* split parallel tool calls for single-call-only chat templates
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Lee Jackson <130007945+Imagineer99@users.noreply.github.com>
The ROCR-vs-HIP masking comments cite "a gfx1103 iGPU under a gfx110X
prebuilt" as an example of a GPU the build has no kernels for, but the
shipping gfx110X prebuilt does build gfx1103: unsloth-prebuilt-rocm.yml
passes -DGPU_TARGETS=gfx1100;gfx1101;gfx1102;gfx1103 on both Linux and
Windows, and the b10079 manifest maps all four. install.sh also routes
gfx1103 to gfx110X-all and is_rdna() includes it.
Swap in gfx1036 under gfx103X, which is genuinely unbuilt: that bundle
maps only gfx1030/1031/1032/1034.
Comment-only, no behavior change.
Comment-only follow-up to #7454. That change carried 523 comment lines, many of
them three and four line preambles where one line says the same thing. This
collapses them and drops the ones restating what the code already says, for a
net 77 lines.
Scope is limited to comments #7454 itself introduced. The files it touched hold
about 3,761 comments in total; the rest predate it and are untouched, verified
by checking that every removed line is one that commit added.
Nothing that records why a non-obvious decision was made was dropped, only
compressed. Still stated: the normcase-before-versus-after Windows separator
trap, the innermost-indexed-model rule for nested directories, an HTTPException
being a decision rather than a failure to decide, that only an explicit False is
anonymous to huggingface_hub while None borrows the server owner's login, the
fail-closed tri-state custom-code gate, and the regressions each test was
written for.
Code is provably unchanged: comment_tools.py check reports 17/17 files
comments-only. Backend CI command 10337 passed, 0 failed. tsc -b clean.
* Add CLI GPU memory mode selection
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Preserve manual GPU layer overrides
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Lee Jackson <130007945+Imagineer99@users.noreply.github.com>
* Vulkan GPUs: real device names and selectable ordinals
Rebases the durable half of #7356 onto the inference_gpu transport #7476
landed on main. Those two PRs solve an overlapping problem and disagree on
the data model, so merging #7356 as-is would ship two parallel Vulkan
device concepts with different index semantics. This keeps main's transport
and adds what #7356 had that #7476 does not.
- _vulkan_probe.py emits a 5th column, ggml's device description, sanitized
for the tab protocol and UTF-8 safe. Reader tolerates 4- or 5-column
output so an older probe still parses.
- llama_cpp gains _run_vulkan_probe (shared parse) and
vulkan_device_inventory (names + is_igpu + real totals).
- get_vulkan_inference_gpu_info reports the real name and an explicit
is_igpu instead of "Vulkan<i>" and a total == 0 guess.
- index_kind becomes "vulkan", not "relative", and gpu_ids picks are
supported on Vulkan builds once the probe enumerated ordinals. The XPU ban
no longer applies to them: a Vulkan pick is a ggml ordinal, not a torch-xpu
index, so it works on an Intel host too.
- Frontend picker reads the Vulkan inventory as the pickable set.
Memory deliberately still comes from _get_gpu_memory, not the inventory.
That path applies _apply_igpu_host_reserve_mib and zeroes a shared total;
budgeting an APU off its raw shared total would hand out the whole machine's
RAM with no OS headroom. Identity is joined onto it by ordinal, so a probe
failure degrades to Vulkan<i> names with the memory readings intact.
Dropped from #7356 as superseded: validate_vulkan_gpu_ids (main's
resolve_requested_gpu_ids already rejects duplicates and
_resolve_gguf_gpu_ids_for_request already probes for existence), the
gguf_devices transport, and the iGPU budget fallback in 71619891e, which
main's aggregateGpuMemoryTotalGb handles better by counting a shared pool
once.
Also keeps #7356's removal of the late diffusion raise, so the graceful
gpu_ids drop stays reachable for a GGUF only classified as diffusion after
download. #7415's real guard, _reject_vulkan_diffusion_gpu_ids_before_
teardown, is untouched.
Verified on Windows + Strix Halo: backend Vulkan/GPU-selection suites at the
same 4 pre-existing failures as main, tests/studio 1671 passed with no new
failures, frontend typecheck clean. Hardware confirmation of the underlying
behavior is on #7356 from @Bebiv24 (RX 9070 XT + RX 480).
Co-authored-by: LeoBorcherding <borchborchmail@gmail.com>
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
---------
Co-authored-by: LeoBorcherding <borchborchmail@gmail.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
* Studio: say which model is missing instead of "No model loaded"
A /v1 request naming a model that is not downloaded returned the generic
"No model loaded. Call POST /inference/load first.", which cannot fix it.
Return 404 model_not_found naming the model and listing what can serve,
and make the API usage examples name a model the server actually has.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: page the API monitor, show model load/unload, pin the example quant
The monitor rendered all 50 retained entries in one scroller: page it 5 at a
time, freezing history while paged back so live traffic cannot reorder it.
Add model load/unload rows so the feed shows what the server is doing, and
stop the header rendering the loaded model as a raw host path. Advertise each
model's GGUF quant on /v1/models so the example pins repo:QUANT, and move the
auto-switch section above the monitor with shorter copy.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: optionally download a model named in an OpenAI API request
Auto-switch only ever loaded models already on disk, so naming one this
server does not have either 404s or, when something else is loaded, gets
quietly answered by the resident model.
Add openai_api_auto_download_model (off by default, gated on auto-switch).
When on, a /v1 request naming a GGUF repo that is not downloaded starts a
background fetch and returns 503 with Retry-After and a typed
model_downloading code. The resident model keeps serving in the meantime,
and the retry after the download completes is served by the new model
through the existing auto-switch path.
The download reuses the Hub manager's service layer, which already does
repo-id validation, casing, claim bookkeeping, disk preflight, resume and
cancel. The in-loader download is deliberately not used: it silently falls
back to a smaller quant under low disk, which is wrong when the caller
named an exact one.
Admission is narrow, since a request only needs an API key:
- namespace/name only, so gpt-4 and other foreign ids fall through to the
resident model exactly as before
- GGUF only, decided from the remote file list rather than the repo name
- anything declaring auto_map is refused, so trust_remote_code stays a
deliberate opt-in in the UI and can never be granted over the API
- a single download at a time, plus a free-disk reserve
- one model_info call answers existence, gating and the quant list, so a
missing repo, a gated repo and a wrong quant each get their own error
With the setting off every one of these paths is byte-identical to before.
Also:
- monitor rows for downloads, with a live percentage
- public_model_id resolves an HF cache snapshot to its repo id, so a
cache-loaded model is no longer labelled with a commit sha; this drops
the duplicate helper added for the monitor and fixes the same leak in
the inference status response
- the unedited sk-unsloth-YOUR_KEY from the copyable examples now says so
instead of "Invalid or expired API key"; every other bad key keeps the
generic message
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: add an Unload button to the API monitor
The monitor names the loaded model but offered no way to free it. Idle
auto-unload is the only existing release path, and it needs a TTL and a
wait.
The button sits next to Refresh, appears only while a model is loaded and
is disabled mid-unload. /unload matches on the internal identifier, which
this response deliberately omits because it would be a host path, so the
click reads it from /api/inference/status the same way the chat runtime
does rather than widening the monitor payload.
Also stamp the manual unload row with the quant, read before the teardown
clears it, so it reads repo:QUANT like the load row it pairs with.
* Studio: keep the API monitor Unload button visible when idle
It only rendered while a model was loaded, which hid the one manual
release path at exactly the moment someone goes looking for it. Render it
always, disabled with a "No model is loaded" tooltip when there is nothing
to free.
* Studio: never answer a named model with a different one
Asking for a model this server is not serving returned 200 from whatever
was resident. Requesting gemma-4-E2B-it-GGUF:UD-Q6_K_XL while UD-Q4_K_XL
was loaded got a confident answer from the wrong quant, with nothing in
the response saying so.
A name carrying a namespace (org/model, optionally :QUANT) is a concrete
reference, so 404 instead, with the reason:
- wrong quant -> names the quants that are actually downloaded
- not on disk -> lists what is available
- on disk but auto-switch off -> says to turn it on
Ids without a namespace (gpt-4, claude-3, default) are foreign labels
rather than references, so they still fall through to the resident model
and drop-in clients are unaffected. A bare org/model is still satisfied by
any loaded quant of that repo; only an explicit :QUANT must match.
The check runs whatever the auto-switch and auto-download toggles are,
since serving the wrong weights is wrong in every configuration. It is
skipped when nothing is loaded, where the existing no-model-loaded error
already says the right thing, and when the model is on disk with
auto-switch on, where a failed swap should still fall back.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: use a simpler prompt in the API usage examples
"What is Unsloth Studio?" rather than "Can Unsloth Studio do API calling?".
One constant feeds all nine snippet tabs.
* Studio: only refuse a model reference meant for this server
A namespace alone was treated as a concrete model reference, so a /v1
request naming anthropic/claude-3.5-sonnet, openai/gpt-4o or any other
LiteLLM or OpenRouter style vendor/model id started returning 404 instead
of being answered by the resident model. Refuse only on evidence the
caller meant this server: an explicit GGUF quant label, or a repo that is
actually on disk here. gpt-4 and vendor/model alike fall through again,
while the wrong-quant and wrong-repo cases this PR exists for still
refuse.
Also from review:
- Release the single download slot by object identity, not repo id. A
stale watcher could clear a newer download of the same repo and let a
second multi-GB fetch start alongside it.
- Catch BaseException around admission: CancelledError is not an
Exception, so a cancelled request stranded the slot for the process
lifetime.
- Honour the download service's accepted=False, which it returns without
raising for a cross-variant conflict, instead of promising a download
that was never dispatched.
- Treat a failed status probe as unknown rather than idle, so a transient
read cannot fail the monitor row and free the slot under a live worker.
- Check gated repos with auth_check. The Hub serves metadata for a gated
repo without granting its files, so the licence gate was being reported
as the unrelated custom-code refusal.
- Size the disk reserve from the download plan, which includes the mmproj
and MTP companions the worker fetches with every quant.
- Never fetch under the server's own HF token. The repo is named by
whoever holds an API key, so the ambient token let that key pull the
owner's private repos.
- Refuse an explicit quant on a backend with no quant identity, gated on
the suffix really being a quant so Ollama style :latest tags still match.
- Raise instead of falling through when the diagnosis fails: the mismatch
is already established by then, only the wording is uncertain.
- Report a failed switch as 503 model_switch_failed rather than answering
as the resident model.
- Fail an open monitor row under the same lock as the check, so a finish
landing in between cannot stamp an error onto a completed row.
- Usage examples never emit a hardcoded model id: the catalog is tri-state
and the panel asks for a model to be loaded instead of printing one the
server cannot serve. It also refreshes when the loaded model changes.
- Keep the monitor pager reachable while frozen entries expire.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: scope the auto-download 404 cache to the caller's credentials
The Hub answers 404 for a private repo the caller cannot see, so caching
that verdict per repo alone let one anonymous request mark a private repo
unservable for everyone for the whole TTL. A later caller sending a valid
X-Unsloth-HF-Token skipped the probe and fell through to the resident
model instead of downloading what it asked for. Keyed on the repo id plus
a digest of the token now, so the token itself is never held.
Two more from the same review:
- Clear the chat runtime checkpoint after unloading from the API monitor,
as the chat eject flow already does. The store went on treating the
freed checkpoint as loaded and the usage examples kept naming it.
- Point gated and not-found callers at the X-Unsloth-HF-Token header.
Automatic download deliberately ignores the server's own Hugging Face
identity, so telling the user to add a token in Studio sent them round
the same 403 forever.
* Studio: tighten the comments added by this branch
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: keep API auto-download off the server's Hugging Face identity
Passing None for the caller's token was not anonymous. spawn_worker
substitutes the backend's HF_TOKEN for a falsy one, and HfApi(token=None)
falls back to a cached login, so a repo named by an API-key holder could
still be fetched under the owner's Hub identity and land in the shared
catalog. The metadata probe and auth_check now pass an explicit False,
and dispatch threads allow_ambient_token=False so the worker stays
anonymous too. The flag defaults to True, so the UI download path keeps
the ambient fallback that private repos rely on.
Three more from the same review:
- Require an exact hf_variant match only when the suffix is really a
quant. The llama.cpp branch still compared Ollama style :latest and :8b
against the loaded quant and refused the resident model, which is the
opposite of what looks_like_quant classifies them as.
- Decode an HF cache repo id only when the models-- component is followed
by snapshots. An ordinary directory whose name merely starts with
models-- was being read as an encoded repo id.
- Return the probing response before consulting the job registry when an
adopted claim has no variant yet. A stale error on the whole-repo key
could otherwise release the slot the first request's probe still holds,
letting a second large download start beside it.
* Studio: stop treating a namespace as what decides model intent
The rule refused a reference only when it carried a namespace, which was
wrong in both directions. vendor/model is how LiteLLM and OpenRouter name
every provider, and a standalone or custom-folder GGUF is advertised
without one, so asking for a path-free local id such as model-Q4_K_M was
answered by whatever else happened to be resident. The slashless early
return is gone and the same evidence test now applies to every id: an
explicit quant, or a model that actually resolves here. gpt-4 and default
still fall through because they are not local, not because of their shape.
Also:
- Recognise bits-per-weight quant labels. _extract_quant_label emits
IQ4_XS-3.53bpw and the resolver and downloader both accept it, but
_GGUF_KNOWN_QUANT_RE has no bpw group, so looks_like_quant rejected a
reference the rest of the machinery understands.
- Upper-case the synthetic names handed to _pick_best_gguf. Its preference
tokens are upper case and matched case-sensitively, so a repo with
lower-case filenames skipped the preference and took the first entry,
which can be F16.
- Only offer a downloaded but unloaded model as a runnable example when
auto-switch is on. It is off by default, so the copied snippet hit the
no-model-loaded error, which is the failure this branch exists to fix.
The tool-passthrough cancel test stubbed asyncio.to_thread module-wide, so
it cancelled at the first thread hop rather than the generation hop it
means to test. Model resolution runs off the loop before the monitor row
opens, so that stub now passes the resolver through.
* Studio: tighten the comments added since the last pass
* Studio: match a resident model through its resolver alias
A manual load stores the model by its on-disk path while the resolver and
/v1/models advertise it as publisher/model, so _loaded_satisfies could not
recognise the alias. Reducing the resolution to a boolean then threw away
the load path that would have proved the match, and the request was
refused with 404 for a model the server was serving at that moment. Common
for LM Studio models and custom-folder aliases. The resolved path is
compared against the resident backend before anything is refused.
Also:
- Size disk admission on what is left to fetch. expected_bytes is the whole
plan, so a resumed quant or a companion already pulled in by another
quant was charged for twice and could 507 a download that fits. Cached
blobs are subtracted through existing_blob_bytes, the same accounting the
worker's own preflight does, and it falls open to the full size when no
blob hashes are available.
- Report a cancelled download as cancelled. The catch-all sent every state
other than complete or idle through fail_open, so a deliberate cancel
rendered as a download failure rather than the monitor's cancelled state.
- Keep polling the servable ids while nothing is loaded. The poll settled
as soon as auto-switch was on, so turning it back off left the examples
naming an unloaded model until something else remounted the panel.
* Studio: shorten the comments added in the last pass
* Studio: keep the FLA fast-path tests hermetic across transformers versions
_discover_fla_model_types scans the *installed* transformers for modeling
files importing `from fla.`, so `models/qwen3_5/` only exists from
transformers 5.x. The backend supports transformers>=4.51, and on a 4.x
install the Qwen3.5 gate returns False, so 14 tests in
test_training_worker_flash_attn.py silently exercised a no-op instead of the
install path and failed their call-count assertions.
Pin the discovered model_type set in those 14 tests, the same way
test_hook_does_not_install_tilelang_for_model_outside_allowlist already pins
it against newly added FLA model_types. Test-only change: the production
gate and the _discover_fla_model_types unit tests are untouched.
* Studio: keep the /v1 admission check off the model-scanning path
The admission check added here runs on every /v1 request, including with
auto-switch off, where the route used to return straight away. It called
resolve_local_gguf, whose index is cached for 5s and otherwise rebuilt by
walking ./models and every HF cache root, under a lock the next caller waits
on. On an install with a large cache that scan measured 6.1s, longer than the
TTL that is meant to amortise it, so steady traffic would keep rebuilding it.
Answer from the last built index instead and never rebuild from the request
path: a stale answer is fine here, since what is on disk barely moves and a
finished download already invalidates the index. The first request, before any
scan has completed, warms the index on a background thread and skips the check
rather than blocking on it. That also makes the lookup a dict read, so it no
longer needs handing to a thread.
Cold resolve on this box goes from 6152495us to 0.4us, and the whole hook now
costs the same for a foreign label as for the resident model.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: fix the admission hook's cold, stale and contended index paths
Five review items, four of them on the admission hook added here.
Skipping the check until the first scan lands also skipped explicit quant
mismatches, so the first request after startup could ask for :Q8_0 while
Q4_K_M was resident and be answered by it. The early return was redundant as
well: with an empty index resolved is None and here is False, so the gate below
already lets a bare name through and refuses an explicit quant, which is what
the except branch has always concluded. Dropped it and index_is_built with it.
index_is_built took _lock, which _index holds for the whole scan, so once a
warm was running every later request blocked on the event loop for exactly as
long as the scan it was there to avoid. The warm now has its own lock and reads
the timestamp unlocked, which is safe because _scan is only ever rebound.
Warming only when the index had never been built left a model fetched in the
Hub UI, or dropped into a scan folder, invisible for the life of the process,
since only the auto-download watcher calls invalidate_index. Warm on staleness
too, and unconditionally, so it refreshes within a TTL without a scan on the
request path. Rescanning is capped at a tenth of the scan's own duration: a big
install takes longer to scan than the TTL, and warming on the TTL alone would
keep a thread scanning continuously.
An Ollama-style tag names no quant, so the resolver misses it and auto-download
saw a model the resident one already answers to, then 404'd it for having no
such quant. Return early when the loaded model satisfies the reference.
Frontend: a cancelled download said "Model download failed", because the label
collapsed everything non-completed into failure.
The backend tests get an autouse fixture that stops the warm from walking the
developer's real HF caches; that scan starved the loop under the timing
sensitive streaming tests.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: make /v1/models and the admission hook agree on what is local
Three review items, all on the seam between the catalog scan and the resolver
index, which run on separate schedules.
/v1/models can advertise a local GGUF the resolver has not indexed yet. A bare
id carries no quant to refuse on, so a client asking for one it had just been
handed was answered by the resident model instead. The hook now reads the
catalog cache as evidence too, never scanning it. It takes the path rather than
a yes/no because the converse also happens: the catalog can list the resident
weights under an alias the loaded entry does not answer to, and those must stay
served.
That alias was also emitted twice by /v1/models, once as the loaded basename a
manual load records and once as publisher/model marked unloaded, because the
dedup only compared ids. Compare the path as well.
A directly loaded standalone .gguf takes its quant from the filename, but the
resolver stores such files with no quants, so the advertised <stem>:<quant>
stopped resolving as soon as anything else loaded. Advertise a quant only when
that reference resolves, and downgrade only on a definite answer so a cold
index leaves the metadata alone.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: tighten the comments this branch adds
Collapse the multi-line notes in the auto-download path, the /v1 admission
hook and their tests to one line each, keeping the reason and dropping the
restatement. No behaviour change.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: four admission and catalog fixes from review
Lowercasing paths in _resolves_to_resident made /srv/models/Foo and
/srv/models/foo the same weights on any case-sensitive filesystem, so a request
for one could be answered by the other and /v1/models could mark the wrong
entry loaded. That helper now backs residency as well as admission, so use
os.path.normcase, which folds case only where the filesystem does.
Advertising a quant whenever the resolver could not disprove it kept the bug it
was meant to fix: a standalone .gguf loaded before the first scan still got
<stem>:<quant> published, and the usage examples persist that. No proof is not
proof, so omit it and warm the index instead.
A 401 from an expired or invalid X-Unsloth-HF-Token skipped the 403 and 404
branches and surfaced as "could not reach Hugging Face, retry shortly". It now
says to replace the token, kept apart from the gated refusal since a rejected
credential is not an unaccepted licence.
An image request naming an undownloaded text-only GGUF started the whole
download and only then hit the capability guard, which never sees a remote
target, so every retry 400d and the bytes were wasted. Thread require_vision
into admission and check it against the mmproj companions the disk preflight
already asks build_gguf_variant_plans for.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: make the Hub error fixture carry a status on both hub majors
The 401 test built HfHubHTTPError directly, which works on 0.x and fails on 1.x
where response is required and keyword-only, so all four Python jobs failed
while the same test passed locally.
_hub_error already handled both constructors, but the 0.x branch left the
exception with no response at all, and hf_error_status reads the status off it
for the types that do not encode it in their name. So it could only produce a
usable error on 1.x, which is why the test bypassed it. Attach the status when
the constructed exception lacks it, and use the helper.
Cover the helper itself against stand-ins for both constructor shapes, since
whichever hub is installed only ever exercises one of them.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: invalidate on every download, resolve bare tags, keep polling
Three review items.
Only the API auto-download watcher dropped the resolver cache, so a GGUF
fetched in the Hub UI stayed absent to the cache-only request path and the
request was answered by whatever was resident. finalize_worker_exit is the one
point every download worker exits through, so invalidate there. That closes the
window without leaning on the TTL, which the scan-duration throttle can stretch
past 5s on an install where the scan itself takes longer than that.
A downloaded but unloaded GGUF asked for as org/model:latest missed the
resolver, since the suffix was always treated as an exact quant. With
auto-download on that probed the Hub and returned a 404 for a quant that was
never a quant; with it off it refused without switching. Fall back to the base
entry when the suffix is not quant-shaped, and keep exact matching for real
quants so a swap can never serve the wrong weights under the right name.
The usage examples stopped polling once a model was resident, but idle unload
frees one without touching the store, so nothing re-ran the effect and the
examples kept naming a model that could no longer be reloaded. Slow the poll to
60s instead of stopping it.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: hold the download slot while it is in use, and keep quants to llama.cpp
_loaded_satisfies refuses a quant reference against the Transformers backend by
name, but the path match did not carry that rule. A Transformers model active
from a directory that also holds GGUF exports therefore matched a request for
one of those quants and answered it with the safetensors weights. Only
llama.cpp has a quant identity, so admission now passes llama_only whenever the
reference is quant-qualified. A bare name still matches either backend, and
/v1/models residency keeps the default so a loaded Transformers model is still
reported loaded.
The 24 hour watch window was bounding ownership of the single-flight slot when
it should only have been bounding progress reporting, so a legitimately slow
download had its slot handed back while the worker was still writing, admitting
a second multi-gigabyte download beside it. Resolve the row on the clock, but
keep the slot on a slower poll until the job is actually terminal. Past the
deadline an unknown state does release it, since it means the worker cannot be
probed and holding it on that forever would wedge auto-download.
* Studio: keep what the resolver already knew when a download lands
Invalidating cleared the index to empty. The request path reads that cache
without scanning, so from a completed download until the rebuild landed it had
no evidence about any local model, not just the new one, and a bare request for
any of them was answered by whatever was resident. Wiring the hook into the
shared completion path in the last commit widened that from auto-download to
every download.
Mark the scan stale and keep the entries instead. Both _index and
warm_index_soon rebuild on a zero stamp, while the request path still sees
everything it knew a moment ago. Only a completed download invalidates, and
that only ever adds models, so nothing retained goes false.
Warm from the completion hook too, so the rebuild starts when the download
lands rather than when the next request happens to need it.
* Studio: match the quant, not just the directory, and default-select bare tags
Two quants of one repo share a directory, so the path match could not tell them
apart and an explicit :Q8_0 was answered by a resident Q4_K_M that
_loaded_satisfies had already refused by name. The llama_only fix in the last
commit only ruled out the wrong backend, not the wrong quant on the right one.
Both path matches now require the resident hf_variant to equal the requested
quant whenever the reference is quantified; a bare name still matches on the
path alone, since it claims nothing about the weights.
The local resolver already treated a tag that names no quant as meaning the
repo, but remote admission still looked for a quant literally called "latest",
so the same reference resolved locally and 404d remotely. Branch on
looks_like_quant there too. A real quant the repo does not have is still a 404
and never a substitution, which is what separates this from the loader's
low-disk fallback.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: one quant preference, and stop trusting a stale checkpoint
list_local_gguf_variants sorts by descending size, so the head of variants was
the biggest quant, often F16, while remote admission and a plain load both rank
through _pick_best_gguf. A bare id therefore meant a different quant depending
on which side answered it, and the local answer was the one that could evict a
working model and then fail or OOM starting an F16 next to a usable Q4.
/v1/models advertised that same head for pinning. Pull the ranking into one
preferred_quant helper and have both sides use it.
The usage examples returned a stored checkpoint without ever consulting
/v1/models, and the polling added last round was gated on not having one, so
for a stored checkpoint it never ran. An idle unload then left the panel
showing a snippet that could not run. Poll whenever mounted, and prefer the
checkpoint only while the catalog still backs it or switching can reload it. A
catalog that has not answered yet is not evidence against it.
The static contract pinned the old dependency array, so it now asserts the
intent it documents: a finished load re-runs the fetch, and the effect is not
gated on having no checkpoint.
* Studio: fix the Windows path compare, and advertise a label the worker knows
The case fix normalized the separator to "/" and then called os.path.normcase,
which on Windows folds case and rewrites the separator back to a backslash, so
the descendant checks compared against a "/" the path no longer had. A manually
loaded GGUF reached through an alias then read as a different model, giving a
false 404 and an alias marked unloaded. Run normcase first and normalize the
separator after it.
There are two quant-label extractors and they only agree while a recognized
quant token is present. With none, _extract_quant_label takes the last
hyphenated segment, "7b" of llama-7b, while build_gguf_variant_plans and the
worker key the whole stem: the plan lookup missed and the job exited on a
variant it had no shards for. Use the canonical extractor for the unrecognized
case only. Checked across real filenames first, the two match on every
recognized quant and part on bpw-qualified labels, which _extract_quant_label
keeps apart on purpose so byteshape's IQ4_XS at 3.53, 3.97 and 4.19 bpw stay
separate variants.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: a stored checkpoint needs catalog evidence, not just the switch setting
Preferring it whenever switching was on short-circuited the catalog check, so a
checkpoint the store still held after the model was deleted or moved kept being
named even though /v1/models had already proved it absent, and the snippets 404d
instead of falling back to a model that is actually there.
A lookup rather than a disjunction, which settles the whole matrix in one place:
no answer yet keeps the checkpoint, since that is not evidence against it; listed
and resident keeps it; listed but unloaded keeps it only when switching can
reload it; absent falls back whatever the setting says.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: normalize the quote style pre-commit would have rewritten
* Studio: cover the model that just landed, and pin the quant the catalog has
Retaining the index on invalidation protects what was already scanned and by
construction cannot contain the model that just finished downloading, so a bare
request for it in the window before the rebuild was still answered by the
resident model. Record the repo at the completion hook and treat that as
admission evidence alongside the resolver and the catalog; the next completed
scan clears the notes, since the index then covers them. Publishing a rebuilt
index before completion becomes observable would have closed it too, but that
blocks the download worker for the length of the scan.
Catalog membership proves the repo, not the saved quant, and the examples then
pinned the stored one. A quant deleted while another quant of the same repo
remained produced repo:deleted-quant, a missing-quant 404 with a runnable
alternative listed right beside it. Pin what the catalog advertises: for a
resident entry that is the resident quant, for an unloaded one it is a quant
actually on disk. The store is only consulted before /v1/models has answered.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: apply three rules everywhere they belong, not only where reported
The trust probe was the last credential handoff still passing a raw token.
huggingface_hub reads None as "use the cached login", so a caller-named repo
was read with this server's Hugging Face identity whenever the caller sent
none, which is exactly the isolation the metadata probe and the worker already
keep. It takes _hub_token now. Enumerated the rest of that path while there:
auth_check, model_info and spawn_worker were already correct.
finalize_worker_exit is shared with dataset downloads, so the resolver hook
fired for every completed dataset, scanning the model directories for nothing
and recording the dataset id as local-model evidence, which turns a bare /v1
request naming that id into a refusal instead of a foreign-id fallthrough.
Gated on repo_type.
_already_serving decided "bare" on the presence of a colon while
_loaded_satisfies and the resolver decide it on whether the suffix names a
quant, so org/model:latest against a serving Q8_0 read as a mismatch and
swapped in the preferred Q4_K_M for a request either one answers. That rule now
lives in four places, each fixed in its own round, so this time I looked for
the rest and found a fifth: describe_local_miss splits on the bare colon and
its docstring claims it splits like the resolver. It no longer did, and would
report a missing quant named "latest". Fixed here too, unreported.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: probe before refusing busy, and scan once when the index is cold
The busy refusal fired before anything established the requested label was a
model at all, so any namespaced id a drop-in client sends was told to wait out
an unrelated download for as long as it ran. Probe first and refuse only a
label the Hub actually serves as GGUF; anything else falls through to the
resident model as before. A probe failure answers "not downloadable", since
stranding ordinary traffic costs more than missing a busy refusal.
Treating an unbuilt index as "nothing here" let the first request after startup
be answered by the resident model under another model's name. That was a
deliberate trade to keep the scan off the request path, and it was the wrong
one. Cold, the scan now runs once on a thread, bounded so a pathological
install falls through rather than hanging the request. Built, the request path
still never scans, so the latency fix stands.
The watcher freed the slot the moment it saw an error, while Retry-After is
thirty times the poll interval, so the client came back to an empty slot and
restarted the same failing download instead of being told. Hold the failure on
the slot until a retry surfaces it, and let another repo take it after three
retry intervals so a client that never returns cannot keep it.
The watcher also invalidated on completion, which now lands after
finalize_worker_exit's warm and marks that fresh scan stale, pushing a
synchronous rescan onto the retry. Removed.
_loaded_satisfies lowercased paths as well as aliases, so it returned satisfied
before the case-preserving compare below could run. Both now go through one
helper: paths compare with normcase, aliases stay case-insensitive.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: an unfinished scan is not absence, and a decided refusal is not a failure
Bounding the cold scan then reading the bound as "not here" left the same hole
one branch over. A timeout now answers 503 model_indexing with a Retry-After
and leaves the warm running. A foreign label sent inside that window is asked
to retry rather than falling through, which is a real cost, but the window is
one request on an install whose scan exceeds ten seconds and it clears itself,
where answering with the wrong weights does not.
That uncovered a worse one. Every check here runs inside a broad except whose
job is "could not verify, so fall through", so an HTTPException raised in the
block was logged as a verification failure and the request was answered by the
resident model. Any refusal decided in there was being swallowed. Re-raise it
ahead of that handler.
Canonicalizing generic labels made them real variant keys, but the matcher
still decided on shape, so repo:llama-13b fell past an exact match and fetched
llama-7b. Match exactly first, whatever the shape; a quant-shaped suffix that
matches nothing is still a miss and never a swap.
Marking a catalog alias loaded while publishing the preferred on-disk quant
claimed alias:Q4 was loaded while Q8 was serving, and requiring the resident
quant to match then made pinning it a 404. Advertise the resident variant when
the entry resolves to the resident model.
* Studio: keep the asyncio.timeout fallback tests runnable on Python 3.10
Both tests deleted asyncio.timeout to force _wall_clock_timeout down its
pre-3.11 branch, but monkeypatch.delattr raises when the attribute is already
absent. On Python 3.10, the one version the fallback exists for, there is
nothing to delete, so the two tests errored with AttributeError before reaching
the code they cover. Passing raising=False makes the deletion a no-op there and
leaves the assertions running against the same branch on every version.
Every other delattr in the repo already passes raising=False for exactly this
reason. Verified with asyncio.timeout removed from the interpreter: the two
tests fail with the CI AttributeError before this change and pass after, and
the file still runs 89 passed on 3.13 where the deletion is real.
* Studio: decide GGUF residency, servability and variant keys by one rule each
Four admission and catalog fixes, each closing a gap between two places that
were answering the same question differently.
The /v1/models catalog asked _resolves_to_resident without llama_only, so a
Transformers model live from a directory that also holds GGUF exports marked a
GGUF alias loaded and gave it a GGUF quant. The usage examples then pinned
alias:quant that nothing could serve with switching off. Every entry in that
loop is advertised as GGUF, so residency there is llama.cpp residency.
The busy probe accepted any .gguf sibling while admission excludes mmproj, MTP
drafters and big-endian builds. A repo holding only companions is not
downloadable, so it was held at model_download_busy for the length of an
unrelated download instead of falling through to the resident model as it does
when no download is running. It now reuses _gguf_variants, the same filter.
split_model_ref refused any slash-bearing suffix, but an unrecognized GGUF below
a subdirectory keys on its path (build/llama-13b), which is_valid_gguf_variant
allows and the catalog advertises. Pinning such a variant could not parse, so
only the default-ranked one was reachable. A slash-bearing suffix is now a
variant exactly when a real Hub repo precedes it, which still leaves
C:/models/x.gguf a path rather than a quant.
The usage examples treated a downloaded-but-unloaded model as runnable only
under auto-switch, but a standalone UNSLOTH_MODEL_IDLE_TTL reloads exactly what
it freed on the next request. The panel hid runnable examples after an idle
unload. Tracked apart from auto-switch, because the stash restores the stored
checkpoint only and never an arbitrary catalog entry.
Also stub the index walk in the three cold-index tests that missed it: a real
multi-root scan inside the cold-wait budget made them time out into a 503 under
load rather than assert what they are there for. One of them flaked locally.
Verified each fix is load-bearing by reverting it and watching its test fail.
Backend CI command: 10195 passed, 0 failed. tsc -b and the frontend build clean.
* Studio: bound the Hub admission probes and stop guessing at nested model paths
Three review fixes plus a test-isolation one.
_resolves_to_resident matched on a path prefix, so two separately indexed models
that nest (/models/A alongside /models/A/sub/B) both satisfied it: loading B
made a request for A resident and answered it with B's weights, and the catalog
marked A loaded. A prefix match now counts only when no catalog entry sits
deeper, which is the innermost indexed model that actually owns the file. With
nothing indexed there is no nesting to tell apart, so the directory-to-weights
match this exists for is unchanged.
auth_check and hf_hub_download take no timeout of their own, and both ran while
the provisional single-flight slot was held, so an unresponsive Hub stalled the
request far past the metadata budget and reported every other model busy for the
duration. Both are bounded now. Each default errs the safe way: an unchecked
repo is not a cleared one, so the custom-code probe refuses on timeout, while a
slow gated-repo check stays inconclusive because the download's own auth is the
real gate.
The usage examples caught a failed refresh into an empty catalog and a disabled
auto-switch, which made a transient error authoritative and blanked every
example while the model was still servable. The catalog is deliberately
tri-state; a failure now keeps the last answer and retries.
Also start the backend tests from a built, empty model index. Stubbing only the
background warm still left the cold path walking real caches synchronously
inside the admission wait, so on a large install a test asserted against a 503
"still indexing" instead of its subject. _build_index is untouched, so the tests
that call it directly still exercise the real walk.
Verified each fix is load-bearing by reverting it and watching its test fail.
tsc -b clean. Backend CI command green apart from two failures reproduced only
on this box (a real model-dir scan and an orphan-process cleanup), neither
touched by this PR; staging CI is the gate for those.
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
* Studio: admission control on /v1/messages, slot pool that tracks --parallel
/v1/chat/completions was gated by the llama admission queue but /v1/messages
was not, so an Anthropic client could oversubscribe llama-server's slots and
stall the backend. Wire the same queue into all six /v1/messages dispatch
sites, and rework the queue itself into an explicit slot pool.
- Queue keyed by base_url, so both API surfaces share one pool of slots.
- Waiting is unbounded by default instead of timing out; the wait line is
sized at 16 x the serving slots so it follows --parallel.
- Neutral UNSLOTH_LLAMA_ADMISSION_* env names, legacy UNSLOTH_OPENAI_COMPAT_*
spellings still honored.
