line1-classifier-nt-encoding-linear

This model is a fine-tuned version of InstaDeepAI/nucleotide-transformer-500m-human-ref on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3788
• F1 Score: 0.8515
• Precision Score: 0.8354
• Recall Score: 0.8747
• Tp: 335
• Tn: 319
• Fp: 66
• Fn: 48
• Line Ratio Reference: 0.4987
• Line Ratio Predictions: 0.5221

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: 0.0001
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	F1 Score	Precision Score	Recall Score	Tp	Tn	Fp	Fn	Line Ratio Reference	Line Ratio Predictions
0.6003	0.0406	100	0.5406	0.7261	0.7084	0.7676	294	264	121	89	0.4987	0.5404
0.5673	0.0812	200	0.6857	0.6227	0.8839	0.3577	137	367	18	246	0.4987	0.2018
0.545	0.1218	300	0.5341	0.7457	0.8556	0.6031	231	346	39	152	0.4987	0.3516
0.4891	0.1623	400	0.4730	0.7744	0.7949	0.7389	283	312	73	100	0.4987	0.4635
0.4933	0.2029	500	0.4979	0.7539	0.7048	0.8851	339	243	142	44	0.4987	0.6263
0.4762	0.2435	600	0.4560	0.7749	0.8487	0.6736	258	339	46	125	0.4987	0.3958
0.4571	0.2841	700	0.4380	0.8007	0.8108	0.7833	300	315	70	83	0.4987	0.4818
0.44	0.3247	800	0.4292	0.8007	0.7919	0.8146	312	303	82	71	0.4987	0.5130
0.4676	0.3653	900	0.4455	0.7907	0.8684	0.6893	264	345	40	119	0.4987	0.3958
0.4327	0.4058	1000	0.4089	0.8081	0.8410	0.7598	291	330	55	92	0.4987	0.4505
0.4356	0.4464	1100	0.4444	0.7848	0.7428	0.8747	335	269	116	48	0.4987	0.5872
0.4432	0.4870	1200	0.4180	0.8059	0.7940	0.8251	316	303	82	67	0.4987	0.5182
0.4272	0.5276	1300	0.3972	0.8149	0.8375	0.7807	299	327	58	84	0.4987	0.4648
0.4174	0.5682	1400	0.4586	0.7955	0.9134	0.6606	253	361	24	130	0.4987	0.3607
0.4161	0.6088	1500	0.3968	0.8110	0.8324	0.7781	298	325	60	85	0.4987	0.4661
0.4282	0.6494	1600	0.3943	0.8268	0.8157	0.8433	323	312	73	60	0.4987	0.5156
0.4256	0.6899	1700	0.3960	0.8136	0.8352	0.7807	299	326	59	84	0.4987	0.4661
0.4018	0.7305	1800	0.3926	0.8307	0.8373	0.8198	314	324	61	69	0.4987	0.4883
0.4195	0.7711	1900	0.3920	0.8294	0.8247	0.8355	320	317	68	63	0.4987	0.5052
0.399	0.8117	2000	0.4236	0.8221	0.7839	0.8903	341	291	94	42	0.4987	0.5664
0.4006	0.8523	2100	0.3959	0.8227	0.8020	0.8564	328	304	81	55	0.4987	0.5326
0.3822	0.8929	2200	0.3932	0.8249	0.8673	0.7676	294	340	45	89	0.4987	0.4414
0.3659	0.9334	2300	0.3931	0.8203	0.8868	0.7363	282	349	36	101	0.4987	0.4141
0.3416	0.9740	2400	0.4383	0.8233	0.7831	0.8956	343	290	95	40	0.4987	0.5703
0.3896	1.0146	2500	0.3925	0.8358	0.8579	0.8042	308	334	51	75	0.4987	0.4674
0.3375	1.0552	2600	0.3863	0.8259	0.8811	0.7546	289	346	39	94	0.4987	0.4271
0.3572	1.0958	2700	0.3672	0.8385	0.8381	0.8381	321	323	62	62	0.4987	0.4987
0.3435	1.1364	2800	0.3661	0.8451	0.8455	0.8433	323	326	59	60	0.4987	0.4974
0.3431	1.1769	2900	0.3935	0.8206	0.8773	0.7467	286	345	40	97	0.4987	0.4245
0.3397	1.2175	3000	0.3727	0.8302	0.8688	0.7781	298	340	45	85	0.4987	0.4466
0.3304	1.2581	3100	0.3882	0.8261	0.8720	0.7650	293	342	43	90	0.4987	0.4375
0.3369	1.2987	3200	0.3750	0.8450	0.8333	0.8616	330	319	66	53	0.4987	0.5156
0.3379	1.3393	3300	0.3646	0.8463	0.8571	0.8303	318	332	53	65	0.4987	0.4831
0.3359	1.3799	3400	0.3653	0.8357	0.8599	0.8016	307	335	50	76	0.4987	0.4648
