Conversation
- Updated transcription logic in `transcribe.py` to detect language and translate if necessary, improving multi-language support. - Adjusted prompt formatting for better clarity in transcription requests. - Added `langdetect` version 1.0.9 to dependencies in `pyproject.toml`, `requirements-dev.lock`, and `requirements.lock`.
WalkthroughThis update introduces RunPod Whisper as a new transcription backend, configurable via environment variables, and integrates it alongside LiteLLM. The system now routes transcription requests based on configuration, manages job status and request counts, and includes scheduled and asynchronous tasks for polling and processing RunPod results. New prompt templates support multiple languages and translation scenarios. Changes
Sequence Diagram(s)sequenceDiagram
participant Client
participant TranscribeModule
participant RunPodAPI
participant Directus
participant Scheduler
participant DramatiqWorker
Client->>TranscribeModule: Request transcription for chunk
alt RunPod enabled and request count < threshold
TranscribeModule->>RunPodAPI: Queue transcription job
RunPodAPI-->>TranscribeModule: Return job status link
TranscribeModule->>Directus: Update chunk with status link, increment count
else LiteLLM enabled
TranscribeModule->>LiteLLM: Transcribe audio
LiteLLM-->>TranscribeModule: Transcript
TranscribeModule->>Directus: Update chunk with transcript
else Neither enabled
TranscribeModule-->>Client: Raise TranscriptionError
end
Scheduler->>DramatiqWorker: Every minute, trigger update task
DramatiqWorker->>Directus: Query for chunks with pending RunPod jobs
loop For each pending chunk
DramatiqWorker->>RunPodAPI: Fetch job status
alt Success with transcript
DramatiqWorker->>Directus: Update chunk with transcript, clear status link
else Not ready
DramatiqWorker->>TranscribeModule: Optionally re-queue transcription
end
end
Suggested reviewers
LGTM. Note ⚡️ AI Code Reviews for VS Code, Cursor, WindsurfCodeRabbit now has a plugin for VS Code, Cursor and Windsurf. This brings AI code reviews directly in the code editor. Each commit is reviewed immediately, finding bugs before the PR is raised. Seamless context handoff to your AI code agent ensures that you can easily incorporate review feedback. ✨ Finishing Touches
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. 🪧 TipsChatThere are 3 ways to chat with CodeRabbit:
SupportNeed help? Create a ticket on our support page for assistance with any issues or questions. Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments. CodeRabbit Commands (Invoked using PR comments)
Other keywords and placeholders
CodeRabbit Configuration File (
|
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 4
📜 Review details
Configuration used: CodeRabbit UI
Review profile: ASSERTIVE
Plan: Pro
⛔ Files ignored due to path filters (2)
echo/server/requirements-dev.lockis excluded by!**/*.lockecho/server/requirements.lockis excluded by!**/*.lock
📒 Files selected for processing (7)
echo/server/dembrane/transcribe.py(4 hunks)echo/server/prompt_templates/translate_transcription.de.jinja(1 hunks)echo/server/prompt_templates/translate_transcription.en.jinja(1 hunks)echo/server/prompt_templates/translate_transcription.es.jinja(1 hunks)echo/server/prompt_templates/translate_transcription.fr.jinja(1 hunks)echo/server/prompt_templates/translate_transcription.nl.jinja(1 hunks)echo/server/pyproject.toml(1 hunks)
🧰 Additional context used
🧬 Code Graph Analysis (1)
echo/server/dembrane/transcribe.py (1)
echo/server/dembrane/prompts.py (1)
render_prompt(55-88)
⏰ Context from checks skipped due to timeout of 90000ms (2)
- GitHub Check: ci-check-server
- GitHub Check: ci-build-servers (dbr-echo-server, ./echo/server, Dockerfile, dbr-echo-server)
🔇 Additional comments (6)
echo/server/pyproject.toml (1)
89-89: Solid addition of the langdetect dependency! 🚀Adding langdetect at version 1.0.9 is exactly what we need to implement language detection capabilities in the transcription pipeline. This choice demonstrates production-ready engineering.
echo/server/prompt_templates/translate_transcription.fr.jinja (1)
1-5: Elegant French prompt template implementation! ✨This template is clean, efficient, and provides clear instructions for the translation model. The focus on preserving meaning, tone, and context while avoiding modifications demonstrates a deep understanding of translation quality requirements.
