Instructions to use chargoddard/MixtralRPChat-ZLoss with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use chargoddard/MixtralRPChat-ZLoss with Transformers:

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

pipe = pipeline("text-generation", model="chargoddard/MixtralRPChat-ZLoss")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("chargoddard/MixtralRPChat-ZLoss")
model = AutoModelForCausalLM.from_pretrained("chargoddard/MixtralRPChat-ZLoss")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use chargoddard/MixtralRPChat-ZLoss with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "chargoddard/MixtralRPChat-ZLoss"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "chargoddard/MixtralRPChat-ZLoss",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/chargoddard/MixtralRPChat-ZLoss

SGLang

How to use chargoddard/MixtralRPChat-ZLoss 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/MixtralRPChat-ZLoss" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "chargoddard/MixtralRPChat-ZLoss",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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/MixtralRPChat-ZLoss" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "chargoddard/MixtralRPChat-ZLoss",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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

QLoRA tuned from mistralai/Mixtral-8x7B-v0.1.

My main reason for training this model was to investigate using an altered router balancing loss combined with the z-loss introduced in ST-MoE: Designing Stable and Transferable Sparse Expert Models. The result is pretty decent, I think! It does a good job of respecting character information in system prompts and performed adequately on a few simple coding tasks.

To train this I used a custom branch of Transformers that adds z-loss and reimplements the router balancing loss based on the version in MegaBlocks. The config used with my custom hacked-up branch of axolotl is available here.

Uses my favorite non-ChatML token-economic chat prompt format. Messages should be prefixed with " ***System:", " ***Query:", or " ***Response:" for system, user, and model messages respectively. No newlines are necessary but the space before the triple asterisk is mandatory.

Downloads last month: 17

Safetensors

Model size

47B params

Tensor type

BF16

Model tree for chargoddard/MixtralRPChat-ZLoss

Base model

mistralai/Mixtral-8x7B-v0.1

Finetuned

(72)

this model

Quantizations

3 models

Datasets used to train chargoddard/MixtralRPChat-ZLoss

Paper for chargoddard/MixtralRPChat-ZLoss

ST-MoE: Designing Stable and Transferable Sparse Expert Models

Paper • 2202.08906 • Published Feb 17, 2022 • 3