Abi Raja
|
701d97ec74
|
add comments
|
2024-07-31 11:27:26 -04:00 |
|
Abi Raja
|
7b2e2963ad
|
print for debugging
|
2024-07-31 11:25:49 -04:00 |
|
Abi Raja
|
637c1b4c1d
|
fix TODO
|
2024-07-31 11:22:25 -04:00 |
|
Abi Raja
|
c61a2ac772
|
fix TODO
|
2024-07-31 11:18:35 -04:00 |
|
Abi Raja
|
bcb89a3c23
|
refactor
|
2024-07-31 11:15:56 -04:00 |
|
Abi Raja
|
3591588e2b
|
abstract out prompt assembly into a separate function
|
2024-07-31 11:07:30 -04:00 |
|
Abi Raja
|
dd7a51dd34
|
refactor
|
2024-07-31 10:16:03 -04:00 |
|
Abi Raja
|
3fbc0f9458
|
refactor
|
2024-07-31 10:15:07 -04:00 |
|
Abi Raja
|
ff12790883
|
refactor
|
2024-07-31 10:14:23 -04:00 |
|
Abi Raja
|
0f731598dd
|
refactor to get .get()
|
2024-07-31 10:10:13 -04:00 |
|
Abi Raja
|
24a123db36
|
refactors
|
2024-07-30 16:37:36 -04:00 |
|
Abi Raja
|
64926408b0
|
refactor
|
2024-07-30 16:29:06 -04:00 |
|
Abi Raja
|
0700de7767
|
standardize to using typed send_message
|
2024-07-30 16:27:04 -04:00 |
|
Abi Raja
|
aff9352dc0
|
set up multiple generations
|
2024-07-30 15:44:48 -04:00 |
|
Abi Raja
|
d7ab620e0b
|
deep copy messages to avoid modifying the original list in the Claude LLM call
|
2024-07-26 11:56:21 -04:00 |
|
Abi Raja
|
54b59c85d6
|
add eval runner for text prompt
|
2024-07-19 09:25:11 -04:00 |
|
Abi Raja
|
9d11866143
|
improve evals code
|
2024-07-19 07:55:44 -04:00 |
|
Abi Raja
|
9f732c4f5d
|
update max tokens for Claude Sonnet 3.5 to newly supported limit (8192)
|
2024-07-15 18:51:22 -04:00 |
|
Abi Raja
|
9b99cad286
|
auto-upgrade usage of older models
|
2024-07-11 16:26:26 -04:00 |
|
Abi Raja
|
9ec3adad19
|
fix bug where a image with dimensions larger than what Claude can accept but under the file size limit Claude can accept was not resized
|
2024-07-03 10:06:04 -04:00 |
|
Abi Raja
|
d2359f644c
|
for images sent to Claude, resize if dimension exceeds limit
|
2024-06-27 17:59:15 +08:00 |
|
Abi Raja
|
38a06517fb
|
add unit tests for the prompt assembly
|
2024-06-27 16:08:41 +08:00 |
|
Abi Raja
|
638053a54c
|
add HTML + CSS prompt for imported code
|
2024-06-27 16:02:06 +08:00 |
|
Abi Raja
|
ec1bada3e2
|
fix type errors introduced with previous changes
|
2024-06-27 16:01:28 +08:00 |
|
Abi Raja
|
0200274e61
|
Merge branch 'main' into pr/227
|
2024-06-27 14:44:24 +08:00 |
|
Abi Raja
|
edfd16ef1d
|
extract only html content
|
2024-06-26 16:36:33 +08:00 |
|
Abi Raja
|
4288cf2088
|
fix imported code prompt for Claude which doesn't allow multiple 'user' role messages in a row
|
2024-06-26 13:21:12 +08:00 |
|
Abi Raja
|
6be83b4a2d
|
process images to be under 5mb for Claude
|
2024-06-25 20:57:41 +08:00 |
|
Abi Raja
|
9a4fbd4591
|
Merge pull request #358 from moas/patch-1
Fix Python version in dockerfile to avoid TypeError
|
2024-06-23 05:48:44 -04:00 |
|
Abi Raja
|
49a24d8b0c
|
support Claude Sonnet 3.5
|
2024-06-22 19:39:49 +08:00 |
|
Oscar Landry L
|
9e2eb0c58c
|
Update Dockerfile
fix error:
TypeError: ForwardRef._evaluate() missing 1 required keyword-only argument: 'recursive_guard'
|
2024-06-11 14:58:18 +02:00 |
|
Abi Raja
|
561ac0b088
|
Merge pull request #348 from naman1608/type-fixes
full typed in image generation file
|
2024-06-05 15:21:41 -04:00 |
|
Abi Raja
|
7cd9b294c3
|
fix another type ignore
|
2024-06-05 15:18:54 -04:00 |
|
Abi Raja
|
8de0e95c54
|
fix up types
|
2024-06-05 15:12:47 -04:00 |
|
Abi Raja
|
5a62fce52d
|
Merge pull request #341 from naman1608/handle-none-types
handle none type in llm.py
|
2024-05-31 15:04:32 -04:00 |
|
Abi Raja
|
f6f8b2f707
|
fix bug
|
2024-05-31 14:24:01 -04:00 |
|
Abi Raja
|
c64968b8f0
|
read anthropic api key from front-end
|
2024-05-31 14:06:08 -04:00 |
|
Naman
|
f0ee686660
|
null check for chunk.choices
|
2024-05-29 13:15:23 +05:30 |
|
Naman
|
c5695975c7
|
handle none type in llm.py
|
2024-05-29 13:15:23 +05:30 |
|
Naman
|
d31ebcaa27
|
types fix in image generation file
|
2024-05-29 07:11:03 +05:30 |
|
Abi Raja
|
43214bbbf8
|
add another unit test for new model
|
2024-05-13 15:26:12 -04:00 |
|
Abi Raja
|
8e6a9c48f8
|
support GPT-4o
|
2024-05-13 15:24:47 -04:00 |
|
Abi Raja
|
a5fe0960d8
|
support best of n evals
|
2024-04-24 14:54:03 -04:00 |
|
Abi Raja
|
bb642b320e
|
improve evaluation docs and the way the model is passed into the evaluation script
|
2024-04-11 10:52:25 -04:00 |
|
Abi Raja
|
6587b626c5
|
clean up model strings and add support for GPT-4 Turbo (Apr 2024)
|
2024-04-11 09:55:55 -04:00 |
|
Abi Raja
|
9e1bcae545
|
remove access code/token functionality
|
2024-04-10 14:51:15 -04:00 |
|
Abi Raja
|
29cea327cc
|
improve error UX (particular when no OpenAI API key is found)
|
2024-04-10 13:20:42 -04:00 |
|
Abi Raja
|
dc52ff4cce
|
store intermediate artifacts for video for easier debugging
|
2024-03-25 14:09:13 -04:00 |
|
Abi Raja
|
6a0201a5c0
|
fix up mock data
|
2024-03-22 11:14:15 -04:00 |
|
Abi Raja
|
7f44228a95
|
fix bug where open AI key is required to use it with Anthropic
|
2024-03-19 13:57:57 -04:00 |
|