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

Libraries

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

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

pipe = pipeline("image-text-to-text", model="cagataydev/loopl-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-2b")
model = AutoModelForMultimodalLM.from_pretrained("cagataydev/loopl-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-2b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "cagataydev/loopl-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-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-2b

SGLang

How to use cagataydev/loopl-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-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-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-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-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-2b with Docker Model Runner:
```
docker model run hf.co/cagataydev/loopl-2b
```

loopl-2b / README.md

cagataydev

card: loopl 2B v1.2 (bf16) — train/cards.py

b9e4648 verified 27 days ago

preview code

Raw History Blame Contribute Delete

12.5 kB

---
base_model: Qwen/Qwen3.5-2B
base_model_relation: finetune
library_name: transformers
pipeline_tag: image-text-to-text
license: apache-2.0
language:
- en
- tr
datasets:
- cagataydev/loopl-train
tags:
- loopl
- on-device
- agent
- tool-calling
- qwen3.5
- qwen3_5
- sft
- vision
- image-text-to-text
---
<p align="center">
  <a href="https://loopl.dev"><img src="https://huggingface.co/cagataydev/loopl-2b/resolve/main/assets/banner.png" alt="loopl — Agents that run on your phone." width="100%"></a>
</p>

<div align="center" style="display:flex;flex-wrap:wrap;gap:6px;justify-content:center"><a href="https://huggingface.co/cagataydev/loopl-2b/blob/main/README.md#license"><img alt="license: Apache 2.0" src="https://img.shields.io/badge/license-Apache%202.0-5d5d63?style=flat-square&labelColor=0b0b0c"></a> <img alt="bf16: transformers · 4.43 GB" src="https://img.shields.io/badge/bf16-transformers%20%C2%B7%204.43%20GB-5d5d63?style=flat-square&labelColor=0b0b0c"> <img alt="vision: ✓ 297 tensors kept" src="https://img.shields.io/badge/vision-%E2%9C%93%20297%20tensors%20kept-5d5d63?style=flat-square&labelColor=0b0b0c"> <img alt="tool calling: native + loopl JSON" src="https://img.shields.io/badge/tool%20calling-native%20%2B%20loopl%20JSON-5d5d63?style=flat-square&labelColor=0b0b0c"> <img alt="params: 2B" src="https://img.shields.io/badge/params-2B-5d5d63?style=flat-square&labelColor=0b0b0c"> <img alt="version: v1.2" src="https://img.shields.io/badge/version-v1.2-5d5d63?style=flat-square&labelColor=0b0b0c"> <a href="https://testflight.apple.com/join/uUNjfzW9"><img alt="Get it on: TestFlight" src="https://img.shields.io/badge/Get%20it%20on-TestFlight-ff5a1f?style=flat-square&labelColor=0b0b0c&logo=apple&logoColor=white"></a> <a href="https://loopl.dev"><img alt="site: loopl.dev" src="https://img.shields.io/badge/site-loopl.dev-c2410c?style=flat-square&labelColor=0b0b0c"></a> <a href="https://huggingface.co/collections/cagataydev/loopl-on-device-agents-swift-6ac51395b13a9c41d4932eaf"><img alt="collection: 6 models" src="https://img.shields.io/badge/collection-6%20models-c2410c?style=flat-square&labelColor=0b0b0c"></a></div>

# loopl 2B · v1.2

**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 Qwen3.5 post-tuned to run loopl's agent loop well, know who it is, and read images.

<table align="center"><tr><td align="center" width="33%"><img src="https://huggingface.co/cagataydev/loopl-2b/resolve/main/assets/see-tool.png" alt="It uses the phone — every tool call is a row." width="250"></td><td align="center" width="33%"><img src="https://huggingface.co/cagataydev/loopl-2b/resolve/main/assets/see-approval.png" alt="It asks before anything leaves the device." width="250"></td><td align="center" width="33%"><img src="https://huggingface.co/cagataydev/loopl-2b/resolve/main/assets/see-vision.png" alt="Show it a photo. It reads it on-device." width="250"></td></tr><tr><td align="center"><sub>It uses the phone — every tool call is a row.</sub></td><td align="center"><sub>It asks before anything leaves the device.</sub></td><td align="center"><sub>Show it a photo. It reads it on-device.</sub></td></tr></table>
<p align="center"><sub>Real captures from the free iPhone app.</sub></p>

## 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 Qwen3.5 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](https://loopl.dev) · [Get it on TestFlight](https://testflight.apple.com/join/uUNjfzW9) · [github.com/cagataycali/loopl](https://github.com/cagataycali/loopl)

## This model

| probe | this model | previous (v1) | base Qwen3.5 4-bit |
|---|---:|---:|---:|
| agent-SDK knowledge quiz (first 472 questions, exact-answer) | **99** | 88 | 56 |
| tool calls, native template dialect | **9/9** | 9/9 | 7/9 |
| tool calls, loopl JSON dialect | **9/9** | 8/9 | 5/9 |
| identity (5 classic probes, with system prompt) | **5/5** | 4/5 | 4/5 |
| identity · bio facts (8, with system prompt) | **8/8** | — | — |
| identity · **no system prompt** (13) | **12/13** | — | — |

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** — 297 `vision_tower.*` tensors in the MLX export, so the app shows the photo button.

