MaxView

← Back to run

Log Summary

XPK Start: Mon Apr 20 22:10:08 UTC 2026
2026-04-20 22:10:12.845138: 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 00:00:1776723012.857706      10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
E0000 00:00:1776723012.861420      10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
W0000 00:00:1776723012.872784      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776723012.872804      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776723012.872807      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776723012.872809      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
2026-04-20 22:10:32.019605: 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)
I0420 22:10:32.545640 135340420060992 max_utils.py:273] Attempting to initialize the jax distributed system...
INFO:2026-04-20 22:10:41,586:jax._src.distributed:140: Starting JAX distributed service on [::]:8482
I0420 22:10:41.586390 135340420060992 distributed.py:140] Starting JAX distributed service on [::]:8482
INFO:2026-04-20 22:10:41,588:jax._src.distributed:157: Connecting to JAX distributed service on mt-16-zarr3-true-sv-ccub4-slice-job-0-0.mt-16-zarr3-true-sv-ccub4:8482
I0420 22:10:41.588954 135340420060992 distributed.py:157] Connecting to JAX distributed service on mt-16-zarr3-true-sv-ccub4-slice-job-0-0.mt-16-zarr3-true-sv-ccub4:8482
I0420 22:10:42.699030 135340420060992 max_utils.py:284] Jax distributed system initialized!
I0420 22:10:49.579529 135340420060992 max_utils.py:800] System Information: Jax Version: 0.8.1
I0420 22:10:49.579632 135340420060992 max_utils.py:801] System Information: Jaxlib Version: 0.8.1
I0420 22:10:49.579672 135340420060992 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
I0420 22:10:49.579706 135340420060992 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0420 22:10:50.197025 135340420060992 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0420 22:10:50.312650 135340420060992 checkpointing.py:688] Setting up checkpoint logger...
I0420 22:10:50.312764 135340420060992 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0420 22:10:50.312811 135340420060992 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None
I0420 22:10:50.313018 135340420060992 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 0x7b16e6d78ec0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0420 22:10:53.256582 135340420060992 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0420 22:10:53.256815 135340420060992 checkpoint_manager.py:709] [process=1][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b039b02c6b0>}, handler_registry=None
I0420 22:10:53.257053 135340420060992 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b039b02c6b0>` 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`.
I0420 22:10:53.257106 135340420060992 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b039b02e000>` 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`.
I0420 22:10:53.257139 135340420060992 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 0x7b039b02c6b0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b039b02c6b0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b039b02e000>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b039b02e000>}).
I0420 22:10:53.257501 135340420060992 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33
I0420 22:10:53.257579 135340420060992 async_checkpointer.py:177] [process=1][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7b024c740f40> timeout: 600 secs and primary_host=0 for async checkpoint writes
I0420 22:10:54.396365 135340420060992 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_true/checkpoints
I0420 22:10:54.415723 135340420060992 checkpoint_manager.py:929] [process=1][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=5), 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_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7b024c7efb30>
I0420 22:10:54.415848 135340420060992 checkpointing.py:302] Checkpoint manager created!
