This model is a fine-tuned version of Qwen/Qwen2.5-1.5B-Instruct on the YuchenLi01/MATH_Qwen2.5-1.5BInstruct_DPO_generatedAndParaphrasedMoreUniqueResponseNoGT dataset. It achieves the following results on the evaluation set:

Loss: 0.5012
Rewards/chosen: -2.1001
Rewards/rejected: -3.4825
Rewards/accuracies: 0.7317
Rewards/margins: 1.3824
Logps/rejected: -82.5847
Logps/chosen: -63.4417
Logits/rejected: -2.4968
Logits/chosen: -2.6520

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-07
train_batch_size: 4
eval_batch_size: 4
seed: 42
distributed_type: multi-GPU
num_devices: 8
total_train_batch_size: 32
total_eval_batch_size: 32
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: cosine
lr_scheduler_warmup_ratio: 0.1
num_epochs: 2.0

Training results

Training Loss	Epoch	Step	Validation Loss	Rewards/chosen	Rewards/rejected	Rewards/accuracies	Rewards/margins	Logps/rejected	Logps/chosen	Logits/rejected	Logits/chosen
0.6949	0.0135	20	0.6936	0.0006	-0.0014	0.5335	0.0020	-47.7731	-42.4340	-2.2068	-2.3128
0.6997	0.0270	40	0.6935	-0.0001	-0.0020	0.5061	0.0020	-47.7796	-42.4410	-2.2042	-2.3097
0.6979	0.0405	60	0.6935	-0.0034	-0.0029	0.4909	-0.0005	-47.7887	-42.4748	-2.2018	-2.3075
0.6946	0.0540	80	0.6933	-0.0039	-0.0052	0.4939	0.0013	-47.8117	-42.4797	-2.1990	-2.3048
0.6873	0.0675	100	0.6924	-0.0135	-0.0153	0.5030	0.0019	-47.9125	-42.5749	-2.1940	-2.2995
0.6895	0.0810	120	0.6909	-0.0253	-0.0329	0.5671	0.0076	-48.0886	-42.6938	-2.1791	-2.2849
0.692	0.0945	140	0.6889	-0.0430	-0.0540	0.5976	0.0111	-48.2996	-42.8701	-2.1662	-2.2726
0.6829	0.1080	160	0.6877	-0.0727	-0.0833	0.5366	0.0107	-48.5927	-43.1672	-2.1438	-2.2499
0.694	0.1215	180	0.6855	-0.1164	-0.1293	0.5671	0.0129	-49.0520	-43.6040	-2.1104	-2.2168
0.6671	0.1350	200	0.6822	-0.1602	-0.1809	0.5762	0.0207	-49.5681	-44.0423	-2.0781	-2.1850
0.657	0.1484	220	0.6775	-0.2293	-0.2571	0.5854	0.0278	-50.3305	-44.7337	-2.0281	-2.1355
0.6713	0.1619	240	0.6742	-0.2995	-0.3347	0.5457	0.0352	-51.1061	-45.4357	-1.9844	-2.0924
0.6626	0.1754	260	0.6694	-0.3826	-0.4307	0.5915	0.0482	-52.0664	-46.2660	-1.9233	-2.0311
0.712	0.1889	280	0.6656	-0.4568	-0.5151	0.5701	0.0583	-52.9107	-47.0087	-1.8818	-1.9902
0.679	0.2024	300	0.6609	-0.5204	-0.5947	0.5976	0.0742	-53.7060	-47.6449	-1.8547	-1.9642
0.5932	0.2159	320	0.6569	-0.5986	-0.6825	0.5793	0.0839	-54.5843	-48.4259	-1.8133	-1.9230
0.636	0.2294	340	0.6531	-0.7563	-0.8653	0.5854	0.1090	-56.4126	-50.0035	-1.7399	-1.8510
