Add nightly test for GLM5.2 LayerSplit (#31512)

test/registered/models_e2e/test_dsa_glm52_cache_layer_split.py

@@ -9,8 +9,7 @@ PD-disaggregated GLM-5.2 deployment: a layer-split prefill worker running
 interleave prefill-CP + layer split, and an ordinary decode worker that receives
 full cache shards via PD transfer.
 
-Sized for the 4-GPU B200 runner (prefill TP=2 + decode TP=2) rather than an
+Runs nightly on an 8-GPU B200 runner (prefill TP=4 + decode TP=4).
-8-GPU deployment, since the 8-gpu-b200 runner is nightly-only.
 """
 
 import unittest
@@ -22,10 +21,9 @@ from sglang.test.server_fixtures.disaggregation_fixture import (
 )
 
 register_cuda_ci(
-    est_time=1200,
+    est_time=450,
-    stage="extra-b",
+    suite="nightly-8-gpu-b200",
-    runner_config="4-gpu-b200",
+    nightly=True,
-    disabled="Temporarily disabled",
 )
 
 
@@ -36,13 +34,15 @@ class TestGLM52DSACacheLayerSplit(PDDisaggregationServerBase, GSM8KMixin):
     gsm8k_accuracy_thres = 0.935
     gsm8k_num_questions = 1319
     gsm8k_num_threads = 200
-    gsm8k_num_shots = 0
+    gsm8k_num_shots = 20
 
-    # Prefill worker: interleave prefill-CP + DSA cache layer split on 2 GPUs
+    # Prefill worker: interleave prefill-CP + DSA cache layer split on 4 GPUs
-    # (TP=2 -> attn_cp_size=2, so KV/indexer layers shard 2-way across CP ranks).
+    # (TP=4 -> attn_cp_size=4, so KV/indexer layers shard 4-way across CP ranks).
     extra_prefill_args = [
         "--tp",
-        "2",
+        "4",
+        "--attn-cp-size",
+        "4",
         "--dsa-prefill-backend",
         "trtllm",
         "--kv-cache-dtype",
@@ -57,20 +57,36 @@ class TestGLM52DSACacheLayerSplit(PDDisaggregationServerBase, GSM8KMixin):
         "4096",
         "--max-prefill-tokens",
         "4096",
+        "--speculative-algorithm",
+        "EAGLE",
+        "--speculative-num-steps",
+        "5",
+        "--speculative-eagle-topk",
+        "1",
+        "--speculative-num-draft-tokens",
+        "6",
     ]
-    # Decode worker: ordinary local decode cache on the other 2 GPUs, receives
+    # Decode worker: ordinary local decode cache, receives full shards via PD
-    # full shards via PD transfer.
+    # transfer.
     extra_decode_args = [
         "--tp",
-        "2",
+        "4",
         "--dsa-decode-backend",
         "trtllm",
         "--kv-cache-dtype",
         "fp8_e4m3",
         "--mem-fraction-static",
         "0.85",
+        "--speculative-algorithm",
+        "EAGLE",
+        "--speculative-num-steps",
+        "5",
+        "--speculative-eagle-topk",
+        "1",
+        "--speculative-num-draft-tokens",
+        "6",
         "--base-gpu-id",
-        "2",
+        "4",
     ]
 
     @classmethod