Skip to content

Qwen3.8-Flash-Next on any consumer hardware: one-click install for Windows / Linux. Strata inference engine, OpenAI/Anthropic API on localhost, optional image input.

C++

launch: model list shows the app's RAM+VRAM figure and the SSD part; strata monitor gets the Connect view (7); Mac env names

Niko122164fff9e14982
diff --git a/strata_cli/launch/monitor_ext.py b/strata_cli/launch/monitor_ext.py
index b387a3100..c13224243 100644
--- a/strata_cli/launch/monitor_ext.py
+++ b/strata_cli/launch/monitor_ext.py
@@ -25,9 +25,10 @@ from strata_cli.render import VIEWS, safe
class LaunchMonitor(MonitorApp):
BINDINGS = MonitorApp.BINDINGS + [Binding("7", "connect", "Connect view", show=False)]
- def __init__(self, endpoint, snapshot, *, info: dict, **kw) -> None:
+ def __init__(self, endpoint, snapshot, *, info: dict, ask_quit: bool = True, **kw) -> None:
super().__init__(endpoint, snapshot, **kw)
self.info = info
+ self.ask_quit = ask_quit # False: plain `strata monitor` (q detaches, nothing to stop)
self.connect_enabled = True
self.connect_open = False
self.connect_index = 0
@@ -41,6 +42,8 @@ class LaunchMonitor(MonitorApp):
def action_quit(self) -> None:
if self.closing:
return
+ if not self.ask_quit:
+ return super().action_quit()
def done(choice: str | None) -> None:
if choice == "keep":
@@ -122,7 +125,9 @@ class LaunchMonitor(MonitorApp):
def _rendered(self) -> dict:
h = self._harness()
key = self.info.get("key") or ""
- return harness.render(h, harness.target(self.info["url"], key, self.info.get("model") or "strata", self.info.get("context")))
+ if not self.info.get("model"):
+ self.info["model"] = ((self.latest or {}).get("model") or {}).get("id") or "strata"
+ return harness.render(h, harness.target(self.info["url"], key, self.info.get("model") or ((self.latest or {}).get("model") or {}).get("id") or "strata", self.info.get("context")))
def _copy(self) -> None:
out = self._rendered()
@@ -185,7 +190,7 @@ class LaunchMonitor(MonitorApp):
def _paint(self, size=None) -> None:
if not getattr(self, "connect_open", False):
super()._paint(size)
- if not self.error and not self.selection_notice and self.connection_state == "connected" and not self.frozen:
+ if self.ask_quit and not self.error and not self.selection_notice and self.connection_state == "connected" and not self.frozen:
try:
self.query_one("#message", Static).update("Read only | q asks: keep serving or stop | 7 connect an app")
except Exception: # noqa: BLE001
diff --git a/strata_cli/launch/screens.py b/strata_cli/launch/screens.py
index 6886a4f12..14a2770eb 100644
--- a/strata_cli/launch/screens.py
+++ b/strata_cli/launch/screens.py
@@ -353,7 +353,7 @@ class ConfirmScreen(ModalScreen):
class ModelScreen(Base):
step_id = "model"
hints = "↑↓ choose Enter next o use my own files Esc back ? help"
- help_text = ("Pick the model. 'Needs' is the graphics memory and RAM it uses on this computer; the n-gram table stays on "
+ help_text = ("Pick the model. 'Needs' is the RAM + graphics memory the model wants (the app's figure); 'uses SSD N GB' is the part that does not fit and is read from the SSD while it answers; the n-gram table stays on "
"your SSD and is not counted. 'Does not fit' is a warning, not a block: you can still try it.")
BINDINGS = Base.BINDINGS + [Binding("o", "own", "Own files", show=False)]
@@ -375,8 +375,8 @@ class ModelScreen(Base):
def widths(self) -> list[tuple[str, int]]:
w = self.width()
if w >= 100:
- return [("Model", 42), ("Download", 10), ("Needs (graphics + RAM)", 24), ("On this computer", 0)]
- return [("Model", 41), ("Download", 9), ("Needs (gfx + RAM)", 18), ("On this computer", 0)]