0.6755	0.2429	360	0.6484	-0.7939	-0.9129	0.6006	0.1190	-56.8879	-50.3793	-1.7252	-1.8370
0.633	0.2564	380	0.6430	-0.8458	-0.9910	0.625	0.1452	-57.6698	-50.8987	-1.7093	-1.8224
0.6464	0.2699	400	0.6374	-0.8640	-1.0294	0.6433	0.1654	-58.0533	-51.0804	-1.7080	-1.8226
0.5798	0.2834	420	0.6315	-0.8993	-1.0877	0.6494	0.1884	-58.6359	-51.4330	-1.6924	-1.8081
0.657	0.2969	440	0.6259	-0.8665	-1.0692	0.6646	0.2026	-58.4511	-51.1057	-1.7170	-1.8350
0.6099	0.3104	460	0.6204	-0.8976	-1.1205	0.6768	0.2229	-58.9641	-51.4164	-1.7137	-1.8322
0.606	0.3239	480	0.6147	-0.8988	-1.1425	0.6799	0.2437	-59.1845	-51.4289	-1.7339	-1.8538
0.5626	0.3374	500	0.6073	-0.9335	-1.2117	0.6799	0.2782	-59.8763	-51.7756	-1.7523	-1.8743
0.5897	0.3509	520	0.6017	-1.0087	-1.3222	0.6768	0.3135	-60.9816	-52.5279	-1.7366	-1.8594
0.5299	0.3644	540	0.5954	-1.0387	-1.3786	0.6982	0.3399	-61.5455	-52.8272	-1.7458	-1.8700
0.5706	0.3779	560	0.5887	-1.0716	-1.4546	0.7043	0.3830	-62.3049	-53.1563	-1.7719	-1.8984
0.6878	0.3914	580	0.5842	-1.1211	-1.5347	0.7073	0.4136	-63.1065	-53.6518	-1.7764	-1.9047
0.5076	0.4049	600	0.5808	-1.1596	-1.5925	0.7256	0.4329	-63.6841	-54.0362	-1.7808	-1.9105
0.5306	0.4184	620	0.5769	-1.2801	-1.7530	0.7195	0.4730	-65.2898	-55.2410	-1.7681	-1.8991
0.5482	0.4318	640	0.5716	-1.3625	-1.8872	0.7256	0.5247	-66.6318	-56.0659	-1.7937	-1.9268
0.6604	0.4453	660	0.5674	-1.3946	-1.9597	0.7104	0.5651	-67.3565	-56.3861	-1.8311	-1.9656
0.5764	0.4588	680	0.5649	-1.4188	-2.0048	0.7043	0.5860	-67.8071	-56.6283	-1.8578	-1.9929
0.5146	0.4723	700	0.5612	-1.4495	-2.0652	0.7104	0.6157	-68.4114	-56.9358	-1.8717	-2.0090
0.478	0.4858	720	0.5569	-1.4166	-2.0453	0.7104	0.6286	-68.2119	-56.6067	-1.8888	-2.0273
0.4643	0.4993	740	0.5533	-1.4500	-2.1055	0.7165	0.6556	-68.8148	-56.9402	-1.9043	-2.0431
0.56	0.5128	760	0.5497	-1.4193	-2.0717	0.7226	0.6524	-68.4768	-56.6334	-1.9119	-2.0507
0.4987	0.5263	780	0.5470	-1.3978	-2.0536	0.7378	0.6558	-68.2953	-56.4183	-1.9213	-2.0614
0.4931	0.5398	800	0.5439	-1.3768	-2.0439	0.7256	0.6671	-68.1987	-56.2083	-1.9359	-2.0771
0.5729	0.5533	820	0.5400	-1.4168	-2.1213	0.7287	0.7045	-68.9718	-56.6083	-1.9445	-2.0875
0.6182	0.5668	840	0.5367	-1.4102	-2.1339	0.7287	0.7238	-69.0985	-56.5419	-1.9709	-2.1154
0.3569	0.5803	860	0.5333	-1.3893	-2.1258	0.7226	0.7365	-69.0169	-56.3329	-2.0001	-2.1450
0.5866	0.5938	880	0.5300	-1.3970	-2.1555	0.7348	0.7585	-69.3145	-56.4109	-2.0157	-2.1616
