COPAS-mms1ball-Nov29

This model is a fine-tuned version of facebook/mms-1b-all on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3656
• Wer: 0.2399

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: 1e-05
• train_batch_size: 3
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 100
• num_epochs: 100
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
39.0459	0.2358	100	32.7908	1.0
31.3076	0.4717	200	24.5080	1.0005
22.6116	0.7075	300	16.6952	1.0
14.3915	0.9434	400	9.8732	1.0
8.4511	1.1792	500	5.7561	1.0
4.8318	1.4151	600	3.8778	1.0
3.5683	1.6509	700	3.2052	1.0
3.1283	1.8868	800	3.0170	1.0
3.0561	2.1226	900	2.9596	1.0
2.9704	2.3585	1000	2.9107	1.0
2.9138	2.5943	1100	2.8762	1.0
2.9195	2.8302	1200	2.8405	1.0
2.8696	3.0660	1300	2.7997	1.0
2.824	3.3019	1400	2.7525	1.0
2.7823	3.5377	1500	2.6803	1.0
2.7284	3.7736	1600	2.5883	0.9997
2.6758	4.0094	1700	2.4691	0.9985
2.5467	4.2453	1800	2.3142	0.9871
2.3888	4.4811	1900	2.1150	0.9593
2.2207	4.7170	2000	1.8880	0.9007
2.0405	4.9528	2100	1.6605	0.8418
1.8474	5.1887	2200	1.4562	0.7697
1.6394	5.4245	2300	1.2640	0.6726
1.4813	5.6604	2400	1.1130	0.5468
1.3794	5.8962	2500	1.0028	0.4876
1.2862	6.1321	2600	0.9112	0.4381
1.1636	6.3679	2700	0.8412	0.4133
1.1022	6.6038	2800	0.7828	0.4019
1.1028	6.8396	2900	0.7427	0.3936
0.9554	7.0755	3000	0.7045	0.3873
0.9922	7.3113	3100	0.6779	0.3830
0.9837	7.5472	3200	0.6560	0.3799
0.9329	7.7830	3300	0.6360	0.3764
0.9373	8.0189	3400	0.6202	0.3749
0.8958	8.2547	3500	0.6060	0.3716
0.8097	8.4906	3600	0.5937	0.3703
0.8469	8.7264	3700	0.5847	0.3696
0.9508	8.9623	3800	0.5755	0.3686
0.7845	9.1981	3900	0.5669	0.3670
0.8738	9.4340	4000	0.5585	0.3648
0.8345	9.6698	4100	0.5528	0.3650
0.8014	9.9057	4200	0.5484	0.3635
0.852	10.1415	4300	0.5417	0.3607
0.8325	10.3774	4400	0.5368	0.3595
0.7868	10.6132	4500	0.5317	0.3562
0.7985	10.8491	4600	0.5285	0.3544
0.7187	11.0849	4700	0.5245	0.3534
0.764	11.3208	4800	0.5222	0.3516
0.7842	11.5566	4900	0.5188	0.3473
0.7214	11.7925	5000	0.5149	0.3478
0.8369	12.0283	5100	0.5108	0.3433
0.8013	12.2642	5200	0.5067	0.3415
0.7132	12.5	5300	0.5055	0.3385
0.7433	12.7358	5400	0.5037	0.3375
0.7471	12.9717	5500	0.4994	0.3357
0.7	13.2075	5600	0.4972	0.3344
0.764	13.4434	5700	0.4946	0.3334
0.7371	13.6792	5800	0.4917	0.3319
0.7365	13.9151	5900	0.4892	0.3274
0.7302	14.1509	6000	0.4870	0.3266
0.6932	14.3868	6100	0.4874	0.3243
0.7475	14.6226	6200	0.4852	0.3228
0.7309	14.8585	6300	0.4816	0.3210
0.7329	15.0943	6400	0.4789	0.3218
0.7439	15.3302	6500	0.4770	0.3195
0.719	15.5660	6600	0.4753	0.3175
0.6595	15.8019	6700	0.4742	0.3145
0.6897	16.0377	6800	0.4725	0.3127
0.689	16.2736	6900	0.4706	0.3097
0.7532	16.5094	7000	0.4683	0.3086
0.7149	16.7453	7100	0.4676	0.3076
0.7224	16.9811	7200	0.4659	0.3056
0.6838	17.2170	7300	0.4640	0.3043
0.6294	17.4528	7400	0.4623	0.3043
0.7538	17.6887	7500	0.4610	0.3054
0.6775	17.9245	7600	0.4604	0.3033
0.7451	18.1604	7700	0.4581	0.3011
