train_copa_1745950331

This model is a fine-tuned version of mistralai/Mistral-7B-Instruct-v0.3 on the copa dataset. It achieves the following results on the evaluation set:

• Loss: 0.1137
• Num Input Tokens Seen: 11206480

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: 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
0.0001	2.2222	200	0.1137	56064
0.0	4.4444	400	0.1598	112064
0.0	6.6667	600	0.1698	168096
0.0	8.8889	800	0.1829	224048
0.0	11.1111	1000	0.1941	280048
0.0	13.3333	1200	0.1876	336032
0.0	15.5556	1400	0.1744	392032
0.0	17.7778	1600	0.2256	448128
0.0	20.0	1800	0.2210	503904
0.0	22.2222	2000	0.2241	559936
0.0	24.4444	2200	0.2302	615968
0.0	26.6667	2400	0.2274	672064
0.0	28.8889	2600	0.2323	728128
0.0	31.1111	2800	0.2305	784032
0.0	33.3333	3000	0.2234	839984
0.0	35.5556	3200	0.2326	896288
0.0	37.7778	3400	0.2253	952128
0.0	40.0	3600	0.2456	1008096
0.0	42.2222	3800	0.2680	1063984
0.0	44.4444	4000	0.2292	1120080
0.0	46.6667	4200	0.2472	1176240
0.0	48.8889	4400	0.2371	1232160
0.0	51.1111	4600	0.2621	1288160
0.0	53.3333	4800	0.2626	1344160
0.0	55.5556	5000	0.2676	1400368
0.0	57.7778	5200	0.2592	1456368
0.0	60.0	5400	0.2542	1512336
0.0	62.2222	5600	0.2448	1568192
0.0	64.4444	5800	0.2488	1624288
0.0	66.6667	6000	0.2498	1680352
0.0	68.8889	6200	0.2324	1736384
0.0	71.1111	6400	0.2660	1792480
0.0	73.3333	6600	0.2607	1848416
0.0	75.5556	6800	0.2732	1904480
0.0	77.7778	7000	0.2680	1960496
0.0	80.0	7200	0.2818	2016368
0.0	82.2222	7400	0.2566	2072400
0.0	84.4444	7600	0.2850	2128384
0.0	86.6667	7800	0.2502	2184416
0.0	88.8889	8000	0.2657	2240512
0.0	91.1111	8200	0.2781	2296496
0.0	93.3333	8400	0.2922	2352560
0.0	95.5556	8600	0.2896	2408640
0.0	97.7778	8800	0.3003	2464672
0.0	100.0	9000	0.2869	2520688
0.0	102.2222	9200	0.3022	2576656
0.0	104.4444	9400	0.2933	2632720
0.0	106.6667	9600	0.3192	2688704
0.0	108.8889	9800	0.2988	2744768
0.0	111.1111	10000	0.3447	2800768
0.0	113.3333	10200	0.3235	2856768
0.0	115.5556	10400	0.3245	2912640
0.0	117.7778	10600	0.3207	2968832
0.0	120.0	10800	0.2889	3024896
0.0	122.2222	11000	0.3321	3081056
0.0	124.4444	11200	0.3126	3136944
0.0	126.6667	11400	0.3366	3192960
0.0	128.8889	11600	0.3456	3248976
0.0	131.1111	11800	0.3281	3305024
0.0	133.3333	12000	0.3490	3361008
0.0	135.5556	12200	0.3237	3417152
0.0	137.7778	12400	0.3755	3472832
0.0	140.0	12600	0.3279	3529008
0.0	142.2222	12800	0.3029	3585200
0.0	144.4444	13000	0.3640	3641200
0.0	146.6667	13200	0.3434	3697232
0.0	148.8889	13400	0.3727	3753168
0.0	151.1111	13600	0.3504	3809136
0.0	153.3333	13800	0.3874	3865216
0.0	155.5556	14000	0.3806	3921216
0.0	157.7778	14200	0.3886	3977312
0.0	160.0	14400	0.3641	4033488
0.0	162.2222	14600	0.3967	4089504
0.0	164.4444	14800	0.4044	4145504
0.0	166.6667	15000	0.4141	4201440
0.0	168.8889	15200	0.3544	4257504
0.0	171.1111	15400	0.3646	4313408
0.0	173.3333	15600	0.3906	4369488
0.0	175.5556	15800	0.3670	4425536
0.0	177.7778	16000	0.3432	4481568
0.0	180.0	16200	0.3773	4537616
0.0	182.2222	16400	0.3990	4593600
0.0	184.4444	16600	0.3898	4649664
0.0	186.6667	16800	0.4068	4705600
0.0	188.8889	17000	0.3945	4761760
0.0	191.1111	17200	0.3966	4817728
0.0	193.3333	17400	0.3901	4873856
0.0	195.5556	17600	0.4018	4929936
0.0	197.7778	17800	0.3664	4985840
0.0	200.0	18000	0.3878	5041920
0.0	202.2222	18200	0.3578	5097872
0.0	204.4444	18400	0.3875	5154064
0.0	206.6667	18600	0.3622	5210112
0.0	208.8889	18800	0.3671	5266064
0.0	211.1111	19000	0.3801	5322160
0.0	213.3333	19200	0.3521	5378224
0.0	215.5556	19400	0.3749	5434432
0.0	217.7778	19600	0.3714	5490352
