IC_ver6F_coco_swin_gpt2_50B_1e

This model is a fine-tuned version of VK246/IC_ver6e_coco_swin_gpt2_50Apc_1e on the coco dataset. It achieves the following results on the evaluation set:

• Loss: 0.7799
• Cider: 5.8986
• Rouge1: 42.1787
• Rouge2: 16.6289
• Rougel: 38.245
• Rougelsum: 38.236
• Bleu-1: 43.2152
• Bleu-2: 25.0563
• Bleu-3: 15.845
• Bleu-4: 10.5042
• Gen Len: 11.3063

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: 5e-05
• train_batch_size: 96
• eval_batch_size: 96
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Cider	Rouge1	Rouge2	Rougel	Rougelsum	Bleu-1	Bleu-2	Bleu-3	Bleu-4	Gen Len
0.6972	0.34	1000	0.8128	5.8314	41.3992	16.1278	37.5675	37.5537	42.6637	24.5815	15.5018	10.2465	11.3063
0.7318	0.68	2000	0.7912	6.9716	41.8244	16.3282	37.9594	37.9525	42.7623	24.7305	15.6458	10.4067	11.3063

Framework versions

• Transformers 4.31.0
• Pytorch 2.0.1+cu118
• Datasets 2.14.4
• Tokenizers 0.13.3