## Use it today

**In the loopl app (iPhone).** Install from [TestFlight](https://testflight.apple.com/join/uUNjfzW9) → Models → the **loopl models** section lists this
size (no account, no token) → download. The app sees the vision tower and shows the photo button; the system prompt is yours to
edit — identity and tool manners are in the weights.

**On a Mac with MLX** (vision included):

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

**With llama.cpp:** the GGUF Q4_K_M export (`cagataydev/loopl-2b-GGUF`) is not public yet; export your own from the bf16 weights with
llama.cpp's `convert_hf_to_gguf.py --no-mtp` (text only — the converter has no vision path).

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

```swift
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-2b](https://huggingface.co/cagataydev/loopl-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-2b`); on a Mac they load with `transformers` ≥ 5 as `Qwen3_5ForConditionalGeneration`. Export your own
quantisation with `mlx_vlm.convert -q` (keeps vision) — `mlx_lm.convert` silently drops the tower.

## Recipe

| | |
|---|---|
| base | [Qwen/Qwen3.5-2B](https://huggingface.co/Qwen/Qwen3.5-2B) (vision tower frozen, `visual.*` never trained) |
| data | [cagataydev/loopl-train](https://huggingface.co/datasets/cagataydev/loopl-train) *(private)* @ `47d27b44` · 13704 rendered rows / 37,226,229 tokens · both tool dialects · oversample `c=2` |
| 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 | ep1 1.654, ep2 1.714 → best 1.654 (eval split of the same dataset revision) |
| compute | a100-large (Hugging Face Jobs), 12636 s train · job `6ac5e6ec404719ba37665446` |
| stack | transformers 5.19.0 · trl 1.14.2 · peft 0.21.2 |
| exports | MLX 4-bit via `mlx_vlm.convert -q` (297 vision tensors of 991, 1.72 GB) · GGUF Q4_K_M via llama.cpp `--no-mtp` (1.27 GB, text only) |

Script: [`train/loopl_sft.py`](https://github.com/cagataycali/loopl-train) — 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

Apache-2.0, inherited from the Qwen3.5 base. The fine-tune, exports and dataset are © Cagatay Cali, same 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 |
|---|---|---|
| **0.8B** | [`cagataydev/loopl-0.8b-4bit`](https://huggingface.co/cagataydev/loopl-0.8b-4bit) · 625 MB | [`cagataydev/loopl-0.8b`](https://huggingface.co/cagataydev/loopl-0.8b) · 1.71 GB |
| **2B** | [`cagataydev/loopl-2b-4bit`](https://huggingface.co/cagataydev/loopl-2b-4bit) · 1.72 GB | [`cagataydev/loopl-2b`](https://huggingface.co/cagataydev/loopl-2b) · 4.43 GB ← this repo |
| **4B** | [`cagataydev/loopl-4b-4bit`](https://huggingface.co/cagataydev/loopl-4b-4bit) · 3.03 GB | [`cagataydev/loopl-4b`](https://huggingface.co/cagataydev/loopl-4b) · 9.08 GB |

All of them: [the loopl collection](https://huggingface.co/collections/cagataydev/loopl-on-device-agents-swift-6ac51395b13a9c41d4932eaf).

## Links

- Site and docs — [loopl.dev](https://loopl.dev) · SDK: [agent loop](https://loopl.dev/sdk/agent/), [tools](https://loopl.dev/sdk/tools/), [human in the loop](https://loopl.dev/sdk/hitl/), [models](https://loopl.dev/models/)
- App — [Get it on TestFlight](https://testflight.apple.com/join/uUNjfzW9) (iOS 18+, free, public beta)
- Code — [github.com/cagataycali/loopl](https://github.com/cagataycali/loopl) · [issues](https://github.com/cagataycali/loopl/issues) · training: [github.com/cagataycali/loopl-train](https://github.com/cagataycali/loopl-train)
- Models — [the loopl collection](https://huggingface.co/collections/cagataydev/loopl-on-device-agents-swift-6ac51395b13a9c41d4932eaf) · data: [cagataydev/loopl-train](https://huggingface.co/datasets/cagataydev/loopl-train) *(private)*
- Built by Cagatay Cali — [@cagataycali](https://github.com/cagataycali)

	---
	base_model: Qwen/Qwen3.5-2B
	base_model_relation: finetune
	library_name: transformers
	pipeline_tag: image-text-to-text
	license: apache-2.0
	language:
	- en
	- tr
	datasets:
	- cagataydev/loopl-train
	tags:
	- loopl
	- on-device
	- agent
	- tool-calling
	- qwen3.5
	- qwen3_5
	- sft
	- vision
	- image-text-to-text
	---
	<p align="center">
	<a href="https://loopl.dev"><img src="https://huggingface.co/cagataydev/loopl-2b/resolve/main/assets/banner.png" alt="loopl — Agents that run on your phone." width="100%"></a>
	</p>

