MaxView

XPK Start: Thu Apr 23 21:20:34 UTC 2026
Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'rope_theta'}
`rope_scaling`'s factor field must be a float >= 1, got 40
`rope_scaling`'s beta_fast field must be a float, got 32
`rope_scaling`'s beta_slow field must be a float, got 1
Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'rope_theta'}
Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'rope_theta'}
Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'rope_theta'}
2026-04-23 21:21:33.370165: 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)
I0423 21:21:33.615033 138982174828352 max_utils.py:273] Attempting to initialize the jax distributed system...
INFO:2026-04-23 21:21:42,656:jax._src.distributed:166: Connecting to JAX distributed service on mt-01-sft-smoke-9wari-slice-job-0-0.mt-01-sft-smoke-9wari:8482
I0423 21:21:42.656563 138982174828352 distributed.py:166] Connecting to JAX distributed service on mt-01-sft-smoke-9wari-slice-job-0-0.mt-01-sft-smoke-9wari:8482
I0423 21:21:58.796566 138982174828352 max_utils.py:284] Jax distributed system initialized!
I0423 21:22:04.095677 138982174828352 max_utils.py:800] System Information: Jax Version: 0.8.3
I0423 21:22:04.095790 138982174828352 max_utils.py:801] System Information: Jaxlib Version: 0.8.3
I0423 21:22:04.095831 138982174828352 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
I0423 21:22:04.099245 138982174828352 maxtext_utils.py:1565] Num_devices: 32, shape (1, 4, 1, 8, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0423 21:22:04.279144 138982174828352 maxtext_utils.py:1565] Num_devices: 32, shape (1, 4, 1, 8, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0423 21:22:05.421166 138982174828352 config.py:112] TensorFlow version 2.20.0 available.
I0423 21:22:05.421674 138982174828352 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(
E0423 21:22:10.694799 138982174828352 packing.py:209] PackAndBatchOperation is deprecated. Please use lazy_dataset.FirstFitPackIterDataset instead.
I0423 21:22:10.695133 138982174828352 data_loader.py:408] Adding CopyNumPyArrayToSharedMemory MapTransform.
I0423 21:22:11.087161 138982174828352 pytree_checkpoint_handler.py:577] save_device_host_concurrent_bytes=None
I0423 21:22:11.087625 138982174828352 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 0x7e66995a5e20>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0423 21:22:11.087678 138982174828352 pytree_checkpoint_handler.py:577] save_device_host_concurrent_bytes=None
I0423 21:22:11.087722 138982174828352 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 0x7e66995a5e20>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0423 21:22:11.087772 138982174828352 checkpoint_manager.py:702] [process=3][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=None, item_handlers={'model_params': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e669790c470>, 'optimizer_state': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e495c0a45f0>, 'custom_metadata': <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e495c0a6630>}, handler_registry=None
I0423 21:22:11.088010 138982174828352 composite_checkpoint_handler.py:237] Deferred registration for item: "model_params". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e669790c470>` 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`.
I0423 21:22:11.088057 138982174828352 composite_checkpoint_handler.py:237] Deferred registration for item: "optimizer_state". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e495c0a45f0>` 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`.
I0423 21:22:11.088085 138982174828352 composite_checkpoint_handler.py:237] Deferred registration for item: "custom_metadata". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e495c0a6630>` 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`.
I0423 21:22:11.088110 138982174828352 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e4a380c6780>` 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`.
I0423 21:22:11.088140 138982174828352 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 0x7e669790c470>, ('model_params', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e669790c470>, ('optimizer_state', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e495c0a45f0>, ('optimizer_state', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e495c0a45f0>, ('custom_metadata', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e495c0a6630>, ('custom_metadata', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e495c0a6630>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e4a380c6780>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e4a380c6780>}).
