Instructions to use NewstaR/Starlight-7B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use NewstaR/Starlight-7B with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="NewstaR/Starlight-7B")# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("NewstaR/Starlight-7B") model = AutoModelForCausalLM.from_pretrained("NewstaR/Starlight-7B") - Notebooks
- Google Colab
- Kaggle
- Local Apps
- vLLM
How to use NewstaR/Starlight-7B with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "NewstaR/Starlight-7B" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "NewstaR/Starlight-7B", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/NewstaR/Starlight-7B
- SGLang
How to use NewstaR/Starlight-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 "NewstaR/Starlight-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": "NewstaR/Starlight-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 "NewstaR/Starlight-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": "NewstaR/Starlight-7B", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use NewstaR/Starlight-7B with Docker Model Runner:
docker model run hf.co/NewstaR/Starlight-7B
Adding Evaluation Results
#1
by leaderboard-pr-bot - opened
README.md
CHANGED
|
@@ -115,4 +115,17 @@ license: other
|
|
| 115 |
archivePrefix={arXiv},
|
| 116 |
primaryClass={cs.CL}
|
| 117 |
}
|
| 118 |
-
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 115 |
archivePrefix={arXiv},
|
| 116 |
primaryClass={cs.CL}
|
| 117 |
}
|
| 118 |
+
```
|
| 119 |
+
# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
|
| 120 |
+
Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_NewstaR__Starlight-7B)
|
| 121 |
+
|
| 122 |
+
| Metric | Value |
|
| 123 |
+
|-----------------------|---------------------------|
|
| 124 |
+
| Avg. | 43.42 |
|
| 125 |
+
| ARC (25-shot) | 53.07 |
|
| 126 |
+
| HellaSwag (10-shot) | 78.57 |
|
| 127 |
+
| MMLU (5-shot) | 46.8 |
|
| 128 |
+
| TruthfulQA (0-shot) | 38.75 |
|
| 129 |
+
| Winogrande (5-shot) | 74.03 |
|
| 130 |
+
| GSM8K (5-shot) | 7.13 |
|
| 131 |
+
| DROP (3-shot) | 5.59 |
|