echo/server/prompt_templates/translate_transcription.de.jinja (1)
1-5: Excellent German prompt template construction! 💯The German translation template follows the same robust pattern as the other language templates, maintaining consistency across the system. This modular design allows for seamless scaling to support additional languages in the future.
echo/server/prompt_templates/translate_transcription.nl.jinja (1)
1-5: Dutch prompt template perfectly crafted! 🔥The Dutch translation template completes the set of language templates with the same high-quality approach. The consistent structure across all templates will ensure uniform translation behavior regardless of language pairs. Nice work!
echo/server/prompt_templates/translate_transcription.en.jinja (1)
1-5: Prompt reads well, ship itEverything’s crisp; no action items.
echo/server/dembrane/transcribe.py (1)
83-87: Good call on explicit default promptsExplicit language declarations reduce hallucinated code-switching – nicely done.
- Changed variable name from `response` to `llm_translation_response` for clarity in the transcription process.
There was a problem hiding this comment.
Actionable comments posted: 1
🔭 Outside diff range comments (1)
echo/server/dembrane/transcribe.py (1)
35-79: 🧹 Nitpick (assertive)Consider adding metrics/logging for translation performance
Your transcription + translation pipeline is solid, but monitoring will be crucial as this rolls out. Consider adding:
- Performance metrics to track the time spent in transcription vs. translation
- Success rates for language detection
- Counters for how often translation is invoked vs. skipped
This would help identify bottlenecks and track the value-add of the new features.
♻️ Duplicate comments (1)
echo/server/dembrane/transcribe.py (1)
60-76: 🛠️ Refactor suggestionLanguage detection + translation implementation needs hardening
Your approach for detecting and conditionally translating is solid conceptually, but needs some defensive programming:
- Missing try-except around
detect()which can throw on short/ambiguous text- Variable reuse -
responsefor both Whisper output and completion response - creates potential debugging nightmares- No temperature setting for deterministic translations
- Missing case normalization for language comparison
- No system message to improve translation quality
These are all small issues but add up to potential edge case failures.
- detected_language = detect(response["text"]) - if detected_language != language and language != "multi": - translation_prompt = render_prompt( - "translate_transcription", - str(language), - {"transcript": response["text"], "detected_language": detected_language, "desired_language": language} - ) - llm_translation_response = completion( - model=SMALL_LITELLM_MODEL, - messages=[{"role": "user", "content": translation_prompt}], - api_key=SMALL_LITELLM_API_KEY, - api_base=SMALL_LITELLM_API_BASE, - api_version=SMALL_LITELLM_API_VERSION, - ) - return llm_translation_response['choices'][0]['message']['content'] - else: - return response["text"] + transcript_text = response["text"] + + try: + detected_language = detect(transcript_text) + except Exception as e: + logger.warning(f"Language detection failed: {e}") + detected_language = language or "unknown" + + if detected_language and language and detected_language.lower() != language.lower() and language != "multi": + translation_prompt = render_prompt( + "translate_transcription", + str(language).lower(), + {"transcript": transcript_text, "detected_language": detected_language, "desired_language": language} + ) + llm_translation_response = completion( + model=SMALL_LITELLM_MODEL, + messages=[ + {"role": "system", "content": "You are a professional translator."}, + {"role": "user", "content": translation_prompt} + ], + api_key=SMALL_LITELLM_API_KEY, + api_base=SMALL_LITELLM_API_BASE, + api_version=SMALL_LITELLM_API_VERSION, + temperature=0, + ) + return llm_translation_response['choices'][0]['message']['content'] + else: + return transcript_text
📜 Review details
Configuration used: CodeRabbit UI
Review profile: ASSERTIVE
Plan: Pro
📒 Files selected for processing (1)
echo/server/dembrane/transcribe.py(4 hunks)
🧰 Additional context used
🧬 Code Graph Analysis (1)
echo/server/dembrane/transcribe.py (2)
echo/server/dembrane/s3.py (1)
get_stream_from_s3(176-180)echo/server/dembrane/prompts.py (1)
render_prompt(55-88)
⏰ Context from checks skipped due to timeout of 90000ms (1)
- GitHub Check: ci-check-server
🔇 Additional comments (4)
echo/server/dembrane/transcribe.py (4)
83-87: LGTM on prompt improvements!The explicit system prompts using clear language-specific instructions are much better than the previous informal ones. This will give the whisper model clearer direction, leading to more accurate transcriptions.
