Skip to content

Commit cfcbddf

Browse files
meizhong986claude
andcommitted
fix: translation temp file overwrites existing subtitles (#259), semantic diag script (#267), tracker rev22
- core.py: set project.subtitles.outputpath immediately after init_project() so ALL intermediate saves (SaveProject after each batch) go to the user's desired output location. Without this, PySubtrans defaults to writing in the input directory — either as *.translated.srt (new project) or *.{target_lang}.srt (resumed from .subtrans via UpdateOutputPath) — which can overwrite existing translations produced by a different provider. Resume capability via .subtrans is fully preserved. (#259) - test_semantic_diag.py: standalone diagnostic script for #267 semantic scene detection hang. Reproduces the exact librosa feature extraction steps outside WhisperJAV's subprocess architecture to isolate whether a hang is system-level (librosa/numpy/BLAS) or subprocess-specific. Accepts a media file as CLI argument, prints step-by-step progress with timing at each stage. - Issue tracker rev22: #267 diagnostic log analyzed (hang at librosa.feature.mfcc, all versions compatible), standalone test sent. #259 new bug (temp file overwrite), fix coded. #272 fix coded, awaiting confirmation. #271 awaiting model info. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
1 parent e50df15 commit cfcbddf

3 files changed

Lines changed: 298 additions & 58 deletions

File tree

docs/ISSUE_TRACKER_v1.8.x.md

Lines changed: 65 additions & 58 deletions
Original file line numberDiff line numberDiff line change
@@ -1,6 +1,6 @@
11
# WhisperJAV Issue Tracker — v1.8.x Cycle
22

3-
> Updated: 2026-03-31 (rev21) | Source: [GitHub Issues](https://github.com/meizhong986/WhisperJAV/issues) | **55 open** on GitHub
3+
> Updated: 2026-04-01 (rev22) | Source: [GitHub Issues](https://github.com/meizhong986/WhisperJAV/issues) | **55 open** on GitHub
44
55
---
66

@@ -21,13 +21,12 @@
2121

2222
| Category | Count | Notes |
2323
|----------|------:|-------|
24-
| Total open on GitHub | **55** | Was 54 at rev20. New: #271, #272. Closed: #269. Net +1. |
24+
| Total open on GitHub | **55** | Unchanged from rev21. No new issues, no closures. |
2525
| **v1.8.10 RELEASED** || 2026-03-30. 39 commits. |
26-
| **FIX CODED (on main, not released)** | 2 | Hallucination filter bundle, Colab install |
27-
| **NEW since rev20** | 2 | #271 (Ollama variant error), #272 (XXL model flag ignored) |
28-
| **NEEDS RESPONSE** | 2 | #271, #272 |
29-
| **NEEDS FOLLOW-UP** | 2 | #265, #267 (debug logs arrived) |
30-
| **AWAITING REPLY** | 5 | #234, #251, #258, #260, #261 |
26+
| **FIX CODED (on main, not released)** | 3 | Hallucination filter, Colab install, XXL model flag (#272) |
27+
| **NEEDS RESPONSE** | 0 | All cleared. |
28+
| **NEEDS FOLLOW-UP** | 1 | #259 (temp file overwrite bug) |
29+
| **AWAITING REPLY** | 7 | #234, #251, #258, #260, #261, #267, #271 |
3130
| Feature requests (open) | 29 | See Feature Requests section |
3231

3332
---
@@ -44,11 +43,11 @@ These fixes are committed to main after v1.8.10 was tagged. Colab/pip users get
4443

4544
| Commit | Item | Issues | Risk |
4645
|--------|------|--------|------|
46+
| `e50df15` | XXL model flag not forwarded to subprocess; semantic scene detect diagnostics v7.1.0 | #272, #267 | LOW — XXL arg fix + diagnostic logging |
4747
| `1bfe548` | Hallucination filter: bundled JSON missing from wheel, stale gist URLs, INFO logging | #265 | LOW — packaging + logging |
4848
| `4a600b2` | `--scene-detection-method` ignored in standard pipeline modes | #269 (closed) | LOW — one-line injection |
4949
| `2533438` | Docs: GUI user guide updated for Ollama + enhance-for-VAD || LOW — docs only |
5050
| `a6b473d` | Colab install: remove venv, reuse system torch, fix llama-cpp filename | #231 | LOW — Colab-tested |
51-
| `a6b473d` | Colab/Kaggle notebooks: remove venv references | #231 | LOW — config changes |
5251
| `f875303` | Issue tracker rev20 |||
5352

