t5-small-finetuned-question-answer-template

This model is a fine-tuned version of t5-small on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2608
• Rouge1: 84.7051
• Rouge2: 67.1537
• Rougel: 80.4837
• Rougelsum: 80.763
• Gen Len: 13.5

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: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	15	2.0280	60.1069	36.344	49.1148	49.3518	11.8077
No log	2.0	30	1.5653	66.3894	39.7781	53.9648	54.1828	11.0769
No log	3.0	45	1.2481	55.2986	33.5161	45.3224	45.3243	12.2308
No log	4.0	60	0.9674	48.2375	31.5412	43.2247	43.6014	13.0
No log	5.0	75	0.7555	48.2945	33.4249	44.2227	44.3073	14.2308
No log	6.0	90	0.6018	61.6858	43.9501	57.9666	58.4171	11.5
No log	7.0	105	0.4910	69.2262	51.5066	66.0391	66.5198	11.4231
No log	8.0	120	0.4109	78.2833	59.9289	73.4671	73.7781	12.5769
No log	9.0	135	0.3641	83.3592	64.287	78.1249	78.6497	13.3846
No log	10.0	150	0.3312	84.3652	66.4705	79.9104	80.3878	13.4615
No log	11.0	165	0.3115	84.3652	66.4705	79.9104	80.3878	13.4615
No log	12.0	180	0.2949	84.3652	66.4705	79.9104	80.3878	13.4615
No log	13.0	195	0.2823	84.2552	66.4705	79.8583	80.2947	13.5
No log	14.0	210	0.2745	84.2552	66.4705	79.8583	80.2947	13.4615
No log	15.0	225	0.2689	84.7051	67.1537	80.4837	80.763	13.5
No log	16.0	240	0.2649	84.7051	67.1537	80.4837	80.763	13.5
No log	17.0	255	0.2630	84.7051	67.1537	80.4837	80.763	13.5
No log	18.0	270	0.2617	84.7051	67.1537	80.4837	80.763	13.5
No log	19.0	285	0.2611	84.7051	67.1537	80.4837	80.763	13.5
No log	20.0	300	0.2608	84.7051	67.1537	80.4837	80.763	13.5

Framework versions

• Transformers 4.45.2
• Pytorch 2.4.1+cpu
• Datasets 3.0.1
• Tokenizers 0.20.1