wiki_lora_hoa1b4_15_epoch

This model is a fine-tuned version of vlsp-2023-vllm/hoa-1b4 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2544

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: 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: 15

Training results

Training Loss	Epoch	Step	Validation Loss
19.6041	1.0	50	18.7356
5.3787	2.0	100	2.7051
3.4559	3.0	150	2.0077
1.7394	4.0	200	1.6610
1.4427	5.0	250	1.3440
1.0363	6.0	300	1.0267
0.9629	7.0	350	0.7544
0.5965	8.0	400	0.5325
0.4756	9.0	450	0.4114
0.299	10.0	500	0.3445
0.3162	11.0	550	0.3028
0.2772	12.0	600	0.2796
0.2617	13.0	650	0.2658
0.2485	14.0	700	0.2569
0.2272	15.0	750	0.2544

Framework versions

• PEFT 0.12.0
• Transformers 4.44.0
• Pytorch 2.4.0
• Datasets 2.21.0
• Tokenizers 0.19.1