glpn-nyu-finetuned-diode-230118-181403

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: 0.4408
Mae: 0.4277
Rmse: 0.6172
Abs Rel: 0.4520
Log Mae: 0.1735
Log Rmse: 0.2270
Delta1: 0.3788
Delta2: 0.6217
Delta3: 0.8047

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: 100
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Mae	Rmse	Abs Rel	Log Mae	Log Rmse	Delta1	Delta2	Delta3
1.0764	1.0	72	0.5030	0.4778	0.6687	0.5507	0.2006	0.2591	0.3021	0.5334	0.8002
0.4775	2.0	144	0.4662	0.4525	0.6313	0.4929	0.1869	0.2383	0.3276	0.5766	0.7713
0.4668	3.0	216	0.4841	0.4709	0.6371	0.5461	0.1961	0.2470	0.3112	0.5262	0.7238
0.4391	4.0	288	0.4571	0.4354	0.6215	0.4835	0.1779	0.2335	0.3605	0.6136	0.7948
0.4621	5.0	360	0.4870	0.4658	0.6352	0.5593	0.1935	0.2480	0.3153	0.5462	0.7390
0.4577	6.0	432	0.4661	0.4411	0.6238	0.5039	0.1805	0.2371	0.3654	0.5885	0.7847
0.4206	7.0	504	0.4889	0.4660	0.6351	0.5636	0.1938	0.2484	0.3212	0.5392	0.7310
0.3938	8.0	576	0.4856	0.4638	0.6369	0.5443	0.1923	0.2474	0.3293	0.5515	0.7313
0.4073	9.0	648	0.4733	0.4538	0.6312	0.5211	0.1867	0.2412	0.3469	0.5653	0.7483
0.399	10.0	720	0.4862	0.4620	0.6345	0.5579	0.1918	0.2476	0.3277	0.5530	0.7396
0.3621	11.0	792	0.4523	0.4307	0.6178	0.4724	0.1749	0.2298	0.3720	0.6131	0.8001
0.4208	12.0	864	0.4790	0.4524	0.6294	0.5339	0.1860	0.2420	0.3503	0.5648	0.7454
0.4365	13.0	936	0.4450	0.4284	0.6175	0.4516	0.1741	0.2269	0.3673	0.6182	0.8091
0.4192	14.0	1008	0.4581	0.4357	0.6214	0.4790	0.1766	0.2316	0.3697	0.6067	0.7763
0.4172	15.0	1080	0.4481	0.4286	0.6276	0.4498	0.1732	0.2309	0.3898	0.6298	0.7964
0.3825	16.0	1152	0.4873	0.4649	0.6342	0.5704	0.1929	0.2482	0.3184	0.5443	0.7413
0.4006	17.0	1224	0.4752	0.4545	0.6298	0.5398	0.1871	0.2423	0.3383	0.5683	0.7496
0.4171	18.0	1296	0.4979	0.4742	0.6410	0.5991	0.1974	0.2544	0.3148	0.5296	0.7145
0.366	19.0	1368	0.4919	0.4659	0.6381	0.5903	0.1936	0.2519	0.3212	0.5474	0.7497
0.4029	20.0	1440	0.4403	0.4257	0.6208	0.4245	0.1712	0.2248	0.3915	0.6344	0.8015
0.3569	21.0	1512	0.4423	0.4251	0.6257	0.4244	0.1709	0.2281	0.4043	0.6405	0.7928
0.3614	22.0	1584	0.4645	0.4402	0.6251	0.5081	0.1800	0.2378	0.3631	0.6089	0.7816
0.3875	23.0	1656	0.4603	0.4384	0.6268	0.4957	0.1789	0.2369	0.3665	0.6138	0.7876
0.4125	24.0	1728	0.4857	0.4628	0.6341	0.5720	0.1918	0.2484	0.3239	0.5471	0.7591
0.3937	25.0	1800	0.4785	0.4555	0.6300	0.5504	0.1882	0.2446	0.3310	0.5617	0.7685
0.3677	26.0	1872	0.4370	0.4225	0.6156	0.4300	0.1702	0.2237	0.3872	0.6348	0.8056
0.3467	27.0	1944	0.4550	0.4321	0.6193	0.4884	0.1758	0.2330	0.3697	0.6155	0.8055
0.3769	28.0	2016	0.5003	0.4765	0.6410	0.6055	0.1988	0.2552	0.3100	0.5227	0.7041
0.3622	29.0	2088	0.4574	0.4442	0.6242	0.4790	0.1818	0.2329	0.3507	0.5814	0.7708
0.3789	30.0	2160	0.4314	0.4174	0.6225	0.4145	0.1671	0.2248	0.4066	0.6706	0.8073
0.3698	31.0	2232	0.4480	0.4270	0.6210	0.4586	0.1731	0.2302	0.3842	0.6308	0.8003