I0420 22:10:55.341749 135340420060992 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0420 22:10:55.341867 135340420060992 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0420 22:10:56.793547 135340420060992 maxtext_utils.py:1749]  decoder/decoder_norm/scale/value
    Shape:     float32[2048]
    Physical:  (None,)
I0420 22:10:56.793667 135340420060992 maxtext_utils.py:1749]  decoder/layers/mlp/wi_0/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0420 22:10:56.793722 135340420060992 maxtext_utils.py:1749]  decoder/layers/mlp/wi_1/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0420 22:10:56.793767 135340420060992 maxtext_utils.py:1749]  decoder/layers/mlp/wo/kernel/value
    Shape:     float32[7168,16,2048]
    Physical:  (None, None, 'fsdp')
I0420 22:10:56.793806 135340420060992 maxtext_utils.py:1749]  decoder/layers/post_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0420 22:10:56.793842 135340420060992 maxtext_utils.py:1749]  decoder/layers/pre_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0420 22:10:56.793885 135340420060992 maxtext_utils.py:1749]  decoder/layers/self_attention/key/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0420 22:10:56.793922 135340420060992 maxtext_utils.py:1749]  decoder/layers/self_attention/out/kernel/value
    Shape:     float32[16,16,128,2048]
    Physical:  (None, None, None, 'fsdp')
I0420 22:10:56.793958 135340420060992 maxtext_utils.py:1749]  decoder/layers/self_attention/query/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0420 22:10:56.793992 135340420060992 maxtext_utils.py:1749]  decoder/layers/self_attention/value/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0420 22:10:56.794024 135340420060992 maxtext_utils.py:1749]  decoder/logits_dense/kernel/value
    Shape:     float32[2048,32000]
    Physical:  ('fsdp', None)

I0420 22:10:56.794057 135340420060992 maxtext_utils.py:1749]  token_embedder/embedding/value
    Shape:     float32[32000,2048]
    Physical:  (None, 'fsdp')
I0420 22:10:57.189992 135340420060992 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0420 22:10:57.190149 135340420060992 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 22:10:57.195601 135340420060992 nnx_decoders.py:616] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0420 22:10:57.195696 135340420060992 nnx_decoders.py:616] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 22:10:57.356234 135340420060992 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0420 22:10:57.356369 135340420060992 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 22:10:57.513023 135340420060992 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0420 22:10:57.513157 135340420060992 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 22:10:57.677573 135340420060992 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 22:10:57.677711 135340420060992 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 22:10:57.827542 135340420060992 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 22:10:57.827676 135340420060992 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 22:10:57.979440 135340420060992 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 22:10:57.979573 135340420060992 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 22:10:58.140607 135340420060992 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0420 22:10:58.140741 135340420060992 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 22:10:58.165602 135340420060992 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0420 22:10:58.165701 135340420060992 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0420 22:11:09.047233 135340420060992 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.
I0420 22:11:09.049743 135340420060992 metric_logger.py:289] number parameters: 1.104 billion
I0420 22:11:20.298771 135340420060992 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0420 22:11:20.449380 135340420060992 checkpointing.py:798] Waited 0.15059494972229004 seconds for step 0 to finish before starting checkpointing.
I0420 22:11:20.451874 135340420060992 checkpoint_manager.py:2013] [process=1][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0420 22:11:20.453628 135340420060992 checkpoint_manager.py:1518] [process=1] Saving checkpoint at step 0
I0420 22:11:20.456457 135340420060992 async_checkpointer.py:452] [process=1] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_true/checkpoints/0.
I0420 22:11:21.152634 135340420060992 signaling_client.py:364] Using JaxDistributedSignalingClient
I0420 22:11:21.153945 135340420060992 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0420 22:11:21.154059 135340420060992 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
I0420 22:11:21.398099 135340420060992 base_pytree_checkpoint_handler.py:153] [process=1][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.246616s
I0420 22:11:21.398264 135340420060992 base_pytree_checkpoint_handler.py:129] [process=1] /jax/checkpoint/write/blocking_gbytes_per_sec: 3.072 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5021657943725586 s) (per-host)
I0420 22:11:21.398315 135340420060992 base_pytree_checkpoint_handler.py:737] [process=1][thread=MainThread] Initiated Pytree async_save. Time taken: 0.502226s (batch_requests_ready=0.249001s, total_serialization_initiated=0.253158s, others=0.000067s)
I0420 22:11:21.398422 135340420060992 composite_checkpoint_handler.py:715] [process=1][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.506310s (all_items=0.000019s, per_item={'items': '0.00001907'}, temp_paths=0.506291)
I0420 22:11:21.399317 135214512707328 async_checkpointer.py:79] [process=1][thread=async_save] Background save thread started.
I0420 22:11:21.399437 135340420060992 async_checkpointer.py:561] Finished blocking save. Time taken: 0.945743s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_true/checkpoints/0.
I0420 22:11:21.886527 135340420060992 checkpoint_manager.py:1566] [process=1][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0420 22:11:21.886911 135212927272704 async_checkpointer.py:265] [process=1][thread=save_finalize] Waiting for background save thread=async_save.