5453
---
@@ -59,14 +58,14 @@ These fixes are committed to main after v1.8.10 was tagged. Colab/pip users get
5958

6059
| # | Title | Reporter | Status | Notes |
6160
|---|-------|----------|--------|-------|
62-
| **#272** | XXL ignores --pass2-model large-v2 flag | TinyRick1489 | `NEEDS RESPONSE` | **NEW.** Confirmed bug with debug log. WhisperJAV logs model=large-v2 but XXL subprocess loads large-v3. Flag not forwarded to XXL command line. |
63-
| **#271** | Ollama "Could Not Create Optimized Variant" | justantopair-ai | `NEEDS RESPONSE` | **NEW.** qwen3:14b error. Also reports gist timeout fetching instructions_standard.txt, plot summary cutoff in GUI, translation errors. 3 screenshots. |
64-
| **#267** | Stuck on streaming features (Qwen+Semantic) | OrangeFarmHorse | `NEEDS FOLLOW-UP` | **Debug logs arrived.** All 3 test runs (short clip, long file, balanced pipeline) hang at exact same point: "[1/5] Streaming features (60s chunks)..." inside SemanticAudioClustering v7.0.0. File length irrelevant. RTX 5060 Ti (Blackwell). |
65-
| **#265** | Hallucination in Chinese translation (large-v2) | yangming2027 | `NEEDS FOLLOW-UP` | User provided 3 concrete post-processing suggestions: (1) trailing period inconsistency, (2) leading dash removal, (3) long subtitle line wrapping. Sample SRT files attached. Also requesting merge module (#230). |
61+
| **#272** | XXL ignores --pass2-model large-v2 flag | TinyRick1489 | `FIX CODED` | Fix on main (`e50df15`). Responded with upgrade instructions. Awaiting user confirmation. |
62+
| **#271** | Ollama "Could Not Create Optimized Variant" | justantopair-ai | `AWAITING REPLY` | Asked which model. Non-fatal variant warning + translation format mismatch — likely model not suited for translation. |
63+
| **#267** | Stuck on streaming features (Qwen+Semantic) | OrangeFarmHorse | `AWAITING REPLY` | Hang pinpointed at `librosa.feature.mfcc()`. Standalone test script (`test_semantic_diag.py`) sent to user to isolate system vs subprocess issue. Waiting on results. |
64+
| **#265** | Hallucination in Chinese translation (large-v2) | yangming2027 | `NEEDS FOLLOW-UP` | 3 post-processing suggestions with sample SRTs: trailing period, leading dash, line wrapping. Also shared Subtitle Edit multi-replace XML. |
6665
| **#263** | GPU not utilized / stuck at VAD | herlong6529424-dot | `AWAITING REPLY` | RTX 3070 8GB. Process hangs during Silero VAD. Asked for debug log + short file test. No response yet. |
6766
| **#261** | Network check: unknown url type https | henry99a | `AWAITING REPLY` | NSIS SSL context issue. User found workaround (run from cmd). |
6867
| **#260** | Uninstall leaves 6GB | hawai-stack | `AWAITING REPLY` | Responded with cache paths. |
69-
| **#259** | Local Translation Issues (hotfix2) | destinyawaits | `RESPONDED` | Responded with v1.8.10 Ollama upgrade instructions. |
68+
| **#259** | Local Translation Issues (hotfix2) | destinyawaits | `NEEDS FOLLOW-UP` | **Ollama confirmed working.** New bug: temp file overwrites existing English subtitles in same folder — save path not respected until task complete. |
7069
| **#258** | 我遇到的问题 (vague) | Uillike | `AWAITING REPLY` | Asked for logs. |
7170
| **#255** | 如何用ollama进行翻译 | cheny7918 | `RESPONDED` | Responded with v1.8.10 Ollama upgrade instructions. 3 users. |
7271
| **#251** | post2 launch failure (fastgit SSL) | zoqapopita93 | `AWAITING REPLY` | Diagnosed: fastgit mirror SSL failure. Provided 3 solutions. |
@@ -130,16 +129,13 @@ These fixes are committed to main after v1.8.10 was tagged. Colab/pip users get
130129

