my_awesome_billsum_model_40

This model is a fine-tuned version of google-t5/t5-small on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1082
• Rouge1: 0.9787
• Rouge2: 0.8875
• Rougel: 0.9329
• Rougelsum: 0.9315
• Gen Len: 5.2708

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 100
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	12	1.8920	0.4209	0.2805	0.384	0.3833	17.2292
No log	2.0	24	1.3065	0.4547	0.3125	0.4113	0.41	16.0
No log	3.0	36	0.8117	0.6973	0.546	0.6397	0.6374	10.3125
No log	4.0	48	0.6088	0.9492	0.7941	0.867	0.8609	5.1458
No log	5.0	60	0.5672	0.9513	0.797	0.8689	0.8631	5.125
No log	6.0	72	0.5178	0.9537	0.8052	0.8814	0.878	5.1458
No log	7.0	84	0.4737	0.9669	0.8387	0.9018	0.8988	5.1458
No log	8.0	96	0.4479	0.9709	0.8452	0.8972	0.8948	5.1667
No log	9.0	108	0.4178	0.9739	0.8595	0.9048	0.9038	5.1875
No log	10.0	120	0.3904	0.9739	0.8595	0.9048	0.9038	5.1875
No log	11.0	132	0.3681	0.9739	0.8595	0.9048	0.9038	5.1875
No log	12.0	144	0.3463	0.9769	0.8601	0.9066	0.9056	5.2083
No log	13.0	156	0.3295	0.9669	0.8253	0.887	0.8832	5.2917
No log	14.0	168	0.3124	0.9648	0.8236	0.8917	0.8885	5.3125
No log	15.0	180	0.3007	0.9648	0.8236	0.8917	0.8885	5.3125
No log	16.0	192	0.2976	0.9692	0.8346	0.8947	0.8908	5.2708
No log	17.0	204	0.2963	0.9671	0.833	0.8986	0.8952	5.2917
No log	18.0	216	0.2911	0.9671	0.833	0.8986	0.8952	5.2917
No log	19.0	228	0.2853	0.9717	0.8469	0.9028	0.9002	5.2917
No log	20.0	240	0.2782	0.9717	0.8469	0.9028	0.9002	5.2917
No log	21.0	252	0.2802	0.97	0.8462	0.9066	0.9043	5.3125
No log	22.0	264	0.2746	0.97	0.8462	0.9066	0.9043	5.3125
No log	23.0	276	0.2615	0.97	0.8462	0.9066	0.9043	5.3125
No log	24.0	288	0.2504	0.97	0.8462	0.9066	0.9043	5.3125
No log	25.0	300	0.2398	0.9656	0.8254	0.8946	0.8916	5.3333
No log	26.0	312	0.2301	0.9656	0.8254	0.8946	0.8916	5.3333
No log	27.0	324	0.2173	0.9656	0.8254	0.8946	0.8916	5.3333
No log	28.0	336	0.2109	0.9632	0.8237	0.8931	0.8899	5.3542
No log	29.0	348	0.2028	0.9632	0.8237	0.8931	0.8899	5.3542
No log	30.0	360	0.2016	0.9632	0.8237	0.8931	0.8899	5.3542
No log	31.0	372	0.1994	0.9632	0.8237	0.8931	0.8899	5.3542
No log	32.0	384	0.1986	0.9632	0.8237	0.8931	0.8899	5.3542
No log	33.0	396	0.1987	0.9632	0.8237	0.8931	0.8899	5.3542
No log	34.0	408	0.1965	0.9632	0.8237	0.8931	0.8899	5.3542
No log	35.0	420	0.1853	0.9632	0.8237	0.8931	0.8899	5.3542
No log	36.0	432	0.1841	0.9657	0.8368	0.9013	0.8982	5.3333
No log	37.0	444	0.1792	0.9657	0.8368	0.9013	0.8982	5.3333
No log	38.0	456	0.1778	0.9681	0.8379	0.8979	0.8954	5.3125
No log	39.0	468	0.1758	0.9657	0.8368	0.9013	0.8982	5.3333
No log	40.0	480	0.1778	0.9657	0.8368	0.9013	0.8982	5.3333
No log	41.0	492	0.1689	0.9638	0.8399	0.9064	0.904	5.3542
0.4636	42.0	504	0.1665	0.9638	0.8399	0.9064	0.904	5.3542
0.4636	43.0	516	0.1629	0.9657	0.8368	0.9013	0.8982	5.3333
0.4636	44.0	528	0.1616	0.9657	0.8472	0.9145	0.9109	5.3333
0.4636	45.0	540	0.1603	0.9681	0.8492	0.9112	0.9079	5.3125
0.4636	46.0	552	0.1592	0.9681	0.8492	0.9112	0.9079	5.3125
0.4636	47.0	564	0.1547	0.9681	0.8492	0.9112	0.9079	5.3125
0.4636	48.0	576	0.1500	0.9681	0.8492	0.9112	0.9079	5.3125
