Instructions to use Yellow-AI-NLP/komodo-7b-base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Yellow-AI-NLP/komodo-7b-base with Transformers:

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

pipe = pipeline("text-generation", model="Yellow-AI-NLP/komodo-7b-base")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Yellow-AI-NLP/komodo-7b-base")
model = AutoModelForCausalLM.from_pretrained("Yellow-AI-NLP/komodo-7b-base")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Yellow-AI-NLP/komodo-7b-base with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Yellow-AI-NLP/komodo-7b-base"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Yellow-AI-NLP/komodo-7b-base",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/Yellow-AI-NLP/komodo-7b-base

SGLang

How to use Yellow-AI-NLP/komodo-7b-base 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 "Yellow-AI-NLP/komodo-7b-base" \
    --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": "Yellow-AI-NLP/komodo-7b-base",
		"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 "Yellow-AI-NLP/komodo-7b-base" \
        --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": "Yellow-AI-NLP/komodo-7b-base",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use Yellow-AI-NLP/komodo-7b-base with Docker Model Runner:
```
docker model run hf.co/Yellow-AI-NLP/komodo-7b-base
```

weird response when using vllm

by raihan2345 - opened May 30, 2024

Discussion

raihan2345

May 30, 2024

I'm using vllm to infer komodo model but I got a weird response.

prompt:

Candi borobudur adalah

response:

salahsatudaridelapan candiyang ada diindonesia. Candi borobudursendiriberada di magelang, jawatengah, diIndonesia
Candi borobudurmemilikitinggi 34,2 meter danpanjang 51 meter. Terdapatjuga 3,4juta ukirandalam candi borobudur. Disekitar candi borobudurjugaterdapat sebuhtaman.
Pada abad ke sembilanatau sebelas masehi, candi borobudurmerupakansebuah candiyangmemiliki kuil. Di kuilnyaterdapatberbagai relief dan arca.Namun candi borobudurrusakolehgempa ditahun1117 Masehi.
Sejaktahun1911, candi borobudurmulai di bangunkembali. Candi borobudursendiriselesaidibangunpadatahun1923 Masehi.
Awalnya candi borobudur dibuatuntuktempat berdoabagiumat budha.Namunsekarang candi borobudurbukanhanyauntukumat budhamelainkan paraumatagamalainnyapundapat berdoa di candi borobudur.
Ada sejutaceritayang ada di candi borobudur. Candi borobudursendiritidaklahpernah di isukanolehsiapapun.Akhirnya,muncullahbanyakceritayangberedar dimasyarakat.
Padasalahsatubagian,terdapatkisah romansaantara tentarayang disebutKurawadengan Arimbi.Padakisahtersebut, tentaraKurawa menyerang 2orangdaridaerah Bharata.Keduaorangtersebutadalah Rama dan Shinta.Mereka menyerang Rama dan Shinta dikarenakan Rama dan Shintaterlalu romantis.
Akhirnya tentaraKurawa menantangakan dikalahkanjikaingin menikahi Arimbi.Akhirnya Rama dan Shintamaukalahkan tentaraKurawa.Merekapunkalahkan tentaraKurawadengantanpamenggunakansenjata.
Darikisah romansatersebut, di bangunlahsebuah candiyang berbentukbunga.Akhirnya,terdapat 2 candi di manasalahsatunya berbentuk candiKurawayang di isukanoleh tentara dan candi Arimbiyang di isukanoleh Rama dan Shinta.

louisowen6

May 30, 2024

Hi @raihan2345 , thanks for trying Komodo! Wondering, did you pass --trust-remote-code as True when initializing in vLLM?

raihan2345

May 30, 2024

•

edited May 30, 2024

I did, this is my command:

python -u -m vllm.entrypoints.openai.api_server \
       --host 0.0.0.0 \
       --port 8030 \
       --model Yellow-AI-NLP/komodo-7b-base \
       --tensor-parallel-size 1  \
       --gpu-memory-utilization 0.9 \
       --trust-remote-code

*I used vllm version 0.3.2

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment