User1342/distinct at main

distinct / distinct_agent

751 kB

Ctrl+K

2 contributors

History: 16 commits

User1342

Keep the sentence a refused tool wrote, so a finished run says why rather than tool_error

2776025 about 1 month ago

isolation
initial commit about 1 month ago
README.md

5.78 kB
Drawn mockups instead of recorded demos; delete conversations; drop Llama; fix the weights cache re-downloading about 1 month ago
__init__.py

5.13 kB
initial commit about 1 month ago
__main__.py

525 Bytes
Library page with tested demos; no-agent veil; energy meters and J/token; 15 new skills; external MCP catalogue; py3.10 gate about 1 month ago
access.py

3.6 kB
Say how a worker is shared: one access code, any number of people about 1 month ago
auth.py

629 Bytes
initial commit about 1 month ago
cli.py

84.3 kB
Hold the sampler to the action schema, so a small model calls the tool instead of describing it about 1 month ago
containment.py

23.5 kB
initial commit about 1 month ago
dashboard.py

12.6 kB
initial commit about 1 month ago
deno.lock

3.2 kB
initial commit about 1 month ago
desktop.py

5.24 kB
initial commit about 1 month ago
energy.py

100 kB
Offload what the card can hold, survive a dropped connection, and stop sending prompts the model cannot read about 1 month ago
fanout.py

10.3 kB
initial commit about 1 month ago
firstrun.py

8.45 kB
initial commit about 1 month ago
gguf.py

6.97 kB
Offload what the card can hold, survive a dropped connection, and stop sending prompts the model cannot read about 1 month ago
guard.py

33.9 kB
initial commit about 1 month ago
harness.py

61.9 kB
Keep the sentence a refused tool wrote, so a finished run says why rather than tool_error about 1 month ago
httpapi.py

6.82 kB
initial commit about 1 month ago
models.py

13.4 kB
Drawn mockups instead of recorded demos; delete conversations; drop Llama; fix the weights cache re-downloading about 1 month ago
offload.py

6.32 kB
Measure the GPU offload instead of estimating it, and keep the development scratch out of the repository about 1 month ago
queue.py

21.4 kB
initial commit about 1 month ago
rlm.py

21.4 kB
initial commit about 1 month ago
runners.py

14.7 kB
Library as its own page; cards show what each member makes; Anthropic-calm palette; artifact previews in answers about 1 month ago
runtime.py

32 kB
Measure the GPU offload instead of estimating it, and keep the development scratch out of the repository about 1 month ago
sandbox.py

4.09 kB
initial commit about 1 month ago
server_runner.py

46.3 kB
Give the grammar one branch per tool, and say out loud when llama.cpp refuses it about 1 month ago
servers.py

30.6 kB
initial commit about 1 month ago
state.py

5.06 kB
initial commit about 1 month ago
tools.py

12.1 kB
Offload what the card can hold, survive a dropped connection, and stop sending prompts the model cannot read about 1 month ago
transport.py

33.7 kB
Workers print their own code and are claimed from the site; fetch and verify llama.cpp about 1 month ago
tui.py

22 kB
initial commit about 1 month ago
weights.py

27.5 kB
Drawn mockups instead of recorded demos; delete conversations; drop Llama; fix the weights cache re-downloading about 1 month ago
worker.py

30.5 kB
initial commit about 1 month ago