Instructions to use nuprl/MultiPL-T-StarCoderBase_15b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use nuprl/MultiPL-T-StarCoderBase_15b with Transformers:

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

pipe = pipeline("text-generation", model="nuprl/MultiPL-T-StarCoderBase_15b")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("nuprl/MultiPL-T-StarCoderBase_15b")
model = AutoModelForCausalLM.from_pretrained("nuprl/MultiPL-T-StarCoderBase_15b")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use nuprl/MultiPL-T-StarCoderBase_15b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "nuprl/MultiPL-T-StarCoderBase_15b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "nuprl/MultiPL-T-StarCoderBase_15b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/nuprl/MultiPL-T-StarCoderBase_15b

SGLang

How to use nuprl/MultiPL-T-StarCoderBase_15b 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 "nuprl/MultiPL-T-StarCoderBase_15b" \
    --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": "nuprl/MultiPL-T-StarCoderBase_15b",
		"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 "nuprl/MultiPL-T-StarCoderBase_15b" \
        --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": "nuprl/MultiPL-T-StarCoderBase_15b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use nuprl/MultiPL-T-StarCoderBase_15b with Docker Model Runner:
```
docker model run hf.co/nuprl/MultiPL-T-StarCoderBase_15b
```

cassanof commited on Aug 19, 2023

Commit

9ee64ae

1 Parent(s): 9a56eb9

Update README.md

Browse files

Files changed (1) hide show

README.md +3 -3

README.md CHANGED Viewed

@@ -48,7 +48,7 @@ model-index:
 ---
 # MultiPLCoder-15b
-15 billion parameter version of MultiPLCoder, a set of StarCoder-based models finetuned on the MultiPL-T dataset.
 These models are state-of-the-art at low-resource languages, such as: Lua, Racket, and OCaml.
 This 15 billion parameter model is the most capable of the MultiPLCoder family. However, it requires a dedicated GPU for inference.
@@ -78,8 +78,8 @@ model = AutoModelForCausalLM.from_pretrained("nuprl/MultiPLCoder-15b", revision=
 Note that the model's default configuration does not enable caching, therefore you must specify to use the cache on generation.
 ```py
-toks = tokenizer.encode("-- Fibonacci iterative", return_tensors="pt")
-out = model.generate(toks, use_cache=True,  do_sample=True, temperature=0.2, top_p=0.95, max_length=50)
 print(tokenizer.decode(out[0], skip_special_tokens=True))
 ```
 ```

 ---
 # MultiPLCoder-15b
+15 billion parameter version of MultiPLCoder, a set of StarCoder-based models finetuned on the [MultiPL-T dataset](https://huggingface.co/datasets/nuprl/MultiPL-T).
 These models are state-of-the-art at low-resource languages, such as: Lua, Racket, and OCaml.
 This 15 billion parameter model is the most capable of the MultiPLCoder family. However, it requires a dedicated GPU for inference.
 Note that the model's default configuration does not enable caching, therefore you must specify to use the cache on generation.
 ```py
+toks = tokenizer.encode("-- Fibonacci iterative", return_tensors="pt").cuda()
+out = model.generate(toks, use_cache=True,  do_sample=True, temperature=0.2, top_p=0.95, max_length=256)
 print(tokenizer.decode(out[0], skip_special_tokens=True))
 ```
 ```