Backend (services/hwfit + routes):
- VRAM column sort now shows global highest first (was special-cased to
ascending then truncated top-N, which made "highest VRAM" mathematically
unreachable). Every column path uses reverse=True for the truncation.
- Hardware probe cache TTL 30min -> 24h so changing filters doesn't keep
re-probing the rig during a session; Rescan button still forces fresh.
- Multi-GPU rigs filter GGUF Q*/IQ quants (vLLM/SGLang can't serve them);
default non-prequantized to BF16 on 2+ GPUs.
- AWQ / AWQ-8bit / GPTQ-8bit get a -1.0 quality penalty so FP8 wins ties.
- Version-aware tiebreaker (parse Mn.n / Vn) — MiniMax-M2.7 ranks above M2.5.
- hf_models.json: zai-org/GLM-5.1 added; zai-org/GLM-5 quantization flipped
Q4_K_M -> BF16. DeepSeek-V4-Flash / -Pro + their -Base variants registered
with new FP4-MoE-Mixed / FP8-Mixed quant keys (calibrated BPP from the
actual 156 GB / 284 GB disk footprints).
- New FP4-MoE-Mixed + FP8-Mixed entries in QUANT_BPP / QUANT_SPEED_MULT /
QUANT_QUALITY_PENALTY / QUANT_BYTES_PER_PARAM / PREQUANTIZED_PREFIXES.
Frontend — Scan/Download:
- Engine + Quant swapped in the toolbar; Quant defaults to "All".
- Ctx (range slider) ported from origin/main: 8k/16k/32k/50k/128k/Max. Drag
re-sorts by vram ascending (smallest fitting first); back to Max → score.
- Ctx slider rail now visible — was background:transparent in a duplicate
later-cascade rule. Hardcoded grey + !important.
- Search input moved to the far right of the toolbar.
- Type/Standard default; "Context" not uppercased; Search placeholder dimmed.
- Engine "?" + Quant "?" inline help chips inside their dropdown boxes.
- Fit-column dot toggles fit-only filter; un-toggling re-sorts by VRAM desc.
- Quant column truncates to 9 chars + ellipsis ("FP4-MoE-M..."), full in
tooltip. Smart title-suffix strips the parts already in the repo name
(QuantTrio/MiniMax-M2-AWQ + quant AWQ-4bit -> just "(4bit)").
- Conditional warning for safetensors models on non-GPU rigs only.
- Dependency Install / Installed / Installed▾ / N/A all 75.85px wide.
- Rebuild llama.cpp moved into the llama_cpp dep row, styled as a tag.
- Foldable Download admin-card (h2 chevron); line under h2 only when folded.
- HF token save gets a green ✓ + "Saved" flash.
- Cached scan no longer counts stalled rows as downloaded.
- Footer: "Request it →" link with GitHub mark to the public discussion
(#1962) for model-add requests.
Frontend — Running tab:
- Strict download-finish check (DOWNLOAD_OK or /snapshots/, not bare
"Download complete"). True overall % for multi-shard downloads:
((N-1)+frac)/total instead of hf_transfer's per-shard aggregate.
- ETA in the uptime ticker: "downloading: 12m 34s · ETA 1h 23m".
- Clear button kills the tmux session too; if the output still shows a
live shard line, the pill is hidden + relabels as "reconnect" + revives
on click.
- Self-heal: on cookbook open AND every bg-monitor cycle (10s, throttled
to 8s), scan persisted done/error/crashed downloads and probe their
tmux session — if alive, flip status back to running and reattach.
- Per-launch zombie probe: clicking Download on a model whose persisted
state is done but tmux is still alive revives the existing task and
refuses to start a duplicate.
- Pre-launch GPU probe: vllm / sglang / diffusers serve check
/api/cookbook/gpus first; warns + confirms if no GPU is visible.
- Server-side state guard: rejects "done" POSTs for downloads lacking
DOWNLOAD_OK / DOWNLOAD_FAILED / /snapshots/ when the last-mentioned
shard is N<total — stale tabs can't poison persisted state any more.
- Running count includes tasks whose output looks active even if persisted
status got stuck. Dir text on the running row, font matched to uptime.