0.0	220.0	19800	0.3600	5546432
0.0	222.2222	20000	0.4014	5602400
0.0	224.4444	20200	0.3992	5658464
0.0	226.6667	20400	0.3898	5714352
0.0	228.8889	20600	0.3508	5770416
0.0	231.1111	20800	0.3746	5826496
0.0	233.3333	21000	0.3576	5882496
0.0	235.5556	21200	0.3688	5938432
0.0	237.7778	21400	0.3703	5994480
0.0	240.0	21600	0.3403	6050656
0.0	242.2222	21800	0.3380	6106736
0.0	244.4444	22000	0.3652	6162896
0.0	246.6667	22200	0.3580	6218976
0.0	248.8889	22400	0.3398	6274960
0.0	251.1111	22600	0.3740	6331008
0.0	253.3333	22800	0.3142	6387152
0.0	255.5556	23000	0.3603	6443200
0.0	257.7778	23200	0.3027	6499088
0.0	260.0	23400	0.3435	6555184
0.0	262.2222	23600	0.3033	6611312
0.0	264.4444	23800	0.3074	6667104
0.0	266.6667	24000	0.3159	6723024
0.0	268.8889	24200	0.3148	6779376
0.0	271.1111	24400	0.3101	6835232
0.0	273.3333	24600	0.3031	6891104
0.0	275.5556	24800	0.3391	6947456
0.0	277.7778	25000	0.3461	7003408
0.0	280.0	25200	0.3379	7059536
0.0	282.2222	25400	0.3488	7115504
0.0	284.4444	25600	0.3670	7171744
0.0	286.6667	25800	0.3277	7227712
0.0	288.8889	26000	0.3057	7283856
0.0	291.1111	26200	0.3654	7339872
0.0	293.3333	26400	0.3226	7395808
0.0	295.5556	26600	0.3204	7451904
0.0	297.7778	26800	0.3143	7507792
0.0	300.0	27000	0.3090	7563888
0.0	302.2222	27200	0.3612	7619872
0.0	304.4444	27400	0.3267	7676016
0.0	306.6667	27600	0.3082	7731872
0.0	308.8889	27800	0.3292	7787920
0.0	311.1111	28000	0.3150	7844080
0.0	313.3333	28200	0.3051	7900064
0.0	315.5556	28400	0.3160	7956016
0.0	317.7778	28600	0.3321	8012160
0.0	320.0	28800	0.3551	8068256
0.0	322.2222	29000	0.3725	8124112
0.0	324.4444	29200	0.3379	8180192
0.0	326.6667	29400	0.3742	8236304
0.0	328.8889	29600	0.3082	8292272
0.0	331.1111	29800	0.3802	8348416
0.0	333.3333	30000	0.3239	8404432
0.0	335.5556	30200	0.3417	8460384
0.0	337.7778	30400	0.3462	8516432
0.0	340.0	30600	0.3548	8572496
0.0	342.2222	30800	0.3141	8628448
0.0	344.4444	31000	0.3564	8684672
0.0	346.6667	31200	0.3087	8740800
0.0	348.8889	31400	0.3289	8796784
0.0	351.1111	31600	0.3282	8852784
0.0	353.3333	31800	0.3820	8909040
0.0	355.5556	32000	0.3516	8965104
0.0	357.7778	32200	0.3089	9021344
0.0	360.0	32400	0.3822	9077456
0.0	362.2222	32600	0.3348	9133648
0.0	364.4444	32800	0.3841	9189616
0.0	366.6667	33000	0.3711	9245504
0.0	368.8889	33200	0.3494	9301520
0.0	371.1111	33400	0.3389	9357712
0.0	373.3333	33600	0.3506	9413712
0.0	375.5556	33800	0.3593	9469696
0.0	377.7778	34000	0.3583	9525760
0.0	380.0	34200	0.3855	9581648
0.0	382.2222	34400	0.3392	9637632
0.0	384.4444	34600	0.3270	9693568
0.0	386.6667	34800	0.3716	9749792
0.0	388.8889	35000	0.3475	9805840
0.0	391.1111	35200	0.3592	9861856
0.0	393.3333	35400	0.3519	9917904
0.0	395.5556	35600	0.3689	9973888
0.0	397.7778	35800	0.3381	10030096
0.0	400.0	36000	0.3452	10086192
0.0	402.2222	36200	0.3348	10142304
0.0	404.4444	36400	0.3222	10198320
0.0	406.6667	36600	0.3729	10254256
0.0	408.8889	36800	0.3392	10310096
0.0	411.1111	37000	0.3408	10366160
0.0	413.3333	37200	0.3526	10422192
0.0	415.5556	37400	0.3385	10478368
0.0	417.7778	37600	0.3754	10534240
0.0	420.0	37800	0.4234	10590208
0.0	422.2222	38000	0.3726	10646384
0.0	424.4444	38200	0.3496	10702336
0.0	426.6667	38400	0.3405	10758400
0.0	428.8889	38600	0.3523	10814480
0.0	431.1111	38800	0.3756	10870400
0.0	433.3333	39000	0.3986	10926320
0.0	435.5556	39200	0.3526	10982240
0.0	437.7778	39400	0.3900	11038352
0.0	440.0	39600	0.3569	11094352
0.0	442.2222	39800	0.3496	11150400
0.0	444.4444	40000	0.3374	11206480

Framework versions

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