flan-t5-base-extraction-cnndm_40000-all-hint_precision-ep50-nonstop

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

• Loss: 1.6648
• Hint Hit Num: 2.3792
• Hint Precision: 0.4254
• Num: 5.5651
• Gen Len: 18.9983

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: 60
• eval_batch_size: 400
• seed: 1799
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Hint Hit Num	Hint Precision	Num	Gen Len
2.1311	0.75	500	1.7662	2.1274	0.4051	5.2401	18.998
1.9622	1.5	1000	1.7248	2.209	0.4135	5.3367	18.9994
1.9108	2.25	1500	1.6995	2.2522	0.4187	5.3647	18.9999
1.8855	3.0	2000	1.6806	2.246	0.4156	5.3827	18.9999
1.8518	3.75	2500	1.6778	2.2829	0.4206	5.4082	18.9999
1.8324	4.5	3000	1.6741	2.2665	0.4175	5.4088	18.9999
1.8211	5.25	3500	1.6639	2.2819	0.4184	5.433	18.9999
1.7971	6.0	4000	1.6594	2.2896	0.4192	5.4375	18.9999
1.7788	6.75	4500	1.6554	2.3157	0.4224	5.4634	19.0
1.7755	7.5	5000	1.6550	2.3118	0.4216	5.4629	18.9999
1.7543	8.25	5500	1.6501	2.3345	0.4235	5.4905	18.9999
1.7491	9.0	6000	1.6534	2.3242	0.422	5.4823	18.9997
1.7317	9.75	6500	1.6483	2.2962	0.4178	5.4673	18.9999
1.7239	10.49	7000	1.6539	2.3283	0.4219	5.4958	18.9999
1.7109	11.24	7500	1.6495	2.3064	0.4198	5.4751	18.9997
1.7072	11.99	8000	1.6519	2.3465	0.4233	5.5209	18.999
1.6938	12.74	8500	1.6561	2.3086	0.4188	5.4821	18.999
1.6862	13.49	9000	1.6487	2.3524	0.423	5.5348	18.9991
1.6777	14.24	9500	1.6584	2.3453	0.4233	5.5088	18.999
1.6745	14.99	10000	1.6519	2.3062	0.418	5.4853	18.999
1.6623	15.74	10500	1.6553	2.3196	0.4202	5.4929	18.9992
1.6518	16.49	11000	1.6523	2.3467	0.4218	5.5332	18.999
1.651	17.24	11500	1.6568	2.36	0.4239	5.5397	18.999
1.6446	17.99	12000	1.6574	2.3526	0.423	5.5349	18.9991
1.6334	18.74	12500	1.6632	2.3106	0.4185	5.4907	18.9986
1.6322	19.49	13000	1.6590	2.3285	0.4199	5.5171	18.9987
1.6218	20.24	13500	1.6601	2.3377	0.4199	5.535	18.9993
1.6189	20.99	14000	1.6596	2.3493	0.4213	5.5447	18.9987
1.61	21.74	14500	1.6648	2.3792	0.4254	5.5651	18.9983
1.6064	22.49	15000	1.6668	2.3556	0.422	5.5521	18.9979
1.6004	23.24	15500	1.6674	2.3374	0.4195	5.5356	18.9987
1.597	23.99	16000	1.6654	2.3487	0.4203	5.5595	18.9987
1.5906	24.74	16500	1.6705	2.3634	0.4227	5.5575	18.9983
1.5851	25.49	17000	1.6690	2.3609	0.4229	5.5495	18.9983
1.5856	26.24	17500	1.6716	2.3444	0.4213	5.5376	18.9987
1.577	26.99	18000	1.6708	2.3693	0.4233	5.5631	18.9987
1.5734	27.74	18500	1.6707	2.3796	0.4236	5.5854	18.9983
1.5665	28.49	19000	1.6694	2.3639	0.4219	5.5698	18.9987
1.5666	29.24	19500	1.6798	2.3609	0.4221	5.5592	18.9987
1.564	29.99	20000	1.6778	2.3535	0.4204	5.5679	18.9987
1.5574	30.73	20500	1.6786	2.3476	0.4196	5.564	18.9987
1.5549	31.48	21000	1.6787	2.3658	0.4213	5.5862	18.999
1.5522	32.23	21500	1.6830	2.356	0.4212	5.5619	18.999
1.5485	32.98	22000	1.6784	2.3659	0.4218	5.5768	18.9987
1.5425	33.73	22500	1.6836	2.371	0.4222	5.5849	18.998
1.5449	34.48	23000	1.6817	2.365	0.4218	5.573	18.9985
1.5395	35.23	23500	1.6855	2.3633	0.4219	5.5694	18.9984
1.5358	35.98	24000	1.6834	2.3674	0.4221	5.5788	18.9988
1.5323	36.73	24500	1.6887	2.3725	0.4225	5.5857	18.9988
1.5298	37.48	25000	1.6861	2.3656	0.4207	5.5888	18.9991
1.526	38.23	25500	1.6905	2.3535	0.4202	5.5687	18.9991
1.5329	38.98	26000	1.6890	2.371	0.4218	5.5905	18.9988
1.5254	39.73	26500	1.6885	2.371	0.4223	5.5827	18.9989
1.5245	40.48	27000	1.6908	2.3615	0.4209	5.5781	18.9988
1.5166	41.23	27500	1.6907	2.3734	0.4214	5.598	18.9989
1.5225	41.98	28000	1.6904	2.3739	0.4219	5.5945	18.9989
1.5149	42.73	28500	1.6916	2.3768	0.4229	5.5913	18.9989
1.5178	43.48	29000	1.6938	2.3654	0.4214	5.5826	18.9991
1.5212	44.23	29500	1.6928	2.3674	0.4219	5.5821	18.9988
1.5136	44.98	30000	1.6917	2.3781	0.4227	5.5952	18.999
1.5097	45.73	30500	1.6923	2.3704	0.4218	5.5896	18.999
1.5183	46.48	31000	1.6935	2.3719	0.4217	5.5931	18.999
1.5092	47.23	31500	1.6935	2.3684	0.4216	5.5868	18.999
1.5127	47.98	32000	1.6943	2.3691	0.4218	5.5863	18.999
1.5124	48.73	32500	1.6940	2.3704	0.422	5.5874	18.9987
1.5117	49.48	33000	1.6948	2.3693	0.4217	5.587	18.9987

Framework versions

• Transformers 4.18.0
• Pytorch 1.10.0+cu111
• Datasets 2.5.1
• Tokenizers 0.12.1