Serve panel:
- Ctx text input always resets to model max on open (default 20000 when
metadata is missing).
- Max Seqs default 8 -> 4. KV Cache dtype select 32px tall.
- Lightning icon on Launch (same as Action toggle).
- Diagnosis card simplified (no fold/copy/dismiss), suggestion font
matches body; action buttons get icons on the left (Retry/Copy/Edit/
Install/Kill/Switch/etc.).
- Incomplete-download serve warning when model status is
downloading / stalled / has_incomplete.
- MTP "?" tooltip ("supported on a few model families … up to ~3× faster").
Backend (services/hwfit + routes):
- rank_models picks visible set by REQUESTED column, not always score —
sorting by Param now shows highest-param models PERIOD (incl. too_tight).
- New fit_only param. Multi-GPU rigs filter GGUF Q*/IQ quants (vLLM/SGLang
cannot serve them); default non-prequantized to BF16 on 2+ GPUs.
- AWQ / GPTQ-8bit get a -1.0 quality penalty (was 0.0, tied with FP8), so
FP8 wins when both fit.
- Version-aware tiebreaker (parse Mn.n / Vn) — MiniMax-M2.7 ranks above
M2.5 on equal composite score; >=100B integers not misread as versions.
- /api/cookbook/hf-latest no longer drops models without an "NB" pattern in
the repo id (MiniMax-M2.7, DeepSeek-V4-Pro etc. were silently filtered).
- Cached-model scan: atexit flushes models JSON even if the script is
killed mid-walk; each scan_dir wrapped in try/except; timeout 60s -> 180s.
- KB granularity for sub-MB sizes (was "0 MB" for 12 KB shells). New
"stalled" status for shells <1 MB with no .incomplete files.
- /api/cookbook/state POST guard: rejects "done" download tasks lacking
DOWNLOAD_OK / DOWNLOAD_FAILED / /snapshots/ when the last-mentioned
shard is N<total — stops stale tabs from poisoning persisted state.
- hf_models.json: add zai-org/GLM-5.1; flip zai-org/GLM-5 quantization
Q4_K_M -> BF16 (it is the native base, not a quant).
Frontend (static/js):
- Scan/Download toolbar: quant defaults to All; ctx slider (8k/16k/32k/
50k/128k/Max) ported from origin/main with sort=fit on drag, sort=score
on Max. GPU toggle commits _activeCount to maxGpu on initial render. Fit
column header tagged with active budget (RAM / GPU / N GPU).
- Foldable Download admin-card: the Download h2 is the chevron trigger;
state persists in localStorage.
- Download card surfaces destination dir (Dir: <path>). Same dir on running
task row, font/color matched to uptime (9px Fira Code muted, opacity .4).
- Serve panel ctx text input always resets to model max on open. Sub-MB
cached models show with red "download stalled" badge.
- Bulk-select Cancel + Delete reset the Select button label on exit.
- Cookbook running: false-finished bug fixed — DOWNLOAD_OK or /snapshots/
required; bare "Download complete" no longer marks the task done after
the first config file. Clear button now sends tmux kill-session too.
True overall % for multi-shard downloads: ((N-1)+frac)/total instead of
hf_transfer per-shard aggregate.
- Diagnosis card simplified: removed fold toggle, copy button, dismiss X.
Suggestion font matches message body (12px).
- HF token field flashes green check + "Saved" on save.
- Cached scan no longer counts stalled rows as downloaded in Scan/Download.
CSS:
- dep Install button width pinned to 76px to match Installed split.
- task-sub row +1px; task-status badge gets margin-right 8px.
- Ctx slider styled like gallery editor sliders (thin pill rail, red thumb).
- Bulk-select cancel button top -3px -> -5px.
The empty-state tip ("Add an AI endpoint from Settings...") shares a 60px
max-height ceiling with the one-line .welcome-sub / .welcome-version. On
narrow phones the welcome block shrink-wraps and the tip wraps to 4-5 lines
(~67px), so the shared ceiling clipped its last line ("...key into the
chat.") - the only setup hint a first-run user gets.
