MaxView

← Back to run

Log Summary

XPK Start: Fri Apr 24 16:36:19 UTC 2026
2026-04-24 16:36:23.426581: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:467] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
WARNING: All log messages before absl::InitializeLog() is called are written to STDERR
E0000 00:00:1777048583.439604      10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
E0000 00:00:1777048583.443378      10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
W0000 00:00:1777048583.455023      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1777048583.455041      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1777048583.455044      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1777048583.455046      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
2026-04-24 16:36:42.674657: 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)
I0424 16:36:43.204968 139664686720832 max_utils.py:273] Attempting to initialize the jax distributed system...
INFO:2026-04-24 16:36:52,247:jax._src.distributed:140: Starting JAX distributed service on [::]:8482
I0424 16:36:52.247682 139664686720832 distributed.py:140] Starting JAX distributed service on [::]:8482
INFO:2026-04-24 16:36:52,252:jax._src.distributed:157: Connecting to JAX distributed service on mt-03-dropout-ifw9o-slice-job-0-0.mt-03-dropout-ifw9o:8482
I0424 16:36:52.252314 139664686720832 distributed.py:157] Connecting to JAX distributed service on mt-03-dropout-ifw9o-slice-job-0-0.mt-03-dropout-ifw9o:8482
I0424 16:36:53.256539 139664686720832 max_utils.py:284] Jax distributed system initialized!
I0424 16:37:00.151793 139664686720832 max_utils.py:800] System Information: Jax Version: 0.8.1
I0424 16:37:00.151899 139664686720832 max_utils.py:801] System Information: Jaxlib Version: 0.8.1
I0424 16:37:00.151940 139664686720832 max_utils.py:802] System Information: Jax Backend: PJRT C API
TFRT TPU v6 lite
Built on Nov 12 2025 14:16:36 (1762985796) cl/831091709
I0424 16:37:00.151974 139664686720832 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0424 16:37:00.772781 139664686720832 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0424 16:37:00.887574 139664686720832 checkpointing.py:688] Setting up checkpoint logger...
I0424 16:37:00.887695 139664686720832 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0424 16:37:00.887740 139664686720832 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None
I0424 16:37:00.887942 139664686720832 base_pytree_checkpoint_handler.py:415] 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 0x7f05b3b4d1c0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0424 16:37:03.826807 139664686720832 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0424 16:37:03.827055 139664686720832 checkpoint_manager.py:709] [process=6][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ef2ad2e3140>}, handler_registry=None
I0424 16:37:03.827306 139664686720832 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ef2ad2e3140>` for item "items" 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`.
I0424 16:37:03.827355 139664686720832 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ef1d0403bf0>` 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`.
I0424 16:37:03.827392 139664686720832 composite_checkpoint_handler.py:505] Initialized registry DefaultCheckpointHandlerRegistry({('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ef2ad2e3140>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ef2ad2e3140>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ef1d0403bf0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ef1d0403bf0>}).
I0424 16:37:03.827740 139664686720832 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33
I0424 16:37:03.827818 139664686720832 async_checkpointer.py:177] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7ef2ac834f40> timeout: 600 secs and primary_host=0 for async checkpoint writes
I0424 16:37:05.299739 139664686720832 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_03_dropout/checkpoints
I0424 16:37:05.326073 139664686720832 checkpoint_manager.py:929] [process=6][thread=MainThread] CheckpointManager created,  primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, 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_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=FixedIntervalPolicy(interval=10), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_03_dropout/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7ef2ad2ea330>
I0424 16:37:05.326231 139664686720832 checkpointing.py:302] Checkpoint manager created!