131130
| Action | Issues | Status |
132131
|---|---|---|
133-
| **Respond to #272** (XXL model flag bug — confirmed with debug log) | #272 | **TODO** |
134-
| **Respond to #271** (Ollama variant error — 3 screenshots, needs diagnosis) | #271 | **TODO** |
135-
| **Investigate #267** (debug logs available — semantic hang at streaming features) | #267 | **TODO** — logs ready for code analysis |
132+
| ~~Analyze #267 diagnostic log~~ (standalone test script sent) | #267 | **DONE** — awaiting user test results |
133+
| **Respond to #259** (temp file overwrite bug — new report from user) | #259 | **TODO** |
136134
| Assess #265 post-processing suggestions (trailing period, leading dash, line wrap) | #265 | TODO — v1.9 scope |
137-
| ~~Respond to #269~~ (scene-detect fix coded) | #269 | **DONE** — issue closed |
138-
| ~~All NEEDS RESPONSE cleared~~ || **DONE** at rev20. 2 new NEEDS RESPONSE since. |
135+
| ~~Respond to #272~~ (fix coded, upgrade instructions given) | #272 | **DONE** — awaiting user confirmation |
136+
| ~~Respond to #271~~ (asked which model, explained variant warning) | #271 | **DONE** — awaiting user reply |
137+
| ~~Respond to #267~~ (added diagnostic logging, asked to upgrade and retest) | #267 | **DONE** — user responded with diagnostic log |
139138
| ~~Follow up #263~~ (requested debug log + short file test) | #263 | **DONE** — waiting on user |
140-
| ~~Follow up #267~~ (requested debug log + short file test) | #267 | **DONE** — user responded with 4 log files |
141-
| ~~Push commits to main~~ || **DONE** |
142-
| ~~Test Colab install fix on Colab~~ || **DONE** |
143139

144140
---
145141

@@ -151,24 +147,19 @@ Colab/pip users get these automatically. .exe installer users get them via `whis
151147

152148
| Item | Commit | Status |
153149
|---|---|---|
150+
| XXL model flag fix (#272) + semantic diagnostics v7.1.0 (#267) | `e50df15` | Pushed |
154151
| Hallucination filter bundled in wheel | `1bfe548` | Pushed |
155152
| --scene-detection-method fix (#269) | `4a600b2` | Pushed. #269 closed. |
156153
| Docs: GUI guide for Ollama + enhance-for-VAD | `2533438` | Pushed |
157-
| Issue tracker rev20 | `f875303` | Pushed |
158154
| Colab install: remove venv, reuse system torch | `a6b473d` | Pushed. Colab-tested. |
159-
| Colab/Kaggle notebook: remove venv refs | `a6b473d` | Pushed. Colab-tested. |
160155

161-
### v1.8.10-hotfix — If Needed
156+
### v1.8.10-hotfix — Not Currently Needed
162157

163-
**Candidate bugs:**
158+
All fixes are on main and accessible via `pip install --no-deps --upgrade`. No blocking bugs requiring a versioned hotfix release. The #272 XXL fix and #267 diagnostics are already available to pip/Colab users.
164159

165-
| Bug | Issue | Severity | Hotfix? |
166-
|---|---|---|---|
167-
| XXL ignores --pass2-model flag | #272 | MEDIUM | **Yes — if fix is small.** Confirmed bug, user on Kaggle CLI. |
168-
| Semantic scene detect hangs on Blackwell GPUs | #267 | HIGH | **No.** Needs investigation first. Workaround exists (auditok). |
169-
| Ollama optimization variant error | #271 | LOW | **No.** Likely Ollama-side issue, not WhisperJAV bug. |
170-
171-
**Recommendation**: Investigate #272 first. If the fix is a one-liner (model flag not forwarded to XXL subprocess args), it's a good hotfix candidate alongside the already-pushed fixes.
160+
A hotfix would only be needed if:
161+
- A blocking bug is found that affects .exe installer users specifically
162+
- The #267 semantic hang turns out to be a code bug (currently investigating)
172163