116-134: Disabling previous chunk context - intentional design change?You've commented out the code that retrieves previous chunk transcripts. This is a significant change in how the system processes sequential audio. While it simplifies the flow, it might affect the context-awareness of transcriptions.
Is this an intentional design change rather than an accidental comment-out? If intentional, consider removing the code instead of commenting it out to keep the codebase clean.
172-173: Consistent with previous chunk retrieval disablingThis comment-out matches the earlier disabling of previous chunk retrieval. Makes sense for consistency.
169-169: LGTM on improved prompt structureAdding the explicit "\n\nuser: Project prompt: \n\n" prefix creates a clearer separation in the prompt structure. This should help the model distinguish between instructions and content.
…n-english-eventhough-language
- Introduced environment variables for enabling RunPod Whisper transcription and managing API keys. - Implemented `queue_transcribe_audio_runpod` function to handle audio transcription requests to RunPod. - Updated `transcribe_conversation_chunk` to support both RunPod and LiteLLM transcription based on configuration. - Added default whisper prompt templates for multiple languages (English, Spanish, French, German, Dutch). - Created new directus field in conversation_chunk to hold runpod job id - Improved error handling and logging for transcription processes.
Pull request was converted to draft
There was a problem hiding this comment.
Actionable comments posted: 5
♻️ Duplicate comments (1)
echo/server/dembrane/transcribe.py (1)
98-129: 🛠️ Refactor suggestionLanguage-detect flow still brittle – let’s bulletproof it
Previous feedback wasn’t fully incorporated. Two quick wins:•
detect()explodes on short text – wrap it.
• Normalise casing + deterministic translation (temperature=0) with asystemseed message for coherence.- detected_language = detect(response["text"]) - if detected_language != language and language != "multi": + transcript_text = response["text"] + + try: + detected_language = detect(transcript_text) + except Exception as exc: # LangDetectException on short/ambiguous input + logger.debug(f"LangDetect failed -> default to requested lang: {exc}") + detected_language = language or "unknown" + + if detected_language.lower() != (language or "").lower() and language != "multi": translation_prompt = render_prompt( "translate_transcription", - str(language), + (language or "en").lower(), { - "transcript": response["text"], + "transcript": transcript_text, "detected_language": detected_language, "desired_language": language, }, ) llm_translation_response = completion( model=SMALL_LITELLM_MODEL, - messages=[{"role": "user", "content": translation_prompt}], + messages=[ + {"role": "system", "content": "You are a professional translator."}, + {"role": "user", "content": translation_prompt}, + ], api_key=SMALL_LITELLM_API_KEY, api_base=SMALL_LITELLM_API_BASE, api_version=SMALL_LITELLM_API_VERSION, + temperature=0, ) return llm_translation_response["choices"][0]["message"]["content"] - else: - return response["text"] + return transcript_text
📜 Review details
Configuration used: CodeRabbit UI
Review profile: ASSERTIVE
Plan: Pro
⛔ Files ignored due to path filters (3)
echo/directus/sync/specs/item.graphqlis excluded by!echo/directus/sync/specs/**echo/directus/sync/specs/openapi.jsonis excluded by!echo/directus/sync/specs/**echo/directus/sync/specs/system.graphqlis excluded by!echo/directus/sync/specs/**
📒 Files selected for processing (8)
echo/directus/sync/snapshot/fields/conversation_chunk/runpod_job_id.json(1 hunks)echo/server/dembrane/config.py(3 hunks)echo/server/dembrane/transcribe.py(7 hunks)echo/server/prompt_templates/default_whisper_prompt.de.jinja(1 hunks)echo/server/prompt_templates/default_whisper_prompt.en.jinja(1 hunks)echo/server/prompt_templates/default_whisper_prompt.es.jinja(1 hunks)echo/server/prompt_templates/default_whisper_prompt.fr.jinja(1 hunks)echo/server/prompt_templates/default_whisper_prompt.nl.jinja(1 hunks)
🧰 Additional context used
🧬 Code Graph Analysis (1)
echo/server/dembrane/transcribe.py (2)
echo/server/dembrane/s3.py (1)
get_signed_url(143-148)echo/server/dembrane/prompts.py (1)
render_prompt(55-88)
⏰ Context from checks skipped due to timeout of 90000ms (1)
- GitHub Check: ci-check-server
🔇 Additional comments (8)
echo/server/prompt_templates/default_whisper_prompt.es.jinja (1)
1-1: Ship it! Solid Spanish prompt template.The Spanish language prompt template follows the same format as other language templates and provides clear instructions to transcribe in Spanish.