- Passthrough retries once against a respawned llama-server, which comes back
on a new ephemeral port.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Fix over-admission on capacity shrink and restore the stream cancel contract
Review of the previous commit turned up two real regressions plus smaller gaps.
- Pool sizing looked only at free slot ids, so when capacity shrank while slots
were held (an unload resets effective_parallel_slots to 1) a freed low id was
handed out even though the holdovers already met the new ceiling. Count every
held slot against capacity instead. A 1-slot backend could run 4 generations.
- The streaming wrapper closed the monitored body with aclose(), delivering
GeneratorExit where _SameTaskStreamingResponse deliberately throws
CancelledError. The monitor entry was never finalized, so it leaked as
"running" for the process lifetime and cancel_event was never set. Close
through the shared helper so cancellation reaches the handler.
- Finalize the monitor when a stream is abandoned before its body starts, and
when a queued non-streaming request is cancelled (which also leaked the
un-awaited generation coroutine).
- Floor the scaled wait line at 64, so a 1-slot backend keeps the depth it had
before scaling existed instead of dropping from 64 to 16.
- Use the canonical Anthropic type map: a full queue is 429 rate_limit_error,
which SDKs back off on; overloaded_error is 529.
- Treat non-positive max_queue/queue_per_slot as unbounded rather than "reject
everything", and reclaim the slot if a waiter's event loop is gone.
Tests: regression tests for both defects, verified to fail without the fix.
Adds env coverage for QUEUE_PER_SLOT and the legacy fallbacks, a structural
check that all six dispatch sites stay admission-wrapped, and clears the new
env var in the isolation fixtures.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Run the response pre-start cleanup when a queued stream is abandoned
From automated review of the earlier commits.
_anthropic_passthrough_stream enters its _TrackedCancel eagerly and relies on
the stream's finally to exit it, but aclose() on an async generator that never
started is a no-op, so that finally never runs. Admission made this reachable:
a client that disconnects while queued leaves the cancel id registered in
_CANCEL_REGISTRY forever.
- Give the passthrough response an unstarted_cleanup hook that exits the
tracker, via a new optional arg on _sse_streaming_response.
- Chain to that hook from the admission wrapper rather than replacing it, and
run it when the wrapper gives up before the body started.
- Defer to an in-progress MTP fallback instead of respawning underneath it;
only the first caller gets True from _maybe_recover_from_mtp_crash.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Restore Python 3.9 support, and stop the floor overriding an explicit setting
Second review round. The first item is a real break shipped by the earlier
commits, the rest are correctness and contract fixes.
- dataclass(slots = True) and int.bit_count() are both 3.10+, but the package
declares requires-python >=3.9 and CI only runs 3.12, so nothing caught it.
Importing the module raised TypeError on 3.9, taking down the whole backend,
not just admission. Drop the dataclass slots and track the popcount in a
counter. A test now asserts neither API comes back.
- The queue-depth floor applied even when an operator set QUEUE_PER_SLOT
explicitly, so asking for a shallow line silently got 64 and, with no queue
timeout, callers blocked instead of failing fast. The floor now only backs
the default multiplier.
- Never let a failing close strand a slot: closing runs in its own try so the
release always happens. A lost slot shrinks the pool permanently.
- Close the generation coroutine when reserving fails for any reason, not only
on a full queue.
- Exit the passthrough cancel tracker if the client drops while the opening SSE
lines are still being sent; those yields sit outside the teardown try.
- snapshot.free now reports what a caller could actually take, so the admission
log cannot show free slots next to queued requests after a shrink.
- Correct the class docstring: the wait line is bounded by default, not
unlimited. Document that abandoning wait() requires cancel(), and pin the
thread assumption in _deliver_lease.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Make the leak guards real, and cover the untested admission branches
Third review round, which attacked the previous round's tests by reverting each
fix. Two guards turned out to be hollow.
- The pre-start cleanup chain could be severed with the suite still green: the
existing test drove the generator finally, never the response hook. A real
pre-start disconnect leaked the passthrough cancel tracker permanently.
Replaced with a test that runs the response hook and asserts _CANCEL_REGISTRY
is empty; verified against both ways of reintroducing the leak.
- The structural check only asserted the unstarted_cleanup keyword was present,
so passing a literal None passed it while leaking. It now asserts the hook is
actually built.
- test_shares_queue_with_openai_by_base_url never touched the OpenAI helper; it
was a duplicate under a misleading name. It now reserves through the same
helper /v1/chat/completions uses, so it fails if either surface ever derives a
different key. That is the PR's central shared-queue claim.
- Cover the passthrough dispatch site, 499 on disconnect-while-queued, and the
streaming admission timeout. Four of six sites previously had only an AST
node count behind them.
- Clear admission env in the autouse fixture rather than per test: an ambient
canonical name silently beat the legacy name a test was exercising.
- Loosen the wall-clock assertion, which guarded against serialising on the
uncontended path, not against a slow runner.
- The class docstring claimed a global concurrency cap; Studio's own chat
endpoint does not reserve, so it is not one.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Keep dataclass slots on 3.10+ via a version gate
Dropping slots = True for 3.9 gave it up everywhere, including the 3.12 CI runs
and every supported interpreter but one. Gate it instead: _SLOTS is
{"slots": True} on 3.10+ and empty below, unpacked into each dataclass.
The AST scan now requires the unpack rather than merely forbidding a literal
slots keyword, so a dataclass added later cannot quietly lose slots. Added a
test that the gate matches the running interpreter, since a gate that never
applies is worse than no gate. Verified the 3.9 branch by forcing _SLOTS empty
and reloading: the full admission suite passes either way.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Fix two slot leaks, and cover the guards that had no test
Third review round, three reviewers working independently on the admission core,
the route wiring, and whether the PR regresses anything it is not about.
Leaks:
- cancel() made the same call_soon_threadsafe as _grant_waiters_locked but
without its RuntimeError guard. Routes cancel from finally blocks, so a closed
loop masked their exception and skipped the release, stranding the slot and
pinning is_idle() false so the queue was never evicted either.
- The pre-start cleanup released the slot after an await that can raise
BaseException, which is swallowed upstream. Nested it in a finally, as the
streaming and OpenAI paths already do.
An unparseable QUEUE_PER_SLOT dropped the 64 floor while falling back to the
default multiplier, quietly giving a 1-slot backend a 16-deep line. Explicit now
means it parsed.
Guards that were correct but had no test. Each was reverted, confirmed the suite
stayed green, then covered and confirmed red:
- the slot released when stream setup raises, which is the reachable one:
count_chat_tokens is a blocking call to llama-server, so a dead backend raises
after the slot is taken and before a body exists to release it
- coro.close() on a cancelled queued request, the api_monitor.fail that
distinguishes an admission timeout from a client hang-up, the MTP fallback
short-circuit, and the BaseException guard around the opening stream lines
- the queue-full test asserted a type string OpenAI's 429 also uses, so it
passed against an OpenAI envelope. It now pins the Anthropic shape.
Anthropic requests were invisible in the admission log while sharing the pool
with chat completions, so the same events are logged there with a mode. Renamed
the helper to match, since it is no longer OpenAI-only.
Corrected two comments that described behaviour the code does not have: the
slot is taken when the streaming response is built, not when the body starts
iterating, and the pool is not a cap on every generation, since /v1/completions,
Studio's chat endpoint and RAG captioning all reach llama-server directly.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Cover the admission telemetry, and drop a dead helper
Fourth review round. No bugs found in the code this time; the finding was that
most of the previous commit's telemetry had no test. Only queue-full was
asserted, so removing any of the other four log calls left the suite green.
All five are covered now, each verified by removing its call and confirming only
its own test reds. Fixing the first attempt turned up a test bug of my own: the
log line carries a queued=N field, so asserting "queued" in the message matched
every admission log ever emitted. It asserts the event name now.
Also covered two guards that were correct but unguarded: waiters whose futures
die out of band stop counting against the queue depth, and a newcomer cannot
barge past a parked waiter. The second is pinned as behaviour rather than as the
`if not self._waiters` check, because that check cannot actually change the
outcome: _take_slot_locked consults _can_admit_locked anyway, so either alone
refuses the newcomer. The test fails only if both go.
_optional_positive_int_env lost its last caller when the env parsing was
rewritten last round. Removed.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
* fetch bare hostnames as https instead of refusing them
* normalize host:port URLs and route schemeless github repos to the readme API
* only rewrite dotted host:port URLs with in-range ports
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* reject relative paths and oversized ports in url normalization
* Match web-fetch ports as ASCII digits so a unicode digit cannot raise
str.isdigit() is True for digit-class characters int() refuses (superscript
two, circled digit one), so _normalize_url_scheme reached int(port) and raised
ValueError out of _fetch_url_raw, which runs before its try block. A
web_search url of "example.com:<superscript two>" surfaced a generic tool
exception instead of the Blocked: message it returned before this branch.
Match the port against an anchored [0-9]{1,5} instead; the five-digit cap that
kept the range check from converting an unbounded integer is now in the
pattern.
* Apply the invalid-port guard to redirect targets too
_fetch_url_raw wraps the initial parsed.port in try/except ValueError, but the
redirect hop reads rp.port unguarded, so a server answering
Location: https://example.org:99999/next fell through to the broad handler as
"Failed to fetch URL: Port out of range 0-65535" rather than a deliberate
block. No request is dispatched either way; this just makes the two paths
report the same way.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Keep the redirect-port test compact
The formatter expands a signature carrying a spaced kwarg default, which put
the stub opener on eleven lines. **kw absorbs the timeout the fetch loop
passes and leaves the whole stub on four.
* Never let a malformed URL escape _fetch_url_raw as an exception
The URL is model-supplied, so every bad form should come back as one of the
documented (error, body, content_type) strings. Three gaps remained:
urlparse itself raises on an unmatched IPv6 bracket and on a netloc that
NFKC-decomposes into a delimiter (//exam(fullwidth-solidus)ple.com), and both
calls sat outside a guard. getaddrinfo raises UnicodeError, which is a
ValueError and not the OSError _validate_and_resolve_host catches, when IDNA
encoding rejects a hostname.
Over a 3158 URL corpus that injects tabs, newlines, C0 controls, delimiters and
NFKC confusables at every position, main raises 42 times and this raises none.
Also strip surrounding whitespace in _normalize_url_scheme. _web_search already
stripped, but normalization moved down to the fetch layer, so a direct
_fetch_page_text caller did not get it.
* Name the host in the status badge and tool card for bare URLs
status_for_tool and the web-search tool card both required an explicit scheme
before reading the hostname, so every URL this branch newly makes fetchable
showed the generic "Reading page..." and "Read page" instead of the host.
Under permission_mode=ask that means the approval card named no destination for
exactly the inputs the branch enables.
The backend reuses _normalize_url_scheme. The frontend cannot, since new URL()
throws on a bare host, so RE_BARE_HOST mirrors the same grammar: only a dotted
host with an optional in-range port gets the https prefix, leaving /login,
javascript: and userinfo forms to render generically as before.
Also mention bare hostnames in the url parameter description, since they are
part of the accepted interface now.
* Do not let a malformed URL in the status badge kill the tool turn
status_for_tool runs inside prepare_call, before the fetch and outside the
handler that wraps tool execution, so a ValueError from urlparse ends the whole
turn instead of letting _fetch_url_raw return its blocked message.
_normalize_url_scheme catches its own parse error and hands back the original
string, so the parse here still has to be guarded.
Reachable with https://[::1 or a host that NFKC-decomposes into a delimiter.
This predates the branch, main raises identically, but the badge is one of the
lines this branch touches and the rest of it already promises no malformed URL
escapes as an exception.
* Tighten the comments added by this branch
* Revert the web_search url description change
The premise of this branch is that models already emit bare hostnames
unprompted, which is why the fetch layer had to stop refusing them. Advertising
the bare form in the tool schema does not enable anything, it just steers models
toward it, and that is the form carrying every edge case: ambiguous with dotted
custom schemes, and unlike an explicit scheme it does not cover IPv6 literals,
IDN or trailing-dot FQDNs.
The fetch layer tolerates bare hosts. The schema should keep recommending a
full URL. This also drops the one change here with no regression test.
* Match the backend port rule in the tool card host
The card's bare-host pattern required at least one digit after the colon, but
the backend fetches an empty port (example.com: and example.com:/path go to the
default HTTPS port), so a successful fetch rendered as "Read page" with no
host.
Allowing an empty port alone would have swung it the other way: example.com:0
is refused by the backend but new URL() accepts it, so the card would have named
a host that is never fetched. That mismatch was there before this change too.
Mirror the backend rule instead, an empty port or one in 1-65535, checked
against every case in the normalizer's own matrix.
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Daniel Han <danielhanchen@gmail.com>
* Let a decode failure degrade instead of escaping a fail-closed helper
Pinning utf-8 makes a read that used to return mojibake on Windows raise
instead. 33 of those reads sit under a handler catching OSError or
json.JSONDecodeError but not UnicodeDecodeError, which subclasses
ValueError, so a corrupt file would now escape a helper written to return
a default. Adds UnicodeDecodeError to those tuples only.
* Treat an undecodable install lock as stale instead of retrying forever
* Pin utf-8 on shipping-code text I/O instead of the operator locale
113 read_text/write_text/open call sites across unsloth, studio and
unsloth_cli let locale.getencoding() decide the encoding. That is utf-8 on
the Linux and macOS runners and cp1252 on a stock Windows install, so the
same file decodes differently for a Windows user and silently produces
mojibake or raises UnicodeDecodeError.
Adds tests/test_runtime_text_encoding.py to keep it that way. It resolves
openers through each file's own imports rather than a fixed list of module
names, so an aliased tarfile.open or a local from PIL.Image import open is
not asked for an encoding it does not take.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Scan tracked files only and resolve the unbound Path calling forms
* Honour PEP 263 when scanning sources and migrate a legacy JSONL before appending
* Scope guard imports lexically and only migrate a legacy file when it round-trips
* Leave a legacy JSONL untouched and resolve path aliases in the foreign-opener check
* Tighten comments
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
* Studio: add durable Deep Research workflows
* Studio: preserve research integration after upstream updates
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: keep research worker compatible with Python 3.11
* Studio: address Deep Research lifecycle review
* Studio: preserve durable research recovery
* Studio: preserve research stream and context
* Studio: harden research sources and limits
* Studio: align research with shared chats
* Studio: guard durable research actions
* Studio: protect durable research turns
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: deepen durable research decisions
* Studio: protect research prompts and queries
* Studio: slim research stream deltas
* Studio: preserve research evidence and citations
* Studio: harden Deep Research (CI, prompt injection, query PII, config, citations)
- Fix backend CI: add research_runs_router to the synthetic routes stub in
test_desktop_auth so studio.backend.main imports under the health-check test.
- Escape prompt-delimiter tags in the decision and synthesis prompts so gathered
web/document content cannot close an <untrusted_...> wrapper and inject
instructions into the local planner/decision/synthesis model.
- Extend the public-query sanitizer to redact Luhn-valid payment cards, phone
numbers, non-global IPs, and labeled private identifiers before a query can
reach web search.
- Reject nested credential keys in inferenceRequest and ragScope, not just
top-level keys, when persisting a durable run config.
- Treat maxSources as one budget shared across web and document sources
(collection and resume paths) instead of per type, which allowed up to 2x the
configured cap.
- Preserve document citations whose filename contains a closing bracket by
tokenizing valid citations before stripping invalid ones.
- Persist Deep Research off when switching to an external model and when enabling
Web Fetch so a refresh cannot rehydrate a mutually-exclusive state.
- Add regression tests for the query, prompt, citation, and config hardening.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: make the research claims table migration atomic
The owner-scoped to global claims migration ran its RENAME, CREATE, INSERT and DROP in autocommit, so an interruption after CREATE left the new table empty, orphaned the rows in the legacy table, and never re-triggered. Wrap the rebuild in an explicit transaction so a crash rolls back cleanly and the migration re-runs on the next boot.
* Studio: block message edits and regeneration during an active research run
After a reload a durable research run is followed by the research store rather than an assistant-ui run, so thread.isRunning is false while research is still active. Message edit, refresh and the edit composer previously gated only on isRunning, which let a normal generation start alongside the running research run. Gate them on the active thread's research state as well.
* Studio: keep the plan review mounted through approval
Keying PlanReview on planRevision remounted it mid-approve when updateResearchPlan bumped the revision, resetting the local pending flag and re-enabling Start research while the approve was still in flight, which allowed a duplicate approve. Key on runId only.
* Studio: drop the redundant deep-research persistence change
setCheckpoint already persists Deep Research off for external models at the top of the function, so the added saveBool was a duplicate, and clearing Deep Research from setWebFetchToolsEnabled guarded a state that is not reachable (Deep Research is local-model only while the Web Fetch pill is external-provider only). Revert both to the pre-hardening version.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: harden Deep Research citations, query privacy, and message protection
Address review findings in the Deep Research backend:
- Escape an unbalanced ")" in citation destinations so a source URL cannot
close the markdown link early and inject a second link, keeping balanced
parentheses literal.
- Match raw-URL citations on whole tokens so a URL sharing another URL's
prefix is no longer partially rewritten.
- Redact non-global IPv6 addresses in public search queries, matching the
existing IPv4 handling.
- Detect credential key names after normalizing case and separators so nested
openaiApiKey, accessToken, and clientSecret values cannot be persisted.
- Reject client edits to server-managed research prompts and reports at the
storage layer; only the internal writers pass allow_research_update.
- Scope research searches to the first allowed domains instead of dropping
site scoping for large allow lists.
- Persist the same fetch evidence bound used during live synthesis so a
resumed run is not shortened.
- Scope run completion so it only replaces this run's message parts.
Add regression tests for the above.
* Studio: fix Deep Research SSE framing, source counts, and favicon privacy
- Normalize the whole SSE buffer so a CRLF split across transport chunks
still frames events.
- Count web and document sources together in the activity header so a
RAG-only run is not shown as zero sources.
- Cap the plan editor at the run's configured maxSteps instead of a
hard-coded 30.
- Add an allowRemoteIcons opt-out to the sources components and disable
third-party favicon requests for research sources so visited domains are
not leaked.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: address final Deep Research review findings
* Studio: fit Deep Research synthesis evidence to loaded context, add opt-in web grounding
Size the synthesis evidence budget to the loaded model context so the prompt is not
silently truncated on small contexts. When the evidence overflowed the window the report
degenerated (it echoed the evidence tail instead of writing); the budget now reserves tokens
for the prompt scaffolding and converts the remainder to chars, keeping the full cap when the
context is unknown.
Add opt-in web grounding for auto-read: read the top search results, ingest them into an
ephemeral RAG scope, hybrid-retrieve the passages most relevant to the question with the
existing knowledge-base retriever, and fold those chunks into the step evidence. The scope is
per call and deleted afterwards, so a user's knowledge base is never touched.
Off by default; enable with UNSLOTH_RESEARCH_AUTO_SCRAPE=1. Gated per run by
budgets["maxAutoScrape"], so runs created without it keep legacy snippet-only behavior, and
grounding is skipped when the loaded context is too small for the prompt.
Add tests for the adaptive evidence budget, scraped-text cleaning, the ephemeral web-RAG
retrieval and scope cleanup, and the auto-read evidence path.
* Studio: read Deep Research synthesis context from the inference orchestrator
Make the adaptive synthesis-evidence budget actually engage in the normal Studio
architecture. _loaded_context_length read core.inference.inference, the low-level backend that
lives in the model subprocess and stays unpopulated in the main web process where the research
supervisor runs, so it returned None and the budget silently fell back to the 32000 character
cap (leaving the report exposed to the truncation this was meant to fix). Read the inference
orchestrator instead, and the llama.cpp backend for GGUF, mirroring
routes.inference._monitor_context_length so the budget sizes to the context the API layer
serves. Verified on a running server: at a 12288 token load the probe now reports 12288 and the
budget adapts to 24576 characters instead of the 32000 fallback.
Also:
- Reserve context for the generated report as well as the prompt scaffolding (raise the reserve
to 4096 tokens) so evidence does not crowd out the output on a small window.
- Honor a numeric UNSLOTH_RESEARCH_AUTO_SCRAPE by passing the per-run maxAutoScrape as the page
cap to the scraper, instead of always reading the maximum.
- Guard the web-RAG connection acquisition so a get_connection failure returns the documented
empty result rather than propagating.
- Add a synthesis-context test that patches the real backend accessor (not the probe itself) so
the production wiring is exercised, plus a scrape page-cap test.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: harden Deep Research query redaction and research autosave
- research_runs: extend the opaque-token allowlist so unlabeled Hugging
Face (hf_) and GitLab (glpat-) tokens are redacted before a query can
reach web search, without over-redacting public model or version ids.
- runtime-provider: for a server-managed research message, echo the
backend-stored metadata verbatim on autosave. Merging the client
metadata re-added client-only fields the server never persisted, so the
server-side guard saw a diff and rejected every streamed or snapshot
update with 409.
* Studio: keep composer tool pills always accessible after merge
The merge left the composer line marked always-expanded (data-expanded
"true") while the inner pill row was still gated behind composerExpanded,
so the Search and Code toggles disappeared once the permission mode was
"off" with no other toggle set. Render the primary tool pills
unconditionally, matching the always-expanded layout, and drop the now
unused composerExpanded and permissionMode locals. Fixes the Chat UI
Playwright check that asserts the Search and Code pills stay visible.
* Studio: update Deep Research composer contract to always-expanded layout
The always-expanded composer no longer routes effectiveDeepResearchEnabled
through a composerExpanded expression, so the frontend contract now checks
that it gates the Deep Research composer button render instead.
* Studio: do not bind a research run to a populated assistant reply
create_run adopted any assistant message under the user turn whose
researchRunId was unset, including a prior answer reused by a retry. On
completion _update_assistant drops the untagged text and source parts, so
that answer was silently overwritten. Only bind to an empty placeholder or
this run's own message, and reject a reply that already carries content.
* Studio: harden Deep Research synthesis budget, prompt shielding, and message protection
- research_runs: split the synthesis evidence budget evenly across notes so a
small context still keeps a slice of every research step instead of dropping
the later steps after the earliest ones fill the budget.
- research_runs: shield the research question and approved plan before placing
them in the decision and synthesis prompts, so a closing delimiter in either
cannot escape its block and inject sibling sections.
- research_runs: redact bearer authorization tokens from public search queries.
- studio_db: include attachments in the research-message change check and guard
direct attachment deletion, so server-managed research prompts and responses
cannot be mutated through the attachment paths.
- chat_history: map the protected-message conflict on attachment deletion to 409.
* Studio: strip invalid document citations that contain brackets
The invalid-citation regex stopped at the first closing bracket, so a
citation whose filename contained brackets left its tail (".pdf, p. 9]") in
the report. Match a balanced bracketed span so the whole invalid citation is
removed; valid citations stay protected by the earlier tokenization pass.
* Studio: free the RAG search slot when a lookup times out or is cancelled
The bounded knowledge-base search held the sole admission slot in a detached
worker until the search returned, so a lookup that outlived its timeout (a
stalled embedding or blocked vector call) kept the slot forever and starved
every later lookup, disabling knowledge-base retrieval globally. Release the
slot from the caller when it stops waiting, exactly once, so a detached worker
finishes without re-holding it.
* Studio: remove Websites label from research composer
* Studio: fix Deep Research review findings (RAG slot bound, orphaned workers, hardening)
- Bound the shared RAG search slot to one running worker. The search that is
doing the embedding/index/GPU work now owns the admission slot until it
finishes, instead of freeing it on caller timeout while the detached worker
keeps running, which let a second search enter and stack concurrent work
behind the capacity-of-one semaphore.
- Cancel active research runs before deleting their thread, project, or all
history. Deleting cascade-drops the run row, but the worker only notices at
its next lease check, so it could keep doing model/web/RAG work for a run
that no longer exists; signalling cancel first shortens that window.
- Shield the planner prompt's conversation and question with _shield_untrusted,
matching the decision and synthesis prompts, so untrusted text cannot forge
planner delimiters.
- Do not let a research key-revocation failure replace a successful
non-streaming completion; log it like the streaming path does.
- Include created_at in the protected research-message guard so a client cannot
reorder server-managed prompt/response messages while leaving the body intact.
- Reject non-scalar ragScope values; a nested container evades the
sensitive-key scan when its inner keys are unlisted and would reach retrieval
code that expects a scalar scope id.
Adds regression tests for each.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: remove research composer globe icon
* Studio: use Hugeicons telescope in research composer
* Studio: use Telescope02 icon in research composer
* Studio: standardize Deep Research telescope icons
* Studio: move Deep Research below web and code tools
* Studio: merge grounded page excerpts with search snippets instead of replacing
When auto-scrape grounding retrieved page-body chunks, it replaced the raw
search-result text for that step. If the retrieved chunk was a distractor or
dropped the key fact, the answer-bearing search snippet was lost and grounded
runs regressed below snippet-only accuracy on factual questions (e.g. returning
Apache 2.0 instead of the Qwen License, 403 instead of 404, or a single mirror
diameter instead of the sum).
Keep the search snippets and append the grounded excerpts as supplementary
evidence via a small _merge_scraped_evidence helper. Grounding stays opt-in and
off by default, so legacy runs are unchanged. Adds regression tests.
* Studio: fix stale website access assertion in Deep Research contract test
The dialog heading was renamed to a DialogTitle, so the contract test still
asserted a <span>Websites</span> that no longer exists and failed on every
branch built on this one. Assert the current heading instead.
* Add AGPL-3.0 SPDX header to the two new test files for PR #7219
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Fix citation loss, effort clamping and nested inferenceRequest for PR #7219
Three review findings, each with a regression test that fails without the fix.
Citation dropped for a bare URL in prose parentheses. _RAW_URL swallows the
closing paren and the old trim set only stripped ".,;:!?", so the catalog
lookup missed and the validator deleted the whole citation, leaving an
unbalanced "(" in the report. New _trim_url_tail follows GFM extended autolink
path validation: one right-to-left pass that interleaves punctuation and
unmatched-")" trimming. Both rules must run in the same loop, else
"https://x/y.)" keeps a stray dot. Balanced parens inside a URL
(Wikipedia-style) still survive. Output verified against cmark-gfm on nine
cases, including "https://x/foo)bar)" which must keep ")bar".
Research runs forwarded reasoningEffort unclamped. The local chat path clamps
to the loaded model's advertised levels; the research branch did not, and the
backend only validates enum membership, so llama.cpp dropped a level the model
lacks and the whole durable run silently fell back to the template default.
Now uses the same helper and the same levels as normal chat. Note this makes
"max" on a gpt-oss low|medium|high model resolve to "low" rather than falling
through to the template default, matching normal chat exactly; the divergence
between the two paths was the bug.
Nested inferenceRequest values were persisted. Every allowed field is a scalar
and the numeric/bool/enum ones reject a container while coercing, but "model"
is stringified with str(), which never raises, so {"auth": "sk-..."} slipped
past the sensitive-key scan ("auth" is not on the list) into the durable run
config as the model id. Mirrors the ragScope guard already in this PR.
Verified: 542 passed across the research/web/sandbox/chat-history backend
suites, frontend contract 10 passed, tsc --noEmit clean.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Fix report-stalling regex, uncataloged KB evidence and bracketed titles for PR #7219
Catastrophic backtracking in _DOCUMENT_CITATION. The alternation
(?:[^\[\]]+|\[[^\[\]]*\])* backtracks exponentially on an unterminated
"[Document:" with no later bare "]", which is ordinary malformed model output
and exactly what this sanitizer exists to handle. Runtime quadrupled every two
characters; one realistic 76-char line did not finish in 90s. It runs
synchronously inside async _research (the line below it uses asyncio.to_thread),
so a single bad report pins the event loop and stalls all of Studio, not just
the run. Replaced with the language-equivalent unrolled form, verified identical
on well-formed inputs including bracketed filenames, and linear: a 20,000-char
tail now takes 0.4ms. Not using possessive quantifiers or atomic groups, which
need Python 3.11 while this package declares >=3.9.
Uncataloged knowledge base evidence reached synthesis. When maxSources is
already full, every returned chunk hits the continue, so accepted_rag_sources
stays empty, the "if accepted_rag_sources" rebuild no-ops and rag_result keeps
the raw KB text. That text has no document_source_catalog entry, so the
validator strips any citation to it and synthesis is left building claims on
private KB chunks it cannot attribute. Cleared, gated on rag_sources so a
text-only KB reply is still passed through. The resume branch built rag_evidence
from all restored sources with the same hole, so it now mirrors the live loop.
Bracketed source titles destroyed their own citation. The catalog gave the model
the raw title while the citation writer stripped brackets. Search titles
routinely carry one ("[PDF] Annual Report"), and the prompt tells the model to
copy the title verbatim, producing a label the validator cannot match. Both
sides now share _citation_title.
Verified: 756 passed across the research/web/sandbox/chat-history/rag backend
suites. Each fix has a regression test that fails without it.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Keep a durable run alive when no model is loaded for PR #7219
A durable run is claimable within the supervisor's poll interval of startup
(main.py starts it in the lifespan, and claim_next takes any 'running' run whose
lease expired), Studio has no startup model auto-load, and the browser is not
connected yet. So restarting Studio mid-run reliably lands the next model call
on the local endpoint's HTTP 400 "No model loaded". That 400 is not retryable:
_completion retries only >= 500, and _stream_completion, which serves both
planning and synthesis, has no retry at all. The run is marked failed, and the
only recovery is retry, which sets report_text NULL and deletes every
research_plan_step, research_source and research_document_source. Up to an hour
of scraping and synthesis is lost on a plain restart, on the feature whose whole
point is surviving one.
Treat only that refusal as transient: wait up to the run's own
modelTimeoutSeconds for a model to come back, then re-send. Any other 400 still
fails immediately, so no behaviour changes on the happy path. The wait polls
_check_active, so cancellation and lease loss are still honoured, and the model
probe fails open, so a probe error can only send a request, never withhold one.
Each wait is bounded by the run timeout and the number of waits per call is
capped, so a model that keeps disappearing cannot re-send forever.
Deliberately not pinning or restoring the model, which the review comment also
suggested. Auto-switch is opt-in, default off, and GGUF-only, so restoring
would silently evict the model the user just loaded from a background worker,
and comparing the configured name to the loaded id is fragile across variant
suffixes and advertised aliases, so it would break working runs.
Verified: 853 passed across the research/web/sandbox/chat-history/rag/inference
backend suites. Eight of the nine new tests fail without the fix.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Make website-policy search reach the whole allowlist and refill past blocks for PR #7219
Two review findings on the website access policy.
Domains past the site: filter cap were undiscoverable. The policy accepts up to
100 allowed domains and the prompt tells the model all of them are searchable,
but scope_search_query always scoped to allowed[:8], so a source in the ninth or
later domain could never be found, and an undiscovered URL cannot be fetched
either. The cap itself is right, search engines stop honouring long OR chains,
so the window now rotates by a hash of the query instead of being a fixed head.
Every allowed domain is reachable across a multi-step run, the same query is
always scoped the same way, and lists at or under the cap are unchanged.
A page of blocked results returned nothing. The policy filters after the search
while DDGS was asked for exactly max_results candidates, so if those happened to
be disallowed the tool reported no results even when valid ones ranked just
below, wasting a research step. Ask for a deeper pool when a policy is set and
stop at max_results allowed entries. No policy means no over-fetch, so ordinary
searches are unchanged.
Verified: 2324 passed across the research/web/sandbox/chat-history/rag/tool
backend suites. The 8 test_studio_api.py failures are pre-existing and need live
OpenAI/Anthropic credentials; they fail identically with these changes stashed.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Only overfetch search results when the website policy restricts for PR #7219
Follow-up to 8be0b3699. Every run stores normalize_website_policy(...), which
returns {"allowedDomains": [], "blockedDomains": []} and is truthy even when
nothing is restricted, so the default unrestricted path asked DDGS for four
times as many results on every step. That is pure added latency and timeout
risk, since the filter passes everything and only max_results entries are
returned either way. Test the domain lists rather than the dict.
* Budget the whole research prompt against the loaded context for PR #7219
Only the synthesis evidence was budgeted, so the budget could not prevent the
overflow it existed to prevent.
Measured at head with a realistic prompt (40-source catalog, 12-step plan): the
untrimmable scaffolding is about 7,900 chars and the conversation context adds
up to 12,000 more. On a 4096-token context, which is the GGUF auto-fit floor and
the transformers default, the synthesis request came to about 1.7x the window.
Worse, _synthesis_evidence_budget computed usable_tokens = 0 at or below the
4,096-token reserve and then returned the 1,500-char floor anyway, so it added
evidence to a prompt that already did not fit. The decision prompt had no
context awareness at all: a fixed evidence[-60000:], roughly ten times a small
window, on every step rather than once at the end.
Overflow is not cosmetic here. It either silently truncates and degenerates the
report, as the comment above these constants already warned, or fails the run,
and a failed run is only recoverable via retry, which deletes every plan step,
source and document source and nulls the report.
Both paths now share _prompt_char_budget plus _trimmable_budget: each trimmable
section is measured against what the rest of the prompt leaves, and can reach 0
instead of a floor, because a shorter report beats a destroyed run. Evidence is
budgeted before the chat history, since the evidence is the report. Unknown
context still keeps the full cap.
At 4096 tokens the synthesis prompt now fits (0.6x). Below that it is still
over, since a 40-source catalog alone exceeds the window; that needs a smaller
maxSources, and the context box does accept values down to 128.
test_synthesis_evidence_budget_tracks_loaded_context asserted the old floor at
2048 tokens, which is the bug, so it now asserts 0 and that the rest of the
prompt counts against the same budget.
Verified: 2325 passed across the research/web/sandbox/chat-history/rag/tool
suites. The test_mcp_stdio_sessions failure is pre-existing and fails
identically with these changes stashed.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Scope replayed research history to its own attempt for PR #7219
A retry deletes the previous attempt's research_plan_steps, research_sources
and research_document_sources rows but keeps its events, and the SSE route
attaches one live run snapshot to every event it emits, replayed history
included. The step.completed payload carries only position, title, action,
input and sourceCount, so that snapshot is the sole source of the excerpt and
evidence.
On any refresh after a retry, a replayed attempt-0 step was therefore matched
against attempt-1's step row by position alone, and start_position resets to 0
after the delete, so the positions line up exactly. The preserved attempt-0
activity then showed attempt-1's excerpt and evidence, or lost them entirely
when attempt 1 had not yet reached that position, under a banner that says
previous activity is preserved. The run.started resumed branch read the same
cross-attempt snapshot and spliced those activities out.
Both are gated on the event's attempt matching the snapshot's retryCount, which
is the same attempt scoping get_reasoning_text already applies server-side. The
excerpt and evidence fall back to what the activity already holds, so a mismatch
is non-destructive rather than blanking it.
Verified: frontend contract 11 passed, tsc -b exit 0, and the new test fails
without the store change.
* Retry pre-stream failures in the research stream for PR #7219
_stream_completion serves planning, every decision step and synthesis, and it
had no transport retry: a connection error or a 5xx raised before any response
byte failed the durable run, and retry then deletes every gathered source,
document source and plan step. _completion already treats the identical
failures on the identical endpoint as retryable, so the two paths disagreed.
This is partly a hole my own 689b06535 opened. After the no-model 400 the body
is read, the connection returns to the pool, and _wait_for_local_model then
sleeps for up to modelTimeoutSeconds before re-sending on the same client.
Uvicorn's keep-alive is 5s, so that pooled connection is essentially always
server-closed by then, and losing the has_expired race raises
RemoteProtocolError, killing the run the wait existed to save. Also reachable
via a read timeout waiting for headers under prompt-eval load.
Retrying is safe only because nothing has been consumed at that point, and that
is structural rather than a convention: with stream=True httpx returns on the
response headers without calling aread(), and raise_for_status() reads no body,
both verified against the installed 0.28.1. The handler is scoped to the inner
try that ends at break, and _iter_stream_lines sits outside the loop with no
path back to send, so a re-send cannot duplicate report text.
