* feat(AI): per-model thinking toggle with global default (off)
Stop forcing thinking on for every capable model. Adds a global default
(ai.autoThinking, ships OFF) and a per-model override in the chat window,
shown only for thinking-capable models and remembered client-side.
The /v1 (OpenAI-compat) endpoint ignores `think`; reasoning_effort is the
real lever. The controller resolves per-request preference -> global default
-> OFF (gated on capability), and the service maps that to reasoning_effort
('none' to suppress on a capable model, 'medium' for gpt-oss, unset to let a
capable model default thinking on). A new thinkingCapable flag keeps
non-Ollama backends from ever receiving reasoning_effort.
installed-models is enriched with a `thinking` flag (checkModelHasThinking,
now memoized) so the picker knows which models get the toggle.
Stacks on #1078 (reasoning-field read + client-disconnect abort).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
* feat(AI): thinking toggle as NOMAD Switch + info tooltip
Address review feedback on the per-model thinking control:
- Use the shared Switch component (matches AI Assistant settings) instead
of a raw checkbox.
- Label "Thinking:" with a colon to match the adjacent "Model:" label.
- Add an InfoTooltip explaining what thinking does and where the default
lives. Extend InfoTooltip with optional `position` ('top'|'bottom') and
`align` ('center'|'right') so it opens downward and expands leftward from
the right-edge header slot instead of being clipped/crushed against the
viewport edge. Defaults preserve existing (benchmark page) behavior.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Step 1: remove the "Additional Tools" (Notes/Data Tools) section from
onboarding and point users to Supply Depot, the browsable app catalog,
for everything beyond the three core capabilities.
Step 2: add a note that individual countries and a full global map can
be installed any time from the Maps Manager.
Step 4: default the KB auto-index policy to "Ask me first" (Manual)
instead of "Yes, always" — auto-indexing has cost/resource implications
a non-technical user won't anticipate from the toggle alone.
ollama_service: the recommended-models fallback only fired on a null
result, so a successful-but-empty upstream response (models: []) showed
"No recommended AI models available" and poisoned the 24h cache. Now the
fallback fires on empty too, empty results are never cached, empty caches
are ignored, and the upstream request has a 10s timeout.
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Two defects made chat hang forever with thinking-capable models on the OpenAI-compat
(/v1) path, which NOMAD uses for both local and remote Ollama:
1. Field mismatch. chatStream()/chat() read `delta.thinking` / `message.thinking`,
but Ollama's /v1 endpoint emits thinking tokens as `reasoning`. All thinking output
was silently dropped, so the SSE stream was nothing but empty content+thinking chunks
and never reached done. Now read `thinking ?? reasoning` in both paths (the inline
<think>-tag parser for other backends is unchanged).
2. No abort on client disconnect. When the user gave up and closed the chat, the
upstream generation kept decoding server-side. With Ollama's default
OLLAMA_NUM_PARALLEL=1 that abandoned request occupied the only slot, so every later
chat/RAG request queued behind it and the whole assistant appeared dead. The
controller now wires an AbortController to the response 'close' event and threads the
signal into the OpenAI SDK request, so a disconnect aborts the upstream generation.
Verified on NOMAD2 (qwen3:0.6b, which reports the `thinking` capability and emits
`reasoning` on /v1): before, the stream was endless empty chunks; after, thinking streams
visibly and reaches done. On disconnect, Ollama's decode counter freezes and the server
logs `cancel task` / `slot release`, freeing the slot for the next request.
Note: thinking is still force-on for capable models here; a user-facing per-model
thinking toggle (default off) is a planned follow-up.
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Dense source content produces chunks that exceed the embedding model's
context window (nomic-embed-text:v1.5 defaults to 2048 tokens). Two paths
hit this even after the prior pre-cap:
- Older Ollama (e.g. 0.18.1, #944) ignores the num_ctx=8192 we send on
/api/embed, so it stays at the model's 2048 default.
- The OpenAI-compat /v1/embeddings fallback didn't pass num_ctx/truncate
at all, so any Ollama drops to 2048 whenever it lands on the fallback.
When a chunk overflowed, the 400 was swallowed and the chunk was silently
dropped from Qdrant. Worse, the failure propagated to EmbedFileJob, which
re-embeds the entire file on each of its 30 BullMQ attempts — the "endless
queue loop" / "api/embed for weeks" / pegged GPU reported in #944/#959.