echo/server/prompt_templates/default_whisper_prompt.en.jinja (1)
1-1: LGTM! Clean English prompt template.Concise, well-structured English transcription instruction following the same pattern as other language templates.
echo/server/prompt_templates/default_whisper_prompt.fr.jinja (1)
1-1: Parfait! French template ready to ship.The French language prompt template aligns perfectly with the project's multilingual approach and maintains consistency with other language templates.
echo/server/dembrane/config.py (2)
279-281: Clean code simplification, LGTM!Nice refactoring of the
LIGHTRAG_LITELLM_AUDIOMODEL_MODELenvironment variable retrieval to a single line, making the code cleaner and more maintainable.
233-259: 🧹 Nitpick (assertive)Missing newline after flag declaration.
Need a blank line after
ENABLE_LITELLM_WHISPER_TRANSCRIPTIONflag for consistency with the pattern used elsewhere in the file.ENABLE_LITELLM_WHISPER_TRANSCRIPTION = os.environ.get( "ENABLE_LITELLM_WHISPER_TRANSCRIPTION", "false" ).lower() in ["true", "1"] logger.debug(f"ENABLE_LITELLM_WHISPER_TRANSCRIPTION: {ENABLE_LITELLM_WHISPER_TRANSCRIPTION}") + LITELLM_WHISPER_API_KEY = os.environ.get("LITELLM_WHISPER_API_KEY")Likely an incorrect or invalid review comment.
echo/server/prompt_templates/default_whisper_prompt.de.jinja (1)
1-1: LGTM – prompt is crisp and correct German
Nothing to tweak here; the wording is clear and forces Whisper to stay in target language.echo/server/prompt_templates/default_whisper_prompt.nl.jinja (1)
1-1: LGTM – Dutch template reads well
The instruction is unambiguous and mirrors the pattern used in other languages. Ship it.echo/server/dembrane/transcribe.py (1)
224-236: LGTM – graceful hand-off between RunPod and LiteLLM
The configuration gate is clean and the Directus update looks atomic.
echo/directus/sync/snapshot/fields/conversation_chunk/runpod_job_id.json
Show resolved
Hide resolved
- Introduced new environment variables for RunPod Whisper configuration, including priority URL and max request threshold. - Added a new scheduled job to update RunPod transcription responses every 1 minute. - Enhanced `queue_transcribe_audio_runpod` to support priority requests based on source and request count. - Implemented `task_process_runpod_chunk_response` to handle chunk status updates and logging. - Improved error handling and logging throughout the transcription process.
There was a problem hiding this comment.
Actionable comments posted: 7
📜 Review details
Configuration used: CodeRabbit UI
Review profile: ASSERTIVE
Plan: Pro
⛔ Files ignored due to path filters (5)
echo/directus/sync/specs/item.graphqlis excluded by!echo/directus/sync/specs/**echo/directus/sync/specs/openapi.jsonis excluded by!echo/directus/sync/specs/**echo/directus/sync/specs/system.graphqlis excluded by!echo/directus/sync/specs/**echo/server/requirements-dev.lockis excluded by!**/*.lockecho/server/requirements.lockis excluded by!**/*.lock
📒 Files selected for processing (6)
echo/directus/sync/snapshot/fields/conversation_chunk/runpod_job_status_link.json(1 hunks)echo/directus/sync/snapshot/fields/conversation_chunk/runpod_request_count.json(1 hunks)echo/server/dembrane/config.py(3 hunks)echo/server/dembrane/scheduler.py(1 hunks)echo/server/dembrane/tasks.py(3 hunks)echo/server/dembrane/transcribe.py(7 hunks)
🔇 Additional comments (11)
echo/directus/sync/snapshot/fields/conversation_chunk/runpod_job_status_link.json (1)
1-43: Field looks clean for storing RunPod job status URLs.The schema for the
runpod_job_status_linktext field is properly configured to store URLs that track the status of RunPod transcription jobs. Making it nullable is a smart design choice since this field will only be populated during active transcription jobs and cleared when complete.echo/directus/sync/snapshot/fields/conversation_chunk/runpod_request_count.json (1)
1-43: Counter field for tracking RunPod API requests looks solid.The integer field with default value of 0 will effectively track transcription request counts. This enables your system to implement retry logic and thresholds to prevent infinite retries for failed transcription attempts. The schema is well-structured and integrates nicely with the RunPod workflow.
