23070d6ef7
Release 0.0.96
2023-06-19 19:24:32 +01:00
29015fbc2c
Nc/llmchain functions ( #1699 )
...
* Refactor openai functions chains to use output parsers
* Lint
* Rename
* Add docs
2023-06-19 19:11:58 +01:00
46f8147fa6
Add TransformChain and OpenAIFunctionsChains ( #1668 )
...
* Add TransformChain and OpenAIFunctionsChains
* Move files
* Add tests
2023-06-19 16:12:51 +01:00
0ea3dcf716
updated tiktoken to add gpt-3.5-16k ( #1679 )
2023-06-19 15:57:01 +01:00
276fd1ce75
Include placeholder value for all secrets used, not just those in kwargs ( #1696 )
...
* Include placeholder value for all secrets used, not just those in kwargs
* Fix test
* Fix test
* Add error handling for openai streaming when errors are sent as events in the SSE stream (#1698 )
* Add error handling for openai streaming when errors are sent as events in the SSE stream
* Lint
2023-06-19 15:41:15 +01:00
8ddf206998
Fix example for chatanthropic, Fix LLMChain cancellation when using memory, Allow any chain to be cancelled ( #1686 )
...
* Fix example for chatanthropic
* Fix llmchain cancellation when using memory
* Implement cancellation for all chains
- SequentialChain and SimpleSequentialChain pass the cancellation tokens to the inner chains (and if those are llmchains the cancellation token is passed to the llm)
- Add support for using chains with memory inside simpleseqchain
2023-06-17 21:31:08 +01:00
aeb6ccc172
Merge branch 'main' into vwp/eval_chains
2023-06-16 14:43:04 -07:00
b334b0c4c9
handle managed motorhead data key ( #1652 )
...
* handle managed data key
* fix linter
2023-06-16 09:00:31 +01:00
5964f97919
add funny /*pure*/ comments
2023-06-15 12:49:26 -07:00
dd7e4ad36f
Add entrypoint and export on top eval level
2023-06-14 21:55:15 -07:00
bdb12b473e
Merge branch 'main' into vwp/eval_chains
2023-06-14 21:43:53 -07:00
5661b2d07c
bugfix: ajust batchSize for azure embeddings ( #1647 )
2023-06-14 22:13:48 +01:00
04e945048f
Remove extra log ( #1655 )
2023-06-14 22:13:31 +01:00
361a3e3840
Release 0.0.95
2023-06-14 15:25:54 +01:00
778147ddc0
Fix using memory in agent with retrurnIntermediateSteps
2023-06-14 14:55:31 +01:00
687f2c8047
Feat: typesense vector store ( #1244 )
...
* chore(langchain): add typesense
* feat(langchain/vectorstores): add typesense vectorstore
* tests(langchain/vectorstore): add tests transform between docs & typesense records
* docs(examples): add typesense vectorstore examples
* feat(vectorstores/typesense): add fromTexts
* docs(examples): add typesense client to example
* docs(examples): add import and similarity search examples
* chore(package): typesense as peerDependency
* chore(create-entrypoints): add vectorstores/typesense
* chore(package): typesense optional dep
* lint(vectorstores/typesense): fix errors
* fix(vectorstores/typesense): throw error instead of catch it
* refactor(vectorstores/typesense): use asyncCaller in default import
* refactor(vectorstores/typesense): impor types with import type
* feat(vectorstores/typesense): add similaritySearchVectorWithScore & similaritySearchWithScore methods
* Fix entrypoints
* Remove unnecessary TypeSense initialisation
* fix(examples/typesense): add missing properties
* feat(typesense): add documentation
* fix deleted entry points by error
* fix deleted entry points by error
* Some fixes
* Fix bug
* refactor(typesense): remove modifySearchParams, use filters
* fix on docs
* feat(typesense): addVectors method implemented
* fix(typesense): change test after change
* Updates
* Update docs
---------
Co-authored-by: Tat Dat Duong <david@duong.cz >
Co-authored-by: Nuno Campos <nuno@boringbits.io >
2023-06-14 14:44:49 +01:00
6e7017f423
Fix token count in map reduce to reflect the actual tokens used in the final prompt ( #1645 )
2023-06-14 11:26:14 +01:00
e5af8d3250
Memory support for OpenAI Functions Agent ( #1638 )
...
