exceptions_exp2_swap_take_to_hit_1032

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

• Loss: 3.5585
• Accuracy: 0.3695

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: 1032
• 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
4.836	0.2911	1000	4.7612	0.2540
4.3431	0.5822	2000	4.2827	0.2996
4.1517	0.8733	3000	4.0997	0.3154
3.9827	1.1642	4000	3.9901	0.3251
3.9243	1.4553	5000	3.9174	0.3314
3.8802	1.7464	6000	3.8595	0.3370
3.7485	2.0373	7000	3.8150	0.3413
3.7568	2.3284	8000	3.7829	0.3444
3.7395	2.6195	9000	3.7568	0.3470
3.7362	2.9106	10000	3.7277	0.3495
3.6397	3.2014	11000	3.7165	0.3513
3.6558	3.4925	12000	3.6975	0.3531
3.6503	3.7837	13000	3.6787	0.3550
3.5379	4.0745	14000	3.6711	0.3562
3.5792	4.3656	15000	3.6623	0.3570
3.5722	4.6567	16000	3.6485	0.3581
3.5834	4.9478	17000	3.6356	0.3594
3.4952	5.2387	18000	3.6365	0.3598
3.5208	5.5298	19000	3.6278	0.3611
3.5306	5.8209	20000	3.6158	0.3618
3.4547	6.1118	21000	3.6218	0.3623
3.4806	6.4029	22000	3.6116	0.3630
3.4883	6.6940	23000	3.6036	0.3637
3.5017	6.9851	24000	3.5934	0.3645
3.4176	7.2760	25000	3.5992	0.3645
3.4489	7.5671	26000	3.5920	0.3651
3.4756	7.8582	27000	3.5824	0.3661
3.3851	8.1490	28000	3.5897	0.3659
3.4082	8.4401	29000	3.5872	0.3661
3.4407	8.7313	30000	3.5751	0.3670
3.3182	9.0221	31000	3.5826	0.3672
3.3674	9.3132	32000	3.5801	0.3674
3.3945	9.6043	33000	3.5724	0.3677
3.4257	9.8954	34000	3.5622	0.3684
3.3368	10.1863	35000	3.5776	0.3681
3.3671	10.4774	36000	3.5694	0.3684
3.3756	10.7685	37000	3.5619	0.3691
3.2966	11.0594	38000	3.5685	0.3687
3.3352	11.3505	39000	3.5673	0.3691
3.3615	11.6416	40000	3.5585	0.3695
3.3579	11.9327	41000	3.5503	0.3702
3.301	12.2236	42000	3.5659	0.3698
3.3297	12.5147	43000	3.5582	0.3702
3.3465	12.8058	44000	3.5466	0.3709
3.27	13.0966	45000	3.5600	0.3702
3.3062	13.3878	46000	3.5603	0.3704
3.3203	13.6789	47000	3.5480	0.3709
3.3477	13.9700	48000	3.5416	0.3716
3.2778	14.2608	49000	3.5573	0.3708
3.3029	14.5519	50000	3.5524	0.3714
3.3323	14.8430	51000	3.5412	0.3721
3.2466	15.1339	52000	3.5578	0.3712
3.2753	15.4250	53000	3.5506	0.3717
3.2864	15.7161	54000	3.5427	0.3722
3.2584	16.0070	55000	3.5479	0.3718
3.2505	16.2981	56000	3.5531	0.3717
3.2848	16.5892	57000	3.5473	0.3723
3.2979	16.8803	58000	3.5364	0.3728
3.2323	17.1712	59000	3.5527	0.3724
3.2628	17.4623	60000	3.5477	0.3725
3.2792	17.7534	61000	3.5385	0.3727
3.1838	18.0442	62000	3.5499	0.3724
3.2446	18.3354	63000	3.5453	0.3727
3.2673	18.6265	64000	3.5376	0.3730
3.2653	18.9176	65000	3.5342	0.3734
3.2098	19.2084	66000	3.5494	0.3727
3.2505	19.4995	67000	3.5428	0.3731
3.2671	19.7906	68000	3.5347	0.3737
3.1706	20.0815	69000	3.5506	0.3729
3.2117	20.3726	70000	3.5454	0.3733
3.2439	20.6637	71000	3.5393	0.3737
3.2467	20.9548	72000	3.5354	0.3738
3.2	21.2457	73000	3.5522	0.3731
3.2198	21.5368	74000	3.5407	0.3737
3.229	21.8279	75000	3.5321	0.3742
3.1638	22.1188	76000	3.5507	0.3732
3.2052	22.4099	77000	3.5449	0.3736
3.2189	22.7010	78000	3.5340	0.3742
3.2276	22.9921	79000	3.5306	0.3745
3.1931	23.2830	80000	3.5466	0.3737
3.2036	23.5741	81000	3.5391	0.3742
3.2033	23.8652	82000	3.5328	0.3747
3.1472	24.1560	83000	3.5504	0.3736
3.1926	24.4471	84000	3.5439	0.3740
3.2035	24.7382	85000	3.5340	0.3746
3.1209	25.0291	86000	3.5451	0.3744
3.167	25.3202	87000	3.5455	0.3745
3.1906	25.6113	88000	3.5412	0.3742
3.2065	25.9024	89000	3.5326	0.3747
3.1426	26.1933	90000	3.5499	0.3739
3.1732	26.4844	91000	3.5390	0.3746
3.1891	26.7755	92000	3.5367	0.3749
3.1057	27.0664	93000	3.5515	0.3743
3.16	27.3575	94000	3.5443	0.3744
3.1697	27.6486	95000	3.5377	0.3751
3.1803	27.9397	96000	3.5311	0.3753
3.1199	28.2306	97000	3.5479	0.3747
3.1493	28.5217	98000	3.5434	0.3746
3.1746	28.8128	99000	3.5322	0.3756

Framework versions

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