MaxView

← Back to run

Log Summary

XPK Start: Wed Apr 22 18:12:55 UTC 2026
2026-04-22 18:12:59.454368: 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 00:00:1776881579.467170      10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
E0000 00:00:1776881579.471127      10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
W0000 00:00:1776881579.482648      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776881579.482667      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776881579.482670      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776881579.482672      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
2026-04-22 18:13:18.675039: 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)
I0422 18:13:19.204091 132084648474432 max_utils.py:273] Attempting to initialize the jax distributed system...
INFO:2026-04-22 18:13:28,246:jax._src.distributed:140: Starting JAX distributed service on [::]:8482
I0422 18:13:28.246492 132084648474432 distributed.py:140] Starting JAX distributed service on [::]:8482
INFO:2026-04-22 18:13:28,248:jax._src.distributed:157: Connecting to JAX distributed service on mt-13-scan-layers-true-5qugm-slice-job-0-0.mt-13-scan-layers-true-5qugm:8482
I0422 18:13:28.248842 132084648474432 distributed.py:157] Connecting to JAX distributed service on mt-13-scan-layers-true-5qugm-slice-job-0-0.mt-13-scan-layers-true-5qugm:8482
I0422 18:13:29.252249 132084648474432 max_utils.py:284] Jax distributed system initialized!
I0422 18:13:36.711360 132084648474432 max_utils.py:800] System Information: Jax Version: 0.8.1
I0422 18:13:36.711464 132084648474432 max_utils.py:801] System Information: Jaxlib Version: 0.8.1
I0422 18:13:36.711507 132084648474432 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
I0422 18:13:36.711540 132084648474432 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0422 18:13:37.403130 132084648474432 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0422 18:13:37.520112 132084648474432 checkpointing.py:688] Setting up checkpoint logger...
I0422 18:13:37.520231 132084648474432 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0422 18:13:37.520275 132084648474432 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None
I0422 18:13:37.520539 132084648474432 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 0x7820dbb79160>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0422 18:13:40.475188 132084648474432 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0422 18:13:40.475623 132084648474432 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 0x780bc029f7d0>}, handler_registry=None
I0422 18:13:40.475912 132084648474432 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x780bc029f7d0>` 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`.
I0422 18:13:40.475971 132084648474432 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x780b9c7e7680>` 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`.
I0422 18:13:40.476011 132084648474432 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 0x780bc029f7d0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x780bc029f7d0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x780b9c7e7680>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x780b9c7e7680>}).
I0422 18:13:40.476426 132084648474432 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33
I0422 18:13:40.476542 132084648474432 async_checkpointer.py:177] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x780b9c708f40> timeout: 600 secs and primary_host=0 for async checkpoint writes
I0422 18:13:43.352834 132084648474432 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/nnx_xpk_feat_nnx_set_defaults_true_20260422_154647_13_scan_layers_true/checkpoints
I0422 18:13:43.355343 132084648474432 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_20260422_154647/nnx_xpk_feat_nnx_set_defaults_true_20260422_154647_13_scan_layers_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x780b9c7605f0>
I0422 18:13:43.355458 132084648474432 checkpointing.py:302] Checkpoint manager created!
I0422 18:13:44.300895 132084648474432 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0422 18:13:44.301020 132084648474432 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0422 18:13:45.731702 132084648474432 maxtext_utils.py:1749]  decoder/decoder_norm/scale/value
    Shape:     float32[2048]
    Physical:  (None,)
I0422 18:13:45.731822 132084648474432 maxtext_utils.py:1749]  decoder/layers/mlp/wi_0/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0422 18:13:45.731875 132084648474432 maxtext_utils.py:1749]  decoder/layers/mlp/wi_1/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0422 18:13:45.731920 132084648474432 maxtext_utils.py:1749]  decoder/layers/mlp/wo/kernel/value
    Shape:     float32[7168,16,2048]
    Physical:  (None, None, 'fsdp')
I0422 18:13:45.731956 132084648474432 maxtext_utils.py:1749]  decoder/layers/post_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0422 18:13:45.731990 132084648474432 maxtext_utils.py:1749]  decoder/layers/pre_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0422 18:13:45.732026 132084648474432 maxtext_utils.py:1749]  decoder/layers/self_attention/key/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0422 18:13:45.732064 132084648474432 maxtext_utils.py:1749]  decoder/layers/self_attention/out/kernel/value
    Shape:     float32[16,16,128,2048]
    Physical:  (None, None, None, 'fsdp')
I0422 18:13:45.732111 132084648474432 maxtext_utils.py:1749]  decoder/layers/self_attention/query/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0422 18:13:45.732148 132084648474432 maxtext_utils.py:1749]  decoder/layers/self_attention/value/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0422 18:13:45.732183 132084648474432 maxtext_utils.py:1749]  decoder/logits_dense/kernel/value
    Shape:     float32[2048,32000]
    Physical:  ('fsdp', None)
I0422 18:13:45.732217 132084648474432 maxtext_utils.py:1749]  token_embedder/embedding/value
    Shape:     float32[32000,2048]
    Physical:  (None, 'fsdp')

I0422 18:13:46.124822 132084648474432 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0422 18:13:46.124973 132084648474432 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0422 18:13:46.131142 132084648474432 nnx_decoders.py:616] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0422 18:13:46.131290 132084648474432 nnx_decoders.py:616] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0422 18:13:46.289638 132084648474432 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0422 18:13:46.289772 132084648474432 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0422 18:13:46.440744 132084648474432 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0422 18:13:46.440877 132084648474432 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0422 18:13:46.609416 132084648474432 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0422 18:13:46.609561 132084648474432 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 18:13:46.757957 132084648474432 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0422 18:13:46.758089 132084648474432 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 18:13:46.907284 132084648474432 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0422 18:13:46.907423 132084648474432 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 18:13:47.071650 132084648474432 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0422 18:13:47.071790 132084648474432 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 18:13:47.095723 132084648474432 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0422 18:13:47.095823 132084648474432 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0422 18:13:58.254791 132084648474432 max_utils.py:791] Total memory size: 1.5 GB, Output size: 0.4 GB, Temp size: 1.1 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB.
