license: apache-2.0
library_name: peft
tags:
- trl
- sft
- generated_from_trainer
base_model: mistralai/Mistral-7B-Instruct-v0.1
datasets:
- generator
model-index:
- name: Mistral-7B-v0.1_Emotion
results: []
Mistral-7B-v0.1_Emotion
This model is a fine-tuned version of mistralai/Mistral-7B-Instruct-v0.1 on the generator dataset.
Model description
Fine tunning: https://github.com/frank-morales2020/MLxDL/blob/main/FineTuning_Mistral_7b_hfdeployment_dataset_Emotion.ipynb
Evaluation: https://github.com/frank-morales2020/MLxDL/blob/main/FineTunning_Testing_For_EmotionQADataset.ipynb
Intended uses & limitations
More information needed
Training and evaluation data
Evaluation: https://github.com/frank-morales2020/MLxDL/blob/main/FineTunning_Testing_For_EmotionQADataset.ipynb
The following hyperparameters were used during training: learning_rate: 0.0002 train_batch_size: 3 eval_batch_size: 8 seed: 42 gradient_accumulation_steps: 2 total_train_batch_size: 6 optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 lr_scheduler_type: constant lr_scheduler_warmup_ratio: 0.03
num_epochs: 1
Accuracy (Eval dataset and predict) for a sample of 2000: 59.45%
num_epochs: 25
Accuracy (Eval dataset and predict) for a sample of 2000: 79.95%
num_epochs: 40
Accuracy (Eval dataset and predict) for a sample of 2000: 80.70%
num_epochs: 40
Accuracy (Eval dataset and predict) for a sample of 2000: 80%
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0002
- train_batch_size: 3
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 6
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant
- lr_scheduler_warmup_ratio: 0.03
- num_epochs: 100
Training results
Framework versions
- PEFT 0.11.1
- Transformers 4.41.2
- Pytorch 2.3.0+cu121
- Datasets 2.19.2
- Tokenizers 0.19.1