Image-Text-to-Text
MLX
Safetensors
qwen3_5
clairic
cloudflare
systemone
structured-output
information-extraction
multimodal
custom-code
conversational
4-bit precision
Instructions to use mlx-community/clairic-flash-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use mlx-community/clairic-flash-4bit with MLX:
# Make sure mlx-vlm is installed # pip install --upgrade mlx-vlm from mlx_vlm import load, generate from mlx_vlm.prompt_utils import apply_chat_template from mlx_vlm.utils import load_config # Load the model model, processor = load("mlx-community/clairic-flash-4bit") config = load_config("mlx-community/clairic-flash-4bit") # Prepare input image = ["http://images.cocodataset.org/val2017/000000039769.jpg"] prompt = "Describe this image." # Apply chat template formatted_prompt = apply_chat_template( processor, config, prompt, num_images=1 ) # Generate output output = generate(model, processor, formatted_prompt, image) print(output) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Pi
How to use mlx-community/clairic-flash-4bit with Pi:
Start the MLX server
# Install MLX LM: uv tool install mlx-lm # Start a local OpenAI-compatible server: mlx_lm.server --model "mlx-community/clairic-flash-4bit"
Configure the model in Pi
# Install Pi: npm install -g @earendil-works/pi-coding-agent # Add to ~/.pi/agent/models.json: { "providers": { "mlx-lm": { "baseUrl": "http://localhost:8080/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "mlx-community/clairic-flash-4bit" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Hermes Agent
How to use mlx-community/clairic-flash-4bit 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 "mlx-community/clairic-flash-4bit"
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 mlx-community/clairic-flash-4bit
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use mlx-community/clairic-flash-4bit with OpenClaw:
Start the MLX server
# Install MLX LM: uv tool install mlx-lm # Start a local OpenAI-compatible server: mlx_lm.server --model "mlx-community/clairic-flash-4bit"
Configure OpenClaw
# Install OpenClaw: npm install -g openclaw@latest # Register the local server and set it as the default model: openclaw onboard --non-interactive --mode local \ --auth-choice custom-api-key \ --custom-base-url http://127.0.0.1:8080/v1 \ --custom-model-id "mlx-community/clairic-flash-4bit" \ --custom-provider-id mlx-lm \ --custom-compatibility openai \ --custom-text-input \ --accept-risk \ --skip-health
Run OpenClaw
openclaw agent --local --agent main --message "Hello from Hugging Face"
File size: 3,406 Bytes
ae97c5d | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 | from __future__ import annotations
from collections.abc import Sequence
from typing import Any
def batch_image_data(media: Sequence[Any]) -> list[Any]:
if all((item is None for item in media)):
return list(media)
return [item if item is not None else [] for item in media]
def _grid_identity(grid: Any) -> Any:
if hasattr(grid, "tolist"):
grid = grid.tolist()
if isinstance(grid, (list, tuple)):
return tuple((_grid_identity(value) for value in grid))
if isinstance(grid, (str, int, float, bool, type(None))):
return grid
return ("opaque", id(grid))
def _media_identity(media: Any) -> tuple[Any, ...]:
if media is None:
return ("text",)
if not isinstance(media, list) or not media:
return ("opaque", id(media))
identities = []
for item in media:
if (
not isinstance(item, dict)
or item.get("format") != "processor_output"
or (not isinstance(item.get("clair_media_digest"), str))
or (item.get("pixel_values") is None)
):
return ("opaque", id(media))
identities.append(
(
item["clair_media_digest"],
id(item["pixel_values"]),
_grid_identity(item.get("image_grid_thw")),
)
)
return ("prepared", tuple(identities))
def _common_prefix(tasks: list[tuple[str, list[int], Any]]) -> list[int]:
(first, last) = (min((task[1] for task in tasks)), max((task[1] for task in tasks)))
length = next(
(i for (i, (left, right)) in enumerate(zip(first, last)) if left != right),
min(len(first), len(last)),
)
return list(first[:length])
def media_warmup_groups(
tasks: list[tuple[str, list[int], Any]],
*,
default_salt: str,
task_salts: dict[str, str] | None = None,
annotations: dict[str, dict[str, Any]] | None = None,
min_prefix_tokens: int = 128,
) -> list[dict[str, Any]]:
(task_salts, annotations) = (task_salts or {}, annotations or {})
partitions: dict[tuple[str, tuple[Any, ...]], dict[str, Any]] = {}
for task in tasks:
(name, _, media) = task
salt = task_salts.get(name, default_salt)
key = (salt, _media_identity(media))
partitions.setdefault(key, {"media": media, "salt": salt, "tasks": []})[
"tasks"
].append(task)
records = []
for partition in partitions.values():
members = partition["tasks"]
if len(members) < 2:
continue
prefix = _common_prefix(members)
if len(prefix) < min_prefix_tokens:
prefix = []
parents: dict[str, list[tuple[str, list[int], Any]]] = {}
for task in members:
parent = annotations.get(task[0], {}).get("parent")
if parent:
parents.setdefault(parent, []).append(task)
subgroups = [
_common_prefix(group) for group in parents.values() if len(group) > 1
]
subgroups = [group for group in subgroups if group and group != prefix]
if prefix or subgroups:
records.append(
{
"prefix_ids": prefix,
"media": partition["media"],
"salt": partition["salt"],
"group_prefix_ids": subgroups,
}
)
return records
|