Instructions to use agentica-org/DeepScaleR-1.5B-Preview with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use agentica-org/DeepScaleR-1.5B-Preview with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="agentica-org/DeepScaleR-1.5B-Preview")# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("agentica-org/DeepScaleR-1.5B-Preview") model = AutoModelForCausalLM.from_pretrained("agentica-org/DeepScaleR-1.5B-Preview") - Inference
- Notebooks
- Google Colab
- Kaggle
- Local Apps
- vLLM
How to use agentica-org/DeepScaleR-1.5B-Preview with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "agentica-org/DeepScaleR-1.5B-Preview" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "agentica-org/DeepScaleR-1.5B-Preview", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/agentica-org/DeepScaleR-1.5B-Preview
- SGLang
How to use agentica-org/DeepScaleR-1.5B-Preview 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 "agentica-org/DeepScaleR-1.5B-Preview" \ --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": "agentica-org/DeepScaleR-1.5B-Preview", "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 "agentica-org/DeepScaleR-1.5B-Preview" \ --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": "agentica-org/DeepScaleR-1.5B-Preview", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use agentica-org/DeepScaleR-1.5B-Preview with Docker Model Runner:
docker model run hf.co/agentica-org/DeepScaleR-1.5B-Preview
Why use a small model like the 1.5B? Instead of a larger one? Is there a reason?
#15
by likewendy - opened
Why use a small model like the 1.5B? Instead of a larger one? Is there a reason?
That's surely about training cost @likewendy , always better to experiment on smaller and if promising, go bigger. I read somewhere that really small LMs can struggle to pick up the RL. I think they targeted a model just above this limit.
I see! I thought of many reasons, but the only one I hadn’t considered was money.
... I hadn’t considered was money.
lol
hahaha