theoracle's picture
Update README.md
624e3e1 verified
metadata
title: Gemma Italian Camoscio Language Model
tags:
  - italian-language-generation
  - camoscio-dataset
  - gemma-2b
  - autotrain
datasets:
  - camoscio
library_name: transformers
model: theoracle/gemma_italian_camoscio
license: other

Overview

theoracle/gemma_italian_camoscio is a cutting-edge model specifically designed for Italian language generation. Leveraging the comprehensive Camoscio dataset, this model enhances the Gemma 2B architecture's capabilities in producing high-quality, contextually accurate Italian text. Developed with AutoTrain, it excels in various Italian text generation tasks, including but not limited to creative writing, article generation, and conversational responses.

Key Features

  • Italian Language Focus: Tailored to understand and generate Italian text, capturing the language's nuances and complexities.
  • Camoscio Dataset Training: Utilizes the rich Camoscio dataset, ensuring the model is well-versed in a wide range of Italian language styles and contexts.
  • Gemma 2B Architecture: Built on the powerful Gemma 2B framework, known for its efficiency and effectiveness in language generation tasks.
  • AutoTrain Enhanced: Benefits from AutoTrain's optimization, making the model both robust and versatile in handling Italian text generation.

Usage

Here's how to use this model for generating Italian text:

from transformers import AutoModelForCausalLM, AutoTokenizer

model_path = "theoracle/gemma_italian_camoscio"
tokenizer = AutoTokenizer.from_pretrained(model_path)
model = AutoModelForCausalLM.from_pretrained(
    model_path,
    device_map="auto",
    torch_dtype='auto'
).eval()

# Example: Generating Italian text
prompt = "Inizia la storia con una giornata soleggiata in Sicilia, dove"

# Tokenize and generate text
encoding = tokenizer(prompt, return_tensors='pt', padding=True, truncation=True, max_length=500, add_special_tokens=True)
input_ids = encoding['input_ids']
attention_mask = encoding['attention_mask']

output_ids = model.generate(
    input_ids.to('cuda'),
    attention_mask=attention_mask.to('cuda'),
    max_new_tokens=300,
    pad_token_id=tokenizer.eos_token_id
)

generated_text = tokenizer.decode(output_ids[0], skip_special_tokens=True)
print(generated_text)

Application Scenarios

This model is ideal for:

  • Content creators looking to produce Italian-language articles, stories, or scripts.
  • Developers creating conversational AI applications in Italian.
  • Educators and language learners seeking tools for Italian language practice and improvement.

Training and Technology

The theoracle/gemma_italian_camoscio model is trained using the AutoTrain platform for optimal performance, ensuring that it is well-suited for a broad spectrum of Italian text generation tasks. The Camoscio dataset provides a solid foundation, offering diverse and extensive coverage of the Italian language, which, combined with the Gemma 2B architecture, enables the model to generate coherent, nuanced, and contextually relevant Italian text.

License

This model is available under an "other" license, facilitating its use in a wide array of applications. Users are encouraged to review the license terms to ensure compliance with their project requirements and intended use cases.