working

This model is a fine-tuned version of microsoft/Phi-3-mini-128k-instruct on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.6374

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: 6
• eval_batch_size: 6
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 24
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 2
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
2.6546	0.92	6	1.7189
1.2076	2.0	13	0.8973
0.7157	2.92	19	0.5511
0.4138	4.0	26	0.4499
0.4018	4.92	32	0.4044
0.3034	6.0	39	0.3793
0.3186	6.92	45	0.3645
0.2451	8.0	52	0.3590
0.2556	8.92	58	0.3660
0.1937	10.0	65	0.3825
0.1993	10.92	71	0.3782
0.1511	12.0	78	0.4275
0.1487	12.92	84	0.4234
0.1098	14.0	91	0.4876
0.1121	14.92	97	0.4675
0.0846	16.0	104	0.5187
0.0869	16.92	110	0.5365
0.0677	18.0	117	0.5372
0.0729	18.92	123	0.5639
0.0587	20.0	130	0.5773
0.0623	20.92	136	0.6006
0.0524	22.0	143	0.6098
0.0599	22.92	149	0.6101
0.0495	24.0	156	0.6204
0.0571	24.92	162	0.6297
0.0475	26.0	169	0.6353
0.0551	26.92	175	0.6374
0.0455	27.69	180	0.6374

Framework versions

• PEFT 0.10.0
• Transformers 4.39.3
• Pytorch 2.1.2
• Datasets 2.18.0
• Tokenizers 0.15.2