metadata

tags:
  - trl
  - sft
  - generated_from_trainer
  - Text Generation
  - llama
  - t5
model-index:
  - name: Prompt-Enhace-T5-base
    results: []
datasets:
  - gokaygokay/prompt-enhancer-dataset
license: apache-2.0
language:
  - en
base_model: google-t5/t5-base
library_name: transformers

omersaidd / Prompt-Enhace-T5-base

This model was trained from scratch on an gokaygokay/prompt-enhancer-dataset dataset.

Bu modelin eğitiminde gokaygokay/prompt-enhancer-dataset veriseti kullanılmşıtır

Model description

This model is trained with the google/t5-base and the database on prompt generation.

Bu model google/t5-base ile prompt üretimek üzerine veriseti ile eğitilmişitir

Intended uses & limitations

More information needed

Training and evaluation data

Kullandığımız verisetimiz gokaygokay/prompt-enhancer-dataset

Our dataset we use gokaygokay/prompt-enhancer-dataset

Training hyperparameters

Eğitim sırasında aşağıdaki hiperparametreler kullanılmıştır:

The following hyperparameters were used during training:

learning_rate: 3e-6
train_batch_size: 256
eval_batch_size: 256
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 500
num_epochs: 3

Framework versions

Transformers 4.43.1
Pytorch 2.1.2+cu121
Datasets 2.19.1
Tokenizers 0.19.1

Test Model Code

model = AutoModelForSeq2SeqLM.from_pretrained(model_checkpoint)

enhancer = pipeline('text2text-generation',
                    model=model,
                    tokenizer=tokenizer,
                    repetition_penalty= 1.2,
                    device=device)

max_target_length = 256
prefix = "enhance prompt: "

short_prompt = "beautiful house with text 'hello'"
answer = enhancer(prefix + short_prompt, max_length=max_target_length)
final_answer = answer[0]['generated_text']
print(final_answer)