Upload 2025-09-10/runs/5072-17607958977/ci_results_run_models_gpu/model_results.json with huggingface_hub

2025-09-10/runs/5072-17607958977/ci_results_run_models_gpu/model_results.json

@@ -0,0 +1,689 @@
+{
+    "models_gpt_oss": {
+        "failed": {
+            "PyTorch": {
+                "unclassified": 0,
+                "single": 78,
+                "multi": 76
+            },
+            "TensorFlow": {
+                "unclassified": 0,
+                "single": 0,
+                "multi": 0
+            },
+            "Flax": {
+                "unclassified": 0,
+                "single": 0,
+                "multi": 0
+            },
+            "Tokenizers": {
+                "unclassified": 0,
+                "single": 0,
+                "multi": 0
+            },
+            "Pipelines": {
+                "unclassified": 0,
+                "single": 0,
+                "multi": 0
+            },
+            "Trainer": {
+                "unclassified": 0,
+                "single": 0,
+                "multi": 0
+            },
+            "ONNX": {
+                "unclassified": 0,
+                "single": 0,
+                "multi": 0
+            },
+            "Auto": {
+                "unclassified": 0,
+                "single": 0,
+                "multi": 0
+            },
+            "Quantization": {
+                "unclassified": 0,
+                "single": 0,
+                "multi": 0
+            },
+            "Unclassified": {
+                "unclassified": 0,
+                "single": 0,
+                "multi": 0
+            }
+        },
+        "errors": 0,
+        "success": 183,
+        "skipped": 315,
+        "time_spent": [
+            2650.45,
+            1988.38
+        ],
+        "failures": {
+            "multi": [
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssModelTest::test_flash_attn_2_inference_equivalence",
+                    "trace": "(line 3609)  AssertionError: Tensor-likes are not close!"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssModelTest::test_flash_attn_2_inference_equivalence_right_padding",
+                    "trace": "(line 3609)  AssertionError: Tensor-likes are not close!"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_matches_original_120b",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.80 GiB is free. Process 90201 has 20.38 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_matches_original_20b",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.54 GiB is free. Process 90201 has 20.64 GiB memory in use. Of the allocated memory 18.92 GiB is allocated by PyTorch, and 1.24 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_00",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.54 GiB is free. Process 90201 has 20.64 GiB memory in use. Of the allocated memory 18.92 GiB is allocated by PyTorch, and 1.24 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_01",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.54 GiB is free. Process 90201 has 20.64 GiB memory in use. Of the allocated memory 18.92 GiB is allocated by PyTorch, and 1.24 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_02",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.54 GiB is free. Process 90201 has 20.64 GiB memory in use. Of the allocated memory 18.92 GiB is allocated by PyTorch, and 1.24 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_03",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.54 GiB is free. Process 90201 has 20.64 GiB memory in use. Of the allocated memory 18.92 GiB is allocated by PyTorch, and 1.24 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_04",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.54 GiB is free. Process 90201 has 20.64 GiB memory in use. Of the allocated memory 18.92 GiB is allocated by PyTorch, and 1.24 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_05",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.54 GiB is free. Process 90201 has 20.64 GiB memory in use. Of the allocated memory 18.92 GiB is allocated by PyTorch, and 1.24 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_06",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.54 GiB is free. Process 90201 has 20.64 GiB memory in use. Of the allocated memory 18.92 GiB is allocated by PyTorch, and 1.24 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_07",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.54 GiB is free. Process 90201 has 20.64 GiB memory in use. Of the allocated memory 18.92 GiB is allocated by PyTorch, and 1.24 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_08",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.54 GiB is free. Process 90201 has 20.64 GiB memory in use. Of the allocated memory 18.92 GiB is allocated by PyTorch, and 1.24 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_09",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.54 GiB is free. Process 90201 has 20.64 GiB memory in use. Of the allocated memory 18.92 GiB is allocated by PyTorch, and 1.24 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_10",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.54 GiB is free. Process 90201 has 20.64 GiB memory in use. Of the allocated memory 18.92 GiB is allocated by PyTorch, and 1.24 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_11",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.54 GiB is free. Process 90201 has 20.64 GiB memory in use. Of the allocated memory 18.92 GiB is allocated by PyTorch, and 1.24 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_12",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.54 GiB is free. Process 90201 has 20.64 GiB memory in use. Of the allocated memory 18.92 GiB is allocated by PyTorch, and 1.24 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_13",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.54 GiB is free. Process 90201 has 20.64 GiB memory in use. Of the allocated memory 18.92 GiB is allocated by PyTorch, and 1.24 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_14",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.54 GiB is free. Process 90201 has 20.64 GiB memory in use. Of the allocated memory 18.92 GiB is allocated by PyTorch, and 1.24 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_15",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.54 GiB is free. Process 90201 has 20.64 GiB memory in use. Of the allocated memory 18.92 GiB is allocated by PyTorch, and 1.24 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_16",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.80 GiB is free. Process 90201 has 20.38 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_17",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.80 GiB is free. Process 90201 has 20.38 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_18",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.80 GiB is free. Process 90201 has 20.38 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_19",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.80 GiB is free. Process 90201 has 20.38 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_20",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.80 