I0424 16:37:05.674901 139664686720832 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0424 16:37:05.675023 139664686720832 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0424 16:37:07.094985 139664686720832 maxtext_utils.py:1749]  decoder/decoder_norm/scale/value
    Shape:     float32[2048]
    Physical:  (None,)
I0424 16:37:07.095118 139664686720832 maxtext_utils.py:1749]  decoder/layers/mlp/wi_0/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0424 16:37:07.095175 139664686720832 maxtext_utils.py:1749]  decoder/layers/mlp/wi_1/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0424 16:37:07.095221 139664686720832 maxtext_utils.py:1749]  decoder/layers/mlp/wo/kernel/value
    Shape:     float32[7168,16,2048]
    Physical:  (None, None, 'fsdp')
I0424 16:37:07.095261 139664686720832 maxtext_utils.py:1749]  decoder/layers/post_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0424 16:37:07.095297 139664686720832 maxtext_utils.py:1749]  decoder/layers/pre_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0424 16:37:07.095336 139664686720832 maxtext_utils.py:1749]  decoder/layers/self_attention/key/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0424 16:37:07.095374 139664686720832 maxtext_utils.py:1749]  decoder/layers/self_attention/out/kernel/value
    Shape:     float32[16,16,128,2048]
    Physical:  (None, None, None, 'fsdp')
I0424 16:37:07.095413 139664686720832 maxtext_utils.py:1749]  decoder/layers/self_attention/query/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0424 16:37:07.095448 139664686720832 maxtext_utils.py:1749]  decoder/layers/self_attention/value/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0424 16:37:07.095483 139664686720832 maxtext_utils.py:1749]  decoder/logits_dense/kernel/value
    Shape:     float32[2048,32000]
    Physical:  ('fsdp', None)

I0424 16:37:07.095517 139664686720832 maxtext_utils.py:1749]  token_embedder/embedding/value
    Shape:     float32[32000,2048]
    Physical:  (None, 'fsdp')
I0424 16:37:07.503535 139664686720832 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0424 16:37:07.503688 139664686720832 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0424 16:37:07.509136 139664686720832 nnx_decoders.py:616] Unknown Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0424 16:37:07.509234 139664686720832 nnx_decoders.py:616] Unknown Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0424 16:37:07.663398 139664686720832 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0424 16:37:07.663534 139664686720832 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0424 16:37:07.819953 139664686720832 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0424 16:37:07.820127 139664686720832 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0424 16:37:07.983562 139664686720832 attentions.py:1150] attentions/query Logical: bfloat16[32,128,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0424 16:37:07.983701 139664686720832 attentions.py:1150] attentions/query Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0424 16:37:08.136792 139664686720832 attentions.py:1151] attentions/key Logical: bfloat16[32,128,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0424 16:37:08.136928 139664686720832 attentions.py:1151] attentions/key Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0424 16:37:08.286861 139664686720832 attentions.py:1152] attentions/value Logical: bfloat16[32,128,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0424 16:37:08.286996 139664686720832 attentions.py:1152] attentions/value Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0424 16:37:08.447614 139664686720832 attentions.py:1193] attentions/out Logical: bfloat16[32,128,16,128]..................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0424 16:37:08.447755 139664686720832 attentions.py:1193] attentions/out Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0424 16:37:08.476769 139664686720832 linears.py:541] linears/x Logical: bfloat16[32,128,7168]....................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0424 16:37:08.476873 139664686720832 linears.py:541] linears/x Physical: bfloat16[32,128,7168]....................................... ('fsdp', None, None).
I0424 16:37:11.838075 139664686720832 max_utils.py:791] Total memory size: 0.8 GB, Output size: 0.4 GB, Temp size: 0.4 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB.
I0424 16:37:11.840516 139664686720832 metric_logger.py:289] number parameters: 1.104 billion
I0424 16:37:15.588274 139664686720832 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0424 16:37:15.893225 139664686720832 checkpointing.py:798] Waited 0.3049287796020508 seconds for step 0 to finish before starting checkpointing.
I0424 16:37:15.895544 139664686720832 checkpoint_manager.py:2013] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0424 16:37:15.897532 139664686720832 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 0
I0424 16:37:15.900392 139664686720832 async_checkpointer.py:452] [process=6] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_03_dropout/checkpoints/0.
