metadata

license: apache-2.0
base_model: mistralai/Mistral-7B-Instruct-v0.2
tags:
  - generated_from_trainer
model-index:
  - name: Mistral-7B-Instruct-v0.2-absa-MT-laptops
    results: []

Mistral-7B-Instruct-v0.2-absa-MT-laptops

This model is a fine-tuned version of mistralai/Mistral-7B-Instruct-v0.2 on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.0060

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: 3e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 32
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 2
training_steps: 1200
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.8786	0.13	40	0.1392
0.0627	0.25	80	0.0165
0.0162	0.38	120	0.0143
0.0139	0.5	160	0.0125
0.0131	0.63	200	0.0110
0.0115	0.75	240	0.0106
0.0111	0.88	280	0.0105
0.0091	1.0	320	0.0093
0.0073	1.13	360	0.0090
0.0079	1.25	400	0.0090
0.0068	1.38	440	0.0083
0.0065	1.5	480	0.0076
0.0071	1.63	520	0.0076
0.0062	1.75	560	0.0077
0.0062	1.88	600	0.0069
0.0058	2.0	640	0.0069
0.0034	2.13	680	0.0070
0.0034	2.25	720	0.0066
0.0034	2.38	760	0.0071
0.0038	2.5	800	0.0064
0.0032	2.63	840	0.0070
0.0031	2.75	880	0.0062
0.0032	2.88	920	0.0058
0.0026	3.0	960	0.0059
0.0018	3.13	1000	0.0058
0.0014	3.26	1040	0.0059
0.0014	3.38	1080	0.0060
0.0012	3.51	1120	0.0060
0.0014	3.63	1160	0.0060
0.001	3.76	1200	0.0060

Framework versions

Transformers 4.38.2
Pytorch 2.2.1+cu121
Datasets 2.18.0
Tokenizers 0.15.2