Instructions to use PosteriorAI/dravida_llama2_7b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use PosteriorAI/dravida_llama2_7b with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="PosteriorAI/dravida_llama2_7b")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("PosteriorAI/dravida_llama2_7b")
model = AutoModelForCausalLM.from_pretrained("PosteriorAI/dravida_llama2_7b")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use PosteriorAI/dravida_llama2_7b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "PosteriorAI/dravida_llama2_7b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "PosteriorAI/dravida_llama2_7b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/PosteriorAI/dravida_llama2_7b

SGLang

How to use PosteriorAI/dravida_llama2_7b with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "PosteriorAI/dravida_llama2_7b" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "PosteriorAI/dravida_llama2_7b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "PosteriorAI/dravida_llama2_7b" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "PosteriorAI/dravida_llama2_7b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use PosteriorAI/dravida_llama2_7b with Docker Model Runner:
```
docker model run hf.co/PosteriorAI/dravida_llama2_7b
```

Model Details

Model Description

The Dravida Llama is a state-of-the-art Multilingual Large Language Model (LLM) fine-tuned for the four South Indian languages (KaTeMaTa) using Meta's Llama-2 as a foundation.

Developed by: PosteriorAI
Model type: Large Language Model (LLM), specifically fine-tuned Llama-2 model for Kannada, Telugu, Malayalam & Tamil.
Language(s) (NLP): Kannada, Tamil, Malayalam & Telugu.
License: Open-source releases on Hugging Face, MIT Licensed.
Finetuned from model [optional]: Llama-2

Model Sources [optional]

Repository: PosteriorAI//dravida_llama2_7b

Uses

This model is intended for various stakeholders, including researchers, developers, and the broader community. It aims to enhance communication, education, and technology access by providing an AI tool that understands and interacts in the four languages. The model's applications range from personal assistance to educational content creation and more. It addresses the gap in AI for Indic languages and is designed to promote inclusivity in technology.

Find mode details in our blog post at Dravida Llama: LLM for South Indian (ಕతెമத) Languages

Downloads last month: 5

Safetensors

Model size

7B params

Tensor type

F16

PosteriorAI
/

dravida_llama2_7b

Model Details

Model Description

Model Sources [optional]

Uses

Dataset used to train PosteriorAI/dravida_llama2_7b