train_boolq_1745950278

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: 0.4582
• 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
0.4557	0.0943	200	0.4984	171472
0.3184	0.1886	400	0.4857	339520
0.4821	0.2829	600	0.4770	509632
0.6649	0.3772	800	0.4744	685120
0.1655	0.4715	1000	0.4742	856144
0.7329	0.5658	1200	0.4678	1024448
0.2518	0.6601	1400	0.4703	1192560
0.5007	0.7544	1600	0.4664	1360304
0.4601	0.8487	1800	0.4668	1535088
0.3674	0.9430	2000	0.4674	1708224
0.2673	1.0372	2200	0.4673	1880336
0.6585	1.1315	2400	0.4638	2048144
0.8909	1.2258	2600	0.4659	2220048
0.2237	1.3201	2800	0.4659	2388128
0.2765	1.4144	3000	0.4645	2559696
0.3239	1.5087	3200	0.4640	2730224
0.3023	1.6030	3400	0.4650	2897744
0.3181	1.6973	3600	0.4661	3067728
0.4192	1.7916	3800	0.4628	3235856
0.6724	1.8859	4000	0.4614	3409536
0.2948	1.9802	4200	0.4632	3581600
0.8057	2.0745	4400	0.4633	3753552
0.5526	2.1688	4600	0.4639	3924224
0.3203	2.2631	4800	0.4626	4093360
0.6014	2.3574	5000	0.4661	4261312
0.9379	2.4517	5200	0.4648	4438224
0.6379	2.5460	5400	0.4632	4608992
0.2241	2.6403	5600	0.4623	4780976
0.4247	2.7346	5800	0.4626	4946848
0.3434	2.8289	6000	0.4623	5120704
0.6418	2.9231	6200	0.4622	5292816
0.3682	3.0174	6400	0.4642	5463728
0.7239	3.1117	6600	0.4632	5634528
0.3848	3.2060	6800	0.4631	5805344
0.7214	3.3003	7000	0.4591	5976160
0.2775	3.3946	7200	0.4626	6147200
0.252	3.4889	7400	0.4627	6315984
0.388	3.5832	7600	0.4622	6484512
0.5885	3.6775	7800	0.4638	6653536
0.5795	3.7718	8000	0.4646	6823184
0.2884	3.8661	8200	0.4637	6991232
0.2278	3.9604	8400	0.4651	7161488
0.727	4.0547	8600	0.4615	7330416
0.5219	4.1490	8800	0.4626	7502960
0.3851	4.2433	9000	0.4636	7675776
0.5972	4.3376	9200	0.4629	7847200
0.3878	4.4319	9400	0.4632	8016080
0.6179	4.5262	9600	0.4582	8189040
0.4354	4.6205	9800	0.4645	8355360
0.2836	4.7148	10000	0.4603	8527824
0.3939	4.8091	10200	0.4631	8697376
0.3165	4.9033	10400	0.4635	8867872
0.3666	4.9976	10600	0.4620	9039888
0.5206	5.0919	10800	0.4621	9209232
0.498	5.1862	11000	0.4610	9384064
0.9534	5.2805	11200	0.4624	9555168
0.592	5.3748	11400	0.4621	9724832
0.1457	5.4691	11600	0.4631	9894608
0.3453	5.5634	11800	0.4642	10067376
0.4376	5.6577	12000	0.4625	10239648
0.6131	5.7520	12200	0.4611	10406400
0.6455	5.8463	12400	0.4635	10578176
0.4769	5.9406	12600	0.4617	10745200
0.2867	6.0349	12800	0.4614	10917056
0.2948	6.1292	13000	0.4625	11091376
0.3291	6.2235	13200	0.4619	11260160
0.4555	6.3178	13400	0.4627	11430512
0.5634	6.4121	13600	0.4641	11598992
0.5669	6.5064	13800	0.4617	11771312
0.6347	6.6007	14000	0.4616	11940256
0.2713	6.6950	14200	0.4621	12108896
0.7112	6.7893	14400	0.4642	12277824
0.6397	6.8835	14600	0.4648	12450224
0.3727	6.9778	14800	0.4654	12618544
0.4762	7.0721	15000	0.4638	12791104
0.8657	7.1664	15200	0.4635	12964976
0.553	7.2607	15400	0.4615	13132848
0.469	7.3550	15600	0.4648	13302528
0.5783	7.4493	15800	0.4605	13471696
0.669	7.5436	16000	0.4617	13643264
0.2304	7.6379	16200	0.4626	13810336
0.5105	7.7322	16400	0.4618	13980096
0.3843	7.8265	16600	0.4615	14150688
0.5705	7.9208	16800	0.4634	14320928
0.2881	8.0151	17000	0.4633	14497120
0.3272	8.1094	17200	0.4622	14667440
0.6435	8.2037	17400	0.4623	14839920
0.6967	8.2980	17600	0.4635	15013152
0.4326	8.3923	17800	0.4615	15178480
0.1335	8.4866	18000	0.4648	15349216
0.8222	8.5809	18200	0.4655	15518736
0.588	8.6752	18400	0.4618	15689568
0.6896	8.7694	18600	0.4611	15859632
0.4868	8.8637	18800	0.4645	16025920
0.3364	8.9580	19000	0.4641	16196560
0.2577	9.0523	19200	0.4617	16368144
0.2998	9.1466	19400	0.4624	16539952
0.2438	9.2409	19600	0.4632	16710384
0.4018	9.3352	19800	0.4629	16878624
0.4669	9.4295	20000	0.4601	17046992
0.536	9.5238	20200	0.4628	17218176
0.346	9.6181	20400	0.4652	17390384
