Instructions to use mmnga/TinyMixtral-x8-Clonebase-7b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use mmnga/TinyMixtral-x8-Clonebase-7b with Transformers:

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

pipe = pipeline("text-generation", model="mmnga/TinyMixtral-x8-Clonebase-7b")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("mmnga/TinyMixtral-x8-Clonebase-7b")
model = AutoModelForCausalLM.from_pretrained("mmnga/TinyMixtral-x8-Clonebase-7b")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use mmnga/TinyMixtral-x8-Clonebase-7b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "mmnga/TinyMixtral-x8-Clonebase-7b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "mmnga/TinyMixtral-x8-Clonebase-7b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/mmnga/TinyMixtral-x8-Clonebase-7b

SGLang

How to use mmnga/TinyMixtral-x8-Clonebase-7b 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 "mmnga/TinyMixtral-x8-Clonebase-7b" \
    --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": "mmnga/TinyMixtral-x8-Clonebase-7b",
		"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 "mmnga/TinyMixtral-x8-Clonebase-7b" \
        --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": "mmnga/TinyMixtral-x8-Clonebase-7b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use mmnga/TinyMixtral-x8-Clonebase-7b with Docker Model Runner:
```
docker model run hf.co/mmnga/TinyMixtral-x8-Clonebase-7b
```

Model Card for TinyMixtral-x8-Clonebase-7b

This model is based on TinyLlama-1.1B, converted to a mistral model, and then placed the clone in mixtral.
This model was created experimentally for training a small mixtral.
Without Train, the performance of this model is the same as TinyLlama.

How it was made

First, since tinyllama is an llama model, I converted it to a mistral model.

After that, I cloned the FFN part and made it experts. Since they are all the same tensor, the performance does not change. All gates have the same value.

How To Convert

use colab cpu-high-memory.
This model was created with experts=8, but since it is a clone, you can create as many experts as you like.

tinyllama_to_mixtral_clonebase.ipynb

revision

main TinyLlama-1.1B-intermediate-step-1431k-3T
old TinyLlama-1.1B-intermediate-step-1195k-token-2.5T

Usage

pip install transformers --upgrade
pip install flash_attn bitsandbytes accelerate

from transformers import AutoTokenizer, AutoModelForCausalLM
import torch

model_name_or_path = "mmnga/TinyMixtral-x8-Clonebase-7b"

tokenizer = AutoTokenizer.from_pretrained(model_name_or_path)
model = AutoModelForCausalLM.from_pretrained(model_name_or_path, device_map="auto", load_in_8bit=True)

prompt = "Introducing the recipe for today's dinner."

with torch.no_grad():
    token_ids = tokenizer.encode(prompt, return_tensors="pt")
    output_ids = model.generate(
        token_ids.to(model.device),
        do_sample=True,
        max_new_tokens=128,
        repetition_penalty=1.5
    )
    output = tokenizer.decode(output_ids[0])
print(output)

Downloads last month: 9

Safetensors

Model size

6B params

Tensor type

F32