echo/server/dembrane/transcribe.py (7)
7-10: Imports look clean, imported functions now reused from shared modulesYou're pulling in
requestsfor direct API calls and importing bothget_signed_urlandget_stream_from_s3from the s3 module. This avoids duplicating code logic - the sign of a senior dev! Kudos for reusing battle-tested utils.
93-93: Cleaner vertical spacingGood improvement to vertical spacing in
transcribe_audio_litellmfunction. Makes the error handling section more visually distinct.
178-178: Dynamic prompt rendering using templating systemNice optimization using
render_promptinstead of hardcoded prompts. This supports the multi-language capability mentioned in the PR objectives.
200-237: RunPod transcription path is well-implementedThe RunPod integration flow is solid:
- Retrieves current request count before queuing
- Implements priority routing based on source
- Updates Directus with job status link for tracking
- Implements threshold check to prevent infinite retries
- Clears job status if threshold is reached
This is 10x engineering - handling all edge cases!
238-254: LiteLLM path uses the correct variable nameYou've renamed
transcriptiontotranscriptto avoid shadowing the imported function, which was flagged in a previous review. The code flow is clean and straightforward. Well done!
255-256: Good fallback error handlingNice to see explicit error handling when no transcription method is enabled. This will prevent silent failures.
134-153:Details
❓ Verification inconclusive
Previous chunk context retrieval is commented out
You've commented out the code that fetched the previous chunk's transcript. Is this intentional? The PR mentions enhancing transcription logic, but removing context from previous chunks might affect transcription quality for conversations that need continuity.
🏁 Script executed:
#!/bin/bash # Search codebase for any new implementation that might replace this functionality echo "Searching for alternative implementations of previous chunk context retrieval..." rg -p "previous.*chunk" --glob "*.py" | grep -v "^#"Length of output: 683
Missing previous-chunk context logic remains commented out
It looks like the code fetchingprevious_chunk_transcript(echo/server/dembrane/transcribe.py lines 134–153) and the prompt append (lines 193–194) is still completely disabled, and there’s no alternative implementation elsewhere. Without cross-chunk context, transcription continuity may suffer. Was this intentional? If not, please restore or replace this logic.• echo/server/dembrane/transcribe.py: lines 134–153
• echo/server/dembrane/transcribe.py: lines 193–194LGTM.
echo/server/dembrane/config.py (2)
243-269: LiteLLM assertions now properly conditionalGood job making the LiteLLM whisper assertions conditional. This follows the pattern you established for RunPod and ensures the server can start even if these variables aren't set, as long as the feature is disabled.
290-290: Cleaner variable assignmentThe variable assignment for
LIGHTRAG_LITELLM_AUDIOMODEL_MODELis now more concise. Clean refactoring.
- Implemented a check for the job status of RunPod transcription, logging progress and returning None if the job is still in progress.
There was a problem hiding this comment.
Actionable comments posted: 2
📜 Review details
Configuration used: CodeRabbit UI
Review profile: ASSERTIVE
Plan: Pro
📒 Files selected for processing (1)
echo/server/dembrane/transcribe.py(7 hunks)
⏰ Context from checks skipped due to timeout of 90000ms (1)
- GitHub Check: ci-check-server
🔇 Additional comments (4)
echo/server/dembrane/transcribe.py (4)
7-23: Imports look solid. LGTM.The RunPod configuration imports and requests library are properly structured. The dual transcription backend support is cleanly separated through feature flags.
36-76: RunPod integration is rock solid. Ship it! 🚀The error handling cascade, conditional payload construction, and priority queue routing are all dialed in. The 600s timeout is hefty but transcription jobs can be chunky. Past review concerns about raise_for_status() and URL construction are properly addressed.
94-103: Config constants properly wired up. LGTM.Clean migration to use the imported configuration constants while preserving the robust error handling.
