Gameur/Beta
1
0
Fork 0
Code Issues 1 Pull requests Projects Releases Packages Wiki Activity Actions

Optimisation performance (30t/s --> 50t/s) & adaptation parsing json --> llama.cpp & Fix système mémoire

Browse source
This commit is contained in:
Mathis 2026-03-22 14:43:53 +01:00
parent e27b60e60c
commit 0ba47f8363
19 changed files with 457 additions and 123 deletions
Download patch file Download diff file Expand all files Collapse all files

7
.env
View file

@ -11,5 +11,8 @@ PRIMARY_ASSETS_IDS=971446412690722826
INTROSPECTION_CHANNEL_ID=1461107327381016629
# Modèles Ollama (Optimisation Radeon 8060S 64GB)
OLLAMA_MODEL_HEAVY=gpt-oss:120b
OLLAMA_MODEL_FAST=deepseek-r1:7b
OLLAMA_MODEL_HEAVY=gpt-oss-120b-mxfp4
OLLAMA_MODEL_FAST=deepseek-r1:7b
# Configuration llama.cpp direct (SSH Tunnel 8080:8080)
LLAMA_CPP_API_URL=http://localhost:8080