train_boolq_1745950277

This model is a fine-tuned version of meta-llama/Meta-Llama-3-8B-Instruct on the boolq dataset. It achieves the following results on the evaluation set:

• Loss: 0.1300
• Num Input Tokens Seen: 34078960

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
0.2949	0.0943	200	0.3250	171472
0.3203	0.1886	400	0.2047	339520
0.213	0.2829	600	0.1600	509632
0.1715	0.3772	800	0.1526	685120
0.0514	0.4715	1000	0.1652	856144
0.2019	0.5658	1200	0.1425	1024448
0.2173	0.6601	1400	0.1650	1192560
0.0799	0.7544	1600	0.1300	1360304
0.041	0.8487	1800	0.1564	1535088
0.3322	0.9430	2000	0.1519	1708224
0.0089	1.0372	2200	0.1838	1880336
0.2081	1.1315	2400	0.2094	2048144
0.38	1.2258	2600	0.2108	2220048
0.1835	1.3201	2800	0.1721	2388128
0.0032	1.4144	3000	0.2218	2559696
0.0865	1.5087	3200	0.1540	2730224
0.0565	1.6030	3400	0.1570	2897744
0.1569	1.6973	3600	0.1477	3067728
0.1551	1.7916	3800	0.1567	3235856
0.1202	1.8859	4000	0.2155	3409536
0.0059	1.9802	4200	0.1554	3581600
0.0025	2.0745	4400	0.2246	3753552
0.001	2.1688	4600	0.2449	3924224
0.006	2.2631	4800	0.2078	4093360
0.1844	2.3574	5000	0.2490	4261312
0.0018	2.4517	5200	0.2213	4438224
0.096	2.5460	5400	0.2142	4608992
0.0021	2.6403	5600	0.2055	4780976
0.0011	2.7346	5800	0.2560	4946848
0.0591	2.8289	6000	0.2815	5120704
0.1611	2.9231	6200	0.2788	5292816
0.0008	3.0174	6400	0.2730	5463728
0.0023	3.1117	6600	0.2917	5634528
0.1115	3.2060	6800	0.2372	5805344
0.1566	3.3003	7000	0.2747	5976160
0.0014	3.3946	7200	0.2818	6147200
0.0018	3.4889	7400	0.2642	6315984
0.0003	3.5832	7600	0.3310	6484512
0.0006	3.6775	7800	0.3094	6653536
0.166	3.7718	8000	0.3492	6823184
0.0006	3.8661	8200	0.2910	6991232
0.0014	3.9604	8400	0.2436	7161488
0.0043	4.0547	8600	0.3267	7330416
0.0002	4.1490	8800	0.4208	7502960
0.0002	4.2433	9000	0.3043	7675776
0.0008	4.3376	9200	0.3460	7847200
0.0001	4.4319	9400	0.4261	8016080
0.1008	4.5262	9600	0.2292	8189040
0.0001	4.6205	9800	0.3800	8355360
0.0002	4.7148	10000	0.3418	8527824
0.0004	4.8091	10200	0.2841	8697376
0.0012	4.9033	10400	0.2540	8867872
0.0013	4.9976	10600	0.3265	9039888
0.1067	5.0919	10800	0.4539	9209232
0.0	5.1862	11000	0.5187	9384064
0.0023	5.2805	11200	0.3918	9555168
0.0	5.3748	11400	0.4749	9724832
0.0001	5.4691	11600	0.3953	9894608
0.1221	5.5634	11800	0.3914	10067376
0.0015	5.6577	12000	0.6157	10239648
0.0001	5.7520	12200	0.4068	10406400
0.0001	5.8463	12400	0.3948	10578176
0.0009	5.9406	12600	0.4626	10745200
0.0	6.0349	12800	0.4054	10917056
0.0	6.1292	13000	0.4441	11091376
0.0038	6.2235	13200	0.3147	11260160
0.0001	6.3178	13400	0.3441	11430512
0.0	6.4121	13600	0.5056	11598992
0.0	6.5064	13800	0.4643	11771312
0.0	6.6007	14000	0.4942	11940256
0.0	6.6950	14200	0.5800	12108896
0.0	6.7893	14400	0.5291	12277824
0.0	6.8835	14600	0.4069	12450224
0.0004	6.9778	14800	0.3468	12618544
0.4969	7.0721	15000	0.4180	12791104
0.0	7.1664	15200	0.4630	12964976
0.0	7.2607	15400	0.4689	13132848
0.0	7.3550	15600	0.4131	13302528
0.0	7.4493	15800	0.5349	13471696
0.0	7.5436	16000	0.4500	13643264
0.0	7.6379	16200	0.3913	13810336
0.0002	7.7322	16400	0.4377	13980096
0.0	7.8265	16600	0.5126	14150688
0.0001	7.9208	16800	0.4151	14320928
0.0	8.0151	17000	0.4575	14497120
0.0001	8.1094	17200	0.4715	14667440
0.0001	8.2037	17400	0.3730	14839920
0.0001	8.2980	17600	0.4248	15013152
0.0	8.3923	17800	0.5033	15178480
0.0	8.4866	18000	0.5056	15349216
0.0	8.5809	18200	0.5075	15518736
0.0	8.6752	18400	0.4976	15689568
0.0011	8.7694	18600	0.3954	15859632
0.0	8.8637	18800	0.4307	16025920
0.0001	8.9580	19000	0.3383	16196560
0.0	9.0523	19200	0.5251	16368144
0.0	9.1466	19400	0.6275	16539952
