exceptions_exp2_swap_0.3_cost_to_push_40817

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

• Loss: 3.5844
• Accuracy: 0.3658

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: 40817
• 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
4.8265	0.2915	1000	4.7497	0.2549
4.3273	0.5831	2000	4.2942	0.2980
4.1605	0.8746	3000	4.1071	0.3145
4.0091	1.1662	4000	4.0002	0.3236
3.9331	1.4577	5000	3.9235	0.3305
3.8914	1.7493	6000	3.8642	0.3358
3.7592	2.0408	7000	3.8235	0.3400
3.752	2.3324	8000	3.7920	0.3430
3.7562	2.6239	9000	3.7604	0.3460
3.7198	2.9155	10000	3.7376	0.3485
3.6419	3.2070	11000	3.7253	0.3502
3.6493	3.4985	12000	3.7058	0.3522
3.6618	3.7901	13000	3.6882	0.3537
3.5637	4.0816	14000	3.6798	0.3548
3.5681	4.3732	15000	3.6698	0.3559
3.584	4.6647	16000	3.6565	0.3571
3.5892	4.9563	17000	3.6440	0.3583
3.4966	5.2478	18000	3.6450	0.3584
3.5194	5.5394	19000	3.6336	0.3597
3.5266	5.8309	20000	3.6241	0.3608
3.456	6.1224	21000	3.6266	0.3611
3.4867	6.4140	22000	3.6202	0.3618
3.5035	6.7055	23000	3.6086	0.3625
3.5093	6.9971	24000	3.5978	0.3633
3.443	7.2886	25000	3.6092	0.3631
3.4494	7.5802	26000	3.5988	0.3638
3.4777	7.8717	27000	3.5891	0.3648
3.399	8.1633	28000	3.6002	0.3648
3.4319	8.4548	29000	3.5918	0.3655
3.4441	8.7464	30000	3.5844	0.3658
3.3316	9.0379	31000	3.5888	0.3661
3.3925	9.3294	32000	3.5875	0.3661
3.402	9.6210	33000	3.5771	0.3669
3.4104	9.9125	34000	3.5731	0.3671
3.344	10.2041	35000	3.5823	0.3669
3.3809	10.4956	36000	3.5759	0.3674
3.3944	10.7872	37000	3.5648	0.3678
3.3132	11.0787	38000	3.5779	0.3676
3.3508	11.3703	39000	3.5730	0.3678
3.3667	11.6618	40000	3.5658	0.3685
3.3751	11.9534	41000	3.5589	0.3689
3.3221	12.2449	42000	3.5716	0.3684
3.3405	12.5364	43000	3.5667	0.3690
3.3628	12.8280	44000	3.5534	0.3695
3.2828	13.1195	45000	3.5728	0.3689
3.3047	13.4111	46000	3.5630	0.3695
3.3289	13.7026	47000	3.5559	0.3700
3.3505	13.9942	48000	3.5535	0.3701
3.2862	14.2857	49000	3.5620	0.3699
3.3113	14.5773	50000	3.5576	0.3702
3.3271	14.8688	51000	3.5523	0.3706
3.2536	15.1603	52000	3.5683	0.3697
3.2932	15.4519	53000	3.5616	0.3704
3.3196	15.7434	54000	3.5499	0.3710
3.2222	16.0350	55000	3.5590	0.3705
3.2633	16.3265	56000	3.5618	0.3708
3.2935	16.6181	57000	3.5542	0.3711
3.2938	16.9096	58000	3.5442	0.3718
3.2302	17.2012	59000	3.5634	0.3708
3.2746	17.4927	60000	3.5547	0.3713
3.2864	17.7843	61000	3.5471	0.3717
3.1957	18.0758	62000	3.5631	0.3713
3.2546	18.3673	63000	3.5533	0.3715
3.2802	18.6589	64000	3.5502	0.3719
3.2958	18.9504	65000	3.5461	0.3720
3.2175	19.2420	66000	3.5578	0.3715
3.2654	19.5335	67000	3.5526	0.3714
3.2579	19.8251	68000	3.5434	0.3723
3.1941	20.1166	69000	3.5589	0.3714
3.2266	20.4082	70000	3.5517	0.3719
3.2384	20.6997	71000	3.5468	0.3722
3.2588	20.9913	72000	3.5394	0.3728
3.1997	21.2828	73000	3.5622	0.3718
3.2248	21.5743	74000	3.5504	0.3723
3.26	21.8659	75000	3.5409	0.3730
3.1912	22.1574	76000	3.5571	0.3721
3.2111	22.4490	77000	3.5539	0.3722
3.2377	22.7405	78000	3.5427	0.3730
3.1442	23.0321	79000	3.5579	0.3725
3.1807	23.3236	80000	3.5556	0.3724
3.2096	23.6152	81000	3.5471	0.3732
3.2324	23.9067	82000	3.5384	0.3733
3.1653	24.1983	83000	3.5580	0.3724
3.1873	24.4898	84000	3.5516	0.3727
3.2106	24.7813	85000	3.5451	0.3733
3.1382	25.0729	86000	3.5556	0.3728
3.1757	25.3644	87000	3.5534	0.3729
3.1969	25.6560	88000	3.5455	0.3733
3.2073	25.9475	89000	3.5384	0.3735
3.1595	26.2391	90000	3.5561	0.3727
3.1857	26.5306	91000	3.5529	0.3734
3.2067	26.8222	92000	3.5444	0.3736
3.1267	27.1137	93000	3.5605	0.3729
3.1584	27.4052	94000	3.5548	0.3733
3.1835	27.6968	95000	3.5479	0.3733
3.209	27.9883	96000	3.5392	0.3742
3.1475	28.2799	97000	3.5545	0.3731
3.1645	28.5714	98000	3.5475	0.3739
3.1716	28.8630	99000	3.5415	0.3742
3.1244	29.1545	100000	3.5598	0.3732
3.1373	29.4461	101000	3.5515	0.3737
3.1661	29.7376	102000	3.5481	0.3736
3.0821	30.0292	103000	3.5584	0.3736
3.1414	30.3207	104000	3.5575	0.3734
3.1502	30.6122	105000	3.5513	0.3737
3.1632	30.9038	106000	3.5421	0.3741
3.1022	31.1953	107000	3.5608	0.3735
3.1365	31.4869	108000	3.5552	0.3737
3.1462	31.7784	109000	3.5486	0.3740

Framework versions

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