0.3678	32.0	2304	0.4224	0.4052	0.6146	0.3880	0.1607	0.2187	0.4340	0.6882	0.8134
0.4244	33.0	2376	0.4622	0.4456	0.6308	0.4903	0.1822	0.2369	0.3543	0.5903	0.7691
0.3715	34.0	2448	0.4541	0.4358	0.6208	0.4726	0.1775	0.2313	0.3689	0.6006	0.7869
0.3372	35.0	2520	0.4557	0.4337	0.6212	0.4803	0.1769	0.2330	0.3701	0.6134	0.7918
0.3465	36.0	2592	0.4446	0.4278	0.6165	0.4522	0.1734	0.2268	0.3725	0.6200	0.8093
0.3493	37.0	2664	0.4778	0.4553	0.6311	0.5503	0.1878	0.2448	0.3374	0.5685	0.7603
0.3797	38.0	2736	0.4586	0.4375	0.6221	0.4959	0.1788	0.2352	0.3609	0.6061	0.7887
0.3671	39.0	2808	0.4629	0.4455	0.6244	0.4989	0.1819	0.2355	0.3509	0.5852	0.7624
0.3907	40.0	2880	0.4574	0.4390	0.6223	0.4844	0.1792	0.2333	0.3563	0.6029	0.7823
0.3672	41.0	2952	0.4461	0.4300	0.6190	0.4500	0.1739	0.2271	0.3787	0.6217	0.7864
0.3506	42.0	3024	0.4432	0.4282	0.6219	0.4470	0.1732	0.2282	0.3896	0.6232	0.7964
0.3589	43.0	3096	0.4293	0.4174	0.6225	0.4029	0.1673	0.2233	0.4131	0.6637	0.7973
0.3333	44.0	3168	0.4618	0.4430	0.6242	0.5015	0.1814	0.2362	0.3535	0.5957	0.7784
0.3432	45.0	3240	0.4611	0.4383	0.6232	0.5032	0.1793	0.2365	0.3593	0.6061	0.7929
0.3603	46.0	3312	0.4558	0.4345	0.6265	0.4737	0.1769	0.2342	0.3738	0.6222	0.7887
0.3703	47.0	3384	0.4571	0.4369	0.6241	0.4841	0.1782	0.2340	0.3642	0.6103	0.7883
0.3403	48.0	3456	0.4360	0.4204	0.6148	0.4355	0.1691	0.2235	0.3943	0.6403	0.8130
0.3574	49.0	3528	0.4379	0.4216	0.6171	0.4375	0.1698	0.2247	0.3923	0.6457	0.8073
0.3337	50.0	3600	0.4447	0.4263	0.6218	0.4465	0.1730	0.2292	0.3830	0.6431	0.8007
0.3437	51.0	3672	0.4474	0.4320	0.6191	0.4661	0.1757	0.2292	0.3645	0.6189	0.7994
0.3283	52.0	3744	0.4377	0.4226	0.6227	0.4241	0.1702	0.2247	0.3868	0.6511	0.8079
0.3344	53.0	3816	0.4459	0.4243	0.6195	0.4629	0.1716	0.2294	0.3804	0.6403	0.8161
0.3246	54.0	3888	0.4408	0.4234	0.6173	0.4517	0.1712	0.2276	0.3845	0.6380	0.8053
0.3517	55.0	3960	0.4536	0.4350	0.6223	0.4891	0.1779	0.2341	0.3610	0.6075	0.8100
0.3305	56.0	4032	0.4571	0.4427	0.6225	0.5007	0.1815	0.2351	0.3445	0.5906	0.7934
0.3215	57.0	4104	0.4447	0.4288	0.6171	0.4612	0.1738	0.2280	0.3781	0.6218	0.8010
0.3318	58.0	4176	0.4416	0.4272	0.6171	0.4486	0.1729	0.2261	0.3754	0.6304	0.7968
0.328	59.0	4248	0.4637	0.4440	0.6241	0.5086	0.1820	0.2371	0.3502	0.5877	0.7787
0.3152	60.0	4320	0.4457	0.4319	0.6193	0.4577	0.1752	0.2282	0.3712	0.6113	0.7961
0.3389	61.0	4392	0.4368	0.4197	0.6151	0.4404	0.1694	0.2252	0.3893	0.6497	0.8116
0.3137	62.0	4464	0.4397	0.4158	0.6174	0.4197	0.1687	0.2299	0.4016	0.6646	0.8184
0.3219	63.0	4536	0.4506	0.4327	0.6207	0.4780	0.1765	0.2323	0.3674	0.6166	0.7982
0.3182	64.0	4608	0.4433	0.4313	0.6177	0.4529	0.1747	0.2264	0.3711	0.6117	0.7967
0.3255	65.0	4680	0.4458	0.4286	0.6218	0.4656	0.1742	0.2311	0.3834	0.6239	0.8014
0.3064	66.0	4752	0.4533	0.4333	0.6203	0.4871	0.1761	0.2330	0.3723	0.6153	0.7911
0.3418	67.0	4824	0.4529	0.4343	0.6218	0.4804	0.1764	0.2318	0.3706	0.6124	0.7936