Fix:
- OllamaService.embed(): on a context-length error, retry once with an
aggressive 2048-safe cap (EMBED_CONTEXT_SAFE_CHARS = 2000) so the chunk
is embedded (start-of-chunk) instead of dropped. Native-path context
errors now bubble to this retry instead of falling through to the
smaller-context fallback. Split the native+fallback attempt into
_embedWithFallback().
- Pass truncate/num_ctx on the /v1/embeddings fallback too (Ollama's
OpenAI-compat shim forwards them).
- EmbedFileJob: classify "input length exceeds context length" as an
UnrecoverableError so one permanently-oversized chunk can't trigger 30
full-file re-embeds.
- Add OllamaService.isContextLengthError() shared by both.
Graceful degradation: a truncated chunk loses its tail but is kept in the
index, which is strictly better than today's silent drop + retry storm.
Refs #881. Supersedes the #369/#670 symptom closures that never fixed the
fallback path.
Lift the hardcoded 'nomic-embed-text:v1.5' string out of both
RagService and OllamaService into a shared EMBEDDING_MODEL_NAME
constant in constants/ollama.ts. The duplicate in OllamaService
existed only to dodge a circular import with RagService; the
constants module has no service imports, so a shared constant
eliminates both the duplication and the drift risk called out
in the inline "keep in sync" comment.
Surfaces NOMAD's previously-silent model-stacking behavior and enforces a
"one chat model in VRAM at a time" invariant (the embedding model is
always exempt). Addresses Chris's NOMAD3 testing observation that
switching the dropdown in the chat header was invisibly slow on low-VRAM
hardware because the prior model was never unloaded — Ollama would
either evict it under memory pressure or load the new one on CPU after
the runner choked.
Three integration points all funnel through one new helper:
- **User changes the model dropdown** in an active chat session →
confirm modal "Switch to {newModel}? Switching to {newModel} will
start a new chat. Your current conversation stays available in the
sidebar." On confirm, fire `keep_alive: 0` against the previous chat
model, clear active session, set the new selection. Cancel snaps the
visible dropdown back to the previous value (no popup state leaks
into `selectedModel`).
- **User clicks a session in the sidebar** → no popup (system-initiated).
Restore the session's stored model into the dropdown and fire
`unloadChatModels(targetModel)` so anything that isn't the target
gets the unload hint.
- **Chat page first mount** → page-load normalization. Anything stacked
from a prior session gets the unload hint with the current selected
model as the target-to-preserve. Guarded by a ref so it only fires
once per page lifetime; gated on `selectedModel` being populated.
Backend surface is a single new helper and a single new route:
`OllamaService.unloadAllChatModelsExcept(targetModel: string | null)`
→ queries `/api/ps`, filters out (a) the embedding model name
(hardcoded `nomic-embed-text:v1.5` to avoid the RagService circular
import) and (b) `targetModel`, fires `POST /api/generate` with empty
prompt + `keep_alive: 0` in parallel against everything else.
Returns the names that were hinted. Best-effort: network or Ollama
errors are logged and swallowed so callers don't fail on housekeeping.
`POST /api/ollama/unload-chat-models` → thin wrapper validating
`{ targetModel?: string | null }`.
Why `keep_alive: 0` is safe against in-flight inference: per Ollama's
scheduler semantics, the hint sets the post-completion eviction timer
to zero — the runner is not terminated. If Session A is mid-response
on gemma when Session B fires the unload, gemma stays resident until
A's request completes, then evicts. The user-visible worst case is the
race where A's longer-running request re-extends the timer back to the
default and the unload is no-op'd; the next transition (or page reload)
gets another chance, and Ollama's own LRU catches up under memory
pressure regardless. Robust in-flight tracking deferred to a follow-up
if we see stale-state in the wild.
Base `rc`: v1.40.0 will inherit everything from rc.6 via the backmerge.