0.3457	0.6073	900	0.5269	-1.3817	-2.1547	0.7256	0.7730	-69.3060	-56.2573	-2.0465	-2.1934
0.4482	0.6208	920	0.5242	-1.4241	-2.2355	0.7195	0.8114	-70.1148	-56.6814	-2.0735	-2.2209
0.4867	0.6343	940	0.5217	-1.4497	-2.2754	0.7348	0.8257	-70.5133	-56.9373	-2.0826	-2.2304
0.6182	0.6478	960	0.5214	-1.4629	-2.3005	0.7165	0.8376	-70.7643	-57.0693	-2.0844	-2.2311
0.5321	0.6613	980	0.5199	-1.4652	-2.3091	0.7195	0.8439	-70.8503	-57.0924	-2.0852	-2.2322
0.4696	0.6748	1000	0.5169	-1.4466	-2.3056	0.7226	0.8590	-70.8152	-56.9060	-2.0982	-2.2452
0.5199	0.6883	1020	0.5164	-1.4724	-2.3495	0.7165	0.8771	-71.2542	-57.1647	-2.1156	-2.2633
0.4838	0.7018	1040	0.5130	-1.4917	-2.3765	0.7226	0.8848	-71.5244	-57.3577	-2.1094	-2.2577
0.3759	0.7152	1060	0.5115	-1.5399	-2.4470	0.7256	0.9071	-72.2293	-57.8394	-2.1164	-2.2643
0.6272	0.7287	1080	0.5125	-1.5563	-2.4728	0.7317	0.9165	-72.4873	-58.0032	-2.1287	-2.2764
0.4223	0.7422	1100	0.5102	-1.5360	-2.4599	0.7287	0.9239	-72.3582	-57.8006	-2.1353	-2.2830
0.4409	0.7557	1120	0.5088	-1.5404	-2.4871	0.7348	0.9467	-72.6306	-57.8447	-2.1485	-2.2968
0.5385	0.7692	1140	0.5061	-1.5226	-2.4686	0.7287	0.9460	-72.4453	-57.6663	-2.1497	-2.2990
0.4458	0.7827	1160	0.5058	-1.4621	-2.3921	0.7317	0.9301	-71.6806	-57.0611	-2.1473	-2.2964
0.3353	0.7962	1180	0.5053	-1.4880	-2.4219	0.7409	0.9339	-71.9782	-57.3206	-2.1304	-2.2795
0.4865	0.8097	1200	0.5060	-1.5249	-2.4739	0.7348	0.9491	-72.4986	-57.6890	-2.1263	-2.2764
0.4198	0.8232	1220	0.5040	-1.5416	-2.5056	0.7287	0.9640	-72.8155	-57.8569	-2.1315	-2.2818
0.3259	0.8367	1240	0.5027	-1.5477	-2.5177	0.7317	0.9699	-72.9360	-57.9177	-2.1416	-2.2918
0.3897	0.8502	1260	0.5000	-1.5626	-2.5506	0.7256	0.9879	-73.2649	-58.0668	-2.1564	-2.3064
0.2842	0.8637	1280	0.4991	-1.5902	-2.5999	0.7378	1.0097	-73.7581	-58.3421	-2.1593	-2.3102
0.3422	0.8772	1300	0.5008	-1.6244	-2.6314	0.7348	1.0070	-74.0730	-58.6841	-2.1537	-2.3047
0.4053	0.8907	1320	0.5003	-1.6397	-2.6572	0.7348	1.0175	-74.3314	-58.8375	-2.1699	-2.3211
0.4892	0.9042	1340	0.4989	-1.6544	-2.6937	0.7409	1.0393	-74.6966	-58.9849	-2.1851	-2.3368
0.3946	0.9177	1360	0.4983	-1.6361	-2.6663	0.7439	1.0302	-74.4222	-58.8013	-2.1717	-2.3226
0.3453	0.9312	1380	0.4969	-1.6455	-2.6840	0.7348	1.0385	-74.5991	-58.8955	-2.1646	-2.3154
0.43	0.9447	1400	0.4970	-1.6513	-2.6934	0.7378	1.0421	-74.6938	-58.9539	-2.1762	-2.3280
0.4527	0.9582	1420	0.4967	-1.6572	-2.6959	0.7348	1.0387	-74.7181	-59.0126	-2.1736	-2.3258