GiB is free. Process 90201 has 20.38 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_21",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.80 GiB is free. Process 90201 has 20.38 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_22",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.80 GiB is free. Process 90201 has 20.38 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_23",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.80 GiB is free. Process 90201 has 20.38 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_24",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.80 GiB is free. Process 90201 has 20.38 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_25",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.80 GiB is free. Process 90201 has 20.38 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_26",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.80 GiB is free. Process 90201 has 20.38 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_27",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.80 GiB is free. Process 90201 has 20.38 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_28",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.80 GiB is free. Process 90201 has 20.38 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_29",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.80 GiB is free. Process 90201 has 20.38 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_30",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.80 GiB is free. Process 90201 has 20.38 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_31",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.80 GiB is free. Process 90201 has 20.38 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_00",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmptycy_o2z_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_01",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmpaqmulmni_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_02",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmpuohiu2kx_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_03",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmpjax84abr_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_04",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmp8tjqub0b_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_05",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmp2ts67axd_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_06",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmpyhgibzov_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_07",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmpjmh8vnbb_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_08",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmp7cz3n7vm_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_09",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmpcgyzpow6_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_10",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmpwi53ymmm_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_11",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmpcr3tx26s_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_12",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmp69owp634_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_13",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmpa1noiqya_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_14",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmpi8fi8hzq_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_15",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmpr80pv8ic_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_16",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmpjqwkp3cl_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_17",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmpafcjvg2g_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_18",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmpv1rkwuzu_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_19",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmpg10bo_86_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_20",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmpl8z8ya1q_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_21",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmpayr9yt1r_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_22",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmpyk0m5j5b_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_23",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmplo58ejpi_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_24",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmpiiabzhrn_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_25",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmpxlu4pu2m_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_26",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmp8apjv0y8_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_27",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmphgyc8656_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_28",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmp0nxci5tx_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_29",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmp3a4tit07_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_30",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmpv_asvy9a_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_31",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=2', '/tmp/tmpowuzoszo_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_training_step_01",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.54 GiB is free. Process 90201 has 20.64 GiB memory in use. Of the allocated memory 18.92 GiB is allocated by PyTorch, and 1.24 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_training_step_03",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.54 GiB is free. Process 90201 has 20.64 GiB memory in use. Of the allocated memory 18.92 GiB is allocated by PyTorch, and 1.24 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_training_step_05",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.54 GiB is free. Process 90201 has 20.64 GiB memory in use. Of the allocated memory 18.92 GiB is allocated by PyTorch, and 1.24 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_training_step_07",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.54 GiB is free. Process 90201 has 20.64 GiB memory in use. Of the allocated memory 18.92 GiB is allocated by PyTorch, and 1.24 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_training_step_17",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.80 GiB is free. Process 90201 has 20.38 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_training_step_19",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.80 GiB is free. Process 90201 has 20.38 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_training_step_21",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.80 GiB is free. Process 90201 has 20.38 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_training_step_23",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.80 GiB is free. Process 90201 has 20.38 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                }