0.0	9.2409	19600	0.6473	16710384
0.0	9.3352	19800	0.4718	16878624
0.0	9.4295	20000	0.5149	17046992
0.0	9.5238	20200	0.5534	17218176
0.0	9.6181	20400	0.6049	17390384
0.0	9.7124	20600	0.6236	17560512
0.0	9.8067	20800	0.4179	17726048
0.0	9.9010	21000	0.4773	17897376
0.0001	9.9953	21200	0.4315	18068096
0.0	10.0896	21400	0.4717	18244704
0.0	10.1839	21600	0.4507	18420464
0.0	10.2782	21800	0.5002	18588480
0.0	10.3725	22000	0.4912	18758992
0.0	10.4668	22200	0.5160	18930688
0.0	10.5611	22400	0.5553	19096400
0.0	10.6554	22600	0.4440	19263456
0.0	10.7496	22800	0.5174	19430576
0.0001	10.8439	23000	0.4635	19599200
0.0	10.9382	23200	0.4561	19770608
0.0	11.0325	23400	0.5011	19942144
0.0	11.1268	23600	0.5298	20112704
0.0	11.2211	23800	0.5707	20282048
0.0	11.3154	24000	0.5005	20456384
0.0	11.4097	24200	0.4854	20624672
0.0	11.5040	24400	0.5311	20796640
0.0	11.5983	24600	0.5520	20964288
0.0	11.6926	24800	0.6044	21132896
0.0	11.7869	25000	0.5289	21304080
0.0	11.8812	25200	0.5572	21471264
0.0	11.9755	25400	0.5636	21642432
0.0	12.0698	25600	0.5867	21811200
0.0	12.1641	25800	0.5944	21983552
0.0	12.2584	26000	0.5938	22155824
0.0	12.3527	26200	0.6187	22329984
0.0	12.4470	26400	0.4904	22499472
0.0	12.5413	26600	0.5263	22669904
0.0	12.6355	26800	0.5404	22837440
0.0	12.7298	27000	0.5808	23008384
0.0	12.8241	27200	0.5923	23177264
0.0001	12.9184	27400	0.6002	23344128
0.0	13.0127	27600	0.6057	23512272
0.0	13.1070	27800	0.6309	23680080
0.0	13.2013	28000	0.6362	23850944
0.0	13.2956	28200	0.6412	24022624
0.0	13.3899	28400	0.6448	24192144
0.0	13.4842	28600	0.6528	24364768
0.0	13.5785	28800	0.5868	24538352
0.0	13.6728	29000	0.5990	24710416
0.0	13.7671	29200	0.6152	24881936
0.0	13.8614	29400	0.6230	25050656
0.0	13.9557	29600	0.6278	25222752
0.0	14.0500	29800	0.6313	25389344
0.0	14.1443	30000	0.6579	25563712
0.0	14.2386	30200	0.6643	25738992
0.0	14.3329	30400	0.6687	25909744
0.0	14.4272	30600	0.6975	26079120
0.0	14.5215	30800	0.7041	26245936
0.0	14.6157	31000	0.7109	26416944
0.0	14.7100	31200	0.7532	26586032
0.0	14.8043	31400	0.7551	26756496
0.0	14.8986	31600	0.7506	26924320
0.0	14.9929	31800	0.7554	27096688
0.0	15.0872	32000	0.7591	27264640
0.0	15.1815	32200	0.7616	27440592
0.0	15.2758	32400	0.7617	27613248
0.0	15.3701	32600	0.7620	27781552
0.0	15.4644	32800	0.7669	27956496
0.0	15.5587	33000	0.7672	28125456
0.0	15.6530	33200	0.7705	28295136
0.0	15.7473	33400	0.7677	28462640
0.0	15.8416	33600	0.7741	28631360
0.0	15.9359	33800	0.7735	28799488
0.0	16.0302	34000	0.7718	28964832
0.0	16.1245	34200	0.7772	29137792
0.0	16.2188	34400	0.7773	29306192
0.0	16.3131	34600	0.7795	29481760
0.0	16.4074	34800	0.7860	29654256
0.0	16.5017	35000	0.7843	29821840
0.0	16.5959	35200	0.7870	29992016
0.0	16.6902	35400	0.7885	30157952
0.0	16.7845	35600	0.7890	30329792
0.0	16.8788	35800	0.7849	30500240
0.0	16.9731	36000	0.7907	30668944
0.0	17.0674	36200	0.7922	30840688
0.0	17.1617	36400	0.7952	31012176
0.0	17.2560	36600	0.7901	31184160
0.0	17.3503	36800	0.7919	31359648
0.0	17.4446	37000	0.7899	31529872
0.0	17.5389	37200	0.7934	31699104
0.0	17.6332	37400	0.7909	31870016
0.0	17.7275	37600	0.7962	32036672
0.0	17.8218	37800	0.7975	32206048
0.0	17.9161	38000	0.7980	32377104
0.0	18.0104	38200	0.7977	32548208
0.0	18.1047	38400	0.8014	32716560
0.0	18.1990	38600	0.7976	32885504
0.0	18.2933	38800	0.7980	33056256
0.0	18.3876	39000	0.7969	33225648
0.0	18.4818	39200	0.8011	33393952
0.0	18.5761	39400	0.8007	33564304
0.0	18.6704	39600	0.8016	33735024
0.0	18.7647	39800	0.8004	33907088
0.0	18.8590	40000	0.7987	34078960

Framework versions

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