Instructions to use cagataydev/loopl-vl-2b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use cagataydev/loopl-vl-2b with Transformers:

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

pipe = pipeline("image-text-to-text", model="cagataydev/loopl-vl-2b")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("cagataydev/loopl-vl-2b")
model = AutoModelForMultimodalLM.from_pretrained("cagataydev/loopl-vl-2b", device_map="auto")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.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=256)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use cagataydev/loopl-vl-2b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "cagataydev/loopl-vl-2b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "cagataydev/loopl-vl-2b",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/cagataydev/loopl-vl-2b

SGLang

How to use cagataydev/loopl-vl-2b 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 "cagataydev/loopl-vl-2b" \
    --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": "cagataydev/loopl-vl-2b",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

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 "cagataydev/loopl-vl-2b" \
        --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": "cagataydev/loopl-vl-2b",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use cagataydev/loopl-vl-2b with Docker Model Runner:
```
docker model run hf.co/cagataydev/loopl-vl-2b
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

loopl 2B · v1

An agent that lives on your phone. loopl runs the model, the loop and the tools on the device. It talks, uses the phone, asks before it acts — and works with the network off. This repo is the merged bf16 weights — the base for Train your own — of the 2B loopl model, a Qwen3-VL post-tuned to run loopl's agent loop well, know who it is, and read images and video.


_{It uses the phone — every tool call is a row.}	_{It asks before anything leaves the device.}	_{Show it a photo. It reads it on-device.}

_{Real captures from the free iPhone app.}

What loopl is

A Swift SDK — Agent(model:tools:) and a small, honest loop: model → tools → model until it answers. Tool errors go back to the model (it fixes the call instead of repeating it); anything that leaves the device waits for a tap. Runtimes for MLX, llama.cpp and Apple's on-device model; hooks, checkpoints, structured output, tracing.
A free iPhone app built on it — pick a model, talk to it in airplane mode, let it use the phone (torch, haptics, translation, speech, image generation, other apps, memory, HTTP), show it photos, record a take, and train it on your own conversations from the phone. No account, nothing collected.
These models — the same Qwen3-VL the app offers, post-tuned so the loop, the identity and the tool manners are in the weights, not in a system prompt you cannot see.

loopl.dev · Get it on TestFlight · github.com/cagataycali/loopl

This model

probe	this model	previous (v1)	base Qwen3-VL 4-bit
agent-SDK knowledge quiz (first 472 questions, exact-answer)	68	—	60
tool calls, native template dialect	8/9	—	7/9
tool calls, loopl JSON dialect	7/9	—	7/9
identity (5 classic probes, with system prompt)	4/5	—	2/5
identity · bio facts (8, with system prompt)	6/8	—	5/8
identity · no system prompt (13)	2/13	—	2/13

Scored by train/eval.py on the 4-bit MLX export (greedy). The quiz asks for exact file-level facts about an agent SDK's source — hard for every small model; the delta over base is the point. Tool probes count a well-formed call with the right arguments. identity_nosys is the one that matters in the app: the system prompt there is the user's to edit.

Vision: the base's tower is frozen and kept — 315 vision_tower.* tensors in the MLX export, so the app shows the photo and video buttons.

loopl-vl reads pictures and video natively (frames go straight to the model; the audio track is transcribed on-device by the app and handed over as text — Qwen3-VL has no audio input). Base: an Instruct base — no thinking blocks.

Use it today

In the loopl app (iPhone). Install from TestFlight → Models → the loopl models section lists this size (no account, no token) → download. The app sees the vision tower and shows the photo and video buttons; the system prompt is yours to edit — identity and tool manners are in the weights.

On a Mac with MLX (vision included):

pip install -U mlx-vlm
python -m mlx_vlm.generate --model cagataydev/loopl-vl-2b-4bit \
  --image photo.jpg --prompt "What is in this picture? Then tell me who built you." --max-tokens 200