Bounded and mirrors _completion: same >= 500 predicate, same 3 attempts, same
2**attempt backoff, lease and cancellation re-checked before re-sending. The
transport counter and the model-wait counter are independent, so they cannot
multiply. The response is closed before every re-send, as manual stream mode
requires.
Note HTTPStatusError is not a TransportError in httpx, so both are caught
explicitly.
Verified: 2330 passed. Five of the new tests fail without the fix; the three
that pass either way are the invariants that must not change (fail fast on a
real 400, never retry once the report has streamed, existing model-wait path).
* Bound the planning prompt to the loaded context for PR #7219
Completes dc16598a4, which budgeted the decision and synthesis prompts but left
planning unbounded. The question reaches the planner verbatim (a pasted document
arrives here as-is) and the history is capped only at the fixed 12,000 chars,
so on a small context planning could overflow before any plan was persisted,
failing the run without doing any research at all.
Same helpers as the other two paths. The question is budgeted before the
history, since the question is the request.
A test now asserts all three prompt paths hold their own context budget, so a
fourth path cannot be added later without one.
Verified: 2331 passed; the new test fails without the change.
* Keep prompt inputs non-empty and fit the source catalog for PR #7219
Two follow-ups to the prompt budgeting, the first a regression I introduced in
dc16598a4.
The output reserve was a flat 4096 tokens, so on any context at or below that,
including the documented 4096-token GGUF floor, the whole prompt budget came out
as 0. Every trimmable section then sliced to nothing: planning_question became
the empty string, so the planner never saw the request at all, and synthesis
dropped all its evidence. Removing the old floor outright went too far; an empty
prompt is worse than the overflow it was avoiding. The reserve is now capped at
half the window, and the question and the evidence each keep a floor, since one
carries the request and the other carries the answer. A truncated completion is
recoverable, a confidently empty report is not.
The source catalog was the one section still inserted whole. It holds up to
maxSources entries with snippets persisted at up to 4000 chars each, so on a
smaller context it alone could exceed the budget while the code responded only
by zeroing the evidence and history. It is now fitted first, dropping whole
entries from the tail rather than slicing mid-entry, because a half-truncated
URL is worse than an absent one: the validator would strip it and the claim
would be left uncited.
Verified: 2333 passed. All three new tests fail without the change; the question
now keeps 1072 chars at a 2048-token context and 4144 at 4096, where both were
previously 0.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Tighten Deep Research comments for PR #7219
Post-convergence comment pass over the 40 source files in the PR diff, limited
to lines the PR itself adds so untouched upstream code in the same files is left
alone. 15 files, 110 insertions, 141 deletions.
The reduction is deliberately small. Almost every comment here records why
something non-obvious is done, a measured result, a spec rule, or the exact bug
it prevents, and those are worth more than the lines they cost, so nearly every
edit is a same-meaning compression rather than a deletion. Kept in full: the GFM
autolink citation for the URL trim, the catastrophic-backtracking note on
_DOCUMENT_CITATION, the prompt-budget notes recording that a reserve at or above
the context leaves nothing, the two measured site: filter findings, and the
remount note on the activity panel key.
Verified comment-only three ways: comment_tools.py reports 15/15 code-unchanged,
and an independent ast.dump comparison with docstrings stripped shows zero of the
12 Python files differing. 421 backend tests and the 11 frontend contract tests
pass, and the phrase the contract test asserts on is still present on one line.
* Harden Deep Research model streams
* Fit Deep Research decision prompts
* Preserve Deep Research follow-up context
* Redact composite credentials from research queries
* Scale Deep Research UI typography
* Address Deep Research refinement review
* Harden Deep Research refinement edge cases
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: danielhanchen <unslothai@gmail.com>
Co-authored-by: Daniel Han <danielhanchen@gmail.com>
* fix(studio): reject Vulkan diffusion gpu_ids before Phase 1 teardown
Classify local GGUF paths (and cached HF downloads when available) for
diffusion before _kill_process() so unsupported gpu_ids requests return
400 without tearing down the active model. Fixes#7205.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* fix(studio): always pre-download HF GGUF before Vulkan diffusion preflight
Reverts the cached-path shortcut so partial split caches still run
_download_gguf before Phase 1 teardown. Header-only classification from
resolve_local_gguf_path() does not prove the variant is complete.
* Fix inaccurate shared-constant comment and cover the local pre-teardown branch for PR #7415
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Add a regression test for the pre-teardown GGUF download for PR #7415
* Tighten the Vulkan diffusion preflight comments for PR #7415
* Trim the Vulkan diffusion preflight comments for PR #7415
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Daniel Han <danielhanchen@gmail.com>
Co-authored-by: danielhanchen <unslothshared@gmail.com>
* Add Agents settings tab for unsloth start
Adds a Settings > Agents tab documenting the `unsloth start` command:
quickstart, supported agents with click-to-copy commands, model
selection, common options, remote Studio setup, argument pass-through,
and a dry-run preview. Agent CLIs found on PATH are badged as installed.
Also removes the "New" badge from the System and Chat tabs.
* Use official brand logos for agents, invert Ollama and OpenRouter in dark mode
Claude Code and OpenAI Codex now use the Anthropic and OpenAI logos from
the provider-logos registry; agents without an official asset keep the
monogram tile. Also inverts the Ollama and OpenRouter logos in dark mode
so their monochrome marks stay visible.
* Title Agents tab "Agents (unsloth start)" and move it below Connections
The in-tab header now reads "Agents (unsloth start)" while the sidebar
label stays "Agents". Reorders the tab to sit below Connections.
* Address review: guard PATH detection, fix copy timeout, OS-aware remote snippet
- Only probe agent PATH in the desktop app on a loopback backend, so
Installed badges are not driven by a remote server's environment.
- Show the "none found" note only when detection actually ran and
returned empty, not when the call failed.
- Share one copy hook that resets its timeout on rapid clicks and clears
it on unmount.
- Render the Remote Studio snippet with PowerShell syntax on Windows.
- Note that --no-launch can still load a model when --model is set.
- Drop unused quickstart translation keys.
* Add interactive Agents command builder
* Add local subagent command guidance
* Add official coding agent icons
* Use client OS for remote commands, fix copy a11y and model wording (#7303)
- Pick the remote snippet shell from the client platform, not the server deviceType
- Single-line the model examples so they paste in POSIX, PowerShell and cmd
- Split the pass-through block into independent one-command copies
- Derive detection visibility instead of clearing state in the effect
- Announce copy success to assistive tech
- Correct the quickstart/model copy: bare start uses the loaded model
* Shell-quote the model, forward the HF token, and fix the quant placeholder
- Quote the --model value in the generated and subagent commands so a local
path with spaces or metacharacters stays a single argument (client-OS aware)
- Pass the saved Hugging Face token to listGgufVariants so gated repos resolve
- Show 'No separate quantization' instead of a stuck 'Loading quantizations...'
when a model has no variants; clear the failure once a later request succeeds
* Fix Agents command discovery and routing
* Unsloth start improvements: download progress, server reuse, and safe model switching (#7313)
* Improve unsloth start runtime lifecycle
* Remove speculative Gemma prompt override
* Polish model download progress output
* Refine unsloth start status output
* Clarify unsloth readiness banner
* Clarify model reuse and switching output
* Queue model switches behind active inference
* Tighten unsloth start model switching
* Reduce model switch bookkeeping
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Fix Studio re-exec compatibility
* Recheck sidecar reservation after inference drain
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Pass start marker through child environment
* Fix key redaction, switch-waiter ordering, and stop/messaging gaps for PR #7313
- Redact minted sk-unsloth keys from the startup-failure log tail: the early
key marker lands in the server log before the model load finishes, so a
load-phase crash printed a live key to the terminal
- Deregister a finished switch waiter before releasing the swap gate so a
swap on another event loop cannot count it as still queued and unload the
model the finished request is about to generate against
- Warn on same-repo quant switches: an explicit variant replaces the resident
weights for every attached session, but the repo ids match so no switch
warning was printed
- Note the agent exit code when it is nonzero so the server keep-alive
message does not read as a successful session
- Use taskkill /T in unsloth studio stop so llama-server children stop too
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Tighten comments in start, studio, and inference changes
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Daniel Han <danielhanchen@gmail.com>
* Unsloth start: add local subagents for Claude Code, Codex, OpenCode and Pi (#7326)
Bring the local-subagent support onto main. The original change (#7316) merged
into the stacked pr/daniel-unsloth-start-audit branch rather than main, and #7313
reached main via squash, so these files never landed on main.
Adds --as-subagent for claude, codex, opencode and pi: the parent agent keeps its
own cloud model while a locally served GGUF is registered as a delegated subagent,
using ephemeral per-session config that never touches the user's real agent config.
* Fix Agents builder defaults and flag validation
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Fix Agents variant and provider fallbacks
* Fix local model and Pi subagent edge cases
* Agents tab: flag the Codex row when the loaded model is not GGUF
* Agents tab: target the active Studio server, wrap narrow rows, index the tab's search terms
* Agents tab: build copied commands from the browser-reachable Studio and show the key placeholder
* Preserve cache load ids and path variants in built commands for PR #7312
A GGUF outside the active Hugging Face cache only loads by its snapshot
path, so keep that load_id for --model while still listing the row by repo
id. Path based models carry their quant in --gguf-variant rather than a
":variant" suffix, and the active selection now keeps the variant inference
status reports for them.
* Agents tab: index the intro for agent-name searches and keep long commands inside the panel
* List GGUF variants from the cache the command loads from for PR #7312
A snapshot outside the active Hugging Face cache was offering the remote
variant list, so a quant absent from that snapshot could be selected and
the generated command would fail to load it.
* Agents tab: omit --api-key so the CLI can replay a saved key for the base
* Agents tab: label the indexed heading rows and fall back to the active desktop API base
* Agents tab: name every supported agent in the indexed intro for PR #7303
* Send the cached GGUF load path and fix the agents tab search targets for PR #7312
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Tighten the agents tab comments for PR #7303
* Build the agents tab example commands from the active Studio base for PR #7303
* Keep the resident model on its active cache load for PR #7312
* Tighten the agents tab and cached GGUF comments for PR #7312
* Take the agent command shell from the Studio host for PR #7303
* Stop emitting snapshot paths as --model and keep unsloth start searchable for PR #7312
* Pick the command shell from where the CLI runs for PR #7303
* Match a path load by its advertised id and follow the resident model for PR #7312
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Keep an explicit quantization and retire superseded native-grant labels for PR #7312
* Scope the remembered quant, stop following unloaded models and keep local GGUF paths for PR #7312
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Stop shadowing the path classifier, match snapshot ordering and sequence status polls for PR #7312
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Release stale native-grant picks, keep local GGUF identities and index snapshot aliases for PR #7312
* Index inactive-cache snapshots, widen local GGUF detection and clear retired quants for PR #7312
* Classify cached repos by snapshot, merge repo ids case-insensitively and keep loose GGUFs variantless for PR #7312
* Fix snapshot alias, partial split and mmproj-only handling for PR #7312
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Trust scanned model_format and drop incomplete snapshot ids for PR #7312
* Exclude mmproj and partial downloads, keep path case and drop duplicate scan for PR #7312
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Restrict revision aliases and require complete snapshot variants for PR #7312
* Index revisions individually and hide partial variants for PR #7312
---------
Co-authored-by: shimmyshimmer <107991372+shimmyshimmer@users.noreply.github.com>
Co-authored-by: Daniel Han <danielhanchen@gmail.com>
Co-authored-by: oobabooga <oobabooga4@gmail.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
* Default tool-call permission to Approve for me, prompting only on high-risk actions
Make "auto" ("Approve for me") the product default permission mode for local
tool calls, and narrow what it prompts on so ordinary development commands run
without interruption.
Before, an omitted permission_mode behaved as "ask" (or ran ungated on a
non-streaming request), and "auto" paused on any call that was not read-only
(pip install, mkdir, cp, python train.py, git commit, any redirect). Now:
- Unset permission_mode normalizes to "auto" at the API boundary and in both
tool loops; the Field defaults are "auto" too. An unrecognized value still
falls back to the stricter "ask".
- "auto" pauses only on genuinely high-risk calls via a new
is_high_risk_tool_call classifier: credential/secret path access, privilege
escalation (sudo/su/doas/pkexec), destructive or persistence commands
(rm/dd/mkfs/crontab/systemctl/recursive chmod, ...), and network exec/exfil
(curl piped to a shell, ssh/scp/nc, curl uploads). Everything else runs.
Python prompts on shell escapes, network egress, sensitive reads, and
dynamically built code; ordinary in-workdir writes run.
- Frontend sends permission_mode for every local chat and omits
confirm_tool_calls for "auto" so the safe-only no-stream exception still
applies; the picker and store describe the new behavior.
The hard-block command set, code-safety static analysis, resource limits,
secret-env stripping, and the per-session sandbox workdir remain in force under
every mode, and "ask" is still available for users who want to confirm every
call.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Keep non-streaming tool requests working under the auto default
The default-permission change made an omitted permission_mode normalize to
auto at the request boundary, so a non-streaming enable_tools request hit the
confirm-without-stream guard and returned 400 instead of running (regression
against the #6570 non-streaming tool-call contract used by non-interactive
clients and health checks).
Keep permission_mode unset at the request boundary (the confirm gate can only
prompt while streaming, so an unset non-streaming request stays lenient and
runs), while the tool loops continue to normalize an unset mode to auto for the
per-call gate. Net: streaming requests default to auto and pause high-risk
calls; non-streaming requests keep the prior run-without-gate behavior.
* Harden the auto high-risk classifier against review-flagged bypasses
Address Codex/Gemini review of the default-permission change by gating the
destructive/exec cases that were reaching auto mode without a prompt:
- Terminal: a non-shell interpreter running inline code (python -c, node -e,
perl -E, php -r), destructive git subcommands (git clean, git reset --hard,
git push --force), and a command synthesized by a command-position
substitution ($(printf rm) -rf build) now prompt. Ordinary python <script>,
git commit/push, and argument-position substitutions (echo $(date)) run.
- Python tool: exec/eval/compile/__import__ invoked by keyword (compile(source=
...), import_module(name=...)) is now caught alongside the positional form.
- MCP: an execution tool (run_command, execute_script, invoke_shell) is gated
like a terminal call, since it runs arbitrary commands on the MCP server
outside the terminal sandbox; ordinary create/list/read tools still run.
The curl/wget exfil and shell eval cases the review raised are already refused
by the sandbox hard-block set, so no gate change was needed there; the PR
description now notes the classifier layers on top of that hard-block.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Recurse shell -c payloads and literal exec source in the high-risk gate
Second review round on the auto high-risk classifier:
- A high-risk command wrapped in a shell -c payload (bash -c 'git clean -fd',
sh -c 'truncate -s 0 x') is now screened by recursing into the payload,
bounded by depth. The sandbox hard-block only recurses for its own smaller
command set, so git/truncate wrapped this way previously ran unprompted.
- A literal exec/eval/compile source is screened for what it runs rather than
assumed harmless: exec('import urllib...urlopen(...)') now prompts, while
exec('x = 1') and a literal __import__('os') name still run.
- git global options that take a value (git -C repo clean, git -c k=v clean)
consume their value before the subcommand is read, so the real subcommand
is judged.
- The network exfil check also runs over the assignment-expanded command, so a
curl/wget name assembled from variables (c=cu d=rl; $c$d -F ...) is seen.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Cover attached inline flags, env -S/-C, camelCase MCP, folded python paths
Third review round on the auto high-risk classifier:
- Interpreter inline code in the attached short form (python -c'...',
node -e'...') is now matched by the -c/-e/-E/-r prefix, not only the exact
flag token.
- env -S / --split-string runs its string as a command (screened recursively)
and env -C / --chdir changes the working directory (asks), so a destructive
command behind env is no longer treated as a plain wrapper.
- camelCase MCP tool names are split on the case boundary (runCommand ->
run_Command) before the execution / sensitive-noun regexes, so camelCase
execution tools are gated like snake_case ones.
- A sensitive path folded across string-literal variables, os.path.join,
sep.join([...]), or an f-string (p='/etc'; open(p+'/shadow')) is now folded
and re-checked; an unresolved fragment folds to a sentinel so a partial fold
never false-positives.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Gate substitution-built shell payloads and keep explicit confirm opt-in
Two auto-mode gaps from review:
- A command substitution stashed in a variable and then executed dynamically
(x=`printf 'git clean -fd'`; bash -c "$x", or ...; $x, or eval "$x") never
appears as literal command text, so the token scan could not see the real
command and git clean ran without a prompt. Fail closed when a command
substitution coincides with a variable executed as a command. Ordinary
substitutions captured into a value/argument (d=$(date); mkdir build_$d) still
run.
- An explicit confirm_tool_calls=True with no permission_mode is the
pre-permission-mode opt-in to confirm every call. It now resolves to "ask" at
the request layer instead of the "auto" product default, so those callers keep
per-call gating rather than only prompting on high-risk calls. A bare unset
request (confirm flag not set) still defaults to auto.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Cover CLI-forced confirm, Windows delete built-ins, and pathlib reads
Three more auto-mode gaps from review:
- An explicit confirm_tool_calls=True with no permission_mode is now resolved to
"ask" regardless of the request-level tool flags, so a process-wide
--enable-tools policy that forces the loop when the request sets neither
enable_tools nor mcp_enabled still gates every call. Setting only the mode is
inert unless the loop runs, so a passthrough request is unaffected;
external-provider requests are still left untouched.
- The Windows cmd.exe delete built-ins del, erase, and rd are added to the
high-risk terminal set. The terminal executor runs cmd /c on Windows and these
are not in the hard-block set, so del /q file.csv would otherwise run in the
workdir without a prompt.
- A sensitive path assembled with pathlib (Path('/etc') / 'passwd', joinpath, or
a Path bound to a variable then joined) is now gated. The python high-risk
folder reuses the shared _folded_path builder plus _folded_is_sensitive, which
already handle the / operator, path constructors, os.path.join, str.join,
f-strings, and %/.format. Relative in-workdir and unknown-base paths still run.
* Gate combined -c, versioned interpreters, busybox, and sensitive chdir
Four more auto-mode classifier gaps from review, plus a sandbox backstop:
- Combined shell flag clusters (bash -lc, bash -xc) and the attached form
(bash -c'...') now have their -c payload screened recursively; the same
cluster handling closes python -Bc inline code. Previously only an exact -c
matched, so bash -lc 'git clean -fd' ran without a prompt.
- Versioned interpreter binaries (python3.11, python2.7, pypy3.10) are recognized
as inline-code interpreters, so python3.11 -c '...' is gated like python3 -c.
- busybox / toybox are treated as command wrappers, so the applet
(busybox rm -rf) is judged instead of the multicall binary, which was slipping
through as an unknown-but-safe command.
- A chdir into a sensitive directory (cd /proc/$PPID; cat environ, cd /etc) is
gated: the read happens after the directory change so no single token spells
out the sensitive path. Ordinary in-workdir chdirs still run.
- Backstop for the /proc/<parent>/environ read: the sandbox now hardens the
Unsloth process against same-UID /proc environ reads in normal sandboxed mode
too, not only in bypass mode, so a classifier miss cannot recover the parent
environment. Best-effort in the sandbox (the child env is already scrubbed), so
a host where prctl is unavailable still runs.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Harden parent proc-env on the sandboxed python path too
The previous commit hardened the Unsloth process against same-UID
/proc/<parent>/environ reads on the sandboxed bash path; apply the same
best-effort hardening on the sandboxed python exec path so both tools are
symmetric. Update test_bypass_exec_hardens_parent_proc_env, which asserted the
sandboxed path never hardened, to expect the backstop on both paths.
* Tighten the curl/wget exfil check for attached and wget upload flags
The network exec/exfil classifier missed a curl upload flag when it was attached
to its value (curl -Ffile=@dump.sql, curl -d@f) because the token was split on =
first, and it did not cover wget's upload flags (--post-data, --post-file,
--body-data, --body-file). curl short upload flags are now matched prefix-wise and
wget's upload flags are checked separately, which also removes a false positive
where a benign wget short option (wget -T timeout, wget -F force-html) was read as
an upload. curl and wget remain hard-blocked by the sandbox regardless; this only
tightens when auto mode pauses for approval.
* Tighten the high-risk auto-mode classifier: wrapper, interpreter, git, python-fs, MCP, and persistence-write gaps
Close reachable gaps where a genuinely dangerous tool call was auto-approved
without a prompt in Approve-for-me mode:
- Process-launch wrappers: setsid/exec/builtin forward the command position, so
screen their child (setsid git clean, exec python -c) instead of the wrapper.
- Inline-code interpreters: node/bun -p/--print evaluate code like -e; pwsh
-Command/-EncodedCommand run inline code (not hard-blocked off Windows).
- Windows cmd.exe /c|/k recurses into the nested command (cmd /c del x).
- git restore (default --worktree) and git checkout -- . / git checkout .
discard tracked edits irrecoverably, same class as the already-gated git clean.
- Python destructive filesystem calls (os.remove, shutil.rmtree, Path.unlink,
os.rmdir/removedirs, incl. bare imports) pair with the terminal rm gate.
- MCP: a read-named tool carrying a destructive payload (DELETE/DROP SQL,
GraphQL mutation, mutating HTTP method) still prompts; honestly-named
create/update/delete MCP calls keep running.
- System persistence writes: a write into /etc/profile.d, /etc/cron*,
/etc/systemd, /etc/ld.so.preload, /etc/rc.local, /etc/init.d installs a
boot/login/preload hook. The sandbox keeps host-fs access, so gate these;
ordinary /etc reads (hostname, resolv.conf) and in-workdir writes still run.
Adds table-driven regression rows for every new prompt case and its
guard-against-over-prompt counterpart.
* Extend the high-risk auto-mode gate: non-curl network clients, destructive MCP verbs, array-fed shell payloads
Round-two Codex hardening on the auto (Approve-for-me) classifier:
- Network exfil beyond curl/wget: gate nc/ncat/netcat/telnet/socat/ssh/scp/sftp
at command position and openssl s_client/s_server. The sandbox has no network
namespace, so tar czf - . | openssl s_client -connect host:443 was streaming
the workdir without a prompt. Local openssl (dgst/enc) and a filename that
merely contains a client name still run.
- Destructive MCP tools: an honestly-named delete_file/delete_repo/drop_table/
purge_index/revoke_token runs outside the terminal sandbox and loses data, so
gate the destructive verb on the name. Non-destructive create/update/list/get
still run; a substring like undelete does not match on the segment boundary.
- Dynamically constructed shell payloads: x=(git clean -fd); bash -c "${x[*]}"
carries no command substitution and is not resolved by assignment expansion,
so it slipped the var-executed check. Fail closed when an array expansion is
run as a command; a benign array print (echo "${a[@]}") is untouched.
Adds regression rows for every new prompt case and its benign counterpart.
* Gate user-level persistence writes in auto mode
Extend the persistence-write gate from the /etc set to user-level startup and
autostart locations: a write into ~/.bashrc, ~/.zshrc, ~/.profile and the other
shell rc/profile files, ~/.config/autostart, ~/.config/systemd/user, or
~/.config/environment.d runs on the next login/session, the same boot-hook risk
but needing no root (Studio commonly runs unprivileged, so this is the more
reachable vector). The sandbox does not confine absolute paths, so an append to
~/.bashrc reaches the real file. A non-persistence ~/.config dir and ordinary
reads still run. Adds regression rows.
* Close three more auto-mode gate gaps: curl destructive methods, the dot source synonym, aliased os.remove
- curl -X DELETE / --request DELETE|PUT|PATCH (separated, attached, and
--request= forms) mutates or deletes a remote resource, so gate it; a plain
download and GET still run.
- The hard-block set blocked source but not its POSIX synonym '.', so
. ./script.sh ran the file's contents past the classifier. Block '.' at
command position too; a path argument (find . -type f, cd .) is unaffected.
- os.remove reached through an aliased module (import os as fs; fs.remove(...))
was missed because only the literal receiver 'os' was recognized; resolve
import os as ... aliases, matching the existing safety analyzer.
Adds regression rows for each case and its benign counterpart.
* Close three more obfuscation bypasses of the auto-mode gate and hard block
- ANSI-C quoting hid the command name: a $'rm' -rf x form tokenized as $rm, so
both the high-risk scan and _find_blocked_commands missed it while Bash ran
rm. Decode ANSI-C ($'...') before classifying, in both the terminal
classifier and the blocklist; an ANSI-C string in argument position stays
benign.
- Process substitution executed as a script (an interpreter consuming a <(...)
whose generated content is unscreenable) ran without a prompt; the prior <(
check was unreachable without curl/wget. Gate a process substitution consumed
by an interpreter; a non-interpreter consumer (diff over two <(sort ...))
still runs.
- os.remove bound to a name (f = os.remove; f(x)) or reached via getattr(os,
'remove') bypassed the direct-attribute scan. Track assignment aliases and
getattr with a literal attribute name; a bound list.remove still runs.
Adds regression rows for each case and its benign counterpart.
* Gate container runtimes, MCP privilege grants, arg-embedded exec, and network listeners
- Container/VM runtimes (docker, podman, nerdctl, ctr, crictl, lxc, machinectl,
kubectl) act through a daemon with host privileges, so a bind mount writes the
real filesystem and escapes the child process workdir and rlimits entirely.
Gated wholesale because the escape lives in the arguments.
- MCP privilege grants: an unambiguous privilege verb (grant/authorize/elevate/
escalate/impersonate) prompts on its own; a softer verb (assign/add/set/
attach/bind/put/update/create) prompts only next to a privilege noun (role,
permission, policy, acl, scope, membership), so assign_issue and add_label
keep running while grant_role and add_permission ask.
- A flag whose value is a command the tool then executes (GNU tar
--checkpoint-action=exec=CMD, --rsh, --rsync-path) hid a payload inside an
argument, past both the classifier and the blocklist. Ordinary archiving runs.
- An interpreter serving on the network (python -m http.server, uvicorn,
gunicorn, waitress) exposes the session workdir since the sandbox keeps no
network namespace. A non-server module (python -m pytest, -m pip) still runs.
Adds regression rows for each case and its benign counterpart.
* Close the parallel-review gaps: over-prompting regressions and asymmetric high-risk omissions
Over-prompting fixes (auto mode was pausing on ordinary work):
- The network-listener check matched a server name ANYWHERE in the command, so
`pip install uvicorn`, `grep uvicorn reqs.txt` and even `echo uvicorn`
prompted. Scope it to the two forms that actually listen: a module after
`-m`, or a server binary at command position.
- Inline-code flags were one shared set, so `python -E` (ignore env) and
`python -Werror` read as eval. Resolve them per interpreter: python -c,
node/deno/bun -e/--eval, ruby -e, perl -e/-E, php -r.
- The curl upload scan read option letters from unrelated commands in the same
line (`ls -T && echo curl`). Scope the scan to the segment whose command is
actually curl/wget.
Under-prompting fixes (destructive actions the narrowed gate stopped catching,
each the twin of something already gated):
- git: switch -f/--force/--discard-changes, stash clear/drop, branch -D/-M,
rm, push --delete/--mirror/--prune and the +src / :dst refspec forms.
- Platform twins: unlink, ftp, tftp, format, diskpart, diskutil, schtasks,
reg, sc, launchctl.
- Python: posix/nt module twins (including bare imports), os.truncate,
os.ftruncate, os.kill, os.killpg, and a file handle's truncate. Gated via the
handle name so pandas DataFrame.truncate() keeps running.
- MCP: clear/reset/empty/flush/prune/expire destructive verbs, promote.
- deno/bun expose inline eval as a subcommand, not a flag.
- A bare redirect (`> file`, `: > file`) truncates; a redirect after a real
command is an ordinary write and still runs.
- A forwarded git command keeps its git context (`find -exec git clean`,
`xargs git clean`), and an unquoted `cmd /c` payload spans the remainder.
Adds regression rows for every case and its benign counterpart.
* Gate shell control flow, bash -c clusters, wrapper option values, and annotated aliases
- `if`/`while`/`until` are followed by a condition the shell runs, so a command
there is at command position. `if rm -rf build; then :; fi` slipped both the
classifier and the blocklist (they share the keyword set, so both are fixed).
- A short letter run after `-c` (bash -ce, bash -cl) is more bash options, not
an attached payload: bash still reads the command string from the next token,
so the real payload was never screened.
- A wrapper option taking a separate value (env -u NAME, stdbuf -o L, timeout
--signal TERM, nice -n 5) had its value read as the wrapped command, so
`env -u FOO rm -rf build` resolved the command `FOO` and never judged `rm`.
env -C/--chdir is deliberately excluded: it is gated as a chdir already.
- An annotated binding (f: object = os.remove) is the same alias as a plain
assignment; only ast.Assign was collected.
Adds regression rows for each case and its benign counterpart.
* Fix two gate regressions and close seven more bypasses
Regressions from the previous round, both caught by review:
- Shell keywords were treated as separators anywhere, so `grep if rm README.md`
resolved `rm` as a command and was blocked. A keyword only separates where a
command may start, so gate the check on command position (all three scanners).
- The wrapper option-value table was shared across wrappers, but `env -i` is
valueless while `stdbuf -i` takes a value. `env -i git clean -fd` therefore
consumed `git` and never judged the subcommand. The table is per wrapper now.
New gaps closed:
- `git -c alias.NAME=PAYLOAD` defines code git then runs. Screen the payload: a
`!` alias as a shell command, a plain one as `git <payload>`.
- A script fed to a shell over a pipe (printf '...' | bash) or a herestring
(bash <<< '...') never appears at command position. Ordinary pipes still run.
- `chroot`, `nsenter` and `unshare` cross a privilege or namespace boundary and
then exec a nested command the wrapper hides.
- A bare runtime name (mcp__srv__python, __node, __code) is an MCP execution
tool even without a verb.
- `m = __import__("os")` binds the module like `import os as m`, and
`getattr(__import__("os"), "remove")` reaches it inline.
Declined: gating every command substitution used as a path argument (would
prompt on `echo $(date)` / `make $(FILES)`), and bare `git checkout <path>`
(statically indistinguishable from the very common `git checkout <branch>`).
Adds regression rows for each case and its benign counterpart.
* Pin the auto-mode contract with benign and dangerous corpora
The value of defaulting to "Approve for me" rests on two properties that pull
in opposite directions: ordinary development work must run silently, and
genuinely dangerous work must still prompt. Every denylist change risks
trading one for the other, and a regression in the benign direction is easy to
miss because nothing fails, the mode just starts nagging.
Add two corpora that pin both directions: 62 ordinary commands, python
snippets and MCP calls that must NOT prompt (package installs, builds, tests,
git workflow, reads, ordinary pipes and redirects), and 55 dangerous ones that
must (credential reads, destructive and persistence changes, privilege
escalation, network exec and exfil, container escapes, obfuscated forms).
125 cases, currently 100 percent in both directions.
* Scope four over-prompting checks and close six more gate gaps
Over-prompting fixes (auto mode was pausing on ordinary work):
- find/fd were marked forwarding from the command itself, so every later
positional looked executable and a search whose pattern happened to equal a
gated command name prompted. They only forward after an explicit
-exec/-execdir/-ok flag now.
- The openssl s_client check was not command-position aware, so grepping for
the string in a README prompted.
- An exec-valued flag (--checkpoint-action, --rsh, --rsync-path) counted no
matter which command owned it, so printf '%s' --rsh prompted. It now
requires the owning utility (tar/rsync/scp/sftp) in the same command.
- A listener behind a wrapper or given by absolute path was missed instead
(env uvicorn, timeout 60 gunicorn, /usr/local/bin/uvicorn); resolving the
binary at command position covers all three.
New gaps closed:
- git checkout <commit> <path> overwrites the file from that commit, as does
--pathspec-from-file. A single positional stays ambiguous with a branch name
and is still left alone.
- git config alias.NAME BODY stores code git runs on the next invocation, so
the body is screened like the -c form.
- systemd-run launches a nested command as a transient unit.
- Version-suffixed perl/ruby/php/node still run inline code with -e/-r.
- A file handle bound by `with open(...) as f` is tracked for truncate, not
just an assigned one.
- Exceeding the shell nesting depth now fails closed, matching the docstring,
instead of letting an unscreened payload through.
Declined: rebinding a command name through the bash hash builtin. Like the
alias/read/awk/coproc family already declined, it is deliberate
self-obfuscation of an already-gated command rather than anything a model
emits, and the always-on backstops cover it.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Scope two more over-prompting checks and close four gate gaps
Over-prompting fixes (auto mode was pausing on ordinary work):
- A recursive flag was looked for across the whole command line, so
`grep -R pattern . && chmod +x build.sh` made the chmod look recursive and
prompted. The flag is now scoped to the segment that owns the command.
- The startup-file names were matched anywhere in the line, so `cat
notes.profile.bak` and `my.zshrc.template` prompted. They now have to sit on
a path boundary, while the real dotfiles still prompt.
New gaps closed:
- A pending wrapper option value leaked past a command separator, so the
command after it was never screened (`env -u` followed by a recursive delete
was missed). The pending state is cleared at every separator now.
- git plumbing and maintenance that loses data: update-ref, reflog, gc, prune
and history rewriting drop refs and unreachable objects, the same loss the
porcelain forms already gate.
- A module pulled in dynamically is screened against the same set as a static
import, so a dynamically imported socket or shutil is treated alike.
- MCP names that move money or ship artefacts (transfer, payout, charge,
refund, wire, publish, deploy) are irreversible for the operator even though
they are not destructive in the filesystem sense.
Declined two items:
- Gating arbitrary interpreters that can shell out (awk BEGIN blocks and
friends). Consistent with the alias/read/coproc/trap family already declined
here: it inverts the denylist into an allowlist and costs real ergonomics for
payloads a model does not emit in normal work.
- Prompting on every write outside the session workdir. Ordinary builds and
scripts write to the standard temp directories constantly, so this would
prompt on routine work. Persistence and credential paths are already gated
specifically.
* Resolve command-position globs and keep quoted data out of shell syntax
- A glob at command position is expanded by bash after this scan runs, so
`/bin/r[m] -rf x` was screened under a name that never executes. The
always-on blocklist now resolves such a pattern against the blocked names,
and the classifier asks when a command word cannot be resolved at all. The
test builtins are excluded, and a pattern carrying no literal character
resolves to nothing in particular.
- A dollar-quoted word expands to a single word, so a newline inside it is
data rather than a separator. Decoding it before tokenization made
`printf '%s'` with multiline data read as two commands and the call was
refused outright. The decoded text can no longer introduce shell syntax,
while an escape-obfuscated command name still resolves.
- An attribute name assembled from literals is folded before it is screened,
so a deletion spelled as a concatenation is treated like the plain form. A
name on a filesystem module that cannot be folded at all fails closed, since
there is nothing left to screen.
- An MCP name with no separators never reached the segment boundaries, so a
server-side execution tool was classified as ordinary even though the
previous classifier failed closed on it. The verb and object compounds are
matched directly now, while a name that merely starts with those letters is
left alone.
Also narrowing a verb pair added in the previous commit: subscribing to a
topic is not a billing subscription, and pub/sub tools should not prompt.
* Screen attached exec values, wrapped openssl, php code flags, worktree removal and sysctl writes
- fd accepts the command attached to the flag (--exec=<cmd>, --exec-batch=),
and that spelling was stripped and discarded without ever being screened.
The value is treated as command position now, in the classifier and in the
always-on blocklist. Only the long spellings are read this way: a short -x
belongs to too many other utilities for its neighbour to be a command.
- The openssl socket check was anchored at command position, so a wrapper in
front of it (env, timeout) hid the very thing it was meant to catch. The
subcommand is checked on the resolved command segment now, so the wrapped
and absolute forms are covered. Local openssl (dgst, enc) still runs.
- php runs code from -B, -R and -E as well as -r, which are begin, per-line
and end blocks. Only -r was listed, so the other three ran inline programs
unscreened.
- git worktree remove --force deletes a linked worktree even when it holds
uncommitted work or is locked, but only the first-level subcommand was read
so the nested action was invisible. An unforced remove refuses on a dirty
worktree and stays out, matching how the checkout and switch discard flags
are handled.
