Instructions to use Technoculture/Medtulu-2x7b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Technoculture/Medtulu-2x7b with Transformers:

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

pipe = pipeline("text-generation", model="Technoculture/Medtulu-2x7b")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Technoculture/Medtulu-2x7b")
model = AutoModelForCausalLM.from_pretrained("Technoculture/Medtulu-2x7b", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Technoculture/Medtulu-2x7b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Technoculture/Medtulu-2x7b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Technoculture/Medtulu-2x7b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/Technoculture/Medtulu-2x7b

SGLang

How to use Technoculture/Medtulu-2x7b 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 "Technoculture/Medtulu-2x7b" \
    --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": "Technoculture/Medtulu-2x7b",
		"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 "Technoculture/Medtulu-2x7b" \
        --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": "Technoculture/Medtulu-2x7b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use Technoculture/Medtulu-2x7b with Docker Model Runner:
```
docker model run hf.co/Technoculture/Medtulu-2x7b
```

Medtulu-2x7b / added_tokens.json

satyamt

Upload folder using huggingface_hub

4a9d3b8 verified over 2 years ago

Raw

History Blame Contribute Delete

344 Bytes

	{
	"<CLS>": 32000,
	"<EOD>": 32002,
	"<MASK>": 32003,
	"<PAD>": 32004,
	"<SEP>": 32001,
	"[/bib]": 32010,
	"[/bib_ref]": 32006,
	"[/fig]": 32012,
	"[/fig_ref]": 32008,
	"[/formula]": 32016,
	"[/table]": 32014,
	"[bib]": 32009,
	"[bib_ref]": 32005,
	"[fig]": 32011,
	"[fig_ref]": 32007,
	"[formula]": 32015,
	"[table]": 32013
	}