178-191: Clean prompt templating implementation. LGTM.The shift to render_prompt for language-specific defaults is solid. The "user:" prefix addition helps with prompt engineering. Note that previous chunk context is disabled - ensure this aligns with your transcription quality requirements.
| collection_name="conversation_chunk", | ||
| item_id=conversation_chunk_id, | ||
| item_data={ | ||
| "runpod_job_status_link": str(RUNPOD_WHISPER_BASE_URL) + "/status/" + job_id, |
There was a problem hiding this comment.
🧹 Nitpick (assertive)
Consider using urljoin for cleaner URL construction.
While the current string concatenation works, using urllib.parse.urljoin would be more robust.
+from urllib.parse import urljoin
...
- "runpod_job_status_link": str(RUNPOD_WHISPER_BASE_URL) + "/status/" + job_id,
+ "runpod_job_status_link": urljoin(str(RUNPOD_WHISPER_BASE_URL), f"status/{job_id}"),📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| "runpod_job_status_link": str(RUNPOD_WHISPER_BASE_URL) + "/status/" + job_id, | |
| # At the top of echo/server/dembrane/transcribe.py, add: | |
| from urllib.parse import urljoin | |
| # … other imports … | |
| # Then, in the RunPod logic where you build the status link: | |
| # (replace the existing line) | |
| "runpod_job_status_link": urljoin( | |
| str(RUNPOD_WHISPER_BASE_URL), | |
| f"status/{job_id}" | |
| ), |
🤖 Prompt for AI Agents
In echo/server/dembrane/transcribe.py at line 238, replace the string
concatenation used to build the "runpod_job_status_link" URL with
urllib.parse.urljoin for more robust URL construction. Import urljoin from
urllib.parse if not already imported, then use it to join
RUNPOD_WHISPER_BASE_URL and "/status/" + job_id to form the complete URL.
| if ENABLE_RUNPOD_WHISPER_TRANSCRIPTION: | ||
| directus_response = directus.get_items( | ||
| "conversation_chunk", | ||
| { | ||
| "query": {"filter": {"id": {"_eq": conversation_chunk_id}}, | ||
| "fields": ["source","runpod_job_status_link","runpod_request_count"]}, | ||
| }, | ||
| ) | ||
| runpod_request_count = (directus_response[0]["runpod_request_count"]) | ||
| source = (directus_response[0]["source"]) | ||
| runpod_job_status_link = (directus_response[0]["runpod_job_status_link"]) | ||
|
|
||
| headers = { | ||
| "Content-Type": "application/json", | ||
| "Authorization": f"Bearer {RUNPOD_WHISPER_API_KEY}", | ||
| } | ||
|
|
||
| if runpod_job_status_link: | ||
| response = requests.get(runpod_job_status_link, headers=headers) | ||
| job_status = response.json()['status'] | ||
| logger.debug(f"job_status: {job_status}") | ||
| if job_status == "IN_PROGRESS": | ||
| logger.info(f"RunPod job {runpod_job_status_link} is in progress") | ||
| return None | ||
|
|
There was a problem hiding this comment.
Add defensive checks for API calls and null fields.
The RunPod status check needs hardening:
sourcecould be None/missing and cause issues on line 226- The requests.get call on line 218 needs error handling
Apply this fix:
source = (directus_response[0]["source"])
runpod_job_status_link = (directus_response[0]["runpod_job_status_link"])
headers = {
"Content-Type": "application/json",
"Authorization": f"Bearer {RUNPOD_WHISPER_API_KEY}",
}
if runpod_job_status_link:
- response = requests.get(runpod_job_status_link, headers=headers)
- job_status = response.json()['status']
+ try:
+ response = requests.get(runpod_job_status_link, headers=headers, timeout=30)
+ response.raise_for_status()
+ job_status = response.json().get('status')
+ except Exception as e:
+ logger.error(f"Failed to check RunPod job status: {e}")
+ job_status = None
logger.debug(f"job_status: {job_status}")
if job_status == "IN_PROGRESS":
logger.info(f"RunPod job {runpod_job_status_link} is in progress")
return None📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| if ENABLE_RUNPOD_WHISPER_TRANSCRIPTION: | |
| directus_response = directus.get_items( | |
| "conversation_chunk", | |
| { | |
| "query": {"filter": {"id": {"_eq": conversation_chunk_id}}, | |
| "fields": ["source","runpod_job_status_link","runpod_request_count"]}, | |
| }, | |
| ) | |
| runpod_request_count = (directus_response[0]["runpod_request_count"]) | |
| source = (directus_response[0]["source"]) | |
| runpod_job_status_link = (directus_response[0]["runpod_job_status_link"]) | |
| headers = { | |
| "Content-Type": "application/json", | |