0.6567	0.9717	1440	0.4964	-1.6448	-2.6927	0.7378	1.0479	-74.6859	-58.8884	-2.1817	-2.3340
0.31	0.9852	1460	0.4955	-1.6422	-2.6865	0.7439	1.0443	-74.6239	-58.8624	-2.1891	-2.3412
0.5545	0.9987	1480	0.4941	-1.6357	-2.6790	0.7409	1.0433	-74.5497	-58.7974	-2.1909	-2.3430
0.3024	1.0121	1500	0.4936	-1.6366	-2.6962	0.7470	1.0596	-74.7217	-58.8067	-2.2150	-2.3670
0.4762	1.0256	1520	0.4934	-1.6723	-2.7736	0.7256	1.1013	-75.4952	-59.1634	-2.2504	-2.4038
0.3378	1.0391	1540	0.4937	-1.7030	-2.8222	0.7470	1.1191	-75.9809	-59.4708	-2.2740	-2.4276
0.293	1.0526	1560	0.4940	-1.7160	-2.8608	0.7470	1.1448	-76.3674	-59.6000	-2.3024	-2.4569
0.3054	1.0661	1580	0.4953	-1.7460	-2.9117	0.7439	1.1657	-76.8763	-59.9006	-2.3126	-2.4677
0.1804	1.0796	1600	0.4957	-1.7699	-2.9665	0.7409	1.1966	-77.4244	-60.1396	-2.3319	-2.4872
0.3666	1.0931	1620	0.4969	-1.7820	-2.9927	0.7378	1.2108	-77.6868	-60.2599	-2.3377	-2.4938
0.3084	1.1066	1640	0.4971	-1.7779	-2.9878	0.7348	1.2099	-77.6372	-60.2192	-2.3473	-2.5031
0.3241	1.1201	1660	0.4983	-1.8401	-3.0607	0.7378	1.2206	-78.3662	-60.8411	-2.3450	-2.5006
0.3452	1.1336	1680	0.4981	-1.9265	-3.1733	0.7287	1.2468	-79.4928	-61.7054	-2.3547	-2.5112
0.3385	1.1471	1700	0.4986	-1.9185	-3.1751	0.7317	1.2566	-79.5099	-61.6251	-2.3615	-2.5176
0.2755	1.1606	1720	0.4984	-1.9373	-3.1973	0.7530	1.2600	-79.7323	-61.8136	-2.3578	-2.5133
0.3262	1.1741	1740	0.4999	-1.9657	-3.2225	0.7378	1.2568	-79.9847	-62.0975	-2.3584	-2.5149
0.2325	1.1876	1760	0.5004	-1.9808	-3.2627	0.7317	1.2819	-80.3864	-62.2486	-2.3829	-2.5390
0.4601	1.2011	1780	0.4995	-1.9740	-3.2654	0.7409	1.2914	-80.4132	-62.1802	-2.3887	-2.5450
0.2212	1.2146	1800	0.4988	-1.9907	-3.2841	0.7348	1.2935	-80.6004	-62.3469	-2.3893	-2.5451
0.2362	1.2281	1820	0.4993	-1.9958	-3.2962	0.7348	1.3003	-80.7212	-62.3989	-2.3919	-2.5479
0.1936	1.2416	1840	0.4995	-1.9981	-3.2994	0.7439	1.3012	-80.7528	-62.4216	-2.4053	-2.5616
0.2773	1.2551	1860	0.4994	-2.0067	-3.3146	0.7409	1.3079	-80.9054	-62.5078	-2.4009	-2.5571
0.3307	1.2686	1880	0.4977	-2.0075	-3.3134	0.7378	1.3059	-80.8932	-62.5156	-2.4085	-2.5651
0.2251	1.2821	1900	0.4989	-2.0109	-3.3070	0.7348	1.2961	-80.8294	-62.5493	-2.4116	-2.5690
0.2819	1.2955	1920	0.4989	-2.0051	-3.2992	0.7317	1.2941	-80.7518	-62.4918	-2.4121	-2.5686
0.34	1.3090	1940	0.4995	-2.0186	-3.3298	0.7378	1.3112	-81.0572	-62.6265	-2.4282	-2.5848
0.2324	1.3225	1960	0.5003	-2.0172	-3.3308	0.7378	1.3136	-81.0678	-62.6128	-2.4337	-2.5900