Give .welcome-tip its own taller max-height (120px), placed above the
@media (max-height: 650px) block so that rule's max-height:0 still collapses
the tip on short viewports. .welcome-sub / .welcome-version are untouched,
and desktop is unchanged (the tip is ~50px there, well under the ceiling).
Sidebar section titles were vertically clipped in Chromium/Edge (fine in
Firefox). Raise line-height 1 → 1.3, mirroring the existing .list-item fix.
The titles are flex-centred in a fixed-height (29px) header, so this adds
glyph headroom without any reflow.
iOS Safari auto-zooms when a focused input has font-size < 16px. Bump
text-entry controls to 16px under (hover: none) and (pointer: coarse) so
desktop sizing is untouched. Date/time inputs and selects are excluded —
they open native pickers and never zoom.
Doc-editor tiers keep their size hierarchy: Large lands at 18px (above the
16px threshold) instead of collapsing onto Medium, and the email rich-body
Large (17px) is left alone since it was already zoom-safe. All three editor
layers (textarea, highlight overlay, line numbers) move together so the
syntax overlay stays metrically aligned.
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
Docking a modal to a window edge pushes the chat aside (body padding via
right-dock-active + --right-dock-w). Three problems on close/reopen:
1. Chat stayed offset after closing a docked modal. The close-watcher only
reacted to the `.hidden` class or DOM removal, but the draggable modals
(calendar, plan, workspace, document, …) close via inline `display:none`.
Watch the `style` attribute too and treat `display:none` as closed.
2. Reopening a previously-docked singleton modal floated it off to the side,
overlapping the chat. The reused element kept its docked inline geometry.
Clear the content's inline position/size on close so it reopens at its CSS
default (centered).
3. Undock wasn't animated. The transition lived on `.right/left-dock-active`,
so removing the class dropped the transition with it and padding snapped to
0. Move the transition to the base `body` so the push animates both ways.
Files: static/js/modalSnap.js, static/style.css.
Checks: node --check static/js/modalSnap.js; verified in-browser (dock → close
→ chat animates back; reopen → centered, no overlap).
Rebased on current main. Integrates with the new Recent/Favorites
system — provider groups appear below Recent and Favorites in browse
mode for large catalogs (>12 models).
Changes:
- Models grouped by canonical provider with collapsible sections
- Chevron animation consistent with sidebar sections
- Domino cascade on expand (only on just-opened group)
- Provider display names (deepseek-ai -> DeepSeek, meta -> Llama, etc.)
- Alias merging (meta + meta-llama -> one Llama group)
- Search includes provider display names for filtering
- Collapsed state persists in localStorage
- No screenshot binary committed
Co-authored-by: danielxb <5981902+danielxb@users.noreply.github.com>
* Cookbook: Engine filter + intelligent hardware-computed serve profiles
Two related Cookbook serving improvements for accurate, hardware-aware model
serving (especially on consumer GPUs that can only run GGUF/llama.cpp).
Engine filter
- New "Engine" dropdown (All / llama.cpp / vLLM / SGLang) beside the quant
picker. Pure client-side view filter over the fetched list via the same
_detectBackend() the serve commands use, so what you filter to is exactly what
would launch. Re-renders from cache (no refetch). Empty-state message + the
instant-cache-paint path account for it too.
Intelligent serve profiles (Quality / Balanced / Speed)
- services/hwfit/profiles.py: compute_serve_profiles() turns detected VRAM +
model size into concrete llama.cpp flags (n_gpu_layers, n_cpu_moe, cache-type,
context). Encodes the by-hand tuning: a too-big MoE offloads experts to CPU
instead of failing; a model that fits stays fully on GPU; quant tracks profile
intent; vision models keep image-encoder headroom. Reuses models.py VRAM math
so filtering and serving agree on what fits. Pure/deterministic (no t/s claims
— partial-offload speed isn't reliably predictable; fit is what's computed).
- /api/hwfit/profiles endpoint returns the profiles + the model's trained
context limit, with loose name matching (strips org/ prefix, -GGUF suffix,
quant tag) so a local GGUF folder name resolves to its catalog entry.
