vowelparrot
Follow
Popular repositories
-
langchain Public
Forked from hwchase17/langchain
⚡ Building applications with LLMs through composability⚡ Python 2
-
pygpt4all Public
Forked from nomic-ai/pygpt4all
Official supported Python bindings for llama.cpp + gpt4all
C++
587 contributions in the last year
Contribution activity
May 2023
Created 111 commits in 2 repositories
Created 1 repository
Created a pull request in hwchase17/langchainjs that received 31 comments
Add V2 Tracer
Add V2 Tracer. Add an 'integration test' (no assertions though)
+564
−227
•
31
comments
Opened 77 other pull requests in 4 repositories
hwchase17/langchain
5
open
3
closed
49
merged
- Add Delete Session Method
- Log warning
- Add Feedback Methods + Evaluation examples
- Add Evaluation / Feedback Examples
- Pass Dataset Name by Name not Position
- Separate Runner Functions from Client
- Add logs command
- [WIP] Update Tracing Docs
- Bold Crumbs
- Add dev option
- Async Run Persistance
- Retry session and tenant
- Change server start name
- Add Support for Flexible Input Format for LLM and Chat Model Runs
- Add 'get_token_ids' method
- Persist Volume After Stop
- Update Tracing Walkthrough
- Fix Async Shared Resource Bug
- Revert "Make serpapi base url configurable via env (#4402)"
- Add Patch to stream runs
- Add Server Command
- Add Environment Info to Run
- Convert Chain to a Chain Factory
- Add Steamship Image Generation Tool
- Dereference Messages
- Some pull requests not shown.
hwchase17/langchainjs
16
merged
1
closed
- Update Port Numbers
- Make API URL optional when specifed via env var
- Add couple of run and session crud
- Always fetch tenant
- Add retries to tracer and client
- Add Invocation Params to onLLMStart
- Fix v1 tracer chat message handling
- Add Environment Info to Run Logs
- Update to Use a Factory Method
- Update prompts to prompt
- Update Chat Model Serialization
- Add Chat Message Handling
- Fix tracing variable check
- Add V2 Env Var for Callbacks
- Run the Client over a Dataset
- Add LCP Client with CRD methods for datasets and examples
- [WIP] Structured tool input agent
vowelparrot/langchain
2
merged
nomic-ai/pygpt4all
1
open
Reviewed 100 pull requests in 2 repositories
hwchase17/langchain
25 pull requests
- standardize json parsing
- Update rellm_experimental.ipynb
- Add Evaluation / Feedback Examples
- Docs: clean up getting started
- Add instructions to pyproject.toml
- Update serpapi.py
- Improved query, print & exception handling in REPL Tool
- update langchainplus client and docker file to reflect port changes
- Make test gha workflow manually runnable
- Remove autoreload in examples
- Chatconv agent: output parser exception
- Bold Crumbs
- Hugging Face Loader: Add lazy load
- Async Run Persistance
- Some notebook and client fixes (add retries, clean up docs, etc)
- Add all to extended tests extra
-
Finish implementing
get_token_ids - Get tokens method for BaseLanguageModel, OpenAI, & ChatOpenAI
-
Replaced
SQLDatabaseChaindeprecated direct initialisation withfrom_llmmethod - fixed on_llm issue
- Make serpapi base url configurable via env
- console callback verbose
- Added support for streaming output response to HuggingFaceTextgenInference LLM class
- Add summarization task type for HuggingFace APIs
- Update click.py to pass errors back to Agent
- Some pull request reviews not shown.
hwchase17/langchainjs
18 pull requests
- Fix redis chat message history message hydration
- Refactor logic, reduce repetition
- Make API URL optional when specifed via env var
- Add verbose env var
- Add couple of run and session crud
- Add Invocation Params to onLLMStart
- Nc/fix tracer chat output
- Update to Use a Factory Method
- Update Chat Model Serialization
- Nc/tracer session ensure
- Adds chat model support in the LC+ dataset runner
- Make dataset runner parallel
- Separate base tracer from langchain tracers, Remove concept of session from base tracer, add missing exports, update console tracer accordingly
- Remove extra entrypoint
- Add Chat Message Handling
- Add V2 Tracer
- Add V2 Env Var for Callbacks
- [WIP] Structured tool input agent