0.2626	1.3360	1980	0.5019	-2.0454	-3.3825	0.7317	1.3371	-81.5843	-62.8942	-2.4487	-2.6048
0.3427	1.3495	2000	0.5006	-2.0568	-3.4018	0.7256	1.3450	-81.7775	-63.0083	-2.4610	-2.6178
0.4233	1.3630	2020	0.5021	-2.0647	-3.4099	0.7317	1.3452	-81.8579	-63.0873	-2.4756	-2.6324
0.3095	1.3765	2040	0.5009	-2.0625	-3.4083	0.7256	1.3458	-81.8427	-63.0654	-2.4682	-2.6249
0.3898	1.3900	2060	0.5024	-2.0826	-3.4298	0.7287	1.3472	-82.0569	-63.2660	-2.4522	-2.6086
0.4208	1.4035	2080	0.5012	-2.0819	-3.4405	0.7256	1.3586	-82.1646	-63.2599	-2.4505	-2.6067
0.4068	1.4170	2100	0.5022	-2.0744	-3.4380	0.7470	1.3636	-82.1390	-63.1842	-2.4654	-2.6220
0.3034	1.4305	2120	0.4997	-2.0437	-3.3971	0.7378	1.3533	-81.7299	-62.8775	-2.4550	-2.6115
0.2935	1.4440	2140	0.4997	-2.0205	-3.3653	0.7317	1.3448	-81.4121	-62.6454	-2.4573	-2.6136
0.326	1.4575	2160	0.5006	-2.0116	-3.3581	0.7317	1.3465	-81.3400	-62.5564	-2.4693	-2.6259
0.2855	1.4710	2180	0.5011	-2.0281	-3.3684	0.7348	1.3404	-81.4435	-62.7211	-2.4571	-2.6132
0.2625	1.4845	2200	0.5005	-2.0293	-3.3834	0.7378	1.3541	-81.5932	-62.7332	-2.4692	-2.6257
0.3465	1.4980	2220	0.5019	-2.0628	-3.4123	0.7287	1.3495	-81.8820	-63.0684	-2.4742	-2.6307
0.3329	1.5115	2240	0.5033	-2.0698	-3.4345	0.7317	1.3647	-82.1046	-63.1383	-2.4826	-2.6385
0.4216	1.5250	2260	0.5032	-2.0750	-3.4368	0.7409	1.3617	-82.1269	-63.1908	-2.4763	-2.6324
0.5066	1.5385	2280	0.5027	-2.0625	-3.4255	0.7470	1.3630	-82.0146	-63.0659	-2.4745	-2.6309
0.319	1.5520	2300	0.5012	-2.0558	-3.4188	0.7287	1.3631	-81.9477	-62.9981	-2.4721	-2.6287
0.3912	1.5655	2320	0.5012	-2.0630	-3.4279	0.7317	1.3649	-82.0381	-63.0702	-2.4766	-2.6333
0.3579	1.5789	2340	0.5013	-2.0652	-3.4322	0.7287	1.3670	-82.0817	-63.0927	-2.4800	-2.6364
0.2608	1.5924	2360	0.5011	-2.0648	-3.4198	0.7317	1.3550	-81.9573	-63.0886	-2.4742	-2.6301
0.2855	1.6059	2380	0.5021	-2.0574	-3.4155	0.7348	1.3580	-81.9138	-63.0148	-2.4747	-2.6304
0.3378	1.6194	2400	0.5005	-2.0567	-3.4286	0.7348	1.3719	-82.0454	-63.0070	-2.4833	-2.6389
0.2207	1.6329	2420	0.5019	-2.0798	-3.4524	0.7287	1.3726	-82.2834	-63.2387	-2.4812	-2.6369
0.2438	1.6464	2440	0.5014	-2.0892	-3.4657	0.7256	1.3765	-82.4162	-63.3319	-2.4842	-2.6411
0.2861	1.6599	2460	0.5008	-2.0844	-3.4532	0.7348	1.3688	-82.2918	-63.2844	-2.4851	-2.6413
0.348	1.6734	2480	0.5011	-2.0950	-3.4613	0.7195	1.3664	-82.3728	-63.3902	-2.4868	-2.6434
0.4291	1.6869	2500	0.5012	-2.0955	-3.4740	0.7317	1.3785	-82.4994	-63.3952	-2.4883	-2.6441