- sysctl -w, --system and -p change kernel parameters, and the assignment form
writes without needing a flag. A read-only query stays automatic.
* Fail closed on unscreenable MCP names, alias bodies and stored lookups
- An MCP name whose verb this classifier does not recognise now asks. MCP
tools run on an external server, outside the terminal sandbox and every
backstop under it, and their names are an open vocabulary rather than the
finite set of POSIX utilities, so the denylists could never be complete: a
name built from an unfamiliar verb sailed through as ordinary. A generous
read and write vocabulary keeps the everyday tools running, and the reverse
or repeat of a recognised verb (undelete, reopen, resend) counts as
recognised too. Measured against thirty tool names taken from the common
servers, one still prompts, and that one is the pre-existing execution rule
rather than this one.
- A shell alias body is a command bash runs when the alias is invoked, so it
is screened as a command in its own right, in the classifier and in the
always-on blocklist. This is the same shape as a git alias body, which was
already handled; leaving the shell form out was inconsistent.
- git --config-env=<key>=<envvar> takes its value from the environment, so an
alias key stores code that never appears in the command text at all. The
attached form was skipped entirely because the parser required no equals
sign. An alias key gates it now; ordinary keys are untouched.
- A destructive lookup stored before it is called (a name bound to
getattr(os, "remove")) matched neither the direct call shape nor the alias
collection, so it ran. The binding is tracked now.
- A credential basename only names a file when it appears in a string, but the
whole Python source was being scanned, so `credentials = {}`, a function
called load_credentials and even a comment mentioning credentials all
prompted while performing no I/O. The check applies to string literals now,
with the raw scan kept for source that does not parse.
* Split git short-option clusters and close five more gate gaps
- Git combines short options, so `git push -qf`, `git checkout -qf` and
`git branch -qD` never matched the exact-string flag sets and ran without a
prompt. Clusters are split before the destructive flags are checked. Also
adds the short `-f` spelling to the branch set, which moves a ref and can
abandon its commits.
- `getent shadow` and `getent gshadow` return password hashes straight from
NSS, so the read never spells out a path for the sensitive-path check to
find. The database name is gated instead; ordinary lookups (hosts, passwd)
still run.
- The account-management set covered useradd and usermod but not adduser,
deluser, addgroup, delgroup, groupmod, gpasswd, newusers or chgpasswd, so
`gpasswd -a user sudo` granted group membership silently.
- at and batch hand a payload to atd, which runs it later as this user and
outside this invocation's blocklist, resource limits, timeout and
cancellation. They belong with crontab.
- A command word bash builds without the NAME=value form (printf -v, read)
left nothing at command position to screen. A bare variable executed as a
command that assignment expansion could not resolve now fails closed. A
variable used as a path prefix is deliberately excluded: ${VENV}/bin/python
still leaves a literal basename the scan can read.
* Stop prompting on six inspection shapes and close eighteen gate gaps
Over-prompting fixes, which matter most here since not interrupting ordinary
work is the point of the change:
- `git clean -n` and `--dry-run` list what would be removed and remove nothing,
so they are inspection commands. The subcommand was gated regardless of its
flags; a dry run is now recognised in the same segment.
- The listener check matched a module name anywhere in the line, so
`echo 'python -m http.server'` and grepping for it prompted. It is anchored at
command position now, like the server-binary check beside it.
- An MCP name that reads names its SUBJECT, not the action: `get_release`,
`get_invoice`, `search_code` and `get_code` were prompting because the impact
and runtime-noun patterns fired on the noun. A read verb now suppresses both,
while an execution verb still wins.
- Free text is not a statement. An issue body or chat message that mentions
DELETE FROM, a credential file or a path was read as an action. Statements are
taken from the query-bearing argument names, and paths are skipped only for
the prose names, since a path can be carried under any other name.
- curl and wget presence was decided by substring, so `grep curl notes.txt &&
wget -T 5 ...` lent curl's option letters to wget.
Gaps closed:
- git checkout-index -f overwrites the working tree from the index; git tag -d
and -f delete or replace a ref; git switch -C and checkout -B reset an
existing branch the way branch -f does.
- Ending a process (kill, pkill, killall, taskkill, tskill) or the machine
(shutdown, reboot, halt, poweroff) was ungated, though the Python os.kill
equivalent already prompted. setcap grants file capabilities without sudo.
- A network client behind a wrapper (env curl -T) was missed because the client
check ran before the wrapper was resolved. slogin is a standard ssh alias and
was in neither set. wget spells the request method --method=DELETE.
- A tracer (strace, ltrace, valgrind, perf) runs the rest of the line as a
child, so the real command sat in argument position behind it.
- A redirection may precede the command word, so `</dev/null` hid what followed
from both scanners. `exec -a NAME cmd` puts a name where the command goes, and
the Windows `if exist FILE cmd` form puts an operand there.
- In Python: a walrus binds a module or a callee just like an assignment,
builtins.__import__ is the attribute form of __import__, and psutil ends a
process exactly as os.kill does. The psutil check is keyed on the import so an
unrelated .kill() on a user object keeps running.
- Over MCP: a credential carried in an argument NAME (Authorization, X-API-Key,
Cookie) goes out whatever its value looks like; collaborator and team-member
grants are access changes like the role verbs; and a recurring subscription
bills repeatedly.
* Bound the classifier's input and stop prompting on four more ordinary shapes
Found by simulating the whole corpus against pre-PR main on Linux, macOS and
Windows tokenizers and diffing the two, then feeding the classifier adversarial
input.
Robustness:
- The credential-path pattern backtracks superlinearly, so a long argument made
a single classification take seconds. Measured on main as well as here, so it
predates this change, but this change makes the auto gate the default and so
runs it on every call. Text far past any real path, and a command far past any
real command, now fail closed: they ask rather than spending unbounded time
deciding. Worst case over the adversarial set drops from a hang to 13 ms.
Over-prompting fixes:
- A container CLI reading its own state (docker ps, docker images, docker logs,
kubectl get) is inspection. The whole CLI was gated because the escape lives
in the arguments of run/exec, so the read subcommands were caught with it. An
unrecognised subcommand still asks, so the list can only be too small.
- A python payload is screened with the same analyzer the python tool uses, so
`python -c 'import torch; print(torch.__version__)'` runs while a destructive
one-liner still asks. A payload that does not parse fails closed, since shell
quoting may have mangled it. The other runtimes have no analyzer here and stay
gated.
- An assignment with no command after it runs nothing: every terminal call gets
its own shell process, so `export PATH=...` on its own dies with that process.
Verified against real bash rather than assumed.
- For the search paths other than PATH (PYTHONPATH and friends), a relative
entry points inside the session workdir, which is the agent's own directory,
so `PYTHONPATH=. pytest` runs. An absolute or escaping entry can shadow a real
module and still asks. PATH itself counts for every value, because a relative
entry there is the sharpest form of the hijack (`PATH=. ls` runs ./ls).
Net effect on the probe corpus, identical on all three platforms: ordinary and
inspection commands go from 99 of 136 prompting to 0, dangerous stays at 99 of
99, and the always-on hard-block set loses nothing and gains six entries.
* Tighten the permission-mode comments
Comment-only pass over the code this branch added. Every explanation is
collapsed to the fewest lines that still read clearly, redundant restatements
of the code are dropped, and a handful of blocks that had drifted away from the
constant or branch they describe are moved back next to it.
The non-obvious behaviours keep their note, just shorter: an unforced
`git worktree remove` refusing on a dirty worktree, a bare `-c` yielding an
empty attached value rather than None, `.` being the POSIX synonym for
`source`, prose keys being skipped rather than path keys allowlisted, and the
route keeping an unset mode lenient so non-streaming clients still work.
No code, string literal or test expectation changed.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Gate the navigation sinks reached by bracket access
The canvas egress check gated location.assign / location.replace and an
assignment to location.href, and it already handled bracket access for the
fetch family, but not for the navigation sinks. So `location['assign'](url)`
and `location['href'] = url` auto-ran and could navigate the preview frame to
an attacker URL with the page contents appended, which is the same egress the
dot forms already gate.
Both bracket forms are covered now, including a fully bracketed host
(`window['location']['href']`). The names are anchored to location so ordinary
bracket keys stay static: a string's own `['replace']`, an object's `['href']`,
and reading `location['href']` all still run without a prompt.
* Gate seven more ways a command reaches the shell in auto mode
git submodule foreach runs its argument in every submodule, so the payload is
a command in its own right; it now recurses through the terminal classifier and
through the hard-block scan. An awk program can shell out with system() or by
piping to "sh", so the program text is screened for those two shapes while
ordinary field work (awk '{print $1}') keeps running.
setpriv changes privilege and then execs what follows, so it is transparent to
the scan (setpriv --nnp rm -f x resolves rm) and its privilege-raising flags
(--reuid, --ambient-caps, --bounding-set) prompt on their own. fallocate
punches, zeroes or collapses a range in place, which destroys file contents,
so those flags prompt while plain allocation (-l SIZE) does not.
vars(os)["remove"] and os.__dict__["unlink"] resolve an attribute the same way
getattr does, so the module namespace dict is screened with the same key rules,
anchored to a filesystem module so an ordinary d["remove"] stays out.
Removing a package (pip uninstall torch, uv pip uninstall, conda remove) tears
down the environment the backend itself runs in; installing into it does not,
and stays automatic.
The listener check was anchored at command position, so a wrapper in front of
it (env python -m http.server, timeout 60 python -m uvicorn) slipped past. The
module after -m is now resolved at the token level, after wrapper resolution.
Adds 54 rows to the classifier tables covering both directions.
---------
Co-authored-by: danielhanchen <unslothshared@gmail.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: danielhanchen <unslothai@gmail.com>
* Fix GGUF tool chat server recovery
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Cover MTP precedence and loosen the replay assertion for PR #7424
Add a regression test for the MTP branch of the tool-loop respawn retry: the
file-wide _make_backend stub forces _maybe_recover_from_mtp_crash to False, so
nothing exercised the case where an MTP crash reload is already claimed and an
ordinary same-config respawn must not run on top of it. Cover both the next
tool-loop request and the final synthesis pass.
Replace the whole-payload equality assertions with a field-wise check. Comparing
the full dict pins max_tokens to the value derived from the dead server's
effective context, so a later fix that rebuilds server-derived defaults after a
respawn would read as a test failure rather than an improvement.
Document that the one-retry budget is per model request, not per chat turn.
* Recover from prefill-time deaths and stop respawn racing the MTP reload
Two gaps in the tool-loop respawn retry, both reproduced before fixing.
A child that exits during prefill has already accepted the socket, so httpx
raises ReadError, WriteError or RemoteProtocolError rather than ConnectError.
Those all arrive before the response opens, which is exactly the window where a
replay is safe, but the helper only caught ConnectError and gave up. Widen the
catch to NetworkError plus RemoteProtocolError. Timeouts stay excluded on
purpose: they mean the server is slow, not dead, and retrying one would spend
the 20 minute first-token budget twice. Windows resets connections where Linux
refuses them, so this also covers the common Windows presentation.
_maybe_recover_from_mtp_crash returns False both when the crash is not an MTP
crash and when an MTP-free reload is already in flight. Callers read that as
permission to respawn, so _respawn_if_dead replayed the crashing MTP kwargs and,
by replacing the process, made the in-flight reload abort on its own newer-load
check. Skip the respawn while that reload owns the corpse. The guard lives in
_respawn_if_dead so the plain chat path gets it too.
Regression tests for both, including a guard against retrying prefill timeouts.
* Release the MTP single-flight claim when the reload never starts
_mtp_runtime_fallback_in_progress is claimed before the reload thread exists, and
only that thread's finally clears it. Two statements ran in between with no unwind
path: re-reading _last_load_kwargs, which an unload can null underneath us, and
Thread.start(), which raises under the thread exhaustion that is exactly the
pressure killing llama-server in the first place. Nothing else ever resets the
flag, so a failure there latched it for the life of the process.
That was survivable before, since respawn ignored the flag. It is not now: the
guard added in db78184be keys off the flag alone, so a latch would silently
disable auto-respawn for every later model, including plain non-MTP ones. Read
the kwargs and process once before claiming, and release the claim if the thread
cannot start.
Restore the whole-payload equality assertions. Comparing field-wise was meant to
leave room for rebuilding server-derived defaults on replay, but the payload is
built once before the retry and re-sent unchanged, so the looser check only
dropped seven real keys and added a vacuous seed comparison.
Also correct the docstring: llama-server flushes its 200 at slot start, so a
death during decode arrives with the response already open. The pre-header window
this covers is an upload still in flight or a request waiting behind busy slots.
* Confirm the child exited before spending the retry
A closing llama-server can beat its own exit status: the socket error arrives while
poll() still reports the process running. _respawn_if_dead then took the alive
branch, handed back the stale _healthy, and the caller read that as a successful
respawn and spent its single retry on the same corpse. When that retry failed,
attempt was no longer 0, so no respawn ever happened and the turn died, with a log
line claiming a respawn that had not occurred. The window matters most for the
pre-header ReadError and RemoteProtocolError shutdowns the retry now covers.
Wait a bounded second for the exit status before calling the child alive. The same
race is already conceded in _maybe_recover_from_mtp_crash, whose recovery thread
polls for 5s because the error can arrive a beat early; 1s here because this runs
on the request path, and a genuinely live server, including one a concurrent caller
has just respawned, still returns promptly.
* Tighten the recovery comments
* Harden the respawn path around concurrent unloads and replacements
Two problems with the reap grace loop, both found by review.
Skip the grace when the server was already replaced. A caller queued on
_respawn_lock behind someone else's respawn woke holding the healthy replacement,
could not tell it from the child its own request had used, and waited out the full
grace. That sleep is under the lock, so the waits serialised: four concurrent
generations cost roughly three grace periods before any retry began. Capture the
process before taking the lock and return early once it has been swapped.
Do not respawn a server that is being torn down on purpose. unload_model() sets
_cancel_event and only clears _last_load_kwargs after the kill, so a request losing
its connection mid-unload could watch that deliberate exit through the grace loop,
read the stale kwargs and load the model straight back; a model switch landing
during the wait was reverted the same way. Re-check the cancel flag and the process
identity under _serial_load_lock before capturing the replay kwargs, matching what
the MTP-crash reload already does.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Tighten the respawn comments
* Do not charge the reap grace to a server that is still serving
The grace loop added for the not-yet-reaped race waits on poll(), which for a
live child never returns, so every transient transport error paid the full
_RESPAWN_REAP_GRACE_S. That sleep is held under _respawn_lock, so the cost
serialised: measured 1002 ms for one caller and 8.02 s for eight concurrent ones,
against 0 ms on main. A working install pays this, not a broken one.
A llama-server's listening socket dies with the process, so a loopback connect
separates the two cases in microseconds. Probe it first and return immediately
when the port still accepts; fall through to the grace only when the port is
gone, which is the case the grace exists for. Back to 0.7 ms for one caller and
0.00 s for eight.
Cross-checked on real hardware over Qwen3.5-2B, Llama-3.2-1B, Gemma-3-4B with
mmproj and Qwen3-30B-A3B: decode throughput within noise of main (-0.06%, -3.71%,
+2.57%, +0.29%, against a 54-232% spread between rounds of a single run), output
byte-identical on every round, tool-path recovery restored on the three families
whose model calls the tool, and plain-chat recovery still working on all four.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Make the respawn lose to a deliberate unload in every window
Two follow-ups on the respawn path, both reproduced first.
Check _cancel_event before the socket fast path. unload_model sets the flag before
it kills, so the child is still accepting when the probe runs; returning the stale
_healthy there aims the retry at a server that is deliberately going away.
Close the unload TOCTOU. The old cancel check sat under _serial_load_lock, which
unload_model never takes, so an unload could land entirely between that check and
load_model and the captured kwargs would restart a model the user had stopped.
Snapshot the kwargs, the flag and a new _unload_epoch together under _lock, the
lock unload does hold, so a teardown is either wholly before the snapshot or
wholly after it. load_model clears _cancel_event on the way in, so the epoch is
the only evidence that survives; when it moves during the reload the replacement
is unloaded again rather than left running.
_lock stays uncontended across load_model, which would deadlock a plain Lock and
block /status for the length of a load. Error-path latency is unchanged: 0.6 ms
for a live server and 0.00 s for eight concurrent callers.
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: danielhanchen <unslothai@gmail.com>
Co-authored-by: danielhanchen <danielhanchen@gmail.com>
* studio: shard export checkpoint loads across all visible GPUs
Export checkpoint loading always used unsloth's from_pretrained default of
device_map="sequential", which stacks the whole model on GPU0. On a multi-GPU
host this OOMs GPU0 while the other GPUs sit empty, so a GGUF export that would
comfortably fit across the machine fails with CUDA out of memory (#7053).
Add _multi_gpu_device_map_kwargs(): when the CUDA/ROCm host exposes more than
one visible GPU and get_device_map resolves to "balanced" (the same policy the
inference loader already uses), pass device_map="balanced" to every
from_pretrained in load_checkpoint. In every other case -- single GPU, CPU,
MLX, or any probe failure -- it returns {} so the loader default is untouched.
Fixes#7053
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* studio/save: reach the UUID/MIG fallback, release sharded models before quantize
Two review fixes on the multi-GPU export sharding:
1. UUID/MIG CUDA_VISIBLE_DEVICES masks resolve to no numeric ids, so the
len(visible) > 1 gate skipped get_device_map entirely and large exports on
those hosts still stacked onto GPU0. An empty id list now routes to
get_device_map(None), whose visible-count fallback exists for exactly this
case; a genuinely GPU-less host still resolves "sequential" and keeps the
loader default.
2. The compressed (FP8/NVFP4) export freed GPU memory before its llm-compressor
subprocess only for single-device models -- a plain .to("cpu") is invalid on
an accelerate-dispatched model, so a multi-GPU-sharded checkpoint stayed
resident on every GPU while the subprocess loaded a second copy. The release
is factored into _offload_model_for_quantize_subprocess /
_restore_model_after_quantize_subprocess: dispatched all-GPU shards get their
accelerate hooks removed, move to CPU, and are re-dispatched over the
recorded hf_device_map afterwards. Maps with cpu/disk targets (already
offloading) and quantized models are left alone, as before.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* studio/save: budget merged tensors per device, restore hooks if CPU offload fails
Two review fixes on the multi-GPU export path:
1. The LoRA-merge save path budgeted every merged tensor against GPU0
(get_device_properties(0) + unqualified memory_allocated()). A merged tensor
lives on the GPU of its source layer, so for a model sharded across GPUs
(the device_map="balanced" this PR enables) GPU1+ could OOM as their weights
accumulated while only GPU0's headroom was checked. Budget against W's own
device via a per-device cache; single-GPU behavior is unchanged (W on GPU0).
2. _offload_model_for_quantize_subprocess removed the accelerate hooks and then
moved a dispatched model to CPU; if that move raised (host RAM too small for
the sharded checkpoint) the model was left hookless and half-moved, breaking
later exports in the same worker. It now re-dispatches (or, for the
single-device path, moves back) on a failed move before aborting the offload.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* studio/save: release sharded models before the torchao reload too
The portable torchao FP8/INT8 export freed the in-memory model only when every
parameter sat on one device, then reloaded a second copy with
device_map="auto". A checkpoint loaded through the new multi-GPU export map is
accelerate-dispatched across several GPUs, so that single-device gate never
fired and the original stayed resident on every GPU during the reload -- an OOM
for exactly the models large enough to have needed the sharded load.
It now uses the same _offload_model_for_quantize_subprocess /
_restore_model_after_quantize_subprocess pair as the compressed export, which
removes the accelerate hooks, moves to CPU, and re-dispatches over the recorded
hf_device_map afterwards. Those helpers are extended to XPU as well, since
torchao also runs on Intel GPUs and the path they replace covered both.
* studio/save: release quantized and cpu-spilled shards before quantize reloads
Two cases the release helper skipped outright, both of which leave GPU memory
held while the compressed subprocess or the torchao device_map="auto" reload
allocates a second copy:
- Quantized models. ExportBackend.load_checkpoint loads 4-bit by DEFAULT, so the
common Studio export hit the is_loaded_in_4bit guard and kept a quantized shard
on every visible GPU. They are now attempted like any other model: transformers
refuses .to() for some bitsandbytes builds, but that refusal raises before
anything moves, so the existing recovery path restores the model and returns
None -- best-effort where the stack allows it, old behaviour where it does not.
- Maps that spill to CPU. Any non-GPU target disqualified the whole model even
though the GPU-mapped modules were still resident and are exactly what needs
reclaiming. A cpu spill is safe to move (those weights are already in host RAM)
and is now released; only disk/meta targets are still skipped, because
accelerate keeps those parameters off the model and moving would try to
materialize the whole checkpoint. An all-CPU map is skipped as a no-op.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Fix multi-GPU offload for PEFT exports and fall back when sharding OOMs (#7215)
The dispatch branch of _offload_model_for_quantize_subprocess never ran for a
PEFT model: the wrapper proxies _hf_hook, so remove_hook_from_submodules raised
AttributeError and the bare except returned None. Studio always loads adapters,
so the new balanced map turned the offload off (0 percent freed against 91.8 on
the sequential path it replaces).
- resolve the real dispatch root before removing or replaying hooks
- snapshot and replay hooks, tensor placements and instance forwards; a plain
re-dispatch rebuilds hooks against the post-PEFT tree (395 to 1379) and drops
the fused kernels accelerate captured into _old_forward before unsloth patched
- drop the accelerator side of tied_params_map so the offload actually frees
- pass skip_keys on the fallback dispatch_model
- log the swallowed exception instead of returning None silently
- guard _unsloth_save_torchao_with_given_config like its two siblings
- retry the export load once on the loader default when the balanced map OOMs,
which happens when a training or chat job already owns the other GPUs
Measured on 4x B200 with Qwen3-0.6B: 89.9 percent freed bf16 and 79.7 percent
4bit under balanced, logits bit-identical, hooks and placements restored
exactly, 184 Params4bit round-tripped unchanged including nested state2.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Keep the original offloaded until the torchao copy is released, and retie shared weights (#7215)
Two follow-ups from review of 8b6b4ca0b.
_unsloth_save_torchao_with_given_config restored the original inside a finally
that ran as soon as from_pretrained returned, so the original and the quantized
copy were both resident while the copy was still being saved. The restore now
sits in an outer finally that covers saving and releasing quantized_model, which
is what the two sibling paths already do.
The dispatch replay did not preserve tied embeddings. A CPU round trip repoints
every tensor and accelerate's tied_params_map is keyed on the old pointer, so
replaying the hooks produced two independent parameters. Reproduced on a tied
Llama: lm_head picked up its own storage, the embedding was duplicated in VRAM,
and an update to one no longer reached the other. The snapshot now records tied
groups (named_parameters(remove_duplicate=False), since the default hides one
half of every pair) and re-ties them after placements are restored.
Verified: tie preserved, no extra storages, live CUDA storage census identical
before and after, updates propagate again, logits bit-identical, and the 4 GPU
invariants unchanged at 89.9 percent freed bf16 and 79.7 percent 4bit.
* Keep meta tensors out of tie groups, restore accelerate move guards, retry CPU spills (#7215)
Four follow-ups from review of a58f1086b.
Meta tensors all report storage pointer 0, and accelerate parks every
CPU-offloaded parameter on meta, so grouping by pointer collapsed them into one
fake tied group. Reproduced with a balanced map that spills two blocks to CPU:
18 meta parameters in a single group with shapes 64x64, 32x64 and 128x64, which
the retie step would have overwritten with the first one. Meta and null-pointer
tensors are now skipped, and the retie also checks shape.
remove_hook_from_submodules deletes the to/cuda/xpu wrappers dispatch_model
installs to stop a caller moving an offloaded model. The snapshot now records
and replays those alongside forward and _old_forward.
The single-device retry only matched OOM, but a balanced map that spills to CPU
is refused by bitsandbytes with a plain ValueError saying modules were dispatched
to the CPU or the disk (transformers quantizers/quantizer_bnb_4bit.py:128), with
no memory wording. That is now retryable too, which matters because Studio loads
4-bit by default and busy secondary GPUs are exactly when balanced spills.
The torchao path dropped the quantized copy at the end of the try, so a failure
in save_pretrained left it resident while the original was restored. The del
moved into the finally, ahead of the restore.
Four regression tests added; suites now 25 and 9.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Retry exports whose multi-GPU load silently offloads to CPU, and clear the failed torchao traceback (#7215)
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Tighten comments for PR #7215
* Keep gradients across the export offload and release the failed torchao copy (#7215)
* Tighten comments for PR #7215
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Daniel Han <danielhanchen@gmail.com>
Co-authored-by: Daniel Han <unslothai@gmail.com>
* fix(studio): support hostname-based enterprise proxies
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* fix(studio): strip userinfo from proxy fetch targets
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Lee Jackson <130007945+Imagineer99@users.noreply.github.com>
* ROCm/AMD CI coverage: arch-table parity, native-Linux lib prepend, RDNA4 grouped_mm, discovery-based shell suite
Three merged ROCm fixes shipped without tests, and the CI wiring that
would have run them was gated on files the fixes do not touch.
Tests added (113):
tests/studio/install/test_rocm_arch_table_parity.py (27)
diffs the four duplicated gfx -> AMD pip-index tables across
install.sh, install.ps1, studio/setup.ps1 and install_python_stack.py,
plus the GPU-name -> arch tables and the torch 2.11 pin allowlist.
tests/studio/install/test_rocm_native_linux_lib_dirs.py (26)
covers #7233: system-ROCm lib dirs prepended ahead of bundled
libggml-hip, the /dev/kfd + not-WSL + libhsa gate, the opt-out env
var, root resolution order, and source parity between the two copies.
studio/backend/tests/test_grouped_mm_rdna4_fallback.py (46)
covers #7292: registration on the CUDA dispatch key, grouped and
ungrouped numerics, bias/dtype promotion, and the Linux HIP<7.13 +
RDNA4 name gate, executed from the shipped source rather than a copy.
tests/studio/test_ci_shell_suite_coverage.py (14)
fails if either shell runner goes back to a hardcoded list or skips
a file without a recorded reason.
CI wiring:
studio-backend-ci.yml: add install.sh / install.ps1 to the path filter
(the suites it runs assert against those two files, so install-only
changes -- the shape most AMD/ROCm routing fixes take -- skipped it),
and replace the 13-file hardcoded shell list with directory
discovery. That list had fallen seven files behind, including
test_strixhalo_wsl_reroute.sh, the only shell coverage of the ROCm
WSL reroute, which had never run on a PR.
tests/run_all.sh: same discovery loop so local and CI agree.
* Test review fixes: assert on outcomes, not on the code under test
Self-review of the previous commit found four tests that passed for the
wrong reason.
1. The arch-table parity test pinned expected gfx ids copied out of the
shipped tables, which enshrined three upstream inaccuracies as
correct: RX 9070 (non-XT) is gfx1201 not gfx1200, RX 7800 XT is
gfx1101 not gfx1100, and PRO V710 is gfx1101 not gfx1102 per AMD's
ROCm compatibility matrix. The expectation is now the AMD pip index
leaf -- the thing the tables exist to produce, and what a wrong
answer costs the user. The three known drifts are listed explicitly
with a test asserting they stay cosmetic, i.e. that the wrong and
right ids still map to the same wheel index. That test turns red the
day one of them starts routing users to the wrong wheel.
2. The RDNA4 device-name test extracted the regex from worker.py and
then matched with it, so it could not fail. Widening the pattern --
the dangerous edit, since it forces the slow Python mm fallback onto
RDNA3 users -- would have been silently accepted. It now reads the
live pattern and checks it against fixed cases, plus asserts the
name match stays guarded by `not _lin_arch` and that the name is
lowercased before matching.
3. The CI-coverage test matched a verbatim line of studio-backend-ci.yml,
so reindenting the step would fail the build while a real regression
to a hardcoded list could slip past a reformat. It now parses the
YAML, finds the step by name, and asserts on the glob plus the
absence of individual filenames. The path-filter test likewise reads
the parsed trigger instead of scanning raw text.
4. A set comprehension in the parity helper had a ternary whose branches
were identical.
Mutation-tested: widening the RDNA4 regex, desyncing one copy of the
name table, dropping install.sh from the path filter, and re-skipping
the ROCm WSL shell suite each fail at least two tests. Verified on
Linux (WSL Ubuntu 24.04) with CI's torch pin: 86 + 48 pass.
* Fix three wrong gfx ids in the GPU-name arch tables
The name -> gfx tables disagreed with AMD's ROCm compatibility matrix on
three entries. Corrected against the "Radeon GPU" list at
rocm.docs.amd.com/en/latest/compatibility/compatibility-matrix.html:
RX 9070, RX 9070 GRE gfx1200 -> gfx1201 (Navi 48, same die as the XT)
RX 7800 XT, RX 7700 XT gfx1100 -> gfx1101 (Navi 32, not Navi 31)
PRO W7700 gfx1100 -> gfx1101
PRO V710 gfx1102 -> gfx1101 (Navi 32, not Navi 33)
No wheel changes for anyone: gfx1200/gfx1201 both resolve to gfx120X-all
and gfx1100/gfx1101/gfx1102 all resolve to gfx110X-all, in all four copies
of the index-family map. That collapse is why the errors survived being
copied into six places -- the leaf-level tests could not see them.
It was not purely cosmetic, though. install.sh's second copy feeds
"Tip: set UNSLOTH_ROCM_GFX_ARCH=<arch>", so a 7800 XT user following the
printed advice exported gfx1100 and made a wrong id authoritative for
every later run. It would also have become a real misroute the moment AMD
split a family across index leaves, as they already do for gfx1151/gfx1150.
Fixed in all six places, which is two more than the table's own "kept in
sync with" comments claim exist:
install.sh _infer_amd_gfx_arch_from_gpu_name
install.sh case "$_gpu_disp_mkt" (banner + env tip; undocumented)
studio/setup.sh
install.ps1
studio/setup.ps1
studio/install_python_stack.py
Ordering is preserved: the gfx1102 arm still precedes gfx1101 in the shell
copies so "RX 7700S" cannot fall onto the "RX 7700" glob, and the
PowerShell copies keep the (?!S) lookahead.
Test changes:
- test_rocm_arch_table_parity.py gains _AMD_DOCUMENTED_ARCH, exact gfx
ids transcribed from AMD rather than from the tables. Agreement between
six copies proves nothing when all six were transcribed from the same
mistake, so the ground truth has to come from outside. Verified it
catches the bug: against the pre-fix tables it fails 6 tests.
- The parity check now covers all six copies. It had four; the two
install.sh copies were being treated as one, and
_WIN_GPU_NAME_ARCH_TABLE was not checked at all.
- test_rocm_support.py's TestGfxArchNameFallback pinned two of the wrong
ids as expected values; updated, and extended with a 9060 XT and a
7900 XTX case so each RDNA3/4 die is represented.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Guard against unregistered copies of the GPU-name arch table
Counting the copies by hand is what let them drift: the in-code "kept in
sync with" comments claimed four, the arch-id fix found six, and scanning
the tree turns up a seventh.
TestNoUnregisteredArchTable rediscovers the copies from the source tree
instead of trusting a hand-maintained list. A table line is one that names
a card and gives its arch; real tables score 9-17 such lines and the only
other hits in the repo are two single-line prose comments, so the
three-line threshold is not load-bearing. A companion test asserts the
scan still finds the known copies, so the heuristic cannot go blind and
pass by finding nothing.
The seventh copy is tests/_zoo_rocm_spoof.py, the fixture other ROCm tests
build their fake AMD host from. It states the mapping backwards (gfx ->
the name torch should report), which makes it an independent witness: it
had gfx1101 -> RX 7800 XT and gfx1201 -> RX 9070 XT right while all six
installer copies were wrong, and nothing compared the two. Now they are
round-tripped against each other.
RX 6700 XT is pinned as a known divergence rather than normalised. AMD's
compatibility matrix documents no consumer RX 6000 card and no gfx1031 at
all, the installer arm is commented "gfx103X family", and gfx1031 appears
only as an index-family key, never as a value a name table emits. With no
external source to correct against, changing shipped behaviour would be
guesswork. A test fails if the divergence ever disappears, so the
exemption cannot go stale.
Also adds the reverse of the AMD-matrix check: a documented card that
matches no arm anywhere is a silent CPU fallback rather than a wrong id.
This cannot detect hardware nobody transcribed, which would need a live
fetch of AMD's matrix and a non-hermetic suite; the docstring says so
rather than implying coverage that is not there.
Verified on Linux: 478 passed, plus all five new guards mutation-tested
to confirm each fails when its invariant is broken.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Docstring said six copies; the list under it now has seven
* tests: run discovered shell tests with bash, not sh
tests/run_all.sh discovered tests/sh/ instead of listing files, but still
invoked each one with sh. Every file there declares a bash shebang, and on
Debian/Ubuntu /bin/sh is dash: test_apt_distro_prompt.sh,
test_studio_home_node_dir.sh and test_with_llama_cpp_dir_link_behavior.sh
fail on bashisms under dash and pass under bash. The old hand-written list
happened to name only dash-clean files, so switching to discovery is what
surfaced it. Backend CI already used bash, so this was a local-only break.
Guarded by a new test asserting both runners invoke tests/sh/ with bash.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Fix Krackan Point (Radeon 860M/840M) routed to the gfx1150 wheel index
The GPU-name tables map 860M/840M and the Ryzen AI 7 350 / AI 5 340 CPU
strings to gfx1150, but Krackan Point is gfx1152. AMD's own lemonade table
(src/cpp/server/system_info.cpp) maps both Krackan iGPUs to gfx1152.
Unlike the three ids already fixed here, this one is not wheel-neutral:
repo.amd.com publishes gfx1150 and gfx1152 as separate index leaves with
separately built torch wheels, so these laptops were installing wheels
built for a different LLVM target. gfx1152 was absent from the codebase
entirely, so it needed the index-family maps, the torch 2.11 floor lists
(same _grouped_mm bug as gfx1150/1151), the Strix reroute set and the
Windows arch allowlist as well as the seven name tables.
The parity test added in this PR did not catch it because its AMD-matrix
expectations stopped at 890M/880M. Added the APU rows, so the case that
actually changes a wheel is now covered: reverting the tables fails 9
tests naming 860M, 840M and Krackan.
gfx1153 (Ryzen AI 5 430 era) is left alone; AMD publishes no gfx1153
wheel family, so there is nothing to route it to.
Verified: bash -n on both shell installers, PowerShell AST parse on both
.ps1 files, python ast.parse on all touched modules, install suite 1334
passed with no new failures against main, shell suite 20 files.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Add gfx1152 to unified-memory classifiers, make parity allowlist set-based
Krackan Point (gfx1152, Radeon 860M/840M) is the third RDNA 3.5 APU and
shares one GPU/system-RAM pool exactly like Strix Point (gfx1150) and
Strix Halo (gfx1151), but only the installers knew about it. The two
runtime classifiers still had two-element arch sets, so a Krackan laptop
got the 0.90 discrete headroom factor on a shared pool and ran llama.cpp
without GGML_CUDA_ENABLE_UNIFIED_MEMORY.
- worker.py _rocm_classify_unified_memory: add gfx1152 to the arch set,
and 860m/840m to the device-name fallback. The NVIDIA GeForce 840M
cannot collide there: the function is only reached under _hw.IS_ROCM.
- llama_cpp.py _amd_apu_wants_unified_memory: add gfx1152 to the arch set.
- Tests for both, including the :sramecc-:xnack- suffix form.
TestGfx211AllowlistParity compared four hardcoded allowlist strings, so
adding gfx1152 to all four installers correctly turned three assertions
red without any installer actually disagreeing with another. Each test
now extracts the set its installer holds and compares it to one EXPECTED
constant. Order and spacing are free, membership is not, and the next
leaf is a one-line edit instead of four.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Daniel Han <danielhanchen@gmail.com>
* Studio: reset quantized KV cache to f16 when flash-attn-off fallback fires
Studio force-enables --flash-attn on for GGUF launches. On a hard startup
or first-decode crash it retries via _with_flash_attn_off, which flipped FA
off but left --cache-type-k/-v untouched. A quantized KV cache (q8_0, q4_0,
q4_1, q5_0, q5_1, iq4_nl) requires flash attention in llama.cpp, so the retry
itself aborted at init with 'V cache quantization requires flash_attn' instead
of recovering.