I0420 22:11:21.887088 135340420060992 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776723080.4518547, 'wait_for_prev_duration_secs': 6.461143493652344e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776723080.4536657, 'checkpointer_blocking_duration_secs': 0.9458770751953125, 'get_old_steps_start_time': 1776723081.3995645, 'get_old_steps_duration_secs': 2.384185791015625e-05, 'checkpoint_manager_blocking_start_time': 1776723080.4498115, 'checkpoint_manager_blocking_duration_secs': 1.437227487564087}
I0420 22:11:21.887232 135340420060992 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0420 22:11:21.887326 135340420060992 max_utils.py:750] 
Memstats: After params initialized:
I0420 22:11:21.887392 135340420060992 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_2(process=1,(2,0,0,0))
I0420 22:11:21.887446 135340420060992 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_3(process=1,(3,0,0,0))
I0420 22:11:21.887475 135340420060992 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_6(process=1,(2,1,0,0))
I0420 22:11:21.887504 135340420060992 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_7(process=1,(3,1,0,0))
I0420 22:11:22.153768 135340420060992 metric_logger.py:185] completed step: 0, seconds: 11.248, TFLOP/s/device: 1.208, Tokens/s/device: 182.077, total_weights: 65536, loss: 10.874
I0420 22:11:22.322222 135340420060992 metric_logger.py:185] completed step: 1, seconds: 1.853, TFLOP/s/device: 7.332, Tokens/s/device: 1105.176, total_weights: 65536, loss: 10.874
I0420 22:11:22.825536 135340420060992 metric_logger.py:185] completed step: 2, seconds: 0.023, TFLOP/s/device: 596.686, Tokens/s/device: 89938.957, total_weights: 65536, loss: 10.560
I0420 22:11:22.973003 135340420060992 metric_logger.py:185] completed step: 3, seconds: 0.495, TFLOP/s/device: 27.459, Tokens/s/device: 4138.921, total_weights: 65536, loss: 9.979
I0420 22:11:22.978784 135340420060992 checkpointing.py:794] Waiting for step 5 to finish before checkpoint...
I0420 22:11:23.268052 135340420060992 checkpointing.py:798] Waited 0.28923916816711426 seconds for step 5 to finish before starting checkpointing.
I0420 22:11:23.270772 135340420060992 checkpoint_manager.py:2024] [process=1][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0420 22:11:27.240516    2597 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0420 22:11:30.890221 135214521100032 array_metadata_store.py:203] [process=1][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_true/checkpoints/0/items/array_metadatas/process_1
I0420 22:11:59.673346 135214512707328 base_pytree_checkpoint_handler.py:129] [process=1] /jax/checkpoint/write/gbytes_per_sec: 40.735 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 38.77720785140991 s) (per-host)
I0420 22:11:59.673475 135214512707328 async_checkpointer.py:90] [process=1][thread=async_save] 3 Handler Commit operations completed. Time taken: 38.274096s.
I0420 22:12:08.862589 135214512707328 async_checkpointer.py:144] [process=1][thread=async_save] Background save thread done. Time taken: 47.463193s.
I0420 22:12:08.862874 135212927272704 async_checkpointer.py:273] [process=1][thread=save_finalize] Done with waiting for background save thread=async_save.
I0420 22:12:08.863001 135212927272704 async_checkpointer.py:283] [process=1][thread=save_finalize] No errors found in background save thread=async_save.
I0420 22:12:08.863051 135212927272704 checkpoint_manager.py:2133] [process=1][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0420 22:12:08.865510 135212927272704 checkpoint_manager.py:2142] [process=1][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0420 22:12:08.865692 135340420060992 checkpoint_manager.py:2036] [process=1][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0420 22:12:08.865822 135340420060992 checkpoint_manager.py:1458] Waiting for previous save to complete took 45.595055 seconds. If this number is high, consider checkpointing less frequently.
I0420 22:12:08.867886 135340420060992 checkpoint_manager.py:1518] [process=1] Saving checkpoint at step 5
I0420 22:12:08.871955 135340420060992 async_checkpointer.py:452] [process=1] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_true/checkpoints/5.
I0420 22:12:09.422371 135340420060992 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0420 22:12:09.422541 135340420060992 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
I0420 22:12:09.465491 135340420060992 base_pytree_checkpoint_handler.py:153] [process=1][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.045030s
I0420 22:12:09.465658 135340420060992 base_pytree_checkpoint_handler.py:129] [process=1] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.031 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.3066103458404541 s) (per-host)
I0420 22:12:09.465711 135340420060992 base_pytree_checkpoint_handler.py:737] [process=1][thread=MainThread] Initiated Pytree async_save. Time taken: 0.306674s (batch_requests_ready=0.256491s, total_serialization_initiated=0.050113s, others=0.000069s)
I0420 22:12:09.465804 135340420060992 composite_checkpoint_handler.py:715] [process=1][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.311422s (all_items=0.000015s, per_item={'items': '0.00001478'}, temp_paths=0.311407)
I0420 22:12:09.466649 135213447358208 async_checkpointer.py:79] [process=1][thread=async_save] Background save thread started.