Video, on a Mac with MLX — frames go straight to the model (2 fps, 8 frames here; the loopl app does the same and transcribes the audio track on-device, Qwen3-VL has no audio input). No GGUF: llama.cpp cannot see.

python -m mlx_vlm.generate --model cagataydev/loopl-vl-2b-4bit \
  --video clip.mp4 --fps 2 --video-max-frames 8 --prompt "What happens in this video?" --max-tokens 200

From Swift with the loopl SDK — the snippet below is docs/start.md § "Run the loop" in the loopl repo, compile-checked in CI against the package (products Loopl + LooplRuntimes); spec is this repo's ModelSpec and dir the downloaded folder:

import Loopl
import LooplRuntimes
import Foundation

func chat(_ spec: ModelSpec, at dir: URL) async throws {
    let model = MLXModel()                                       // Metal — a real device
    try await model.load(spec, from: dir) { _ in }

    let agent = Agent(model: model,
                      tools: [CurrentTimeTool(), CalculatorTool()],
                      systemPrompt: "You are a concise assistant on the user's phone.")

    for try await event in agent.stream("What time is it in Istanbul, and what is 23 × 47?") {
        switch event {
        case .text(let t):            print(t, terminator: "")
        case .toolStarted(let use):   print("\n→ \(use.name)")
        case .toolResult(let r):      print("← \(r.content)")
        default: break
        }
    }
}

Keep post-tuning it. The bf16 weights (cagataydev/loopl-vl-2b) are the base for continual post-tuning: in the app, Models › Train your own takes a loopl model as the base and trains on your own shared conversations (train/loopl_sft.py --base cagataydev/loopl-vl-2b); on a Mac they load with transformers ≥ 5 as Qwen3VLForConditionalGeneration. Export your own quantisation with mlx_vlm.convert -q (keeps vision) — mlx_lm.convert silently drops the tower.

Recipe


base	Qwen/Qwen3-VL-2B-Instruct (vision tower frozen, `visual.*` never trained)
data	cagataydev/loopl-train (private) @ `?` · 9208 rendered rows / 15,010,935 tokens · both tool dialects · oversample `—`
method	full fine-tune of the language model · lr 1.5e-05 · batch 2×8 · max_len 4096 · 2.0 epochs, best-epoch checkpoint kept
eval loss	— → best — (eval split of the same dataset revision)
compute	a100-large (Hugging Face Jobs), 6284 s train · job `6ac5abc1fbc85ba6823bb686`
stack	transformers 5.19.0 · trl 1.14.2 · peft 0.21.2
exports	MLX 4-bit via `mlx_vlm.convert -q` (315 vision tensors of 1019, 1.78 GB) · GGUF Q4_K_M via llama.cpp `--no-mtp` (0 MB, text only)

Script: train/loopl_sft.py — the same single file the loopl app launches when you tap Models › Train your own on your own conversations. Assistant turns that seed a bad tool call carry weight: 0 so the model learns the recovery, not the mistake. Tokenizer files are normalised to the base's (Qwen2Tokenizer, chat_template.jinja with tools) so Swift loaders accept them; train/check_mlx_repo.py gates every export on that plus the vision triad (vision_config ⇔ vision_tower.* ⇔ preprocessor_config.json).

Limitations

A 2B model: fluent and well-behaved in the loop, not an encyclopedia. It will get arithmetic and obscure facts wrong; give it tools (calculator, http, recall) and it does better.
Identity holds in most no-prompt probes (see the scores), not all; a one-line system prompt ("You are loopl…") makes it consistent.
The GGUF export is text-only. Vision needs the MLX export (or the bf16 weights).
Trained on English plus a little Turkish; other languages are the base's.
Knowledge about agent SDKs is file-level and dated to the dataset revision; it does not know your repo.

Data and privacy

Every training row is agent-synthesised from public sources (the owner's public GitHub repos and the agent-SDK source they build on), critic-reviewed (score ≥ 4/5 kept) and filtered by a denylist for secrets, private names, phone numbers and addresses — see the dataset card. No user conversations from the app are in this model.

License

The family

Six public repos, one recipe. The app lists the three MLX rows under Models › loopl models (2B is the sweet spot for speed on an iPhone 15/16; 4B is the most knowledgeable; 0.8B fits anywhere). Every bf16 repo is a valid --base for the next round.

	MLX 4-bit · vision · what the app downloads	bf16 · the base to keep training
2B	`cagataydev/loopl-vl-2b-4bit` · 1.78 GB	`cagataydev/loopl-vl-2b` · 4.26 GB ← this repo
4B	`cagataydev/loopl-vl-4b-4bit` · 3.09 GB	`cagataydev/loopl-vl-4b` · 8.88 GB

All of them: the loopl collection.

Model tree for cagataydev/loopl-vl-2b

Base model

Qwen/Qwen3-VL-2B-Instruct

Finetuned

(271)

this model

Quantizations

3 models

cagataydev
/

loopl-vl-2b