flan-t5-base-Mistral7BI-LORA-V1

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: 0.6273
• Exact Match: 32.8431
• Gen Len: 3.8500

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: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 12

Training results

Training Loss	Epoch	Step	Validation Loss	Exact Match	Gen Len
0.6768	1.0	4246	0.7026	24.8937	3.9547
0.6585	2.0	8492	0.6736	27.4687	3.9347
0.6501	3.0	12738	0.6543	29.961	3.8808
0.7615	4.0	16984	0.6446	29.7189	3.8876
0.5709	5.0	21230	0.6413	30.9828	3.8304
0.6062	6.0	25476	0.6387	31.7328	3.8798
0.5539	7.0	29722	0.6318	31.4907	3.8315
0.5708	8.0	33968	0.6319	32.2939	3.8669
0.6859	9.0	38214	0.6269	32.6069	3.8267
0.6026	10.0	42460	0.6271	32.6246	3.8338
0.5558	11.0	46706	0.6272	32.7014	3.8460
0.566	12.0	50952	0.6273	32.8431	3.8500

Framework versions

• PEFT 0.11.1
• Transformers 4.41.2
• Pytorch 2.2.1
• Datasets 2.19.2
• Tokenizers 0.19.1