Frontend tests deferred to a follow-up PR; existing inertia tsconfig
errors are pre-existing and unrelated.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The OpenAI-compatible /v1/embeddings fallback path can't pass
`truncate:true` / `num_ctx:8192` to the model, so any chunk that
exceeds the model's loaded context_length (often 2048 for
nomic-embed-text:v1.5) returns a 400 BadRequestError and is silently
dropped from Qdrant. Two CPU-only ingestion runs on NOMAD1 hit this
on dense technical content (medlineplus, arduino.stackexchange) even
after PR #763's num_ctx fix on the native path.
Pre-cap each input string at 4000 chars before either backend call.
That's ~1000-2000 tokens depending on density, comfortably under the
model's 2048 default. The chunker in RagService is sized for
MAX_SAFE_TOKENS=1600 (3200 chars at its conservative 2 chars/token
estimate), so well-formed inputs are never touched; this is purely a
runtime safety net for the edge cases that slip through.
Also stop swallowing the original error in the catch. The bare
`} catch {}` here has masked recurring "input length exceeds context
length" failures for months (#369, #670, #881). Capture and warn-log
the message so future investigations see why we fell back.
Same root cause as #369 and #670 which were closed without an actual
fix to the fallback path.
Stacks on top of the multi-batch ZIM ingestion fix. After that fix,
multi-batch ZIM ingestion completes correctly — but on installs where
Ollama runs the embedding model on CPU (currently every AMD ROCm
install, since Ollama's ROCm build doesn't accelerate nomic-bert),
the now-correct sustained 100% CPU saturation across all cores can
starve other services hard enough to take the box down. Confirmed
on a Threadripper 3960X + RX 6800 NOMAD: a wikipedia-class ZIM
ingestion pegged 48 threads cleanly enough that sshd lost
banner-exchange responsiveness and the box ultimately required a
power-cycle.
NVIDIA installs aren't affected — nomic-embed-text:v1.5 runs at
100% GPU on RTX 5060 (verified via `ollama ps`).
Detect placement at runtime, pace only when needed:
1. OllamaService.isEmbeddingGpuAccelerated() — queries /api/ps and
returns true if any loaded embedding model reports size_vram > 0.
Fails closed (returns false) if /api/ps is unreachable or no embed
model is loaded yet — over-pacing is safer than crashing.
2. EmbedFileJob.handle() — between batches (hasMoreBatches: true
branch), check placement and `await setTimeout(CPU_BATCH_DELAY_MS)`
when CPU-only. CPU_BATCH_DELAY_MS = 1000 (1s) — enough to give the
OS scheduler a window for sshd/disk-collector/etc., small enough
that total ingestion time isn't meaningfully affected (each batch
is ~60-90s of work).
GPU-accelerated installs see zero behavior change.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Some Ollama installs ship nomic-embed-text:v1.5 with the embedding
model's default num_ctx=2048, which the RAG chunker (sized for ~1500
tokens of estimated content with ratio=2 chars/token) can exceed on
dense PDFs. The result is `400 the input length exceeds the context
length` from /api/embed, which then hits the OpenAI-compatible
fallback (which also errors), and surfaces as a BadRequestError.
Pass options.num_ctx=8192 (nomic-embed-text v1.5's RoPE-extrapolated
max) and truncate=true (silent truncation safety net) on every
embed call so we don't depend on the local modelfile defaults.
Reported on #756 by @NC4WD; same root cause as #369 and #670 which
were closed without an actual fix.
Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds a cancel button to in-progress Ollama model downloads and unifies
the Active Model Downloads card layout with the Active Downloads card
used for ZIMs, maps, and pmtiles (byte counts, progress bar, live speed,
status indicator).
Closes#676.
Corrupted ZIM files cause a native C++ abort (ZimFileFormatError) that
bypasses JS try/catch and kills the process. Add magic number validation
before passing files to @openzim/libzim so invalid files are skipped
gracefully. Also deduplicate Ollama download progress broadcasts — both
within a single stream (skip unchanged percentages) and across concurrent
callers (share one download promise per model).
Co-authored-by: aegisman <aegis@manicode.com>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Model downloads that fail (e.g., when Ollama is too old for a model)
were silently retrying 40 times with no UI feedback. Now errors are
broadcast via SSE and shown in the Active Model Downloads section.
Version mismatch errors use UnrecoverableError to fail immediately
instead of retrying. Stale failed jobs are cleared on retry so users
aren't permanently blocked.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Jake Turner
chriscrosstalk
Chris Sherwood
jakeaturner
Henry Estela