Spaces:

FINAL-Bench
/

POCKET-35B-CPU

Running on CPU Upgrade

App Files Files Community

28.1 kB

Ctrl+K

Ctrl+K

1 contributor

History: 23 commits

SeaWolf-AI's picture

A/B: raise max_tokens 256->512 so POCKET's reasoning + final answer complete (not cut)

336526e verified about 6 hours ago

.gitattributes

1.52 kB
initial commit about 9 hours ago
Dockerfile

1.48 kB
add Tab2 live A/B vs Bonsai (sequential: Bonsai first, then POCKET); 2nd llama-server for Bonsai Q1_0; README=BONSAI vs POCKET about 7 hours ago
README.md

1.31 kB
A/B is now the landing tab (swap order); add prism-ml/Bonsai-27B-gguf to models relation about 6 hours ago
app.py

5.7 kB
A/B: show reasoning again (remove no_think prefill); raise A/B max_tokens 96->256 so thinking completes about 6 hours ago
index.html

16.9 kB
A/B: raise max_tokens 256->512 so POCKET's reasoning + final answer complete (not cut) about 6 hours ago
requirements.txt

81 Bytes
rebuild backend on upstream llama.cpp master (qwen35moe support) via llama-server; FastAPI proxies; portable CPU build about 8 hours ago
start.sh

1.07 kB
add Tab2 live A/B vs Bonsai (sequential: Bonsai first, then POCKET); 2nd llama-server for Bonsai Q1_0; README=BONSAI vs POCKET about 7 hours ago