Reset any quantized --cache-type-k/-v to f16 in the FA-off fallback path so
the retry can actually launch. Non-quantized types (f16, bf16, f32) run fine
without flash attention and are left unchanged. Handles long and short flag
forms and both space and equals syntax, rewriting in place to preserve list
length. Adds pytest coverage.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: FA-off fallback resets only the quantized V cache and drops env-only V cache
Only the V cache requires flash attention in llama.cpp; a quantized K cache
runs fine without it. Restrict the FA-off crash-recovery reset to the V axis
(main and draft) so a memory-constrained config keeps its quantized K cache
instead of risking an OOM on the recovery. Also drop an inherited quantized V
cache set purely through the environment (LLAMA_ARG_CACHE_TYPE_V /
LLAMA_ARG_SPEC_DRAFT_CACHE_TYPE_V) at the FA-off retry sites, which the argv
rewrite cannot reach, so the child falls back to the f16 default rather than
aborting.
* Studio: normalize underscore V-cache aliases in the FA-off fallback
llama.cpp rewrites '_' to '-' for any '--' long option before matching,
so a pass-through --cache_type_v q8_0 enables a quantized V cache just
like --cache-type-v. The FA-off crash-recovery reset only matched the
hyphenated spelling, so the underscore alias slipped through and the
retry still aborted with "V cache quantization requires flash_attn".
Canonicalize the flag name the same way before matching (short flags and
the type value are untouched).
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
* Fix reasoning-only Qwen3.6 completions in Studio
* Address reasoning-only review findings
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
* feat(studio): add DoRA support to studio
* fix: added use_dora fast encoder LoraConfig and gated use_dora on AdapterMethod
* fix(studio) serverside normalization for use_dora=true - add note documenting use_dora is silently dropped on diffusion
* fix: dora button disabled on mac, add preflight guard on GGUF lora export, mismatch now correctly falls through to existing error instead of silently no-opping
* Studio: add dora to the WizardState LoRA variant union for consistency
* Reject --use_dora on the MLX (Apple Silicon) CLI path
---------
Co-authored-by: danielhanchen <unslothai@gmail.com>
* fix(studio): stop false MTP/vision capability reports (#7302)
MTP probing only inspected the first physical --spec-type help line and
treated empty/crash --help output as "lacks MTP", which false-warned on
otherwise capable builds. Parse the full --spec-type help block, fail open
when the probe is inconclusive, and stop blaming bare mmproj crashes on a
projector-format mismatch when the text-only retry also fails.
Fixes#7302
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* fix(studio): tighten MTP probe semantics per Codex review (#7302)
Treat nonempty --help without --spec-type as definitive no-MTP, keep only
empty/crash probes inconclusive, skip binary_no_mtp UI hint on inconclusive
loads, and stop reporting supports_mtp=True in /status for unknown probes.
* Treat failed llama-server --help probes as inconclusive (#7302)
Gate definitive no-MTP results on a zero exit code so crash diagnostics with
nonempty stderr do not re-enable the false lacks-MTP warning path.
* Add returncode to probe test mock so probe_ok gating passes
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Fail open in /status when the MTP probe is inconclusive
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Report missing llama-server as lacking MTP in /status
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Tighten comments in MTP/mmproj probe changes
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Daniel Han <danielhanchen@gmail.com>
---------
Co-authored-by: Daniel Han <danielhanchen@gmail.com>
Co-authored-by: Roland Tannous <115670425+rolandtannous@users.noreply.github.com>
Co-authored-by: oobabooga <112222186+oobabooga@users.noreply.github.com>
* fix(studio): resolve bare git on Windows sandbox PATH
Sandboxed terminal tools rebuilt PATH as venv + System32 only, so
user-installed Git under Program Files never resolved by bare name.
Append absolute host PATH dirs after the curated prefix and inherit
PATHEXT on Windows (#7317).
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* fix(studio): restrict sandbox PATH inheritance to Windows Git dirs (#7323)
Only append Git-for-Windows install directories from the host PATH on
Windows, instead of every absolute entry. This fixes bare `git` resolution
(#7317) without letting user-writable dirs (venv, node_modules/.bin)
shadow auto-safe terminal commands.
* Pin sandbox PATHEXT to block cwd script hijacks (#7317)
Use a fixed .EXE;.COM list instead of inheriting the host PATHEXT so
cmd cannot resolve auto-approved bare names from workdir .BAT/.CMD stubs.
* Resolve sandbox git dir via shutil.which and disable cwd exe lookup
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Keep non-exe git launchers resolvable under restricted PATHEXT
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Restrict inherited sandbox git dir to system install roots
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Drop SystemRoot trust and canonicalize short paths for sandbox git
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Resolve Program Files via known-folder API and append canonical git dir
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Trust native Program Files on 32-bit Windows and stub program roots in tests
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Scan PATH for a trusted git and derive native Program Files root
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Drop ProgramFiles env from the trusted-root fallback
* Fail closed when trusted Program Files root cannot be resolved
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Daniel Han <danielhanchen@gmail.com>
* Studio STT: only load safetensors weights for custom dictation models
The STT sidecar accepts arbitrary Hugging Face owner/model repos for
custom dictation models and, when safetensors were absent, downloaded
and loaded pytorch_model.bin through WhisperForConditionalGeneration
.from_pretrained. PyTorch checkpoints are pickles that execute code
during deserialization, and this path does not run the malware gate the
normal model loader applies, so an authenticated client on an exposed
Studio instance could load a crafted Whisper-looking repo and run code
in the backend.
Restrict custom STT repos to safetensors: the snapshot selector no
longer falls back to pytorch_model.bin(.index.json), the cached-snapshot
completeness check ignores pickle weights, and the load forces
use_safetensors so a stray cached pickle still cannot execute. The five
curated Whisper defaults already ship safetensors only, so this changes
nothing for the built-in models.
* STT: reject safetensors indexes that reference non-safetensors shards
A safetensors index (model.safetensors.index.json) is attacker-supplied
JSON and can name pytorch_model-*.bin shards in its weight_map.
Transformers dispatches shard loading per file by extension, so those
.bin shards still load through torch.load (pickle) even with
use_safetensors set. Require every weight_map value to end in
.safetensors in both the snapshot selector and the completeness check so
no pickle shard is downloaded or reused.
* Studio: add Voice settings tab (dictation, dictionary, read aloud)
New Voice tab in Settings, placed just before About:
- Dictation: microphone picker, browser STT engine, recognition language,
and an inline mic test with a live transcript
- Dictation dictionary: entries rewrite matching speech to their exact
spelling and casing, applied in both dictation paths
- Recent dictations: last 20 final transcripts with copy and clear, so
text can be recovered if it lands in the wrong place
- Read aloud: optional button on assistant responses with two engines,
curated system voices (novelty and legacy voices filtered, quality
ranked, capped at 20) or the TTS audio model loaded in Unsloth via
/audio/generate (e.g. Orpheus), plus speed, pitch, volume and preview
Settings persist in localStorage (unsloth_voice_settings) and are read
at call time so changes apply without reloading the runtime. Adds en
keys plus the tab label for ja, zh-CN and pt-BR.
* Studio: drop the single option STT engine select, rename TTS option
The STT engine dropdown only had one entry, so it added noise without
giving a real choice. The engine row can come back once local STT
models land. Also renames the TTS engine option Unsloth TTS model to
Load TTS model to make the action clearer.
* Studio: harden Voice settings against edge cases found in simulation
Simulated the feature across Chromium, Firefox and WebKit plus node
level unit runs and backend contract checks. Fixes from the findings:
- Dictionary rewrite used a replacement string, so entries containing
dollar patterns corrupted transcripts (A$$AP became A$AP, $& injected
the match). Switched to the callback form of String.replace
- Persisted voice settings now validate types on hydration: non string
micDeviceId, dictationLanguage and ttsVoiceURI, and non boolean
ttsEnabled fall back to defaults instead of flowing into the UI
- Dictionary entries are trimmed, capped at 120 chars and re-sanitized
on hydration
- The Test dictation panel now falls back to the default microphone
when the saved device is unplugged, matching the composer adapter
Test coverage: 46 unit assertions (dictionary regex edge cases across
unicode, word boundaries and injection, voice curation for simulated
macOS, Windows and Linux voice inventories, corrupt storage merge),
13 backend contract checks against /audio/generate on an isolated
instance, and 60 browser assertions across the three engines covering
rendering, degradation without SpeechRecognition, curation in a real
DOM, dictionary persistence with unicode and dollar entries, the
no-model preview error path and corrupt localStorage recovery.
* Studio: address Voice settings review feedback
Verified each review comment before acting. Confirmed and fixed:
- Editing a dictionary entry was broken in two ways: the store trimmed
on every keystroke so spaces could not be typed, and clearing the
field deleted the entry and unmounted the input mid edit. Updates now
keep the raw value and a blur commit trims or removes the entry
- The unplugged mic fallback checked instanceof DOMException, but a
cross browser probe showed Firefox and WebKit throw
OverconstrainedError objects that are not DOMExceptions, so the
fallback never fired there. Matching on the error name now
- When the browser ended a dictation test on its own (silence timeout),
the mic stream stayed open. All recognition end paths now stop the
tracks and save the transcript through a single finalize path
- The studio TTS audio element now releases its WAV data URL as soon as
playback ends, fails or is cancelled
- Allow microphone now reports insecure contexts (no mediaDevices)
accurately instead of claiming access was blocked
- Voice tab copy moved into i18n keys per src/i18n/AGENTS.md, so locale
overlays can translate it; en is the baseline and parity passes
- unsloth_voice_settings added to the Reset all local preferences key
list so voice preferences obey the reset
- Non default microphones note that the system default is used when the
browser speech engine cannot bind a specific device, since browsers
without the start(track) overload ignore the argument silently
Re-ran the full simulation set after the changes: 46 unit assertions,
13 backend contract checks and 60 browser assertions across Chromium,
Firefox and WebKit all pass, plus a dedicated browser probe for the
dictionary editing behavior.
* Studio: use the chat mic icon in Voice settings for consistency
The Voice tab and its buttons used the hugeicons Mic02 glyph while the
chat composer uses a custom filled mic. Extract that composer icon into
a shared lib/mic-icon component, drop the duplicate inline copies in
thread.tsx and shared-composer.tsx, and use it for the Voice tab icon
and the tab's mic buttons so the microphone looks the same everywhere.
* Studio: address second round of Voice settings review feedback
Verified each new comment against the current code first. One item was
already fixed in the previous round (recording transcripts when the
browser ends a dictation test on its own). Confirmed and fixed:
- The microphone row showed a picker with generic names when browsers
enumerate unlabeled devices before permission, leaving no way to
grant access from the row. It now branches on whether labels are
visible and shows Allow microphone otherwise
- Compare chat dictation ignored the selected microphone. It now opens
the chosen device with the same fallback rules as the main adapter,
passes the track to recognition where supported and releases the
stream when recognition ends
- Closing the Voice tab cancelled the shared speechSynthesis even when
read aloud was playing a chat message. Cleanup now only cancels when
the tab owns an active preview
- Double clicking Start test could race two recognizers and leak the
first stream. A starting flag set before the getUserMedia await makes
start reentrancy safe
- Turning off the read aloud setting mid playback removed the only stop
control. The stop button now renders whenever a message is speaking
- When an engine lacks the start(track) overload, both dictation paths
now release the selected device stream before retrying with the
default microphone instead of holding it open
- Read aloud support no longer requires Web Speech synthesis: the
Unsloth TTS engine only needs audio playback, so it stays available
in WebViews without speechSynthesis, with a clear error if the system
engine is chosen there
Not addressed here: cancelling in flight backend TTS generation on
stop. The route runs generation in a worker thread without a
cancellation path, which is shared pre existing behavior with audio
chat generation and belongs in a backend change.
All suites re-run green: 46 unit, 13 backend contract and 60 browser
matrix assertions across Chromium, Firefox and WebKit, plus probes for
the unlabeled device branch and the double click race.
* Studio: drop empty and duplicate voiceURIs so the Voice tab never renders a crashing Select item
* Studio: guard dictation mic lifecycle in Voice test and Compare composer
Release a microphone opened after the component unmounts, and stop Compare
dictation on a permission or security failure instead of silently recording
from the default device, matching the main chat adapter.
* Studio: fix dictation and read-aloud lifecycle edge cases in Voice settings
- Join final dictation chunks with a space so recorded transcripts do not merge words
- Ignore a stale recognizer onend so a quick stop then restart is not torn down
- Use previewingRef so a double click on TTS preview does not orphan the first request
- Keep the read-aloud stop control visible when a new run starts while a message is spoken
- Stop the dictionary remove button from deleting an adjacent entry on a blur then click race
* Studio: trim redundant Voice settings comments
* Studio: fix Voice preview and Compare dictation edge cases
- Only cancel the shared speechSynthesis for a system-voice preview, so stopping
a Studio preview no longer stops an unrelated chat read-aloud
- Release the Studio preview audio and its WAV data URL on normal completion
- Iterate every finalized result in Compare dictation so batched phrases are kept
- Cap persisted recent dictations to the last 20 on hydration
* Studio: use clipboard fallback for recents and release failed preview audio
- Copy recent dictations via the copyToClipboard helper so the execCommand
fallback works in Safari and insecure http LAN contexts
- Release the Studio preview audio when play() rejects, not just on ended/error
* Studio: add local speech-to-text dictation engine
Add an offline dictation engine that transcribes with a local faster-whisper
model, alongside the existing browser (Web Speech) engine. The browser engine
streams audio to Apple or Google speech services and needs internet; the new
engine runs on the server, works offline, and drives any chat model without
evicting it (it loads in the backend process, separate from the model
subprocess). It also gives Firefox dictation, which has no Web Speech support.
Backend: a lazily-loaded, kept-warm faster-whisper sidecar and three routes
under /api/inference/audio (stt/status, stt/load, transcribe). faster-whisper
is torch-free, so this does not disturb the existing model stack.
Frontend: a Dictation engine setting (browser or local model), a curated model
picker with sizes, and MediaRecorder capture posted to the transcribe route.
The model warms automatically when the engine is selected, with live status.
* Studio: stream local STT transcription as you speak
Local dictation showed nothing until you stopped, because the whole clip was
transcribed once on stop. Now the growing recording is re-transcribed on a
fast pass every second and emitted as live interim text, with an accurate
final pass on stop. Partial recordings decode fine, and the model refines
earlier words as more audio arrives.
Adds an interim flag to the transcribe route (beam 1, no VAD) for the fast
preview pass; the final stop uses the accurate path.
* Studio: make local dictation stop instant and reliable
Stopping local dictation waited for a final network transcription before the
session ended, so the stop button did not flip and a second click ended the
session early and dropped the text. Now stop commits the live transcript
immediately, releases the mic at once, and ignores a second stop while
finalizing. Previews run more often so the committed text is current.
* Studio: record local dictation in short clips for reliable streaming
Re-transcribing a growing buffer every second got slower as it grew, flooded
the backend, showed stale words, and could leave the stop button stuck waiting
on a backlog. Record short independent clips instead and transcribe each once,
appending the text as you speak. Work per clip is bounded, so stopping is
prompt (with a hard timeout as a safety net) and long dictations stay smooth.
* Studio: dictate then transcribe once on stop, ChatGPT style
Local STT dictation streamed by re-transcribing the growing clip, which
was quadratic and saturated the backend (multi-second lag), and stop only
halted the recorder without releasing the mic, so it kept recording. Record
the microphone continuously, release it the instant the user stops, and
transcribe the whole clip once. Stopping is immediate and the transcript
lands in about a second. Also add the tiny model for the fastest option.
* Studio: surface dictation and read-aloud failures instead of failing silently
- Compare dictation reports microphone and speech-recognition errors via toast,
reusing the main chat adapter's describeMediaError and describeSpeechError
- Read-aloud toasts genuine model or synthesis failures while ignoring cancellations
* Studio: ChatGPT-style recording bar for dictation
Clicking the mic now drops the composer into a dedicated recording bar
with a live waveform, a discard (X) and a confirm (tick), instead of a
plain stop button. The tick stops recording and transcribes the clip;
the X throws the recording away and keeps whatever text was already in
the composer. The model adapter taps the mic with an analyser to drive
the waveform, and the router tracks the live session so the X can cancel
it without transcribing.
* Studio: transcribe dictation while speaking, ChatGPT layout
Match ChatGPT's recording layout: the bar now renders in place of the
input with the left plus button kept, the waveform in the middle, and
the discard and confirm buttons together on the right.
Cut the post-confirm delay by transcribing in the background as the user
talks. The audio is split at natural pauses (voice-activity detection off
the same analyser that drives the waveform) and each clip is transcribed
as it is cut, so confirming only has to finish the short final tail. The
model is also warmed when recording starts so the first run never pays a
cold load.
* Studio: ChatGPT waveform, hide tools while dictating, faster STT
Make the recording UI read like ChatGPT: the waveform is now a dense row
of round dots that rise into thin centered bars, and while dictating only
the plus button shows, with the mode badge and tool toggles hidden so the
bar is just the waveform and controls.
Speed up transcription: decode greedily (beam_size=1), which is several
times faster on CPU with negligible accuracy loss on short dictation
clips, and cap background segments at 6s so the final tail after confirm
stays short.
* Studio: finish ChatGPT voice bar and low-latency STT
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: full-width waveform with a timer that freezes on stop
Use the full-width waveform for the recording bar: brighter, bigger bars
that advance on a fixed cadence (keeping peaks between advances) so they
glide instead of racing by, inset from the composer edges. Keep a visible
timer and the green confirm button, matching the ChatGPT reference, and
freeze the timer and waveform the moment the user confirms.
* Studio: fix multilingual local dictation
* Studio: speed up dictation and release local STT
* Studio: harden dictation finalization and STT decoding
* Studio: restore Firefox dictation fallback
* Studio: add dictation history manager
* Studio: manage speech model downloads
* Studio: remove em dash from voice model label
* Studio: move dictation history into Voice
* Studio: source local STT from Unsloth Whisper models
Point the dictation STT sidecar and its Model Hub download entries at
Unsloth's Hugging Face Whisper repos (small, large-v3-turbo, large-v3)
and run them through Transformers, so Studio only ever downloads
Unsloth-uploaded weights. Drop faster-whisper and the Systran/mobiuslabs
repos; keep the Model Hub as the only download path via local_files_only,
and keep PyAV for audio decoding.
Device selection uses float16 on CUDA and float32 on MPS and CPU, since
Whisper's decoder is unstable in float16 on MPS and repeats tokens.
Shorten the model picker labels to name plus download size and update the
STT tests for the new backend.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: smooth dictation waveform and keep pill height
* Studio: align STT model dropdown width and tidy voice copy
* Studio: guide to local engine when browser dictation is offline
* Studio: clarify voice section and STT model copy
* Studio: keep STT warm with training-aware eviction
* Harden STT lifecycle and browser compatibility
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Fix model discovery test lint
* Harden cross-browser microphone errors
* Harden cross-browser microphone errors
* Surface voice test recognition errors and fall back to Studio TTS
- Voice test now toasts non-abort speech-recognition failures instead of
ending silently, matching the main and Compare dictation paths.
- Read-aloud routes to the backend model when the runtime lacks Web Speech
synthesis (audio-only WebView), so it no longer errors immediately.
* Fix reviewed STT lifecycle races
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Fix read-aloud fallback controls
* Guard read-aloud stop when deleting a non-speaking message
aui.message().stopSpeaking() throws unless this message is the one being
read aloud, so calling it unconditionally rejected the delete handler before
the message was removed. Only stop speech when this message is speaking.
* Cap recent dictation transcript length before persisting
Recent dictations only limited entry count, so a long transcript stored the
full text in the persisted voice settings and a few could exceed the
localStorage quota, throwing synchronously from the uncaught dictation cleanup
path. Truncate each entry on save and on hydration, matching the dictionary cap.
* Studio: keep dictation mic clickable and guide to local model
Register the dictation adapter unconditionally so the mic stays enabled
for any engine and starts working right after switching to the local
model on an already-open thread.
When the browser engine cannot run (Firefox, Brave, non-secure origins),
clicking the mic shows a toast that points to the local speech-to-text
model instead of leaving a disabled button. The toast stacks its action
below the text with a fully rounded button.
* Studio: add bottom padding below the dictation guidance toast button
* Studio: increase bottom padding under the dictation toast button
* Studio: add bottom padding inside the dictation toast button
* Studio: add five Whisper defaults and custom model search
Add private UnslothAI Tiny and Base mirrors to the curated local STT choices while keeping Small as the default. Let users search or paste a Transformers-compatible Whisper repository and validate it end to end.
Keep short dictations in one clip to avoid repeated padded encoder work, then split longer recordings near Whisper's 30-second boundary.
Update hidden model filters and tests, including the CPU-only CI runtime stub for PyAV.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: use public Unsloth Whisper repositories
Point the Tiny and Base dictation defaults to the public unsloth repositories and remove the private mirror references from model filtering and tests.
* Studio: update Whisper download sizes
Reflect the cleaned public Tiny and Base repositories in the curated model labels.
* Studio: right-align STT model size, fix dropdown wheel scroll, refresh sizes
- Show the download size on the right of each model row so long names
like Whisper Large v3 Turbo no longer hide it
- Update curated Whisper sizes to the safetensors weights actually
downloaded: Tiny 151 MB, Base 290 MB, Small 967 MB
- Drive the model list scroll from a wheel handler so the mouse wheel
scrolls it inside the Settings dialog, not just the scrollbar
- Add a search icon and shorten the placeholder to Search model
* Studio: do not search when a dictation model is picked, shrink repo label
- Treat the filled-in model text as a selection, not a query, so choosing
a model no longer kicks off a Hugging Face search
- Make the repository line under each model name smaller
* Studio: tighten dictation model and local engine descriptions
* Studio: keep model display on pick instead of the query, shrink row text
- Guard the combobox input so selecting a model shows its name and does
not echo the typed query back or start a search
- Map the item label to the friendly display so picks fill the field
- Reduce the model name and size text in each row
* Studio: show only the model name in the dictation field, shrink size label
- Drop the download size from the search field; the name alone is shown
once a model is selected, with sizes kept in the dropdown list
- Reduce the size label text in each row
* Studio: clarify the dictation model description
* Studio: drop Hugging Face from the dictation model description
* Studio: move the dictation dictionary to its own Manage subpage
- Replace the inline entry list with a Manage row, matching Dictation
history, so a long dictionary no longer crowds Voice settings
- Add a DictationDictionaryView subpage that holds the entry editor
* Studio: match STT field font, use best voice for System default
- Bump the dictation model field text to text-sm so it matches the
engine dropdown next to it
- Resolve the System default read-aloud voice to the top curated voice
instead of the browser default, which is a robotic legacy voice on macOS
* Studio: rerank read-aloud voices and drop duplicate voice entries
- Rank by vendor quality, then the user's locale, then a preferred list of
natural voices, so the best voice leads instead of the first alphabetically
- Collapse voices that macOS reports twice under one name and language
* Studio: fold dictionary and recents into the dictation section
- Drop the separate Dictation dictionary and Recent dictations headings;
their Manage rows now sit under Dictation, split by the row divider
- Shorten the custom spellings description
* Studio: add search and sort to dictation history
- Filter saved dictations by text with a search field
- Sort by newest, oldest, or A to Z; show a no-matches message
- Keep Clear all available regardless of the current filter
* Studio: settle cancelled STT loads before training and fix dictation review items
Wait for a cancelled STT load to exit and release its memory before
reporting it freed for training, so the loader cannot still be inside
from_pretrained()/.to(device) holding VRAM when the training subprocess
starts. A load that finishes before observing the cancel now gets
unloaded so the memory is actually reclaimed.
Clear the accelerator cache before the CPU fallback in load() so a failed
CUDA/MPS load does not strand reserved VRAM once the sidecar is marked
CPU-resident.
Send the saved Hugging Face token when polling STT download progress so a
gated or private repo resolves and shows the correct Load/Downloaded
state instead of reporting missing.
Mark the composer Dictate button as type="button" so clicking it does not
also submit the draft when the composer already has text or attachments.
* Studio: pin dictation settings per session and close STT startup races
Capture the STT model and language when a dictation session starts and
pass them to every queued segment and the warm-up load, so changing the
model or language mid-recording no longer transcribes the same clip with
the wrong model or a model that is not downloaded.
Check the local runtime at the top of transcribe(), before the model
cache lookup and the bounded audio decode, so a server missing PyTorch or
Transformers returns 501 up front instead of decoding a long clip first.
Treat the training startup window as active for STT device selection.
start_training frees VRAM in before_spawn but only assigns _proc later, so
a concurrent STT load could take the GPU that was just cleared. A startup
flag now reports training active from the free until the process is live,
forcing those loads to CPU; a finally clears it on every exit.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: stub the STT runtime check in transcribe orchestration tests
transcribe() now verifies the local runtime up front, so the unit tests
that exercise transcription orchestration must treat the runtime as
present to keep passing where PyTorch, Transformers, and PyAV are not
installed. Stub ensure_stt_available in the shared fixture and restore
the real check in the availability and load-rejection tests.
* Harden custom Whisper dictation models
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Add whisper.cpp dictation engine with per-engine downloads and history rework
Engines
- New GGML STT sidecar that runs a managed whisper-server subprocess with
idle unload, plus a pinned static build script (scripts/build_whisper_cpp.sh)
- Dictation engine picker now offers Browser, Local transcription
(whisper.cpp), and Local transcription (Transformers)
- Both local engines serve the same five curated Whisper models and download
them directly with byte-level progress reported by /audio/stt/status
- Models auto load on selection and when their download finishes
- Unload and training admission account for both engines
Benchmarks (Apple Silicon, greedy, warm, same checkpoints)
- whisper.cpp transcribes 2.4x to 5x faster than Transformers and loads in
about 0.45s vs 0.86s for Whisper Small
- whisper.cpp GGUF path is unchanged by the Transformers addition
(load 0.445s -> 0.444s, short clip 0.391s -> 0.347s, long 1.197s -> 1.129s)
Voice settings UI
- Plain curated model select replaces the searchable combobox
- Single download progress bar with transfer rate for both engines
- Dictation history now stores every dictation with Show more pagination,
a top Clear history action, and links back to the chat it was spoken into
- Archived chats dialog gets the same pagination
- Delete dialog offers deleting a dictation together with its chat
Tests: 88 backend STT tests pass, including new snapshot download coverage.
Frontend typecheck, lint, i18n parity, and production build pass.
* Merge local engines into one option and source GGML models from unslothai
Engine selection
- The dictation engine dropdown is back to two choices: Browser and Local
transcription. The selected model decides the backend: curated ids run
GGML checkpoints through whisper.cpp, searched Hugging Face repositories
run safetensors through Transformers
- Model picker lists the curated models and searches Hugging Face for other
Whisper repositories, validating them before selection. The trigger is a
plain button so the selection never renders inside a text input
- /audio/stt/status accepts a model query param so downloaded state works
for custom repositories; the engine param on load, transcribe, and
download routes is derived from the model everywhere
Model source
- Curated GGML checkpoints now download from the Unsloth-hosted
unslothai/whisper-*-GGUF repositories (one repo per model) instead of
ggerganov/whisper.cpp; cache lookups, progress totals, and in-flight blob
tracking are per-model
Fixes
- Voice settings and dictation history were not persisting: the quota-safe
localStorage wrapper was declared after the store that uses it, so the
persist storage factory failed silently. Every settings write also threw
mid-click, which kept the model picker popover from closing on selection
- is_model_downloaded now verifies config, preprocessor config, and real
weight files instead of trusting an offline snapshot lookup, so a partial
download left by an aborted fetch shows the Download button instead of
failing to load
- Removed whisper.cpp mentions from user-facing text: the ready status
shows Loaded instead of the runtime name, picker rows show the source
repository, and runtime error messages say local transcription runtime
Verified with automated browser sessions and live API checks: selection
closes the picker with no page errors, persisted settings hydrate on
reload, a stale partial snapshot triggers download then loads on MPS and
transcribes, and curated models download from the unslothai repos. 88
backend STT tests, typecheck, lint, i18n parity, and build pass.
* Skip the duplicate source line for custom models in the STT picker
A custom repository's display name is its id, so search results and the
appended current selection rendered the same string twice. The source
line now only renders when it differs from the name; curated rows keep
their name, unslothai source repository, and download size.
* Verify every shard of a sharded checkpoint in the downloaded check
A snapshot holding one of N shards (or a corrupt shard index) passed the
downloaded check and then failed at load. When model.safetensors.index.json
exists, every shard in its weight map must now be present. Found by
simulation; covered by a regression test.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Rename stale _starting references in the pump resilience tests
The startup flag on TrainingBackend was renamed to _spawn_in_progress but
two tests added alongside it still asserted on the old name, failing the
Python 3.11 to 3.13 CI jobs.
* Make the selected model row clearly highlighted in the STT picker
The current selection was a faint background tint. It now uses the accent
background with a medium weight name. Two line rows use a small corner
radius; single line custom repo rows keep the pill shape.
* Address review feedback on STT snapshot checks, VRAM release, and dictation UX
Verify snapshot completeness in the load preflight so a partial download
fails before the audio is decoded, for curated and custom repos alike.
Drop the failed accelerator traceback before the CPU retry so the cache
clear can actually release that memory. Keep unloading the GGUF sidecar
after cancelling an in-flight Transformers load; both engines can hold
memory at once. Allow Auto language with English-only .en checkpoints,
matching the backend which sends no forced language. Keep the discard
button usable while a transcription is pending so a slow or hung request
cannot trap the composer in dictation mode. Stop linking Compare and
settings test dictations to the unrelated active single chat thread.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Move the CPU retry out of the exception handler
On Python 3.10 the interpreter exception state keeps its own reference
to the traceback, so dropping it from the caught exception was not
enough to release the failed accelerator load during the retry. Leaving
the handler before clearing the cache works on every supported version.
* Address review feedback on session handoff, chat pinning, and server lifetime
Starting a dictation from a second entry point now cancels the session
it replaces, so the old recording cannot keep the microphone open or
save a transcript with no discard button pointing at it. The linked
chat is pinned when recording starts, so switching threads while a
transcription finalizes cannot relink the transcript to the newly
opened chat. whisper-server is now bound to Studio's lifetime like the
other long-lived children: PDEATHSIG on Linux, the parent job object on
Windows, and pid adoption so the shutdown sweep reaps it; before this
it survived a Ctrl+C exit as an orphan still holding the model.
* Remove the dictation mic test from Voice settings
The composer dictate button covers the same check, so the test row, its
transcript panel, the unsupported fallback row, and their strings and
search entry are gone.
* Studio STT: gate GGUF whisper-server on training and fix dictation retry and dictionary edits
GGUF (whisper.cpp) sidecar:
- Launch whisper-server with --no-gpu while training is active, mirroring the Transformers sidecar's CPU device choice, so a mid-training dictation cannot reclaim the VRAM training just freed.
- Report is_loading() during whisper-server startup so training VRAM admission accounts for the accelerator memory it is about to bind.
- Require PyAV in is_available() so /audio/stt/status reports the engine unavailable when uploads cannot be decoded, instead of loading fine and then 501ing at transcription.
- Reject a missing model before decoding audio, matching the Transformers download preflight.
Voice settings:
- The download Retry button now restarts the download; the sidecar error is sticky until a new start(), so re-polling alone never cleared it.
Dictation dictionary:
- Tabbing from an emptied entry to its remove button no longer commit-splices the row first, which shifted indices and deleted the wrong entry.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio STT: fix curated GGUF whisper filenames to match hosted repos
The unslothai/whisper-*-GGUF repos host the checkpoint as whisper-<id>.bin,
not ggml-<id>.bin, so every curated dictation download and cached-path
lookup 404'd and the whisper.cpp engine could never load a model. Point
GGML_STT_MODELS at the real filenames and guard the naming with a test.
* Studio STT: validate a custom dictation repo before downloading it
The Transformers STT engine accepts an arbitrary owner/model repo, but the
download route handed it straight to snapshot_download, pulling a possibly large
non-Whisper repository into the shared HF cache. Confirm the repo is a Whisper
checkpoint first with the existing metadata-only validate_remote_model (no
weights); curated ids short-circuit and the GGUF engine (curated-only) is
unaffected. A non-Whisper repo now 422s before any download.
* Studio STT: preempt a still-loading GGUF server for training admission
A whisper-server still in its startup window binds accelerator memory but has no
loaded_model yet, so training admission could miss it and launch into an OOM.
Make the GGUF startup cancellable (cancel_pending_load signals an abort event and
terminates the starting process without the load lock; _wait_for_server observes
it and raises SttLoadCancelledError; wait_for_load_to_settle blocks on the lock
until the killed server is reaped), and always fold the GGUF sidecar into the
resident-STT summary so a resident Transformers model cannot mask a loading GGUF
server. free_stt_model_for_training now cancels an in-flight load and waits for it
to settle before training claims the memory.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio STT: fall back to Transformers when whisper-server is absent
A curated dictation model (including the default small) hard-pinned the GGUF
engine, but standard installs do not ship whisper-server, so every recording
501'd instead of using the Transformers engine that serves the same checkpoint
-- the GGUF sidecar's own documented contract. Add _resolve_serving_stt_engine:
a GGUF request for a curated id (the only ids GGUF accepts, all Transformers-
servable) downgrades to Transformers when whisper-server is unavailable, applied
consistently to download, load and transcribe (not unload, which targets a
specific engine). The Voice tab likewise falls back to the Transformers status so
the model is not shown unavailable and download is not blocked.
* Studio STT: hide custom Whisper caches from the legacy model pickers
The legacy /cached-models (and /cached-gguf) routes called is_hidden_model with
only the owner/model id, which cannot reach the config-based Whisper check, so a
downloaded custom (non-curated) Whisper checkpoint was still offered as a chat
model. Pass the cached snapshot path so _path_is_whisper_model inspects the repo
config and hides it, matching the discovery route.
* Studio STT: hide GGUF dictation repos, lock-free status, unload fallback, split training eviction
- Hide the curated GGUF dictation repos (unslothai/whisper-*-GGUF) from the chat
model inventory and pickers, backend and frontend. Only their Transformers
safetensors companions were hidden; the GGUF repos use a different org and a
-GGUF suffix and carry a raw .bin with no whisper config.json, so they leaked
into chat pickers.
- Make the GGUF sidecar loaded_model/device accessors lock-free, mirroring the
Transformers sidecar. transcribe() holds self._lock across the whole inference
call, so /audio/stt status polls and training admission previously blocked
behind an in-flight transcription.
- stt_unload resolves through the serving resolver: a "gguf" pick on a host
without whisper-server is served by the Transformers fallback, so unload must
target that engine or the resident model is never freed. Unload also attempts
every engine even if one raises, so a failure freeing one backend no longer
skips the other.
- free_stt_model_for_training frees the Transformers and GGUF sidecars under
independent exception boundaries so a failure unloading one no longer skips
the other before training claims the memory.
Adds tests/test_stt_review_fixes.py covering all four.
* Studio STT: resolve Auto dictation language for the model engine + snapshot process liveness
- The model dictation adapter sent the raw setting (the literal "auto") to the
backend, while the browser engine resolves Auto via resolveDictationLanguage.
A batch of non-English voice notes came back mostly English on Auto. Add
resolveModelDictationLanguage: only the literal "auto" is resolved to a
concrete locale, gated so it becomes a language the model AND Whisper can
honor (mirroring the backend's known-whisper-languages set); an explicit
language, or a locale Whisper cannot honor, stays unchanged/auto-detect. Wire
it into both adapter call sites.
- GgmlSttSidecar._process_alive() read self._process twice; a concurrent
unload() nulls it under the lock while loaded_model/device read lock-free, so
a null between the two reads called None.poll(). Snapshot once. Adds a
deterministic regression test.