+ return [("Model", 42), ("Download", 10), ("Needs (RAM + VRAM)", 20), ("On this computer", 0)]
+ return [("Model", 41), ("Download", 9), ("Needs (RAM+VRAM)", 17), ("On this computer", 0)]
def build_menu(self, initial: bool = False) -> None:
s = self.ctx.settings
@@ -389,8 +389,9 @@ class ModelScreen(Base):
for i, r in enumerate(self.rows):
col = {"fits": "", "tight": WARN, "no": BAD}[r.verdict]
tag = (" (recommended)" if r.recommended else "") + (" (installed)" if r.installed else "") + (" [experimental]" if r.experimental else "")
- need = f"{r.need_vram_gb:.0f} GB + {r.need_ram_gb:.0f} GB" if r.need_vram_gb or r.need_ram_gb else "-"
- items.append(Item([r.title, f"{r.download_gb:.0f} GB", need, Cell(r.words + tag, col)], key=(r.family, r.model)))
+ need = f"{r.memory_gb:.0f} GB" if r.memory_gb else "-"
+ ssd = f" · uses SSD {r.spill_gb:.0f} GB" if r.spill_gb >= 1 else ""
+ items.append(Item([r.title, f"{r.download_gb:.0f} GB", need, Cell((r.words.split(",")[0].split(":")[0] if ssd and "SSD" in r.words else r.words) + ssd + tag, col)], key=(r.family, r.model)))
if (s.family, s.model) == (r.family, r.model):
sel = i
m = self.query_one(Menu)
diff --git a/strata_cli/monitor_commands.py b/strata_cli/monitor_commands.py
index f11477378..ead6429cd 100644
--- a/strata_cli/monitor_commands.py
+++ b/strata_cli/monitor_commands.py
@@ -193,8 +193,14 @@ def run(a: argparse.Namespace) -> int:
from strata_cli.monitor import MonitorApp
except ImportError:
return _error("monitor component missing; install the optional terminal UI dependency", 5, machine)
- MonitorApp(endpoint, None, interval=a.interval, requests=a.requests,
- theme=a.theme, color=a.color, glyphs=a.glyphs).run()
+ try: # the monitor with its Connect view (key 7); q just detaches here
+ from strata_cli.launch.monitor_ext import LaunchMonitor
+ LaunchMonitor(endpoint, None, info={"url": endpoint.base_url, "urls": [endpoint.base_url], "key": endpoint.credential or "", "model": "",
+ "profile": None, "name": None}, ask_quit=False, interval=a.interval, requests=a.requests,
+ theme=a.theme, color=a.color, glyphs=a.glyphs).run()
+ except ImportError:
+ MonitorApp(endpoint, None, interval=a.interval, requests=a.requests,
+ theme=a.theme, color=a.color, glyphs=a.glyphs).run()
return 0
except ClientError as e:
return _error(str(e), e.exit_code, machine)
diff --git a/tests/terminal/test_launch_tui.py b/tests/terminal/test_launch_tui.py
index 2ff66e880..7f4dba424 100644
--- a/tests/terminal/test_launch_tui.py
+++ b/tests/terminal/test_launch_tui.py
@@ -222,15 +222,17 @@ class ModelStep(Base):
await press(pilot, "enter")
t = text_of(app)
self.assertIn("Which model?", t)
- for col in ("Model", "Download", "Needs (graphics + RAM)", "On this computer"):
+ for col in ("Model", "Download", "Needs (RAM + VRAM)", "On this computer"):
self.assertIn(col, t)
self.assertIn("IQ3_XXS", t)
line = next(l for l in t.splitlines() if "IQ3_XXS" in l and "Qwen3.8-Flash-Next IQ3_XXS" in l)
self.assertIn("76 GB", line) # setup.py's download_gb 75.8
- self.assertIn("12 GB + 36 GB", line) # fitOn: 12 on the card, 35.9 in RAM
+ self.assertIn("44 GB", line) # the app's figure: experts 42.9 + 1 GB, RAM + VRAM together
self.assertIn("(recommended)", line)
self.assertIn("› ", line) # the cursor starts on the recommendation
self.assertIn("does not fit (can still try)", t) # NVFP4: a warning, not a block
+ big = next(l for l in t.splitlines() if "UD-Q5_K_XL" in l)
+ self.assertIn("uses SSD", big) # a model bigger than RAM says how much is read from the SSD, as the app does
self.assertIn("The n-gram table stays on your SSD", t)
self.run_app(case)