Instructions to use cnfusion/Mellum-4b-sft-python-mlx-8Bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use cnfusion/Mellum-4b-sft-python-mlx-8Bit with Transformers:

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

pipe = pipeline("text-generation", model="cnfusion/Mellum-4b-sft-python-mlx-8Bit")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("cnfusion/Mellum-4b-sft-python-mlx-8Bit")
model = AutoModelForCausalLM.from_pretrained("cnfusion/Mellum-4b-sft-python-mlx-8Bit")

MLX

How to use cnfusion/Mellum-4b-sft-python-mlx-8Bit with MLX:

# Make sure mlx-lm is installed
# pip install --upgrade mlx-lm
# if on a CUDA device, also pip install mlx[cuda]

# Generate text with mlx-lm
from mlx_lm import load, generate

model, tokenizer = load("cnfusion/Mellum-4b-sft-python-mlx-8Bit")

prompt = "Once upon a time in"
text = generate(model, tokenizer, prompt=prompt, verbose=True)

Notebooks
Google Colab
Kaggle
Local Apps
LM Studio

vLLM

How to use cnfusion/Mellum-4b-sft-python-mlx-8Bit with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "cnfusion/Mellum-4b-sft-python-mlx-8Bit"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "cnfusion/Mellum-4b-sft-python-mlx-8Bit",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/cnfusion/Mellum-4b-sft-python-mlx-8Bit

SGLang

How to use cnfusion/Mellum-4b-sft-python-mlx-8Bit 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 "cnfusion/Mellum-4b-sft-python-mlx-8Bit" \
    --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": "cnfusion/Mellum-4b-sft-python-mlx-8Bit",
		"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 "cnfusion/Mellum-4b-sft-python-mlx-8Bit" \
        --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": "cnfusion/Mellum-4b-sft-python-mlx-8Bit",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

MLX LM

How to use cnfusion/Mellum-4b-sft-python-mlx-8Bit with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Generate some text
mlx_lm.generate --model "cnfusion/Mellum-4b-sft-python-mlx-8Bit" --prompt "Once upon a time"

Docker Model Runner
How to use cnfusion/Mellum-4b-sft-python-mlx-8Bit with Docker Model Runner:
```
docker model run hf.co/cnfusion/Mellum-4b-sft-python-mlx-8Bit
```

cnfusion/Mellum-4b-sft-python-mlx-8Bit

The Model cnfusion/Mellum-4b-sft-python-mlx-8Bit was converted to MLX format from JetBrains/Mellum-4b-sft-python using mlx-lm version 0.22.3.

Use with mlx

pip install mlx-lm

from mlx_lm import load, generate

model, tokenizer = load("cnfusion/Mellum-4b-sft-python-mlx-8Bit")

prompt="hello"

if hasattr(tokenizer, "apply_chat_template") and tokenizer.chat_template is not None:
    messages = [{"role": "user", "content": prompt}]
    prompt = tokenizer.apply_chat_template(
        messages, tokenize=False, add_generation_prompt=True
    )

response = generate(model, tokenizer, prompt=prompt, verbose=True)

Downloads last month: 14

Safetensors

Model size

1B params

Tensor type

F16

U32

MLX

Hardware compatibility

8-bit

Model tree for cnfusion/Mellum-4b-sft-python-mlx-8Bit

Base model

JetBrains/Mellum-4b-base

Finetuned

JetBrains/Mellum-4b-sft-python

Quantized

(7)

this model

Datasets used to train cnfusion/Mellum-4b-sft-python-mlx-8Bit

Evaluation results

EM on RepoBench 1.1 (Python)
self-reported

0.284
EM ≤ 8k on RepoBench 1.1 (Python)
self-reported

0.299
EM on RepoBench 1.1 (Python)
self-reported

0.292
EM on RepoBench 1.1 (Python)
self-reported

0.306
EM on RepoBench 1.1 (Python)
self-reported

0.298
EM on RepoBench 1.1 (Python)
self-reported

0.268
EM on RepoBench 1.1 (Python)
self-reported

0.254
pass@1 on SAFIM
self-reported

0.421
pass@1 on SAFIM
self-reported

0.332
pass@1 on SAFIM
self-reported

0.361
pass@1 on SAFIM
self-reported

0.571
pass@1 on HumanEval Infilling (Single-Line)
self-reported

0.804
pass@1 on HumanEval Infilling (Single-Line)
self-reported

0.482
pass@1 on HumanEval Infilling (Single-Line)
self-reported

0.377