0.664	18.3962	7800	0.4582	0.3008
0.6857	18.6321	7900	0.4550	0.3008
0.6746	18.8679	8000	0.4531	0.2983
0.6475	19.1038	8100	0.4540	0.3003
0.6577	19.3396	8200	0.4525	0.2973
0.7043	19.5755	8300	0.4510	0.2988
0.6841	19.8113	8400	0.4504	0.2963
0.6529	20.0472	8500	0.4498	0.2963
0.7497	20.2830	8600	0.4466	0.2940
0.6469	20.5189	8700	0.4457	0.2932
0.6383	20.7547	8800	0.4453	0.2922
0.6848	20.9906	8900	0.4441	0.2922
0.6669	21.2264	9000	0.4429	0.2930
0.6405	21.4623	9100	0.4417	0.2917
0.6643	21.6981	9200	0.4417	0.2915
0.6853	21.9340	9300	0.4383	0.2894
0.6366	22.1698	9400	0.4385	0.2864
0.6469	22.4057	9500	0.4381	0.2864
0.6843	22.6415	9600	0.4356	0.2872
0.642	22.8774	9700	0.4364	0.2864
0.7332	23.1132	9800	0.4338	0.2861
0.6393	23.3491	9900	0.4348	0.2864
0.6332	23.5849	10000	0.4324	0.2856
0.6038	23.8208	10100	0.4324	0.2851
0.6561	24.0566	10200	0.4311	0.2824
0.6515	24.2925	10300	0.4293	0.2826
0.6436	24.5283	10400	0.4285	0.2806
0.5898	24.7642	10500	0.4279	0.2808
0.6363	25.0	10600	0.4260	0.2801
0.6533	25.2358	10700	0.4254	0.2783
0.6512	25.4717	10800	0.4243	0.2788
0.6041	25.7075	10900	0.4235	0.2786
0.6423	25.9434	11000	0.4234	0.2770
0.5831	26.1792	11100	0.4225	0.2750
0.6029	26.4151	11200	0.4213	0.2738
0.6696	26.6509	11300	0.4210	0.2740
0.657	26.8868	11400	0.4202	0.2722
0.6323	27.1226	11500	0.4190	0.2743
0.6333	27.3585	11600	0.4180	0.2738
0.635	27.5943	11700	0.4175	0.2715
0.6309	27.8302	11800	0.4189	0.2715
0.6286	28.0660	11900	0.4182	0.2707
0.5783	28.3019	12000	0.4188	0.2682
0.627	28.5377	12100	0.4163	0.2707
0.6092	28.7736	12200	0.4148	0.2687
0.6429	29.0094	12300	0.4133	0.2695
0.6148	29.2453	12400	0.4128	0.2700
0.6757	29.4811	12500	0.4118	0.2685
0.6041	29.7170	12600	0.4124	0.2682
0.6144	29.9528	12700	0.4117	0.2682
0.5617	30.1887	12800	0.4100	0.2672
0.5981	30.4245	12900	0.4100	0.2674
0.6428	30.6604	13000	0.4089	0.2664
0.5913	30.8962	13100	0.4096	0.2652
0.6583	31.1321	13200	0.4078	0.2662
0.6472	31.3679	13300	0.4073	0.2664
0.5755	31.6038	13400	0.4061	0.2649
0.6032	31.8396	13500	0.4063	0.2639
0.5765	32.0755	13600	0.4062	0.2631
0.6069	32.3113	13700	0.4060	0.2624
0.6172	32.5472	13800	0.4055	0.2626
0.603	32.7830	13900	0.4044	0.2624
0.6575	33.0189	14000	0.4033	0.2614
0.6387	33.2547	14100	0.4023	0.2611
0.5852	33.4906	14200	0.4028	0.2614
0.5802	33.7264	14300	0.4017	0.2616
0.6124	33.9623	14400	0.4010	0.2604
0.6248	34.1981	14500	0.4006	0.2606
0.6169	34.4340	14600	0.4009	0.2599
0.564	34.6698	14700	0.4007	0.2581
0.5788	34.9057	14800	0.3997	0.2591
0.6364	35.1415	14900	0.3988	0.2581
0.6257	35.3774	15000	0.3985	0.2576
0.5516	35.6132	15100	0.3984	0.2588
0.6263	35.8491	15200	0.3963	0.2594
0.5153	36.0849	15300	0.3962	0.2578
0.5622	36.3208	15400	0.3960	0.2581
0.5238	36.5566	15500	0.3970	0.2566
0.6735	36.7925	15600	0.3957	0.2571
0.6307	37.0283	15700	0.3958	0.2573
0.6231	37.2642	15800	0.3945	0.2576
0.5458	37.5	15900	0.3931	0.2558
0.6256	37.7358	16000	0.3913	0.2563
