exceptions_exp2_swap_take_to_drop_5039

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

• Loss: 3.5560
• Accuracy: 0.3700

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: 5039
• 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
4.822	0.2911	1000	0.2557	4.7456
4.3327	0.5822	2000	0.2994	4.2819
4.1442	0.8733	3000	0.3152	4.0968
3.988	1.1642	4000	0.3252	3.9879
3.9266	1.4553	5000	0.3323	3.9127
3.8707	1.7464	6000	0.3372	3.8542
3.748	2.0373	7000	0.3417	3.8133
3.7503	2.3284	8000	0.3444	3.7823
3.7339	2.6195	9000	0.3474	3.7529
3.7242	2.9106	10000	0.3499	3.7271
3.6359	3.2014	11000	0.3520	3.7129
3.6477	3.4925	12000	0.3536	3.6930
3.6404	3.7837	13000	0.3551	3.6773
3.5411	4.0745	14000	0.3562	3.6695
3.5634	4.3656	15000	0.3576	3.6564
3.5771	4.6567	16000	0.3588	3.6422
3.5833	4.9478	17000	0.3600	3.6300
3.5064	5.2387	18000	0.3604	3.6340
3.5224	5.5298	19000	0.3613	3.6236
3.5265	5.8209	20000	0.3622	3.6129
3.4555	6.1121	21000	0.3619	3.6225
3.4688	6.4032	22000	0.3631	3.6104
3.4815	6.6943	23000	0.3641	3.5999
3.4938	6.9854	24000	0.3646	3.5898
3.412	7.2763	25000	0.3649	3.5982
3.4507	7.5674	26000	0.3651	3.5910
3.4643	7.8585	27000	0.3662	3.5809
3.3746	8.1493	28000	0.3662	3.5871
3.4129	8.4404	29000	0.3666	3.5818
3.4204	8.7315	30000	0.3672	3.5708
3.3191	9.0224	31000	0.3674	3.5789
3.3652	9.3135	32000	0.3676	3.5797
3.3979	9.6046	33000	0.3681	3.5693
3.42	9.8957	34000	0.3686	3.5598
3.3209	10.1866	35000	0.3682	3.5719
3.3613	10.4777	36000	0.3690	3.5639
3.3789	10.7688	37000	0.3695	3.5575
3.2783	11.0597	38000	0.3694	3.5668
3.3382	11.3508	39000	0.3694	3.5635
3.3576	11.6419	40000	0.3700	3.5560
3.3676	11.9330	41000	0.3707	3.5506
3.3003	12.2239	42000	0.3700	3.5619
3.3319	12.5150	43000	0.3706	3.5557
3.3495	12.8061	44000	0.3709	3.5473
3.2541	13.0969	45000	0.3704	3.5610
3.2963	13.3880	46000	0.3710	3.5569
3.3284	13.6791	47000	0.3709	3.5485
3.3499	13.9702	48000	0.3718	3.5377
3.2809	14.2611	49000	0.3710	3.5558
3.3139	14.5522	50000	0.3718	3.5462
3.3249	14.8433	51000	0.3722	3.5382
3.2315	15.1342	52000	0.3717	3.5541
3.2842	15.4253	53000	0.3721	3.5459
3.2969	15.7164	54000	0.3727	3.5421
3.2481	16.0073	55000	0.3721	3.5455
3.2577	16.2984	56000	0.3721	3.5497
3.2754	16.5895	57000	0.3726	3.5411
3.2806	16.8806	58000	0.3731	3.5362
3.2295	17.1715	59000	0.3723	3.5508
3.2494	17.4626	60000	0.3727	3.5434
3.2762	17.7537	61000	0.3732	3.5344
3.1888	18.0445	62000	0.3730	3.5466
3.2432	18.3356	63000	0.3728	3.5458
3.2445	18.6267	64000	0.3735	3.5376
3.274	18.9179	65000	0.3737	3.5312
3.2197	19.2087	66000	0.3731	3.5478
3.2441	19.4998	67000	0.3735	3.5384
3.2563	19.7909	68000	0.3740	3.5328
3.1666	20.0818	69000	0.3736	3.5489
3.2238	20.3729	70000	0.3738	3.5431
3.2472	20.6640	71000	0.3740	3.5372
3.2567	20.9551	72000	0.3745	3.5299
3.1901	21.2460	73000	0.3735	3.5467
3.2194	21.5371	74000	0.3741	3.5403
3.2313	21.8282	75000	0.3745	3.5312
3.1707	22.1191	76000	0.3735	3.5494
3.1912	22.4102	77000	0.3743	3.5406
3.223	22.7013	78000	0.3744	3.5362
3.2529	22.9924	79000	0.3747	3.5275
3.1893	23.2832	80000	0.3740	3.5449
3.2118	23.5743	81000	0.3744	3.5361
3.2235	23.8655	82000	0.3747	3.5327
3.1507	24.1563	83000	0.3742	3.5461
3.1847	24.4474	84000	0.3746	3.5410
3.2091	24.7385	85000	0.3749	3.5348
3.1147	25.0294	86000	0.3741	3.5474
3.1497	25.3205	87000	0.3742	3.5473
3.1849	25.6116	88000	0.3749	3.5387
3.1967	25.9027	89000	0.3755	3.5313
3.1474	26.1936	90000	0.3743	3.5463
3.1539	26.4844	91000	3.5466	0.3746
3.1673	26.7755	92000	3.5414	0.3748
3.1	27.0667	93000	3.5534	0.3743
3.1525	27.3578	94000	3.5462	0.3746
3.164	27.6489	95000	3.5375	0.3750
3.1906	27.9400	96000	3.5308	0.3755
3.1249	28.2308	97000	3.5490	0.3750
3.142	28.5219	98000	3.5417	0.3752
3.1525	28.8131	99000	3.5335	0.3756
3.0956	29.1039	100000	3.5490	0.3748
3.1377	29.3950	101000	3.5450	0.3748
3.1595	29.6861	102000	3.5364	0.3756
3.1699	29.9772	103000	3.5272	0.3761
3.1199	30.2681	104000	3.5475	0.3749
3.1472	30.5592	105000	3.5370	0.3758
3.1579	30.8503	106000	3.5318	0.3758
3.0871	31.1412	107000	3.5485	0.3752
3.1203	31.4323	108000	3.5436	0.3753
3.1341	31.7234	109000	3.5362	0.3759
3.0667	32.0143	110000	3.5461	0.3755
3.0948	32.3054	111000	3.5434	0.3756
3.1198	32.5965	112000	3.5402	0.3758
3.146	32.8876	113000	3.5340	0.3764
3.0662	33.1784	114000	3.5494	0.3756
3.1008	33.4696	115000	3.5428	0.3759
3.1246	33.7607	116000	3.5370	0.3759
3.05	34.0515	117000	3.5519	0.3754
3.0861	34.3426	118000	3.5441	0.3756
3.0992	34.6337	119000	3.5400	0.3760
3.1145	34.9248	120000	3.5341	0.3765
3.0726	35.2157	121000	3.5509	0.3756
3.1016	35.5068	122000	3.5436	0.3759
3.124	35.7979	123000	3.5352	0.3763

Framework versions

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