* studio: tighten comments and docstrings in the dictation modules
* studio: harden dictation model downloads, GGML readiness, and recording paths
Address review findings on the STT dictation feature:
- build_whisper_cpp.sh refuses to delete a whisper.cpp tree under a custom
Studio home unless it carries the Studio ownership marker, matching the
setup.sh policy, and marks trees it creates
- _snapshot_is_complete validates every shard of a sharded PyTorch
(pytorch_model.bin.index.json) checkpoint like the safetensors path, and
requires tokenizer assets (tokenizer.json or vocab.json + merges.txt)
- custom-repo downloads pin the revision resolved at validation time and
restrict snapshot_download to the model/tokenizer/config/preprocessor file
classes Studio loads
- the GGML sidecar holds its port reservation until just before spawning
whisper-server and only accepts readiness from a responder that both looks
like whisper.cpp's server and belongs to the still-running managed child,
probing twice, so mic audio cannot be posted to a foreign local process
- the recording adapter transcribes every non-empty segment; the RMS meter
only shapes segment boundaries and can no longer discard quiet speech
- Compare-pane dictation can cancel a pending transcription on second click,
with the button relabeled while finalizing
- localStorage quota recovery halves the dictation history until the save
fits, so small histories shrink too
- the System default TTS voice resolves to the platform default voice
- new dictation UI imports go through the chat and hub feature barrels
Regression tests cover the build-script gate, sharded PyTorch and tokenizer
completeness, revision pinning and allow patterns, and the whisper-server
readiness probe.
* Fix STT download and voice picker follow-ups
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Add dictation button regression coverage
* Studio: prebuilt whisper.cpp via the shared llama.cpp install core, slim bundles paired to the llama prebuilt (#7294)
* Studio STT: add prebuilt whisper.cpp (whisper-server) installer
New install_whisper_prebuilt.py downloads a per-platform whisper-server
bundle published by the unslothai/whisper.cpp prebuilt CI into the managed
whisper.cpp dir (build/bin/whisper-server) so local dictation needs no
compiler. Mirrors install_node_prebuilt.py / install_llama_prebuilt.py:
host + backend detection, sha256 pins (whisper_prebuilt_pins.json) as the
trust anchor, staging + install lock + atomic swap, traversal-safe extract,
co-located shared libs (RUNPATH=$ORIGIN), an UNSLOTH_WHISPER_PREBUILT_INFO.json
marker with idempotent "already matches", and exit codes 0/1/2/3. Not wired
into setup yet; the pins ship empty so every asset fails closed until the
first fork release is published and its digests are reviewed in.
* Studio STT: install prebuilt whisper.cpp during setup and update
Add a fail-open whisper.cpp block to setup.sh after the llama.cpp section so
`unsloth studio update` (and a fresh install) fetch the prebuilt whisper-server
into the managed whisper.cpp dir the sidecar discovers. It skips a user-set
WHISPER_SERVER_PATH/UNSLOTH_WHISPER_CPP_PATH, honors UNSLOTH_SKIP_WHISPER_INSTALL,
forwards the resolved ROCm gfx, and never aborts setup: a busy install keeps the
existing runtime, and an unavailable prebuilt stays quiet (source build is opt-in
via UNSLOTH_WHISPER_FORCE_COMPILE) since Transformers STT and browser dictation
remain. Register UNSLOTH_WHISPER_PREBUILT_INFO.json as Studio-owned evidence.
* Studio STT: harden whisper-server child env + WSL ROCm detection
- Sidecar spawns whisper-server with a scrubbed child env that prepends the
binary dir (co-located GPU libs) to the loader path, and on WSL2 ROCm loads
the system HIP first (HSA_ENABLE_DXG_DETECTION=1) so a bundle's bare-metal HIP
does not segfault on /dev/dxg. Secret-bearing vars are dropped from the child.
- find_whisper_server_binary now requires an executable, not just a file.
- Installer rocm probe passes HSA_ENABLE_DXG_DETECTION and falls back to
/opt/rocm/bin/rocminfo so a WSL ROCm host is not misdetected as CPU-only;
gfx parsing skips the gfx000 CPU agent and generic ISA lines.
- Tests for the child env (secret scrub, lib dir, WSL HIP precedence), the
executable check, and the WSL rocm detection.
* Studio STT: in-app whisper.cpp prebuilt update stack + ship pins in the wheel
Mirror the llama.cpp update stack for the whisper.cpp prebuilt so Studio can
detect and install a newer whisper-server release from inside the app:
- backend/utils/whisper_cpp_freshness.py: read UNSLOTH_WHISPER_PREBUILT_INFO.json
and compare the installed release against the newest unslothai/whisper.cpp
release. Whisper tags are v<upstream>-unsloth.<N>, so is_behind compares a
(major, minor, patch, serial) key with a strict downgrade guard; 24h cache;
fail-open.
- backend/utils/whisper_cpp_update.py: run install_whisper_prebuilt.py to fetch
and atomically swap the newest bundle, unloading the warm GGUF sidecar first.
- backend/routes/whisper.py mounted at /api/whisper (update-status + update).
- pyproject: add whisper_prebuilt_pins.json to studio package-data so the
installer's trust anchor ships in the wheel (it is a data file, not a .py
module, so package discovery alone does not include it; node_prebuilt_pins.json
is listed for the same reason). Without this a pip-installed wheel had no pins
and the prebuilt install aborted to Transformers STT.
Adds test_whisper_cpp_freshness.py (version parser, is_behind matrix + downgrade
guard, marker layouts, stale decision, fail-open).
* Studio STT: verify whisper prebuilts via the release checksum index, like llama.cpp
Re-align the whisper.cpp prebuilt installer to install_llama_prebuilt.py's trust
model: instead of a committed whisper_prebuilt_pins.json, verify every download
against the release's own whisper-prebuilt-sha256.json checksum index, fetched
from the same GitHub release.
- parse_release_checksums / fetch_release_checksums / expected_sha256_for replace
the pins layer. The index is validated for schema/component and that its
release_tag matches the resolved release; an asset absent from it, a release
that does not publish it, or a manifest sha256 that disagrees with it all fail
closed to a source build.
- resolve_release_tag now resolves the newest published release at runtime (or an
explicit --published-release-tag), matching llama and the freshness check;
removed the pinned-default and the UNSLOTH_WHISPER_ALLOW_UNVERIFIED opt-in.
- Delete studio/whisper_prebuilt_pins.json and drop its pyproject package-data
entry (nothing to ship now, same as llama which has no committed pins).
- Adds test_install_whisper_prebuilt_checksums.py (index parser, fail-closed on
uncovered asset, tampered-manifest guard, newest-release resolution).
This is a same-origin checksum (integrity, not authenticity), identical to the
llama.cpp installer; pair releases with GitHub artifact attestations for provenance.
* Resolve whisper prebuilt release via the download host (no GitHub API)
Mirror install_llama_prebuilt.py's fast path: resolve the release tag from
the releases/latest redirect and fetch the manifest + checksum index from
constructed releases/download URLs, so the common install path makes zero
api.github.com calls (unauthenticated api.github.com is capped at 60 req/hour
per IP; the download host is not). Fall back to the GitHub API only on a 404,
malformed asset, or tag mismatch.
* Studio STT: coverage-aware whisper prebuilt selection via a shared core
whisper's select_artifact returned the first os/arch/backend manifest match and
ignored the SM-coverage fields the release manifest already carries, so a
Blackwell B200 (sm_100) was served cuda12-legacy (sms 50-61) -- runnable only via
forward PTX JIT. install_llama_prebuilt.py on the same host correctly picks
cuda13-newer.
Extract the coverage-aware selection into a shared, component-agnostic core under
studio/backend/utils/prebuilt/ (selection + GPU host-capability detection), lifted
from llama's linux_cuda_choice_from_release / _artifact_covers_sms / _sm_range and
generalised over a normalised artifact. whisper's HostInfo now records the GPU
compute caps + driver CUDA version (honoring CUDA_VISIBLE_DEVICES), and
select_artifact routes CUDA/ROCm through the shared selector: every visible SM
must be covered, the tightest-covering profile wins (Blackwell-aware runtime-line
ordering), ROCm matches the gfx target exactly, and an uncovered GPU falls back to
the CPU bundle. CPU/Metal/Vulkan keep first-match. The resolver JSON, exit codes,
and "already matches" contract are unchanged.
On the B200 the installer now resolves cuda13-newer, matching llama.
* Studio STT: gate whisper CUDA selection on the on-disk runtime, like llama
The prebuilt CUDA bundles are dynamically linked and intentionally do NOT ship
libcudart/libcublas -- they load the same runtime the host already has. So the
driver's advertised CUDA version is only an upper bound: a cuda13 bundle still
needs cuda13 runtime libraries present on disk. Port llama's on-disk runtime
scan (detected_linux_runtime_lines / detected_windows_runtime_lines) into the
shared core and intersect it with the driver-compatible lines in
select_cuda_attempts. A host with a cuda13 driver but only cuda12 runtime (e.g.
torch-cuda12) now correctly gets a cuda12 bundle instead of an unloadable cuda13
one; a host with no CUDA runtime at all falls back to CPU.
Fixes a glob bug in the port (any(Path(d).glob(p) for d in dirs) tests generator
truthiness, not a match) that made every major report present; add a real
filesystem test that exercises the scan.
* studio: harden shared prebuilt core to full llama parity
Apply the review findings on the shared coverage-aware prebuilt-consumer
core so whisper.cpp selection is exactly equivalent to the llama.cpp path.
hosts.py: port llama's CUDA_VISIBLE_DEVICES handling. A GPU hidden by an
index/UUID selector now reports has_usable_nvidia False instead of staying
usable, via supports_explicit_visible_device_matching plus the physical /
explicit-match branches, and _select_visible_rows now matches rows the way
llama does (index or UUID, gpu- prefix optional) and skips unmatched tokens
rather than keeping all rows. Adds the Linux /proc/driver/nvidia/gpus
fallback and has_physical_nvidia. Adds parse_macos_version.
runtime_libs.py: the Linux on-disk scan now requires the exact libcudart /
libcublas SONAME (libcudart.so.13), not a libcudart.so.13* glob, so a bare
versioned file without the SONAME symlink no longer counts as loadable.
Hardens the ldconfig parse against an empty left-hand side.
selection.py: fix the Blackwell/torch reordering so it keys on the covering
runtime lines (falls through to the torch preference when the covering lines
were filtered out), matching linux_cuda_choice_from_release. Corrects the
compatible_runtime_lines_for_driver docstring: the bundles do not ship the
CUDA runtime, so the driver version is only an upper bound and the caller
must intersect with the on-disk scan.
install_whisper_prebuilt.py: enforce a macOS artifact's min_os (new
HostInfo.macos_version) so a bundle that cannot load on the host OS version
is dropped. Keep resolver stdout to only the JSON line by leaving logs on
stderr in --resolve-prebuilt mode, and map an unexpected probe failure to
prebuilt_available False instead of a traceback.
Tests: new host-probe suite for the visible-device logic, exact-SONAME
runtime-scan cases, macOS min_os filtering, resolver stdout-only-JSON,
exit-code mapping, and the repo key.
* studio: fix whisper prebuilt selection + launch parity gaps from review
A parallel review surfaced integration defects where the whisper path could
select or launch a bundle that cannot run on a concrete host. Each is fixed to
match install_llama_prebuilt.py.
macOS min_os: the manifest labels macOS requirements as macos-<version>
(e.g. macos-14.0), which the version parser could not read, so the guard was a
no-op and a macOS-13 host would install the macos-14 Metal bundle. Strip the
platform prefix before parsing.
ROCm gfx detection: _detect_rocm_gfx returned the first gfx token and ignored
HIP_VISIBLE_DEVICES / ROCR_VISIBLE_DEVICES / CUDA_VISIBLE_DEVICES. Since exact
ROCm matching treats that token as the active GPU, a mixed APU + dGPU host
(gfx1151 + gfx1100) with HIP_VISIBLE_DEVICES=1 installed the wrong archive. Route
through a shared pick_rocm_gfx_target (lifted from llama) that parses per-GPU
sections and honors the visibility vars (empty / -1 -> no AMD GPU).
--rocm-gfx override: recording the arch without setting has_rocm left the host on
its CUDA/CPU path so the ROCm bundle was never picked. --rocm-gfx now implies
has_rocm and clears NVIDIA state, like llama's _apply_host_overrides.
CUDA launch env: a CUDA bundle ships the ggml CUDA backend but not
libcudart/libcublas, and the sidecar launch env exposed only the bundle dir, so
on a host whose CUDA runtime lives only in the PyTorch wheels the selection would
gate cuda usable but the server could not load it. Add the CUDA-from-PyTorch
runtime dirs to the child loader path for CUDA bundles (bundle dir still first),
mirroring binary_env.
Also normalize a manifest artifact's supported_sms defensively (parity with
llama's parser) and document that blackwell_min_toolkit_for_caps is retained for
the Phase B llama Windows path.
Not changed (verified parity, not defects): Linux/Windows min_os is enforced
nowhere in llama (macOS only); the resolver is optimistic about the checksum
index and the install path verifies.
* studio: tighten prebuilt-core code comments
* studio: lift shared prebuilt installer core out of the whisper installer
* studio: reuse the llama.cpp prebuilt installer machinery for whisper
* studio: unify llama and whisper prebuilt installers on a shared descriptor core
* studio: consolidate prebuilt installer tests into the shared core suite
Grow tests/studio/install/test_prebuilt_core.py from 62 to 164 tests so every
component-agnostic behavior runs against both descriptors: the full seven
profile CUDA release matrix (multi-GPU, on-disk runtime gating, shuffle
stability, missing SM metadata, dotted SM normalization, no-driver fallback
policy), the ROCm gfx family matrix, macOS min_os gating and its helper,
backend resolution incl. cpu-fallback precedence and Intel-mac auto detect,
checksum-index non-object and plain-lookup cases, the tar symlink/hardlink
extraction guards moved from the llama suite, and the compute-cap, visible
device, runtime-line and Blackwell helper value tables moved verbatim from
the llama characterization suites.
Delete only tests whose exact behavior the master now asserts for the same
component: 40 pure-alias helper cases in test_selection_logic.py (replaced by
value-identical master tables plus an alias-identity pin), 6 extraction moves
and the master-absorbed zip-symlink case in the llama logic suite, 3 routing
twins in test_rocm_support.py already pinned byte-for-byte in
test_selection_logic.py, the 2 Blackwell helper tables in the backend resolve
suite, 28 whisper logic tests and 10 whisper checksum tests re-asserted by
the master whisper parameterization. Wrapper wiring pins, the llama release
plan dialect, fingerprints and every llama-only behavior stay untouched.
* studio: dedupe sidecar and update helpers into the backend prebuilt package
* studio: chain whisper.cpp prebuilt updates onto the llama.cpp update flow
* studio: consume paired slim whisper prebuilts via the llama ggml runtime
* studio: serve every whisper backend from slim prebuilts
* studio: drop the whisper fat per-accelerator selection chain
unslothai/whisper.cpp releases are slim-only from v1.9.1-unsloth.2: one
ggml-less bundle per os/arch, paired to the llama.cpp prebuilt that provides
every ggml backend. Delete the whisper-side fat CUDA/ROCm/metal/vulkan
selection glue; keep slim selection + pairing, link_ggml_runtime, and one
legacy shape, the published fat CPU bundle of an explicitly pinned pre-slim
release. Exit 2 now reads as prebuilt unavailable (whisper never source
builds); setup already treats it that way.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Wire libomp runtime DLL alongside ggml in slim whisper installs
llama's clang-built windows-arm64 ggml-base.dll imports
libomp140.aarch64.dll, shipped in the llama bundle but not a system DLL.
Without it next to whisper-server.exe the loader fails with
STATUS_DLL_NOT_FOUND before main. MSVC x64 links vcomp140.dll from
System32 and Linux ggml uses system libgomp.so.1, so only windows-arm64
was affected. The empty-runtime guard still requires a real ggml
library; libomp alone is not a pairing.
* studio: drop whisper-side fat-selection support structure
Slim whisper bundles are selected per os/arch only; all accelerator
capability comes from the installed llama.cpp prebuilt, whose installer
already did the coverage-aware selection. Remove the machinery that only
existed to pick among fat per-accelerator whisper bundles:
- prebuilt_core: delete the generic CUDA/ROCm coverage selection
(select_cuda_artifact, select_rocm_artifact, ArtifactView adapters,
detected_cuda_runtime_lines, the exact-SONAME linux probe) that no
shipped component routes through; llama keeps its own selection chain
and whisper shadows select_artifact with the slim-only version.
select_artifact is now a plain os/arch/backend first-match.
- install_whisper_prebuilt: drop the HostInfo CUDA fields
(compute_caps, driver_cuda_version, torch_runtime_line) and the torch
runtime probe that populated them; nothing reachable reads them, and
the resolver payload sources runtime_line from the artifact.
- whisper_cpp_update: delete the standalone start_update job worker;
whisper applies only run as the chained phase of the combined
llama+whisper update. The status payload keeps its job field (idle).
- routes/whisper: drop the progress logger that could never fire.
- tests: remove tests of the deleted paths and tests duplicating the
descriptor-parameterized core suite or the llama freshness suite.
Contracts unchanged: resolver JSON keys, exit codes, marker fields,
pairing logs, and the pinned pre-slim fat CPU escape hatch.
* Address review feedback on the whisper prebuilt update and install paths
- Pin the chained whisper phase to the release the freshness check
offered, so the download-host latest pointer cannot reinstall an
older build in a loop
- Wire the whisper prebuilt install into setup.ps1 (Windows setup
previously skipped it entirely)
- Treat a non-executable server or missing wired ggml libraries as a
broken install instead of reporting already matches
- Keep whisper sidecar reloads out of the job-level reload flag and
resync chat state after a partial chained update that unloaded llama
- Repoint home and profile vars for the whisper-server subprocess at a
managed scratch dir and drop credential-store pointers
- Clear the prebuilt marker before the opt-in source build overwrite
- Write the prebuilt marker with explicit utf-8 encoding
* Tighten comments in the whisper prebuilt consumer
* Harden the Windows whisper setup phase and the chained update edges
- setup.ps1: honor WHISPER_SERVER_PATH / UNSLOTH_WHISPER_CPP_PATH /
UNSLOTH_SKIP_WHISPER_INSTALL, run the custom-home ownership guard
before the atomic install, and forward the release-tag pin and ROCm
hints like setup.sh
- sidecar: a cpu-selected install launches whisper-server with --no-gpu
(slim wiring links every llama backend, so the flag is what keeps a
deliberate CPU choice off the GPU)
- chained update: leave whisper unpinned on macOS (the llama phase can
walk back there, and a newest-tag pin could be an impossible pairing
on every retry) and treat installer exit 2 as kept-existing-runtime
instead of failing the combined job
- job.to_tag now comes only from the llama phase, so a whisper-only
round cannot report a llama update that never ran
* Fix slim whisper runtime follow-ups
* Address remaining whisper update reviews
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Address remaining prebuilt update reviews
* Fix remaining chained update reviews
* Fix remaining whisper runtime review edges
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
---------
Co-authored-by: danielhanchen <unslothai@gmail.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: oobabooga <112222186+oobabooga@users.noreply.github.com>
---------
Co-authored-by: danielhanchen <danielhanchen@gmail.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Unsloth <michaelhan@Michaels-MacBook-Pro.local>
Co-authored-by: oobabooga <112222186+oobabooga@users.noreply.github.com>
Adds a configurable Hugging Face model download cache location to Unsloth Studio, selectable from Settings, with per-cache download manifests, scoped deletion, and read-only inventory of previously selected caches.
* Studio: mask AMD GPU pins via ROCR so an unsupported iGPU can't crash llama-server
On a mixed AMD host (e.g. a discrete gfx1102 GPU next to a gfx1103 iGPU)
the bundled rocm-gfx110X llama.cpp build segfaults during HSA device
enumeration on the unsupported iGPU -- before llama-server prints a line,
so every model load fails with a bare signal and empty logs.
The GPU-subset pin masked visibility with HIP_VISIBLE_DEVICES, but HIP
filtering runs only after the HSA runtime has already enumerated (and
crashed on) every agent. Mask the subset via ROCR_VISIBLE_DEVICES (the
ROCr/HSA layer) instead, so a deselected/unsupported GPU is never
enumerated. Exactly one layer is masked (HIP cleared) to avoid the
double-mask reindex that would otherwise drop the child to CPU. The
whole-set tensor-split path and the CPU-only sentinel keep their existing
HIP behavior.
Also stop misreporting the resulting startup segfault as a vision
projector incompatibility: when the text-only mmproj retry also hard-
crashes with a signal, surface a GPU/driver init crash (with the ROCR
hint) instead of blaming the projector.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Tighten _emit_child_gpu_visibility comments for #7272
Comment and docstring only: condense the ROCR-vs-HIP masking rationale from ~22 to ~14 lines and the call-site note from 5 to 3, keeping every technical point (HSA enumeration segfault, physical ids, the -1 sentinel). Logic is unchanged, verified by an AST compare with docstrings stripped and by exercising _emit_child_gpu_visibility against a torch/HIP stub.
* Detect AMD SDK ROCm wheels (hip=None) in _emit_child_gpu_visibility (Codex P2)
The ROCm branch gated only on torch.version.hip, but AMD SDK wheels leave that unset while encoding 'rocm' in __version__ (detect_hardware handles this the same way). On such a wheel the masking was skipped entirely, leaving only CUDA_VISIBLE_DEVICES, so on a mixed AMD box the unsupported deselected iGPU still enumerated and could crash llama-server. Now the branch also treats 'rocm' in torch.__version__ as ROCm, mirroring detect_hardware. Adds tests for the hip=None SDK wheel (ROCR + default paths) and a CUDA guard so the version-string check can't false-positive.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Remap CUDA_VISIBLE_DEVICES to post-ROCR ordinals on prefer_rocr for PR #7272 (Codex P1)
On the prefer_rocr path _emit_child_gpu_visibility set ROCR_VISIBLE_DEVICES to the
physical id and cleared HIP_VISIBLE_DEVICES, but left CUDA_VISIBLE_DEVICES at the
physical id. ROCR re-indexes the visible agents from 0 and, with HIP cleared, HIP
honours CUDA_VISIBLE_DEVICES -- so a non-zero pick (e.g. GPU 1) pointed out of
range, HIP saw 0 devices, and the child fell back to CPU, defeating GPU-picker
selections other than physical GPU 0. Remap CUDA to the post-ROCR ordinals
(0..N-1); GPU 0 is unchanged, the default (HIP) path and the CPU sentinel are
untouched, and non-AMD wheels never enter this branch.
* Detect AMD SDK wheels in _resolve_visible_physical_ids for PR #7272 (Codex P2)
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Keep the HIP mask on Windows ROCm in prefer_rocr for PR #7272 (Codex P2)
* Ignore ROCR_VISIBLE_DEVICES in _resolve_visible_physical_ids on Windows for PR #7272 (Codex P2)
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Preserve inherited ROCR masks in the tensor-split pin for PR #7272 (Codex P2)
---------
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Daniel Han <danielhanchen@gmail.com>
Co-authored-by: Leo Borcherding <borchborchmail@gmail.com>
* studio: classify embedding models from the HF cache and honor offline mode
is_embedding_model() went straight to huggingface_hub.model_info() for any repo
id, so in offline mode (no DNS, or HF_HUB_OFFLINE set) selecting an
already-downloaded model hung on network retries that could never succeed and
training/export never started (#6817).
Check the local HF cache first: a sentence-transformers repo carries
modules.json in its snapshot (the same marker used for local paths), so a cached
model is classified with no network call. When HF_HUB_OFFLINE / TRANSFORMERS_OFFLINE
is set, anything not positively an embedding model returns False without a
network call instead of retrying a doomed request. Online, uncached lookups still
fall through to model_info(), so tag-only embedding models (feature-extraction)
are unaffected.
Adds _embedding_marker_in_hf_cache() over the existing _iter_hf_cache_snapshots.
* studio: judge the active cached revision, harden the cache probe, stop stub leaks
Three review fixes on the cache-first embedding detection:
1. Prefer the revision refs/main resolves to. The HF cache keeps snapshots of
older revisions, so an any-snapshot scan could classify a repo by a stale
revision -- e.g. a repo that used to be a sentence-transformers model would
short-circuit even the online lookup. When refs/main is recorded, only its
snapshot is consulted; the newest-first scan remains the fallback for caches
with no ref.
2. Keep the cache probe inside the detection error boundary. The snapshot
iterator stat()s entries and could raise if a cached model is deleted
concurrently, propagating a 500 out of the config/check-embedding routes.
_embedding_marker_in_hf_cache now catches everything and reads as
not-cached, so callers keep their normal Hub/offline fallback.
3. Stub loggers/structlog in the test only when the real modules are absent
(try-import, mirroring test_windows_gpu_detection_mock), so collecting this
file first can no longer shadow the real packages for later tests in the
same pytest process.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* studio: treat a missing active-ref snapshot as a cache miss, don't cache offline misses
Two review fixes on the cache-first embedding detection:
1. When refs/main is recorded but points at a commit whose snapshot dir is
absent (partial download / cache pruning), the recorded ref is still
authoritative: return None (cache miss) instead of falling through to scan
older snapshots, which could report a stale historical revision's
modules.json as the active one -- the same stale-cache class this helper
avoids.
2. Do not cache the offline negative. When HF_HUB_OFFLINE/TRANSFORMERS_OFFLINE
is set and the repo is not positively an ST model from modules.json,
is_embedding_model stored False under the (model_name, hf_token) key shared
with online lookups; after the env var cleared in the same process, a
tag-only (feature-extraction) embedder returned the cached False and never
reached model_info(). The offline negative is now returned without caching.
* studio: defer online embedding detection to the Hub, re-probe offline
The local modules.json marker short-circuited is_embedding_model() even
online, so a repo that dropped (or added) the marker since it was cached
was judged by its stale local revision instead of the current remote one.
Online now treats model_info() as authoritative and uses the cache marker
only as an uncached fallback when the Hub is unreachable, so a transient
failure never poisons the memo. Offline re-probes the marker on every call
without consulting or populating the memo, so a model downloaded later in
the session (or a cached online negative that predates the download) is
detected. _embedding_marker_in_hf_cache() now treats an unreadable refs/main
(a non-FileNotFoundError OSError) as a cache miss rather than scanning stale
history -- only a genuinely missing ref enables the fallback scan.
* studio: harden offline embedding detection against empty refs, offline flips, and cache casing
- _embedding_marker_in_hf_cache: an existing-but-empty/whitespace refs/main
(a partial write or in-progress truncate-and-rewrite) now reads as a cache
miss (None) instead of falling through to scan stale snapshots; only a
genuinely missing ref enables the historical scan.
- is_embedding_model: while offline, retain a positive already confirmed online
this session (model_info only ever memoizes Hub-derived results), so
_hf_offline_if_dns_dead() flipping the process to offline mid-load can't
downgrade a verified tag-only embedder to False. Cached negatives are still
bypassed and re-probed.
- resolve_cached_repo_casing + settings route: persist the embedding model in
the casing its local HF cache dir uses. Validation accepts a case-insensitive
cache hit, but an offline SentenceTransformer load resolves the cache by exact
case, so storing the requested spelling (baai/bge-m3 vs models--BAAI--bge-m3)
made the model fail to load on a case-sensitive filesystem.
* studio: reuse the exact-match-first case resolver and preserve the default
Replace the ad-hoc resolve_cached_repo_casing with the existing
resolve_cached_repo_id_case, which already prefers the exact-case cache dir
before any case variant and tie-breaks variants deterministically -- so an
exact requested id is never rewritten to a differently cased directory just
because iterdir() happened to yield it first.
Skip the normalization entirely when the submitted model equals the default:
rewriting its casing would make set_rag_embedding_model()'s exact-string
default comparison treat it as a custom override, pinning it so later changes
to the configured default stop taking effect.
* studio: don't let a stale cache marker mask a permanent Hub error
is_embedding_model's Hub-failure fallback consulted the local modules.json
marker for ANY model_info() exception, so a permanent error -- a deleted repo,
a gated repo without credentials, or a typo that matches stale cache casing --
could pass online validation on a stale marker instead of returning the
documented 409, and the persisted model could then fail when the loader
refreshes from the Hub. Classify permanent Hub errors (RepositoryNotFound,
GatedRepo, RevisionNotFound, EntryNotFound) as False, matching the nearby
GGUF/vision detectors, and reserve the cache fallback for transient/5xx failures.
* studio: honor TRANSFORMERS_OFFLINE in the embedding preflight, skip casing for local paths
- The embedding-model save reached the offline-aware is_embedding_model() only
after two preflight helpers made direct huggingface_hub calls that honor just
HF_HUB_OFFLINE: _st_module_subdirs() downloads modules.json and the security
scan fetches Hub metadata twice. In a TRANSFORMERS_OFFLINE-only session those
blocked on network timeouts before the offline return, so saving an already
cached model stalled. Both now consult a canonical hf_env_offline() helper --
the download passes local_files_only, and the metadata-only security scan
short-circuits to its documented fail-open instead of burning both timeouts.
- Skip cache-casing normalization for local paths: a relative directory such as
"org/model" is loaded from disk, so rewriting it to a case-insensitive HF
cache collision ("Org/model") would stop resolving to that directory and be
read as a Hub repo id instead.
* studio: never skip the security scan on TRANSFORMERS_OFFLINE alone
The previous commit skipped the Hub security scan whenever either offline flag
was set, but huggingface_hub honors only HF_HUB_OFFLINE: under a
TRANSFORMERS_OFFLINE-only session the later SentenceTransformer load still
reaches the network, so the scan was being skipped while the repo's pickle could
still be downloaded and deserialized -- waving through exactly what
_guard_model_security exists to block.
Split the flags: hf_hub_offline() (HF_HUB_OFFLINE, the only one that actually
prevents a fetch) gates the security short-circuit, while hf_env_offline()
(either flag, the user's intent) is used only where local-only behavior is
forced explicitly. The SentenceTransformer load now passes local_files_only from
that intent, so TRANSFORMERS_OFFLINE genuinely stops the loader fetching instead
of merely being assumed to.
* studio: short-circuit the security preflight under either offline flag
With the loader now pinned to the local cache by local_files_only =
hf_env_offline(), a TRANSFORMERS_OFFLINE-only session can no longer fetch
anything -- yet the preflight still fell through to two model_info() attempts on
10s and 20s timeouts, stalling every save and load of an already-cached embedder
for half a minute before failing open anyway.
Skip the metadata-only scan whenever either flag is set. The scan's job is to
stop a poisoned pickle being downloaded and deserialized, and nothing can be
downloaded under that predicate; the residual case -- a model cached BEFORE it
was flagged -- is the same fail-open this function has always documented for an
unavailable scan, and is exactly what HF_HUB_OFFLINE already did.
That safety argument depends on every loader behind the gate honoring the same
predicate, so it is pinned as a test invariant instead of a comment: removing
local_files_only from the SentenceTransformer construction now fails the suite.
Drops the short-lived hf_hub_offline() helper, which no longer has a caller.
* studio: scope the offline scan bypass to callers that load local-only
The previous commit put the offline short-circuit inside _fetch_security_status,
which is the malware gate shared by every loader -- so TRANSFORMERS_OFFLINE=1
disabled it for all of them, while only the RAG embedder had been changed to
pass local_files_only. MLX inference (core/inference/worker.py -> FastMLXModel
.from_pretrained), training and export call from_pretrained with no local-only
argument, and huggingface_hub ignores that flag, so those paths could still
fetch and deserialize an unscanned model with the gate switched off.
The bypass is now an explicit local_only_load argument, defaulting to False, and
only the two RAG embedding callers -- whose loader is pinned to the local cache
by the same predicate -- opt in. Tests pin both halves: the shared gate must
still scan under either offline flag by default, and no other caller may pass
local_only_load without constraining its loader.
* studio: capture offline state once, and probe the ST cache root
Two holes in the offline embedding path:
- _get() read hf_env_offline() twice: once inside _guard_model_security and
again for local_files_only. _hf_offline_if_dns_dead() mutates the process-wide
offline vars and restores them on exit, so a concurrent load could see True in
the guard -- skipping the Hub malware scan -- and False by the time the
constructor ran, fetching and deserializing the unscanned repo and breaking
the very invariant that licenses the bypass. The value is now read once in
_get() and passed to both; _guard_model_security takes it as an argument
instead of re-deriving it.
- The cache probe searched only HF_HUB_CACHE. SentenceTransformer downloads into
SENTENCE_TRANSFORMERS_HOME when that is set, using the same
models--org--name/snapshots layout under a different root, so a model fully
present there looked uncached and was rejected with a 409 offline even though
the local-only loader could load it. Snapshot lookup now covers both roots.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* studio: probe the cache the ST loader actually uses, and require it be loadable
Adding SENTENCE_TRANSFORMERS_HOME to the shared snapshot iterator was too broad
in one direction and too narrow in another:
- _get() builds SentenceTransformer with no cache_folder, so with ST_HOME set it
searches THAT root only, never the Hub cache. Probing the union let offline
validation pass on a repo cached only in the Hub cache, after which the loader
looked in ST_HOME and failed. The Sentence-Transformers probe now resolves to
exactly one root: ST_HOME when set, the Hub cache otherwise.
- The shared iterator is also used by the GGUF detectors, whose downloads go
through hf_hub_download with no cache_dir and therefore really do use the Hub
cache. It is back to Hub-cache-only so detection cannot pick a snapshot the
GGUF load will not find.
- Casing normalization ran through resolve_cached_repo_id_case, which scans the
Hub cache, so with ST_HOME set the requested spelling was persisted unchanged
and the exact-case offline load missed the differently cased directory that
detection had just accepted. It now resolves against the same roots detection
uses, exact match first.
- A snapshot carrying only modules.json no longer counts as cached: the online
security preflight downloads that single file itself, and a partial download
leaves it behind, so validation passed for a snapshot with no weights and the
first RAG load then failed. A hit now requires the marker plus a config and at
least one weight file.
* studio: thread the captured offline state into the module probe, fix the gate shard
- _st_module_subdirs() re-read the process env for its local_files_only. With
_hf_offline_if_dns_dead() flipping those vars from another thread, a load that
captured local_only=False could still force this probe local-only, get () back
because modules.json is not cached, and leave the scan with NO module load
roots -- a Hub-flagged pickle under 0_Transformer/ would then pass as an
unreferenced nested artifact while the loader fetched and deserialized it. It
now takes the captured predicate as an argument, and the settings route reads
the state once and uses that single value for both the probe and the scan.
- Skip ST-cache casing on the llama-server backend. Nothing there loads through
SentenceTransformer: the embedder derives a GGUF companion from the saved
spelling and fetches it from the HUB cache, so normalizing to an ST_HOME
spelling would point it at a repo _hf_gguf_backend_error() never validated
(BAAI/bge-m3-GGUF instead of the checked baai/bge-m3-GGUF).
- Fix the security-gate shard, which the signature change had broken: the direct
_guard_model_security / _st_module_subdirs callers now pass the new argument
(they were raising TypeError before reaching any assertion), and the casing
tests patch utils.models.resolve_st_cached_repo_id_case, which the route
actually calls, instead of the Hub-only resolver it no longer uses -- those
patches were being silently ignored.
* studio: accept only torch-loadable weights in the offline ST probe; fix re-export lint
_snapshot_is_loadable_st_model accepted a cached snapshot whose only weights
were .onnx (or .pt), but the RAG loader builds SentenceTransformer with the
default torch backend, so such a snapshot passed offline validation and then
failed on the first load, the exact validate-then-fail this helper exists to
prevent. Restrict _ST_WEIGHT_SUFFIXES to .safetensors and .bin and add a
regression test for an ONNX-only snapshot.
Also teach scripts/verify_import_hoist.py that names listed in a module-level
__all__ are uses, so the legitimately added resolve_st_cached_repo_id_case
re-export in utils/models/__init__.py no longer trips HOISTED-IMPORT-UNUSED.
