Hebrew Gpt Neo Small

🚀 Hebrew GPT Neo Small

A Hebrew text generation model based on EleutherAI's gpt - neo, offering unique capabilities for Hebrew text generation.

🚀 Quick Start

This is a Hebrew text generation model based on [EleutherAI's gpt - neo](https://github.com/EleutherAI/gpt - neo). Each model was trained on a TPUv3 - 8, which was made available via the TPU Research Cloud Program.

✨ Features

• Utilizes advanced architecture from EleutherAI's gpt - neo for Hebrew text generation.
• Trained on a variety of Hebrew datasets to ensure high - quality output.

📦 Installation

The installation steps can be found in the Google Colab Notebook. You can access it [here](https://colab.research.google.com/github/Norod/hebrew - gpt_neo/blob/main/hebrew - gpt_neo - small/Norod78_hebrew_gpt_neo_small_Colab.ipynb).

💻 Usage Examples

Basic Usage

The following is a simple usage sample code:

!pip install tokenizers==0.10.2 transformers==4.6.0

from transformers import AutoTokenizer, AutoModelForCausalLM
  
tokenizer = AutoTokenizer.from_pretrained("Norod78/hebrew-gpt_neo-small")
model = AutoModelForCausalLM.from_pretrained("Norod78/hebrew-gpt_neo-small", pad_token_id=tokenizer.eos_token_id)

prompt_text = "אני אוהב שוקולד ועוגות"
max_len = 512
sample_output_num = 3
seed = 1000

import numpy as np
import torch

device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
n_gpu = 0 if torch.cuda.is_available()==False else torch.cuda.device_count()

print(f"device: {device}, n_gpu: {n_gpu}")

np.random.seed(seed)
torch.manual_seed(seed)
if n_gpu > 0:
    torch.cuda.manual_seed_all(seed)

model.to(device)

encoded_prompt = tokenizer.encode(
    prompt_text, add_special_tokens=False, return_tensors="pt")

encoded_prompt = encoded_prompt.to(device)

if encoded_prompt.size()[-1] == 0:
        input_ids = None
else:
        input_ids = encoded_prompt

print("input_ids = " + str(input_ids))

if input_ids != None:
  max_len += len(encoded_prompt[0])
  if max_len > 2048:
    max_len = 2048

print("Updated max_len = " + str(max_len))

stop_token = "<|endoftext|>"
new_lines = "\n\n\n"

sample_outputs = model.generate(
    input_ids,
    do_sample=True, 
    max_length=max_len, 
    top_k=50, 
    top_p=0.95, 
    num_return_sequences=sample_output_num
)

print(100 * '-' + "\n\t\tOutput\n" + 100 * '-')
for i, sample_output in enumerate(sample_outputs):

  text = tokenizer.decode(sample_output, skip_special_tokens=True)
  
  # Remove all text after the stop token
  text = text[: text.find(stop_token) if stop_token else None]

  # Remove all text after 3 newlines
  text = text[: text.find(new_lines) if new_lines else None]

  print("\n{}: {}".format(i, text))
  print("\n" + 100 * '-')

📚 Documentation

Datasets

1. An assortment of various Hebrew corpuses - Available [here](https://mega.nz/folder/CodSSA4R#4INvMes - 56m_WUi7jQMbJQ).
2. oscar / unshuffled_deduplicated_he - [Homepage](https://oscar - corpus.com) | Dataset Permalink. The Open Super - large Crawled ALMAnaCH coRpus is a huge multilingual corpus obtained by language classification and filtering of the Common Crawl corpus using the goclassy architecture.
3. CC100 - Hebrew Dataset [Homepage](https://metatext.io/datasets/cc100 - hebrew). Created by Conneau & Wenzek et al. in 2020, this dataset is one of the 100 corpora of monolingual data that was processed from the January - December 2018 Commoncrawl snapshots from the CC - Net repository. The size of this corpus is 6.1G, in Hebrew language.

Training Config

The training configuration is available [here](https://github.com/Norod/hebrew - gpt_neo/tree/main/hebrew - gpt_neo - small/configs).

Google Colab Notebook

You can access the Google Colab Notebook [here](https://colab.research.google.com/github/Norod/hebrew - gpt_neo/blob/main/hebrew - gpt_neo - small/Norod78_hebrew_gpt_neo_small_Colab.ipynb).

📄 License

This project is licensed under the MIT license.