Instructions to use vilm/Mixsmol-4x400M-v0.1-epoch2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use vilm/Mixsmol-4x400M-v0.1-epoch2 with Transformers:

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

pipe = pipeline("text-generation", model="vilm/Mixsmol-4x400M-v0.1-epoch2")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("vilm/Mixsmol-4x400M-v0.1-epoch2")
model = AutoModelForCausalLM.from_pretrained("vilm/Mixsmol-4x400M-v0.1-epoch2", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use vilm/Mixsmol-4x400M-v0.1-epoch2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "vilm/Mixsmol-4x400M-v0.1-epoch2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "vilm/Mixsmol-4x400M-v0.1-epoch2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/vilm/Mixsmol-4x400M-v0.1-epoch2

SGLang

How to use vilm/Mixsmol-4x400M-v0.1-epoch2 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 "vilm/Mixsmol-4x400M-v0.1-epoch2" \
    --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": "vilm/Mixsmol-4x400M-v0.1-epoch2",
		"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 "vilm/Mixsmol-4x400M-v0.1-epoch2" \
        --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": "vilm/Mixsmol-4x400M-v0.1-epoch2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use vilm/Mixsmol-4x400M-v0.1-epoch2 with Docker Model Runner:
```
docker model run hf.co/vilm/Mixsmol-4x400M-v0.1-epoch2
```

Mixsmol-4x400M-v0.1 by Ontocord

This is the first checkpoint (Epoch 1) of Mixsmol-4x400M-v0.1 Note that this is an experimental in data mixing. Therefore, we only trained the model on 50B tokens (95% English and 5% Vietnamese) to test the following:

Reasoining capabilities through high-quality synthetic textbooks data pretraining
Crosslingual understanding through machine translation and multilingual + multiple tasks pretraining

After verifying our hypothesis with this run, we will schedule a second run on bigger data and compute for it to achieve its maximum capability.

Data

Synthetic Textbooks: 8M samples
RefinedWeb: 1M samples
RedPajama-v2: 500K samples
MathPile: Everything
ThePile: MiniPile Subset
GoodWiki
The Stack Smol XL
The Vault: train_small split
Instruction Pretraining: 250k samples

Downloads last month: 49

Safetensors

Model size

2B params

Tensor type

BF16

Collection including vilm/Mixsmol-4x400M-v0.1-epoch2

Mixsmol

Collection

Smol MoE series - in collaboration with Ontocord • 3 items • Updated Feb 10, 2024