	<div align="center" style="display:flex;flex-wrap:wrap;gap:6px;justify-content:center"><a href="https://huggingface.co/cagataydev/loopl-2b/blob/main/README.md#license"><img alt="license: Apache 2.0" src="https://img.shields.io/badge/license-Apache%202.0-5d5d63?style=flat-square&labelColor=0b0b0c"></a> <img alt="bf16: transformers · 4.43 GB" src="https://img.shields.io/badge/bf16-transformers%20%C2%B7%204.43%20GB-5d5d63?style=flat-square&labelColor=0b0b0c"> <img alt="vision: ✓ 297 tensors kept" src="https://img.shields.io/badge/vision-%E2%9C%93%20297%20tensors%20kept-5d5d63?style=flat-square&labelColor=0b0b0c"> <img alt="tool calling: native + loopl JSON" src="https://img.shields.io/badge/tool%20calling-native%20%2B%20loopl%20JSON-5d5d63?style=flat-square&labelColor=0b0b0c"> <img alt="params: 2B" src="https://img.shields.io/badge/params-2B-5d5d63?style=flat-square&labelColor=0b0b0c"> <img alt="version: v1.2" src="https://img.shields.io/badge/version-v1.2-5d5d63?style=flat-square&labelColor=0b0b0c"> <a href="https://testflight.apple.com/join/uUNjfzW9"><img alt="Get it on: TestFlight" src="https://img.shields.io/badge/Get%20it%20on-TestFlight-ff5a1f?style=flat-square&labelColor=0b0b0c&logo=apple&logoColor=white"></a> <a href="https://loopl.dev"><img alt="site: loopl.dev" src="https://img.shields.io/badge/site-loopl.dev-c2410c?style=flat-square&labelColor=0b0b0c"></a> <a href="https://huggingface.co/collections/cagataydev/loopl-on-device-agents-swift-6ac51395b13a9c41d4932eaf"><img alt="collection: 6 models" src="https://img.shields.io/badge/collection-6%20models-c2410c?style=flat-square&labelColor=0b0b0c"></a></div>

	# loopl 2B · v1.2

	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 Qwen3.5 post-tuned to run loopl's agent loop well, know who it is, and read images.

	<table align="center"><tr><td align="center" width="33%"><img src="https://huggingface.co/cagataydev/loopl-2b/resolve/main/assets/see-tool.png" alt="It uses the phone — every tool call is a row." width="250"></td><td align="center" width="33%"><img src="https://huggingface.co/cagataydev/loopl-2b/resolve/main/assets/see-approval.png" alt="It asks before anything leaves the device." width="250"></td><td align="center" width="33%"><img src="https://huggingface.co/cagataydev/loopl-2b/resolve/main/assets/see-vision.png" alt="Show it a photo. It reads it on-device." width="250"></td></tr><tr><td align="center"><sub>It uses the phone — every tool call is a row.</sub></td><td align="center"><sub>It asks before anything leaves the device.</sub></td><td align="center"><sub>Show it a photo. It reads it on-device.</sub></td></tr></table>
	<p align="center"><sub>Real captures from the free iPhone app.</sub></p>

	## 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 Qwen3.5 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](https://loopl.dev) · [Get it on TestFlight](https://testflight.apple.com/join/uUNjfzW9) · [github.com/cagataycali/loopl](https://github.com/cagataycali/loopl)

	## This model

	\| probe \| this model \| previous (v1) \| base Qwen3.5 4-bit \|
	\|---\|---:\|---:\|---:\|
	\| agent-SDK knowledge quiz (first 472 questions, exact-answer) \| 99 \| 88 \| 56 \|
	\| tool calls, native template dialect \| 9/9 \| 9/9 \| 7/9 \|
	\| tool calls, loopl JSON dialect \| 9/9 \| 8/9 \| 5/9 \|
	\| identity (5 classic probes, with system prompt) \| 5/5 \| 4/5 \| 4/5 \|
	\| identity · bio facts (8, with system prompt) \| 8/8 \| — \| — \|
	\| identity · no system prompt (13) \| 12/13 \| — \| — \|

	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 — 297 `vision_tower.*` tensors in the MLX export, so the app shows the photo button.

	## Use it today

	In the loopl app (iPhone). Install from [TestFlight](https://testflight.apple.com/join/uUNjfzW9) → Models → the loopl models section lists this
	size (no account, no token) → download. The app sees the vision tower and shows the photo button; the system prompt is yours to
	edit — identity and tool manners are in the weights.

	On a Mac with MLX (vision included):

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

	With llama.cpp: the GGUF Q4_K_M export (`cagataydev/loopl-2b-GGUF`) is not public yet; export your own from the bf16 weights with
	llama.cpp's `convert_hf_to_gguf.py --no-mtp` (text only — the converter has no vision path).

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