173164
### v1.9.0 — Next Major Release
174165

@@ -179,6 +170,7 @@ Colab/pip users get these automatically. .exe installer users get them via `whis
179170
| **P0** | Ollama full migration (deprecate llama-cpp-python) | #132, #233, #255, #259 | Large | Remove ~1500 LOC. llama-cpp already broken on Colab. |
180171
| **P0** | Chinese UI (partial i18n) | #175, #180 | Medium | 40%+ issues are Chinese. Biggest support burden reducer. |
181172
| **P0** | Unified CLI override layer | #269 | Small | Fix architectural divergence between standard/ensemble override paths. |
173+
| **P1** | Translation temp file overwrite fix | #259 | Small | Save path not respected until task complete — overwrites existing SRTs. |
182174
| **P1** | Post-translation hallucination filter (Chinese) | #265 | Medium | User contributed 130-rule XML. Automate as post-process option. |
183175
| **P1** | Post-processing polish (trailing period, leading dash, line wrapping) | #265 | Small | 3 concrete suggestions from yangming2027 with sample SRTs. |
184176
| **P1** | Standalone merge CLI | #230 | Medium | `whisperjav-merge`. Active demand (yangming2027, weifu8435). |
@@ -187,7 +179,7 @@ Colab/pip users get these automatically. .exe installer users get them via `whis
187179
| **P1** | AMD ROCm support | #142, #114, #239 | Medium | Document + test. |
188180
| **P1** | Additional translation targets | #268 | Small | Thai, Korean requested. May be PySubtrans config only. |
189181
| **P1** | Full dual-track enhance-for-VAD || Medium | ASR module separate VAD/ASR audio paths. |
190-
| **P2** | Semantic scene detection hanging investigation | #267 | Medium | Hangs at "[1/5] Streaming features" on Blackwell GPU. Debug logs available. Needs code-level investigation. |
182+
| **P2** | Semantic scene detection: librosa.feature.mfcc hang | #267 | Medium | Diagnostic log pinpointed hang at first `librosa.feature.mfcc()` call. Versions all compatible. Suspected numba JIT issue — needs further investigation. |
191183
| **P2** | GUI settings persistence | #96 | Medium | Long-standing request. |
192184
| **P2** | Vocal separation | #224, #254 | Medium | BS-RoFormer / UVR integration. |
193185
| **P2** | Grey out incompatible options | #206 | Small | Prevent invalid GUI combos. |
@@ -222,65 +214,80 @@ Colab/pip users get these automatically. .exe installer users get them via `whis
222214

223215
## Emerging Patterns
224216

225-
### Semantic Scene Detection Hanging (#267) — DEBUG LOGS AVAILABLE
217+
### Semantic Scene Detection Hanging (#267) — PINPOINTED
218+
219+
**Symptom**: Process hangs at `librosa.feature.mfcc()` — the first librosa feature extraction call inside SemanticAudioClustering.
226220

227-
**Symptom**: Process hangs at "[1/5] Streaming features (60s chunks)..." inside SemanticAudioClustering v7.0.0.
221+
**Diagnostic log (2026-04-01, v7.1.0)**:
222+
```
223+
[Semantic Diag] numpy=2.2.6 librosa=0.11.0 numba=0.64.0 soundfile=0.13.1
224+
[Semantic Diag] ffmpeg: ffmpeg version 7.1.1
225+
[Semantic Diag] All version checks passed.
226+
[1/5 diag] Step 1 OK: WAV PCM_16, 16000Hz, 1ch, 120.0s (0.00s)
227+
[1/5 diag] Step 2 OK: 120.0s, 16000Hz, block_size=960000, total_blocks=2 (0.00s)
228+
[1/5 diag] Step 3 OK: First block received (shape=(960000, 1), 0.00s)
229+
[1/5 diag] Step 5: librosa feature extraction (first chunk)...
230+
← HANGS HERE
231+
```
228232