- _buildServeCmd (llama.cpp) now emits --n-cpu-moe / --flash-attn /
--cache-type-k/v when set, with llama-cpp-python fallback equivalents. It
previously only set -ngl/-c, which is why it OOM'd or ran slow.
- Serve panel: profile chips that fill the fields on click, plus CPU-MoE / KV
Cache / Flash Attn fields. Context is clamped to the model's trained limit
(and an absolute 1M sanity ceiling) on type/blur/profile-load and at launch —
fixes a crash where a stale 256k/16M preset + quantized KV cache caused an
amdgpu ErrorDeviceLost.
Tests: tests/test_serve_profiles.py (7) — offload vs full-GPU fit, never exceed
VRAM, context cap, launchable flags, vision headroom, no-GPU empty.
Checks: py_compile + node --check pass; pytest test_serve_profiles + test_hwfit_amd
green; verified live on an RDNA4 box (gfx1200) — Balanced lands ~ncm18 q4 128k,
matching hand-tuning.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* Cookbook: make column-header sorting discoverable (incl. Newest)
Sorting in Cookbook is via clickable column headers (pewds' design), but the
headers had no visual cue that they're interactive — so sorting in general, and
the Newest sort on the Model header specifically, was undiscoverable.
- Style sortable headers as interactive: pointer cursor, hover underline, and
the active sort column bolded/highlighted. There was no CSS for
.hwfit-sortable / .hwfit-sort-active at all; this helps every existing sort,
not just Newest.
- The Model column header sorts by release_date (newest first), reusing the
existing header-click sort wiring and the "newest" SORT_KEY.
No new sort control — uses the existing column-header paradigm.
Checks: node --check passes.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* Cookbook serve profiles: keep the on-disk file's quant fixed (don't propose Q6/Q2)
In the Serve tab the model is a specific GGUF file already on disk, so its quant
can't change — but the profiles were suggesting "Quality · Q6_K" / "Speed · Q2_K"
as if you could re-quantize it. That's meaningless when serving a fixed file.
- compute_serve_profiles gains serve_weights_gb / serve_quant. When set (SERVE
mode), the quant is locked to the file's and profiles differ only in the real
serving knobs — n_cpu_moe, KV-cache type, context. _weights_gb / _cpu_moe_for_budget
use the file's actual size instead of a quant-derived estimate. DOWNLOAD mode
(no override) still varies the quant to show download options.
- /api/hwfit/profiles accepts serve_weights_gb & serve_quant.
- The Serve panel parses the file's size (from m.size "20.6 GB") and quant (from
the repo/file name) and passes them, so profiles match what's actually served.
Result for a 20.6 GB Q4_K_M file: all three profiles stay Q4_K_M and differ by
KV/ctx/offload (Quality q8 KV 128k ncm21, Balanced q4 128k ncm17, Speed q4 32k
ncm15) — no nonsensical quant changes.
Tests: test_serve_mode_keeps_fixed_quant. Full serve-profile suite green (9).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* Cookbook serve: Vision toggle (auto-find mmproj) + live VRAM/RAM-spillover monitor
Two serve-panel additions:
1. **Vision toggle.** A "Vision" checkbox that serves the model with its
multimodal projector so it can read images. The mmproj path is resolved at
runtime (find mmproj-*.gguf next to the model), so dropping an mmproj file in
the model folder makes the toggle just work; `--mmproj … --image-max-tokens
1024` (native) / `--clip_model_path` (llama-cpp-python) only when on + found.
2. **Live GPU-memory monitor.** A readout that polls /api/cookbook/gpus every 4s
while the panel is open and shows VRAM used/total/%, free, and — crucially on
a discrete card — **RAM spillover** (AMD gtt_used_mb), with a plain-language
health hint: green/healthy, amber/tight, red/"spilled to RAM — slow (raise
CPU MoE or lower context)". Surfaces gtt_used_mb from the gpus endpoint
(previously read for total only and discarded for 'used').
Lets you see at a glance whether a config fits VRAM (fast) or is paging to system
RAM over PCIe (slow) instead of guessing.
Checks: node --check + py_compile pass.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Two small polish items in the Cookbook Serve panel.
