Instructions to use chargoddard/mixtralnt-4x7b-test with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use chargoddard/mixtralnt-4x7b-test with Transformers:

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

pipe = pipeline("text-generation", model="chargoddard/mixtralnt-4x7b-test")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("chargoddard/mixtralnt-4x7b-test")
model = AutoModelForCausalLM.from_pretrained("chargoddard/mixtralnt-4x7b-test")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use chargoddard/mixtralnt-4x7b-test with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "chargoddard/mixtralnt-4x7b-test"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "chargoddard/mixtralnt-4x7b-test",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/chargoddard/mixtralnt-4x7b-test

SGLang

How to use chargoddard/mixtralnt-4x7b-test 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 "chargoddard/mixtralnt-4x7b-test" \
    --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": "chargoddard/mixtralnt-4x7b-test",
		"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 "chargoddard/mixtralnt-4x7b-test" \
        --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": "chargoddard/mixtralnt-4x7b-test",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use chargoddard/mixtralnt-4x7b-test with Docker Model Runner:
```
docker model run hf.co/chargoddard/mixtralnt-4x7b-test
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Mixtraln't 4x7B

Oh boy, a new model architecture in Transformers! Time to do profane things with it.

What if instead of training a MoE from scratch, we took some pre-trained Mistral models and shoved them in a little clown car?

Uses parts from the following models:

Works and generates coherent text. The big question here is if the hack I used to populate the MoE gates works well enough to take advantage of all of the experts. Let's find out!

Prompt format: maybe alpaca??? or chatml??? life is full of mysteries

Downloads last month: 221

Safetensors

Model size

24B params

Tensor type

BF16

Model tree for chargoddard/mixtralnt-4x7b-test

Quantizations

4 models

Collection including chargoddard/mixtralnt-4x7b-test

Frankenmodels

Collection

They're not supposed to be that size! Neat, right? • 8 items • Updated Dec 12, 2023 • 3