229-
**Evidence from 4 log files (2026-03-31)**:
230-
- 2-minute clip hangs identically to 1.9-hour file → **file length is NOT the cause**
231-
- Both Qwen and Balanced pipelines hang → **pipeline is NOT the cause**
232-
- Auditok works perfectly on same files → **audio content is NOT the cause**
233-
- `SemanticClusteringAdapter initialized` successfully — engine loads, hangs during feature extraction
234-
- RTX 5060 Ti (Blackwell architecture) — possibly GPU/driver-specific
235-
- `--debug` flag produces no additional output from inside the streaming step
233+
**What is ruled out**:
234+
- numpy/librosa/numba version incompatibility (all versions compatible, checks passed)
235+
- soundfile/libsndfile issues (WAV reads fine, 0.00s)
236+
- ffmpeg issues (not involved — audio is already 16kHz WAV)
237+
- File length (2min = 1.9h hang identically)
238+
- Pipeline (Qwen + Balanced both hang)
239+
- Audio content (auditok processes same files fine)
240+
- Resampling (native SR = target SR, no resample triggered)
236241

237-
**Hypothesis (unverified)**: The feature streaming step may use a GPU operation (torch audio feature extraction) that is incompatible with the Blackwell/RTX 50 series drivers or CUDA version. The hang occurs inside a library call, not in WhisperJAV code.
242+
**What remains**: The hang is at the first `librosa.feature.mfcc()` call. This function uses numpy/scipy for FFT and DCT — no numba directly. However, librosa may trigger numba JIT compilation through internal utility imports. numba 0.64.0 is very recent and could have platform-specific issues.
238243

239-
**Next step**: Read the semantic scene detection source code to identify what "[1/5] Streaming features" does — what library calls, what GPU operations. Determine if there's a known Blackwell compatibility issue.
244+
**Next diagnostic step**: Ask user to test with `NUMBA_DISABLE_JIT=1` environment variable to confirm/eliminate numba as the cause.
240245

241246
### Silero VAD Hanging (#263)
242247

243-
**Symptom**: Process hangs during Silero VAD initialization/inference. Log shows "Starting speech segmentation with: silero-v4.0" with no "complete" message. User has RTX 3070 8GB, processing a 2.4h file (314 scenes).
244-
**Status**: Waiting on user debug log + short file test. No response since 2026-03-31 request.
248+
**Symptom**: Process hangs during Silero VAD. Log shows "Starting speech segmentation with: silero-v4.0" with no "complete" message. RTX 3070 8GB.
249+
**Status**: Waiting on user debug log + short file test. No response since 2026-03-31.
245250

246-
### XXL Model Flag Ignored (#272) — CONFIRMED BUG
251+
### Translation Temp File Overwrite (#259) — NEW BUG
247252

248-
**Symptom**: `--pass2-model large-v2` is logged correctly by WhisperJAV but not forwarded to the Faster-Whisper-XXL subprocess. XXL loads large-v3 by default.
249-
**Evidence**: Debug log shows `[Worker 3816] Pass 2: BYOP XXL (exe=..., model=large-v2)` but XXL subprocess output says `'large-v3' model may produce worse results`.
250-
**Root cause**: The `--model` flag is likely missing from the XXL subprocess command construction in `pass_worker.py` or `xxl_runner.py`.
253+
**Symptom**: During AI SRT Translate, the temp file is created in the same folder as the input SRT and overwrites existing English subtitles generated earlier (e.g., by DeepSeek). The user-specified save path is not used until the task completes.
254+
**Reporter**: destinyawaits (also confirmed Ollama working after v1.8.10 upgrade).
255+
**Status**: Needs investigation — is this PySubtrans behavior or WhisperJAV's translation wrapper?
251256

252-
### Hallucination in Chinese Translation (#265)
257+
### Ollama Model Suitability (#271)
253258

