llamantino7b_2_10_question-answering

This model is a fine-tuned version of swap-uniba/LLaMAntino-2-7b-hf-ITA on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.9768
• Rouge1: 19.35
• Rouge2: 10.89
• Rougel: 18.01
• Rougelsum: 18.19
• R: 16.06
• Gen Len: 1.0
• R@1: 0.0
• R@3: 0.0
• R@5: 0.0
• R@10: 0.0
• R@20: 0.0

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: 0.0002
• 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
• lr_scheduler_warmup_ratio: 0.03
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	R	Gen Len	R@1	R@3	R@5	R@10	R@20
1.7082	1.0	23	1.4263	16.0	5.83	14.16	14.7	11.97	1.0	0.0	0.0	0.0	0.0	0.0
1.1279	2.0	46	1.3286	17.65	8.25	16.23	16.85	14.02	1.0	0.0	0.0	0.0	0.0	0.0
0.6508	3.0	69	1.3710	18.49	9.03	16.79	17.44	14.74	1.0	0.0	0.0	0.0	0.0	0.0
0.3175	4.0	92	1.4644	19.18	9.83	17.33	17.52	15.42	1.0	0.0	0.0	0.0	0.0	0.0
0.1619	5.0	115	1.5859	18.93	10.18	17.36	17.62	15.47	1.0	0.0	0.0	0.0	0.0	0.0
0.0925	6.0	138	1.6658	19.06	10.78	17.65	17.86	15.81	1.0	0.0	0.0	0.0	0.0	0.0
0.0571	7.0	161	1.7225	18.85	10.28	17.28	17.54	15.45	1.0	0.0	0.0	0.0	0.0	0.0
0.0279	8.0	184	1.8907	18.94	10.96	17.61	17.74	15.82	1.0	0.0	0.0	0.0	0.0	0.0
0.0172	9.0	207	1.9558	19.32	10.79	18.01	18.18	16.02	1.0	0.0	0.0	0.0	0.0	0.0
0.0127	10.0	230	1.9768	19.35	10.89	18.01	18.19	16.06	1.0	0.0	0.0	0.0	0.0	0.0

Framework versions

• PEFT 0.8.2
• Transformers 4.38.0.dev0
• Pytorch 2.0.1+cu117
• Datasets 2.16.1
• Tokenizers 0.15.2