Text Generation
GGUF
Japanese
English
llama.cpp
Mixture of Experts
expert-pruning
intel-mac
cpu
local-agent
imatrix
conversational
Instructions to use miutti/intel-mac-local-llm with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use miutti/intel-mac-local-llm with llama.cpp:
Install (macOS, Linux)
curl -LsSf https://llama.app/install.sh | sh # Start a local OpenAI-compatible server with a web UI: llama serve -hf miutti/intel-mac-local-llm:UD-Q2_K_XL # Run inference directly in the terminal: llama cli -hf miutti/intel-mac-local-llm:UD-Q2_K_XL
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf miutti/intel-mac-local-llm:UD-Q2_K_XL # Run inference directly in the terminal: llama cli -hf miutti/intel-mac-local-llm:UD-Q2_K_XL
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 miutti/intel-mac-local-llm:UD-Q2_K_XL # Run inference directly in the terminal: ./llama-cli -hf miutti/intel-mac-local-llm:UD-Q2_K_XL
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 miutti/intel-mac-local-llm:UD-Q2_K_XL # Run inference directly in the terminal: ./build/bin/llama-cli -hf miutti/intel-mac-local-llm:UD-Q2_K_XL
Use Docker
docker model run hf.co/miutti/intel-mac-local-llm:UD-Q2_K_XL
- LM Studio
- Jan
- vLLM
How to use miutti/intel-mac-local-llm with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "miutti/intel-mac-local-llm" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "miutti/intel-mac-local-llm", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/miutti/intel-mac-local-llm:UD-Q2_K_XL
- Ollama
How to use miutti/intel-mac-local-llm with Ollama:
ollama run hf.co/miutti/intel-mac-local-llm:UD-Q2_K_XL
- Unsloth Desktop
- Pi
How to use miutti/intel-mac-local-llm with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf miutti/intel-mac-local-llm:UD-Q2_K_XL
Configure the model in Pi
# Install Pi: npm install -g @earendil-works/pi-coding-agent # Add to ~/.pi/agent/models.json: { "providers": { "llama-cpp": { "baseUrl": "http://localhost:8080/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "miutti/intel-mac-local-llm:UD-Q2_K_XL" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Docker Model Runner
How to use miutti/intel-mac-local-llm with Docker Model Runner:
docker model run hf.co/miutti/intel-mac-local-llm:UD-Q2_K_XL
- Lemonade
How to use miutti/intel-mac-local-llm with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull miutti/intel-mac-local-llm:UD-Q2_K_XL
Run and chat with the model
lemonade run user.intel-mac-local-llm-UD-Q2_K_XL
List all available models
lemonade list
- Hermes Agent
How to use miutti/intel-mac-local-llm with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf miutti/intel-mac-local-llm:UD-Q2_K_XL
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 miutti/intel-mac-local-llm:UD-Q2_K_XL
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use miutti/intel-mac-local-llm with OpenClaw:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf miutti/intel-mac-local-llm:UD-Q2_K_XL
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 "miutti/intel-mac-local-llm:UD-Q2_K_XL" \ --custom-provider-id llama-cpp \ --custom-compatibility openai \ --custom-text-input \ --accept-risk \ --skip-health
Run OpenClaw
openclaw agent --local --agent main --message "Hello from Hugging Face"
File size: 11,294 Bytes
df41178 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 | #!/usr/bin/env python3
"""server.py の、画面から使う新しい入口だけをモデル無しで検査する。"""
from __future__ import annotations
import http.client
import json
import os
from pathlib import Path
import sys
import tempfile
import threading
import time
import unittest
from types import SimpleNamespace
from urllib.request import Request, urlopen
HERE = Path(__file__).resolve().parents[1]
sys.path.insert(0, str(HERE))
import server # noqa: E402
class ServerApiTests(unittest.TestCase):
def setUp(self):
self.tmp = tempfile.TemporaryDirectory()
root = Path(self.tmp.name)
self.old_dir, self.old_db, self.old_examples, self.old_ready = (
