ci: add LongCat-Flash-Lite-FP8 8-GPU nightly test + fix NextN rope_theta (#32125)

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com> Co-authored-by: Xiaoyu Zhang <1182563586@qq.com>
2026-07-24 19:45:28 +08:00
parent 841fa293b5
commit 8389d79e43
2 changed files with 89 additions and 2 deletions
@@ -131,8 +131,12 @@ class LongcatFlashDenseDecoderLayer(nn.Module):
            v_head_dim=config.v_head_dim,
            q_lora_rank=config.q_lora_rank,
            kv_lora_rank=config.kv_lora_rank,
-            rope_theta=config.rope_parameters["rope_theta"],
-            rope_scaling=None,
+            rope_theta=(
+                config.rope_parameters["rope_theta"]
+                if "rope_theta" in getattr(config, "rope_parameters", {})
+                else config.rope_theta
+            ),
+            rope_scaling=getattr(config, "rope_scaling", None),
            max_position_embeddings=config.max_position_embeddings,
            quant_config=quant_config,
            layer_id=layer_id,
@@ -0,0 +1,83 @@
+import unittest
+
+from sglang.test.accuracy_test_runner import AccuracyTestParams
+from sglang.test.ci.ci_register import register_cuda_ci
+from sglang.test.performance_test_runner import PerformanceTestParams
+from sglang.test.run_combined_tests import run_combined_tests
+from sglang.test.test_utils import ModelLaunchSettings
+
+# Runs on both H200 and B200 via the nightly-8-gpu-common suite.
+register_cuda_ci(est_time=1200, suite="nightly-8-gpu-common", nightly=True)
+
+# LongCat-Flash-Lite-FP8 is the smallest member of the LongCat family
+# (~138 GB FP8 weights, hidden=3072, 14 layers, 256 routed + 128 zero
+# experts, moe_topk=12). It fits comfortably on a single 8-GPU node
+# (~17 GB/GPU of weights at tp8), so it is the CI regression proxy for the
+# much larger LongCat-2.0-FP8 (2 TB, needs >=16 GPUs and cannot run on any
+# current single-node CUDA runner).
+#
+# It shares the SAME sglang model file (models/longcat_flash.py, arch name
+# LongcatFlashNgramForCausalLM is remapped to model_type "longcat_flash")
+# with LongCat-Flash-Chat-FP8 and LongCat-2.0-FP8, so it guards the shared
+# code paths that recent LongCat EP fixes touched:
+#   - the scheduler moe-topk gate for --moe-a2a-backend (PR #30975)
+#   - ScMoE dense-branch gather (RoPE) + the MoE-vs-DeepEPMoE double
+#     all_reduce fix (PR #31311)
+#   - the zero-expert (identity) compute path (zero_expert_num=128) and
+#     NgramEmbedding (ngram_vocab_size_ratio) — 2.0-specific features that
+#     Flash-Chat-FP8 does NOT have, but Lite DOES.
+# It does NOT cover DSA (LongCat Sparse Attention) or MTP/NextN, which are
+# unique to LongCat-2.0 and require a >=16-GPU e2e run.
+LONGCAT_FLASH_LITE_FP8_MODEL_PATH = "meituan-longcat/LongCat-Flash-Lite-FP8"
+
+COMMON_ARGS = [
+    "--trust-remote-code",
+    "--mem-fraction-static=0.85",
+]
+
+
+class TestLongCatFlashLiteFp8(unittest.TestCase):
+    """LongCat-Flash-Lite-FP8 on H200/B200 (8-GPU).
+
+    Two variants exercise the two MoE all-to-all backends that the LongCat
+    EP fixes gate on:
+      - EP8 + deepep : real expert parallelism (the path #30975/#31311 fix)
+      - EP8 + none   : EP-over-TP baseline (all_reduce / gather correctness)
+    """
+
+    def test_longcat_flash_lite_fp8(self):
+        variants = [
+            ModelLaunchSettings(
+                LONGCAT_FLASH_LITE_FP8_MODEL_PATH,
+                tp_size=8,
+                extra_args=COMMON_ARGS + ["--ep=8"],
+                variant="TP8+EP8+none",
+            ),
+            ModelLaunchSettings(
+                LONGCAT_FLASH_LITE_FP8_MODEL_PATH,
+                tp_size=8,
+                extra_args=COMMON_ARGS + ["--ep=8", "--moe-a2a-backend=deepep"],
+                variant="TP8+EP8+deepep",
+            ),
+        ]
+
+        run_combined_tests(
+            models=variants,
+            test_name="LongCat-Flash-Lite-FP8",
+            # Measured 2026-07-22 on 8xH100-80GB, gsm8k 200q, 5-shot, greedy:
+            #   TP8+EP8+none   -> 0.840
+            #   TP8+EP8+deepep -> 0.820
+            # Floor set to 0.78 to absorb 200-sample noise.
+            accuracy_params=AccuracyTestParams(
+                dataset="gsm8k",
+                baseline_accuracy=0.78,
+                num_examples=200,
+            ),
+            performance_params=PerformanceTestParams(
+                profile_dir="performance_profiles_longcat_flash_lite_fp8",
+            ),
+        )
+
+
+if __name__ == "__main__":
+    unittest.main()