fix(recipes): MB-922 reduce DeepSeek VR200 memory pressure

src/megatron/bridge/perf_recipes/deepseek/vr200/deepseek_v3.py

@@ -243,6 +243,7 @@ def deepseek_v3_pretrain_128gpu_vr200_nvfp4_config() -> ConfigContainer:
 def deepseek_v3_pretrain_256gpu_vr200_bf16_config() -> ConfigContainer:
     """DeepSeek V3 pretrain: 256× VR200, BF16 (alias of GB300)."""
     cfg = deepseek_v3_pretrain_256gpu_gb300_bf16_config()
+    cfg.dataset.num_workers = 0
     # Keep process settings next to the recipe so users can see the exact benchmark environment.
     cfg.env_vars = {
         **COMMON_PERF_ENV_VARS,
@@ -329,6 +330,7 @@ def deepseek_v3_pretrain_256gpu_vr200_fp8mx_config() -> ConfigContainer:
 def deepseek_v3_pretrain_256gpu_vr200_nvfp4_config() -> ConfigContainer:
     """DeepSeek V3 pretrain: 256× VR200, NVFP4 (alias of GB300)."""
     cfg = deepseek_v3_pretrain_256gpu_gb300_nvfp4_config()
+    cfg.dataset.num_workers = 0
     # Keep process settings next to the recipe so users can see the exact benchmark environment.
     cfg.env_vars = {
         **COMMON_PERF_ENV_VARS,

tests/unit_tests/recipes/test_deepseek_perf_recipes.py

@@ -0,0 +1,54 @@
+# Copyright (c) 2026, NVIDIA CORPORATION.  All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+"""Unit tests for DeepSeek flat performance recipes."""
+
+import pytest
+
+from megatron.bridge.perf_recipes.deepseek import (
+    deepseek_v3_pretrain_256gpu_vr200_bf16_config,
+    deepseek_v3_pretrain_256gpu_vr200_nvfp4_config,
+)
+from tests.unit_tests.recipes.recipe_test_utils import patch_recipe_construction_dependencies
+
+
+pytestmark = pytest.mark.unit
+
+
+@pytest.fixture(autouse=True)
+def _keep_recipe_construction_offline(monkeypatch: pytest.MonkeyPatch) -> None:
+    patch_recipe_construction_dependencies(monkeypatch)
+
+
+def test_deepseek_v3_256gpu_vr200_bf16_limits_host_memory() -> None:
+    cfg = deepseek_v3_pretrain_256gpu_vr200_bf16_config()
+
+    assert cfg.model.pipeline_model_parallel_size == 4
+    assert cfg.model.virtual_pipeline_model_parallel_size == 4
+    assert cfg.model.expert_model_parallel_size == 64
+    assert cfg.train.micro_batch_size == 1
+    assert cfg.model.recompute_modules == ["moe_act"]
+    assert cfg.dataset.num_workers == 0
+
+
+def test_deepseek_v3_256gpu_vr200_nvfp4_limits_host_memory() -> None:
+    cfg = deepseek_v3_pretrain_256gpu_vr200_nvfp4_config()
+
+    assert cfg.model.pipeline_model_parallel_size == 2
+    assert cfg.model.virtual_pipeline_model_parallel_size == 8
+    assert cfg.model.expert_model_parallel_size == 32
+    assert cfg.train.micro_batch_size == 2
+    assert cfg.model.recompute_modules == ["mla_up_proj"]
+    assert cfg.dataset.num_workers == 0
+    assert cfg.env_vars["NUM_OF_HYBRID_EP_RANKS_PER_NVLINK_DOMAIN"] == 32