	```swift
	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-2b](https://huggingface.co/cagataydev/loopl-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-2b`); on a Mac they load with `transformers` ≥ 5 as `Qwen3_5ForConditionalGeneration`. Export your own
	quantisation with `mlx_vlm.convert -q` (keeps vision) — `mlx_lm.convert` silently drops the tower.

	## Recipe

	\| \| \|
	\|---\|---\|
	\| base \| [Qwen/Qwen3.5-2B](https://huggingface.co/Qwen/Qwen3.5-2B) (vision tower frozen, `visual.*` never trained) \|
	\| data \| [cagataydev/loopl-train](https://huggingface.co/datasets/cagataydev/loopl-train) (private) @ `47d27b44` · 13704 rendered rows / 37,226,229 tokens · both tool dialects · oversample `c=2` \|
	\| 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 \| ep1 1.654, ep2 1.714 → best 1.654 (eval split of the same dataset revision) \|
	\| compute \| a100-large (Hugging Face Jobs), 12636 s train · job `6ac5e6ec404719ba37665446` \|
	\| stack \| transformers 5.19.0 · trl 1.14.2 · peft 0.21.2 \|
	\| exports \| MLX 4-bit via `mlx_vlm.convert -q` (297 vision tensors of 991, 1.72 GB) · GGUF Q4_K_M via llama.cpp `--no-mtp` (1.27 GB, text only) \|

	Script: [`train/loopl_sft.py`](https://github.com/cagataycali/loopl-train) — 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

	Apache-2.0, inherited from the Qwen3.5 base. The fine-tune, exports and dataset are © Cagatay Cali, same 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 \|
	\|---\|---\|---\|
	\| 0.8B \| [`cagataydev/loopl-0.8b-4bit`](https://huggingface.co/cagataydev/loopl-0.8b-4bit) · 625 MB \| [`cagataydev/loopl-0.8b`](https://huggingface.co/cagataydev/loopl-0.8b) · 1.71 GB \|
	\| 2B \| [`cagataydev/loopl-2b-4bit`](https://huggingface.co/cagataydev/loopl-2b-4bit) · 1.72 GB \| [`cagataydev/loopl-2b`](https://huggingface.co/cagataydev/loopl-2b) · 4.43 GB ← this repo \|
	\| 4B \| [`cagataydev/loopl-4b-4bit`](https://huggingface.co/cagataydev/loopl-4b-4bit) · 3.03 GB \| [`cagataydev/loopl-4b`](https://huggingface.co/cagataydev/loopl-4b) · 9.08 GB \|

	All of them: [the loopl collection](https://huggingface.co/collections/cagataydev/loopl-on-device-agents-swift-6ac51395b13a9c41d4932eaf).

	## Links

	- Site and docs — [loopl.dev](https://loopl.dev) · SDK: [agent loop](https://loopl.dev/sdk/agent/), [tools](https://loopl.dev/sdk/tools/), [human in the loop](https://loopl.dev/sdk/hitl/), [models](https://loopl.dev/models/)
	- App — [Get it on TestFlight](https://testflight.apple.com/join/uUNjfzW9) (iOS 18+, free, public beta)
	- Code — [github.com/cagataycali/loopl](https://github.com/cagataycali/loopl) · [issues](https://github.com/cagataycali/loopl/issues) · training: [github.com/cagataycali/loopl-train](https://github.com/cagataycali/loopl-train)
	- Models — [the loopl collection](https://huggingface.co/collections/cagataydev/loopl-on-device-agents-swift-6ac51395b13a9c41d4932eaf) · data: [cagataydev/loopl-train](https://huggingface.co/datasets/cagataydev/loopl-train) (private)
	- Built by Cagatay Cali — [@cagataycali](https://github.com/cagataycali)