0.2938	1.7004	2520	0.5008	-2.1066	-3.4951	0.7256	1.3885	-82.7108	-63.5064	-2.4878	-2.6431
0.2641	1.7139	2540	0.5007	-2.0981	-3.4876	0.7287	1.3895	-82.6352	-63.4209	-2.4912	-2.6464
0.3547	1.7274	2560	0.5005	-2.0981	-3.4842	0.7378	1.3861	-82.6013	-63.4209	-2.4933	-2.6490
0.3516	1.7409	2580	0.5025	-2.0901	-3.4695	0.7317	1.3794	-82.4539	-63.3413	-2.4912	-2.6466
0.2512	1.7544	2600	0.5003	-2.0899	-3.4789	0.7378	1.3890	-82.5487	-63.3397	-2.4951	-2.6510
0.2101	1.7679	2620	0.5014	-2.0927	-3.4734	0.7317	1.3807	-82.4929	-63.3674	-2.4932	-2.6486
0.2488	1.7814	2640	0.5012	-2.0904	-3.4767	0.7256	1.3863	-82.5259	-63.3440	-2.4996	-2.6551
0.3607	1.7949	2660	0.5020	-2.0936	-3.4806	0.7317	1.3870	-82.5656	-63.3767	-2.4979	-2.6538
0.4349	1.8084	2680	0.5011	-2.0926	-3.4821	0.7317	1.3895	-82.5806	-63.3668	-2.5023	-2.6582
0.2374	1.8219	2700	0.5026	-2.1029	-3.4853	0.7287	1.3823	-82.6122	-63.4699	-2.4961	-2.6513
0.3131	1.8354	2720	0.5025	-2.1034	-3.4903	0.7378	1.3869	-82.6621	-63.4740	-2.4960	-2.6517
0.2307	1.8489	2740	0.5018	-2.0985	-3.4863	0.7226	1.3878	-82.6223	-63.4256	-2.4959	-2.6514
0.3829	1.8623	2760	0.5020	-2.1036	-3.4865	0.7256	1.3829	-82.6244	-63.4760	-2.4951	-2.6504
0.2897	1.8758	2780	0.5012	-2.0994	-3.4912	0.7378	1.3918	-82.6709	-63.4340	-2.4959	-2.6516
0.4397	1.8893	2800	0.5019	-2.1020	-3.4885	0.7348	1.3865	-82.6441	-63.4601	-2.4947	-2.6496
0.3054	1.9028	2820	0.5013	-2.0950	-3.4898	0.7378	1.3948	-82.6570	-63.3905	-2.4970	-2.6525
0.2605	1.9163	2840	0.5013	-2.1024	-3.4881	0.7317	1.3857	-82.6401	-63.4641	-2.4978	-2.6533
0.3978	1.9298	2860	0.5024	-2.1026	-3.4889	0.7317	1.3863	-82.6483	-63.4665	-2.4985	-2.6542
0.1836	1.9433	2880	0.5027	-2.1029	-3.4904	0.7348	1.3875	-82.6633	-63.4694	-2.4945	-2.6494
0.1722	1.9568	2900	0.5015	-2.0995	-3.4892	0.7226	1.3897	-82.6516	-63.4353	-2.5001	-2.6556
0.461	1.9703	2920	0.5018	-2.1045	-3.4866	0.7409	1.3822	-82.6258	-63.4852	-2.4956	-2.6513
0.4559	1.9838	2940	0.5014	-2.0918	-3.4861	0.7348	1.3943	-82.6199	-63.3583	-2.4962	-2.6515
0.283	1.9973	2960	0.5012	-2.0987	-3.4833	0.7348	1.3846	-82.5923	-63.4273	-2.4968	-2.6520

Framework versions

Transformers 4.45.2
Pytorch 2.5.1+cu121
Datasets 3.5.0
Tokenizers 0.20.3

YuchenLi01/genParaMoreUniqueResNoGT_Qwen2.5-1.5BInstruct_dpo_ebs32_lr5e-07_beta0.1_42

genParaMoreUniqueResNoGT_Qwen2.5-1.5BInstruct_dpo_ebs32_lr5e-07_beta0.1_42

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

YuchenLi01/
genParaMoreUniqueResNoGT_Qwen2.5-1.5BInstruct_dpo_ebs32_lr5e-07_beta0.1_42