exceptions_exp2_swap_take_to_push_5039

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

• Loss: 3.5565
• 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
4.8275	0.2911	1000	0.2557	4.7510
4.3366	0.5822	2000	0.2984	4.2895
4.1458	0.8733	3000	0.3153	4.0971
3.9919	1.1642	4000	0.3248	3.9935
3.9299	1.4553	5000	0.3318	3.9165
3.8737	1.7464	6000	0.3370	3.8594
3.7502	2.0373	7000	0.3414	3.8157
3.7525	2.3284	8000	0.3440	3.7854
3.7369	2.6195	9000	0.3471	3.7542
3.726	2.9106	10000	0.3496	3.7301
3.6384	3.2014	11000	0.3517	3.7162
3.6497	3.4925	12000	0.3534	3.6961
3.6427	3.7837	13000	0.3550	3.6786
3.5432	4.0745	14000	0.3562	3.6696
3.5661	4.3656	15000	0.3572	3.6586
3.5785	4.6567	16000	0.3587	3.6440
3.5853	4.9478	17000	0.3599	3.6337
3.5068	5.2387	18000	0.3604	3.6340
3.5222	5.5298	19000	0.3613	3.6250
3.5268	5.8209	20000	0.3620	3.6156
3.4494	6.1118	21000	0.3623	3.6182
3.4639	6.4029	22000	0.3631	3.6105
3.48	6.6940	23000	0.3642	3.6004
3.4942	6.9851	24000	0.3647	3.5903
3.414	7.2760	25000	0.3646	3.5988
3.4482	7.5671	26000	0.3653	3.5895
3.4635	7.8582	27000	0.3664	3.5793
3.3745	8.1490	28000	0.3661	3.5876
3.4127	8.4401	29000	0.3666	3.5825
3.4209	8.7313	30000	0.3671	3.5728
3.3206	9.0221	31000	0.3670	3.5786
3.3656	9.3132	32000	0.3676	3.5788
3.3969	9.6043	33000	0.3682	3.5695
3.4205	9.8954	34000	0.3686	3.5613
3.321	10.1863	35000	0.3681	3.5747
3.3602	10.4774	36000	0.3688	3.5663
3.3797	10.7685	37000	0.3693	3.5585
3.2791	11.0594	38000	0.3692	3.5690
3.3373	11.3505	39000	0.3691	3.5661
3.3598	11.6416	40000	0.3700	3.5565
3.3676	11.9327	41000	0.3705	3.5502
3.2981	12.2236	42000	0.3698	3.5641
3.3283	12.5147	43000	0.3704	3.5565
3.3464	12.8058	44000	0.3709	3.5471
3.2542	13.0966	45000	0.3704	3.5623
3.2961	13.3878	46000	0.3708	3.5589
3.3298	13.6789	47000	0.3711	3.5486
3.3472	13.9700	48000	0.3719	3.5395
3.2784	14.2608	49000	0.3710	3.5565
3.3166	14.5519	50000	0.3716	3.5482
3.3255	14.8430	51000	0.3718	3.5417
3.232	15.1339	52000	0.3715	3.5570
3.2841	15.4250	53000	0.3719	3.5489
3.2959	15.7161	54000	0.3723	3.5438
3.2495	16.0070	55000	0.3719	3.5521
3.2564	16.2981	56000	0.3718	3.5546
3.2723	16.5892	57000	0.3724	3.5431
3.2848	16.8803	58000	0.3728	3.5366
3.2261	17.1712	59000	0.3720	3.5537
3.2481	17.4623	60000	0.3727	3.5436
3.2767	17.7534	61000	0.3730	3.5371
3.1865	18.0442	62000	0.3727	3.5515
3.2425	18.3354	63000	0.3724	3.5483
3.2443	18.6265	64000	0.3733	3.5432
3.2743	18.9176	65000	0.3737	3.5323
3.2169	19.2084	66000	0.3729	3.5509
3.2448	19.4995	67000	0.3733	3.5435
3.2561	19.7906	68000	0.3737	3.5353
3.1648	20.0815	69000	0.3731	3.5504
3.2214	20.3726	70000	0.3734	3.5442
3.2448	20.6637	71000	0.3736	3.5375
3.2538	20.9548	72000	0.3744	3.5317
3.1887	21.2457	73000	0.3734	3.5490
3.222	21.5368	74000	0.3739	3.5425
3.2289	21.8279	75000	0.3740	3.5379
3.168	22.1188	76000	0.3737	3.5510
3.1926	22.4099	77000	0.3740	3.5464
3.2217	22.7010	78000	0.3742	3.5397
3.2512	22.9921	79000	0.3746	3.5279
3.1883	23.2830	80000	0.3739	3.5469
3.1876	23.5741	81000	3.5502	0.3738
3.1969	23.8652	82000	3.5425	0.3741
3.1555	24.1563	83000	3.5525	0.3738
3.1854	24.4474	84000	3.5491	0.3741
3.2108	24.7385	85000	3.5381	0.3744
3.1158	25.0294	86000	3.5545	0.3737
3.1504	25.3205	87000	3.5512	0.3738
3.1867	25.6116	88000	3.5389	0.3747
3.1968	25.9027	89000	3.5352	0.3751
3.1459	26.1936	90000	3.5523	0.3740
3.1692	26.4847	91000	3.5428	0.3748
3.1866	26.7758	92000	3.5373	0.3748
3.0939	27.0667	93000	3.5543	0.3742
3.1471	27.3578	94000	3.5456	0.3745
3.1629	27.6489	95000	3.5446	0.3746
3.191	27.9400	96000	3.5350	0.3752
3.1242	28.2308	97000	3.5557	0.3745
3.1423	28.5219	98000	3.5447	0.3748
3.1517	28.8131	99000	3.5374	0.3752
3.0951	29.1039	100000	3.5492	0.3745

Framework versions

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