umyunsang commited on
Commit
741d10d
·
verified ·
1 Parent(s): 0104403

fix: KV_CACHE_DTYPE=fp8 제거 (A100 80GB 여유 충분) + ninja-build 추가

Browse files
Files changed (1) hide show
  1. Dockerfile +1 -2
Dockerfile CHANGED
@@ -11,7 +11,6 @@ ENV PYTHONUNBUFFERED=1 \
11
  GPU_UTILIZATION=0.90 \
12
  MAX_MODEL_LEN=8192 \
13
  ENFORCE_EAGER=false \
14
- KV_CACHE_DTYPE=fp8 \
15
  ADAPTER_PATHS="civil=umyunsang/govon-civil-adapter,legal=siwo/govon-legal-adapter"
16
 
17
  WORKDIR /app
@@ -19,7 +18,7 @@ WORKDIR /app
19
  COPY --from=ghcr.io/astral-sh/uv:0.7.12 /uv /bin/uv
20
 
21
  RUN apt-get update && apt-get install -y --no-install-recommends \
22
- python3.10 python3.10-dev git gcc \
23
  && rm -rf /var/lib/apt/lists/*
24
 
25
  COPY requirements.txt .
 
11
  GPU_UTILIZATION=0.90 \
12
  MAX_MODEL_LEN=8192 \
13
  ENFORCE_EAGER=false \
 
14
  ADAPTER_PATHS="civil=umyunsang/govon-civil-adapter,legal=siwo/govon-legal-adapter"
15
 
16
  WORKDIR /app
 
18
  COPY --from=ghcr.io/astral-sh/uv:0.7.12 /uv /bin/uv
19
 
20
  RUN apt-get update && apt-get install -y --no-install-recommends \
21
+ python3.10 python3.10-dev git gcc ninja-build \
22
  && rm -rf /var/lib/apt/lists/*
23
 
24
  COPY requirements.txt .