I0422 18:13:58.257240 132084648474432 metric_logger.py:289] number parameters: 1.104 billion
I0422 18:14:09.395240 132084648474432 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0422 18:14:09.545535 132084648474432 checkpointing.py:798] Waited 0.15027356147766113 seconds for step 0 to finish before starting checkpointing.
I0422 18:14:09.547758 132084648474432 checkpoint_manager.py:2013] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0422 18:14:09.549558 132084648474432 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 0
I0422 18:14:09.552208 132084648474432 async_checkpointer.py:452] [process=6] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/nnx_xpk_feat_nnx_set_defaults_true_20260422_154647_13_scan_layers_true/checkpoints/0.
I0422 18:14:10.467795 132084648474432 signaling_client.py:364] Using JaxDistributedSignalingClient
I0422 18:14:10.468718 132084648474432 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0422 18:14:10.468774 132084648474432 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
I0422 18:14:10.717438 132084648474432 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.250824s
I0422 18:14:10.717619 132084648474432 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.907 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5307011604309082 s) (per-host)
I0422 18:14:10.717673 132084648474432 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.530765s (batch_requests_ready=0.273293s, total_serialization_initiated=0.257399s, others=0.000072s)
I0422 18:14:10.717768 132084648474432 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.534877s (all_items=0.000019s, per_item={'items': '0.00001907'}, temp_paths=0.534858)
I0422 18:14:10.718781 131961897072384 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started.
I0422 18:14:10.718947 132084648474432 async_checkpointer.py:561] Finished blocking save. Time taken: 1.169316s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/nnx_xpk_feat_nnx_set_defaults_true_20260422_154647_13_scan_layers_true/checkpoints/0.
I0422 18:14:11.157478 132084648474432 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0422 18:14:11.157869 131957090412288 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0422 18:14:11.158033 132084648474432 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/nnx_xpk_feat_nnx_set_defaults_true_20260422_154647_13_scan_layers_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776881649.54774, 'wait_for_prev_duration_secs': 6.4849853515625e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776881649.5495977, 'checkpointer_blocking_duration_secs': 1.169456958770752, 'get_old_steps_start_time': 1776881650.7190747, 'get_old_steps_duration_secs': 3.933906555175781e-05, 'checkpoint_manager_blocking_start_time': 1776881649.5459206, 'checkpoint_manager_blocking_duration_secs': 1.6120707988739014}
I0422 18:14:11.158158 132084648474432 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0422 18:14:11.158238 132084648474432 max_utils.py:750] 
Memstats: After params initialized:
I0422 18:14:11.158294 132084648474432 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_24(process=6,(0,6,0,0))
I0422 18:14:11.158328 132084648474432 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_25(process=6,(1,6,0,0))
I0422 18:14:11.158355 132084648474432 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_28(process=6,(0,7,0,0))
I0422 18:14:11.158378 132084648474432 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_29(process=6,(1,7,0,0))
I0422 18:14:11.419741 132084648474432 metric_logger.py:185] completed step: 0, seconds: 11.137, TFLOP/s/device: 1.220, Tokens/s/device: 183.892, total_weights: 65536, loss: 10.874
I0422 18:14:11.598945 132084648474432 metric_logger.py:185] completed step: 1, seconds: 2.023, TFLOP/s/device: 6.717, Tokens/s/device: 1012.475, total_weights: 65536, loss: 10.874
I0422 18:14:12.011832 132084648474432 metric_logger.py:185] completed step: 2, seconds: 0.034, TFLOP/s/device: 398.625, Tokens/s/device: 60085.081, total_weights: 65536, loss: 10.262
I0422 18:14:12.159194 132084648474432 metric_logger.py:185] completed step: 3, seconds: 0.403, TFLOP/s/device: 33.740, Tokens/s/device: 5085.735, total_weights: 65536, loss: 9.732
I0422 18:14:12.454730 132084648474432 metric_logger.py:185] completed step: 4, seconds: 0.163, TFLOP/s/device: 83.182, Tokens/s/device: 12538.033, total_weights: 65536, loss: 9.274
I0422 18:14:12.461895 132084648474432 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.399, Tokens/s/device: 13927.425, total_weights: 65536, loss: 8.888
I0422 18:14:31.359380 132084648474432 metric_logger.py:185] completed step: 6, seconds: 0.296, TFLOP/s/device: 45.941, Tokens/s/device: 6924.721, total_weights: 65536, loss: 8.589
I0422 18:14:31.506551 132084648474432 metric_logger.py:185] completed step: 7, seconds: 18.751, TFLOP/s/device: 0.725, Tokens/s/device: 109.219, total_weights: 65536, loss: 8.380
I0422 18:14:31.653967 132084648474432 metric_logger.py:185] completed step: 8, seconds: 0.153, TFLOP/s/device: 88.576, Tokens/s/device: 13351.065, total_weights: 65536, loss: 8.251
I0422 18:14:31.802441 132084648474432 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0422 18:14:31.803051 132084648474432 checkpointing.py:798] Waited 0.0006237030029296875 seconds for step 9 to finish before starting checkpointing.
