Repository navigation
update venv_create.bat so it works a little smoother - #1
Merged
Merged
Conversation
joeblowma
commented
Oct 3, 2026
joeblowma
commented
Owner
- codex reworked py version detection so it's hopefully less likely to break again
- added a bunch of (likely ineffective) errorlevel checks, probably breaks somewhere for someone...
- added venv_delete to clean up an installed venv (making it easier to debug prior point)
- made venv_activate static rather than generated
- set it up so venv is set consistently with installation options
- asked myself multiple times... cmd shell script?- codex reworked py version detection so it's hopefully less likely to break again
- added a bunch of (likely ineffective) errorlevel checks, probably breaks somewhere for someone...
- added venv_delete to clean up an installed venv (making it easier to debug prior point)
- made venv_activate static rather than generated
- set it up so venv is set consistently with installation options
- asked myself multiple times... cmd shell script?
- codex reworked py version detection so it's hopefully less likely to break again - added a bunch of (likely ineffective) errorlevel checks, probably breaks somewhere for someone... - added venv_delete to clean up an installed venv (making it easier to debug prior point) - made venv_activate static rather than generated - set it up so venv is set consistently with installation options - asked myself multiple times... cmd shell script?
revert some UI up/down arrows changed from a prior commit
- [x] Add a CLI option such as `--write-modelinfo-json`. - [x] Reuse the existing `.modelinfo` naming schema and append `.json`, for example `model.safetensors.modelinfo.json`. - [x] Serialize existing inspection output with the standard `json` module unless a stronger need appears. - [x] Add matching GUI setting/action for JSON dump generation. - [x] Verify output is stable, pretty-printed, and contains parsed metadata, tensor summaries, architecture details, and warnings.
- [x] Add a reader abstraction that can dispatch by extension: `.safetensors` first, then `.gguf`. - [x] Use the installed `gguf` package from `.venv\Lib\site-packages\gguf`. - [x] Review `.venv\Lib\site-packages\gguf\scripts\gguf_editor_gui.py` for PySide/PyQt patterns, metadata parsing, and tensor listing behavior. - [x] Implement GGUF metadata extraction, tensor summaries, dtype counts, architecture hints, and model size reporting. - [x] Keep GGUF support read-only for now; do not expose editing behavior.
todo 3 - [x] Investigate how Windows symlinks behave with `Path.resolve()` and file IDs. - [x] Decide whether `.modelinfo` should be written beside the user-provided path or the resolved target path. - [x] Preserve the user-provided path for default `.modelinfo` output and expose `resolved_filepath` separately. todo 7 - [x] Add CLI and GUI smoke checks with at least one representative `.gguf` file.
- [x] Add a cache for parsed model summaries to speed up repeated launches. - [x] Store enough identity data to detect moved, deleted, or changed files, such as canonical path, size, modified time, and format. - [x] Default settings/cache to a portable app-local data directory with relocation overrides. - [x] Use a cache index with per-entry files instead of one monolithic cache blob.
add keyboard iteration of raw item list via up/down page up/down irgnore test/
- [x] Add an option if both behaviors are useful, for example `--resolve-output-path`. - [x] Handle failures conservatively and show the chosen output path in CLI/GUI results.
- [x] Add format filters for `.safetensors`, `.gguf`, `.ckpt`, `.pt`, and `.pth` in a separated section of the filters popup. - [x] Keep architecture/type filters separate from file-format filters so users can combine both.
2 - [x] Replace vague progress percentages with visible metrics, such as files discovered, files parsed, bytes scanned, and current directory. - [x] Keep the progress bar visible during long directory scans and model-info loading. - [x] Surface cancellation and partial-results behavior for long scans. 9 - [x] Evaluate the official `safetensors` Python package as a replacement or supplement for the current custom header reader. - [x] Confirm whether `safe_open` can provide metadata, keys, tensor shapes, and dtypes without loading full tensor payloads or requiring Torch. - [x] Test `framework="numpy"` and metadata-only/key-only flows to avoid a full Torch dependency. - [x] If unsuitable, document why the custom header parser remains preferable.
5 - [x] Audit current `AnalysisWorker` behavior and CLI folder scanning for safe concurrency points. - [x] Add a setting/CLI option such as `--threads N` for batch inspection. - [x] Use bounded worker pools only for independent file reads; avoid parallel UI mutation. 9 - [x] Add a cache for scanned directories to speed up repeated launches. - [x] Add a `load default libraries on startup` setting backed by the cached directory list. - [x] Add optional startup pruning for missing files, with a warning before removing stale entries. - [x] Add a settings action to clear the cache entirely. - [x] Decide whether cache writes happen immediately after each scan or only after successful batch completion.
- [x] Research safe metadata-only handling for PyTorch checkpoint formats. - [x] Treat pickle-based formats as unsafe by default; require explicit opt-in before loading. - [x] Add clear warnings in CLI and GUI when a format may execute pickle deserialization.
- [x] Benchmark on large folders before enabling a default above one thread.
- [x] Sort filter menus with `ERROR` first, then `Unknown`, then alphabetical values. - [x] Keep card and Raw dropdown ordering aligned with the current Data table sort order. - [x] Organize settings into General, Cards, and Data Columns tabs.
- fixed in app icon on titlebar - updated splash screen to a bitmap - updated splash resizer to retain 0xFF00FF for transparencies - separated requirements into runtime/build(dev)
- rewrite splash converter bmp->resize->png - lint with ruff - start setting up to build wheels as well clean up unused imports
fixing up the model size display
- build targets confirmed working - switch filenames to use short 7 first char of sha - add mypy to requirements-dev.txt
- added lineclean.py - ran lineclean: cleanup of line end whitespace and line endings - restore some images to README.md
PR Summary by QodoExpand ModelInspector formats, GUI workflows, packaging, and releases
AI Description
Diagram
High-Level Assessment
Files changed (183)
|
Code Review by Qodo
1.
|
![qodo-code-review[bot]](https://avatars.githubusercontent.com/in/484649?s=60&v=4){kind=small}
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.