I0423 21:22:11.088338 138982174828352 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.28
I0423 21:22:11.088430 138982174828352 async_checkpointer.py:177] [process=3][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7e48fc146b60> timeout: 600 secs and primary_host=0 for async checkpoint writes
I0423 21:22:14.201516 138982174828352 checkpoint_manager.py:1788] Found 0 checkpoint steps in gs://lance-maxtext/pt_ckpt_xpk_test_pipeline_scan_nnx_20260423_211626/pt_sft_linen_xpk_test_pipeline_scan_nnx_20260423_211626_01_sft_smoke/checkpoints
I0423 21:22:14.645002 138982174828352 checkpoint_manager.py:921] [process=3][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_test_pipeline_scan_nnx_20260423_211626/pt_sft_linen_xpk_test_pipeline_scan_nnx_20260423_211626_01_sft_smoke/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7e495c0a5520>
I0423 21:22:14.645354 138982174828352 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))
I0423 21:22:15.056616 138982174828352 peft_trainer.py:600] Compiled train_step cache size: 0

Training:   0%|          | 0/5 [00:00<?, ?step/s]I0423 21:22:15.058829 138982174828352 metric_logger.py:301] number parameters: 0.000 billion
I0423 21:22:15.061190 138819164690176 grain_pool.py:367] Grain pool will use 1 processes.
I0423 21:22:15.087374 138819164690176 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
I0423 21:22:15.092790 138819164690176 grain_pool.py:448] Grain pool started all child processes.
2026-04-23 21:22:19.005084: 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`.
2026-04-23 21:22:19.049370: 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.
2026-04-23 21:22:20.027456: 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`.
Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'rope_theta'}
`rope_scaling`'s factor field must be a float >= 1, got 40
`rope_scaling`'s beta_fast field must be a float, got 32
`rope_scaling`'s beta_slow field must be a float, got 1
Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'rope_theta'}
Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'rope_theta'}
Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'rope_theta'}
2026-04-23 21:22:24.640388: 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)
I0423 21:22:30.359984 138982174828352 utils.py:86] Train loop finished in: 15.3001 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 217, 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 213, in main
    train(mt_config, goodput_recorder)
  File "/deps/src/maxtext/trainers/post_train/sft/train_sft.py", line 190, in train
    trainer = train_model(mt_config, trainer, mesh)
              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/deps/src/maxtext/trainers/post_train/sft/train_sft.py", line 176, 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=31, process_index=7, coords=(3,7,0), core_on_chip=0), TpuDevice(id=9, process_index=2, coords=(1,2,0), core_on_chip=0), TpuDevice(id=7, process_index=1, coords=(3,1,0), core_on_chip=0), TpuDevice(id=26, process_index=7, coords=(2,6,0), core_on_chip=0), 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=27, process_index=7, coords=(3,6,0), core_on_chip=0), TpuDevice(id=0, process_index=0, coords=(0,0,0), core_on_chip=0), TpuDevice(id=12, process_index=2, coords=(0,3,0), core_on_chip=0), TpuDevice(id=13, process_index=2, coords=(1,3,0), 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=11, process_index=3, coords=(3,2,0), core_on_chip=0), TpuDevice(id=19, process_index=5, coords=(3,4,0), core_on_chip=0), TpuDevice(id=3, process_index=1, coords=(3,0,0), core_on_chip=0), TpuDevice(id=1, process_index=0, coords=(1,0,0), core_on_chip=0), TpuDevice(id=5, process_index=0, coords=(1,1,0), 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=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=8, process_index=2, coords=(0,2,0), core_on_chip=0), TpuDevice(id=6, process_index=1, coords=(2,1,0), core_on_chip=0), TpuDevice(id=10, process_index=3, coords=(2,2,0), core_on_chip=0), TpuDevice(id=4, process_index=0, coords=(0,1,0), core_on_chip=0), TpuDevice(id=30, process_index=7, coords=(2,7,0), 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)}
I0423 21:22:30.702380 138819164690176 grain_pool.py:542] Grain pool is exiting.
I0423 21:22:30.702491 138819164690176 grain_pool.py:547] Shutting down multiprocessing system.
I0423 21:22:36.589716 138819164690176 grain_pool.py:547] Shutting down multiprocessing system.

Training:   0%|          | 0/5 [00:25<?, ?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: Thu Apr 23 21:22:45 UTC 2026
EXIT_CODE=1

XPK Start: Thu Apr 23 21:29:16 UTC 2026
Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'rope_theta'}
`rope_scaling`'s factor field must be a float >= 1, got 40
`rope_scaling`'s beta_fast field must be a float, got 32
`rope_scaling`'s beta_slow field must be a float, got 1
Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'rope_theta'}
Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'rope_theta'}
Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'rope_theta'}
2026-04-23 21:29:46.004889: 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)
I0423 21:29:46.251108 139484222990144 max_utils.py:273] Attempting to initialize the jax distributed system...
