gahow/xserv
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

tools: default dash5 paths to home layout

Browse Source
This commit is contained in:
Gahow Wang
2026-07-10 11:22:41 +08:00
parent 6309dc1181
commit 588bfd9df3
5 changed files with 10 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
tools/bench/pp_clean_bench.sh
View File

@@ -9,8 +9,8 @@ set -u
cd "$(dirname "$0")/../.."
export PATH=$HOME/.cargo/bin:/usr/local/cuda-12.9/bin:$PATH
export CUDA_HOME=${CUDA_HOME:-/usr/local/cuda-12.9}
MODEL=${MODEL:-/opt/wjh/models/qwen3-8b}
GGUF=${GGUF:-/opt/wjh/models/qwen3-8b/qwen3-8b-bf16.gguf}
MODEL=${MODEL:-$HOME/models/qwen3-8b}
GGUF=${GGUF:-$HOME/models/qwen3-8b/qwen3-8b-bf16.gguf}
LLAMA_BIN=${LLAMA_BIN:-third_party/llama.cpp/build/bin/llama-server}
XBIN=./target/release/xserv-server
PPS=${PPS:-1 2 4}

4
tools/bench/run_pp_parallel.sh
View File

@@ -10,8 +10,8 @@
# Run from the repo root on the GPU host. Produces bench-out/pp{1,2,4}-{xserv,llama}.
set -u
MODEL="${MODEL:-/opt/wjh/models/qwen3-8b}"
GGUF="${GGUF:-/opt/wjh/models/qwen3-8b/qwen3-8b-bf16.gguf}"
MODEL="${MODEL:-$HOME/models/qwen3-8b}"
GGUF="${GGUF:-$HOME/models/qwen3-8b/qwen3-8b-bf16.gguf}"
LIMIT="${LIMIT:-20}"
MAXSEQ="${MAXSEQ:-2048}"
PPS="${PPS:-1 2 4}"

4
tools/bench/run_tp_parallel.sh
View File

@@ -7,8 +7,8 @@
# Run from the repo root on the GPU host. Produces bench-out/tp{1,2,4}-{xserv,llama}.
set -u
MODEL="${MODEL:-/opt/wjh/models/qwen3-8b}"
GGUF="${GGUF:-/opt/wjh/models/qwen3-8b/qwen3-8b-bf16.gguf}"
MODEL="${MODEL:-$HOME/models/qwen3-8b}"
GGUF="${GGUF:-$HOME/models/qwen3-8b/qwen3-8b-bf16.gguf}"
LIMIT="${LIMIT:-30}"
MAXSEQ="${MAXSEQ:-2048}"
TPS="${TPS:-1 2 4}"