0.3339	1.4205	3500	0.3798	0.8355	0.8746	0.7833	300	342	43	83	0.4987	0.4466
0.3441	1.4610	3600	0.3712	0.8331	0.8571	0.7990	306	334	51	77	0.4987	0.4648
0.3288	1.5016	3700	0.3672	0.8501	0.8743	0.8172	313	340	45	70	0.4987	0.4661
0.3322	1.5422	3800	0.3575	0.8382	0.8711	0.7937	304	340	45	79	0.4987	0.4544
0.334	1.5828	3900	0.3691	0.8344	0.8122	0.8695	333	308	77	50	0.4987	0.5339
0.3215	1.6234	4000	0.3674	0.8358	0.8173	0.8642	331	311	74	52	0.4987	0.5273
0.3187	1.6640	4100	0.3766	0.8326	0.8829	0.7676	294	346	39	89	0.4987	0.4336
0.3123	1.7045	4200	0.3965	0.8404	0.8919	0.7755	297	349	36	86	0.4987	0.4336
0.3135	1.7451	4300	0.3708	0.8486	0.8847	0.8016	307	345	40	76	0.4987	0.4518
0.3344	1.7857	4400	0.3954	0.8367	0.8	0.8982	344	299	86	39	0.4987	0.5599
0.3208	1.8263	4500	0.3645	0.8380	0.8798	0.7833	300	344	41	83	0.4987	0.4440
0.3234	1.8669	4600	0.3685	0.8445	0.8905	0.7859	301	348	37	82	0.4987	0.4401
0.326	1.9075	4700	0.3506	0.8463	0.8571	0.8303	318	332	53	65	0.4987	0.4831
0.3194	1.9481	4800	0.3582	0.8423	0.8195	0.8773	336	311	74	47	0.4987	0.5339
0.296	1.9886	4900	0.3560	0.8503	0.8490	0.8512	326	327	58	57	0.4987	0.5
0.2809	2.0292	5000	0.3788	0.8515	0.8354	0.8747	335	319	66	48	0.4987	0.5221
0.272	2.0698	5100	0.3511	0.8528	0.8689	0.8303	318	337	48	65	0.4987	0.4766
0.2556	2.1104	5200	0.3718	0.8463	0.8630	0.8225	315	335	50	68	0.4987	0.4753
0.2358	2.1510	5300	0.3761	0.8514	0.8747	0.8198	314	340	45	69	0.4987	0.4674
0.2738	2.1916	5400	0.3681	0.8539	0.8839	0.8146	312	344	41	71	0.4987	0.4596
0.2538	2.2321	5500	0.3959	0.8449	0.8220	0.8799	337	312	73	46	0.4987	0.5339
0.2602	2.2727	5600	0.4058	0.8487	0.8232	0.8877	340	312	73	43	0.4987	0.5378
0.2768	2.3133	5700	0.3697	0.8472	0.8866	0.7963	305	346	39	78	0.4987	0.4479
0.2637	2.3539	5800	0.3622	0.8529	0.8497	0.8564	328	327	58	55	0.4987	0.5026
0.252	2.3945	5900	0.3607	0.8580	0.8703	0.8407	322	337	48	61	0.4987	0.4818
0.2573	2.4351	6000	0.3544	0.8541	0.8613	0.8433	323	333	52	60	0.4987	0.4883
0.2736	2.4756	6100	0.3495	0.8515	0.8645	0.8329	319	335	50	64	0.4987	0.4805
0.2563	2.5162	6200	0.3536	0.8541	0.8672	0.8355	320	336	49	63	0.4987	0.4805
0.2307	2.5568	6300	0.3612	0.8489	0.8638	0.8277	317	335	50	66	0.4987	0.4779
0.2499	2.5974	6400	0.3586	0.8527	0.875	0.8225	315	340	45	68	0.4987	0.4688
0.2552	2.6380	6500	0.3558	0.8541	0.8733	0.8277	317	339	46	66	0.4987	0.4727
0.2503	2.6786	6600	0.3505	0.8528	0.8709	0.8277	317	338	47	66	0.4987	0.4740
0.2646	2.7192	6700	0.3493	0.8592	0.8809	0.8303	318	342	43	65	0.4987	0.4701
0.2485	2.7597	6800	0.3523	0.8554	0.8676	0.8381	321	336	49	62	0.4987	0.4818
0.2556	2.8003	6900	0.3612	0.8501	0.8743	0.8172	313	340	45	70	0.4987	0.4661
0.2297	2.8409	7000	0.3624	0.8567	0.8740	0.8329	319	339	46	64	0.4987	0.4753
0.2595	2.8815	7100	0.3608	0.8528	0.8649	0.8355	320	335	50	63	0.4987	0.4818
0.2523	2.9221	7200	0.3587	0.8554	0.8716	0.8329	319	338	47	64	0.4987	0.4766
0.2647	2.9627	7300	0.3569	0.8541	0.8652	0.8381	321	335	50	62	0.4987	0.4831

Framework versions

• PEFT 0.18.0
• Transformers 4.57.3
• Pytorch 2.9.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.1