.0	4464	0.4422	0.4238	0.6150	0.4582	0.1717	0.2275	0.3829	0.6331	0.8189
0.3037	63.0	4536	0.4306	0.4142	0.6132	0.4240	0.1663	0.2223	0.3992	0.6677	0.8193
0.309	64.0	4608	0.4450	0.4277	0.6162	0.4625	0.1736	0.2282	0.3710	0.6242	0.8169
0.3096	65.0	4680	0.4442	0.4277	0.6169	0.4601	0.1736	0.2283	0.3714	0.6262	0.8144
0.3049	66.0	4752	0.4449	0.4278	0.6166	0.4622	0.1737	0.2285	0.3725	0.6273	0.8122
0.3324	67.0	4824	0.4416	0.4264	0.6159	0.4511	0.1728	0.2265	0.3731	0.6284	0.8145
0.3183	68.0	4896	0.4405	0.4243	0.6149	0.4501	0.1718	0.2261	0.3786	0.6315	0.8173
0.3178	69.0	4968	0.4397	0.4240	0.6150	0.4488	0.1716	0.2259	0.3779	0.6332	0.8168
0.3159	70.0	5040	0.4365	0.4212	0.6138	0.4396	0.1701	0.2242	0.3836	0.6404	0.8173
0.3266	71.0	5112	0.4397	0.4244	0.6145	0.4480	0.1718	0.2256	0.3773	0.6312	0.8161
0.3234	72.0	5184	0.4384	0.4237	0.6144	0.4451	0.1714	0.2251	0.3761	0.6346	0.8177
0.3108	73.0	5256	0.4371	0.4219	0.6144	0.4429	0.1705	0.2250	0.3820	0.6395	0.8174
0.3184	74.0	5328	0.4351	0.4206	0.6138	0.4381	0.1697	0.2240	0.3850	0.6430	0.8182
0.3152	75.0	5400	0.4360	0.4211	0.6143	0.4394	0.1700	0.2243	0.3835	0.6419	0.8181

Framework versions

Transformers 4.24.0
Pytorch 1.12.1+cu116
Datasets 2.8.0
Tokenizers 0.13.2

glpn-nyu-finetuned-diode-230115-063851

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

NSDT 3DConvert

UnrealSynth

DreamTexture.js