0.5762	37.9717	16100	0.3912	0.2551
0.5254	38.2075	16200	0.3917	0.2548
0.6052	38.4434	16300	0.3912	0.2535
0.5942	38.6792	16400	0.3911	0.2538
0.5946	38.9151	16500	0.3902	0.2538
0.5595	39.1509	16600	0.3901	0.2530
0.5734	39.3868	16700	0.3899	0.2528
0.6011	39.6226	16800	0.3899	0.2528
0.6162	39.8585	16900	0.3886	0.2525
0.559	40.0943	17000	0.3885	0.2513
0.5645	40.3302	17100	0.3866	0.2513
0.6226	40.5660	17200	0.3871	0.2515
0.5847	40.8019	17300	0.3864	0.2505
0.5887	41.0377	17400	0.3861	0.2513
0.5798	41.2736	17500	0.3851	0.2515
0.5874	41.5094	17600	0.3846	0.25
0.5508	41.7453	17700	0.3849	0.2492
0.5838	41.9811	17800	0.3845	0.2490
0.5456	42.2170	17900	0.3831	0.2492
0.5564	42.4528	18000	0.3830	0.2480
0.5869	42.6887	18100	0.3824	0.2462
0.5954	42.9245	18200	0.3830	0.2482
0.6119	43.1604	18300	0.3812	0.2492
0.567	43.3962	18400	0.3812	0.2472
0.5582	43.6321	18500	0.3814	0.2462
0.5892	43.8679	18600	0.3810	0.2465
0.5518	44.1038	18700	0.3806	0.2472
0.5711	44.3396	18800	0.3807	0.2475
0.5688	44.5755	18900	0.3805	0.2475
0.5578	44.8113	19000	0.3806	0.2470
0.6209	45.0472	19100	0.3794	0.2467
0.5857	45.2830	19200	0.3804	0.2465
0.5627	45.5189	19300	0.3804	0.2467
0.5763	45.7547	19400	0.3788	0.2454
0.5156	45.9906	19500	0.3788	0.2462
0.5729	46.2264	19600	0.3787	0.2454
0.5487	46.4623	19700	0.3774	0.2444
0.5907	46.6981	19800	0.3777	0.2454
0.5644	46.9340	19900	0.3768	0.2452
0.5931	47.1698	20000	0.3767	0.2449
0.5478	47.4057	20100	0.3771	0.2429
0.5243	47.6415	20200	0.3771	0.2442
0.5685	47.8774	20300	0.3757	0.2419
0.5726	48.1132	20400	0.3754	0.2442
0.5173	48.3491	20500	0.3760	0.2439
0.5968	48.5849	20600	0.3765	0.2429
0.5777	48.8208	20700	0.3748	0.2401
0.5744	49.0566	20800	0.3750	0.2409
0.5613	49.2925	20900	0.3752	0.2414
0.5391	49.5283	21000	0.3744	0.2406
0.5792	49.7642	21100	0.3740	0.2419
0.5557	50.0	21200	0.3738	0.2424
0.5856	50.2358	21300	0.3736	0.2401
0.5424	50.4717	21400	0.3729	0.2412
0.5288	50.7075	21500	0.3729	0.2396
0.5527	50.9434	21600	0.3720	0.2394
0.615	51.1792	21700	0.3722	0.2389
0.5514	51.4151	21800	0.3719	0.2396
0.5393	51.6509	21900	0.3719	0.2396
0.5501	51.8868	22000	0.3714	0.2391
0.5798	52.1226	22100	0.3700	0.2396
0.5434	52.3585	22200	0.3702	0.2391
0.5552	52.5943	22300	0.3693	0.2401
0.546	52.8302	22400	0.3692	0.2394
0.5097	53.0660	22500	0.3695	0.2394
0.5282	53.3019	22600	0.3698	0.2396
0.5435	53.5377	22700	0.3692	0.2389
0.5104	53.7736	22800	0.3695	0.2394
0.5664	54.0094	22900	0.3685	0.2384
0.5502	54.2453	23000	0.3686	0.2389
0.558	54.4811	23100	0.3682	0.2396
0.5888	54.7170	23200	0.3679	0.2389
0.4898	54.9528	23300	0.3680	0.2379
0.6259	55.1887	23400	0.3676	0.2371
0.5581	55.4245	23500	0.3663	0.2386
0.5439	55.6604	23600	0.3672	0.2391
0.5254	55.8962	23700	0.3666	0.2391
0.5214	56.1321	23800	0.3659	0.2384
0.5219	56.3679	23900	0.3657	0.2391
0.564	56.6038	24000	0.3656	0.2399

Framework versions

• Transformers 4.43.4
• Pytorch 2.4.1
• Datasets 3.0.0
• Tokenizers 0.19.1