INFO:2026-04-23 21:29:55,291:jax._src.distributed:149: Starting JAX distributed service on [::]:8482
I0423 21:29:55.291442 139484222990144 distributed.py:149] Starting JAX distributed service on [::]:8482
INFO:2026-04-23 21:29:55,293:jax._src.distributed:166: Connecting to JAX distributed service on mt-01-sft-smoke-rl7dy-slice-job-0-0.mt-01-sft-smoke-rl7dy:8482
I0423 21:29:55.293942 139484222990144 distributed.py:166] Connecting to JAX distributed service on mt-01-sft-smoke-rl7dy-slice-job-0-0.mt-01-sft-smoke-rl7dy:8482
I0423 21:29:56.554928 139484222990144 max_utils.py:284] Jax distributed system initialized!
I0423 21:30:02.857969 139484222990144 max_utils.py:800] System Information: Jax Version: 0.8.3
I0423 21:30:02.858078 139484222990144 max_utils.py:801] System Information: Jaxlib Version: 0.8.3
I0423 21:30:02.858118 139484222990144 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
I0423 21:30:02.861550 139484222990144 maxtext_utils.py:1565] Num_devices: 32, shape (1, 4, 1, 8, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0423 21:30:02.959833 139484222990144 maxtext_utils.py:1565] Num_devices: 32, shape (1, 4, 1, 8, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0423 21:30:04.059411 139484222990144 config.py:112] TensorFlow version 2.20.0 available.
I0423 21:30:04.059947 139484222990144 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(
E0423 21:30:09.239800 139484222990144 packing.py:209] PackAndBatchOperation is deprecated. Please use lazy_dataset.FirstFitPackIterDataset instead.
I0423 21:30:09.240124 139484222990144 data_loader.py:408] Adding CopyNumPyArrayToSharedMemory MapTransform.
I0423 21:30:09.629795 139484222990144 pytree_checkpoint_handler.py:577] save_device_host_concurrent_bytes=None
I0423 21:30:09.630309 139484222990144 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 0x7edb7ddd3980>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0423 21:30:09.630357 139484222990144 pytree_checkpoint_handler.py:577] save_device_host_concurrent_bytes=None
I0423 21:30:09.630397 139484222990144 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 0x7edb7ddd3980>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0423 21:30:09.630443 139484222990144 checkpoint_manager.py:702] [process=5][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=None, item_handlers={'model_params': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ec2441ab770>, 'optimizer_state': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ec3767af200>, 'custom_metadata': <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ebdec1c18b0>}, handler_registry=None
I0423 21:30:09.630665 139484222990144 composite_checkpoint_handler.py:237] Deferred registration for item: "model_params". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ec2441ab770>` 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`.
I0423 21:30:09.630711 139484222990144 composite_checkpoint_handler.py:237] Deferred registration for item: "optimizer_state". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ec3767af200>` 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`.
I0423 21:30:09.630739 139484222990144 composite_checkpoint_handler.py:237] Deferred registration for item: "custom_metadata". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ebdec1c18b0>` 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`.
I0423 21:30:09.630764 139484222990144 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ebdec1c1790>` 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`.
I0423 21:30:09.630792 139484222990144 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 0x7ec2441ab770>, ('model_params', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ec2441ab770>, ('optimizer_state', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ec3767af200>, ('optimizer_state', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ec3767af200>, ('custom_metadata', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ebdec1c18b0>, ('custom_metadata', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ebdec1c18b0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ebdec1c1790>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ebdec1c1790>}).
I0423 21:30:09.630990 139484222990144 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.28
I0423 21:30:09.631044 139484222990144 async_checkpointer.py:177] [process=5][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7ebe56f2b560> timeout: 600 secs and primary_host=0 for async checkpoint writes
I0423 21:30:12.208957 139484222990144 checkpoint_manager.py:1788] Found 0 checkpoint steps in gs://lance-maxtext/pt_ckpt_xpk_test_pipeline_scan_nnx_20260423_211626/pt_sft_nnx_xpk_test_pipeline_scan_nnx_20260423_211626_01_sft_smoke/checkpoints
I0423 21:30:12.645482 139484222990144 checkpoint_manager.py:921] [process=5][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_test_pipeline_scan_nnx_20260423_211626/pt_sft_nnx_xpk_test_pipeline_scan_nnx_20260423_211626_01_sft_smoke/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7ebdec1c0b00>
I0423 21:30:12.645884 139484222990144 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))
I0423 21:30:13.063138 139484222990144 peft_trainer.py:600] Compiled train_step cache size: 0

Training:   0%|          | 0/5 [00:00<?, ?step/s]I0423 21:30:13.067335 139484222990144 metric_logger.py:301] number parameters: 0.000 billion
I0423 21:30:13.069706 139332177413888 grain_pool.py:367] Grain pool will use 1 processes.
