Add DeepSeek-V4-Pro FP8 MI355X SGLang benchmark

.github/configs/amd-master.yaml

@@ -1472,3 +1472,20 @@ dsr1-fp4-mi355x-sglang-disagg-mtp:
         - "DECODE_NODES=1"
         - "DECODE_MTP_SIZE=1"
 
+dsv4-fp8-mi355x-sglang:
+  image: rocm/sgl-dev:deepseek-v4-mi35x
+  model: sgl-project/DeepSeek-V4-Pro-FP8
+  model-prefix: dsv4
+  runner: mi355x
+  precision: fp8
+  framework: sglang
+  multinode: false
+  seq-len-configs:
+  - isl: 1024
+    osl: 1024
+    search-space:
+    - { tp: 8, conc-start: 4, conc-end: 64 }
+  - isl: 8192
+    osl: 1024
+    search-space:
+    - { tp: 8, conc-start: 4, conc-end: 64 }

AGENTS.md

@@ -112,6 +112,7 @@ When working with benchmark configurations, use these valid values:
 
 **Models (model-prefix)**:
 - `dsr1` - DeepSeek-R1-0528
+- `dsv4` - DeepSeek-V4-Pro
 - `gptoss` - GPT-OSS-120B
 
 **Precisions**:

benchmarks/single_node/dsv4_fp8_mi355x.sh

@@ -0,0 +1,120 @@
+#!/usr/bin/env bash
+
+source "$(dirname "$0")/../benchmark_lib.sh"
+
+check_env_vars \
+    MODEL \
+    TP \
+    CONC \
+    ISL \
+    OSL \
+    RANDOM_RANGE_RATIO \
+    RESULT_FILENAME
+
+if [[ -n "$SLURM_JOB_ID" ]]; then
+  echo "JOB $SLURM_JOB_ID running on $SLURMD_NODENAME"
+fi
+
+hf download "$MODEL"
+
+# Transformers in the container doesn't recognize the `deepseek_v4` model_type.
+# PR #23608's fallback in hf_transformers_utils.get_config tries to handle this
+# by writing a patched config to /tmp, but in practice isn't catching the error
+# in this image. Patch the cached config.json directly instead: set model_type
+# to `deepseek_v3` so AutoConfig.from_pretrained succeeds, and keep
+# architectures=['DeepseekV4ForCausalLM'] so SGLang dispatches to its native
+# DSv4 model class (python/sglang/srt/models/deepseek_v4.py).
+python3 << PYEOF
+import json
+from huggingface_hub import hf_hub_download
+path = hf_hub_download(repo_id="$MODEL", filename="config.json")
+with open(path) as f:
+    config = json.load(f)
+if config.get("model_type") == "deepseek_v4":
+    config["model_type"] = "deepseek_v3"
+    with open(path, "w") as f:
+        json.dump(config, f, indent=2)
+    print(f"Patched {path}: model_type deepseek_v4 -> deepseek_v3")
+else:
+    print(f"No patch needed: model_type is {config.get('model_type')!r}")
+PYEOF
+
+# DSv4-specific SGLang env vars (from sgl-project/sglang#23608)
+export SGLANG_OPT_USE_FUSED_COMPRESS=false
+export SGLANG_OPT_USE_OLD_COMPRESSOR=true
+export SGLANG_OPT_USE_TILELANG_SWA_PREPARE=false
+export SGLANG_OPT_USE_JIT_KERNEL_FUSED_TOPK=false
+export SGLANG_OPT_USE_FUSED_HASH_TOPK=false
+export SGLANG_HACK_FLASHMLA_BACKEND=torch
+export SGLANG_OPT_DEEPGEMM_HC_PRENORM=false
+export SGLANG_OPT_USE_TILELANG_MHC_PRE=false
+export SGLANG_OPT_USE_TILELANG_MHC_POST=false
+export SGLANG_ENABLE_THINKING=1
+export SGLANG_USE_AITER=1
+export SGLANG_USE_ROCM700A=1
+export SGLANG_TOPK_TRANSFORM_512_TORCH=1
+export SGLANG_FP8_PAGED_MQA_LOGITS_TORCH=1
+export SGLANG_DSV4_FP4_EXPERTS=false
+export SGLANG_OPT_DPSK_V4_RADIX=0
+export SGLANG_OPT_USE_OVERLAP_STORE_CACHE=false
+export SGLANG_OPT_USE_FUSED_STORE_CACHE=false
+export SGLANG_FORCE_TRITON_MOE_FP8=1
+
+SERVER_LOG=/workspace/server.log
+PORT=${PORT:-8888}
+
+EVAL_CONTEXT_ARGS=""
+if [ "${EVAL_ONLY}" = "true" ]; then
+    setup_eval_context
+    EVAL_CONTEXT_ARGS="--context-length $EVAL_MAX_MODEL_LEN"
+fi
+# Start GPU monitoring (power, temperature, clocks every second)
+start_gpu_monitor
+
+python3 -m sglang.launch_server \
+    --model-path $MODEL \
+    --host=0.0.0.0 \
+    --port $PORT \
+    --tensor-parallel-size $TP \
+    --dp $TP \
+    --enable-dp-attention \
+    --trust-remote-code \
+    --disable-radix-cache \
+    --attention-backend compressed \
+    --max-running-request 256 \
+    --page-size 256 \
+    --chunked-prefill-size 4096 \
+    --disable-shared-experts-fusion \
+    --disable-cuda-graph \
+    --tool-call-parser deepseekv4 \
+    --reasoning-parser deepseek-v4 \
+    --mem-fraction-static 0.88 \
+    --max-total-tokens $((CONC * (ISL + OSL) + 200)) \
+    --watchdog-timeout 1800 $EVAL_CONTEXT_ARGS > $SERVER_LOG 2>&1 &
+
+SERVER_PID=$!
+
+# Wait for server to be ready
+wait_for_server_ready --port "$PORT" --server-log "$SERVER_LOG" --server-pid "$SERVER_PID"
+
+run_benchmark_serving \
+    --model "$MODEL" \
+    --port "$PORT" \
+    --backend vllm \
+    --input-len "$ISL" \
+    --output-len "$OSL" \
+    --random-range-ratio "$RANDOM_RANGE_RATIO" \
+    --num-prompts "$((CONC * 10))" \
+    --max-concurrency "$CONC" \
+    --result-filename "$RESULT_FILENAME" \
+    --result-dir /workspace/
+
+# After throughput, run evaluation only if RUN_EVAL is true
+if [ "${RUN_EVAL}" = "true" ]; then
+    run_eval --framework lm-eval --port "$PORT"
+    append_lm_eval_summary
+fi
+
+# Stop GPU monitoring
+stop_gpu_monitor
+set +x

perf-changelog.yaml

@@ -1766,3 +1766,12 @@
     - "Prefix caching disabled, no speculative decoding"
     - "Configs: 1k1k conc 4-1024, 8k1k conc 4-512"
   pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/1143
+
+- config-keys:
+    - dsv4-fp8-mi355x-sglang
+  description:
+    - "Day 0 DeepSeek-V4-Pro FP8 MI355X SGLang benchmark"
+    - "Image: rocm/sgl-dev:deepseek-v4-mi35x (from sgl-project/sglang#23608)"
+    - "Model: sgl-project/DeepSeek-V4-Pro-FP8"
+    - "https://github.com/sgl-project/sglang/pull/23608#issuecomment-4311952977"
+  pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/1134