learning_rate: 5e-05
train_batch_size: 4
eval_batch_size: 1
seed: 42
gradient_accumulation_steps: 8
total_train_batch_size: 32
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Chrf++	Gen Len
4.4398	1.0	2	1.6731	22.409	131.9054
4.4398	2.0	4	1.6730	22.6373	125.1149
4.4398	3.0	6	1.6722	22.4006	127.9932
4.4398	4.0	8	1.6716	22.5914	130.5068
4.4398	5.0	10	1.6701	22.8162	125.2568
4.4398	6.0	12	1.6696	22.7978	124.5811
4.4398	7.0	14	1.6679	22.9677	123.7027
4.4398	8.0	16	1.6657	23.1412	118.7162
4.4398	9.0	18	1.6644	23.3956	112.5743
4.4398	10.0	20	1.6626	23.5915	105.5203
4.4398	11.0	22	1.6606	23.7557	108.7027
4.4398	12.0	24	1.6587	23.81	108.7095
4.4398	13.0	26	1.6581	23.353	116.9595
4.4398	14.0	28	1.6565	23.9004	115.6486
4.4398	15.0	30	1.6551	23.9943	113.4189
4.4398	16.0	32	1.6531	24.2434	108.0541
4.4398	17.0	34	1.6514	23.774	115.0608
4.4398	18.0	36	1.6516	23.9001	119.2297
4.4398	19.0	38	1.6512	23.3543	117.4189
4.4398	20.0	40	1.6521	23.2647	123.8378
4.4398	21.0	42	1.6499	23.4415	115.027
4.4398	22.0	44	1.6483	23.563	114.0473
4.4398	23.0	46	1.6487	23.9621	109.5
4.4398	24.0	48	1.6505	24.1548	108.9459
4.4398	25.0	50	1.6503	23.8819	116.2095
4.4398	26.0	52	1.6523	23.7761	115.6284
4.4398	27.0	54	1.6543	24.1968	114.4122
4.4398	28.0	56	1.6606	23.5004	127.8243
4.4398	29.0	58	1.6593	24.3851	111.027
4.4398	30.0	60	1.6645	24.4127	111.1824
4.4398	31.0	62	1.6685	24.5705	111.8108
4.4398	32.0	64	1.6710	24.8052	114.5541
4.4398	33.0	66	1.6762	24.6648	119.0743
4.4398	34.0	68	1.6768	24.6401	113.5541
4.4398	35.0	70	1.6779	24.2439	128.4392
4.4398	36.0	72	1.6856	24.8706	110.75
4.4398	37.0	74	1.6937	24.8547	116.8919
4.4398	38.0	76	1.6999	24.9912	112.2027
4.4398	39.0	78	1.7069	25.0511	112.9122
4.4398	40.0	80	1.7148	24.8773	114.1284
4.4398	41.0	82	1.7169	24.9462	117.7095
4.4398	42.0	84	1.7261	25.4464	111.8446
4.4398	43.0	86	1.7348	25.2694	116.9257
4.4398	44.0	88	1.7428	25.3079	114.5068
4.4398	45.0	90	1.7467	25.0552	116.3986
4.4398	46.0	92	1.7478	25.203	111.2297
4.4398	47.0	94	1.7565	25.2615	117.8514
4.4398	48.0	96	1.7623	25.1844	117.6554
4.4398	49.0	98	1.7681	24.7638	125.0541
1.3129	50.0	100	1.7785	24.9174	120.6892
1.3129	51.0	102	1.7852	24.7326	120.9257
1.3129	52.0	104	1.7866	24.7287	127.2297
1.3129	53.0	106	1.7936	24.7578	120.0405
1.3129	54.0	108	1.8004	25.1902	122.277
1.3129	55.0	110	1.8062	24.8639	121.723
1.3129	56.0	112	1.8043	26.0527	109.9054
1.3129	57.0	114	1.8106	25.6328	117.6486
1.3129	58.0	116	1.8145	26.2787	111.3649
1.3129	59.0	118	1.8234	26.3914	111.2568
1.3129	60.0	120	1.8328	25.8916	113.2973
1.3129	61.0	122	1.8328	26.0354	117.5541
1.3129	62.0	124	1.8385	26.0422	112.1014
1.3129	63.0	126	1.8421	26.5003	113.3649
1.3129	64.0	128	1.8481	26.8912	106.6554
1.3129	65.0	130	1.8613	26.1604	119.4054
1.3129	66.0	132	1.8621	26.35	112.8986
1.3129	67.0	134	1.8572	26.4157	111.6959
1.3129	68.0	136	1.8567	26.515	112.0
1.3129	69.0	138	1.8547	26.1988	114.9527
1.3129	70.0	140	1.8549	26.8537	108.3784
1.3129	71.0	142	1.8697	26.5113	112.7162
1.3129	72.0	144	1.8844	26.0838	112.9527
1.3129	73.0	146	1.8875	26.3631	115.5608
1.3129	74.0	148	1.8983	26.3795	116.7973
1.3129	75.0	150	1.9092	26.1195	112.1284
1.3129	76.0	152	1.9146	25.9709	109.25
1.3129	77.0	154	1.9263	26.5445	109.5541
1.3129	78.0	156	1.9204	26.4612	106.0811
1.3129	79.0	158	1.9318	26.4628	112.3986

Framework versions

Transformers 5.12.1
Pytorch 2.6.0+cu124
Datasets 5.0.0
Tokenizers 0.22.2

Tsedeniya/finetuned_glosslm_kmr

finetuned_glosslm_kmr

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Tsedeniya/
finetuned_glosslm_kmr