Saved-config badge
The little count badge next to the Save button ("3 ▾" etc.) had a
generic "Saved launch configs" tooltip, so the number reads like a
notification dot. Make it spell out what it is and what clicking does:
"3 saved launch configs for <model> — click ▾ to load or delete"
(and "No saved launch configs for <model> yet — click Save to add
one" when empty). Tooltip stays in sync via _updateSavedToggleLabel
so save/delete updates both the count and the hint.
GPU chip on mixed-GPU boxes (#711)
The chip label was `${gpuCount}x ${gpu_name}`, where gpu_name is
just gpus[0].name — so a 4090 + 3060 reads as "2x RTX 4090". The
backend already emits gpu_groups (identical cards grouped, used by
the serve flow to pin CUDA_VISIBLE_DEVICES) and a per-card gpus[]
array, so use them:
- Label renders each homogeneous pool: "1× RTX 4090 + 1× RTX 3060".
Homogeneous setups keep the existing "2× RTX 4090" form.
- Tooltip lists each GPU with its index + VRAM, useful for picking
the right device when launching.
Refs #711.
First incremental pass at issue #86, focused on the universal entry
points and primary navigation. All changes verified in-browser with the
axe-core engine (0 violations on the surfaces below) plus manual keyboard
testing, on both desktop (1280px) and mobile (390px).
Login / first-run setup (static/login.html)
- Add a real <h1>, wrap content in <main> + <footer> landmarks.
- Mark the decorative boat SVG aria-hidden.
- Errors now use role="alert" so screen readers announce them.
- "Remember me" checkbox is keyboard-focusable (was display:none) with an
accessible name and a focus ring; dynamic 2FA field gets a linked label.
- Darken the brand-red submit button so white text clears WCAG AA 4.5:1
(was ~3.2:1); add visible :focus-visible rings.
App shell (static/index.html, static/style.css)
- Remove invalid role="region" from the <main> chat container (it was
overriding the implicit main landmark).
- Add a persistent, visually-hidden <h1> inside <main> so the page always
exposes one logical level-1 heading — works even on mobile where the
sidebar (with the visible brand) is hidden off-canvas.
- Add a reusable .a11y-visually-hidden utility.
- Raise chat-title, model-picker, settings-helper and notes text contrast
above 4.5:1 (were 2.8-3.9:1).
Keyboard nav + dialogs (static/js/a11y.js - new)
- Make the click-only <div> sidebar navigation (New Chat, Search, Brain,
Calendar, Compare, Cookbook, Deep Research, Gallery, Library, Notes,
Tasks, Theme, account) focusable and Enter/Space-activatable, announced
as buttons (skipping role=button where a nested control would create a
nested-interactive violation). Visible focus ring reused from existing
.list-item:focus-visible.
- Upgrade modals (.modal-content and the docked .notes-pane) to labelled
role="dialog" + aria-modal, and normalise their title to heading level 2
so heading order stays valid. A MutationObserver covers runtime-rendered
rows and modals.
Decorative background canvases (static/js/theme.js)
- Mark all 7 bg-effect canvases aria-hidden.
Notes & Tasks (static/js/notes.js, static/js/tasks.js)
- Label the icon-only Note/To-do toggle pills (fixes a critical
button-name issue) and track aria-pressed state.
- Improve Notes header-button + empty-state contrast.
- Give the Tasks sort <select> an accessible name (fixes a critical
select-name issue).
Remaining data-dense tool modals (Tasks cards, Calendar, Gallery, Email,
Cookbook, Compare, Deep Research) still have muted-text contrast to polish
and are the next incremental step, per the issue's own guidance.
Replace the flat dump of every model in the chat-input picker with a
quick-switch. Opening the picker now shows a search box, an auto-tracked
Recent list (last 5 picks), and a manual Favorites list instead of every
available model crammed into a 280px dropdown. With large catalogs
(e.g. OpenRouter's 350+ models) this was unusable as both a quick-switch
and a browser.
- Recent: each pick is recorded most-recent-first (capped at 5) under a
new odysseus-model-recent key, so the next open has it one click away.
- Favorites: an inline star on every row toggles favorite state and
writes the existing odysseus-model-favorites key, so the sidebar Models
section stays in sync. The star toggles only — it never picks the model.
