Modelpublic

RyanZZZZZ/mental_health_llm_baseline_sft_Llama-3.1-8B_lora_QA_YT_0223

sourceHugging Facellama3.1updated 7mo agoView on Hugging Face

0likes10downloads

Model Card

mentalhealthllmbaselinesftLlama-3.1-8BloraYTs2_0223-3

This model is a fine-tuned version of meta-llama/Llama-3.1-8B-Instruct on the None dataset. It achieves the following results on the evaluation set:

—Loss: 2.3455
—Bleu: 0.0104
—Rougel: 0.2219
—Meteor: 0.2155
—Bertscore Precision: 0.6791
—Bertscore Recall: 0.8505
—Bertscore F1: 0.7545

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
—trainbatchsize: 8
—evalbatchsize: 8
—seed: 42
—optimizer: Use OptimizerNames.ADAMWTORCHFUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
—lrschedulertype: linear
—lrschedulerwarmup_ratio: 0.05
—num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Rougel	Meteor	Bertscore Precision	Bertscore Recall	Bertscore F1
2.671	0.1961	10	2.6227	0.0093	0.2141	0.2037	0.6797	0.8467	0.7532
2.5343	0.3922	20	2.5686	0.0095	0.2178	0.2055	0.6789	0.8468	0.7527
2.5861	0.5882	30	2.5097	0.0094	0.2197	0.2031	0.6735	0.8479	0.7498
2.3769	0.7843	40	2.4689	0.0093	0.2183	0.1990	0.6695	0.8467	0.7468
2.4214	0.9804	50	2.4345	0.0099	0.2197	0.2060	0.6706	0.8479	0.7481
2.4351	1.1765	60	2.4094	0.0104	0.2211	0.2123	0.6764	0.8497	0.7524
2.4466	1.3725	70	2.3896	0.0101	0.2172	0.2142	0.6792	0.8499	0.7543
2.3013	1.5686	80	2.3772	0.0107	0.2185	0.2178	0.6791	0.8506	0.7545
2.3168	1.7647	90	2.3673	0.0106	0.2181	0.2158	0.6811	0.8501	0.7556
2.286	1.9608	100	2.3601	0.0107	0.2206	0.2201	0.6813	0.8504	0.7559
2.3821	2.1569	110	2.3553	0.0105	0.2211	0.2173	0.6808	0.8509	0.7557
2.306	2.3529	120	2.3513	0.0105	0.2216	0.2173	0.6810	0.8502	0.7555
2.3682	2.5490	130	2.3486	0.0105	0.2225	0.2159	0.6806	0.8501	0.7552
2.288	2.7451	140	2.3457	0.0106	0.2241	0.2174	0.6791	0.8509	0.7546
2.2992	2.9412	150	2.3455	0.0104	0.2219	0.2155	0.6791	0.8505	0.7545

Framework versions

—PEFT 0.18.1
—Transformers 4.57.3
—Pytorch 2.10.0+cu128
—Datasets 4.0.0
—Tokenizers 0.22.2