michael/AI-Profile-Router
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Fix Athena recovery and installation consistency

Browse Source
This commit is contained in:
Mikei386
2026-09-01 10:19:13 +02:00
parent 7e14da77f9
commit e30f4f024c
10 changed files with 179 additions and 27 deletions
Download Patch File Download Diff File Expand all files Collapse all files
install.sh
+21 -1
View File
@@ -42,6 +42,9 @@ required=(AI_HOSTNAME ADMIN_USER MODEL_DIR FAST_MODEL_FILE
FAST_MODEL_URL FAST_MODEL_SHA256 MEDIUM_MODEL_FILE MEDIUM_MODEL_URL
MEDIUM_MODEL_SHA256 LARGE_MODEL_FILE LARGE_MODEL_URL LARGE_MODEL_SHA256
ULTRA_MODEL_FILE ULTRA_MODEL_URL ULTRA_MODEL_SHA256
UNCENSORED_MODEL_FILE UNCENSORED_MODEL_URL UNCENSORED_MODEL_SHA256
UNCENSORED_PROJECTOR_FILE UNCENSORED_PROJECTOR_URL
UNCENSORED_PROJECTOR_SHA256
VISION_PROJECTOR_FILE VISION_PROJECTOR_URL VISION_PROJECTOR_SHA256)
for name in "${required[@]}"; do
[[ -n "${!name:-}" ]] || die "Pflichtwert $name fehlt."
@@ -322,10 +325,25 @@ UNCENSORED_MODEL_FILE=$UNCENSORED_MODEL_FILE
UNCENSORED_PROJECTOR_FILE=$UNCENSORED_PROJECTOR_FILE
VISION_PROJECTOR_FILE=$VISION_PROJECTOR_FILE
FAST_CONTEXT=${FAST_CONTEXT:-76800}
FAST_BATCH_SIZE=${FAST_BATCH_SIZE:-64}
FAST_UBATCH_SIZE=${FAST_UBATCH_SIZE:-32}
FAST_PARALLEL_SLOTS=${FAST_PARALLEL_SLOTS:-1}
MEDIUM_CONTEXT=${MEDIUM_CONTEXT:-160000}
MEDIUM_BATCH_SIZE=${MEDIUM_BATCH_SIZE:-2048}
MEDIUM_UBATCH_SIZE=${MEDIUM_UBATCH_SIZE:-128}
MEDIUM_PARALLEL_SLOTS=${MEDIUM_PARALLEL_SLOTS:-1}
LARGE_CONTEXT=${LARGE_CONTEXT:-192000}
LARGE_BATCH_SIZE=${LARGE_BATCH_SIZE:-2048}
LARGE_UBATCH_SIZE=${LARGE_UBATCH_SIZE:-128}
LARGE_PARALLEL_SLOTS=${LARGE_PARALLEL_SLOTS:-1}
ULTRA_CONTEXT=${ULTRA_CONTEXT:-262144}
ULTRA_BATCH_SIZE=${ULTRA_BATCH_SIZE:-2048}
ULTRA_UBATCH_SIZE=${ULTRA_UBATCH_SIZE:-128}
ULTRA_PARALLEL_SLOTS=${ULTRA_PARALLEL_SLOTS:-1}
UNCENSORED_CONTEXT=${UNCENSORED_CONTEXT:-80000}
UNCENSORED_BATCH_SIZE=${UNCENSORED_BATCH_SIZE:-2048}
UNCENSORED_UBATCH_SIZE=${UNCENSORED_UBATCH_SIZE:-128}
UNCENSORED_PARALLEL_SLOTS=${UNCENSORED_PARALLEL_SLOTS:-1}
FAST_GPU_DEVICES=${TEXT_GPU_DEVICES:-0}${SECONDARY_GPU_DEVICES:+,$SECONDARY_GPU_DEVICES}
MEDIUM_GPU_DEVICES=${TEXT_GPU_DEVICES:-0}${SECONDARY_GPU_DEVICES:+,$SECONDARY_GPU_DEVICES}
MEDIUM_TENSOR_SPLIT=${MEDIUM_TENSOR_SPLIT:-90,10}
@@ -340,6 +358,8 @@ IMAGE_GPU_DEVICES=${IMAGE_GPU_DEVICES:-${TEXT_GPU_DEVICES:-0}}
FLUX_MODEL_DIR=${FLUX_MODEL_DIR:-/data/models/FLUX.2-klein-4B}
LLAMA_THREADS=${LLAMA_THREADS:-6}
LLAMA_THREADS_BATCH=${LLAMA_THREADS_BATCH:-6}
LLAMA_CACHE_RAM_MIB=${LLAMA_CACHE_RAM_MIB:-24576}
DEFAULT_REASONING_EFFORT=${DEFAULT_REASONING_EFFORT:-off}
EOF
chmod 0600 $SECRETS_DIR/stack.env
}
@@ -468,7 +488,7 @@ build_and_start() {
# Portable MCPs and Hermes live on Unraid and are restored through Appdata.
"$STACK_DIR/platform/mcp/install-tools.sh"
docker compose --env-file "$SECRETS_DIR/stack.env" --profile inference create \
llama-fast llama-medium llama-large llama-ultra llama-experimental
llama-fast llama-medium llama-large llama-ultra llama-uncensored
docker compose --env-file "$SECRETS_DIR/stack.env" --profile image create image-worker
docker compose --env-file "$SECRETS_DIR/stack.env" up -d --build \
wireguard-gateway xtts piper tts-gateway profile-controller router llama-dashboard backup