Commit Graph

378 Commits

Author SHA1 Message Date
arcticfly
896c8c5c57 Update README.md 2023-08-02 00:51:57 -07:00
arcticfly
ec5547d0b0 Update README.md with new features and gifs (#110) 2023-08-02 00:46:48 -07:00
Kyle Corbitt
77e4e3b8c3 mobile styles 2023-08-01 23:08:35 -07:00
Kyle Corbitt
a1b03ddad1 Merge pull request #109 from OpenPipe/debug-prompts
Add debug modal for output cells
2023-08-01 22:51:39 -07:00
Kyle Corbitt
6be32bea4c Add debug modal for output cells
See the actual input that a model got for a specific cell. The formatting isn't great right now; should probably iterate on that.
2023-08-01 22:49:38 -07:00
arcticfly
72c70e2a55 Improve conversion to/from Claude (#108)
* Increase min width of prompt variant

* Increase width of custom instructions input

* Start recording API docs

* Provide better instructions for converting to/from Claude

* Fix prettier
2023-08-01 21:03:23 -07:00
arcticfly
026532f2c2 Model selection styling changes (#107)
* Model selection styling changes

* Fix prettier
2023-08-01 18:45:15 -07:00
Kyle Corbitt
f88538336f fix types 2023-08-01 18:31:34 -07:00
Kyle Corbitt
3c7178115e Merge pull request #105 from OpenPipe/bump-models
Bump Replicate models
2023-08-01 18:26:16 -07:00
Kyle Corbitt
292aaf090a Merge pull request #106 from OpenPipe/dark-mode
Update global background color in ChakraThemeProvider
2023-08-01 18:25:57 -07:00
Kyle Corbitt
d9915dc41b Update global background color in ChakraThemeProvider 2023-08-01 18:25:29 -07:00
David Corbitt
3560bcff14 Correct time stamps on waiting message 2023-08-01 18:09:23 -07:00
Kyle Corbitt
6982339a1a Bump Replicate models 2023-08-01 18:08:02 -07:00
arcticfly
d348b130d5 Add navbar to world-champs (#104)
* Add navbar to world-champs

* Move TopNavbar to signup.tsx
2023-08-01 16:59:46 -07:00
Kyle Corbitt
bf67580991 Merge pull request #103 from OpenPipe/world-champs
add created_at and updated_at to users
2023-08-01 16:49:11 -07:00
Kyle Corbitt
156f248c3a add created_at and updated_at to users 2023-08-01 16:43:35 -07:00
Kyle Corbitt
6184498810 Merge pull request #102 from OpenPipe/world-champs
world champs signup
2023-08-01 13:09:06 -07:00
Kyle Corbitt
65a76cddc5 world champs signup
Basic landing page to sign up for the "world champs"
2023-08-01 13:07:14 -07:00
David Corbitt
c88266bcd4 Add twitter card 2023-08-01 12:47:05 -07:00
David Corbitt
1bf9554eca Comment out getServerSideProps in [id].tsx 2023-08-01 01:29:28 -07:00
arcticfly
1fb428ef4a Add scenario editing modal, twitter sentiment seeding (#101)
* testing agi-eval benchmark

* Add scenario modal editor

* Add initial values to ScenarioEditorModal

* Add seedTwitterSentiment.ts

---------

Co-authored-by: Kyle Corbitt <kyle@corbt.com>
2023-08-01 01:26:43 -07:00
Kyle Corbitt
6316eaae6d dummy key at build time 2023-07-31 18:03:51 -07:00
Kyle Corbitt
8513924ea5 give the openai lib a dummy default value to try to fix the build 2023-07-31 17:39:45 -07:00
arcticfly
51d64baae9 Add og image for experiments (#100)
* Add og image for experiments

* Update meta tags

* Fix prettier

* Add key to meta tags

* Remove ngrok from og:image
2023-07-31 16:50:15 -07:00
arcticfly
26b6fa4f0c Requeue rate-limited query model tasks (#99)
* Continue polling stats until all evals complete

* Return evaluation changes early, before it has run

* Add task for running new eval

* requeue rate-limited tasks

* Fix prettier
2023-07-26 16:30:50 -07:00
Kyle Corbitt
807665fdc1 Merge pull request #98 from OpenPipe/anthropic
Add Anthropic provider
2023-07-26 15:44:39 -07:00
Kyle Corbitt
d6597d2c8a Add Anthropic provider
Adds support for Claude 1 and 2 in experiments! 🎉
2023-07-26 15:37:27 -07:00
arcticfly
566d67bf48 Change prompt variant section background color (#97)
* Change backgroundColor of variant editor section

