Bggpt 7B Instruct V0.1

🚀 INSAIT-Institute/BgGPT-7B-Instruct-v0.1

Meet BgGPT-7B, a Bulgarian language model trained from mistralai/Mistral-7B-v0.1 and distributed under Apache 2.0 license.

This model was created by INSAIT Institute, part of Sofia University, in Sofia, Bulgaria.

🚀 Quick Start

Use in Transformers

First, install the direct dependencies:

pip install transformers torch accelerate

If you want faster inference using flash-attention2, you need to install these dependencies:

pip install packaging ninja
pip install flash-attn

Then, load the model in transformers:

from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
model = AutoModelForCausalLM.from_pretrained(
    model="INSAIT-Institute/BgGPT-7B-Instruct-v0.1",
    device_map="auto",
    torch_dtype=torch.bfloat16,
    use_flash_attn_2=True # optional
)

Use with GGML / llama.cpp

The model in GGUF format INSAIT-Institute/BgGPT-7B-Instruct-v0.1-GGUF

✨ Features

• Trained from mistralai/Mistral-7B-v0.1 to enhance Bulgarian language capabilities.
• Distributed under the Apache 2.0 license.
• The tokenizer is extended for more efficient encoding of Bulgarian words in Cyrillic, improving throughput and performance.

📦 Installation

Transformers Installation

pip install transformers torch accelerate

Optional Flash-Attention2 Installation

pip install packaging ninja
pip install flash-attn

💻 Usage Examples

Instruction Format

In order to leverage instruction fine-tuning, your prompt should be surrounded by [INST] and [/INST] tokens. The very first instruction should begin with a begin of sentence token <s>. Following instructions should not. The assistant generation will be ended by the end-of-sentence token.

E.g.

text = "<s>[INST] –ö–æ–≥–∞ –µ –æ—Å–Ω–æ–≤–∞–Ω –°–æ—Ñ–∏–π—Å–∫–∏—è—Ç —É–Ω–∏–≤–µ—Ä—Å–∏—Ç–µ—Ç? [/INST]"
"Софийският университет „Св. Климент Охридски“ е създаден на 1 октомври 1888 г.</s> "
"[INST] –ö–æ–π –≥–æ –µ –æ—Å–Ω–æ–≤–∞–ª? [/INST]"

This format is available as a chat template via the apply_chat_template() method.

📚 Documentation

Model description

The model is fine-tuned to improve its Bulgarian language capabilities using multiple datasets, including Bulgarian web crawl data, a range of specialized Bulgarian datasets sourced by INSAIT Institute, and machine translations of popular English datasets. This Bulgarian data was augmented with English datasets to retain English and logical reasoning skills.

The model's tokenizer has been extended to allow for a more efficient encoding of Bulgarian words written in Cyrillic. This not only increases throughput of Cyrillic text but also performance.

Benchmarks

The model comes with a set of Benchmarks that are translations of the corresponding English-benchmarks. These are provided at https://github.com/insait-institute/lm-evaluation-harness-bg

📄 License

This model is distributed under the Apache 2.0 license.

📋 Summary