Instructions to use l3cube-pune/marathi-gpt-gemma-7b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use l3cube-pune/marathi-gpt-gemma-7b with Transformers:

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

pipe = pipeline("text-generation", model="l3cube-pune/marathi-gpt-gemma-7b")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("l3cube-pune/marathi-gpt-gemma-7b")
model = AutoModelForCausalLM.from_pretrained("l3cube-pune/marathi-gpt-gemma-7b")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use l3cube-pune/marathi-gpt-gemma-7b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "l3cube-pune/marathi-gpt-gemma-7b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "l3cube-pune/marathi-gpt-gemma-7b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/l3cube-pune/marathi-gpt-gemma-7b

SGLang

How to use l3cube-pune/marathi-gpt-gemma-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 "l3cube-pune/marathi-gpt-gemma-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": "l3cube-pune/marathi-gpt-gemma-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 "l3cube-pune/marathi-gpt-gemma-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": "l3cube-pune/marathi-gpt-gemma-7b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use l3cube-pune/marathi-gpt-gemma-7b with Docker Model Runner:
```
docker model run hf.co/l3cube-pune/marathi-gpt-gemma-7b
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

MahaGemma-7B

MahaGemma-7B is a Marathi Gemma model. It is a Gemma 7B (google/gemma-7b) model LoRA fine-tuned on translated Marathi datasets. [dataset link] (https://github.com/l3cube-pune/MarathiNLP)

This is part of the MahaNLP initiative. More details coming soon.

Prompt format:

<bos>\n### Instruction:\nमहाराष्ट्राची राजधानी काय आहे?\n\n### Input:\n\n\n### Response:\nमहाराष्ट्राची राजधानी मुंबई आहे

Citing

@article{joshi2022l3cube,
  title={L3cube-mahanlp: Marathi natural language processing datasets, models, and library},
  author={Joshi, Raviraj},
  journal={arXiv preprint arXiv:2205.14728},
  year={2022}
}

Model Family:
MahaGemma-2B
MahaGemma-7B

Downloads last month: 24

Safetensors

Model size

9B params

Tensor type

BF16

Model tree for l3cube-pune/marathi-gpt-gemma-7b

Quantizations

1 model

Space using l3cube-pune/marathi-gpt-gemma-7b 1

Paper for l3cube-pune/marathi-gpt-gemma-7b

L3Cube-MahaNLP: Marathi Natural Language Processing Datasets, Models, and Library

Paper • 2205.14728 • Published May 29, 2022