+            ],
+            "single": [
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssModelTest::test_flash_attention_2_padding_matches_padding_free_with_position_ids",
+                    "trace": "(line 4304)  AssertionError: Tensor-likes are not close!"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssModelTest::test_flash_attention_2_padding_matches_padding_free_with_position_ids_and_fa_kwargs",
+                    "trace": "(line 4304)  AssertionError: Tensor-likes are not close!"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssModelTest::test_flash_attn_2_inference_equivalence",
+                    "trace": "(line 3609)  AssertionError: Tensor-likes are not close!"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssModelTest::test_flash_attn_2_inference_equivalence_right_padding",
+                    "trace": "(line 3609)  AssertionError: Tensor-likes are not close!"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_matches_original_120b",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.90 GiB is free. Process 69563 has 20.28 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_matches_original_20b",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.59 GiB is free. Process 69563 has 20.58 GiB memory in use. Of the allocated memory 17.38 GiB is allocated by PyTorch, and 2.83 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_00",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.59 GiB is free. Process 69563 has 20.58 GiB memory in use. Of the allocated memory 17.38 GiB is allocated by PyTorch, and 2.83 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_01",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.59 GiB is free. Process 69563 has 20.58 GiB memory in use. Of the allocated memory 17.38 GiB is allocated by PyTorch, and 2.83 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_02",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.59 GiB is free. Process 69563 has 20.58 GiB memory in use. Of the allocated memory 17.38 GiB is allocated by PyTorch, and 2.83 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_03",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.59 GiB is free. Process 69563 has 20.58 GiB memory in use. Of the allocated memory 17.38 GiB is allocated by PyTorch, and 2.83 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_04",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.59 GiB is free. Process 69563 has 20.58 GiB memory in use. Of the allocated memory 17.38 GiB is allocated by PyTorch, and 2.83 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_05",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.59 GiB is free. Process 69563 has 20.58 GiB memory in use. Of the allocated memory 17.38 GiB is allocated by PyTorch, and 2.83 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_06",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.59 GiB is free. Process 69563 has 20.58 GiB memory in use. Of the allocated memory 17.38 GiB is allocated by PyTorch, and 2.83 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_07",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.59 GiB is free. Process 69563 has 20.58 GiB memory in use. Of the allocated memory 17.38 GiB is allocated by PyTorch, and 2.83 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_08",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.59 GiB is free. Process 69563 has 20.58 GiB memory in use. Of the allocated memory 17.38 GiB is allocated by PyTorch, and 2.83 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_09",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.59 GiB is free. Process 69563 has 20.58 GiB memory in use. Of the allocated memory 17.38 GiB is allocated by PyTorch, and 2.83 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_10",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.59 GiB is free. Process 69563 has 20.58 GiB memory in use. Of the allocated memory 17.38 GiB is allocated by PyTorch, and 2.83 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_11",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.59 GiB is free. Process 69563 has 20.58 GiB memory in use. Of the allocated memory 17.38 GiB is allocated by PyTorch, and 2.83 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_12",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.59 GiB is free. Process 69563 has 20.58 GiB memory in use. Of the allocated memory 17.38 GiB is allocated by PyTorch, and 2.83 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_13",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.59 GiB is free. Process 69563 has 20.58 GiB memory in use. Of the allocated memory 17.38 GiB is allocated by PyTorch, and 2.83 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_14",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.59 GiB is free. Process 69563 has 20.58 GiB memory in use. Of the allocated memory 17.38 GiB is allocated by PyTorch, and 2.83 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_15",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.59 GiB is free. Process 69563 has 20.58 GiB memory in use. Of the allocated memory 17.38 GiB is allocated by PyTorch, and 2.83 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_16",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.90 GiB is free. Process 69563 has 20.28 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_17",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.90 GiB is free. Process 69563 has 20.28 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_18",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.90 GiB is free. Process 69563 has 20.28 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_19",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.90 GiB is free. Process 69563 has 20.28 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_20",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.90 GiB is free. Process 69563 has 20.28 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_21",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.90 GiB is free. Process 69563 has 20.28 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_22",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.90 GiB is free. Process 69563 has 20.28 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_23",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.90 GiB is free. Process 69563 has 20.28 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_24",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.90 GiB is free. Process 69563 has 20.28 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_25",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.90 GiB is free. Process 69563 has 20.28 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_26",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.90 GiB is free. Process 69563 has 20.28 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_27",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.90 GiB is free. Process 69563 has 20.28 