Instructions to use GenVRadmin/AryaBhatta-GemmaUltra-Merged with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use GenVRadmin/AryaBhatta-GemmaUltra-Merged with Transformers:

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

pipe = pipeline("text-generation", model="GenVRadmin/AryaBhatta-GemmaUltra-Merged")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("GenVRadmin/AryaBhatta-GemmaUltra-Merged")
model = AutoModelForCausalLM.from_pretrained("GenVRadmin/AryaBhatta-GemmaUltra-Merged")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use GenVRadmin/AryaBhatta-GemmaUltra-Merged with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/GenVRadmin/AryaBhatta-GemmaUltra-Merged

SGLang

How to use GenVRadmin/AryaBhatta-GemmaUltra-Merged 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 "GenVRadmin/AryaBhatta-GemmaUltra-Merged" \
    --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": "GenVRadmin/AryaBhatta-GemmaUltra-Merged",
		"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 "GenVRadmin/AryaBhatta-GemmaUltra-Merged" \
        --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": "GenVRadmin/AryaBhatta-GemmaUltra-Merged",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use GenVRadmin/AryaBhatta-GemmaUltra-Merged with Docker Model Runner:
```
docker model run hf.co/GenVRadmin/AryaBhatta-GemmaUltra-Merged
```

Base model: CorticalStack/gemma-7b-ultrachat-sft

This is finetuned from above base model and to be used for multi-turn chat based use-cases. Unlike our AryaBhatta-GemmaOrca model which is skilled in science, literature and finetuned on Orca datasets, this model is fine-tuned on Ultra-Chat datasets. And show improved performance over AryaBhatta-GemmaOrca on Hellaswag datasets and in multi-turn conversations. It is finetuned on 9 Indian languages (Hindi, Tamil, Punjabi, Bengali, Gujarati, Oriya, Telugu, Kannada, Malayalam) plus English.

Benchmarked on Indic LLM leaderboard: https://huggingface.co/spaces/Cognitive-Lab/indic_llm_leaderboard

Release post: https://www.linkedin.com/feed/update/urn:li:activity:7184856055565180928

Downloads last month: 14

Safetensors

Model size

9B params

Tensor type

F16