glpn-nyu-finetuned-diode-230119-100058

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: 0.4305
Mae: 0.4203
Rmse: 0.6123
Abs Rel: 0.4280
Log Mae: 0.1694
Log Rmse: 0.2214
Delta1: 0.3813
Delta2: 0.6446
Delta3: 0.8152

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: 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
1.2807	1.0	72	0.9866	0.8312	1.0131	0.7179	0.5655	0.5924	0.0087	0.0200	0.0552
0.7396	2.0	144	0.4976	0.4741	0.6670	0.5279	0.1989	0.2567	0.3070	0.5470	0.7943
0.5018	3.0	216	0.4811	0.4630	0.6367	0.5198	0.1929	0.2446	0.3211	0.5440	0.7506
0.482	4.0	288	0.4726	0.4556	0.6337	0.4951	0.1893	0.2410	0.3306	0.5636	0.7663
0.4874	5.0	360	0.4813	0.4662	0.6355	0.5265	0.1941	0.2446	0.3179	0.5385	0.7278
0.4648	6.0	432	0.4681	0.4512	0.6309	0.4783	0.1869	0.2383	0.3430	0.5757	0.7527
0.4346	7.0	504	0.4637	0.4499	0.6292	0.4710	0.1859	0.2357	0.3453	0.5671	0.7644
0.4018	8.0	576	0.4790	0.4638	0.6349	0.5161	0.1928	0.2436	0.3255	0.5408	0.7338
0.4092	9.0	648	0.4559	0.4449	0.6267	0.4540	0.1827	0.2319	0.3541	0.5814	0.7692
0.3891	10.0	720	0.4619	0.4433	0.6259	0.4748	0.1823	0.2351	0.3579	0.5870	0.7742
0.3707	11.0	792	0.4624	0.4500	0.6269	0.4828	0.1851	0.2350	0.3421	0.5672	0.7638
0.4129	12.0	864	0.4648	0.4468	0.6265	0.4836	0.1836	0.2358	0.3533	0.5786	0.7625
0.4108	13.0	936	0.4474	0.4312	0.6187	0.4501	0.1752	0.2280	0.3801	0.6088	0.7887
0.3948	14.0	1008	0.4619	0.4498	0.6263	0.4853	0.1844	0.2344	0.3401	0.5721	0.7645
0.4009	15.0	1080	0.4619	0.4440	0.6244	0.4889	0.1820	0.2351	0.3563	0.5841	0.7751
0.3657	16.0	1152	0.4636	0.4491	0.6260	0.4936	0.1846	0.2360	0.3422	0.5734	0.7644
0.3605	17.0	1224	0.4353	0.4255	0.6153	0.4248	0.1715	0.2218	0.3844	0.6207	0.8008
0.3937	18.0	1296	0.4756	0.4609	0.6310	0.5281	0.1909	0.2423	0.3220	0.5461	0.7538
0.3453	19.0	1368	0.4698	0.4517	0.6270	0.5145	0.1863	0.2392	0.3360	0.5702	0.7689
0.3883	20.0	1440	0.4349	0.4240	0.6145	0.4311	0.1712	0.2230	0.3841	0.6321	0.8030
0.3482	21.0	1512	0.4339	0.4209	0.6146	0.4223	0.1694	0.2223	0.3967	0.6337	0.8036
0.3374	22.0	1584	0.4400	0.4289	0.6167	0.4431	0.1737	0.2254	0.3743	0.6191	0.7971
0.3516	23.0	1656	0.4395	0.4280	0.6171	0.4426	0.1737	0.2259	0.3710	0.6241	0.7998
0.3901	24.0	1728	0.4444	0.4324	0.6184	0.4562	0.1758	0.2280	0.3665	0.6118	0.7991
0.3587	25.0	1800	0.4326	0.4200	0.6129	0.4281	0.1690	0.2222	0.3920	0.6403	0.8073
0.3425	26.0	1872	0.4371	0.4231	0.6152	0.4341	0.1709	0.2242	0.3852	0.6372	0.7974
0.3252	27.0	1944	0.4381	0.4225	0.6140	0.4399	0.1705	0.2245	0.3851	0.6396	0.8065
0.3586	28.0	2016	0.4441	0.4304	0.6162	0.4488	0.1746	0.2258	0.3674	0.6179	0.7929
0.3389	29.0	2088	0.4240	0.4112	0.6100	0.4017	0.1640	0.2173	0.4152	0.6599	0.8128
0.3418	30.0	2160	0.4312	0.4195	0.6126	0.4211	0.1687	0.2206	0.3899	0.6435	0.8123
0.3454	31.0	2232	0.4301	0.4176	0.6126	0.4167	0.1674	0.2203	0.3974	0.6479	0.8089
0.3499	32.0	2304	0.4262	0.4154	0.6115	0.4081	0.1661	0.2184	0.3997	0.6578	0.8083
0.3649	33.0	2376	0.4429	0.4313	0.6171	0.4507	0.1753	0.2263	0.3641	0.6134	0.7982
0.3341	34.0	2448	0.4292	0.4207	0.6127	0.4161	0.1689	0.2192	0.3874	0.6415	0.8007
0.3323	35.0	2520	0.4402	0.4266	0.6148	0.4434	0.1728	0.2247	0.3754	0.6254	0.7983
0.3374	36.0	2592	0.4336	0.4233	0.6139	0.4277	0.1706	0.2219	0.3810	0.6362	0.8008
0.334	37.0	2664	0.4310	0.4230	0.6138	0.4240	0.1703	0.2209	0.3826	0.6345	0.8034
