XPK Start: Sat Apr 25 15:20:25 UTC 2026 2026-04-25 15:20:54.464649: E external/local_xla/xla/stream_executor/cuda/cuda_platform.cc:51] failed call to cuInit: INTERNAL: CUDA error: Failed call to cuInit: UNKNOWN ERROR (303) I0425 15:20:54.687363 139339677632320 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-25 15:21:03,729:jax._src.distributed:149: Starting JAX distributed service on [::]:8482 I0425 15:21:03.729579 139339677632320 distributed.py:149] Starting JAX distributed service on [::]:8482 INFO:2026-04-25 15:21:03,731:jax._src.distributed:166: Connecting to JAX distributed service on mt-03-sft-linen-ckpt-vgx7q-slice-job-0-0.mt-03-sft-linen-ckpt-vgx7q:8482 I0425 15:21:03.731930 139339677632320 distributed.py:166] Connecting to JAX distributed service on mt-03-sft-linen-ckpt-vgx7q-slice-job-0-0.mt-03-sft-linen-ckpt-vgx7q:8482 I0425 15:21:04.552050 139339677632320 max_utils.py:284] Jax distributed system initialized! I0425 15:21:10.572360 139339677632320 max_utils.py:800] System Information: Jax Version: 0.8.3 I0425 15:21:10.572468 139339677632320 max_utils.py:801] System Information: Jaxlib Version: 0.8.3 I0425 15:21:10.572517 139339677632320 max_utils.py:802] System Information: Jax Backend: PJRT C API TFRT TPU v6 lite Built on Dec 15 2025 14:03:46 (1765836226) cl/844590465 I0425 15:21:10.575888 139339677632320 maxtext_utils.py:1631] Num_devices: 32, shape (1, 4, 1, 8, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0425 15:21:10.669638 139339677632320 maxtext_utils.py:1631] Num_devices: 32, shape (1, 4, 1, 8, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0425 15:21:10.769145 139339677632320 maxtext_utils.py:1631] Num_devices: 32, shape (1, 4, 1, 8, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0425 15:21:11.810186 139339677632320 pytree_checkpoint_handler.py:577] save_device_host_concurrent_bytes=None I0425 15:21:11.810631 139339677632320 base_pytree_checkpoint_handler.py:411] Created BasePyTreeCheckpointHandler: use_ocdbt=True, use_zarr3=True, pytree_metadata_options=PyTreeMetadataOptions(support_rich_types=False), array_metadata_store=<orbax.checkpoint._src.metadata.array_metadata_store.Store object at 0x7eb9d65661e0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0425 15:21:11.810692 139339677632320 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.28 W0425 15:21:12.366390 139339677632320 checkpoint.py:202] Metadata file does not exist: gs://lance-maxtext/pt_seed_ckpts/pt_seed_ckpts/pt_seed_ckpt_gpt352k_linen/checkpoints/9/items/_CHECKPOINT_METADATA I0425 15:21:12.886984 1940 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0425 15:21:14.029629 139339677632320 checkpointer.py:304] Restoring checkpoint from gs://lance-maxtext/pt_seed_ckpts/pt_seed_ckpts/pt_seed_ckpt_gpt352k_linen/checkpoints/9/items. W0425 15:21:15.845864 139339677632320 transform_utils.py:230] The transformations API will eventually be replaced by an upgraded design. The current API will not be removed until this point, but it will no longer be actively worked on. I0425 15:21:15.846306 139339677632320 transform_utils.py:288] The following keys are not loaded from the original tree after applying specified transforms: params/params/decoder/dropout/rngs/aqt/count, params/params/decoder/dropout/rngs/aqt/key, params/params/decoder/dropout/rngs/dropout/count, params/params/decoder/dropout/rngs/dropout/key, params/params/decoder/dropout/rngs/params/count, params/params/decoder/dropout/rngs/params/key, params/params/decoder/layers/dropout/rngs/aqt/count, params/params/decoder/layers/dropout/rngs/aqt/key, params/params/decoder/layers/dropout/rngs/dropout/count, params/params/decoder/layers/dropout/rngs/dropout/key, params/params/decoder/layers/dropout/rngs/params/count, params/params/decoder/layers/dropout/rngs/params/key, params/params/decoder/layers/mlp/dropout/rngs/aqt/count, params/params/decoder/layers/mlp/dropout/rngs/aqt/key, params/params/decoder/layers/mlp/dropout/rngs/dropout/count, params/params/decoder/layers/mlp/dropout/rngs/dropout/key, params/params/decoder/layers/mlp/dropout/rngs/params/count, params/params/decoder/layers/mlp/dropout/rngs/params/key, params/params/decoder/layers/rngs/aqt/count, params/params/decoder/layers/rngs/aqt/key, params/params/decoder/layers/rngs/dropout/count, params/params/decoder/layers/rngs/dropout/key, params/params/decoder/layers/rngs/params/count, params/params/decoder/layers/rngs/params/key, params/params/decoder/rngs/aqt/count, params/params/decoder/rngs/aqt/key, params/params/decoder/rngs/dropout/count, params/params/decoder/rngs/dropout/key, params/params/decoder/rngs/params/count, params/params/decoder/rngs/params/key