---
base_model: hivaze/ParaLex-Llama-3-8B-SFT
library_name: peft
tags:
- generated_from_trainer
model-index:
- name: results_packing
  results: []
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# results_packing

This model is a fine-tuned version of [hivaze/ParaLex-Llama-3-8B-SFT](https://huggingface.co/hivaze/ParaLex-Llama-3-8B-SFT) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 1.8083

## 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: 2e-05
- train_batch_size: 1
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 17
- total_train_batch_size: 17
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- num_epochs: 8

### Training results

| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 7.3306        | 1.25  | 5    | 5.9428          |
| 5.4669        | 2.5   | 10   | 4.3334          |
| 4.0282        | 3.75  | 15   | 3.1156          |
| 2.9271        | 5.0   | 20   | 2.3114          |
| 2.3074        | 6.25  | 25   | 1.9202          |
| 1.9795        | 7.5   | 30   | 1.8083          |


### Framework versions

- PEFT 0.11.1
- Transformers 4.41.2
- Pytorch 2.3.0+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1