MSR-VTT-real-moderate_kl05-model-run2

This model is a fine-tuned version of gpt2 on an unknown dataset. It achieves the following results on the evaluation set:

Model description

More information needed

More information needed

More information needed

The following hyperparameters were used during training:

learning_rate: 1e-05
train_batch_size: 8
eval_batch_size: 16
seed: 42
optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 100
num_epochs: 50
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss
4.2843	1.0	309	3.9968
3.992	2.0	618	3.8891
3.8682	3.0	927	3.8550
3.6733	4.0	1236	3.8397
3.5029	5.0	1545	3.8347
3.3741	6.0	1854	3.8386
3.6111	7.0	2163	3.8416
3.3825	8.0	2472	3.8575
3.3059	9.0	2781	3.8732
3.2055	10.0	3090	3.9035
3.1009	11.0	3399	3.9143
3.096	12.0	3708	3.9303
2.9366	13.0	4017	3.9587
3.0297	14.0	4326	3.9803
2.9667	15.0	4635	4.0008
2.867	16.0	4944	4.0216
2.7298	17.0	5253	4.0639
2.903	18.0	5562	4.0928
2.6696	19.0	5871	4.1225
2.7352	20.0	6180	4.1452
2.4967	21.0	6489	4.1795
2.6842	22.0	6798	4.2077
2.524	23.0	7107	4.2342
2.3755	24.0	7416	4.2657
2.4914	25.0	7725	4.2914
2.4262	26.0	8034	4.3363
2.5525	27.0	8343	4.3569
2.3333	28.0	8652	4.3795
2.3502	29.0	8961	4.4104
2.2461	30.0	9270	4.4472
2.2232	31.0	9579	4.4459
2.2551	32.0	9888	4.4897
2.139	33.0	10197	4.5207
2.0815	34.0	10506	4.5432
2.1642	35.0	10815	4.5530
2.1724	36.0	11124	4.5782
2.1013	37.0	11433	4.6007
2.0314	38.0	11742	4.6085
2.1594	39.0	12051	4.6321
2.0138	40.0	12360	4.6428
1.9364	41.0	12669	4.6672
2.0273	42.0	12978	4.6664
2.0119	43.0	13287	4.6873
1.9308	44.0	13596	4.7023
1.9809	45.0	13905	4.7079
1.8919	46.0	14214	4.7118
2.0521	47.0	14523	4.7211
1.9894	48.0	14832	4.7239
1.998	49.0	15141	4.7285
2.0072	50.0	15450	4.7298

Safetensors

Model size

0.1B params

Tensor type

F32

Base model

Finetuned

(2021)

this model