I0425 15:21:15.969077 139339677632320 checkpointer.py:318] Finished restoring checkpoint in 2.31 seconds from gs://lance-maxtext/pt_seed_ckpts/pt_seed_ckpts/pt_seed_ckpt_gpt352k_linen/checkpoints/9/items. I0425 15:21:16.036340 139339677632320 config.py:112] TensorFlow version 2.20.0 available. I0425 15:21:16.036850 139339677632320 config.py:125] JAX version 0.8.3 available. /deps/src/maxtext/input_pipeline/input_pipeline_utils.py:467: UserWarning: WARNING: Inefficient dataloading. Your train or eval dataset contains 3 shards, smaller than number of host loading data. This is known to lead to inefficient dataloading. Seegithub.com/google/maxtext/blob/main/getting_started/Data_Input_Pipeline.md#multihost-dataloading-best-practice warnings.warn( E0425 15:21:21.622207 139339677632320 packing.py:209] PackAndBatchOperation is deprecated. Please use lazy_dataset.FirstFitPackIterDataset instead. I0425 15:21:21.622436 139339677632320 data_loader.py:408] Adding CopyNumPyArrayToSharedMemory MapTransform. I0425 15:21:22.016263 139339677632320 pytree_checkpoint_handler.py:577] save_device_host_concurrent_bytes=None I0425 15:21:22.016411 139339677632320 base_pytree_checkpoint_handler.py:411] Created BasePyTreeCheckpointHandler: use_ocdbt=True, use_zarr3=False, pytree_metadata_options=PyTreeMetadataOptions(support_rich_types=False), array_metadata_store=<orbax.checkpoint._src.metadata.array_metadata_store.Store object at 0x7eb9d65661e0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0425 15:21:22.016458 139339677632320 pytree_checkpoint_handler.py:577] save_device_host_concurrent_bytes=None I0425 15:21:22.016493 139339677632320 base_pytree_checkpoint_handler.py:411] Created BasePyTreeCheckpointHandler: use_ocdbt=True, use_zarr3=False, pytree_metadata_options=PyTreeMetadataOptions(support_rich_types=False), array_metadata_store=<orbax.checkpoint._src.metadata.array_metadata_store.Store object at 0x7eb9d65661e0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0425 15:21:22.016537 139339677632320 checkpoint_manager.py:702] [process=6][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=None, item_handlers={'model_params': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7eb74f9120c0>, 'optimizer_state': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ea19e074770>, 'custom_metadata': <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ea19e074bc0>}, handler_registry=None I0425 15:21:22.016756 139339677632320 composite_checkpoint_handler.py:237] Deferred registration for item: "model_params". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7eb74f9120c0>` for item "model_params" and save args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>` to `_handler_registry`. I0425 15:21:22.016801 139339677632320 composite_checkpoint_handler.py:237] Deferred registration for item: "optimizer_state". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ea19e074770>` for item "optimizer_state" and save args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>` to `_handler_registry`. I0425 15:21:22.016829 139339677632320 composite_checkpoint_handler.py:237] Deferred registration for item: "custom_metadata". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ea19e074bc0>` for item "custom_metadata" and save args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>` to `_handler_registry`. I0425 15:21:22.016855 139339677632320 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7eb9db486270>` for item "metrics" and save args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>` to `_handler_registry`. I0425 15:21:22.016883 139339677632320 composite_checkpoint_handler.py:505] Initialized registry DefaultCheckpointHandlerRegistry({('model_params', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7eb74f9120c0>, ('model_params', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7eb74f9120c0>, ('optimizer_state', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ea19e074770>, ('optimizer_state', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ea19e074770>, ('custom_metadata', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ea19e074bc0>, ('custom_metadata', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ea19e074bc0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7eb9db486270>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7eb9db486270>}). I0425 15:21:22.017289 