0.4636	49.0	588	0.1405	0.9681	0.8379	0.8979	0.8954	5.3125
0.4636	50.0	600	0.1316	0.9681	0.8379	0.8979	0.8954	5.3125
0.4636	51.0	612	0.1338	0.9681	0.8492	0.9112	0.9079	5.3125
0.4636	52.0	624	0.1351	0.9681	0.8492	0.9112	0.9079	5.3125
0.4636	53.0	636	0.1376	0.9681	0.8492	0.9112	0.9079	5.3125
0.4636	54.0	648	0.1349	0.9681	0.8492	0.9112	0.9079	5.3125
0.4636	55.0	660	0.1349	0.9681	0.8492	0.9112	0.9079	5.3125
0.4636	56.0	672	0.1319	0.9681	0.8492	0.9112	0.9079	5.3125
0.4636	57.0	684	0.1264	0.9681	0.8492	0.9112	0.9079	5.3125
0.4636	58.0	696	0.1223	0.9739	0.875	0.9282	0.926	5.2708
0.4636	59.0	708	0.1215	0.9739	0.875	0.9282	0.926	5.2708
0.4636	60.0	720	0.1233	0.9739	0.875	0.9282	0.926	5.2708
0.4636	61.0	732	0.1225	0.9739	0.875	0.9282	0.926	5.2708
0.4636	62.0	744	0.1201	0.9739	0.875	0.9282	0.926	5.2708
0.4636	63.0	756	0.1217	0.9739	0.875	0.9282	0.926	5.2708
0.4636	64.0	768	0.1220	0.9739	0.875	0.9282	0.926	5.2708
0.4636	65.0	780	0.1227	0.9739	0.875	0.9282	0.926	5.2708
0.4636	66.0	792	0.1215	0.9739	0.875	0.9282	0.926	5.2708
0.4636	67.0	804	0.1192	0.9787	0.8875	0.9329	0.9315	5.2708
0.4636	68.0	816	0.1171	0.9787	0.8875	0.9329	0.9315	5.2708
0.4636	69.0	828	0.1146	0.9821	0.8958	0.9424	0.9408	5.2917
0.4636	70.0	840	0.1129	0.9821	0.8958	0.9424	0.9408	5.2917
0.4636	71.0	852	0.1120	0.9821	0.8958	0.9424	0.9408	5.2917
0.4636	72.0	864	0.1098	0.9816	0.9101	0.9459	0.9455	5.2917
0.4636	73.0	876	0.1091	0.9722	0.8833	0.9304	0.9289	5.3125
0.4636	74.0	888	0.1086	0.9757	0.8976	0.9329	0.9325	5.3333
0.4636	75.0	900	0.1076	0.9816	0.9101	0.9459	0.9455	5.2917
0.4636	76.0	912	0.1080	0.9783	0.8958	0.9433	0.9419	5.2708
0.4636	77.0	924	0.1095	0.9821	0.8958	0.9424	0.9408	5.2917
0.4636	78.0	936	0.1112	0.9821	0.8958	0.9424	0.9408	5.2917
0.4636	79.0	948	0.1109	0.9821	0.8958	0.9424	0.9408	5.2917
0.4636	80.0	960	0.1101	0.9821	0.8958	0.9424	0.9408	5.2917
0.4636	81.0	972	0.1111	0.9821	0.8958	0.9424	0.9408	5.2917
0.4636	82.0	984	0.1102	0.9821	0.8958	0.9424	0.9408	5.2917
0.4636	83.0	996	0.1083	0.9821	0.911	0.9474	0.9464	5.2917
0.1189	84.0	1008	0.1084	0.9821	0.911	0.9474	0.9464	5.2917
0.1189	85.0	1020	0.1085	0.9851	0.9244	0.9502	0.9498	5.3125
0.1189	86.0	1032	0.1085	0.9816	0.9244	0.9508	0.9508	5.2917
0.1189	87.0	1044	0.1087	0.9816	0.9244	0.9508	0.9508	5.2917
0.1189	88.0	1056	0.1076	0.9816	0.9244	0.9508	0.9508	5.2917
0.1189	89.0	1068	0.1085	0.9788	0.9018	0.9364	0.9359	5.2708
0.1189	90.0	1080	0.1081	0.9823	0.9018	0.9359	0.9349	5.2917
0.1189	91.0	1092	0.1075	0.9788	0.9018	0.9364	0.9359	5.2708
0.1189	92.0	1104	0.1084	0.9823	0.9018	0.9359	0.9349	5.2917
0.1189	93.0	1116	0.1086	0.9823	0.9018	0.9359	0.9349	5.2917
0.1189	94.0	1128	0.1084	0.9787	0.8875	0.9329	0.9315	5.2708
0.1189	95.0	1140	0.1088	0.9787	0.8875	0.9329	0.9315	5.2708
0.1189	96.0	1152	0.1086	0.9787	0.8875	0.9329	0.9315	5.2708
0.1189	97.0	1164	0.1085	0.9787	0.8875	0.9329	0.9315	5.2708
0.1189	98.0	1176	0.1083	0.9787	0.8875	0.9329	0.9315	5.2708
0.1189	99.0	1188	0.1082	0.9787	0.8875	0.9329	0.9315	5.2708
0.1189	100.0	1200	0.1082	0.9787	0.8875	0.9329	0.9315	5.2708

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1