train_copa_1745950322

This model is a fine-tuned version of google/gemma-3-1b-it on the copa dataset. It achieves the following results on the evaluation set:

• Loss: 0.1742
• Num Input Tokens Seen: 11200800

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: 0.3
• train_batch_size: 2
• eval_batch_size: 2
• seed: 123
• gradient_accumulation_steps: 2
• total_train_batch_size: 4
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• training_steps: 40000

Training results

Training Loss	Epoch	Step	Validation Loss	Input Tokens Seen
0.1585	2.2222	200	0.1960	56176
0.1537	4.4444	400	0.1932	112064
0.1745	6.6667	600	0.1866	168112
0.1615	8.8889	800	0.1828	224048
0.1705	11.1111	1000	0.1820	279904
0.1756	13.3333	1200	0.1769	336032
0.1546	15.5556	1400	0.1742	391904
0.1698	17.7778	1600	0.1744	448112
0.1734	20.0	1800	0.1756	503920
0.1819	22.2222	2000	0.1746	560016
0.1603	24.4444	2200	0.1819	615952
0.177	26.6667	2400	0.1764	672080
0.176	28.8889	2600	0.1817	728000
0.1629	31.1111	2800	0.1786	783872
0.1461	33.3333	3000	0.1881	839808
0.1762	35.5556	3200	0.1912	896064
0.1672	37.7778	3400	0.1763	951888
0.1666	40.0	3600	0.1825	1007760
0.1347	42.2222	3800	0.2474	1063648
0.1005	44.4444	4000	0.2612	1119744
0.0669	46.6667	4200	0.2604	1175680
0.1043	48.8889	4400	0.2414	1231696
0.0445	51.1111	4600	0.3379	1287744
0.0243	53.3333	4800	0.4504	1343760
0.0371	55.5556	5000	0.5648	1399856
0.0138	57.7778	5200	0.5205	1455808
0.0002	60.0	5400	0.7808	1511856
0.0001	62.2222	5600	0.7838	1567808
0.0	64.4444	5800	0.8040	1623744
0.0	66.6667	6000	0.8280	1679888
0.0	68.8889	6200	0.8334	1735952
0.0	71.1111	6400	0.8474	1791904
0.0	73.3333	6600	0.8694	1847888
0.0	75.5556	6800	0.8714	1904000
0.0	77.7778	7000	0.8972	1959920
0.0	80.0	7200	0.9046	2015792
0.0	82.2222	7400	0.9208	2071808
0.0	84.4444	7600	0.9380	2127808
0.0	86.6667	7800	0.9472	2183888
0.0	88.8889	8000	0.9389	2239840
0.0	91.1111	8200	0.9594	2295888
0.0	93.3333	8400	0.9713	2351872
0.0	95.5556	8600	0.9810	2407824
0.0	97.7778	8800	0.9830	2463744
0.0	100.0	9000	1.0046	2519680
0.0	102.2222	9200	1.0120	2575584
0.0	104.4444	9400	1.0307	2631680
0.0	106.6667	9600	1.0327	2687728
0.0	108.8889	9800	1.0424	2743792
0.0	111.1111	10000	1.0433	2799840
0.0	113.3333	10200	1.0506	2855808
0.0	115.5556	10400	1.0754	2911648
0.0	117.7778	10600	1.0816	2967856
0.0	120.0	10800	1.0875	3023792
0.0	122.2222	11000	1.1016	3079920
0.0	124.4444	11200	1.1077	3135904
0.0	126.6667	11400	1.1129	3191808
0.0	128.8889	11600	1.1311	3247840
0.0	131.1111	11800	1.1461	3303712
0.0	133.3333	12000	1.1524	3359680
0.0	135.5556	12200	1.1528	3415824
0.0	137.7778	12400	1.1764	3471520
0.0	140.0	12600	1.1755	3527664
0.0	142.2222	12800	1.2023	3583696
0.0	144.4444	13000	1.1986	3639680
0.0	146.6667	13200	1.2131	3695712
0.0	148.8889	13400	1.2381	3751728
0.0	151.1111	13600	1.2481	3807744
0.0	153.3333	13800	1.2522	3863664
0.0	155.5556	14000	1.2715	3919584
0.0	157.7778	14200	1.2780	3975568
0.0	160.0	14400	1.3001	4031632
0.0	162.2222	14600	1.3049	4087632
0.0	164.4444	14800	1.3115	4143664
0.0	166.6667	15000	1.3477	4199552
0.0	168.8889	15200	1.3329	4255584
0.0	171.1111	15400	1.3379	4311504
0.0	173.3333	15600	1.3553	4367408
0.0	175.5556	15800	1.3785	4423376
0.0	177.7778	16000	1.3628	4479456
0.0	180.0	16200	1.3936	4535504
0.0	182.2222	16400	1.3908	4591504
0.0	184.4444	16600	1.4268	4647424
0.0	186.6667	16800	1.4218	4703376
0.0	188.8889	17000	1.4472	4759552
0.0	191.1111	17200	1.4649	4815552
0.0	193.3333	17400	1.4669	4871600
0.0	195.5556	17600	1.4431	4927696
0.0	197.7778	17800	1.4888	4983424
0.0	200.0	18000	1.5016	5039536
0.0	202.2222	18200	1.4928	5095376
0.0	204.4444	18400	1.5293	5151440
0.0	206.6667	18600	1.5467	5207488
0.0	208.8889	18800	1.5432	5263360
0.0	211.1111	19000	1.5500	5319344