I0422 18:14:31.804953 132084648474432 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0422 18:14:32.843562    2639 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0422 18:14:34.884340 131961905465088 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_20260422_154647/nnx_xpk_feat_nnx_set_defaults_true_20260422_154647_13_scan_layers_true/checkpoints/0/items/array_metadatas/process_6
I0422 18:15:06.709305 131961897072384 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 27.946 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 56.5223445892334 s) (per-host)
I0422 18:15:06.709428 131961897072384 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 55.990540s.
I0422 18:15:15.069685 131961897072384 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 64.350780s.
I0422 18:15:15.069980 131957090412288 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0422 18:15:15.070133 131957090412288 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0422 18:15:15.070187 131957090412288 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0422 18:15:15.071461 131957090412288 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0422 18:15:15.071635 132084648474432 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.
W0422 18:15:15.071775 132084648474432 checkpoint_manager.py:1458] Waiting for previous save to complete took 43.266828 seconds. If this number is high, consider checkpointing less frequently.
I0422 18:15:15.073567 132084648474432 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 9
I0422 18:15:15.076977 132084648474432 async_checkpointer.py:452] [process=6] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/nnx_xpk_feat_nnx_set_defaults_true_20260422_154647_13_scan_layers_true/checkpoints/9.
I0422 18:15:16.593958 132084648474432 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0422 18:15:16.594128 132084648474432 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
I0422 18:15:16.643857 132084648474432 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.052104s
I0422 18:15:16.644029 132084648474432 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 1.937 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.7964975833892822 s) (per-host)
I0422 18:15:16.644081 132084648474432 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.796559s (batch_requests_ready=0.739721s, total_serialization_initiated=0.056768s, others=0.000070s)
I0422 18:15:16.644193 132084648474432 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.800580s (all_items=0.000016s, per_item={'items': '0.00001574'}, temp_paths=0.800564)
I0422 18:15:16.645117 131958659061504 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started.
I0422 18:15:16.645284 132084648474432 async_checkpointer.py:561] Finished blocking save. Time taken: 1.571639s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/nnx_xpk_feat_nnx_set_defaults_true_20260422_154647_13_scan_layers_true/checkpoints/9.
I0422 18:15:16.647325 132084648474432 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0422 18:15:16.647626 131957090412288 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0422 18:15:16.647787 132084648474432 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/nnx_xpk_feat_nnx_set_defaults_true_20260422_154647_13_scan_layers_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776881671.8049228, 'wait_for_prev_duration_secs': 43.26682758331299, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776881715.0736063, 'checkpointer_blocking_duration_secs': 1.5717887878417969, 'get_old_steps_start_time': 1776881716.6454153, 'get_old_steps_duration_secs': 2.6464462280273438e-05, 'checkpoint_manager_blocking_start_time': 1776881671.8032897, 'checkpoint_manager_blocking_duration_secs': 44.844462871551514}
I0422 18:15:16.647898 132084648474432 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0422 18:15:16.647945 132084648474432 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0422 18:15:21.676974 131961905465088 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_20260422_154647/nnx_xpk_feat_nnx_set_defaults_true_20260422_154647_13_scan_layers_true/checkpoints/9/items/array_metadatas/process_6
I0422 18:15:59.094681 131958659061504 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 36.525 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 43.247108936309814 s) (per-host)
I0422 18:15:59.094800 131958659061504 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.449571s.
I0422 18:16:05.723489 131958659061504 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 49.078243s.
I0422 18:16:05.723793 131957090412288 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0422 18:16:05.723913 131957090412288 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0422 18:16:05.723961 131957090412288 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0422 18:16:05.726329 131957090412288 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0422 18:16:05.726511 132084648474432 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.
I0422 18:16:05.726686 132084648474432 checkpoint_manager.py:2013] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0422 18:16:05.727394 132084648474432 metric_logger.py:185] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.560, Tokens/s/device: 13951.619, total_weights: 65536, loss: 8.175
Per train step:
 Total TFLOPs: 13.59 
 split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Wed Apr 22 18:16:15 UTC 2026
EXIT_CODE=0