kjh01's picture
Udate README.md
3e2b144
metadata
base_model: meta-llama/Llama-2-7b-chat-hf
tags:
  - generated_from_trainer
model-index:
  - name: dataset_infos_llama_2
    results: []

dataset_infos_llama_2

This model is a fine-tuned version of meta-llama/Llama-2-7b-chat-hf on an unknown dataset.

Model description

Llama-2-7b-chat-hf은 metaμ—μ„œ κ°œλ°œν•œ μ‚¬μ „ν•™μŠ΅ ν…μŠ€νŠΈ 생성 μ–Έμ–΄λͺ¨λΈ μž…λ‹ˆλ‹€. λ¬Έμžμ—΄μ„ μž…λ ₯으둜 ν•˜λ©°, λ¬Έμžμ—΄μ„ μƒμ„±ν•©λ‹ˆλ‹€. ν•΄λ‹Ή λͺ¨λΈ(meta-llama/Llama-2-7b-chat-hf)을 베이슀 λͺ¨λΈλ‘œ ν•˜μ—¬ λ―Έμ„ΈνŠœλ‹μ„ μ§„ν–‰ν•˜μ˜€μŠ΅λ‹ˆλ‹€.

'Llama-2-7b-chat-hf' is a pre-trained text generation language model developed by Meta. It takes a string as input and generates text. We fine-tuned this model based on it(meta-llama/Llama-2-7b-chat-hf).

Intended uses & limitations

nsmc λ°μ΄ν„°μ…‹μ˜ μ‚¬μš©μžκ°€ μž…λ ₯ν•œ 리뷰 λ¬Έμž₯을 λΆ„λ₯˜ν•˜λŠ” μ—μ΄μ „νŠΈμž…λ‹ˆλ‹€. μ‚¬μš©μž 리뷰 λ¬Έμž₯μœΌλ‘œλΆ€ν„° '긍정' λ˜λŠ” 'λΆ€μ •'을 νŒλ‹¨ν•©λ‹ˆλ‹€.

This agent classifies user-input review sentences from NSMC dataset. It determines whether the user review is 'positive' or 'negative' based on the input review sentence.

Training and test data

Training 및 test λ°μ΄ν„°λŠ” nsmc 데이터 μ…‹μ—μ„œ λ‘œλ”©ν•΄ μ‚¬μš©ν•©λ‹ˆλ‹€. (elvaluation λ°μ΄ν„°λŠ” μ‚¬μš©ν•˜μ§€ μ•ŠμŠ΅λ‹ˆλ‹€.)

We load and use training and test data from the NSMC dataset. (We do not use an evaluation data.)

Training procedure

μ‚¬μš©μžμ˜ μ˜ν™” 리뷰 λ¬Έμž₯을 μž…λ ₯으둜 λ°›μ•„ λ¬Έμž₯을 '긍정(1)' λ˜λŠ” 'λΆ€μ •(0)'으둜 λΆ„λ₯˜ν•©λ‹ˆλ‹€.

Accepts movie review sentences from the user as input and classifies the sentences as 'Positive (1)' or 'Negative (0)'.

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 0.0001
  • train_batch_size: 1
  • eval_batch_size: 1
  • seed: 42
  • gradient_accumulation_steps: 2
  • total_train_batch_size: 2
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_ratio: 0.03
  • training_steps: 900
  • mixed_precision_training: Native AMP

Training results

  • Binary Confusion Matrix | | TP | TN |

|:-----|:------------:|:------------:| | PP | 425 | 67 | | PN | 66 | 442 |

  • Accuracy: 0.894

Framework versions

  • Transformers 4.35.2
  • Pytorch 2.1.0+cu118
  • Datasets 2.15.0
  • Tokenizers 0.15.0