0.0	213.3333	19200	1.5504	5375280
0.0	215.5556	19400	1.5739	5431520
0.0	217.7778	19600	1.5765	5487472
0.0	220.0	19800	1.5911	5543504
0.0	222.2222	20000	1.5940	5599440
0.0	224.4444	20200	1.5977	5655424
0.0	226.6667	20400	1.6347	5711344
0.0	228.8889	20600	1.6275	5767376
0.0	231.1111	20800	1.6913	5823264
0.0	233.3333	21000	1.6944	5879248
0.0	235.5556	21200	1.6750	5935168
0.0	237.7778	21400	1.6816	5991232
0.0	240.0	21600	1.6530	6047376
0.0	242.2222	21800	1.6663	6103328
0.0	244.4444	22000	1.6708	6159376
0.0	246.6667	22200	1.6437	6215360
0.0	248.8889	22400	1.6692	6271232
0.0	251.1111	22600	1.6101	6327136
0.0	253.3333	22800	1.6198	6383248
0.0	255.5556	23000	1.5668	6439168
0.0	257.7778	23200	1.5763	6495280
0.0	260.0	23400	1.5500	6551264
0.0	262.2222	23600	1.5696	6607424
0.0	264.4444	23800	1.5225	6663168
0.0	266.6667	24000	1.5554	6719216
0.0	268.8889	24200	1.5902	6775344
0.0	271.1111	24400	1.4873	6831344
0.0	273.3333	24600	1.5270	6887344
0.0	275.5556	24800	1.6768	6943632
0.0	277.7778	25000	1.6876	6999632
0.0	280.0	25200	1.5999	7055664
0.0	282.2222	25400	1.6702	7111664
0.0	284.4444	25600	1.6623	7167744
0.0	286.6667	25800	1.5950	7223696
0.0	288.8889	26000	1.6427	7279760
0.0	291.1111	26200	1.7028	7335792
0.0	293.3333	26400	1.6055	7391808
0.0	295.5556	26600	1.5844	7447808
0.0	297.7778	26800	1.6416	7503824
0.0	300.0	27000	1.7239	7559856
0.0	302.2222	27200	1.6796	7615904
0.0	304.4444	27400	1.5742	7672000
0.0	306.6667	27600	1.6184	7727808
0.0	308.8889	27800	1.5991	7783744
0.0	311.1111	28000	1.5418	7839808
0.0	313.3333	28200	1.6170	7895872
0.0	315.5556	28400	1.6391	7951664
0.0	317.7778	28600	1.6250	8007744
0.0	320.0	28800	1.6387	8063616
0.0	322.2222	29000	1.6151	8119520
0.0	324.4444	29200	1.6220	8175584
0.0	326.6667	29400	1.6851	8231760
0.0	328.8889	29600	1.6862	8287696
0.0	331.1111	29800	1.6520	8343760
0.0	333.3333	30000	1.6907	8399696
0.0	335.5556	30200	1.6143	8455776
0.0	337.7778	30400	1.6979	8511760
0.0	340.0	30600	1.6834	8567792
0.0	342.2222	30800	1.6586	8623728
0.0	344.4444	31000	1.6678	8679920
0.0	346.6667	31200	1.7132	8736032
0.0	348.8889	31400	1.7332	8791888
0.0	351.1111	31600	1.6019	8847728
0.0	353.3333	31800	1.6742	8903952
0.0	355.5556	32000	1.7011	8959920
0.0	357.7778	32200	1.6816	9016096
0.0	360.0	32400	1.7048	9072192
0.0	362.2222	32600	1.7152	9128272
0.0	364.4444	32800	1.7085	9184240
0.0	366.6667	33000	1.7312	9240064
0.0	368.8889	33200	1.7539	9295952
0.0	371.1111	33400	1.7249	9352016
0.0	373.3333	33600	1.7504	9407968
0.0	375.5556	33800	1.7512	9463920
0.0	377.7778	34000	1.7510	9519984
0.0	380.0	34200	1.7470	9575936
0.0	382.2222	34400	1.7684	9631952
0.0	384.4444	34600	1.7600	9687936
0.0	386.6667	34800	1.7497	9743968
0.0	388.8889	35000	1.7552	9800016
0.0	391.1111	35200	1.7796	9856016
0.0	393.3333	35400	1.7958	9912112
0.0	395.5556	35600	1.7898	9968112
0.0	397.7778	35800	1.7811	10024160
0.0	400.0	36000	1.7936	10080240
0.0	402.2222	36200	1.7834	10136208
0.0	404.4444	36400	1.7830	10192208
0.0	406.6667	36600	1.8050	10248192
0.0	408.8889	36800	1.7914	10304144
0.0	411.1111	37000	1.8239	10360192
0.0	413.3333	37200	1.7780	10416288
0.0	415.5556	37400	1.7846	10472368
0.0	417.7778	37600	1.7938	10528352
0.0	420.0	37800	1.7924	10584384
0.0	422.2222	38000	1.7995	10640496
0.0	424.4444	38200	1.8110	10696528
0.0	426.6667	38400	1.7964	10752640
0.0	428.8889	38600	1.8125	10808672
0.0	431.1111	38800	1.7951	10864512
0.0	433.3333	39000	1.8056	10920608
0.0	435.5556	39200	1.7908	10976624
0.0	437.7778	39400	1.8039	11032608
0.0	440.0	39600	1.7930	11088720
0.0	442.2222	39800	1.7873	11144688
0.0	444.4444	40000	1.7878	11200800

Framework versions

• PEFT 0.15.2.dev0
• Transformers 4.51.3
• Pytorch 2.6.0+cu124
• Datasets 3.5.0
• Tokenizers 0.21.1