Covered by two new self-test cases.
* studio: probe the exact repo dir and revision an offline load resolves
The cache probe modelled the cache loosely rather than modelling what
SentenceTransformer actually does with local_files_only=True:
- It merged snapshots across every case-variant repo dir and then read refs/main
from whichever held the newest one. With both models--baai--bge-m3 and
models--BAAI--bge-m3 present, a complete embedding snapshot in the directory
the loader opens could be judged by a newer partial snapshot in the other,
failing validation for a usable model. It now selects the ONE directory the
loader opens, by the same exact-case-first rule resolve_st_cached_repo_id_case
uses to choose the spelling that gets persisted.
- It fell back to scanning historical snapshots when refs/main was absent. With
local_files_only the default revision is resolved THROUGH that ref, so a
snapshot directory alone is not discoverable: the settings request succeeded
and the loader then failed at first indexing. A missing, empty or unreadable
ref is now a cache miss, and the historical scan is gone.
The tests exercise the real lookup against a built cache tree instead of
patching the snapshot iterator, so they now cover the directory selection and
ref resolution the loader depends on.
* studio: record refs/main in the ONNX-only probe test
The ONNX-only regression test predates the refs/main requirement, so after that
change it returned None (a cache miss for want of a ref) before ever reaching
the weight-format check it exists to make. Recording the ref restores its
intent: the snapshot resolves, and the answer is False because an ONNX export is
not loadable by the RAG loader's default Torch backend.
* studio: recognize base-model weight files and gate the offline positive on a materialized snapshot
_snapshot_is_loadable_st_model matched any .safetensors/.bin by suffix, so a
partial cache carrying only a commonly published non-weight bin such as
training_args.bin (or an adapter-only artifact) passed offline validation and
then failed the local_files_only load at first indexing. Match recognized Torch
base-model weight filenames (model / pytorch_model, including sharded) by name.
is_embedding_model retained an online-confirmed positive offline even when no
files were cached, so a metadata-only /check-embedding result let an uncached
repo be saved and then fail at first indexing. Retain the positive only when the
active revision is materialized locally, which still covers a downloaded tag-only
embedder whose snapshot carries no modules.json.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* studio: require a complete weight set offline and persist embedder verdicts across restarts
Two follow-ups to the offline embedding-model classifier:
- _snapshot_is_loadable_st_model now requires a COMPLETE Torch base-model
weight set in one snapshot directory, not just any single recognized weight
file. A partially downloaded sharded model (model-00001-of-00002 without its
sibling) no longer passes offline validation and then fails at first indexing
under local_files_only. Weight files are grouped by directory and a directory
counts only when it holds a single model.safetensors / pytorch_model.bin or a
full shard set whose indices cover 1..total.
- Online-confirmed embedder verdicts are now recorded under the resolved Studio
home (embedding_verdicts.json). The session memo is lost on exit, so a
downloaded tag-only feature-extraction embedder (snapshot present but no
modules.json) was misclassified as non-embedding the first offline call after
a restart. The offline branch consults this durable allowlist in addition to
the memo, still gated on the active revision being materialized on disk, so an
uncached repo is never trusted. Writes are best-effort and only positive
verdicts are stored.
* studio: require complete weights (with shard index) and resolve default casing offline
Follow-ups to the offline embedding-model classifier from the latest review:
- Trust a recorded embedder verdict (session memo or persisted allowlist) offline
only when the active snapshot carries a COMPLETE, loadable weight set, not merely
that it is materialized. A partial download (config present, weights missing or an
incomplete shard set) makes _embedding_marker_in_hf_cache read False rather than
None, so the previous marker-is-not-None gate wrongly returned True and the
local_files_only load then failed. Split out _snapshot_has_complete_weights (config
plus complete weights, modules.json aside) and _active_snapshot_dir, and gate the
known-embedder positive on the weight set.
- Require a sharded checkpoint's index map (model.safetensors.index.json /
pytorch_model.bin.index.json) in addition to every shard before accepting it:
transformers discovers and wires shards through that index, so a complete shard set
without it fails the local-only load.
- Resolve the embedding model name to its exact cache casing in the RAG loader before
constructing SentenceTransformer. The settings route persists that spelling for a
custom override but deliberately leaves the configured default verbatim, so a
default whose casing differs from the cache dir would miss it and fail offline.
Resolving at load time covers the default too; a no-op for a local path or when
nothing case-matching is cached, and idempotent for an already-normalized override.
Adds regression tests for the partial-snapshot verdict, the missing shard index, and
the loader casing resolution; updates the offline-invariant source assertion to the
resolved-name variable.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* studio: require a tokenizer, case-fold verdict ids, and serialize verdict writes
Three follow-ups to the offline embedding-model classifier from the latest review:
- _snapshot_has_complete_weights now also requires a tokenizer asset. A
SentenceTransformer Transformer module builds an AutoTokenizer, so a snapshot with
a complete weight set but no tokenizer.json / tokenizer_config.json / vocab still
fails the local_files_only load. The check is a permissive union over the common
fast-tokenizer, config, and WordPiece/BPE/SentencePiece assets, so an unusual but
valid layout is not rejected -- only a genuinely tokenizer-less partial download.
- The persisted embedder allowlist is now keyed case-insensitively. model_info() is
queried under the requested casing while the settings route saves the cache-resolved
casing, so an exact-string lookup missed the persisted positive after a restart
(baai/model recorded, BAAI/model looked up) and a loadable tag-only embedder was
rejected. Both persist and lookup case-fold the id.
- _persist_embedder serializes its read-modify-write under a lock and writes through a
per-thread temp file, so concurrent confirmations of different embedders no longer
drop each other's entry or collide on the temp path. Cross-process writers stay
best-effort (os.replace is atomic; a dropped verdict is only an optimization miss a
later online re-confirmation heals).
Adds regression tests for the missing-tokenizer reject, alternate tokenizer assets,
cross-casing verdict match, and concurrent verdict writes; updates the snapshot test
helpers to materialize a tokenizer alongside config and weights.
* studio: tighten comments in the offline embedding-model classifier
Comment-only pass over the PR's changed files. Collapse the long block
comments and docstrings around is_embedding_model, the cache-snapshot and
weight-completeness helpers, the embedder-verdict persistence, the offline
security gate, and the offline/casing tests to short one- or two-line forms.
Preserve the rationale (issue #6817, the local_files_only invariant, the
casing and weight-gate reasons) in far fewer words. No code changes.
* studio: drop redundant comments in the offline embedding-model classifier
Second comment-reduction pass over the offline embedding-model cache work:
delete comments and trailing notes that restate the adjacent code or an
assertion, and trim the remaining docstrings and rationale comments to their
load-bearing invariants. Comments and docstrings only; no code changes.
* studio: pin embedder verdicts to a revision, canonicalize default aliases
- A persisted verdict recorded that the Hub tagged ONE revision an embedder, but
was stored per repo. Once refs/main advanced to a complete but non-embedding
Transformer snapshot, the offline path still returned True: the settings route
accepted the updated model without force and RAG could silently load it as an
embedder. Verdicts now carry the commit they were confirmed at and are trusted
only while the active revision matches. One confirmed before the repo was
cached has no revision to compare, so the first revision observed afterwards is
pinned then -- which is what lets a later advance be caught. The persisted file
gains a {id: commit} form and still reads the previous list format.
- tokenizer_config.json no longer counts as a tokenizer asset. It only DESCRIBES
a tokenizer, so a snapshot with config, weights and just that file passed
validation and then failed AutoTokenizer.from_pretrained(local_files_only=True)
at first indexing for common BERT/GPT-style models.
- A casing-only alias of the default is canonicalized to the default up front.
Repo ids are case-insensitive but every gate here compares exact strings, so
saving "Unsloth/bge-m3" against a default of "unsloth/bge-m3" ran the
verification and scan for a custom model and then persisted an override --
after which later changes to the configured default stopped applying.
- verify_import_hoist.py replays __all__ assignments in order instead of unioning
them. Only the final value exports anything, so a later plain "=" that drops a
name must leave its import counted as unused; "+=" still extends, and an
unreadable rebind keeps the earlier names rather than flagging real re-exports.
* studio: validate the real ST load root, and pin verdicts to the Hub revision
Four ways the offline probe still disagreed with what the loader does:
- Verdicts were pinned to the LOCAL refs/main, but model_info() describes the
current HUB revision. With a stale cache the two differ, so an older snapshot
nobody verified was allowlisted. The pin is now info.sha, taken from the
ModelInfo that produced the positive. A verdict carrying no revision (a legacy
entry) is no longer trusted at all -- trusting it meant pinning whatever
happened to be cached, which is the same bug; the next online check re-records
it properly.
- config, tokenizer and weights had to exist somewhere in the snapshot, not
together. modules.json can send SentenceTransformer at 0_Transformer/, which is
loaded FROM that directory, so a cache with the config at the root and only
0_Transformer/model.safetensors passed and then failed the local-only load.
Each directory is now checked as a complete load root, which covers both the
plain HF layout and the ST module layout.
- vocab.json and merges.txt counted independently, but BPE needs the pair unless
a serialized tokenizer.json is present, so half a pair validated and then
failed AutoTokenizer.from_pretrained(local_files_only=True).
- A slashless short name like all-MiniLM-L6-v2 is a supported ST alias that the
loader resolves through the sentence-transformers/ organization, so its
snapshot is cached under that full id. Probing only the bare name reported a
miss and 409'd a model that was cached and loadable; the bare id is still tried
first, matching the loader's own order.
* studio: fail closed for an offline security scan instead of failing open
A local_only (offline) load cannot fetch Hugging Face's malware scan, and the previous
behaviour skipped the scan and failed OPEN, so a cached repo with a poisoned pickle weight
could deserialize under SentenceTransformer(local_files_only=True). Evaluate it fail-CLOSED
against the cached files instead: block a base-model pickle weight the load would deserialize
(pytorch_model.bin and its shards, in a directory with no safetensors alternative) and allow a
pickle-free (safetensors / gguf are inert) cache. A cached pickle model must be reloaded online
once to be scanned, or shipped as safetensors. Nothing cached is not a security event.
_fetch_security_status no longer needs the local_only_load skip (the offline branch is handled
in evaluate_file_security). Adds a regression test covering the safetensors-allow and
pickle-block paths with no Hub call.
* studio: only suppress an offline pickle when a loadable safetensors weight exists
The offline security gate treated any .safetensors in a directory as covering a
pickle weight, so a cache with pytorch_model.bin beside a bare adapter_model.safetensors
(or an orphan shard with no index) passed the fail-closed check even though
from_pretrained still selects and deserializes the pickle. Require a genuinely loadable
safetensors weight -- an unsharded base file or a complete indexed shard set -- before
treating the pickle as covered.
Also make the import-hoist analyzer preserve uncertainty when __all__ is extended by a
value it cannot read statically (__all__ += dynamic()), matching how it already handles
an unreadable rebind, so a dynamically-supplied re-export is not flagged HOISTED-IMPORT-UNUSED.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* studio: scope the offline pickle scan to load paths; reset __all__ opacity on rebind
Address three review follow-ups on the offline security gate and the import-hoist analyzer:
- The offline pickle scan walked the whole snapshot, so a stray pickle in a non-load
subdirectory (archive/, nemo/) that SentenceTransformer never deserializes was blocked.
Scope it to real from_pretrained load roots -- the snapshot root, or a subdir that holds
its own config.json -- matching the online scan's load-path scoping.
- _collect_dunder_all kept a sticky opaque flag: a readable replacing assignment after an
unreadable extend (__all__ += dynamic(); __all__ = []) still credited every import, so a
genuinely unused hoist went unreported. A replacing assignment now resets opacity.
- A bare __all__: list[str] annotation has no runtime value; it was treated as an unreadable
assignment and marked the export set opaque. Skip annotation-only declarations.
* studio: recase slashless ST aliases and accept a pinned embedder after a transient failure
Two offline-detection gaps on well-formed input:
- resolve_st_cached_repo_id_case bailed on every slashless name, so a differently-cased
short alias (all-minilm-l6-v2) validated case-insensitively but was loaded verbatim; the
SentenceTransformer loader rewrites it to sentence-transformers/all-minilm-l6-v2 and looks
it up case-sensitively, missing the canonical sentence-transformers/all-MiniLM-L6-v2 cache
dir. Resolve through _st_cache_repo_dir, which follows the same org alias, and hand back the
on-disk casing.
- On a transient (non-permanent) Hub failure, is_embedding_model only accepted a cached
modules.json marker, so a downloaded tag-only embedder (no modules.json) with a verdict
pinned to the active revision was rejected even though the offline branch accepts the
identical cache. Mirror the offline branch's pinned-verdict acceptance.
* studio: scan modules.json-declared module roots in the offline pickle gate
The offline pickle scan treated only the snapshot root and config.json-bearing subdirs as
load roots, so a pickle in a non-Transformer SentenceTransformer module directory that has no
config.json (e.g. a 0_WordEmbeddings/ module: wordembedding_config.json + pytorch_model.bin)
was skipped even though the loader deserializes it. Parse modules.json (and thread through
load_subdirs) to treat every declared module directory as a load root, so such a pickle is
scanned and fail-closed offline.
* studio: classify cached non-Transformer SentenceTransformer models offline
_snapshot_has_complete_weights recognized only a Transformer-shaped load root (config +
tokenizer + weights co-located), so a fully-cached model built from a non-Transformer module
(0_WordEmbeddings uses wordembedding_config.json + embedding weights and its own tokenizer, no
HF config.json; BoW keeps its vocab in config.json) was classified non-embedding offline and
the settings endpoint returned 409.
Add _snapshot_modules_all_loadable, which parses modules.json and accepts a snapshot when every
declared module's path directory carries the files that module class's own load() reads (a
Transformer/root module still needs the full HF load root; a WordEmbeddings module needs its
config plus a complete weight set; other modules need their *_config.json), and at least one
embedding-producing module is present. It is OR-ed after the Transformer check, so it only ever
accepts more and cannot regress the existing path or reject a pruned cache.
* studio: scan PEFT adapter pickle weights in the offline security gate
from_pretrained auto-detects an adapter_config.json in the load root and deserializes the
adapter weights on top of the base model, so adapter_model.bin is a separate pickle RCE vector
that a safetensors base weight does not cover. The offline scan matched only base-model pickle
names, so an offline local-only load with safetensors base weights plus a cached
adapter_model.bin was allowed despite the live adapter pickle. Scan adapter pickles too, scoped
to a load root where adapter_config.json is present and no adapter_model.safetensors exists.
* studio: require weights for Dense/CNN/LSTM SentenceTransformer modules offline
_module_dir_is_loadable accepted a Dense, CNN, or LSTM module dir with only its config, but
those modules' load() hard-load model.safetensors else pytorch_model.bin (verified against
sentence-transformers source: no fallback, raises if neither exists) -- exactly like
WordEmbeddings. A cache with such a module's config but no weights would validate and then
fail the local_files_only load. Require a complete weight set for every weighted module, not
just WordEmbeddings.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* studio: scan root-index subdir pickle shards offline; handle __all__.append/.extend
- The offline pickle scan followed only load-root directories, so a shard mapped by a root
pytorch_model.bin.index.json into a non-root subdirectory was skipped even though
from_pretrained follows the index weight_map and deserializes it (a layout an attacker can
craft to evade the scanner). Read the local index and scan its referenced pickle shards,
covered by a loadable base safetensors at the index root -- mirroring the online scan.
- The import-hoist analyzer ignored __all__.append("X") / __all__.extend([...]) runtime
re-export mutators, so an import added solely for one tripped HOISTED-IMPORT-UNUSED. Read
their string args like +=, and treat any other __all__ method call as opaque.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* studio: classify StaticEmbedding offline, require WordEmbeddings tokenizer, bound model_info
- A StaticEmbedding module (e.g. sentence-transformers/static-retrieval-mrl-en-v1's
0_StaticEmbedding/) holds tokenizer.json + weights and NO config, so the config-gated
non-Transformer path 409'd it offline. Recognize it by what StaticEmbedding.load() reads: a
tokenizer.json plus a complete Torch weight set.
- WordEmbeddings.load() rebuilds its tokenizer via the configured tokenizer_class.load() from the
module dir, so a WordEmbeddings module now also requires a tokenizer artifact
(whitespacetokenizer_config.json / phrasetokenizer_config.json, or a shared HF tokenizer asset),
not just its config + weights.
- With neither offline env var set, an unbounded model_info() could hang on connect/DNS retries
for networkless users (the #6817 symptom). Bound it with a 15s timeout so a dead network fails
fast and the existing transient-failure cache fallback resolves a cached model, while a
reachable Hub still wins. (Documented caveat: a stalled DNS getaddrinfo may exceed this.)
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Resolve indexed safetensors shards relative to their index
_safetensors_index_complete compared shard basenames against the flat
set of files in the index directory, so an index whose weight_map names
shards in a subdirectory was treated as incomplete whenever a legacy
pytorch_model.bin sat beside it. That falsely blocked a snapshot whose
pickle weights are fully covered by a complete, loadable safetensors
shard set. Resolve each shard path relative to the index directory
instead, and add a regression test for the subdir-mapped shard case.
* Restrict offline weight-completeness check to declared load roots
_snapshot_has_complete_weights scanned every directory in a snapshot and
accepted it when ANY directory was a complete Transformer load root. When
modules.json is present a SentenceTransformer load only opens the declared
module paths, so a snapshot whose declared modules are incomplete but which
happens to contain an unrelated complete directory was accepted offline and
then failed at the first local_files_only load. Restrict the candidate
directories to the roots a load actually opens: the snapshot root plus each
modules.json module path. For a well-formed snapshot the verdict is
unchanged; only a complete directory at an undeclared path no longer vouches
for an otherwise-incomplete snapshot.
* Scan SentenceTransformer Router child module weights offline
A Router (legacy Asym) snapshot declares its child sub-modules only in
router_config.json, not the top-level modules.json, and Router.load()
deserializes each child's weights from its own subdir. A config.json-less
child such as query_0_WordEmbeddings (wordembedding_config.json plus a
pickle pytorch_model.bin loaded via torch.load) was therefore neither a
modules.json-declared load root nor a config.json-bearing dir, so the
offline gate skipped its pickle even though the loader deserializes it.
Parse router_config.json at each load root and treat every declared child
subdir as a load root (bounded BFS, so nested routers are covered), so
those child pickles are scanned. Add Router regression tests: a pickle
child blocks, a safetensors child is allowed, and a Router in a declared
subfolder is followed.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Do not treat an unreferenced config subdir as an offline load root
The offline pickle gate skipped a directory only when it was neither a
declared load root nor held a config.json. Because _st_load_roots already
resolves every real load root (snapshot root, modules.json / load_subdirs
dirs, Router children), the config.json fallback only ever promoted an
UNREFERENCED subdir -- a nested checkpoint-500/ or archive/ that ships its
own config.json + pytorch_model.bin -- to a load root. from_pretrained
never descends into such a subdir and the online scan ignores the same
unindexed pickle, so offline mode wrongly blocked a model the loader reads
from a clean safetensors root. Scope the pickle to directory in roots
only, and add a regression test (a stray checkpoint-500/ no longer blocks;
a modules.json-declared module dir still does).
* Classify a root Router (Asym) model as loadable offline
_module_dir_is_loadable applied Transformer root requirements (config +
tokenizer + weights) to every root module, so a Router saved at the
snapshot root -- which carries only modules.json + router_config.json and
loads its weights from child subdirs -- was classified not loadable
offline, and is_embedding_model missed a cached Router embedder. Dispatch
on the module class before the root Transformer fallback: a Router/Asym
dir is loadable when router_config.json parses and every declared child
subdir is loadable (validated recursively through _module_dir_is_loadable,
so nested routers and every child type are covered) with at least one
embedding-producing child. This also tightens a non-root Router, which
previously validated on the mere presence of router_config.json without
checking its children. Add Router regression tests (root and declared
subfolder, complete and incomplete-child).
* Require every declared module before accepting an offline cache
_snapshot_is_loadable_st_model returned has_complete_weights OR
modules_all_loadable, so a complete 0_Transformer short-circuited the or
and vouched for the whole snapshot even when a declared sibling module was
missing its serialized weights; SentenceTransformer builds every module in
modules.json, so that snapshot passed offline validation and then failed
the local-only load. When modules.json declares a non-empty list it is now
authoritative (modules_all_loadable validates every declared module);
has_complete_weights stays the fallback only for an empty/non-list
modules.json (the plain from_pretrained root). Also add the weight-bearing
modules whose load() hard-loads via load_torch_weights and previously fell
to the config-only path -- LayerNorm, WeightedLayerPooling, SparseAutoEncoder
-- to _ST_WEIGHTED_MODULE_NAMES, with source citations and the deliberate
exclusions (Pooling/Normalize/BoW/WordWeights read no weights on load).
Add parametrized regression tests over LayerNorm/WeightedLayerPooling/Dense
(a weightless sibling rejects, a complete sibling accepts).
* Reject self-referential Router children instead of recursing forever
_router_dir_is_loadable validates each router_config.json child through
_module_dir_is_loadable, which re-enters _router_dir_is_loadable for a
Router child. A malformed types entry naming the router's own directory
(a key of ".", which normalizes to the same dir) made that recursion
never descend, so it looped until RecursionError -- breaking the
documented never-raises contract and turning a crafted/corrupted cached
model into a 500 from is_embedding_model instead of a graceful
unverifiable result. A real child reference is a subdir and always
resolves deeper, so reject any child whose resolved path is the router
dir itself. Add a regression test (a router_config naming "." as a
Router child returns False without raising).
* Treat a destructuring __all__ assignment as opaque
_collect_dunder_all detected __all__ only as a direct ast.Name assignment
target, so a binding through a destructuring target (__all__, meta = [...],
v -> an ast.Tuple) was skipped entirely, leaving an empty, non-opaque
export set. A newly hoisted import re-exported only through that assignment
was then falsely flagged HOISTED-IMPORT-UNUSED. Its value cannot be mapped
statically, so mark the export set opaque when __all__ is reached only
through a destructuring / item / attr target, matching how the collector
already handles other unreadable __all__ forms. Add a self-test case.
* Canonicalize declared module paths before scoping the offline pickle gate
A repo could declare a traversing module path such as 0/../evil in
modules.json (or a router_config child), which SentenceTransformer resolves
to evil/ and deserializes evil/pytorch_model.bin. _st_load_roots recorded
the raw snap/"0/../evil", which never equals the snap/evil that rglob
yields, so the offline pickle gate skipped that directory and a malicious
repo slipped a pickle past the newly added gate. Add _canonical_load_dir
to collapse ./ and ../ components lexically and reject an upward escape,
and route the modules.json paths, load_subdirs and router children through
it so the gate scopes the same normalized directory the loader opens. Add
regression tests for a traversing modules.json path and router child.
* Close offline embedding-classification completeness gaps
Five real offline misclassifications, each a false negative (the #6817 hang
recurs) or false positive (accepted then 409s at the local_files_only load).
Dispatch _module_dir_is_loadable on the module class before the root
Transformer fallback. A module with save_in_root=True (every InputModule:
WordEmbeddings, StaticEmbedding, SparseStaticEmbedding, Transformer, Router)
is saved at the snapshot root, so a root WordEmbeddings was wrongly held to
Transformer requirements (an HF tokenizer it never writes) and classified not
loadable.
CLIPModel is Transformer-shaped: CLIPModel.load() reads AutoModel weights plus
AutoProcessor, so a config-only CLIP dir must not validate.
SparseStaticEmbedding needs a tokenizer plus either idf.json or a complete
torch weight set (conditionally weight-bearing); a config alone is not enough.
A present but empty or malformed modules.json is not loadable and does not fall
back to a root Transformer: with modules.json present the loader never takes
the plain-Transformer path (base/model.py _load_config_modules). The tag-only
no-modules.json embedder is classified separately via
_snapshot_has_complete_weights.
Validate a sharded weight index against its weight_map (every mapped shard
present, resolved relative to the index dir) instead of trusting the index
file's mere existence, mirroring the security-side check.
Add regression tests for all five.
* Close case-folding and online-traversal holes in the offline pickle gate
Two gate bypasses where the security scan credited or scoped a path
differently from what the loader actually resolves:
The safetensors credit was case-folded. _cached_pickle_weight_files lowercases
every filename, and the loadable-safetensors and adapter checks tested those
folded keys against the exact-lowercase names. On a case-sensitive filesystem
(Linux, the Studio default) a crafted repo shipping Model.SafeTensors plus a
malicious pytorch_model.bin makes transformers and sentence-transformers miss
the exact-name model.safetensors and deserialize the pickle, while the gate
credited an inert safetensors and did not block. Credit safetensors
case-sensitively against real filenames, and drop pytorch_model.safetensors
from the credit set (transformers loads only model.safetensors, never that
name). Pickle matching stays case-insensitive (over-blocking a mis-cased
pickle the loader would not load is the safe direction).
The online scan did not canonicalize traversing paths while the offline gate
did. A repo-controlled modules.json path (threaded into the online scan via
the RAG guard) or a weight_map shard entry like 0/../evil / ../evil was
compared verbatim, so a flagged evil/pytorch_model.bin never matched and
evaded the online scan though the loader resolves and deserializes it.
Canonicalize the repo-controlled load-subdir prefixes and weight_map shards
the same way the offline gate does, so offline and online agree.
Add regression tests for both bypasses.
* Treat a conditional __all__ mutation as opaque in the import-hoist linter
_collect_dunder_all replayed only top-level module statements, so an __all__
assignment or mutation inside a module-level if / try / for / while / with /
match (or a deeper scope) was ignored, leaving the export set understated. A
newly hoisted import re-exported only through such a conditional __all__ was
then falsely flagged HOISTED-IMPORT-UNUSED, blocking a valid change. A
conditional value cannot be replayed statically, so mark the export set opaque
when __all__ is bound or mutated anywhere other than a top-level statement.
Add a self-test case.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Scope Router child sub-modules as load roots in the online embedding scan
The RAG embedding security guard unions the SentenceTransformer module dirs
from modules.json into the load roots it scopes for the Hub scan, so a flagged
pickle directly under a Transformer module blocks. A Router (legacy Asym)
module declares its child sub-modules only in router_config.json, not in
modules.json, and Router.load() deserializes each child from its own subdir.
The online scan therefore dropped a flagged child pickle (for example
query_0_WordEmbeddings/pytorch_model.bin) as an unreferenced nested shard while
the loader still deserialized it, the counterpart to the offline gate which
already expands router children via _router_child_dirs.
_st_module_subdirs now reads router_config.json for any Router-typed module and
adds each declared child (joined onto the module path, canonicalized so a
traversing entry is dropped) to the load roots. The config is read only for a
Router-typed module, so a plain embedder pays no extra fetch, and every failure
path still returns () so the guard never bricks the embedder.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Allow a recorded-clean pickle embedder to load offline
The offline embedding security gate is fail-closed: with no network to reach
Hugging Face's scan, a cached pickle weight cannot be verified, so it is blocked
and a model the user already downloaded and used online will not load offline.
This adds a persistent cache of clean Hub verdicts so that exact content can load
offline, without weakening the gate for an unknown or never-scanned pickle.
When an embedding repo is loaded online and HF's scan returns a completed clean
verdict, the load roots are hashed and recorded under the scanned commit as an
exact map of snapshot-relative pickle name to sha256, in a per-user JSON store at
studio_root()/security/embedding_scan_verdicts.json (atomic write, 0600, thread
and cross-process locked, 30-day TTL). Offline, a cached pickle model loads only
when the active cached commit and every load-root pickle's sha256 match the
recorded verdict; a missing record, moved commit, changed or added pickle,
expired record, or any error keeps blocking. Online loads always re-query the
Hub and an authoritative unsafe verdict deletes any stale record, so a
now-flagged commit cannot keep loading on an old clean record.
The store binds repo id, full commit, and a per-file sha256 map so a locally
swapped pickle at the same commit, a branch advance, or an added load-relevant
pickle is detected. A same-user attacker who can rewrite the model cache or the
store is outside the enforceable boundary and this is documented; the sha256 is
computed just before load, so a narrow verify-to-load window remains, and a Hub
scanner false negative is recorded faithfully (safetensors stays the stronger
defense).
Recording is triggered post-load in the RAG embedder because the settings route
only validates and the pre-load guard runs before the constructor downloads;
recording is skipped when the loaded commit differs from the scanned commit. The
blocked-pickle enumerator now returns snapshot-relative Paths so two module dirs
that ship the same pickle basename are hashed and reported distinctly.
* Harden the embedding verdict cache against review findings
Tighten the offline verdict cache and its enumeration so every uncertain or
malformed input fails closed and the recorded hashes always match the files the
loader reads:
- Hash every case-colliding pickle in a load root, not one representative. On a
case-sensitive filesystem pytorch_model.bin and PYTORCH_MODEL.BIN are distinct
files; keying by lowered name dropped one and could hash a decoy instead of the
loader's target. The enumerator now returns every variant Path.
- Only persist a clean verdict for a COMPLETED, entirely-benign scan. Require
scansDone to be the boolean True (not a truthy string), filesWithIssues to be a
well-formed list, and every flagged file to be a definitively-safe level; a
pending, error, unknown, or malformed entry no longer records as clean. The
online block decision is unchanged.
- Fail closed when the offline cache cannot be inspected: an rglob error now
propagates and blocks instead of reading as pickle-free, and a snapshot that
errors on resolution (vs a clean not-cached) blocks. The offline guard also
raises instead of returning when its own inspection throws, so the constructor
never deserializes an unverified cached pickle.
- Expand online Router children recursively (bounded BFS with a seen set),
mirroring the offline load-root expansion, so a flagged grandchild pickle is
scoped online and cannot be recorded clean.
- Reject absolute and drive/UNC declared paths in the load-root canonicalizers;
the loader would resolve them outside the snapshot, so collapsing them to an
in-snapshot relative dir scoped the wrong place.
- Pin verdict recording to the scanned commit's snapshot and take the offline
verify commit from the snapshot directory name, removing a second refs/main read
and the skew it allowed.
- Drop the now-unused pickle-name wrapper.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Tighten offline embedding classification and the pickle gate
Close a set of offline edge cases where validation accepted a cache the
local_files_only load then rejects, and one gate bypass:
- Credit a sharded model.safetensors.index.json for a pickle sibling only at a
from_pretrained root. A non-Transformer SentenceTransformer module (Dense,
WordEmbeddings, StaticEmbedding) loads via Module.load_torch_weights, which
reads model.safetensors then pytorch_model.bin and never the index, so a sharded
safetensors index in such a module dir must not vouch for its pytorch_model.bin.
- Stop counting pytorch_model.safetensors as loadable in the offline classifier:
the loader probes model.safetensors (then its index) or pytorch_model.bin, never
pytorch_model.safetensors, matching the gate that already treats it as a decoy.
- Treat a present but unreadable weight index as incomplete: transformers opens
and parses any present index, so a malformed one or one without a weight_map
fails the load rather than falling back to filename-numbered shards.
- Require the CLIP image-processor config (preprocessor_config.json) for a CLIP
module: CLIPModel.load builds a CLIPProcessor that needs it, so a tokenizer
alone is not enough.
- Require a SparseStaticEmbedding config to actually select idf.json (a path
ending .json) or ship loadable weights; a bare idf.json the config does not name
falls through to load_torch_weights and raises.
- Do not use the tag-only recorded-verdict fallback when modules.json is present:
with the file present the loader takes the modules.json path, so a present but
empty or malformed manifest must not be validated as a plain root Transformer.
- Import-hoist linter: only a module-level conditional mutation or a function that
declares global __all__ makes the export set opaque; a __all__ bound as a local
in a nested function or class no longer masks a genuinely unused hoisted import.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Scope Router-child pickles to their deepest load root and gate the ST offline kwarg
The online scan stripped the first matching load-subdir prefix from a flagged file, so a
nested Router child pickle (0_Router/query_0_WordEmbeddings/pytorch_model.bin) matched the
parent 0_Router root, looked like an unreferenced nested shard, and slipped the gate even
though Router.load() deserializes that child directly. Match the deepest (longest) load
subdir instead, so the child becomes root-level under its own load root and blocks.
pyproject sets no lower bound on sentence-transformers and the local_files_only constructor
arg is absent on older releases, so always forwarding it broke every embedder warm on those
installs. Pass it only for an offline load; an online warm never forwards it and works as
before, while the offline capability still requires a version that supports it.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Reject snapshot-escaping shard paths and credit Transformer submodule safetensors
The offline pickle enumerator joined a weight-index weight_map value straight to the load
root and followed it, so a repo-controlled index mapping "../.." into a sibling snapshot
made an offline from_pretrained deserialize an out-of-snapshot pickle, and an online load
would then hash and record that external file as the scanned commit's clean content. Reject
any shard path that escapes the snapshot root and fail closed, mirroring the canonical-root
check the online shard scan already applies.
A complete model.safetensors.index.json was credited over a sibling pickle only at the
snapshot root, but a Transformer module subdirectory (0_Transformer/) is loaded via
AutoModel.from_pretrained, which honors that shard set and never reads the pickle. Credit the
sharded index for Transformer-typed modules declared in modules.json so a cached model that
ships both a sharded safetensors checkpoint and an unused PyTorch checkpoint is no longer
falsely blocked offline. Non-Transformer modules (Dense, WordEmbeddings, StaticEmbedding) read
a flat weight with no index and keep their pickle blocked.
Limit the import-hoist verifier's global __all__ scan to the declaring function's own scope so
a nested inner-scope local __all__ no longer marks the module export set opaque and mask an
unused hoisted import.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Scope Router children against the snapshot and mirror the ST alias rewrite
Router.load resolves each child at Path(subfolder, model_id) relative to the Router dir, so a
nested 1_Router with a "../evil" child points at evil/ inside the snapshot and the loader
deserializes evil/pytorch_model.bin. The offline enumerator canonicalized the child against
the Router dir alone and dropped anything with "..", so that pickle was never scanned and the
gate reported the cache pickle-free. Canonicalize router children against the snapshot,
retaining in-snapshot siblings as load roots and failing closed on a child that escapes the
snapshot itself, matching the online scan which already joins the prefix before normalizing.
The security gate resolved a slashless model id by probing the bare cache dir first, but the
SentenceTransformer constructor rewrites a non-basic slashless name to sentence-transformers/
<name> and loads THAT snapshot (only the basic ORIGINAL_TRANSFORMER_MODELS load bare). With
both models--<name> and models--sentence-transformers--<name> cached, the gate inspected the
bare dir while the loader read the namespaced one, so a pickle there bypassed the local-only
gate. Mirror the constructor: try the namespaced candidate first for non-basic slashless names.
Add the same not (snapshot / modules.json).is_file() guard to the transient-Hub-failure
tag-only fallback that the offline branch already carries, so a cache whose present manifest is
empty or malformed is no longer reported as a loadable embedder.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Tighten root shard credit, module-path escapes, and weight-set probe order
Credit the sharded safetensors index at the snapshot ROOT only when the root is actually loaded
through an AutoModel/from_pretrained path. A modules.json root module of a non-Transformer type
(StaticEmbedding / WordEmbeddings / Dense) loads via load_torch_weights, which reads
pytorch_model.bin and ignores the index, so crediting a root shard index there suppressed a live
root pickle and let the offline gate report the cache pickle-free.
Recognize the Transformer subclasses CLIPModel and MLMTransformer as index-honoring load roots
(they load via from_pretrained), so a sharded-safetensors CLIP/MLM submodule with a legacy
pytorch_model.bin sibling is no longer falsely blocked offline. Mirrors the classifier dispatch.
Fail closed on an absolute or snapshot-escaping modules.json module path (or load_subdirs entry)
instead of silently dropping it: SentenceTransformer resolves such a path outside the snapshot and
would deserialize an external pytorch_model.bin the gate cannot scan.
