Commit Graph

5 Commits

Author SHA1 Message Date
Stijnus
b5d9055851 🔧 Fix Token Limits & Invalid JSON Response Errors (#1934)
ISSUES FIXED:
-  Invalid JSON response errors during streaming
-  Incorrect token limits causing API rejections
-  Outdated hardcoded model configurations
-  Poor error messages for API failures

SOLUTIONS IMPLEMENTED:

🎯 ACCURATE TOKEN LIMITS & CONTEXT SIZES
- OpenAI GPT-4o: 128k context (was 8k)
- OpenAI GPT-3.5-turbo: 16k context (was 8k)
- Anthropic Claude 3.5 Sonnet: 200k context (was 8k)
- Anthropic Claude 3 Haiku: 200k context (was 8k)
- Google Gemini 1.5 Pro: 2M context (was 8k)
- Google Gemini 1.5 Flash: 1M context (was 8k)
- Groq Llama models: 128k context (was 8k)
- Together models: Updated with accurate limits

�� DYNAMIC MODEL FETCHING ENHANCED
- Smart context detection from provider APIs
- Automatic fallback to known limits when API unavailable
- Safety caps to prevent token overflow (100k max)
- Intelligent model filtering and deduplication

🛡️ IMPROVED ERROR HANDLING
- Specific error messages for Invalid JSON responses
- Token limit exceeded warnings with solutions
- API key validation with clear guidance
- Rate limiting detection and user guidance
- Network timeout handling

 PERFORMANCE OPTIMIZATIONS
- Reduced static models from 40+ to 12 essential
- Enhanced streaming error detection
- Better API response validation
- Improved context window display (shows M/k units)

🔧 TECHNICAL IMPROVEMENTS
- Dynamic model context detection from APIs
- Enhanced streaming reliability
- Better token limit enforcement
- Comprehensive error categorization
- Smart model validation before API calls

IMPACT:
 Eliminates Invalid JSON response errors
 Prevents token limit API rejections
 Provides accurate model capabilities
 Improves user experience with clear errors
 Enables full utilization of modern LLM context windows
2025-08-29 20:53:57 +02:00
Stijnus
9a5076d8c6 feat: lock files (#1681)
* Add persistent file locking feature with enhanced UI

* Fix file locking to be scoped by chat ID

* Add folder locking functionality

* Update CHANGES.md to include folder locking functionality

* Add early detection of locked files/folders in user prompts

* Improve locked files detection with smarter pattern matching and prevent AI from attempting to modify locked files

* Add detection for unlocked files to allow AI to continue with modifications in the same chat session

* Implement dialog-based Lock Manager with improved styling for dark/light modes

* Add remaining files for file locking implementation

* refactor(lock-manager): simplify lock management UI and remove scoped lock options

Consolidate lock management UI by removing scoped lock options and integrating LockManager directly into the EditorPanel. Simplify the lock management interface by removing the dialog and replacing it with a tab-based view. This improves maintainability and user experience by reducing complexity and streamlining the lock management process.

Change Lock & Unlock action to use toast instead of alert.

Remove LockManagerDialog as it is now tab based.

* Optimize file locking mechanism for better performance

- Add in-memory caching to reduce localStorage reads
- Implement debounced localStorage writes
- Use Map data structures for faster lookups
- Add batch operations for locking/unlocking multiple items
- Reduce polling frequency and add event-based updates
- Add performance monitoring and cross-tab synchronization

* refactor(file-locking): simplify file locking mechanism and remove scoped locks

This commit removes the scoped locking feature and simplifies the file locking mechanism. The `LockMode` type and related logic have been removed, and all locks are now treated as full locks. The `isLocked` property has been standardized across the codebase, replacing the previous `locked` and `lockMode` properties. Additionally, the `useLockedFilesChecker` hook and `LockAlert` component have been removed as they are no longer needed with the simplified locking system.

This gives the LLM a clear understanding of locked files and strict instructions not to make any changes to these files

* refactor: remove debug console.log statements

---------

Co-authored-by: KevIsDev <zennerd404@gmail.com>
2025-05-08 00:07:32 +02:00
Anirban Kar
3c56346e83 feat: enhance context handling by adding code context selection and implementing summary generation (#1091) #release
* feat: add context annotation types and enhance file handling in LLM processing

* feat: enhance context handling by adding chatId to annotations and implementing summary generation

* removed useless changes

* feat: updated token counts to include optimization requests

* prompt fix

* logging added

* useless logs removed
2025-01-22 22:48:13 +05:30
fernsdavid25
f7615093dd Update max_tokens in constants.ts
max_tokens for llama 3.1 models must be less than or equal to 8000 but it is set to 8192. just change it to 8000 and the error is fixed.
2024-10-21 19:32:57 +05:30
Sam Denty
6fb59d2bc5 fix: remove monorepo 2024-09-25 19:54:09 +01:00