OpenBench

OpenBench Testing Framework

Author	Menkib
Creation	1778076873
Source	https://github.com/Menkib64/lc0
Sha	028effcb490d83fc8ee598d95b4fdcd999c135c5
Branch	uncertainty_weighting_logistic
Bench	1
Network	tf13-375
Options	Threads=2 Hash=1
Time	20.0+0.20
Opening Book	UHO_Lichess_4852_v1.epd
Upload PGNs	FALSE
Syzygy WDL	DISABLED
Syzygy ADJ.	OPTIONAL
Win ADJ.	movecount=3 score=400
Draw ADJ.	movenumber=40 movecount=8 score=10
Scaling	DEV 6250 NPS
Alpha	0.63
Gamma	0.13
A-Ratio	0.1
Reporting	BATCHED
Distribution	SINGLE
Pairs Per Point	16

uncertainty_weighting_lo... vs uncertainty_weighting_lo...

Approve Restart Delete

UncertaintyWeightingMidPoint, float, 0.04, 0.0, 1.0, 0.002, 0.002
UncertaintyWeightingMinusExponent, float, 75.0, 0.0, 1000.0, 0.5, 0.5
UncertaintyWeightingScale, float, 0.395, 0.0, 10.0, 0.01, 0.01
UncertaintyWeightingBase, float, 0.68, 0.0, 10.0, 0.01, 0.01

Copy SPSA Inputs

UncertaintyWeightingMidPoint, 0.04015149079777302
UncertaintyWeightingMinusExponent, 105.04547795631156
UncertaintyWeightingScale, 0.4050108431669857
UncertaintyWeightingBase, 0.6773005440944668

Copy SPSA Outputs

Tuning 4 Parameters
673/25000 Iterations
21560/800000 Games Played

Name	Curr	Start	Max	C	C_end	R	R_end
UncertaintyWeightingMidPoint	0.0402	0.0400	1.0000	0.0032	0.0020	0.0030	0.0020
UncertaintyWeightingMinusExponent	105.0455	75.0000	1000.0000	0.7997	0.5000	0.7617	0.5000
UncertaintyWeightingScale	0.4050	0.3950	10.0000	0.0160	0.0100	0.0152	0.0100
UncertaintyWeightingBase	0.6773	0.6800	10.0000	0.0160	0.0100	0.0152	0.0100

#	User	Updated	Games	Won	Lost	Drawn	Crash
1271	Kovax	1778132184	518	118	130	270	0
236	Ploot	1778132162	426	116	99	211	0
1281	Kovax	1778163548	480	108	116	256	0
1256	Kovax	1778163202	980	232	244	504	1
1275	Kovax	1778163216	970	239	228	503	0
1278	Kovax	1778163166	570	151	140	279	0
1286	Kovax	1778132223	520	129	143	248	0
1264	Kovax	1778163122	940	232	220	488	0
1273	Kovax	1778132214	524	118	123	283	0
1289	Kovax	1778130779	480	133	137	210	0
1276	Kovax	1778163187	990	260	252	478	0
1266	Kovax	1778163149	934	239	234	461	0
1258	Kovax	1778163191	992	246	255	491	0
1267	Kovax	1778163202	974	229	236	509	0
1261	Kovax	1778163165	578	139	149	290	0
1290	Kovax	1778132198	514	125	116	273	0
1282	Kovax	1778163535	512	114	129	269	0
239	Ploot	1778163177	800	198	201	401	0
1269	Kovax	1778163212	538	152	147	239	0
1255	Kovax	1778132218	512	112	130	270	0
1272	Kovax	1778163245	780	193	188	399	0
443	Kovax	1778132188	288	64	77	147	0
1257	Kovax	1778132168	336	83	82	171	0
444	Kovax	1778163134	652	161	151	340	0
237	Ploot	1778163218	662	151	155	356	0
414	Kovax	1778163132	368	95	98	175	1
441	Kovax	1778163164	370	88	87	195	0
1262	Kovax	1778163184	462	116	115	231	0
1301	Kovax	1778163212	460	124	104	232	1
1274	Kovax	1778163202	440	115	116	209	0
1263	Kovax	1778163228	454	118	120	216	0
1253	Kovax	1778163205	458	117	123	218	0
1302	Kovax	1778163205	438	108	90	240	0
1303	Kovax	1778163239	462	117	112	233	0
238	Ploot	1778163237	376	96	91	189	0
1299	Kovax	1778163182	56	12	15	29	0
413	Kovax	1778163187	46	10	12	24	0
442	Kovax	1778163202	46	14	16	16	0
1283	Kovax	1778163202	54	13	16	25	0
1268	Kovax	1778163236	58	15	10	33	0
1280	Kovax	1778163248	56	11	11	34	0
415	Kovax	1778163201	46	9	12	25	0
1277	Kovax	1778163236	58	15	14	29	0
1304	Kovax	1778163205	52	7	9	36	0
1300	Kovax	1778163208	58	15	15	28	0
1298	Kovax	1778163226	56	12	12	32	0
1265	Kovax	1778163232	56	12	12	32	0
1252	Kovax	1778163127	56	20	14	22	0
416	Kovax	1778163202	48	11	10	27	0
1279	Kovax	1778163238	56	14	16	26	0
1307	Kovax	1778163534	0	0	0	0	0
1308	Kovax	1778163540	0	0	0	0	0
1309	Kovax	1778163545	0	0	0	0	0
1310	Kovax	1778163551	0	0	0	0	0
1311	Kovax	1778163558	0	0	0	0	0
1312	Kovax	1778163564	0	0	0	0	0