I0424 16:37:16.455347 139664686720832 signaling_client.py:364] Using JaxDistributedSignalingClient
I0424 16:37:16.456343 139664686720832 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0424 16:37:16.456404 139664686720832 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False
I0424 16:37:16.706315 139664686720832 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.252098s
I0424 16:37:16.706492 139664686720832 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.995 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5150575637817383 s) (per-host)
I0424 16:37:16.706541 139664686720832 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.515116s (batch_requests_ready=0.256425s, total_serialization_initiated=0.258625s, others=0.000067s)
I0424 16:37:16.706634 139664686720832 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.519342s (all_items=0.000019s, per_item={'items': '0.00001907'}, temp_paths=0.519323)
I0424 16:37:16.707522 139533782181632 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started.
I0424 16:37:16.707687 139664686720832 async_checkpointer.py:561] Finished blocking save. Time taken: 0.810080s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_03_dropout/checkpoints/0.
I0424 16:37:16.720417 139664686720832 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0424 16:37:16.720705 139538110342912 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0424 16:37:16.720861 139664686720832 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_03_dropout/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777048635.8955262, 'wait_for_prev_duration_secs': 6.437301635742188e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777048635.8975742, 'checkpointer_blocking_duration_secs': 0.8102250099182129, 'get_old_steps_start_time': 1777048636.7078214, 'get_old_steps_duration_secs': 2.5272369384765625e-05, 'checkpoint_manager_blocking_start_time': 1777048635.8936763, 'checkpoint_manager_blocking_duration_secs': 0.8271470069885254}
I0424 16:37:16.720968 139664686720832 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0424 16:37:16.721040 139664686720832 max_utils.py:750] 
Memstats: After params initialized:
I0424 16:37:16.721089 139664686720832 max_utils.py:756] 	Using (GB) 0.41 / 31.25 (1.312000%) on TPU_24(process=6,(0,6,0,0))
I0424 16:37:16.721136 139664686720832 max_utils.py:756] 	Using (GB) 0.41 / 31.25 (1.312000%) on TPU_25(process=6,(1,6,0,0))
I0424 16:37:16.721163 139664686720832 max_utils.py:756] 	Using (GB) 0.41 / 31.25 (1.312000%) on TPU_28(process=6,(0,7,0,0))
I0424 16:37:16.721188 139664686720832 max_utils.py:756] 	Using (GB) 0.41 / 31.25 (1.312000%) on TPU_29(process=6,(1,7,0,0))
I0424 16:37:16.991391 139664686720832 metric_logger.py:185] completed step: 0, seconds: 3.747, TFLOP/s/device: 0.214, Tokens/s/device: 34.163, total_weights: 4096, loss: 10.874
I0424 16:37:17.098355 139664686720832 metric_logger.py:185] completed step: 1, seconds: 1.401, TFLOP/s/device: 0.572, Tokens/s/device: 91.342, total_weights: 4096, loss: 10.882
I0424 16:37:17.542707 139664686720832 metric_logger.py:185] completed step: 2, seconds: 0.039, TFLOP/s/device: 20.759, Tokens/s/device: 3317.867, total_weights: 4096, loss: 8.922
I0424 16:37:17.612741 139664686720832 metric_logger.py:185] completed step: 3, seconds: 0.434, TFLOP/s/device: 1.847, Tokens/s/device: 295.127, total_weights: 4096, loss: 7.313
I0424 16:37:17.754041 139664686720832 metric_logger.py:185] completed step: 4, seconds: 0.087, TFLOP/s/device: 9.210, Tokens/s/device: 1472.043, total_weights: 4096, loss: 6.002
I0424 16:37:17.761257 139664686720832 metric_logger.py:185] completed step: 5, seconds: 0.070, TFLOP/s/device: 11.508, Tokens/s/device: 1839.265, total_weights: 4096, loss: 4.934
I0424 16:37:26.851191 139664686720832 metric_logger.py:185] completed step: 6, seconds: 0.142, TFLOP/s/device: 5.646, Tokens/s/device: 902.330, total_weights: 4096, loss: 4.196
I0424 16:37:26.921369 139664686720832 metric_logger.py:185] completed step: 7, seconds: 9.021, TFLOP/s/device: 0.089, Tokens/s/device: 14.189, total_weights: 4096, loss: 3.620
I0424 16:37:26.991564 139664686720832 metric_logger.py:185] completed step: 8, seconds: 0.076, TFLOP/s/device: 10.523, Tokens/s/device: 1681.821, total_weights: 4096, loss: 3.238
I0424 16:37:27.062789 139664686720832 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0424 16:37:27.063388 139664686720832 checkpointing.py:798] Waited 0.0006105899810791016 seconds for step 9 to finish before starting checkpointing.