0.3215	68.0	4896	0.4434	0.4318	0.6186	0.4576	0.1751	0.2272	0.3675	0.6096	0.8017
0.3268	69.0	4968	0.4429	0.4283	0.6165	0.4540	0.1735	0.2266	0.3754	0.6177	0.7967
0.3181	70.0	5040	0.4399	0.4242	0.6158	0.4522	0.1713	0.2265	0.3821	0.6369	0.8066
0.3312	71.0	5112	0.4467	0.4310	0.6192	0.4642	0.1752	0.2294	0.3682	0.6164	0.7995
0.329	72.0	5184	0.4469	0.4304	0.6187	0.4671	0.1749	0.2294	0.3668	0.6181	0.8051
0.3114	73.0	5256	0.4507	0.4331	0.6196	0.4775	0.1764	0.2315	0.3641	0.6143	0.8026
0.3405	74.0	5328	0.4486	0.4323	0.6193	0.4667	0.1759	0.2299	0.3666	0.6149	0.7967
0.3192	75.0	5400	0.4491	0.4279	0.6187	0.4737	0.1738	0.2313	0.3782	0.6310	0.7995
0.3071	76.0	5472	0.4487	0.4287	0.6200	0.4695	0.1741	0.2309	0.3812	0.6239	0.8028
0.3173	77.0	5544	0.4459	0.4297	0.6199	0.4603	0.1745	0.2295	0.3786	0.6207	0.8008
0.318	78.0	5616	0.4395	0.4258	0.6180	0.4450	0.1723	0.2264	0.3822	0.6321	0.8025
0.3248	79.0	5688	0.4429	0.4262	0.6185	0.4598	0.1729	0.2292	0.3858	0.6277	0.8076
0.3182	80.0	5760	0.4506	0.4344	0.6204	0.4778	0.1772	0.2321	0.3653	0.6084	0.7958
0.3235	81.0	5832	0.4427	0.4243	0.6176	0.4556	0.1714	0.2282	0.3918	0.6346	0.8029
0.3353	82.0	5904	0.4494	0.4342	0.6205	0.4715	0.1770	0.2311	0.3656	0.6085	0.7980
0.3006	83.0	5976	0.4459	0.4334	0.6200	0.4620	0.1763	0.2293	0.3642	0.6111	0.8011
0.3154	84.0	6048	0.4477	0.4335	0.6190	0.4691	0.1767	0.2300	0.3662	0.6064	0.8013
0.3263	85.0	6120	0.4433	0.4303	0.6184	0.4583	0.1747	0.2281	0.3697	0.6172	0.8068
0.3028	86.0	6192	0.4432	0.4291	0.6184	0.4603	0.1744	0.2289	0.3746	0.6181	0.8089
0.3066	87.0	6264	0.4362	0.4245	0.6170	0.4404	0.1718	0.2252	0.3834	0.6326	0.8104
0.3084	88.0	6336	0.4359	0.4218	0.6151	0.4395	0.1705	0.2248	0.3882	0.6375	0.8113
0.2976	89.0	6408	0.4391	0.4268	0.6158	0.4488	0.1729	0.2259	0.3764	0.6246	0.8062
0.3239	90.0	6480	0.4413	0.4278	0.6178	0.4533	0.1736	0.2274	0.3777	0.6207	0.8090
0.321	91.0	6552	0.4390	0.4259	0.6172	0.4469	0.1726	0.2264	0.3801	0.6269	0.8106
0.3094	92.0	6624	0.4427	0.4301	0.6190	0.4552	0.1745	0.2277	0.3757	0.6132	0.8034
0.3215	93.0	6696	0.4389	0.4256	0.6166	0.4466	0.1724	0.2261	0.3846	0.6256	0.8060
0.3074	94.0	6768	0.4435	0.4298	0.6180	0.4563	0.1746	0.2279	0.3764	0.6153	0.8011
0.3049	95.0	6840	0.4443	0.4311	0.6185	0.4598	0.1752	0.2285	0.3728	0.6133	0.8015
0.3198	96.0	6912	0.4385	0.4261	0.6163	0.4468	0.1726	0.2260	0.3802	0.6271	0.8085
0.2957	97.0	6984	0.4426	0.4300	0.6186	0.4566	0.1746	0.2279	0.3745	0.6167	0.8029
0.3035	98.0	7056	0.4427	0.4287	0.6174	0.4556	0.1741	0.2276	0.3765	0.6195	0.8042
0.329	99.0	7128	0.4386	0.4263	0.6174	0.4458	0.1725	0.2260	0.3824	0.6278	0.8058
0.3087	100.0	7200	0.4408	0.4277	0.6172	0.4520	0.1735	0.2270	0.3788	0.6217	0.8047

Framework versions

Transformers 4.24.0
Pytorch 1.12.1+cu116
Datasets 2.8.0
Tokenizers 0.13.2

glpn-nyu-finetuned-diode-230118-181403

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

NSDT 3DConvert

UnrealSynth

DreamTexture.js