On the classifier side, walk the weight set in the exact from_pretrained probe order
(model.safetensors, its index, pytorch_model.bin, its index) so a pickle behind a malformed
safetensors index is no longer accepted as complete, and restrict shard names to the loader-probed
stem/ext pairs so a decoy model-*.bin / pytorch_model-*.safetensors set is not treated as loadable.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Restore scripts/verify_import_hoist.py to main
The offline embedding cache fix does not depend on the __all__ scope
handling that had accumulated in this linter, so revert the file to its
main version and keep the PR focused on the feature. The feature modules
still pass the existing import hoist check unchanged.
* Reuse a shared HF cache skeleton in the offline classification tests
Extract _mk_repo and _activate helpers for the repeated snapshot cache
setup that every per-type builder duplicated, and fold the two
StaticEmbedding missing-asset cases into one parametrized test. Same 125
collected items, all still passing.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Reclassify embedding models from the cache on every offline call
is_embedding_model consulted its process memo before the offline branch, so an
online lookup that memoized True from tags (without caching any weights) was returned
unchanged once the session went offline -- the studio flips HF_HUB_OFFLINE in-process
on a dead DNS, and the ungated check-embedding route can populate the memo. Settings
would then accept a repo the offline loader cannot open. Run the offline
cache-marker reclassification ahead of the memo and never record it, so an offline
verdict always reflects the local cache and a later cache materialization is not
masked by a stale negative. Add regression tests.
* Tighten comments on the offline embedding path
Condense the offline-embedding helper docstrings and inline comments added in
this PR to fewer, clearer lines, keeping the non-obvious security and offline
rationale. Comments and docstrings only; no code change.
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: danielhanchen <unslothai@gmail.com>
* Studio: reuse MLX prompt cache across turns instead of re-prefilling
* clean up
* key prompt cache on what the KV covers
* skip windowed KV caches past their window
* verify prefix coverage before caching KV
* Recognize Radeon 8065S (Gorgon Halo / Ryzen AI Max 400) as gfx1151
* Classify Radeon 8065S (Gorgon Halo) as unified memory in ROCm OOM guard
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
* studio: extend the _grouped_mm null-kernel guard to Linux ROCm RDNA4
torch._grouped_mm has a null HIP kernel on RDNA4 (gfx1200/gfx1201) at
ROCm <= 7.12 (fixed in 7.13; ROCm/TheRock #5284). The existing guard that
registers a Python mm/bmm fallback was win32-only, so Linux gfx1201 (e.g.
R9700 Pro on Ubuntu) hits the null kernel -> illegal instruction during
training.
Extract the fallback registration into a module-level helper
(_install_grouped_mm_cpu_fallback) and add a Linux branch that installs it,
gated on gfx1200/gfx1201 AND HIP < 7.13 so NVIDIA/CUDA and every non-RDNA4
AMD arch are untouched, and it is a no-op on fixed runtimes. The Windows
path now calls the same helper with identical behavior.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Resolve HIP version from torch.__version__ when version.hip is unset for PR #7292
AMD SDK / Radeon ROCm wheels leave torch.version.hip empty and encode the
version only in torch.__version__ (e.g. +rocm7.12). The Linux gfx120X guard
parsed version.hip only, so those affected installs skipped the fallback and
still hit the null _grouped_mm kernel. Mirror the Windows parse: version.hip,
then the embedded rocmX.Y, then assume affected unless a post-fix rocmsdk wheel.
* Scan all GPUs and add RDNA4 name fallback for _grouped_mm guard in PR #7292
* worker.py: tighten gfx120X Linux guard comments (no code change)
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
* fix(rocm): prepend system ROCm libs on native Linux to avoid bundled HIP crash
Prebuilt llama.cpp bundles ship their own ROCR/HIP runtime which can be
incompatible with the host's amdkfd kernel driver, causing hsa_init()
to crash or report zero devices. The llama-server then silently falls
back to CPU while the UI reports GPU.
The existing workaround (_wsl_system_rocm_lib_dirs) that prepends
/opt/rocm/lib to LD_LIBRARY_PATH was gated on WSL (/dev/dxg) only,
leaving native Linux AMD hosts unprotected.
This commit adds _native_linux_system_rocm_lib_dirs(), a parallel
helper gated on:
- Linux platform (not WSL)
- /dev/kfd present (bare-metal AMD compute)
- Bundle contains bundled HIP libs (libggml-hip.so)
- System has libhsa-runtime64.so(.1)
It is called from both _llama_server_env_for_binary (serve-time)
and binary_env (install-time validation), directly after the WSL
block in both paths.
Fixes#7208Fixes#7208
* Add UNSLOTH_LLAMA_NO_SYSTEM_ROCM opt-out to native-Linux system ROCm preference for PR #7233
Lets a host where the bundled runtime works but system ROCm is mismatched keep
the bundle. Mirrored in llama_cpp.py and install_llama_prebuilt.py.
* Prefer env-configured ROCm root over /opt/rocm fallback for PR #7233
Put HIP_PATH/HIP_PATH_57/ROCM_PATH-derived roots before /opt/rocm so a stale
/opt/rocm can't shadow the driver-matching install the env vars point at.
Mirrored in llama_cpp.py and install_llama_prebuilt.py.
* Match versioned libggml-hip.so via glob so the native-Linux ROCm fix fires for PR #7233
* Clarify native-Linux ROCm prepend uses the consistent system stack for PR #7233
* llama_cpp: tighten native-Linux ROCm prepend comments (no code change)
---------
Co-authored-by: Daniel Han <danielhanchen@gmail.com>
* Fix resume training crash recovery and MLX checkpoints
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Address review: preserve interrupted stop-and-save output_dir, verify MLX checkpoint
- finish_run: add clear_output_dir flag; preserve output_dir for stopped/error
unless cancel explicitly clears it (fixes pump finalization wiping persisted path).
- training pump: pass interrupted stop-and-save context into finalize_run_in_db.
- MLX stop-and-save: verify resumable checkpoint exists before sending complete;
return bool from _write_mlx_stop_checkpoint and add regression tests.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Address Codex review: MLX current-step checkpoint and cancel error finalize
- Only skip MLX stop checkpoint write when checkpoint-{current_step} exists;
stale periodic checkpoints no longer mask missing stop saves.
- Pass clear_output_dir through error-event finalization so Stop-without-save
cannot leave a persisted output_dir that still offers Resume.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Address review
* Address more reviews
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* more reviews
* clear in-memory output_dir on interrupted cancel
* allow resuming errored runs at the final step
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Clear persisted output_dir in cancel watchdog path
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Write MLX stop checkpoint in stop path, keep output_dir on crash finalize
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* fix(studio): harden resumable run finalization
* fix(studio): defer safetensors checkpoint import
* fix(studio): reject stale training cancellation
* fix(studio): replay null resume targets
* fix(studio): serialize terminal cancellation
* Harden resume checkpoint validation and fix stop-save cleanup
- Reject unrecognized shard formats and keep indexed shard paths inside the checkpoint dir
- Require a non-empty tensor record when validating .pt/.bin optimizer and model state
- Always finalize TensorBoard and W&B on stop-save-failure exits
- Refuse writing an MLX stop checkpoint through a symlinked directory
- Clarify the resume rejection message to cover errored runs
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Tighten resume/checkpoint comments
* Recover resumability when a valid stop checkpoint landed
- Re-validate the current-step checkpoint in the dead-worker and error finalization paths so a stop-and-save that actually wrote a valid checkpoint is not wrongly marked error/resume_blocked
- Accept a valid tensor-free optimizer state (e.g. SGD without momentum); the model-state check still requires real tensors
- Include errored runs in the frontend resume rejection message
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Lyxot <longyixing331@gmail.com>
Co-authored-by: Lee Jackson <130007945+Imagineer99@users.noreply.github.com>
Co-authored-by: danielhanchen <danielhanchen@gmail.com>
* Studio (Windows): keep prompt caching on full GPU offload (#5692 follow-up)
The #5692 full-offload tuning also added --no-cache-prompt, which disables
in-VRAM prompt-prefix reuse. That is unrelated to the host-RAM KV checkpoints
#5692 fixed (--cache-ram 0 / --ctx-checkpoints 0): a fully offloaded model keeps
its KV cache in VRAM, so reusing a common prefix does not copy to system RAM and
does not cause the PCI-E overhead. --no-cache-prompt only forces every request to
re-prefill the whole prompt, which is small for short chats but severe for large
stable system prompts reused across calls (coding agents, long multi-turn chats).
Remove --no-cache-prompt; keep the checkpoint disables and the thread/OMP tuning.
_prompt_cache_disabled stays False (its default), so slot save/restore is intact.
Verified on a fully offloaded gemma GGUF: an identical repeated prompt reprefills
1 token instead of 2220.
* Guard against re-adding --no-cache-prompt to any llama-server command
Add a backend-wide test that AST-scans studio/backend and fails if
--no-cache-prompt is appended/extended/+= into a command. This locks in
the #7260 fix across every code path, not just load_model. Detecting the
flag or honouring a user-supplied one stays allowed.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
* Studio: make Stop and stall deadlines interrupt a wedged stream portably
The cancel watcher unblocks a stalled read by shutting the socket down from
another thread, which works on POSIX but not reliably on native Windows, where
Winsock does not dependably wake a recv() already in progress on another thread.
Wrap the httpcore network stream so the reader loops each read in short slices
and polls the cancel event itself. Stop and the stall deadlines now interrupt a
wedged mid-stream read without any cross-thread socket teardown, and a slow but
still-alive stream is never torn down. The POSIX shutdown path is preserved.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Honor the post-first-token stall timeout in the cancel-aware read
httpcore snapshots request.extensions timeout read once when the body
starts, so lowering it to the stall timeout after the first token never
reached the socket read and a one-token-then-silent server hung for the
full prefill window. Re-read the live extensions timeout per call and
bound each read by it, falling back to the httpcore-passed timeout when
absent so prefill and normal completion are unchanged.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* studio: tighten comments in the llama.cpp stall timeout path
* Tighten comments in the stream stall cancel path
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: danielhanchen <unslothshared@gmail.com>
* Studio: Data settings tab, uploaded files manager, quant pinning, image preview fix
Settings
- New Data tab in the settings sidebar, under Connections. Chat data
management (archived chats, confirm before deleting, exports, import,
clear all) moved there from the Chat tab.
- New Archive all chats action with confirmation. Archives every chat in
Recents and Projects; compare pairs count as one chat.
- New Uploaded files manager listing RAG documents (chats, projects,
knowledge bases) and chat message attachments with location, size and
date. Files can be opened in a new tab or deleted. Deleting a chat
attachment keeps the message text.
Backend
- GET /api/rag/documents lists all uploaded RAG documents with file size
plus KB and project names.
- GET /api/chat/attachments lists chat message attachments; per
attachment file and delete endpoints included.
Model selector
- Downloaded GGUF quants can be pinned from the quant row (next to the
settings and delete actions). Pinned quants show at the top of On
Device under a Pinned heading as model name plus a grey quant chip and
load directly with one click. Non GGUF cached repos pin as a whole.
- Toned down the green of the downloaded label.
Fix
- Clicking an image attachment in chat now opens the preview overlay.
The tooltip trigger wrapper called preventDefault before composed
handlers ran, which made Radix DialogTrigger skip opening.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: image previews and file type chips in uploaded files list
Image attachments now show a small thumbnail (lazy loaded from the
stored bytes, object URL revoked on unmount) and every row shows a grey
uppercase type chip derived from the extension or content type. Non
image rows keep a file icon. Name cell floors its width and clips
overflow so narrow dialogs stay aligned.
* Harden attachment serving, add tests, and polish pinned rows and previews
- Strict base64 decoding for attachment files: corrupt payloads now return
422 instead of silently serving empty or garbled bytes; whitespace,
missing padding, the URL-safe alphabet, and RFC 2397 percent-encoded
data URLs are all handled
- New backend test suite covering attachment listing, size accounting,
malformed rows, deletion semantics, and every file-serving edge case
- Pinned quant rows show a Loaded tag when that exact quant is active,
and reveal unpin, settings, and delete actions on hover
- Uploaded files dialog is wider and chat locations link straight to the
thread the attachment belongs to
- Chat image preview is now a chrome-free lightbox: dimmed backdrop,
rounded image, corner close button, click outside to dismiss
- File opens go through a synchronous window.open so Safari and Firefox
popup blockers do not eat them
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Uploaded files: click a file to jump to its chat, square thumbs, new Data icon
- Clicking a file row (thumbnail or name) now goes straight to the chat it
belongs to; files without a chat open directly as before
- File thumbnails pin a small 7px radius: the theme scales rounded-md up
to a near circle at this size
- Settings Data tab now uses the database-setting icon
* Uploaded files is now a Data tab subpage instead of a popup
- Manage swaps the tab body for an inline Uploaded files page with a back
header, matching the rest of settings navigation
- Size column header and values are left aligned like the other columns
- Column widths tightened so the table fits the settings panel
* Lightbox polish and Data tab row order
- Image preview close button is transparent until hovered
- Preview image no longer rounds its corners
- Import chats now sits below Clear all chats in the Data tab
* Data tab: export chats as fine-tuning data and open them in Recipes
- New Fine-tuning section in Settings > Data converts every chat into a
JSONL dataset in the OpenAI messages format, one conversation per line
with string-only system/user/assistant turns
- The Train tab detects this file as chatml natively: no column mapping
and no standardization pass, and it works with train on completions
since every assistant turn sits behind the chat template response marker
- Consecutive same-role turns merge, trailing turns without an assistant
reply drop, and reasoning, tool calls, and images are excluded so chat
templates format the data cleanly
- Open in Recipes stages the JSONL as a local seed upload, creates a new
Data Recipe with the seed block preconfigured, and jumps to the editor
* Data tab: load chats straight into the Train tab, row moved to the top
- New Load in Train tab button uploads the fine-tuning JSONL through the
training dataset endpoint, selects it in the training config store, and
opens the Train tab with the dataset loaded and format-checked
- Use chats as training data now sits at the very top of the Data tab
- The Chats subheading is gone; chat rows flow directly under it
* Address review findings on the uploads manager and quant pins
- Deleting the last attachment stores '[]' instead of NULL: a NULL reads
back as a missing field and triggers the legacy IndexedDB backfill,
which resurrected the deleted attachment on the next chat load
- The attachment file endpoint now serves audio: adapter parts store
{data, format} raw base64 and compare chats store a bare base64 string;
media type comes from the attachment contentType or the format
- Compare-chat uploads live in message content parts, not attachments;
the uploads list now includes those blobs via synthetic content-part
ids that the same get and delete routes resolve
- Deleting a quant from the expanded repo row also unpins it so a pinned
row cannot try to load a file that no longer exists
- Thumbnails in the uploads list fetch their blob only once the row is
visible, so a long screenshot history does not download everything
- Nine new backend tests cover audio serving, content-part listing,
serving, deletion, and the empty-list delete behavior
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Data tab: single action dropdown with format choices for chat training data
- The three fine-tune buttons collapse into one dropdown plus a run
button; pick Load in Train tab, Open in Recipes, or Export JSONL,
then click the arrow to run it
- The dropdown's Format section adds ShareGPT and Alpaca alongside the
default OpenAI messages format, ticked like a checklist; all three
shapes are auto-detected by the Train tab's format check
- Alpaca is single-turn, so each user to assistant pair becomes its own
record with the system prompt and earlier turns carried in the input
column
- Shorter description on the training data row
- Uploaded files rows show the size under the file name instead of a
separate column, matching the tighter layout
* Polish the training data action control
- Run button is a true circle (icon-sm plus rounded-full) with a
heavier arrow stroke
- Dropdown trigger uses the shared standard chevron and a fixed width
so switching actions no longer resizes the control
* Shorten the training data row description
* Use the standard chevron for the run button and enlarge the ticks
- Run button uses the shared standard right chevron so it matches the
dropdown chevron instead of the hugeicons arrow
- Dropdown ticks bumped up a size for legibility
* Reword the training data row description
* Shorten Data Recipes to Recipes in the training data description
* List Export JSONL first and rename the default format to Chat Completions
* Handle legacy string content in fine-tune exports and gate Train on chat-only hosts
- messageToPlainText now accepts plain-string message content, the shape
legacy and imported histories store, so those conversations export
instead of being skipped as having no exchange
- The Load in Train tab action is disabled on chat-only hosts the same
way the sidebar gates Train; the default action falls back to Export
JSONL there so the run button never uploads a dataset that /studio
would immediately redirect away from
* Narrow the training data action dropdown slightly
* Drop the format picker from the training data dropdown
Chat Completions (OpenAI messages) is the only export format we ship, so
the ShareGPT and Alpaca options and the Format section are removed. The
export always uses the OpenAI messages shape.
* Address the second round of review findings
Security
- Chat attachment data URLs no longer echo their embedded media type:
anything that is not a plain raster image serves as octet-stream, so
imported text/html or SVG payloads cannot render under the app origin
- Uploaded .html/.htm RAG documents serve as text/plain for the same
reason; the preview sheet only uses the file URL for PDFs
Uploads manager
- Remote image URLs in imported chats are no longer listed as stored
uploads (nothing to serve, and delete would strip the chat reference);
the delete guard mirrors the same data:-only rule
- Deleting a content-part upload refetches the list since the remaining
parts re-index, keeping sibling row ids current
- Deleting a project document from the Data tab invalidates the project
sources cache like the sources panel does
- Data-tab deletions now patch the loaded thread's in-memory copy via a
small event, so a later repo sync cannot write the attachment back
Fine-tune export
- Branch siblings from retries stay out of the exported conversation;
only the selected chain converts (full exports still keep everything)
- Assistant turns before the first user turn drop, preserving leading
system prompts, so no unconditioned assistant targets are emitted
Four new backend tests cover the media type clamp and remote-URL rows;
two existing tests updated for the clamped types
* Fix uploaded file lifecycle and model state
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Make archived chats a Data settings subpage
* Studio: fix attachment route tests and pinned quant edge cases
- test_chat_attachments: drop asyncio.run around the synchronous
/attachments routes (list/get/delete are plain def, so asyncio.run
raised 'a coroutine was expected' and failed the Repo tests CI job).
- test_chat_attachments: align compare-chat content-part assertions with
the stable content-hash id scheme (content-part-sha256-...) instead of
the removed array-index ids; resolve ids from the listing.
- pickers: pass disabled={deleteDisabled} to the pinned-quant delete
action so a quant cannot be deleted mid model-load, matching the
expanded variant rows.
- pickers: build the pinned-quant existence set from the query-unfiltered
cached GGUF repos (format filter still applied) so a pinned quant stays
findable when the search term matches only its quant name.
* Fix Studio review regressions
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Guard fine-tune export content blocks
* Add Export button for archived chats
Adds an Export action to the Archived chats view in Settings > Data that
downloads only the archived chats as a JSON backup (their threads, messages
and projects). The button sits in the archived header row and appears only
when archived chats exist.
* Refactor archived export into pure, testable units
Split the archived-chats export into a dependency-free filter
(archived-chat-export.ts) and a shared JSON download helper
(download-json.ts). Skip the download when nothing is archived so a
stray call never drops an empty file. No behavior change to the button.
---------
Co-authored-by: shimmyshimmer <shimmyshimmer@users.noreply.github.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Etherll <61019402+Etherll@users.noreply.github.com>
Co-authored-by: Unsloth <michaelhan@Michaels-MacBook-Pro.local>
Co-authored-by: Lee Jackson <130007945+Imagineer99@users.noreply.github.com>
* Fix text-only VLM CPT packing truncation
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Handle streaming vision datasets in packing
* Harden multimodal packing detection
* Preserve safe packing boundaries
* Scope stream packing checks to VLMs
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Narrow VLM packing detection
* Align packing mode and eval safety
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Add qwen3_5/qwen3_next to PADDING_FREE_BLOCKLIST to avoid packed-sequence contamination
* Detect hybrid linear-attention models structurally instead of by name for packing guard
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Install wrapped-packing setup at the signature, not the Zoo license comment
The _unsloth_wrapped_packing / _inspect setup block was injected by matching the
exact 'All Unsloth Zoo code licensed under LGPLv3' comment line in the sourced
sft_prepare_dataset. The unsloth_zoo dependency is only lower-bounded, so a newer
Zoo that moves or drops that header made the setup a silent no-op while the
truncation and pack_dataset rewrites still emitted references to those names,
raising NameError on every SFT dataset preparation.
Anchor the setup on the function signature instead (a structural location that
always exists) and fail loudly if it cannot be found, so the helper variables are
always defined before they are referenced across Zoo versions.
Adds a regression test that patches in a Zoo source without the license header.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Etherl <61019402+Etherll@users.noreply.github.com>
Co-authored-by: danielhanchen <danielhanchen@gmail.com>
* Studio: persist llama.cpp KV cache across idle auto-unload (slot save/restore)
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: address KV persistence review feedback
* Studio: guard KV restore on launch config
* Studio: fix KV resume purge race, fingerprint requested ctx, purge on disable
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: re-check idle/keep-KV settings after slot save, ns file identity
* Studio: shard-aware KV guard, honor user --no-cache-prompt, early save cap
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: honor LLAMA_ARG_CACHE_PROMPT env in slot-save guard
* Studio: derive prompt-cache state from final argv for slot saves
* Studio: stat LoRA/control-vector sidecars in KV restore fingerprint
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: parse csv and FNAME:SCALE sidecar syntax in KV fingerprint
* Studio: address codex review on idle-unload KV resume
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: harden slot-save cleanup, cap accounting, stale-KV guard, save timeout
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: treat unavailable KV estimate as full-cap for slot-save disk check
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Daniel Han <danielhanchen@gmail.com>
* Studio: GPU memory dropdown — llama.cpp --fit on and manual gpu-layers/cpu-moe
* Studio: simplify GPU memory changes (reuse ParamSlider, GPU_LAYERS_ALL, loadedGpuMemoryFields helper)
* Studio: GPU picker — choose which GPUs a GGUF model loads on (gpu_ids)
* Studio: simplify GPU picker (share /api/system fetch, validate gpu_ids)
* Studio: GPU picker review fixes (gate relative indices, no cross-model leak, validate, types)
* Studio: group GPU controls under a collapsible GPU section
* Studio: GPU feature review fixes (fix fit-ctx test, behavior-test the floor, comment accuracy)
* Studio: make GPU a top-level settings section (not nested under Model)
* Studio: flatten GPU controls into the Model section, group by GPU/context/generation
* Studio: move GPU Memory to the bottom of Model with its dependent controls beneath it
* Studio: move GPU Memory below Tensor Parallelism and GPUs below GPU Memory
* Studio: tighten GPU Memory and GPU Layers tooltip copy
* Studio: fix fit-mode context slider track-click, restore GPU Memory tooltip, shorten fit dropdown label
* Studio: GPU Memory tooltip one mode per line, briefer
* Studio: note HIP_VISIBLE_DEVICES (ROCm) in the GPUs picker tooltip
* Studio: narrow the GPU Memory dropdown to fit the shortened label
* Studio: use 'llama.cpp --fit' in the GPU Memory tooltip for consistency
* Studio: allow Tensor Parallelism in Manual GPU mode
* Studio: graduated MoE-on-CPU offload (--n-cpu-moe) replacing the all-or-nothing toggle
* Studio: size the MoE-offload slider for staged (deferred-load) models
* Studio: share one GGUF header walk for the context-length and MoE-count readers
* Studio: size the GPU Layers slider for staged models (one staged-header read)
* Studio: move Tensor Parallelism below the GPUs picker
* Studio: GPU split (--tensor-split) per-GPU model share in Manual mode
* Studio: tolerate whitespace in GPU split input, move it below GPU Layers
* Studio: rename the GPU split control to "Split ratio"
* Studio: Split ratio sends explicit even input; fix blank=free-VRAM (not even) copy
* Studio: tighten llama.cpp --fit VRAM margin with --fit-target 512
* Studio: GPU memory review fixes (rollback re-baseline, single-GPU TP gate, accurate copy)
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: move Split ratio below MoE Layers on CPU
* Studio: address PR review (fix GPU-info hydration race, share fit context-length across load paths)
* Studio: address codex review (manual single-GPU TP guard, GPU-aware spec defaults in fit/manual, GGUF-only context/preference)
* Studio: address codex review round 2 (gpu_present seed, single-GPU tensor-split guard, staged manual-knob reset, strip inherited offload flags)
* Studio: address codex review round 3 (strip inherited --n-cpu-moe, CPU-fallback warning in Manual mode)
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: address codex review round 4 (preserve pinned fit context across a later Apply)
* Studio: address codex review round 5 (honor GPU picker for diffusion GGUFs, clear fit pin on cross-model switch)
* Studio: preserve the pending GPU Memory mode when staging a model
* Studio: pin diffusion GPU device order and reset GPU-memory state for diffusion loads
* Studio: address codex review round 6 (fit-Auto rollback context, preserve manual non-tensor split modes, persist GPU mode on load not select)
* Studio: persist the applied GPU Memory mode, not the requested one (skip diffusion loads)
* Studio: replace Manual-mode split-ratio field with per-GPU layer sliders
* Studio: clarify per-GPU layer split hint for tensor-parallel mode
* Studio: address codex review round 7 (allow GGUF gpu_ids past the legacy guard, replay GPU-memory fields on respawn)
* Studio: address codex review round 8 (size the validate preflight like the load in fit mode, across both load paths)
* Studio: skip the training-OOM guard for llama.cpp --fit GGUF loads (they spill to RAM)
* Studio: drop the now-redundant compare-path validate sizing (the --fit guard skip makes it moot)
* Studio: address codex review round 9 (keep the training guard for fit loads, forward gpu_ids to validate, strip inherited manual tensor-split)
* Studio: address codex review round 10 (gate GPU-memory adoption on is_gguf, record manual knobs only in Manual mode)
* Studio: handle diffusion GGUFs symmetrically in the GPU Memory controls (preserve the standing mode preference, hide the inapplicable mode/TP controls)
* Studio: remember the GPU Memory settings per model
* Studio: consolidate --fit mode and Manual mode into a single Manual mode
* Studio: preserve the per-GPU layer split across GPU Layers changes
* Studio: trim overly long GPU Memory comments
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* address GPU memory config review comments
* trim redundant GPU memory tests
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Reconcile manual-mode TP drops with the #6659 drop-site invariants
* Preserve quantized KV in manual --fit, charge GGUF companions in full, reconcile GPU pick on load
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Clear stale GPU baseline on non-GGUF loads so it can't read as dirty
* Fix no-context-shift test for the conditional -c flag
* Credit manual GPU-layer offload for cached HF GGUFs
* Reset per-model load knobs on GGUF quant switch
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Strip inherited tensor-split when manual ratio is cleared
* Match auto-load validation to safetensors placement
* Reset editable manual knobs after Auto GGUF loads
* Record a single device for diffusion GPU picks
* Reset per-model GPU knobs before applying saved settings
* Address review comments
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Guard manual tensor splits and keep remembered context on auto-load
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Snapshot compare knobs, seed splits from free VRAM, flag zero-offload loads
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Exempt CPU-only loads from the guard floor and harden compare and reseed paths
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Reach full offload from the layers slider and charge extras drafters in the guard
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Warm the GPU device cache before pick reconciles and disable staged GPU controls
* Align the training guard with inherited extras, spec mode, and compare targets
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Hide GPUs from companion-less zero-offload loads
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Size diffusion picks per device, own manual offload flags, reject XPU picks
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Drop tensor flags at zero layers and exempt CPU-pinned drafters
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Allowlist the zero-layer tensor parallel drop site
* Keep validate and load guards on the same extras and refresh stale baselines
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Drop mismatched manual tensor splits before launch
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Gate XPU picks on the real backend field and harden split and hydration paths
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Weight full GPUs as zero, clamp split shares, and refine the zero-layer mask gate
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Carry fit context across mode changes and align drafter and picker gates
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Catch variant switches, uncached diffusion repos, and text-only mmproj skips
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Check companions on the first device and size native and remote zero-layer loads
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Replace the training guard's precise VRAM modeling with a conservative bound
* Baseline context pins on non-GGUF hydration and reprobe list-seeded staged GGUFs
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Size manual splits by their largest share and preserve resolved context from Default
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Default-deny unsized required companions and price KV at the effective cache dtype
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Reserve MTP draft KV and MLA target-copy in the training guard
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Size tensor-parallel loads per device and show GPU controls for native GGUFs
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Reserve MTP overhead for uncached remote GGUFs and the mmproj runtime factor
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Drop the training-coexistence VRAM estimation this PR added
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Gate remembered load settings to GGUF picks
* Lock the remaining load-time controls during a staged load
* Clear the stale native-path token on compare loads
* Drop a stale guard reference from the zero-offload masking comment
* Seed GPU baselines from the rollback response and drop never-emitted offload flags
* Match validate's training guard to load and keep the native reload token
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Trim verbose GPU-memory comments
* Thread the variants header walk off the event loop, honor device pins on zero-offload, and hold staged GPU edits
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Honor manual placement and classify pinned zero-offload loads
* Close diffusion admission and status hydration gaps
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Check the actual diffusion GPU during training
* Align staged baselines and manual reload dedupe
* Fix GGUF placement and rollback state
* Harden manual GGUF placement boundaries
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Remove unused resolve_tensor_parallel import in llama_cpp.py
The name is used only in llama_server_args.py, routes/inference.py, and tests,
not in llama_cpp.py; the unused hoisted import trips the import-hoist verifier
in the source-lint CI job.
* Fix diffusion GPU dedup and training guard for non-numeric device tokens
The diffusion runner drives only its single lowest device and the backend
records that one device (self._gpu_ids = [sorted(gpu_ids)[0]]), but the reload
dedupe compared it against the full requested list, so a multi-GPU pick that
resolves to the same device forced a needless reload. Normalize the request the
same way for a loaded diffusion model in both _already_in_target_state and the
route _request_matches_loaded_settings.
The chat-during-training coexistence guard called int() on the single-device
token and hard-rejected when it could not parse. A non-numeric token (a CUDA
UUID / MIG handle) now sizes against the whole visible pool like the GGUF guard
instead of falsely blocking the load, and an empty token (a CPU-only runner such
as a CPU diffusion GGUF) is allowed outright since it uses no GPU VRAM.
* Tighten comments added by the GPU memory config changes
* Harden GGUF placement from independent review: VRAM sizing, diffusion TP reset, tensor_split validation
- Training coexistence guard: a single-device runner pinned through an
unresolvable UUID/MIG token was sized against the aggregate visible-VRAM pool,
so a load could pass on capacity it cannot use and then OOM active training.
Size against the worst-case visible device (min free) instead, keeping the
guard's documented default-deny contract. The empty-token (CPU-only runner)
allow path is unchanged.
- Diffusion startup: _start_diffusion_server now resets self._tensor_parallel to
False alongside the other placement resets. A prior tensor-parallel chat load
(process killed but not fully unload-reset) otherwise left /status misreporting
tensor parallelism and made an identical diffusion re-Apply reload against the
stale state.
- tensor_split: reject negative / non-finite / all-zero splits up front. They
were dropped at launch but still compared raw in the reload dedupe, so an
identical Apply reloaded indefinitely.
- Tests: the shared httpx stub was incomplete and, installed via setdefault
before real httpx loaded, broke a combined pytest run (collection errors on
httpx.Response). Import the real installed httpx instead.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: danielhanchen <unslothshared@gmail.com>
Co-authored-by: danielhanchen <danielhanchen@gmail.com>
* Studio: hide infra models from the hub cached inventory
The hub inventory scans behind /api/hub/cached-gguf and /api/hub/cached-models
returned the llama.cpp install validation probe (ggml-org/models) and the RAG
embedder (unsloth/bge-small-en-v1.5[-GGUF]) as on-device models. Share the
hidden-model check from routes/models.py via utils/models/hidden_models.py and
apply it in both scans. A GGUF infra repo stays visible when the user
explicitly downloaded a variant through the Hub, since variant manifests only
exist for user-initiated downloads.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Studio: make On Device trust the hub inventory, match repo ids exactly, lighten the hidden-model import
Follow-up on the hub cached-inventory hidden-model change, addressing the review.
On Device now trusts the Hub inventory API for cached rows. The backend already
hides the RAG embedder and the llama.cpp probe and re-includes a GGUF infra repo
once the user downloads a variant through the Hub, but the frontend was
re-hiding it by repo id, so the user-downloaded variant never appeared in the On
Device list or the count. isVisibleInventoryRow now short-circuits cached rows
(kind === "cache") to visible and keeps client-side needle hiding only for local
filesystem rows and Discover.
is_hidden_model matches Hub repo ids exactly (case-insensitive) against the probe
plus the effective embedder and its GGUF companion, instead of substring
matching the configured-embedder basename. A custom embedder with a generic
basename like org/model no longer hides unrelated cached repos such as
user/model-chat or org/model-instruct. The probe filename and local-path
embedders keep exact matching.
The helper moves to utils/hidden_models.py and is imported at module scope in the
hub cache scanner, so it no longer pulls in utils/models/__init__ (the eager
model-config/checkpoint stack) and a broken import fails at startup instead of
being swallowed per-repo and silently emptying the inventory. routes.models
keeps the _is_hidden_model and _safe_resolve aliases and drops the unused
_HF_REPO_ID_RE re-export that was failing source lint.
Tests: exact repo-id matching with a custom embedder, the cached-models scan
keeping an unrelated repo, and a clean-interpreter check that the helper imports
without the model-config stack.
* Studio: match the llama.cpp probe filename on both path separators
The hidden-model check compared the probe's on-disk filename with
Path(value).name, which on a POSIX interpreter does not split a Windows-style
path ("...\stories260K.gguf") and would let the probe through. Split on both
separators so the probe is matched regardless of which OS produced the path,
matching the tolerance of the previous substring check. Adds a Windows-path
assertion to the probe test.
* Studio: harden hidden infra model handling
* Fix hidden cache row confirmation
* Fix hidden local rows and confirmed hint merges
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Handle snapshot-configured hidden models
* Hide basename-only default embedders
* Fix dynamic embedder inventory filtering
* Studio: hide the configured RAG embedder from Discover and feed rows
---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: danielhanchen <unslothshared@gmail.com>
Co-authored-by: Daniel Han <23090290+danielhanchen@users.noreply.github.com>
* Replace standalone Studio wording with Unsloth
Replace the single word Studio with Unsloth wherever it is used as
shorthand for Unsloth Studio in docs, CLI output, UI strings, i18n
locales, workflow display names, comments and docstrings.
Kept unchanged: the full name Unsloth Studio, third party product
names (LM Studio, Visual Studio, Mac Studio), feature names
(Recipe Studio, Fine-tuning Studio and its translations), and all
identifiers such as env vars, commands, paths and filenames.
* Address review feedback on the Studio wording rename
Use "an" before Unsloth where the rename left the article as "a".
Restore the split brand where Unsloth and Studio render as two halves
of the full product name: the onboarding sidebar subtitle and the
IPv6 localhost warning. Scope two messages to the full name Unsloth
Studio where plain Unsloth was misleading: the AMD README bullet and
the CLI studio setup error.
* fix(studio): add MLX adapter state control
* fix(studio): honor MLX adapter comparison state
* fix(studio): keep enabled MLX adapters permissive
* Studio: preserve public error message on MLX compare-mode adapter failures
generate_with_adapter_control raised a plain RuntimeError, which the compare
route handled with the generic handler that drops the operational message.
Raise GenStreamErrorRaised(public=chunk.public) instead and catch it in the
streaming and non-streaming consumers, matching the safetensors tool loop, so
errors like 'model is being unloaded' surface their real message.
* Studio: re-emit VLM think prefill inside the adapter context
The compare-mode merge dropped _generate_vlm's upfront yield of the prefilled
<think> block. Restore it as the first snapshot inside the lock+adapter context
(matching _generate_text) so the UI renders the thinking block during prefill
and a cancel/error before the first token does not drop it. Adds a regression
test asserting the prefill is emitted first, after entering the adapter context.
---------
Co-authored-by: danielhanchen <unslothshared@gmail.com>
The Connections form hid the API key field for the Ollama preset, which
blocked Ollama cloud (it requires a key). Show the optional field for
Ollama; the backend already sends Authorization: Bearer when a key is
set and omits the header when empty, so local keyless servers are
unaffected.
Fixes#7163
Long Yixing