Repository navigation
autotrainer 0.16.0
·
4 commits
to 70c6fbf40538fc5c27ba86d7344a5f903fe25062
since this release
[0.16.0] - 2026-08-20
Added
- Hardware Telemetry Curves & Resource Diagnostics (
#hwChart):- Live system resource meter footer with green pulse status indicator in sidebar.
- Dedicated hardware telemetry card displaying GPU VRAM (GB / %), GPU model name, and Host RAM / CPU utilization curves over training steps and epochs.
- REST endpoint
GET /api/hardwarewith cross-platform memory and GPU detection.
- Run Tagging, Star Favoriting & Sidebar Organization:
- Star favorite toggle (⭐) on individual runs with persistent metadata state.
- Sidebar filter tabs:
[ All ]and[ ⭐ Starred ]with glow accents. - Inline run tag chip management:
+ Tagcreator,✕tag pill deletion, and dynamic sidebar tag filter pills bar. - REST endpoints:
POST /api/runs/<run_id>/tagsandPOST /api/runs/<run_id>/favorite.
- Hyperparameter & Metric Parallel Coordinates Chart (
#parallelCanvas):- High-performance Canvas multi-axis parallel coordinates visualization in Compare Mode.
- Visualizes complex relationships and trajectories across
Learning Rate->Batch Size->First Loss->Training Loss->Validation Lossacross multiple selected runs.
- Run Annotations & Markdown Notes Drawer (
#notesCard):- Embedded Markdown drawer with
Editand livePreviewrendering tabs. - Auto-persisted to both
run.jsonand<run_dir>/notes.mdon disk. - REST endpoint
POST /api/runs/<run_id>/notes.
- Embedded Markdown drawer with
- 1-Click Run Deletion & Archive:
- Clean deletion of experiment runs and log directories from disk with modal confirmation.
- REST endpoints
DELETE /api/runs/<run_id>andPOST /api/runs/<run_id>/archive.
- Refined KPI Stat Cards:
- Displays
FIRST LOSS,TRAINING LOSS,VALIDATION LOSS, andEPOCHSwith glowing accents.
- Displays
Security
The Web UI added in 0.15.0 served a filesystem-mutating API with no
authentication on every network interface. Each item below was confirmed by
exploiting a running server, and is now pinned by a test in
tests/test_ui_security.py.
- Unauthenticated remote directory deletion.
run_idwas joined onto the
logs root with no containment check, soDELETE /api/runs/../<path>
resolved outside it and reachedshutil.rmtree. A live server answered
{"success": true}and the directory was gone. All six lookups — plus a
seventh inlined in_get_run_dataand two more behind the HTML/Markdown
export writes, which made it an arbitrary-write primitive too — now go
through one_safe_run_dir()that rejects separators and anything
resolving outside a configured root. autotrainer uibound0.0.0.0. On a shared SLURM login node that
published rename/archive/delete to every user who could route to the port.
It now binds127.0.0.1;--hostwidens it deliberately and prints a
warning naming what is exposed.- No authentication. A session token is now generated per run and printed
as part of the URL, carried afterwards by anHttpOnly; SameSite=Strict
cookie (also accepted as?token=orX-Autotrainer-Token) and compared
withsecrets.compare_digest.--no-tokenopts out.SameSiteplus an
Origincheck on mutating verbs closes the CSRF hole that let any page a
user visited drive the delete endpoint. POST /api/sourcesaccepted any absolute path, turning the dashboard
into a filesystem browser — a home directory listed 45 "runs". It now
requires a directory that actually contains run folders.- Stored XSS via run tags. Tags were interpolated raw into
innerHTML,
including insideonclick="removeTagFromCurrentRun('${tag}')". Tags and
usernames are escaped,escapeHtmlnow covers quotes, and the tag remove
control binds a listener instead of building a JS string.
Fixed
AutotrainerCallback.on_epoch_end()crashed on its own defaults.
Omittingtrain_loss— normal for a callback that only has a validation
number — reachedfloat(None)and raisedTypeError, killing the run it
was meant to be observing.log_epochnow treats every metric as optional
and reads its three call shapes through one code path.optimize=Truesilently disabled per-epoch reshuffling.prepare()
wraps a sharded loader in_AutoEpochDataLoaderso
DistributedSampler.set_epoch()advances on its own, but theoptimize
loader-defaults step and theauto_bssweep each build a fresh
DataLoader, which the wrapper did not survive.set_epochwas then never
called and every epoch replayed one fixed permutation on every rank —
silently, with nothing raised. It fired on the most common configuration
there is: multi-GPU,optimize=True(the default), and an untuned loader,
sincebuild_loader_defaults()only has work to do whennum_workersis 0
orpin_memoryis off. The wrapper is now applied once, after every
rebuild. Loaders the user sharded themselves are still passed through
unwrapped, andshuffle=Falsestill means a fixed order.train()synced the GPU twice per step.loss.item()was called once
for the running total and again for the summary; each is a device sync, so
the loop stalled twice per iteration in the entry point whose selling point
is throughput. Read once, reused.
Changed
- Docs now describe the auto-managed epoch shuffling that 0.14.0 added.
docs/guide/gpu-optimization.mdstill showed the pre-0.14.0 log line
(- call autotrainer.set_epoch(loader, epoch) ...) where the code prints
- epoch shuffling auto-managed, and the guides still presented the manual
call as required. They now say it is optional for loadersprepare()
sharded, and — worth knowing — that the automatic counter overwrites an
explicit epoch number, so resuming at a specific epoch needs the loader you
sharded yourself.set_epoch's own docstring says the same. - Removed the
suhasandsuhas-goravale-siddaramuoptional-dependency
groups frompyproject.toml. Both were["psutil"]and shipped as real
installable extras (pip install autotrainer[suhas]) in the published
metadata. Theautotrainer suhasCLI credits command is unrelated and
stays. ThroughputMonitoris now in__all__. It was documented in
docs/guide/monitors.mdwith a worked example while absent from__all__,
which the README declares the boundary of the stable public API — so the
docs taught something the policy called internal. Also added to the README's
entry-point table.tests/test_docs.pynow checks this direction too; it
only ever verified that__all__members appear in the docs, never that
documented names are exported.