learning_rate: 0.0006
train_batch_size: 16
eval_batch_size: 16
seed: 2128
gradient_accumulation_steps: 5
total_train_batch_size: 80
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.98) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 100
num_epochs: 50.0
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
4.8251	0.2911	1000	4.7547	0.2538
4.3285	0.5822	2000	4.2816	0.2998
4.146	0.8733	3000	4.0958	0.3156
3.9912	1.1642	4000	3.9907	0.3253
3.931	1.4553	5000	3.9154	0.3321
3.8782	1.7464	6000	3.8533	0.3375
3.7487	2.0373	7000	3.8109	0.3417
3.7602	2.3284	8000	3.7832	0.3448
3.7441	2.6195	9000	3.7549	0.3477
3.7275	2.9106	10000	3.7243	0.3500
3.6281	3.2014	11000	3.7152	0.3516
3.6473	3.4925	12000	3.6941	0.3537
3.6327	3.7837	13000	3.6770	0.3551
3.5308	4.0745	14000	3.6733	0.3565
3.5666	4.3656	15000	3.6601	0.3575
3.5771	4.6567	16000	3.6448	0.3587
3.5623	4.9478	17000	3.6332	0.3600
3.5009	5.2387	18000	3.6344	0.3606
3.5136	5.5298	19000	3.6242	0.3613
3.5273	5.8209	20000	3.6119	0.3621
3.4407	6.1118	21000	3.6195	0.3625
3.4662	6.4029	22000	3.6101	0.3635
3.4862	6.6940	23000	3.5993	0.3642
3.4945	6.9851	24000	3.5902	0.3650
3.4247	7.2760	25000	3.6004	0.3647
3.4431	7.5671	26000	3.5893	0.3658
3.4692	7.8582	27000	3.5804	0.3662
3.3743	8.1490	28000	3.5928	0.3660
3.4148	8.4401	29000	3.5866	0.3668
3.4272	8.7313	30000	3.5738	0.3671
3.3246	9.0221	31000	3.5837	0.3671
3.376	9.3132	32000	3.5770	0.3675
3.391	9.6043	33000	3.5704	0.3679
3.4221	9.8954	34000	3.5613	0.3687
3.3295	10.1863	35000	3.5744	0.3682
3.365	10.4774	36000	3.5670	0.3685
3.387	10.7685	37000	3.5599	0.3692
3.2772	11.0594	38000	3.5699	0.3692
3.3402	11.3505	39000	3.5686	0.3692
3.3618	11.6416	40000	3.5564	0.3697
3.3757	11.9327	41000	3.5474	0.3703
3.3099	12.2236	42000	3.5660	0.3696
3.3302	12.5147	43000	3.5534	0.3701
3.3526	12.8058	44000	3.5516	0.3707
3.2625	13.0966	45000	3.5597	0.3703
3.3032	13.3878	46000	3.5549	0.3707
3.3322	13.6789	47000	3.5511	0.3712
3.3339	13.9700	48000	3.5426	0.3714
3.271	14.2608	49000	3.5618	0.3707
3.3065	14.5519	50000	3.5518	0.3713
3.3203	14.8430	51000	3.5447	0.3718
3.2281	15.1339	52000	3.5599	0.3712
3.2869	15.4250	53000	3.5515	0.3713
3.2894	15.7161	54000	3.5417	0.3721
3.26	16.0070	55000	3.5530	0.3716
3.2488	16.2981	56000	3.5516	0.3720
3.2734	16.5892	57000	3.5446	0.3726
3.3024	16.8803	58000	3.5364	0.3726
3.2226	17.1712	59000	3.5517	0.3722
3.2674	17.4623	60000	3.5471	0.3723
3.2954	17.7534	61000	3.5406	0.3729
3.1943	18.0442	62000	3.5554	0.3722
3.2312	18.3354	63000	3.5506	0.3725
3.259	18.6265	64000	3.5441	0.3727
3.2747	18.9176	65000	3.5354	0.3737
3.2114	19.2084	66000	3.5522	0.3725
3.2536	19.4995	67000	3.5453	0.3730
3.2448	19.7906	68000	3.5363	0.3734
3.1741	20.0815	69000	3.5539	0.3727
3.215	20.3726	70000	3.5489	0.3729
3.2384	20.6637	71000	3.5401	0.3736
3.2507	20.9548	72000	3.5342	0.3740
3.1906	21.2457	73000	3.5473	0.3732
3.2117	21.5368	74000	3.5460	0.3737
3.2276	21.8279	75000	3.5347	0.3743
3.1557	22.1188	76000	3.5557	0.3731
3.1989	22.4099	77000	3.5477	0.3736
3.2202	22.7010	78000	3.5414	0.3742
3.2437	22.9921	79000	3.5309	0.3746
3.1818	23.2830	80000	3.5501	0.3737
3.1982	23.5741	81000	3.5418	0.3740
3.2273	23.8652	82000	3.5339	0.3744
3.152	24.1560	83000	3.5505	0.3740
3.1798	24.4471	84000	3.5480	0.3737
3.2121	24.7382	85000	3.5390	0.3742
3.1196	25.0291	86000	3.5522	0.3738
3.1639	25.3202	87000	3.5503	0.3738
3.181	25.6113	88000	3.5432	0.3742
3.2067	25.9024	89000	3.5347	0.3749
3.1334	26.1933	90000	3.5539	0.3738
3.1573	26.4844	91000	3.5476	0.3744
3.1951	26.7755	92000	3.5388	0.3750
3.109	27.0664	93000	3.5516	0.3738
3.1523	27.3575	94000	3.5495	0.3745
3.1663	27.6486	95000	3.5406	0.3746
3.1853	27.9397	96000	3.5330	0.3752
3.1239	28.2306	97000	3.5534	0.3743
3.1455	28.5217	98000	3.5473	0.3749
3.1642	28.8128	99000	3.5373	0.3751

Framework versions

Transformers 4.55.2
Pytorch 2.8.0+cu128
Datasets 4.0.0
Tokenizers 0.21.4

craa/exceptions_exp2_swap_take_to_push_2128

exceptions_exp2_swap_take_to_push_2128

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

craa/
exceptions_exp2_swap_take_to_push_2128