t5_recommendation_sports_equipment_english

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

• Loss: 0.4438
• Rouge1: 72.2222
• Rouge2: 66.6667
• Rougel: 72.2222
• Rougelsum: 72.2222
• Gen Len: 4.0952

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
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 40

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	1	9.9716	12.6943	0.0	12.4298	12.3728	19.0
No log	2.0	2	10.1466	10.0457	0.0	9.8413	9.8341	19.0
No log	3.0	3	8.3378	10.7204	0.0	10.4782	10.4681	19.0
No log	4.0	4	7.3021	10.7204	0.0	10.4782	10.4681	19.0
No log	5.0	5	6.3242	10.5739	0.0	10.3628	10.3550	19.0
No log	6.0	6	5.4331	10.3340	0.7937	10.3340	10.2659	19.0
No log	7.0	7	4.7152	10.8955	0.7937	10.9896	10.8598	18.9524
No log	8.0	8	3.9937	14.1311	3.5923	14.2840	13.9026	15.0952
No log	9.0	9	3.1163	16.2812	1.0025	16.1905	16.0614	6.4762
No log	10.0	10	2.3306	23.1746	7.1429	23.1746	23.6508	4.1429
No log	11.0	11	1.9695	21.4286	7.1429	21.4286	21.7460	4.0476
No log	12.0	12	1.5552	24.1270	7.1429	23.9683	24.1270	3.9048
No log	13.0	13	0.8986	9.0476	0.0	9.0476	9.0476	3.7619
No log	14.0	14	0.7398	18.2540	2.3810	18.2540	18.2540	4.1905
No log	15.0	15	0.6966	12.6984	0.0	11.9048	12.6984	3.6667
No log	16.0	16	0.6352	32.5397	14.2857	32.5397	31.7460	3.7619
No log	17.0	17	0.5722	43.6508	23.8095	43.6508	43.6508	4.0952
No log	18.0	18	0.5628	43.6508	23.8095	43.6508	43.6508	3.8571
No log	19.0	19	0.5526	43.1746	23.8095	43.0159	42.8571	3.8571
No log	20.0	20	0.5522	48.4127	38.0952	48.4127	48.4127	3.7619
No log	21.0	21	0.5201	42.8571	28.5714	42.6190	42.6984	4.2381
No log	22.0	22	0.5262	36.9841	19.0476	36.9841	36.9841	4.2857
No log	23.0	23	0.5093	38.0952	23.8095	37.5397	37.9365	4.1429
No log	24.0	24	0.4818	45.6349	33.3333	45.0794	45.3968	4.1429
No log	25.0	25	0.4547	50.7937	38.0952	50.0	50.7937	4.1429
No log	26.0	26	0.4455	50.7937	38.0952	50.0	50.7937	4.1429
No log	27.0	27	0.4660	53.1746	42.8571	53.1746	53.1746	4.0476
No log	28.0	28	0.4825	53.1746	42.8571	53.1746	53.1746	4.0
No log	29.0	29	0.4928	53.1746	42.8571	53.1746	53.1746	4.0476
No log	30.0	30	0.4838	57.4603	42.8571	57.1429	57.1429	4.0476
No log	31.0	31	0.4955	60.3175	47.6190	60.3175	60.3175	4.0476
No log	32.0	32	0.5066	62.6984	52.3810	62.6984	62.6984	4.1429
No log	33.0	33	0.5189	62.6984	52.3810	62.6984	62.6984	4.1905
No log	34.0	34	0.5234	62.6984	52.3810	62.6984	62.6984	4.1905
No log	35.0	35	0.5225	62.6984	52.3810	62.6984	62.6984	4.1905
No log	36.0	36	0.5225	62.6984	52.3810	62.6984	62.6984	4.1905
No log	37.0	37	0.5058	62.2222	52.3810	61.9048	61.9048	4.1429
No log	38.0	38	0.4861	70.6349	61.9048	69.8413	69.8413	4.1905
No log	39.0	39	0.4625	70.6349	61.9048	69.8413	69.8413	4.1905
No log	40.0	40	0.4438	72.2222	66.6667	72.2222	72.2222	4.0952

Framework versions

• Transformers 4.26.0
• Pytorch 2.3.0+cu121
• Datasets 2.8.0
• Tokenizers 0.13.3