Instructions to use facebook/opt-30b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use facebook/opt-30b with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="facebook/opt-30b")# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("facebook/opt-30b") model = AutoModelForCausalLM.from_pretrained("facebook/opt-30b") - Notebooks
- Google Colab
- Kaggle
- Local Apps
- vLLM
How to use facebook/opt-30b with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "facebook/opt-30b" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "facebook/opt-30b", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/facebook/opt-30b
- SGLang
How to use facebook/opt-30b 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 "facebook/opt-30b" \ --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": "facebook/opt-30b", "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 "facebook/opt-30b" \ --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": "facebook/opt-30b", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use facebook/opt-30b with Docker Model Runner:
docker model run hf.co/facebook/opt-30b
Adding `safetensors` variant of this model
#30 opened about 2 years ago
by
SFconvertbot
[AUTOMATED] Model Memory Requirements
#29 opened about 2 years ago
by
model-sizer-bot
Adding Evaluation Results
#28 opened over 2 years ago
by
leaderboard-pr-bot
Add evaluation results on the top_en_ config and test split of futin/feed
#27 opened over 3 years ago
by
autoevaluator
Add evaluation results on the sen_en_ config and test split of futin/feed
#26 opened over 3 years ago
by
autoevaluator
Add evaluation results on the top_en config and test split of futin/feed
#25 opened over 3 years ago
by
autoevaluator
Add evaluation results on the sen_en config and test split of futin/feed
#24 opened over 3 years ago
by
autoevaluator
Add evaluation results on the top_vi config and test split of futin/feed
#23 opened over 3 years ago
by
autoevaluator
Add evaluation results on the sen_vi config and test split of futin/feed
#22 opened over 3 years ago
by
autoevaluator
Add evaluation results on the vi config and test split of futin/guess
#21 opened over 3 years ago
by
autoevaluator
Add evaluation results on the vi_3 config and test split of futin/guess
#20 opened over 3 years ago
by
autoevaluator
Add evaluation results on the en config and test split of futin/guess
#19 opened over 3 years ago
by
autoevaluator
Add evaluation results on the en_3 config and test split of futin/guess
#18 opened over 3 years ago
by
autoevaluator
Script/example for converting metaseq PyTorch checkpoints to HF
#17 opened over 3 years ago
by
diwank
Remove unused `activation_dropout`
#8 opened over 3 years ago
by
shijie-wu
Possibility to download backend-specific weights only
#7 opened almost 4 years ago
by
dennlinger
Bug fix!
#3 opened almost 4 years ago
by
patrickvonplaten