254-
**Symptom**: After Japanese transcription + DeepSeek translation to Chinese, hallucination phrases appear in Chinese output.
255-
**Assessment**: Two contributing factors: (1) hallucination filter was not bundled in wheel — fix committed; (2) no post-translation sanitization step exists. User provided 3 concrete post-processing suggestions with sample SRTs.
259+
**Symptom**: Translation format errors with Ollama. Model output doesn't match PySubtrans expected format.
260+
**Assessment**: Likely model not suited for structured translation tasks. Asked user which model they're using. Non-fatal "optimized variant" warning is separate (Ollama rejected Modelfile create with HTTP 400).
256261

257262
---
258263

259264
## Duplicate / Related Issue Map
260265

261266
| Cluster | Issues | Primary | Action |
262267
|---------|--------|---------|--------|
263-
| **Local LLM / Ollama** | #271, #259, #255, #233, #132 | #132 | v1.8.10 ships Ollama GUI. #271 new (variant error). |
268+
| **Local LLM / Ollama** | #271, #259, #255, #233, #132 | #132 | #259 Ollama working, new temp file bug. #271 awaiting model info. |
264269
| **Diarization** | #248, #252 | #248 | v1.9+. Both responded. |
265270
| **AMD/Intel GPU** | #239, #142, #114, #213 | #142 | v1.9+. |
266271
| **i18n** | #175, #180 | #180 | v1.9.0 P0. |
267272
| **Speech enhancement** | #254, #224 | #224 | v1.9. |
268-
| **XXL** | #272, #242, #237, #223 | #242 | #272 is a bug (model flag). #242 is feature (Pass 1). |
273+
| **XXL** | #272, #242, #237, #223 | #242 | #272 fix coded. #242 is feature (Pass 1). |
269274
| **Install/Network** | #261, #251, #243, #240, #225, #217 || Individual. |
270275
| **Model management** | #264, #250 | #250 | Docs / FAQ. |
271276
| **Translation targets** | #268 | #268 | Thai + Korean. v1.9+. |
272-
| **Scene detection** | #267 || Debug logs available. Needs code investigation. |
277+
| **Scene detection** | #267 || Hang pinpointed at librosa.feature.mfcc(). Next: test NUMBA_DISABLE_JIT=1. |
273278
| **Hallucination** | #265, #246 | #265 | Post-translation filter for v1.9. |
274279
| **Merge module** | #230, #265 | #230 | v1.9.0 P1. yangming2027 strong advocate. |
280+
| **Translation UX** | #259 | #259 | Temp file overwrite bug. New. |
275281

276282
---
277283

278284
## Changelog
279285

280286
| Date | Changes |
281287
|------|---------|
282-
| **2026-03-31** | **rev21.** 54→55 open. New: #271 (Ollama variant error), #272 (XXL model flag bug). Closed: #269. #267 debug logs arrived — hangs at "[1/5] Streaming features" regardless of file length/pipeline, Blackwell GPU. #272 confirmed with debug log — model flag not forwarded to XXL subprocess. |
283-
| 2026-03-31 | **rev20.** 55→54 open. New: #265, #267, #268, #269. Closed: #218, #221, #236, #244. All fixes pushed to main (4 commits). Colab tested. All NEEDS RESPONSE cleared. |
288+
| **2026-04-01** | **rev22.** 55 open (unchanged). Committed `e50df15`: XXL model flag fix (#272) + semantic diagnostics v7.1.0 (#267). Responded to #271, #272, #267. #267 diagnostic log arrived — hang pinpointed at `librosa.feature.mfcc()` first call. All library versions compatible. Next step: test NUMBA_DISABLE_JIT=1. #259 user confirmed Ollama working + reported new temp file overwrite bug. All NEEDS RESPONSE cleared. |
289+
| 2026-03-31 | **rev21.** 54→55 open. New: #271, #272. Closed: #269. #267 debug logs arrived. #272 confirmed. |
290+
| 2026-03-31 | **rev20.** 55→54 open. New: #265, #267, #268, #269. Closed: #218, #221, #236, #244. All NEEDS RESPONSE cleared. |
284291
| 2026-03-30 | **rev19.** v1.8.10 released. 54 open. #253 closed. |
285292
| 2026-03-29 | **rev18.** 55 open. Session work: aggressive retune, Ollama GUI, bug fixes. |
286293
| 2026-03-28 | **rev17.** Closed 11. Responded 7. |

0 commit comments

Comments
 (0)