* Memory support for OpenAI Functions Agent
* Add test, fix .run() method
* Update executor.ts
* Update langchain/src/agents/tests/openai.int.test.ts
* Fix type
---------
Co-authored-by: Nuno Campos <nuno@boringbits.io >
2023-06-14 10:29:25 +01:00
9b17b5e2fc
Fix calling chain.run() with memory ( #1642 )
2023-06-14 10:29:11 +01:00
b696686744
Export the agent ( #1636 )
2023-06-14 08:25:40 +01:00
9a3ca0382f
Release 0.0.94
2023-06-13 22:41:37 +01:00
990d11065a
Nc/openai functions ( #1631 )
...
* wip
* Implement openai agent, add support for passing tools to chatopenai calls
* Fix example import
* Lint
* Update tests
* Update name
* Improve streaming logic
* Add missing param
2023-06-13 22:33:58 +01:00
051a05bfac
Merge branch 'main' into vwp/eval_chains
2023-06-13 11:21:47 -07:00
4d3b0a163a
Release 0.0.93
2023-06-13 18:32:27 +01:00
4bcb7e94d7
Replace string case deps again
2023-06-13 18:17:53 +01:00
7c98d80166
Fix string case dependency, speed up publish command
2023-06-13 18:03:52 +01:00
b2c17771d4
Remove build files
2023-06-13 17:25:04 +01:00
9977c96998
Fix weaviate integration tests, add to docker compose ( #1625 )
...
* Fix weaviate integration tests, add to docker compose
* Fix more int tests
2023-06-13 17:07:02 +01:00
8652e9c4ba
Store events on each run ( #1548 )
...
* Store events on each run
* Fix test
2023-06-13 17:06:48 +01:00
6fc517ce30
New Feat: Added Weaviate self-query retriever ( #1541 )
...
* added weaviate self-query
* Update weaviate_translator.ts
Make the AllowedOperator consistent with the other types.
* replaced neq with ne for translator
* Removed in and nin as comparator
* Update docs and entrypoint
* Fix api key
---------
Co-authored-by: jacoblee93 <jacoblee93@gmail.com >
Co-authored-by: Nuno Campos <nuno@boringbits.io >
2023-06-13 16:08:11 +01:00
252b532acf
Add support for tags ( #1593 )
...
* Add support for tags
* Fix tag assignment
* Update sdk
* Uase run schema from sdk
* Remove extra dep on lcp sdk
* Fix yarn
* Fix after merge
* Fix types
* Fix test
* Lint
2023-06-13 14:23:06 +01:00
1def2ba6a3
Add Aleph Alpha LLM ( #1609 )
...
* PR test
* AlephAlpha integration
* Eslint
* Some changes
* Update import map
---------
Co-authored-by: Nuno Campos <nuno@boringbits.io >
2023-06-13 13:19:06 +01:00
85de6e16b9
feat: Add support for the Solidity language ( #1616 )
2023-06-13 10:32:03 +01:00
0088ba98bf
Merge pull request #1615 from skarard/compatibility
...
Improved build environment compatibility
2023-06-13 09:13:46 +01:00
3e57f4bc68
Merge branch 'vwp/upgrade_sdk' into vwp/eval_chains
2023-06-13 00:01:51 -07:00
3ee5b9a5b1
Upgrade SDK
2023-06-12 23:41:55 -07:00
e77f11503f
update version
2023-06-12 23:31:02 -07:00
6de7e43dda
rename file
2023-06-12 13:30:25 -07:00
d4f19d1461
Implement Eval Chains
2023-06-12 13:30:25 -07:00
671ea7bc58
Merge branch 'main' into compatibility
2023-06-12 17:34:36 +01:00
7d88b89264
Improved build environment compatibility
2023-06-12 17:12:11 +01:00
38364af704
Update autogen files
2023-06-12 15:50:55 +01:00
92a7eee90c
Add anthropic serialization
2023-06-12 15:35:23 +01:00
031719aeb8
Update snapshots
2023-06-12 15:35:23 +01:00
ceb16e77de
Fixes for py compat
2023-06-12 15:35:23 +01:00
15556146c5
Fix import
2023-06-12 15:35:23 +01:00
c60a9730c8
Get latest values for each arg from instance if saved with same name, add test
2023-06-12 15:35:23 +01:00
0747ede85f
Re-use lc_attributes to remove args from serialization
2023-06-12 15:35:23 +01:00
f2d0e82dc2
Make serialization opt-in for each subclass, exclude callbacks/verbose from serialization, add an alias for cohere api key
2023-06-12 15:35:23 +01:00
5f7ddfc90d
Convert keys to snake case, add key aliases to allow matching key names with python
2023-06-12 15:35:23 +01:00
Nuno Campos