exceptions_exp2_swap_0.7_last_to_push_2128

This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 3.5665
• Accuracy: 0.3685

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: 0.0006
• 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	Accuracy	Validation Loss
4.8405	0.2915	1000	0.2538	4.7545
4.3421	0.5830	2000	0.2983	4.2930
4.1511	0.8745	3000	0.3143	4.1041
4.0075	1.1659	4000	0.3236	4.0015
3.9398	1.4574	5000	0.3304	3.9234
3.8836	1.7489	6000	0.3355	3.8672
3.7667	2.0402	7000	0.3395	3.8254
3.7647	2.3317	8000	0.3430	3.7946
3.7476	2.6233	9000	0.3456	3.7627
3.7253	2.9148	10000	0.3480	3.7392
3.6522	3.2061	11000	0.3497	3.7253
3.66	3.4976	12000	0.3516	3.7062
3.6552	3.7891	13000	0.3531	3.6898
3.5443	4.0805	14000	0.3548	3.6803
3.5694	4.3720	15000	0.3561	3.6704
3.5998	4.6635	16000	0.3568	3.6596
3.5871	4.9550	17000	0.3585	3.6424
3.4993	5.2463	18000	0.3590	3.6442
3.5276	5.5378	19000	0.3596	3.6372
3.5397	5.8293	20000	0.3607	3.6235
3.4575	6.1207	21000	0.3610	3.6287
3.4711	6.4122	22000	0.3618	3.6190
3.4998	6.7037	23000	0.3626	3.6091
3.5047	6.9952	24000	0.3634	3.6004
3.4571	7.2866	25000	0.3633	3.6089
3.4564	7.5781	26000	0.3641	3.6010
3.4619	7.8696	27000	0.3645	3.5920
3.3915	8.1609	28000	0.3646	3.5990
3.4238	8.4524	29000	0.3648	3.5954
3.433	8.7439	30000	0.3657	3.5863
3.3406	9.0353	31000	0.3655	3.5907
3.3767	9.3268	32000	0.3661	3.5872
3.4124	9.6183	33000	0.3664	3.5781
3.4173	9.9098	34000	0.3671	3.5719
3.353	10.2011	35000	0.3668	3.5851
3.386	10.4927	36000	0.3673	3.5757
3.3967	10.7842	37000	0.3681	3.5682
3.3008	11.0755	38000	0.3674	3.5771
3.3421	11.3670	39000	0.3676	3.5748
3.3865	11.6585	40000	0.3685	3.5665
3.3826	11.9500	41000	0.3691	3.5591
3.3125	12.2414	42000	0.3687	3.5711
3.3482	12.5329	43000	0.3687	3.5677
3.3548	12.8244	44000	0.3694	3.5561
3.2723	13.1157	45000	0.3688	3.5734
3.3128	13.4072	46000	0.3696	3.5666
3.3554	13.6988	47000	0.3700	3.5555
3.3418	13.9903	48000	0.3703	3.5509
3.2873	14.2816	49000	0.3699	3.5642
3.3214	14.5731	50000	0.3699	3.5598
3.3372	14.8646	51000	0.3704	3.5496
3.2634	15.1560	52000	0.3700	3.5655
3.3005	15.4475	53000	0.3703	3.5598
3.3022	15.7390	54000	0.3707	3.5546
3.2172	16.0303	55000	0.3706	3.5590
3.2679	16.3218	56000	0.3707	3.5607
3.2927	16.6133	57000	0.3713	3.5518
3.3154	16.9049	58000	0.3717	3.5446
3.2442	17.1962	59000	0.3711	3.5609
3.2761	17.4877	60000	0.3714	3.5523
3.291	17.7792	61000	0.3719	3.5461
3.1997	18.0705	62000	0.3712	3.5632
3.2501	18.3621	63000	0.3716	3.5541
3.2619	18.6536	64000	0.3718	3.5468
3.2919	18.9451	65000	0.3722	3.5423
3.2264	19.2364	66000	0.3715	3.5561
3.2394	19.5279	67000	0.3720	3.5495
3.2665	19.8194	68000	0.3725	3.5456
3.1939	20.1108	69000	0.3716	3.5588
3.2149	20.4023	70000	0.3719	3.5556
3.2447	20.6938	71000	0.3722	3.5452
3.2733	20.9853	72000	0.3730	3.5387
3.2131	21.2766	73000	0.3721	3.5581
3.2312	21.5682	74000	0.3727	3.5495
3.2569	21.8597	75000	0.3731	3.5411
3.1789	22.1510	76000	0.3719	3.5589
3.202	22.4425	77000	0.3724	3.5496
3.2272	22.7340	78000	0.3730	3.5428
3.1514	23.0254	79000	0.3726	3.5549
3.1818	23.3169	80000	0.3724	3.5531
3.1943	23.6084	81000	3.5582	0.3727
3.1961	23.8999	82000	3.5507	0.3727
3.1603	24.1915	83000	3.5631	0.3723
3.1921	24.4830	84000	3.5538	0.3728
3.2218	24.7745	85000	3.5425	0.3734
3.1492	25.0659	86000	3.5576	0.3726
3.1824	25.3574	87000	3.5505	0.3731
3.201	25.6489	88000	3.5468	0.3735
3.2114	25.9404	89000	3.5391	0.3738
3.1724	26.2318	90000	3.5586	0.3729
3.1831	26.5233	91000	3.5528	0.3733
3.2083	26.8148	92000	3.5445	0.3736
3.1286	27.1061	93000	3.5576	0.3732
3.1577	27.3976	94000	3.5532	0.3733
3.1856	27.6891	95000	3.5450	0.3739
3.211	27.9806	96000	3.5370	0.3744
3.1377	28.2720	97000	3.5558	0.3734
3.1703	28.5635	98000	3.5500	0.3738
3.1761	28.8550	99000	3.5446	0.3739
3.1259	29.1463	100000	3.5599	0.3731
3.1412	29.4378	101000	3.5544	0.3736
3.1537	29.7294	102000	3.5457	0.3741
3.0842	30.0207	103000	3.5555	0.3737
3.1355	30.3122	104000	3.5562	0.3735
3.1478	30.6037	105000	3.5494	0.3740
3.166	30.8952	106000	3.5430	0.3743
3.113	31.1866	107000	3.5579	0.3739
3.1249	31.4781	108000	3.5542	0.3737
3.1627	31.7696	109000	3.5448	0.3745
3.0785	32.0609	110000	3.5565	0.3738
3.1107	32.3524	111000	3.5553	0.3741
3.1327	32.6439	112000	3.5484	0.3742
3.1514	32.9355	113000	3.5464	0.3744
3.1022	33.2268	114000	3.5567	0.3740
3.1056	33.5183	115000	3.5551	0.3739
3.1411	33.8098	116000	3.5420	0.3748

Framework versions

• Transformers 4.55.2
• Pytorch 2.8.0+cu128
• Datasets 4.0.0
• Tokenizers 0.21.4