684f3df38e
fix defect/regression with linear activities
...
modified: guarded_ai.py
2024-07-28 11:59:40 -04:00
7e7b226fb1
print content blocks and skip the llm logic if no question in step.
...
modified: research/guarded_ai.py
2024-07-28 11:28:44 -04:00
02c97beebe
fix defect when step has no question.
...
modified: app.py
modified: research/activity13-choose-adventure.yaml
2024-07-28 11:04:30 -04:00
cfa5d77a42
Allowing for non-linear progression depending on buckets category
...
new field in the transitions is next_section_and_step which is optional
and defaults to None but also can be "section_3:step_1"
modified: app.py
new file: research/activity0.yaml
new file: research/activity13-choose-adventure.yaml
modified: research/guarded_ai.py
2024-07-28 10:36:09 -04:00
7e97acf477
content_blocks and questions support markdown values.
...
modified: research/activity10.yaml
2024-07-27 15:49:19 -04:00
93f274ef39
modified: app.py
2024-07-27 15:37:29 -04:00
692e34d090
repeat the question if not correct, after AI feedback
...
modified: app.py
2024-07-27 14:14:34 -04:00
ee5721a1fb
default rubric for grading and scoring trivia.
...
displays at the end of game or when running `/activity info`.
modified: app.py
2024-07-27 13:56:20 -04:00
5583e34e51
prompt engineering.
...
modified: research/guarded_ai.py
2024-07-27 12:22:51 -04:00
6dff9dc6e9
Integrated gaurded ai via websocket frontend.
...
modified: app.py
new file: migrations/versions/d04950c5a624_add_activitystate_table2.py
new file: migrations/versions/d3631b8bb652_add_activitystate_table.py
2024-07-27 10:30:07 -04:00
15ac4fa015
algo for guarding AI on rails.
...
new file: research/activity.yaml
new file: research/activity10.yaml
new file: research/activity11.yaml
new file: research/activity12.yaml
new file: research/activity2.yaml
new file: research/activity3.yaml
new file: research/activity4.yaml
new file: research/activity5.yaml
new file: research/activity6.yaml
new file: research/activity7.yaml
new file: research/activity8.yaml
new file: research/activity9.yaml
new file: research/guarded_ai.py
2024-07-27 08:57:44 -04:00
Russell Ballestrini
698455fb40
gpt-4o-mini as gpt-mini alias
...
modified: README.rst
modified: app.py
2024-07-18 17:12:05 -04:00
Russell Ballestrini
da5a8cbcfa
gpt-4o for titles and add claude-opus
...
modified: README.rst
modified: app.py
2024-07-08 17:33:12 -04:00
393769bcf7
Merge pull request #1 from russellballestrini/search
...
Keyword search across all chatrooms to find across conversation history
2024-07-04 12:14:58 -04:00
57d8e701f2
remove username and snippet from SERP page
2024-07-04 12:04:46 -04:00
db3c15ea0d
hide snippets pass room list in search page
...
modified: app.py
modified: templates/base.html
modified: templates/search.html
2024-07-04 11:35:22 -04:00
926262cdaa
keep keywords in search form.
...
modified: app.py
modified: templates/base.html
2024-07-04 11:12:21 -04:00
5577821855
save username when searching
...
I CAN program offline without an LLM. ; )
modified: app.py
modified: templates/base.html
modified: templates/search.html
2024-07-04 10:57:30 -04:00
77c2ef1d83
messing around with keyword search working but ugly!
...
modified: app.py
new file: templates/base.html
modified: templates/chat.html
new file: templates/search.html
2024-07-04 10:21:37 -04:00
e5940d3020
Fix Claude consecutive roles issue
...
This commit addresses the issue where the Claude model was throwing an error due to multiple consecutive "user" roles in the chat history. The following changes have been made:
- Implemented a new function `group_consecutive_roles` to group consecutive messages of the same role into one.
- Updated the `chat_claude`, function to use the `group_consecutive_roles` function before sending the chat history to the respective models.
By grouping consecutive messages of the same role, the chat history now alternates between "user" and "assistant" roles, resolving the "roles must alternate between 'user' and 'assistant'" error from the Claude model.
