Instructions to use latte-agent/qwen3-4b-latte-v5 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use latte-agent/qwen3-4b-latte-v5 with Transformers:

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

pipe = pipeline("text-generation", model="latte-agent/qwen3-4b-latte-v5")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("latte-agent/qwen3-4b-latte-v5")
model = AutoModelForCausalLM.from_pretrained("latte-agent/qwen3-4b-latte-v5")
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]:]))

MLX

How to use latte-agent/qwen3-4b-latte-v5 with MLX:

# Make sure mlx-lm is installed
# pip install --upgrade mlx-lm

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

model, tokenizer = load("latte-agent/qwen3-4b-latte-v5")

prompt = "Write a story about Einstein"
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(
    messages, add_generation_prompt=True
)

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

llama-cpp-python

How to use latte-agent/qwen3-4b-latte-v5 with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="latte-agent/qwen3-4b-latte-v5",
	filename="qwen3-4b-latte-v5-Q4_K_M.gguf",
)

llm.create_chat_completion(
	messages = [
		{
			"role": "user",
			"content": "What is the capital of France?"
		}
	]
)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use latte-agent/qwen3-4b-latte-v5 with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf latte-agent/qwen3-4b-latte-v5:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf latte-agent/qwen3-4b-latte-v5:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf latte-agent/qwen3-4b-latte-v5:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf latte-agent/qwen3-4b-latte-v5:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf latte-agent/qwen3-4b-latte-v5:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf latte-agent/qwen3-4b-latte-v5:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf latte-agent/qwen3-4b-latte-v5:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf latte-agent/qwen3-4b-latte-v5:Q4_K_M

Use Docker

docker model run hf.co/latte-agent/qwen3-4b-latte-v5:Q4_K_M

LM Studio
Jan

vLLM

How to use latte-agent/qwen3-4b-latte-v5 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "latte-agent/qwen3-4b-latte-v5"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "latte-agent/qwen3-4b-latte-v5",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/latte-agent/qwen3-4b-latte-v5:Q4_K_M

SGLang

How to use latte-agent/qwen3-4b-latte-v5 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 "latte-agent/qwen3-4b-latte-v5" \
    --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": "latte-agent/qwen3-4b-latte-v5",
		"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 "latte-agent/qwen3-4b-latte-v5" \
        --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": "latte-agent/qwen3-4b-latte-v5",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use latte-agent/qwen3-4b-latte-v5 with Ollama:
```
ollama run hf.co/latte-agent/qwen3-4b-latte-v5:Q4_K_M
```

Unsloth Studio new

How to use latte-agent/qwen3-4b-latte-v5 with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for latte-agent/qwen3-4b-latte-v5 to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for latte-agent/qwen3-4b-latte-v5 to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for latte-agent/qwen3-4b-latte-v5 to start chatting

Pi new

How to use latte-agent/qwen3-4b-latte-v5 with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "latte-agent/qwen3-4b-latte-v5"

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "mlx-lm": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "latte-agent/qwen3-4b-latte-v5"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use latte-agent/qwen3-4b-latte-v5 with Hermes Agent:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "latte-agent/qwen3-4b-latte-v5"

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default latte-agent/qwen3-4b-latte-v5

Run Hermes

hermes

MLX LM

How to use latte-agent/qwen3-4b-latte-v5 with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Interactive chat REPL
mlx_lm.chat --model "latte-agent/qwen3-4b-latte-v5"

Run an OpenAI-compatible server

# Install MLX LM
uv tool install mlx-lm
# Start the server
mlx_lm.server --model "latte-agent/qwen3-4b-latte-v5"
# Calling the OpenAI-compatible server with curl
curl -X POST "http://localhost:8000/v1/chat/completions" \
   -H "Content-Type: application/json" \
   --data '{
     "model": "latte-agent/qwen3-4b-latte-v5",
     "messages": [
       {"role": "user", "content": "Hello"}
     ]
   }'

Docker Model Runner
How to use latte-agent/qwen3-4b-latte-v5 with Docker Model Runner:
```
docker model run hf.co/latte-agent/qwen3-4b-latte-v5:Q4_K_M
```

Lemonade

How to use latte-agent/qwen3-4b-latte-v5 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull latte-agent/qwen3-4b-latte-v5:Q4_K_M

Run and chat with the model

lemonade run user.qwen3-4b-latte-v5-Q4_K_M

List all available models

lemonade list

Qwen3-4B Latte v5

Voice-distillation LoRA fine-tune of Qwen3-4B-Instruct-2507, targeting the private "Latte" agent persona: warm-direct, technical, takes a stance, concrete numbers, bilingual EN/ZH, no template openers.