I0420 22:12:09.466810 135340420060992 async_checkpointer.py:561] Finished blocking save. Time taken: 0.598853s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_true/checkpoints/5.
I0420 22:12:09.898680 135340420060992 checkpoint_manager.py:1566] [process=1][thread=MainThread][step=5] Starting CheckpointManager Save Finalize thread=save_finalize
I0420 22:12:09.899082 135212927272704 async_checkpointer.py:265] [process=1][thread=save_finalize] Waiting for background save thread=async_save.
I0420 22:12:09.899258 135340420060992 standard_logger.py:34] {'step': 5, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776723083.2707434, 'wait_for_prev_duration_secs': 45.59505486488342, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776723128.8679237, 'checkpointer_blocking_duration_secs': 0.5989990234375, 'get_old_steps_start_time': 1776723129.4669425, 'get_old_steps_duration_secs': 2.4557113647460938e-05, 'checkpoint_manager_blocking_start_time': 1776723083.2684138, 'checkpoint_manager_blocking_duration_secs': 46.63080549240112}
I0420 22:12:09.899437 135340420060992 checkpointing.py:409] Started an asynchronous checkpoint save for step 5
I0420 22:12:09.900251 135340420060992 metric_logger.py:185] completed step: 4, seconds: 0.162, TFLOP/s/device: 83.619, Tokens/s/device: 12604.008, total_weights: 65536, loss: 9.457
I0420 22:12:09.907913 135340420060992 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.594, Tokens/s/device: 13956.848, total_weights: 65536, loss: 8.955
I0420 22:12:29.681932 135340420060992 metric_logger.py:185] completed step: 6, seconds: 46.927, TFLOP/s/device: 0.290, Tokens/s/device: 43.642, total_weights: 65536, loss: 8.465
I0420 22:12:29.829341 135340420060992 metric_logger.py:185] completed step: 7, seconds: 19.629, TFLOP/s/device: 0.692, Tokens/s/device: 104.337, total_weights: 65536, loss: 7.998
I0420 22:12:29.976865 135340420060992 metric_logger.py:185] completed step: 8, seconds: 0.154, TFLOP/s/device: 88.514, Tokens/s/device: 13341.846, total_weights: 65536, loss: 7.567
I0420 22:12:30.125265 135340420060992 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0420 22:12:30.125884 135340420060992 checkpointing.py:798] Waited 0.0006291866302490234 seconds for step 9 to finish before starting checkpointing.
I0420 22:12:30.128351 135340420060992 checkpoint_manager.py:2024] [process=1][thread=MainThread][step=5][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0420 22:12:33.165085 135214521100032 array_metadata_store.py:203] [process=1][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_true/checkpoints/5/items/array_metadatas/process_1
I0420 22:13:09.938335 135213447358208 base_pytree_checkpoint_handler.py:129] [process=1] /jax/checkpoint/write/gbytes_per_sec: 25.989 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 60.7792444229126 s) (per-host)
I0420 22:13:09.938467 135213447358208 async_checkpointer.py:90] [process=1][thread=async_save] 3 Handler Commit operations completed. Time taken: 60.471705s.
I0420 22:13:18.474422 135213447358208 async_checkpointer.py:144] [process=1][thread=async_save] Background save thread done. Time taken: 69.007633s.
I0420 22:13:18.474716 135212927272704 async_checkpointer.py:273] [process=1][thread=save_finalize] Done with waiting for background save thread=async_save.
I0420 22:13:18.474879 135212927272704 async_checkpointer.py:283] [process=1][thread=save_finalize] No errors found in background save thread=async_save.
I0420 22:13:18.474937 135212927272704 checkpoint_manager.py:2133] [process=1][thread=save_finalize][step=5] CheckpointManager Save Finalize is syncing with other hosts...
I0420 22:13:18.476455 135212927272704 checkpoint_manager.py:2142] [process=1][thread=save_finalize][step=5] CheckpointManager Save Finalize is done on all hosts.