0.6091	9.7124	20600	0.4649	17560512
0.5173	9.8067	20800	0.4618	17726048
0.3656	9.9010	21000	0.4649	17897376
0.3738	9.9953	21200	0.4632	18068096
0.4799	10.0896	21400	0.4635	18244704
0.4816	10.1839	21600	0.4618	18420464
0.4667	10.2782	21800	0.4636	18588480
0.2609	10.3725	22000	0.4619	18758992
0.3969	10.4668	22200	0.4612	18930688
0.444	10.5611	22400	0.4632	19096400
0.4167	10.6554	22600	0.4621	19263456
0.2856	10.7496	22800	0.4645	19430576
0.5746	10.8439	23000	0.4634	19599200
0.4028	10.9382	23200	0.4611	19770608
0.397	11.0325	23400	0.4633	19942144
1.079	11.1268	23600	0.4653	20112704
0.4449	11.2211	23800	0.4622	20282048
0.53	11.3154	24000	0.4628	20456384
0.2677	11.4097	24200	0.4596	20624672
0.4282	11.5040	24400	0.4603	20796640
0.2601	11.5983	24600	0.4639	20964288
0.3788	11.6926	24800	0.4618	21132896
0.6241	11.7869	25000	0.4627	21304080
0.4434	11.8812	25200	0.4627	21471264
0.5039	11.9755	25400	0.4653	21642432
0.8613	12.0698	25600	0.4636	21811200
1.1598	12.1641	25800	0.4654	21983552
0.5584	12.2584	26000	0.4646	22155824
0.8694	12.3527	26200	0.4675	22329984
0.2481	12.4470	26400	0.4657	22499472
0.3135	12.5413	26600	0.4616	22669904
0.5644	12.6355	26800	0.4649	22837440
0.5779	12.7298	27000	0.4621	23008384
0.7482	12.8241	27200	0.4633	23177264
0.5542	12.9184	27400	0.4646	23344128
0.6394	13.0127	27600	0.4654	23512272
0.6867	13.1070	27800	0.4623	23680080
0.3736	13.2013	28000	0.4623	23850944
0.3404	13.2956	28200	0.4623	24022624
0.8621	13.3899	28400	0.4623	24192144
0.3994	13.4842	28600	0.4623	24364768
0.4172	13.5785	28800	0.4623	24538352
0.3462	13.6728	29000	0.4646	24710416
0.7374	13.7671	29200	0.4646	24881936
0.2591	13.8614	29400	0.4646	25050656
0.3487	13.9557	29600	0.4646	25222752
0.6304	14.0500	29800	0.4646	25389344
0.3926	14.1443	30000	0.4646	25563712
0.4027	14.2386	30200	0.4646	25738992
0.597	14.3329	30400	0.4646	25909744
0.6683	14.4272	30600	0.4646	26079120
0.3154	14.5215	30800	0.4646	26245936
0.1877	14.6157	31000	0.4646	26416944
0.5388	14.7100	31200	0.4646	26586032
0.5175	14.8043	31400	0.4646	26756496
0.2724	14.8986	31600	0.4646	26924320
0.7219	14.9929	31800	0.4646	27096688
0.4891	15.0872	32000	0.4646	27264640
0.531	15.1815	32200	0.4646	27440592
0.7268	15.2758	32400	0.4646	27613248
0.8139	15.3701	32600	0.4646	27781552
0.4931	15.4644	32800	0.4646	27956496
0.3387	15.5587	33000	0.4646	28125456
0.2053	15.6530	33200	0.4646	28295136
0.3154	15.7473	33400	0.4646	28462640
0.3329	15.8416	33600	0.4646	28631360
0.706	15.9359	33800	0.4646	28799488
0.9518	16.0302	34000	0.4646	28964832
0.5958	16.1245	34200	0.4646	29137792
0.7452	16.2188	34400	0.4646	29306192
0.574	16.3131	34600	0.4646	29481760
0.6025	16.4074	34800	0.4646	29654256
0.9419	16.5017	35000	0.4646	29821840
0.3235	16.5959	35200	0.4646	29992016
0.42	16.6902	35400	0.4646	30157952
0.0703	16.7845	35600	0.4646	30329792
0.7282	16.8788	35800	0.4646	30500240
0.5623	16.9731	36000	0.4646	30668944
0.8665	17.0674	36200	0.4646	30840688
0.447	17.1617	36400	0.4646	31012176
0.8911	17.2560	36600	0.4646	31184160
0.3361	17.3503	36800	0.4646	31359648
0.4889	17.4446	37000	0.4646	31529872
0.3732	17.5389	37200	0.4646	31699104
1.0273	17.6332	37400	0.4646	31870016
0.4019	17.7275	37600	0.4646	32036672
0.3667	17.8218	37800	0.4646	32206048
0.2637	17.9161	38000	0.4646	32377104
0.3267	18.0104	38200	0.4646	32548208
0.4919	18.1047	38400	0.4646	32716560
0.5915	18.1990	38600	0.4646	32885504
0.8913	18.2933	38800	0.4646	33056256
0.4376	18.3876	39000	0.4646	33225648
0.5159	18.4818	39200	0.4646	33393952
0.5483	18.5761	39400	0.4646	33564304
0.4287	18.6704	39600	0.4646	33735024
0.2049	18.7647	39800	0.4646	33907088
0.2385	18.8590	40000	0.4646	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