GUE_prom_prom_core_notata-seqsight_4096_512_46M-L8_f

This model is a fine-tuned version of mahdibaghbanzadeh/seqsight_4096_512_46M on the mahdibaghbanzadeh/GUE_prom_prom_core_notata dataset. It achieves the following results on the evaluation set:

• Loss: 0.3732
• F1 Score: 0.8451
• Accuracy: 0.8451

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: 0.0005
• train_batch_size: 128
• eval_batch_size: 128
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• training_steps: 10000

Training results

Training Loss	Epoch	Step	Validation Loss	F1 Score	Accuracy
0.4849	0.6	200	0.3996	0.8184	0.8185
0.4175	1.2	400	0.3830	0.8277	0.8278
0.3993	1.81	600	0.3797	0.8292	0.8293
0.3909	2.41	800	0.3694	0.8344	0.8344
0.383	3.01	1000	0.3659	0.8403	0.8404
0.3765	3.61	1200	0.3609	0.8391	0.8391
0.3792	4.22	1400	0.3657	0.8349	0.8349
0.3787	4.82	1600	0.3606	0.8408	0.8408
0.3656	5.42	1800	0.3801	0.8337	0.8340
0.3728	6.02	2000	0.3631	0.8396	0.8396
0.3688	6.63	2200	0.3582	0.8420	0.8421
0.3632	7.23	2400	0.3628	0.8431	0.8432
0.3651	7.83	2600	0.3620	0.8423	0.8423
0.3578	8.43	2800	0.3633	0.8426	0.8428
0.3639	9.04	3000	0.3591	0.8427	0.8427
0.3559	9.64	3200	0.3590	0.8442	0.8442
0.3546	10.24	3400	0.3612	0.8438	0.8438
0.353	10.84	3600	0.3598	0.8436	0.8436
0.3518	11.45	3800	0.3592	0.8429	0.8428
0.3512	12.05	4000	0.3574	0.8431	0.8430
0.3473	12.65	4200	0.3593	0.8451	0.8451
0.3488	13.25	4400	0.3598	0.8424	0.8425
0.3509	13.86	4600	0.3601	0.8475	0.8476
0.3471	14.46	4800	0.3589	0.8492	0.8493
0.3437	15.06	5000	0.3577	0.8466	0.8466
0.3406	15.66	5200	0.3582	0.8488	0.8489
0.3433	16.27	5400	0.3622	0.8451	0.8451
0.3414	16.87	5600	0.3656	0.8457	0.8461
0.3373	17.47	5800	0.3634	0.8453	0.8455
0.3475	18.07	6000	0.3605	0.8451	0.8453
0.3369	18.67	6200	0.3579	0.8486	0.8487
0.3393	19.28	6400	0.3588	0.8457	0.8457
0.339	19.88	6600	0.3589	0.8460	0.8461
0.332	20.48	6800	0.3609	0.8452	0.8453
0.3415	21.08	7000	0.3592	0.8456	0.8457
0.337	21.69	7200	0.3605	0.8470	0.8470
0.331	22.29	7400	0.3590	0.8488	0.8489
0.3313	22.89	7600	0.3626	0.8461	0.8462
0.3318	23.49	7800	0.3614	0.8460	0.8461
0.3358	24.1	8000	0.3623	0.8486	0.8487
0.3355	24.7	8200	0.3616	0.8468	0.8470
0.3265	25.3	8400	0.3658	0.8444	0.8445
0.3346	25.9	8600	0.3607	0.8490	0.8491
0.3311	26.51	8800	0.3616	0.8485	0.8485
0.3307	27.11	9000	0.3607	0.8474	0.8474
0.3341	27.71	9200	0.3618	0.8484	0.8485
0.3214	28.31	9400	0.3636	0.8463	0.8464
0.3288	28.92	9600	0.3634	0.8482	0.8483
0.3325	29.52	9800	0.3626	0.8479	0.8479
0.324	30.12	10000	0.3628	0.8477	0.8477

Framework versions

• PEFT 0.9.0
• Transformers 4.38.2
• Pytorch 2.2.0+cu121
• Datasets 2.17.1
• Tokenizers 0.15.2