* Move refresh control out of scrollable area
2023-07-26 10:51:23 -07:00
arcticfly
d4fb8b689a Ensure evals run properly (#96)
* Run evals against llama output

* Continue polling in OutputCell until evals complete

* Remove unnecessary check
2023-07-25 20:01:58 -07:00
arcticfly
98b231c8bd Store multiple ModelResponses (#95)
* Store multiple ModelResponses

* Fix prettier

* Add CellContent container
2023-07-25 18:54:38 -07:00
Kyle Corbitt
45afb1f1f4 Merge pull request #94 from OpenPipe/scenario-shape
updated replicate llama2 models
2023-07-25 17:40:05 -07:00
Kyle Corbitt
2bffb03766 updated replicate llama2 models 2023-07-25 16:40:12 -07:00
arcticfly
223b990005 Change experiment button styling (#93)
* Change button styling

* Fix prettier
2023-07-25 11:41:02 -07:00
Kyle Corbitt
fa61c9c472 prettier 2023-07-25 09:24:28 -07:00
Kyle Corbitt
1309a6ec5d Merge pull request #91 from OpenPipe/better-streaming
Better streaming
2023-07-25 06:46:01 -07:00
arcticfly
17a6fd31a5 Update HeaderButtons.tsx marginTop (#92) 2023-07-24 21:09:08 -07:00
Kyle Corbitt
e1cbeccb90 Better streaming
- Always stream the visible scenarios, if the modelProvider supports it
 - Never stream the invisible scenarios

Also actually runs our query tasks in a background worker, which we weren't quite doing before.
2023-07-24 18:34:30 -07:00
arcticfly
d6b97b29f7 Allow experiment forking (#89)
* Move DeleteButton into a separate file

* Rename plural relations

* Add ability to fork

* Fork automatically after auth upon return

* Add experiment card skeleton

* Create HeaderButtons component

* return no header buttons while experiment loading

* Fix prettier

* Remove unused variable

* Remove newline

* Default json values to undefined

* Change header styles

* Fix prettier

* Give AddScenario icon less width

* Move useEffect

* Skip invalidating experiments list after forking

* Require user to be able to view experiment to fork it

* Move experiment creation into same transaction

* Only return the forked experiment id

* Put delete button in experiment settings drawer

* Move useEffect hook
2023-07-24 18:10:59 -07:00
Kyle Corbitt
09140f8b5f Merge pull request #90 from OpenPipe/simplify-model
Only pass in model and provider
2023-07-24 17:43:24 -07:00
Kyle Corbitt
9952dd93d8 Only pass in model and provider
I got somewhat confused by the extra fields, sorry.

Also makes some frontend changes to track that state more directly although in retrospect not sure the frontend changes make things any better.
2023-07-24 17:21:25 -07:00
Kyle Corbitt
e0b457c6c5 Merge pull request #88 from OpenPipe/load-delete
Bugfix: show prompts for view-only experiments
2023-07-24 14:58:58 -07:00
Kyle Corbitt
0c37506975 Bugfix: show prompts for view-only experiments
Accidentally put the Monaco loading code in the wrong place! This fixes.
2023-07-24 14:57:33 -07:00
arcticfly
2b2e0ab8ee Define refinement actions in the model providers (#87)
* Add descriptions of fields in llama 2 input schema

* Let GPT-4 know when the provider stays the same

* Allow refetching in the event of any errors

* Define refinement actions in model providers

* Fix prettier
2023-07-23 17:37:08 -07:00
arcticfly
3dbb06ec00 Give negative margin to account for border (#86) 2023-07-23 16:49:12 -07:00
David Corbitt
85d42a014b Fix padding on AddVariant button 2023-07-22 16:27:59 -07:00
arcticfly
7d1ded3b18 Improve menu styling (#85) 2023-07-22 16:22:00 -07:00
Kyle Corbitt
b00f6dd04b Merge pull request #84 from OpenPipe/paginated-scenarios
Paginate scenarios
2023-07-22 16:12:02 -07:00
Kyle Corbitt
2e395e4d39 Paginate scenarios
Show 10 scenarios at a time and let the user paginate through them to keep the interface responsive with potentially 1000s of scenarios.
2023-07-22 16:10:16 -07:00
Kyle Corbitt
4b06d05908 Merge pull request #82 from OpenPipe/space-out-scenarios
Separate scenarios from prompts in outputs table
2023-07-22 14:44:51 -07:00
Kyle Corbitt
aabf355b81 Merge pull request #81 from OpenPipe/fullscreen-editor
Fullscreen editor
2023-07-22 14:44:42 -07:00