| "Authorization": f"Bearer {RUNPOD_WHISPER_API_KEY}", | |
| } | |
| if runpod_job_status_link: | |
| response = requests.get(runpod_job_status_link, headers=headers) | |
| job_status = response.json()['status'] | |
| logger.debug(f"job_status: {job_status}") | |
| if job_status == "IN_PROGRESS": | |
| logger.info(f"RunPod job {runpod_job_status_link} is in progress") | |
| return None | |
| if ENABLE_RUNPOD_WHISPER_TRANSCRIPTION: | |
| directus_response = directus.get_items( | |
| "conversation_chunk", | |
| { | |
| "query": {"filter": {"id": {"_eq": conversation_chunk_id}}, | |
| "fields": ["source","runpod_job_status_link","runpod_request_count"]}, | |
| }, | |
| ) | |
| runpod_request_count = directus_response[0]["runpod_request_count"] | |
| source = directus_response[0]["source"] | |
| runpod_job_status_link = directus_response[0]["runpod_job_status_link"] | |
| headers = { | |
| "Content-Type": "application/json", | |
| "Authorization": f"Bearer {RUNPOD_WHISPER_API_KEY}", | |
| } | |
| if runpod_job_status_link: | |
| try: | |
| response = requests.get(runpod_job_status_link, headers=headers, timeout=30) | |
| response.raise_for_status() | |
| job_status = response.json().get("status") | |
| except Exception as e: | |
| logger.error(f"Failed to check RunPod job status: {e}") | |
| job_status = None | |
| logger.debug(f"job_status: {job_status}") | |
| if job_status == "IN_PROGRESS": | |
| logger.info(f"RunPod job {runpod_job_status_link} is in progress") | |
| return None |
🤖 Prompt for AI Agents
In echo/server/dembrane/transcribe.py around lines 200 to 224, add defensive
checks to handle cases where 'source' might be None or missing after fetching
from Directus, and wrap the requests.get call in a try-except block to catch and
handle potential network or HTTP errors gracefully. Ensure that before using
'source' or other fields, you verify they are not None, and log or handle errors
from the API call to prevent crashes.
spashii
changed the title
spashii
changed the title
* Add langdetect dependency and enhance transcription logic - Updated transcription logic in `transcribe.py` to detect language and translate if necessary, improving multi-language support. - Adjusted prompt formatting for better clarity in transcription requests. - Added `langdetect` version 1.0.9 to dependencies in `pyproject.toml`, `requirements-dev.lock`, and `requirements.lock`. * Refactor variable naming in transcription logic - Changed variable name from `response` to `llm_translation_response` for clarity in the transcription process. * Add RunPod Whisper transcription support and enhance language handling - Introduced environment variables for enabling RunPod Whisper transcription and managing API keys. - Implemented `queue_transcribe_audio_runpod` function to handle audio transcription requests to RunPod. - Updated `transcribe_conversation_chunk` to support both RunPod and LiteLLM transcription based on configuration. - Added default whisper prompt templates for multiple languages (English, Spanish, French, German, Dutch). - Created new directus field in conversation_chunk to hold runpod job id - Improved error handling and logging for transcription processes. * cosmetic formatting correction * Add RunPod Whisper transcription enhancements and new scheduling task - Introduced new environment variables for RunPod Whisper configuration, including priority URL and max request threshold. - Added a new scheduled job to update RunPod transcription responses every 1 minute. - Enhanced `queue_transcribe_audio_runpod` to support priority requests based on source and request count. - Implemented `task_process_runpod_chunk_response` to handle chunk status updates and logging. - Improved error handling and logging throughout the transcription process. * directus updates for runpod * Remove langdetect dependency and related code from transcription module * Enhance transcription process by adding RunPod job status check - Implemented a check for the job status of RunPod transcription, logging progress and returning None if the job is still in progress.
3 participants
transcribe.pyto detect language and translate if necessary, improving multi-language support.langdetectversion 1.0.9 to dependencies inpyproject.toml,requirements-dev.lock, andrequirements.lock.Summary by CodeRabbit
New Features
Configuration
Bug Fixes