exceptions_exp2_swap_0.7_resemble_to_hit_3591

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: 0.3687

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: 3591
• 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
4.8474	0.2915	1000	4.7781	0.2517
4.3601	0.5831	2000	4.2994	0.2976
4.1592	0.8746	3000	4.1059	0.3143
4.0079	1.1662	4000	4.0015	0.3243
3.9411	1.4577	5000	3.9215	0.3309
3.8798	1.7493	6000	3.8664	0.3358
3.7565	2.0408	7000	3.8226	0.3403
3.7471	2.3324	8000	3.7919	0.3431
3.7565	2.6239	9000	3.7590	0.3463
3.7274	2.9155	10000	3.7342	0.3489
3.6486	3.2070	11000	3.7218	0.3505
3.6498	3.4985	12000	3.7029	0.3519
3.6509	3.7901	13000	3.6846	0.3539
3.5443	4.0816	14000	3.6775	0.3551
3.5816	4.3732	15000	3.6670	0.3563
3.5931	4.6647	16000	3.6545	0.3570
3.5863	4.9563	17000	3.6412	0.3585
3.524	5.2478	18000	3.6425	0.3591
3.516	5.5394	19000	3.6323	0.3599
3.5351	5.8309	20000	3.6227	0.3608
3.4464	6.1224	21000	3.6252	0.3615
3.4841	6.4140	22000	3.6175	0.3618
3.5022	6.7055	23000	3.6078	0.3625
3.493	6.9971	24000	3.5993	0.3634
3.4368	7.2886	25000	3.6062	0.3635
3.4613	7.5802	26000	3.5971	0.3642
3.4725	7.8717	27000	3.5871	0.3650
3.3951	8.1633	28000	3.5948	0.3648
3.4278	8.4548	29000	3.5899	0.3652
3.4365	8.7464	30000	3.5803	0.3660
3.3323	9.0379	31000	3.5879	0.3661
3.3835	9.3294	32000	3.5886	0.3662
3.4017	9.6210	33000	3.5771	0.3668
3.4294	9.9125	34000	3.5684	0.3672
3.334	10.2041	35000	3.5815	0.3670
3.3771	10.4956	36000	3.5744	0.3677
3.3887	10.7872	37000	3.5681	0.3681
3.2991	11.0787	38000	3.5759	0.3676
3.3498	11.3703	39000	3.5710	0.3680
3.3671	11.6618	40000	3.5665	0.3687
3.375	11.9534	41000	3.5586	0.3691
3.3224	12.2449	42000	3.5697	0.3687
3.3425	12.5364	43000	3.5624	0.3692
3.3555	12.8280	44000	3.5539	0.3696
3.2828	13.1195	45000	3.5679	0.3693
3.3191	13.4111	46000	3.5637	0.3696
3.3406	13.7026	47000	3.5533	0.3700
3.3622	13.9942	48000	3.5463	0.3705
3.277	14.2857	49000	3.5618	0.3698
3.3235	14.5773	50000	3.5577	0.3701
3.3302	14.8688	51000	3.5479	0.3705
3.259	15.1603	52000	3.5620	0.3701
3.2934	15.4519	53000	3.5579	0.3705
3.3156	15.7434	54000	3.5521	0.3707
3.2151	16.0350	55000	3.5630	0.3704
3.2638	16.3265	56000	3.5580	0.3708
3.2972	16.6181	57000	3.5499	0.3712
3.3058	16.9096	58000	3.5419	0.3714
3.231	17.2012	59000	3.5595	0.3711
3.2717	17.4927	60000	3.5527	0.3712
3.2823	17.7843	61000	3.5432	0.3719
3.2039	18.0758	62000	3.5580	0.3713
3.2459	18.3673	63000	3.5557	0.3714
3.2632	18.6589	64000	3.5453	0.3718
3.2792	18.9504	65000	3.5387	0.3725
3.2232	19.2420	66000	3.5540	0.3719
3.254	19.5335	67000	3.5492	0.3722
3.2682	19.8251	68000	3.5415	0.3725
3.202	20.1166	69000	3.5595	0.3720
3.2338	20.4082	70000	3.5539	0.3721
3.2566	20.6997	71000	3.5413	0.3726
3.2621	20.9913	72000	3.5376	0.3730
3.2017	21.2828	73000	3.5579	0.3719
3.2353	21.5743	74000	3.5456	0.3727
3.2591	21.8659	75000	3.5399	0.3731
3.1687	22.1574	76000	3.5549	0.3724
3.2161	22.4490	77000	3.5488	0.3728
3.2278	22.7405	78000	3.5401	0.3733
3.1324	23.0321	79000	3.5554	0.3727
3.1933	23.3236	80000	3.5547	0.3729
3.22	23.6152	81000	3.5437	0.3733
3.2157	23.9067	82000	3.5392	0.3736
3.178	24.1983	83000	3.5546	0.3726
3.2107	24.4898	84000	3.5486	0.3729
3.2139	24.7813	85000	3.5459	0.3732
3.1336	25.0729	86000	3.5556	0.3727
3.1801	25.3644	87000	3.5552	0.3728
3.1951	25.6560	88000	3.5485	0.3735
3.2053	25.9475	89000	3.5388	0.3741
3.1536	26.2391	90000	3.5538	0.3733
3.1865	26.5306	91000	3.5487	0.3735
3.1955	26.8222	92000	3.5392	0.3742

Framework versions

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