- Search filters a flat list across the whole catalog; favorited rows
keep their filled star while filtered.
- Small catalogs (<=12 models) still list everything in browse mode so
tiny installs aren't forced to search for a model.
- Touch friendly: stars are always visible (no hover-reveal) and tap
targets grow on narrow screens.
No changes to sidebar visibility defaults.
Closes#399
- Turn the "/setup" text on the welcome screen and fallback state into a clickable link that automatically runs the setup command.
- Add an interactive down-arrow "Use in Chat" button next to copy button on typewriter-generated setup code blocks.
- Programmatically trim the "..." placeholder when inserting API keys, focusing the cursor right after "sk-".
- Implement click-delegation for supported provider spans and raw code elements inside the setup guide to instantly pre-populate the input bar.
Library, Notes, and the other floating tool windows (Tasks, Calendar,
Gallery, Email, Cookbook, Brain, Settings, Theme, Compare, Research,
Sessions) could be moved and snapped but never resized — there were no
resize handles and dragging the edges did nothing.
Add a shared makeWindowResizable() helper and wire it into the existing
makeWindowDraggable() so every draggable window gains native-style
edge/corner resizing from one place:
- Grab any of the four edges or four corners to resize; the cursor
reflects the active handle (ew/ns/nwse/nesw-resize).
- Detects pointer proximity to the border instead of injecting handle
elements, so it works regardless of each window's overflow model
(.modal-content scrolls its body; .notes-pane scrolls an inner el).
- Min-size clamp (320x200) and viewport clamping so a window can't be
collapsed to nothing or dragged off-screen.
- Per-window size is remembered and restored on reopen.
- Disabled on mobile (windows are full-screen sheets there) and while a
window is docked or fullscreen-snapped.
- Touch supported at tablet width and up; self-heals a missed pointer-up
so a lost mouseup can't leave a window stuck in resize mode.
The photo-detail view is an absolutely-positioned (inset:0) overlay
inside .gallery-images-container, so its height resolved to the photo
grid sitting behind it. When the library has only a few photos that grid
is short, which crushed the detail view: the image was clipped and the
metadata sidebar (overflow-y:auto) was squeezed into a tiny,
internally-scrolling strip. With a large library the grid is tall, which
is why the panel looked fine in the demo video but cramped for users with
few photos.
When the detail view is open, hide the grid-view siblings and drop the
overlay into normal flow so the container -- and the window, up to its
existing 92vh max-height -- sizes to the detail's own content (image +
metadata). Nothing is clipped or squeezed regardless of how many photos
exist. Works on both desktop and the mobile full-screen sheet; the grid,
albums and editor views keep sizing to their own content.
Also add before/after comparison screenshots (docs/gallery-314-*.png).
The Settings window inherited the base `.modal` vertical centering
(`align-items:center`). Its height is content-driven, so every tab is a
different height — and a vertically centered window grows and shrinks
around its own midpoint, making the in-modal nav rail (and the whole
window) appear to jump vertically when switching between pages.
Top-anchor the Settings window on desktop (`align-items:flex-start` plus
a fixed `margin-top`) so the top edge stays put and the panel only ever
grows downward. Scoped to desktop only — on mobile the panel is a
full-height bottom sheet that is already stable. Opening and dragging the
window both clear the inline margin/top, so window placement is otherwise
unchanged.
Fixes#208
Two bugs hid the popup that opens on double-click (or right-click) of
a GPU button in the Serve panel:
1. z-index 240 vs the cookbook modal at 260 — popup rendered behind
the modal it was spawned from.
2. Horizontal position was just `button.left`, with no clamp against
the viewport. GPU buttons sit near the right edge of the modal, so
the popup got anchored at a left that pushed most of its body past
the viewport's right edge.
Switch the popup to position:fixed (escapes scrolling / transform
stacking contexts on any ancestor), bump z-index to 10010 (above the
themed-confirm / overlay layer that sits around 9000-10000), and
clamp left/top after measuring the rendered size — including flipping
above the button if there isn't room below. The popup is now fully
visible regardless of which GPU button it's anchored to or how
narrow the viewport is.