server.chats.DIR, server.chats.DB, server.chats.TITLE_EXAMPLES, server.chats._READY,
)
self.old_seiri_oboe = server.chats.SEIRI_OBOE
self.old_settings = server.CTX.get("設定")
server.CTX["設定"] = {"モード": "試験", "会話の長さ": 8}
server.chats.DIR = str(root / "legacy")
server.chats.DB = str(root / "chats.sqlite3")
server.chats.TITLE_EXAMPLES = str(root / "title_examples.json")
server.chats.SEIRI_OBOE = str(root / "seiri_oboe.json")
server.chats._READY = False
# 10/2: 整理は頭脳を起こして Qwen3.6 に聞くようになった。試験では本物の頭脳に触れない(規則の案だけ)。
self.old_youi, self.old_ask_seiri = server.moderu_youi, server.chats._ask_seiri_qwen
server.moderu_youi = lambda key: (True, "試験")
server.chats._ask_seiri_qwen = lambda entries: None
self.httpd = server.Server(("127.0.0.1", 0), server.Handler)
self.base = "http://127.0.0.1:%d" % self.httpd.server_address[1]
self.thread = threading.Thread(target=self.httpd.serve_forever, daemon=True)
self.thread.start()
def tearDown(self):
self.httpd.shutdown()
self.httpd.server_close()
self.thread.join(timeout=2)
server.chats.DIR, server.chats.DB, server.chats.TITLE_EXAMPLES, server.chats._READY = (
self.old_dir, self.old_db, self.old_examples, self.old_ready,
)
server.chats.SEIRI_OBOE = self.old_seiri_oboe
server.moderu_youi, server.chats._ask_seiri_qwen = self.old_youi, self.old_ask_seiri
server.CTX["設定"] = self.old_settings
self.tmp.cleanup()
def post(self, path, body):
request = Request(
self.base + path,
data=json.dumps(body).encode("utf-8"),
headers={"Content-Type": "application/json", "X-Token": server.TOKEN},
method="POST",
)
with urlopen(request, timeout=3) as response:
return json.loads(response.read().decode("utf-8"))
def test_chat_delete_and_restore_uses_sqlite(self):
created = self.post("/chat/new", {})
cid = created["会話"]
deleted = self.post("/chat/delete", {"id": cid})
self.assertEqual(deleted, {"ok": True, "取り消せる": True})
self.assertIsNone(server.chats.load(cid))
self.assertTrue(self.post("/chat/restore", {"id": cid})["ok"])
self.assertIsNotNone(server.chats.load(cid))
def test_seiri_suggests_three_kinds_learns_rejection_and_restores_trash(self):
now = time.time()
def conversation(title, messages=2, age_days=0):
chat = server.chats.create()
server.chats.rename(chat["id"], title)
for i in range(messages):
server.chats.add_turn(chat["id"], "user" if i % 2 == 0 else "bot", f"本文{i}-{title}")
with server.chats._db() as conn:
conn.execute("UPDATE conversations SET created=?, updated=? WHERE id=?",
(now-age_days*86400, now-age_days*86400, chat["id"]))
return chat["id"]
old_trial = conversation("試験: 古い確認", age_days=40)
old_active = conversation("昔の議事録", messages=4, age_days=40)
left = conversation("京都旅行の予定", messages=4)
right = conversation("京都旅行の持ち物", messages=4)
suggested = self.post("/chat/seiri/an", {})["案"]
self.assertEqual({x["区分"] for x in suggested}, {"ゴミ箱", "しまう", "組"})
# しまう候補を外すと覚え、他のチェック済みだけ実行する。
items = [{**x, "checked": x["id"] != old_active} for x in suggested]
result = self.post("/chat/seiri/suru", {"案": items})
self.assertGreaterEqual(result["実行"], 3)
self.assertNotIn(old_active, {x["id"] for x in self.post("/chat/seiri/an", {})["案"]})
self.assertTrue(self.post("/chat/restore", {"id": old_trial})["ok"])
def test_seiri_duplicate_first_message_keeps_newest_and_group_name_is_reused(self):
first, second = server.chats.create(), server.chats.create()
for chat in (first, second):
server.chats.rename(chat["id"], "同じ相談")
server.chats.add_turn(chat["id"], "user", "同じ最初の発言")
with server.chats._db() as conn:
conn.execute("UPDATE conversations SET updated=? WHERE id=?", (time.time()-20, first["id"]))
ideas = self.post("/chat/seiri/an", {})["案"]
trash = [x for x in ideas if x["区分"] == "ゴミ箱"]
self.assertEqual([x["id"] for x in trash], [first["id"]])
server.chats.remove(first["id"])
self.assertTrue(self.post("/chat/restore", {"id": first["id"]})["ok"])
a, b = server.chats.create(), server.chats.create()
for chat, title in ((a, "星空写真の整理"), (b, "星空写真の共有")):
server.chats.rename(chat["id"], title)