I0424 16:37:27.066006 139664686720832 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0424 16:37:28.440116    2685 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0424 16:37:30.370346 139535376770816 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_03_dropout/checkpoints/0/items/array_metadatas/process_6
I0424 16:37:57.650078 139533782181632 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 38.101 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.45860266685486 s) (per-host)
I0424 16:37:57.650207 139533782181632 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 40.942572s.
I0424 16:38:06.463816 139533782181632 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 49.756165s.
I0424 16:38:06.464129 139538110342912 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0424 16:38:06.464248 139538110342912 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0424 16:38:06.464315 139538110342912 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0424 16:38:06.465758 139538110342912 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0424 16:38:06.465890 139664686720832 checkpoint_manager.py:2036] [process=6][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0424 16:38:06.465957 139664686720832 checkpoint_manager.py:1458] Waiting for previous save to complete took 39.399972 seconds. If this number is high, consider checkpointing less frequently.
I0424 16:38:06.468001 139664686720832 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 9
I0424 16:38:06.471405 139664686720832 async_checkpointer.py:452] [process=6] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_03_dropout/checkpoints/9.
I0424 16:38:07.047706 139664686720832 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0424 16:38:07.047861 139664686720832 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False
I0424 16:38:07.086594 139664686720832 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.040934s
I0424 16:38:07.086761 139664686720832 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 4.931 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.3128080368041992 s) (per-host)
I0424 16:38:07.086812 139664686720832 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.312869s (batch_requests_ready=0.267189s, total_serialization_initiated=0.045610s, others=0.000070s)
I0424 16:38:07.086904 139664686720832 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.316915s (all_items=0.000015s, per_item={'items': '0.00001502'}, temp_paths=0.316900)
I0424 16:38:07.087799 139539729348352 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started.
I0424 16:38:07.087965 139664686720832 async_checkpointer.py:561] Finished blocking save. Time taken: 0.619896s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_03_dropout/checkpoints/9.
I0424 16:38:07.340162 139664686720832 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0424 16:38:07.340523 139538110342912 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0424 16:38:07.340700 139664686720832 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_03_dropout/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777048647.0659642, 'wait_for_prev_duration_secs': 39.39997220039368, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777048686.4680386, 'checkpointer_blocking_duration_secs': 0.6200354099273682, 'get_old_steps_start_time': 1777048687.0881064, 'get_old_steps_duration_secs': 2.574920654296875e-05, 'checkpoint_manager_blocking_start_time': 1777048647.0636184, 'checkpoint_manager_blocking_duration_secs': 40.27704453468323}
I0424 16:38:07.340812 139664686720832 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0424 16:38:07.340856 139664686720832 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0424 16:38:12.185971 139512529274624 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_03_dropout/checkpoints/9/items/array_metadatas/process_6
I0424 16:38:48.494847 139539729348352 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 37.861 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.72085452079773 s) (per-host)
I0424 16:38:48.494972 139539729348352 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.407064s.
I0424 16:38:54.348323 139539729348352 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 47.260399s.
I0424 16:38:54.348585 139538110342912 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0424 16:38:54.348641 139538110342912 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0424 16:38:54.348702 139538110342912 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0424 16:38:54.351207 139538110342912 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0424 16:38:54.351467 139664686720832 checkpoint_manager.py:2036] [process=6][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0424 16:38:54.351634 139664686720832 checkpoint_manager.py:2013] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0424 16:38:54.352327 139664686720832 metric_logger.py:185] completed step: 9, seconds: 0.070, TFLOP/s/device: 11.493, Tokens/s/device: 1836.890, total_weights: 4096, loss: 3.085
Per train step:
 Total TFLOPs: 0.80 
 split as 99.60% learnable weight flops and 0.40% attention flops
XPK End: Fri Apr 24 16:39:06 UTC 2026
EXIT_CODE=0