139339677632320 async_checkpointer.py:177] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7ea19e81c5e0> timeout: 600 secs and primary_host=0 for async checkpoint writes I0425 15:21:24.662272 139339677632320 checkpoint_manager.py:1788] Found 0 checkpoint steps in gs://lance-maxtext/pt_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150421/pt_sft_nnx_xpk_feat_nnx_set_defaults_true_20260425_150421_03_sft_linen_ckpt/checkpoints I0425 15:21:25.102928 139339677632320 checkpoint_manager.py:921] [process=6][thread=MainThread] CheckpointManager created, primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=10000, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_hns=False, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=None, preservation_policy=None, prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False), root_directory=gs://lance-maxtext/pt_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150421/pt_sft_nnx_xpk_feat_nnx_set_defaults_true_20260425_150421_03_sft_linen_ckpt/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7ea19e074c20> I0425 15:21:25.103297 139339677632320 peft_trainer.py:590] Training with mesh: Mesh('diloco': 1, 'data': 4, 'stage': 1, 'fsdp': 8, 'fsdp_transpose': 1, 'sequence': 1, 'context': 1, 'context_autoregressive': 1, 'tensor': 1, 'tensor_transpose': 1, 'tensor_sequence': 1, 'expert': 1, 'autoregressive': 1, axis_types=(Auto, Auto, Auto, Auto, Auto, Auto, Auto, Auto, Auto, Auto, Auto, Auto, Auto)) I0425 15:21:25.516936 139339677632320 peft_trainer.py:600] Compiled train_step cache size: 0 Training: 0%| | 0/5 [00:00<?, ?step/s]I0425 15:21:25.521145 139339677632320 metric_logger.py:289] number parameters: 0.000 billion I0425 15:21:25.523363 139185566865152 grain_pool.py:367] Grain pool will use 1 processes. I0425 15:21:25.549615 139185566865152 grain_pool.py:440] Grain pool will start child processes. Per train step: Total TFLOPs: 0.00 split as 54.29% learnable weight flops and 45.71% attention flops I0425 15:21:25.554783 139185566865152 grain_pool.py:448] Grain pool started all child processes. 2026-04-25 15:21:29.488402: I tensorflow/core/util/port.cc:153] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`. 2026-04-25 15:21:29.534289: I tensorflow/core/platform/cpu_feature_guard.cc:210] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations. To enable the following instructions: AVX2 AVX512F AVX512_VNNI AVX512_BF16 FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags. 2026-04-25 15:21:30.524543: I tensorflow/core/util/port.cc:153] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`. 2026-04-25 15:21:34.748021: E external/local_xla/xla/stream_executor/cuda/cuda_platform.cc:51] failed call to cuInit: INTERNAL: CUDA error: Failed call to cuInit: UNKNOWN ERROR (303) I0425 15:21:40.726112 139339677632320 utils.py:86] Train loop finished in: 15.2038 seconds Traceback (most recent call last): File "<frozen runpy>", line 198, in _run_module_as_main File "<frozen runpy>", line 88, in _run_code File "/deps/src/maxtext/trainers/post_train/sft/train_sft.py", line 216, in <module> app.run(main) File "/usr/local/lib/python3.12/site-packages/absl/app.py", line 316, in run _run_main(main, args) File "/usr/local/lib/python3.12/site-packages/absl/app.py", line 261, in _run_main sys.exit(main(argv)) ^^^^^^^^^^ File "/deps/src/maxtext/trainers/post_train/sft/train_sft.py", line 212, in main train(mt_config, goodput_recorder) File "/deps/src/maxtext/trainers/post_train/sft/train_sft.py", line 189, in train trainer = train_model(mt_config, trainer, mesh) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/deps/src/maxtext/trainers/post_train/sft/train_sft.py", line 175, in train_model trainer.train(trainer.data_hooks.train_data_iterator, trainer.data_hooks.eval_data_iterator) File "/usr/local/lib/python3.12/site-packages/tunix/sft/peft_trainer.py", line 659, in train train_example = sharding_utils.shard_input( ^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/local/lib/python3.12/site-packages/tunix/sft/sharding_utils.py", line 58, in shard_input return jax.tree.map( ^^^^^^^^^^^^^ File "/usr/local/lib/python3.12/site-packages/jax/_src/tree.py", line 155, in map return tree_util.tree_map(f, tree, *rest, is_leaf=is_leaf) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/local/lib/python3.12/site-packages/jax/_src/tree_util.py", line 362, in tree_map return treedef.unflatten(f(*xs) for xs in zip(*all_leaves)) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/local/lib/python3.12/site-packages/jax/_src/tree_util.py", line 