I0423 21:30:13.095160 139332177413888 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
I0423 21:30:13.100572 139332177413888 grain_pool.py:448] Grain pool started all child processes.
2026-04-23 21:30:16.993552: 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`.
2026-04-23 21:30:17.038283: 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.
2026-04-23 21:30:18.023452: 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`.
Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'rope_theta'}
`rope_scaling`'s factor field must be a float >= 1, got 40
`rope_scaling`'s beta_fast field must be a float, got 32
`rope_scaling`'s beta_slow field must be a float, got 1
Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'rope_theta'}
Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'rope_theta'}
Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'rope_theta'}
2026-04-23 21:30:22.564219: 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)
I0423 21:30:28.236885 139484222990144 utils.py:86] Train loop finished in: 15.1685 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 217, 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 213, in main
    train(mt_config, goodput_recorder)
  File "/deps/src/maxtext/trainers/post_train/sft/train_sft.py", line 190, in train
    trainer = train_model(mt_config, trainer, mesh)
              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/deps/src/maxtext/trainers/post_train/sft/train_sft.py", line 176, 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=10, process_index=3, coords=(2,2,0), core_on_chip=0), TpuDevice(id=18, process_index=5, coords=(2,4,0), core_on_chip=0), TpuDevice(id=12, process_index=2, coords=(0,3,0), core_on_chip=0), TpuDevice(id=7, process_index=1, coords=(3,1,0), core_on_chip=0), TpuDevice(id=2, process_index=1, coords=(2,0,0), core_on_chip=0), TpuDevice(id=3, process_index=1, coords=(3,0,0), 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=11, process_index=3, coords=(3,2,0), core_on_chip=0), TpuDevice(id=19, process_index=5, coords=(3,4,0), core_on_chip=0), TpuDevice(id=6, process_index=1, coords=(2,1,0), core_on_chip=0), TpuDevice(id=1, process_index=0, coords=(1,0,0), core_on_chip=0), TpuDevice(id=27, process_index=7, coords=(3,6,0), core_on_chip=0), TpuDevice(id=0, process_index=0, coords=(0,0,0), core_on_chip=0), TpuDevice(id=28, process_index=6, coords=(0,7,0), core_on_chip=0), TpuDevice(id=20, process_index=4, coords=(0,5,0), 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=30, process_index=7, coords=(2,7,0), core_on_chip=0), TpuDevice(id=8, process_index=2, coords=(0,2,0), core_on_chip=0), TpuDevice(id=5, process_index=0, coords=(1,1,0), core_on_chip=0), TpuDevice(id=31, process_index=7, coords=(3,7,0), core_on_chip=0), TpuDevice(id=29, process_index=6, coords=(1,7,0), core_on_chip=0), TpuDevice(id=13, process_index=2, coords=(1,3,0), core_on_chip=0), TpuDevice(id=21, process_index=4, coords=(1,5,0), core_on_chip=0), TpuDevice(id=26, process_index=7, coords=(2,6,0), core_on_chip=0), TpuDevice(id=15, process_index=3, coords=(3,3,0), core_on_chip=0), TpuDevice(id=9, process_index=2, coords=(1,2,0), core_on_chip=0), TpuDevice(id=23, process_index=5, coords=(3,5,0), core_on_chip=0), TpuDevice(id=4, process_index=0, coords=(0,1,0), core_on_chip=0)}
I0423 21:30:28.579102 139332177413888 grain_pool.py:542] Grain pool is exiting.
I0423 21:30:28.579203 139332177413888 grain_pool.py:547] Shutting down multiprocessing system.
I0423 21:30:34.424702 139332177413888 grain_pool.py:547] Shutting down multiprocessing system.

Training:   0%|          | 0/5 [00:25<?, ?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: Thu Apr 23 21:30:43 UTC 2026
EXIT_CODE=1