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_28",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.90 GiB is free. Process 69563 has 20.28 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_29",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.90 GiB is free. Process 69563 has 20.28 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_30",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.90 GiB is free. Process 69563 has 20.28 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_31",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.90 GiB is free. Process 69563 has 20.28 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_00",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmpg6p8blkw_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_01",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmppwy9yn13_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_02",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmpaqijxrey_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_03",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmp0sfq4tlo_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_04",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmprtm6tq5w_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_05",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmp7q7tgg15_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_06",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmpicqtbtoh_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_07",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmpzbgufi33_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_08",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmp0ks0p8oh_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_09",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmp3y1i9cwd_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_10",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmp0w96ib9d_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_11",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmp34oaqojc_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_12",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmpjiktwwo7_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_13",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmpgbf2h54l_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_14",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmpv3p78egd_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_15",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmp17n2kvpo_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_16",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmpogwu8r51_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_17",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmpm_1fz4fv_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_18",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmpn1ga7zba_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_19",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmp2f4yib4l_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_20",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmpi7un05dp_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_21",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmp16gppz91_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_22",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmppyqrzmda_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_23",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmpkw6z7ldg_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_24",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmplfwft9jc_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_25",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmpuhgzu7ds_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_26",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmpktd5gsek_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_27",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmpikv1ea2__worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_28",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmp2ym_23yt_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_29",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmpwic9ogra_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_30",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmpsyfshwxy_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_model_outputs_distributed_31",
+                    "trace": "(line 526)  subprocess.CalledProcessError: Command '['torchrun', '--nproc_per_node=1', '/tmp/tmpvfv41a6b_worker.py']' returned non-zero exit status 1."
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_training_step_01",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.59 GiB is free. Process 69563 has 20.58 GiB memory in use. Of the allocated memory 17.38 GiB is allocated by PyTorch, and 2.83 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_training_step_03",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.59 GiB is free. Process 69563 has 20.58 GiB memory in use. Of the allocated memory 17.38 GiB is allocated by PyTorch, and 2.83 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_training_step_05",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.59 GiB is free. Process 69563 has 20.58 GiB memory in use. Of the allocated memory 17.38 GiB is allocated by PyTorch, and 2.83 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_training_step_07",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.98 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.59 GiB is free. Process 69563 has 20.58 GiB memory in use. Of the allocated memory 17.38 GiB is allocated by PyTorch, and 2.83 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_training_step_17",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.90 GiB is free. Process 69563 has 20.28 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_training_step_19",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.90 GiB is free. Process 69563 has 20.28 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_training_step_21",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.90 GiB is free. Process 69563 has 20.28 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                },
+                {
+                    "line": "tests/models/gpt_oss/test_modeling_gpt_oss.py::GptOssIntegrationTest::test_training_step_23",
+                    "trace": "(line 117)  torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 4.00 GiB. GPU 0 has a total capacity of 22.18 GiB of which 1.90 GiB is free. Process 69563 has 20.28 GiB memory in use. Of the allocated memory 17.72 GiB is allocated by PyTorch, and 2.18 GiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation.  See documentation for Memory Management  (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)"
+                }
+            ]
+        },
+        "job_link": {
+            "multi": "https://github.com/huggingface/transformers/actions/runs/17607958977/job/50023238960",
+            "single": "https://github.com/huggingface/transformers/actions/runs/17607958977/job/50023238940"
+        }
+    }
+}