0.3471	38.0	2736	0.4372	0.4250	0.6144	0.4397	0.1720	0.2240	0.3780	0.6303	0.8046
0.3283	39.0	2808	0.4421	0.4301	0.6168	0.4497	0.1743	0.2259	0.3654	0.6209	0.7993
0.3418	40.0	2880	0.4340	0.4224	0.6137	0.4334	0.1703	0.2228	0.3857	0.6351	0.8054
0.3455	41.0	2952	0.4294	0.4174	0.6118	0.4212	0.1675	0.2203	0.3959	0.6469	0.8109
0.3229	42.0	3024	0.4291	0.4165	0.6121	0.4199	0.1671	0.2207	0.4035	0.6464	0.8103
0.352	43.0	3096	0.4393	0.4266	0.6154	0.4462	0.1729	0.2253	0.3744	0.6287	0.8049
0.3163	44.0	3168	0.4250	0.4113	0.6098	0.4112	0.1647	0.2187	0.4041	0.6620	0.8201
0.3284	45.0	3240	0.4358	0.4245	0.6138	0.4379	0.1716	0.2233	0.3745	0.6306	0.8106
0.3359	46.0	3312	0.4321	0.4217	0.6124	0.4283	0.1699	0.2210	0.3770	0.6412	0.8129
0.3406	47.0	3384	0.4238	0.4127	0.6104	0.4084	0.1653	0.2183	0.3982	0.6617	0.8177
0.3207	48.0	3456	0.4375	0.4275	0.6147	0.4435	0.1733	0.2243	0.3658	0.6262	0.8071
0.3338	49.0	3528	0.4331	0.4223	0.6142	0.4310	0.1705	0.2228	0.3846	0.6374	0.8071
0.3203	50.0	3600	0.4308	0.4212	0.6136	0.4253	0.1695	0.2213	0.3878	0.6407	0.8054
0.3238	51.0	3672	0.4379	0.4267	0.6148	0.4416	0.1727	0.2241	0.3723	0.6244	0.8036
0.3209	52.0	3744	0.4289	0.4187	0.6121	0.4178	0.1681	0.2198	0.3920	0.6461	0.8096
0.3198	53.0	3816	0.4376	0.4264	0.6145	0.4402	0.1724	0.2237	0.3708	0.6279	0.8066
0.3137	54.0	3888	0.4294	0.4180	0.6115	0.4242	0.1681	0.2208	0.3888	0.6494	0.8152
0.3238	55.0	3960	0.4416	0.4294	0.6158	0.4521	0.1743	0.2261	0.3645	0.6205	0.8069
0.3173	56.0	4032	0.4257	0.4142	0.6116	0.4145	0.1661	0.2198	0.4016	0.6586	0.8136
0.3173	57.0	4104	0.4303	0.4193	0.6123	0.4246	0.1687	0.2210	0.3879	0.6451	0.8118
0.3297	58.0	4176	0.4302	0.4219	0.6132	0.4259	0.1700	0.2211	0.3792	0.6394	0.8122
0.3261	59.0	4248	0.4319	0.4220	0.6131	0.4312	0.1702	0.2221	0.3781	0.6407	0.8142
0.3082	60.0	4320	0.4340	0.4234	0.6136	0.4346	0.1710	0.2228	0.3754	0.6373	0.8106
0.31	61.0	4392	0.4225	0.4120	0.6104	0.4073	0.1646	0.2181	0.4054	0.6626	0.8168
0.3065	62.0	4464	0.4313	0.4197	0.6125	0.4280	0.1690	0.2216	0.3854	0.6472	0.8127
0.3046	63.0	4536	0.4316	0.4202	0.6127	0.4268	0.1691	0.2213	0.3849	0.6448	0.8131
0.303	64.0	4608	0.4352	0.4241	0.6137	0.4373	0.1712	0.2231	0.3760	0.6364	0.8097
0.3094	65.0	4680	0.4318	0.4205	0.6128	0.4304	0.1695	0.2220	0.3828	0.6438	0.8140
0.3035	66.0	4752	0.4351	0.4233	0.6136	0.4386	0.1709	0.2235	0.3781	0.6388	0.8099
0.327	67.0	4824	0.4307	0.4203	0.6131	0.4280	0.1693	0.2216	0.3828	0.6463	0.8143
0.3175	68.0	4896	0.4325	0.4219	0.6137	0.4314	0.1701	0.2222	0.3809	0.6406	0.8135
0.3188	69.0	4968	0.4299	0.4203	0.6126	0.4271	0.1694	0.2214	0.3827	0.6440	0.8141
0.3158	70.0	5040	0.4304	0.4203	0.6126	0.4274	0.1694	0.2215	0.3832	0.6443	0.8133
0.3298	71.0	5112	0.4315	0.4219	0.6135	0.4292	0.1700	0.2218	0.3792	0.6423	0.8136
0.3246	72.0	5184	0.4323	0.4219	0.6129	0.4322	0.1703	0.2223	0.3769	0.6418	0.8133
0.3116	73.0	5256	0.4301	0.4198	0.6124	0.4264	0.1691	0.2213	0.3833	0.6459	0.8141
0.3192	74.0	5328	0.4301	0.4200	0.6125	0.4266	0.1691	0.2213	0.3819	0.6464	0.8156
0.3172	75.0	5400	0.4305	0.4203	0.6123	0.4280	0.1694	0.2214	0.3813	0.6446	0.8152

Framework versions

Transformers 4.24.0
Pytorch 1.12.1+cu116
Datasets 2.8.0
Tokenizers 0.13.2

glpn-nyu-finetuned-diode-230119-100058

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

NSDT 3DConvert

UnrealSynth

DreamTexture.js