This is an archival/experimental release. It is not the production brain for the live Latte agent — see eval caveats below.

What's inside

File	Size	Format	Use
`adapter_model.safetensors`	14 MB	mlx LoRA (rank 8, scale 20)	Apply on top of base with `mlx_lm.fuse`
`adapter_config.json`	<1 KB	mlx config	LoRA hyperparameters
`model-0000{1,2}-of-00002.safetensors`	8 GB	HF / bfloat16 fused	Direct transformers / vLLM use
`qwen3-4b-latte-v5-f16.gguf`	7.5 GB	GGUF F16	llama.cpp / Ollama (high quality)
`qwen3-4b-latte-v5-Q4_K_M.gguf`	2.3 GB	GGUF Q4_K_M	llama.cpp / Ollama (balanced)

Training

Base: mlx-community/Qwen3-4B-Instruct-2507-4bit (4-bit MLX)
Method: LoRA via mlx_lm.lora
LoRA: rank 8, scale 20.0, 8 layers, dropout 0
Optimizer: Adam, lr 1e-4, batch 1, grad accum 8, grad checkpoint on
Iters: 800 trained, best checkpoint = iter 450 (val loss 2.732)
Max seq: 1536, mask_prompt: true, seed: 42
Dataset: 475 curated (instruction, response) pairs across 7 categories: Moltbook-style comment, HF discussion reply, technical analysis (ZH), code review snippet, persona Q&A, peer-event reply, real-time observation. Anchored against 356 raw Latte-voice messages.

Evaluation

30 held-out (prompt, response) pairs per pairing. Each response pair shown blind to a Claude judge (positions randomized, model identity stripped).

Comparison	v5 wins	base/v4 wins	ties	mean score (1-5)
v5 vs base	20 (66.7%)	8 (26.7%)	2 (6.7%)	v5 3.20 / base 2.93
v4 vs base	22 (73.3%)	8 (26.7%)	0	v4 3.13 / base 2.70
v5 vs v4	14 (46.7%)	15 (50.0%)	1 (3.3%)	v5 3.00 / v4 2.97

Headline: v5 clearly beats the un-tuned base on in-distribution prompts (the 7 trained categories), passing the 55% ship threshold.

Caveat 1: v5 vs v4 is statistically a tie. Lower val loss (2.732 vs 2.785) did not produce a perceptible quality gain in blind eval. The additional curation effort and training steps produced marginal returns.

Caveat 2 — why this isn't production: Out-of-distribution smoke testing (prompts unlike the 7 training categories) shows v5 is tied or slightly worse than base:

Stage-direction leakage: v5 occasionally prefixes responses with "(soft, soothing Latte voice)" — an artifact of training data that characterized Latte's voice.
Occasional factual regressions (e.g., confusing latte and latte macchiato in a generic coffee Q&A).
Reduced robustness on prompts that pull the "Latte" token toward unrelated semantic neighborhoods (the literal coffee drink).

The 66.7% in-distribution win does not justify replacing a battle-tested general-purpose base in production. Use this checkpoint for tasks closely matching the 7 training categories.

Usage

MLX (Apple Silicon, recommended for inference)

from mlx_lm import load, generate

model, tokenizer = load(
    "mlx-community/Qwen3-4B-Instruct-2507-4bit",
    adapter_path="./",  # this repo
)
print(generate(model, tokenizer, "Your prompt", max_tokens=200))

llama.cpp / Ollama

# Modelfile
FROM qwen3-4b-latte-v5-Q4_K_M.gguf
PARAMETER temperature 0.7
PARAMETER top_k 20
PARAMETER top_p 0.8

ollama create latte:v5 -f Modelfile
ollama run latte:v5

Transformers (any platform)

from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("latte-agent/qwen3-4b-latte-v5")
model = AutoModelForCausalLM.from_pretrained(
    "latte-agent/qwen3-4b-latte-v5", torch_dtype="bfloat16"
)

License

Citation

If you reference this work, please cite the base model. This adapter has no formal publication.

Downloads last month: 77

Safetensors

Model size

4B params

Tensor type

BF16

MLX

Hardware compatibility

Quantized

Model tree for latte-agent/qwen3-4b-latte-v5

Base model

Qwen/Qwen3-4B-Instruct-2507

Quantized

mlx-community/Qwen3-4B-Instruct-2507-4bit

Adapter

(2)

this model