exceptions_exp2_swap_take_to_push_3591

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

• Loss: 3.5538
• Accuracy: 0.3701

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
4.8294	0.2911	1000	4.7448	0.2558
4.333	0.5822	2000	4.2753	0.2998
4.1415	0.8733	3000	4.0907	0.3161
3.9891	1.1642	4000	3.9886	0.3261
3.9304	1.4553	5000	3.9144	0.3320
3.8614	1.7464	6000	3.8540	0.3374
3.7284	2.0373	7000	3.8088	0.3420
3.7392	2.3284	8000	3.7766	0.3452
3.726	2.6195	9000	3.7497	0.3481
3.7113	2.9106	10000	3.7223	0.3503
3.6261	3.2014	11000	3.7093	0.3520
3.6235	3.4925	12000	3.6899	0.3541
3.632	3.7837	13000	3.6733	0.3554
3.531	4.0745	14000	3.6652	0.3568
3.5497	4.3656	15000	3.6533	0.3578
3.5631	4.6567	16000	3.6390	0.3592
3.5799	4.9478	17000	3.6274	0.3604
3.4926	5.2387	18000	3.6316	0.3605
3.5165	5.5298	19000	3.6198	0.3618
3.5263	5.8209	20000	3.6089	0.3626
3.4335	6.1118	21000	3.6114	0.3629
3.4665	6.4029	22000	3.6052	0.3634
3.4795	6.6940	23000	3.5951	0.3643
3.4842	6.9851	24000	3.5897	0.3648
3.4084	7.2760	25000	3.5951	0.3651
3.4372	7.5671	26000	3.5851	0.3658
3.4624	7.8582	27000	3.5757	0.3665
3.3668	8.1490	28000	3.5880	0.3664
3.399	8.4401	29000	3.5823	0.3664
3.428	8.7313	30000	3.5709	0.3677
3.313	9.0221	31000	3.5755	0.3677
3.3647	9.3132	32000	3.5726	0.3680
3.3893	9.6043	33000	3.5644	0.3687
3.4037	9.8954	34000	3.5593	0.3689
3.3285	10.1863	35000	3.5704	0.3687
3.353	10.4774	36000	3.5609	0.3689
3.3777	10.7685	37000	3.5549	0.3697
3.2807	11.0594	38000	3.5631	0.3696
3.317	11.3505	39000	3.5619	0.3695
3.3542	11.6416	40000	3.5538	0.3701
3.3574	11.9327	41000	3.5434	0.3707
3.2942	12.2236	42000	3.5574	0.3701
3.3173	12.5147	43000	3.5530	0.3709
3.3501	12.8058	44000	3.5423	0.3713
3.2594	13.0966	45000	3.5594	0.3707
3.2944	13.3878	46000	3.5518	0.3711
3.3138	13.6789	47000	3.5467	0.3716
3.3309	13.9700	48000	3.5373	0.3720
3.2652	14.2608	49000	3.5530	0.3715
3.2915	14.5519	50000	3.5454	0.3719
3.3289	14.8430	51000	3.5339	0.3725
3.2271	15.1339	52000	3.5523	0.3719
3.2716	15.4250	53000	3.5465	0.3722
3.2879	15.7161	54000	3.5418	0.3725
3.2611	16.0070	55000	3.5481	0.3722
3.2312	16.2981	56000	3.5468	0.3727
3.282	16.5892	57000	3.5414	0.3727
3.2917	16.8803	58000	3.5306	0.3733
3.2145	17.1712	59000	3.5497	0.3722
3.2592	17.4623	60000	3.5467	0.3730
3.2712	17.7534	61000	3.5315	0.3735
3.1677	18.0442	62000	3.5441	0.3731
3.2276	18.3354	63000	3.5478	0.3731
3.2524	18.6265	64000	3.5376	0.3734
3.2711	18.9176	65000	3.5303	0.3740
3.1879	19.2084	66000	3.5476	0.3728
3.2276	19.4995	67000	3.5405	0.3733
3.2617	19.7906	68000	3.5351	0.3738
3.154	20.0815	69000	3.5467	0.3733
3.2018	20.3726	70000	3.5408	0.3740
3.2386	20.6637	71000	3.5344	0.3738
3.2336	20.9548	72000	3.5302	0.3744
3.1873	21.2457	73000	3.5451	0.3737
3.2165	21.5368	74000	3.5370	0.3741
3.2324	21.8279	75000	3.5326	0.3746
3.164	22.1188	76000	3.5445	0.3739
3.1883	22.4099	77000	3.5440	0.3742
3.2239	22.7010	78000	3.5326	0.3745
3.2313	22.9921	79000	3.5230	0.3754
3.1688	23.2830	80000	3.5427	0.3742
3.2001	23.5741	81000	3.5363	0.3745
3.2049	23.8652	82000	3.5289	0.3750
3.1424	24.1560	83000	3.5453	0.3741
3.1721	24.4471	84000	3.5398	0.3747
3.1952	24.7382	85000	3.5324	0.3752
3.1083	25.0291	86000	3.5467	0.3745
3.1467	25.3202	87000	3.5426	0.3744
3.1692	25.6113	88000	3.5349	0.3748
3.2055	25.9024	89000	3.5309	0.3753
3.1306	26.1933	90000	3.5457	0.3746
3.1589	26.4844	91000	3.5411	0.3749
3.1932	26.7755	92000	3.5326	0.3754
3.1	27.0664	93000	3.5464	0.3747
3.1413	27.3575	94000	3.5418	0.3749
3.1617	27.6486	95000	3.5378	0.3754
3.1821	27.9397	96000	3.5290	0.3757
3.1096	28.2306	97000	3.5468	0.3750
3.1535	28.5217	98000	3.5427	0.3747
3.1564	28.8128	99000	3.5326	0.3756

Framework versions

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