tutrinh/intention_speaker_task

1---2library_name: peft3tags:4- generated_from_trainer5base_model: mistralai/Mistral-7B-Instruct-v0.26model-index:7- name: speaker_task8  results: []9---10 11<!-- This model card has been generated automatically according to the information the Trainer had access to. You12should probably proofread and complete it, then remove this comment. -->13 14# speaker_task15 16This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.2](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2) on the None dataset.17It achieves the following results on the evaluation set:18- Loss: 1.042919 20## Model description21 22More information needed23 24## Intended uses & limitations25 26More information needed27 28## Training and evaluation data29 30More information needed31 32## Training procedure33 34### Training hyperparameters35 36The following hyperparameters were used during training:37- learning_rate: 3e-0538- train_batch_size: 139- eval_batch_size: 140- seed: 4241- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-0842- lr_scheduler_type: cosine43- lr_scheduler_warmup_ratio: 0.06544- num_epochs: 345- mixed_precision_training: Native AMP46 47### Training results48 49| Training Loss | Epoch | Step | Validation Loss |50|:-------------:|:-----:|:----:|:---------------:|51| No log        | 1.0   | 396  | 1.3015          |52| 1.3232        | 2.0   | 792  | 0.9507          |53| 0.4955        | 3.0   | 1188 | 1.0429          |54 55 56### Framework versions57 58- PEFT 0.7.2.dev059- Transformers 4.37.160- Pytorch 2.1.0+cu12161- Datasets 2.9.062- Tokenizers 0.15.1