Spaces:
Running
Running
Commit History
Workflow to ensure codebase integrity
5e2289d
cost_benefit file deleted
c978556
Merge branch 'main' of github.com:kunpai/HASHIRU
2155e91
Added expense budget
bf722a2
Enable conditional memory usage in GeminiManager
82fe65b
Kunal Pai
commited on
MPS support for memories for Mac users
ab59793
Kunal Pai
commited on
Add os import and ensure data directory exists before saving memory
941e057
Kunal Pai
commited on
Updated memory manager
0ab24d1
updated prompt
93470ad
Updated memory related tasks
167a572
Updates to the prompt
50fb0a0
Merge branch 'main' of github.com:kunpai/HASHIRU
eb1b69a
Added new prompt
9f97a28
models_cost file updated
b915149
Cleaning up debug messages
0144997
Merge branch 'main' of github.com:kunpai/HASHIRU
53ee793
Giving agent access to the model costs
1476939
update agent creator to specify base model
6115f71
Harshil Patel
commited on
Add code to make total budget expand with total resources
0e323dc
Harshil Patel
commited on
Add base models for Employee agents
03de09a
Kunal Pai
commited on
Memory function description added to paper
58408af
Kunal Pai
commited on
Make paper more concise
adfd14d
Kunal Pai
commited on
Update HASHIRU_ARCH.png with new architecture diagram
7040aa5
Kunal Pai
commited on
model_cost file added
705389b
Updating threshold
344c9c4
Merge branch 'main' of github.com:kunpai/HASHIRU
e7a610c
Adding memory management using sentenceTransformer
9d9d55a
cost_benefit modified
703f2c8
Memory actions work
b3d741d
Updated prompt to work with Updating memory
bf8a152
Cleaning up models.json
36dcd43
Fixing tool deletion
fde43e7
Some more refactor
0e58feb
refactor
2f85c93
Implemented thinking and fixed a bug with agent creator
60b4d0f
History and multiple conversations work!
d648fe6
current check for testing cost
1674315
agent creation for testing cost
9725104
updated cost_benefit
800db6b
updated model selector
82a76a7
Testing creation of agent with runtime detector and model selector
69dc898
removed base_model from required fields and added choose_best_model
5d97677
updated model_selector to choose model based on runtime/env to create agents
a5ec5ae
added runtime/env check and api key check
6d068ed
Pushing UI updates
4af653a
QOL updates and refactoring. Also fixed the tool/agent budgeting
6900003
Add benchmarking functionality for Globle game
cab9413
Kunal Pai
commited on
Add benchmarking script for Wordle game
e09bf50
Kunal Pai
commited on
Add benchmarking functionality for NYT Connections dataset
0577af4
Kunal Pai
commited on