server.chats.add_turn(chat["id"], "user", title)
group_ideas = [x for x in self.post("/chat/seiri/an", {})["案"] if x["区分"] == "組"]
selected_name = next(x["組"] for x in group_ideas if x["id"] == a["id"])
self.post("/chat/seiri/suru", {"案": [{**x, "checked": x["id"] == a["id"]} for x in group_ideas]})
c, d = server.chats.create(), server.chats.create()
for chat, title in ((c, f"{selected_name}の追加案"), (d, f"{selected_name}を共有")):
server.chats.rename(chat["id"], title)
server.chats.add_turn(chat["id"], "user", title)
reused = [x for x in self.post("/chat/seiri/an", {})["案"] if x["区分"] == "組" and x["id"] in {c["id"], d["id"]}]
self.assertEqual({x["組"] for x in reused}, {selected_name})
def test_model_title_strips_empty_think_block(self):
"""10/1 本番: --reasoning-format none で「<think> </think> デスク」が題名になった。"""
import io, json as _json, unittest.mock as um
chat = server.chats.create()
server.chats.add_turn(chat["id"], "user", "えっと、デスクトップにあるフォルダの数を教えて")
server.chats.add_turn(chat["id"], "bot", "3つです")
replies = [io.BytesIO(_json.dumps([{"is_processing": False}]).encode()),
io.BytesIO(_json.dumps({"choices": [{"message": {"content": "<think>\n\n</think>\n\nデスクトップのフォルダ数"}}]}).encode())]
def fake_urlopen(*_a, **_k):
body = replies.pop(0)
body.__enter__ = lambda self=body: self
body.__exit__ = lambda *a: False
return body
with um.patch("time.sleep"), um.patch("urllib.request.urlopen", side_effect=fake_urlopen):
server._make_title(chat["id"])
self.assertEqual(server.chats.load(chat["id"])["題"], "デスクトップのフォルダ数")
def test_screen_only_new_chat_creates_no_database_row(self):
empty = server.chats.create()
with server.chats._db() as conn: # 10分より前に作られた空の会話にする(作った直後は消さない)
conn.execute("UPDATE conversations SET created = created - 900 WHERE id = ?", (empty["id"],))
result = self.post("/chat/new", {"画面だけ": True})
self.assertEqual(result, {"ok": True, "会話": None})
self.assertIsNone(server.chats.load(empty["id"]))
self.assertEqual(server.chats.listing(), [])
def test_pick_file_returns_the_native_choice(self):
selected = Path(self.tmp.name) / "選んだ.txt"
selected.write_text("ok", encoding="utf-8")
old_run = server.subprocess.run
server.subprocess.run = lambda *args, **kwargs: SimpleNamespace(
returncode=0, stdout=str(selected) + "\n", stderr=""
)
try:
result = self.post("/pick-file", {})
finally:
server.subprocess.run = old_run
self.assertEqual(result["パス"], str(selected))
self.assertEqual(result["名"], "選んだ.txt")
def test_client_disconnect_marks_the_streamed_answer_stopped_and_saves_it(self):
old_handler = server.handle_text_nagashi
old_queue_title = server._queue_title
server._queue_title = lambda _cid: None
def fake_handler(text, queue, stop, michi=None, rireki=None):
# 切断を検知するまで何度か書く。Handler が BrokenPipe を受け取ると
# stop が立ち、下の回答が履歴へ保存される。
import teachers
teachers.mado_settei(tomeru=stop)
try:
for _ in range(200):
if stop.is_set():
break
queue.put({"文字": "x"})
time.sleep(0.01)
return {
"出力": "(ここで止めた)" if stop.is_set() else "完了",
"経過": "", "ミリ秒": 0, "モード": "試験", "止めた": stop.is_set(),
}
finally:
teachers.mado_settei(None, None)
server.handle_text_nagashi = fake_handler
try:
host, port = "127.0.0.1", self.httpd.server_address[1]
conn = http.client.HTTPConnection(host, port, timeout=3)
body = json.dumps({"text": "止める試験"}).encode("utf-8")
conn.request("POST", "/ask/stream", body=body, headers={
"Content-Type": "application/json", "X-Token": server.TOKEN,
})
response = conn.getresponse()
# 最初の会話 ID を受け取ってから接続を切る。
self.assertIn("会話", json.loads(response.readline().decode("utf-8")))
response.close()
conn.close()
deadline = time.time() + 4
saved = None
while time.time() < deadline:
listed = server.chats.listing()
if listed:
saved = server.chats.load(listed[0]["id"])
if saved and len(saved["やりとり"]) == 2:
break
time.sleep(0.02)
self.assertIsNotNone(saved)
self.assertEqual(saved["やりとり"][1]["文"], "(ここで止めた)")
finally:
server.handle_text_nagashi = old_handler
server._queue_title = old_queue_title
if __name__ == "__main__":
unittest.main()
|