MSc_llama3_finetuned_model_secondData

This model is a fine-tuned version of meta-llama/Meta-Llama-3-8B-Instruct on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.5909

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: 0.0001
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.03
• training_steps: 250

Training results

Training Loss	Epoch	Step	Validation Loss
3.3698	1.36	10	2.0432
1.3777	2.71	20	1.0067
0.8126	4.07	30	0.7822
0.6642	5.42	40	0.7281
0.5708	6.78	50	0.7218
0.5062	8.14	60	0.7360
0.4379	9.49	70	0.7781
0.3924	10.85	80	0.8310
0.3435	12.2	90	0.8856
0.3041	13.56	100	1.0389
0.2787	14.92	110	1.0664
0.2553	16.27	120	1.1655
0.2388	17.63	130	1.2397
0.2288	18.98	140	1.2049
0.2128	20.34	150	1.2746
0.2081	21.69	160	1.3889
0.1998	23.05	170	1.3942
0.1909	24.41	180	1.4383
0.188	25.76	190	1.5012
0.1841	27.12	200	1.5246
0.18	28.47	210	1.5528
0.1794	29.83	220	1.5662
0.1773	31.19	230	1.5788
0.1751	32.54	240	1.5889
0.1756	33.9	250	1.5909

Framework versions

• PEFT 0.4.0
• Transformers 4.38.2
• Pytorch 2.4.0+cu121
• Datasets 2.13.1
• Tokenizers 0.15.2