362, in <genexpr> return treedef.unflatten(f(*xs) for xs in zip(*all_leaves)) ^^^^^^ File "/usr/local/lib/python3.12/site-packages/tunix/sft/sharding_utils.py", line 59, in <lambda> lambda x: jax.make_array_from_process_local_data( ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/local/lib/python3.12/site-packages/jax/_src/array.py", line 986, in make_array_from_process_local_data out = [_array_from_process_local_data(data, s, shape) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/local/lib/python3.12/site-packages/jax/_src/array.py", line 1048, in _array_from_process_local_data return make_array_from_callback(global_shape, sharding, cb) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/local/lib/python3.12/site-packages/jax/_src/array.py", line 845, in make_array_from_callback per_device_values = api.device_put(per_device_values, devices) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/local/lib/python3.12/site-packages/jax/_src/api.py", line 2729, in device_put out_flat = dispatch._batched_device_put_impl( ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/local/lib/python3.12/site-packages/jax/_src/dispatch.py", line 558, in _batched_device_put_impl y = _device_put_impl(x, device=device, src=src, copy=cp, aval=aval) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/local/lib/python3.12/site-packages/jax/_src/dispatch.py", line 545, in _device_put_impl return _device_put_sharding_impl(x, aval, device, copy) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/local/lib/python3.12/site-packages/jax/_src/dispatch.py", line 487, in _device_put_sharding_impl raise ValueError( ValueError: device_put's first argument must be a fully addressable array, but got value with devices {TpuDevice(id=16, process_index=4, coords=(0,4,0), core_on_chip=0), TpuDevice(id=24, process_index=6, coords=(0,6,0), core_on_chip=0), TpuDevice(id=2, process_index=1, coords=(2,0,0), core_on_chip=0), TpuDevice(id=18, process_index=5, coords=(2,4,0), core_on_chip=0), TpuDevice(id=7, process_index=1, coords=(3,1,0), core_on_chip=0), TpuDevice(id=12, process_index=2, coords=(0,3,0), core_on_chip=0), TpuDevice(id=8, process_index=2, coords=(0,2,0), core_on_chip=0), TpuDevice(id=27, process_index=7, coords=(3,6,0), core_on_chip=0), TpuDevice(id=17, process_index=4, coords=(1,4,0), core_on_chip=0), TpuDevice(id=25, process_index=6, coords=(1,6,0), core_on_chip=0), TpuDevice(id=30, process_index=7, coords=(2,7,0), core_on_chip=0), TpuDevice(id=11, process_index=3, coords=(3,2,0), core_on_chip=0), TpuDevice(id=13, process_index=2, coords=(1,3,0), core_on_chip=0), TpuDevice(id=19, process_index=5, coords=(3,4,0), core_on_chip=0), TpuDevice(id=5, process_index=0, coords=(1,1,0), core_on_chip=0), TpuDevice(id=3, process_index=1, coords=(3,0,0), core_on_chip=0), TpuDevice(id=20, process_index=4, coords=(0,5,0), core_on_chip=0), TpuDevice(id=28, process_index=6, coords=(0,7,0), core_on_chip=0), TpuDevice(id=1, process_index=0, coords=(1,0,0), core_on_chip=0), TpuDevice(id=14, process_index=3, coords=(2,3,0), core_on_chip=0), TpuDevice(id=22, process_index=5, coords=(2,5,0), core_on_chip=0), TpuDevice(id=9, process_index=2, coords=(1,2,0), core_on_chip=0), TpuDevice(id=10, process_index=3, coords=(2,2,0), core_on_chip=0), TpuDevice(id=0, process_index=0, coords=(0,0,0), core_on_chip=0), TpuDevice(id=21, process_index=4, coords=(1,5,0), core_on_chip=0), TpuDevice(id=29, process_index=6, coords=(1,7,0), core_on_chip=0), TpuDevice(id=15, process_index=3, coords=(3,3,0), core_on_chip=0), TpuDevice(id=23, process_index=5, coords=(3,5,0), core_on_chip=0), TpuDevice(id=6, process_index=1, coords=(2,1,0), core_on_chip=0), TpuDevice(id=4, process_index=0, coords=(0,1,0), core_on_chip=0), TpuDevice(id=31, process_index=7, coords=(3,7,0), core_on_chip=0), TpuDevice(id=26, process_index=7, coords=(2,6,0), core_on_chip=0)} I0425 15:21:41.068649 139185566865152 grain_pool.py:542] Grain pool is exiting. I0425 15:21:41.068753 139185566865152 grain_pool.py:547] Shutting down multiprocessing system. I0425 15:21:46.987698 139185566865152 grain_pool.py:547] Shutting down multiprocessing system. Training: 0%| | 0/5 [00:24<?, ?step/s] /usr/local/lib/python3.12/multiprocessing/resource_tracker.py:279: UserWarning: resource_tracker: There appear to be 15 leaked shared_memory objects to clean up at shutdown warnings.warn('resource_tracker: There appear to be %d ' XPK End: Sat Apr 25 15:21:56 UTC 2026 EXIT_CODE=1