Mistral-7B-v0.1_cola_sparse_swiglu_ignore_0_1

This model is a fine-tuned version of mistralai/Mistral-7B-v0.1 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.4277
• Accuracy: {'accuracy': 0.8212616822429907}
• Matthews Correlation: 0.5699

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: 1e-05
• train_batch_size: 8
• eval_batch_size: 16
• seed: 2
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• total_eval_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Matthews Correlation
1.6028	0.17	20	1.5539	{'accuracy': 0.5417066155321189}	0.0764
0.9736	0.33	40	0.9708	{'accuracy': 0.6864813039309684}	0.1852
0.7146	0.5	60	0.7850	{'accuracy': 0.713326941514861}	0.3141
0.6892	0.66	80	0.6674	{'accuracy': 0.7238734419942474}	0.3498
0.6792	0.83	100	0.6401	{'accuracy': 0.7411313518696069}	0.3977
0.6233	1.0	120	0.6104	{'accuracy': 0.7574304889741131}	0.3784
0.4778	1.16	140	0.5641	{'accuracy': 0.7948226270373921}	0.4874
0.4792	1.33	160	0.5961	{'accuracy': 0.7746883988494727}	0.4284
0.5573	1.49	180	0.5210	{'accuracy': 0.8034515819750719}	0.5126
0.4464	1.66	200	0.5716	{'accuracy': 0.7871524448705657}	0.5601
0.4541	1.83	220	0.5130	{'accuracy': 0.8015340364333653}	0.5046
0.4989	1.99	240	0.4648	{'accuracy': 0.8149568552253116}	0.5452
0.3891	2.16	260	0.4566	{'accuracy': 0.8207094918504314}	0.5856
0.336	2.32	280	0.4516	{'accuracy': 0.822627037392138}	0.5657
0.3854	2.49	300	0.4224	{'accuracy': 0.8322147651006712}	0.6066
0.3917	2.66	320	0.4247	{'accuracy': 0.837967401725791}	0.6125
0.3779	2.82	340	0.4177	{'accuracy': 0.8302972195589645}	0.5897
0.3462	2.99	360	0.4649	{'accuracy': 0.8207094918504314}	0.5584
0.3448	3.15	380	0.4182	{'accuracy': 0.8293384467881112}	0.5837
0.3894	3.32	400	0.4388	{'accuracy': 0.8302972195589645}	0.5893

Framework versions

• PEFT 0.7.1
• Transformers 4.36.2
• Pytorch 2.1.2+cu121
• Datasets 2.15.0
• Tokenizers 0.15.0