2024-06-22 09:47:04 -04:00
Russell Ballestrini
4575f2531f
upgrade to gpt-4o
...
modified: app.py
2024-05-14 10:18:43 -04:00
d882a14a8c
update to hermes 2 llama 3 8B
...
modified: README.rst
modified: app.py
2024-05-08 17:35:37 -04:00
Russell Ballestrini
a61c339671
allow mistral client to work with many models at the same time.
2024-05-01 10:43:38 -04:00
Russell Ballestrini
76e7cb83af
make line numbers align on firefox
...
modified: templates/chat.html
2024-04-21 11:46:49 -04:00
Russell Ballestrini
046576e128
upgrade gpt-4 to gpt-4-turbo
...
modified: app.py
2024-04-21 10:22:37 -04:00
dbcdad6838
Additional groq models
...
modified: README.rst
modified: app.py
2024-04-21 09:46:35 -04:00
a11df86f6c
cut over from eventlet to gevent. use monkey patching
...
modified: app.py
modified: requirements.txt
2024-04-20 13:40:26 -04:00
dabce4a954
allow user to prevent autoscrolling streamed chunks.
...
modified: templates/chat.html
2024-04-04 10:01:32 -04:00
4037872e7e
anthropic.claude-3-haiku-20240307-v1:0
...
modified: README.rst
modified: app.py
2024-03-15 10:12:58 -04:00
Russell Ballestrini
f365e45821
Claude 3 sonnet working properly now.
...
modified: README.rst
modified: app.py
2024-03-11 10:35:05 -04:00
Russell Ballestrini
e81a68c5d9
claude 3 sonnet
...
modified: app.py
2024-03-04 11:57:24 -05:00
Russell Ballestrini
36336234c5
claude-sonnet
...
modified: README.rst
modified: app.py
2024-03-04 11:11:41 -05:00
ad84d9a1ec
nvim refactors
...
modified: app.py
2024-03-04 08:36:58 -05:00
f4da537d88
mistral-large-latest is the actual model name.
...
modified: app.py
2024-03-02 11:05:11 -05:00
751dd80cc1
mistral-large
...
modified: README.rst
modified: app.py
2024-03-02 10:46:41 -05:00
Russell Ballestrini
0c856a17a4
clean up import
...
modified: app.py
2024-03-01 10:33:26 -05:00
Russell Ballestrini
23e844baf9
Added groq platform support for ultra fast LLM inference
...
modified: README.rst
modified: app.py
modified: requirements.txt
2024-02-28 10:32:31 -05:00
Russell Ballestrini
6cc066ecb4
upgrade to the newest gpt-4
...
modified: app.py
2024-01-26 08:21:42 -05:00
9418fe5673
black
2024-01-25 08:25:31 -05:00
1b605568b2
vllm support hacked in
...
modified: README.rst
modified: app.py
2024-01-25 08:12:01 -05:00
a72511be5d
messing with running vLLM open hermes
...
modified: app.py
2024-01-23 18:06:19 -05:00
0efbe93910
modified: app.py
2024-01-19 08:59:11 -05:00
f256d65391
modified: app.py
2024-01-13 11:06:41 -05:00
2f8a6c502d
hacked in ability to run local llama2 models like mistral
...
modified: app.py
new file: install-llama.sh
modified: requirements.txt
2024-01-13 10:05:02 -05:00
6a42dacf65
localhost openchat is working. lol
...
I replaced gpt-3.5-turbo workloads with openchat a local GPU powered inference server
The openchat inference server supports using the latest and official openai python client.
This means you can replace both standard and streaming workloads with an "offline" LLM.
modified: README.rst
modified: app.py
2024-01-07 14:57:40 -05:00
560d0a1ae5
small refactor
2024-01-06 11:34:53 -05:00
47d0b24fc4
add a link to the docs
...
modified: templates/chat.html
2024-01-06 08:37:46 -05:00
6f70dcfc42
allow cancel on all LLMs
...
modified: app.py
2024-01-06 08:12:54 -05:00
c40f961fc3
Update README.rst
2023-12-29 14:12:36 -05:00
Russell Ballestrini
d666efb162
implement together/solar
...
modified: README.rst
modified: app.py
2023-12-29 13:28:53 -05:00