I0420 22:13:18.476655 135340420060992 checkpoint_manager.py:2036] [process=1][thread=MainThread][step=5][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=5.
W0420 22:13:18.476784 135340420060992 checkpoint_manager.py:1458] Waiting for previous save to complete took 48.348444 seconds. If this number is high, consider checkpointing less frequently.
I0420 22:13:18.478354 135340420060992 checkpoint_manager.py:1518] [process=1] Saving checkpoint at step 9
I0420 22:13:18.481713 135340420060992 async_checkpointer.py:452] [process=1] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_true/checkpoints/9.
I0420 22:13:19.048256 135340420060992 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0420 22:13:19.048351 135340420060992 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
I0420 22:13:19.090986 135340420060992 base_pytree_checkpoint_handler.py:153] [process=1][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.044704s
I0420 22:13:19.091151 135340420060992 base_pytree_checkpoint_handler.py:129] [process=1] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.160 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2989230155944824 s) (per-host)
I0420 22:13:19.091202 135340420060992 base_pytree_checkpoint_handler.py:737] [process=1][thread=MainThread] Initiated Pytree async_save. Time taken: 0.298985s (batch_requests_ready=0.249536s, total_serialization_initiated=0.049382s, others=0.000067s)
I0420 22:13:19.091301 135340420060992 composite_checkpoint_handler.py:715] [process=1][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.303177s (all_items=0.000012s, per_item={'items': '0.00001168'}, temp_paths=0.303166)
I0420 22:13:19.092225 135214512707328 async_checkpointer.py:79] [process=1][thread=async_save] Background save thread started.
I0420 22:13:19.092353 135340420060992 async_checkpointer.py:561] Finished blocking save. Time taken: 0.613930s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_true/checkpoints/9.
I0420 22:13:19.107475 135340420060992 checkpoint_manager.py:1566] [process=1][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0420 22:13:19.107703 135212927272704 async_checkpointer.py:265] [process=1][thread=save_finalize] Waiting for background save thread=async_save.
I0420 22:13:19.107858 135340420060992 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776723150.1283236, 'wait_for_prev_duration_secs': 48.34844446182251, 'time_between_consecutive_saves_sec': 21.262773513793945, 'checkpointer_blocking_start_time': 1776723198.4783926, 'checkpointer_blocking_duration_secs': 0.6141352653503418, 'get_old_steps_start_time': 1776723199.0925415, 'get_old_steps_duration_secs': 2.0742416381835938e-05, 'checkpoint_manager_blocking_start_time': 1776723150.126106, 'checkpoint_manager_blocking_duration_secs': 48.98172116279602}
I0420 22:13:19.107969 135340420060992 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0420 22:13:19.108012 135340420060992 checkpoint_manager.py:2024] [process=1][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0420 22:13:24.387298 135214521100032 array_metadata_store.py:203] [process=1][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_true/checkpoints/9/items/array_metadatas/process_1
I0420 22:14:01.207395 135214512707328 base_pytree_checkpoint_handler.py:129] [process=1] /jax/checkpoint/write/gbytes_per_sec: 37.241 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.41512751579285 s) (per-host)
I0420 22:14:01.207535 135214512707328 async_checkpointer.py:90] [process=1][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.115096s.
I0420 22:14:11.036876 135214512707328 async_checkpointer.py:144] [process=1][thread=async_save] Background save thread done. Time taken: 51.944422s.
I0420 22:14:11.037149 135212927272704 async_checkpointer.py:273] [process=1][thread=save_finalize] Done with waiting for background save thread=async_save.
I0420 22:14:11.037268 135212927272704 async_checkpointer.py:283] [process=1][thread=save_finalize] No errors found in background save thread=async_save.
I0420 22:14:11.037310 135212927272704 checkpoint_manager.py:2133] [process=1][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0420 22:14:11.040177 135212927272704 checkpoint_manager.py:2142] [process=1][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0420 22:14:11.040341 135340420060992 checkpoint_manager.py:2036] [process=1][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0420 22:14:11.040521 135340420060992 checkpoint_manager.py:2013] [process=1][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0420 22:14:11.041207 135340420060992 metric_logger.py:185] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.524, Tokens/s/device: 13946.204, total_weights: 65536, loss: 7.176
Per train step:
 Total TFLOPs: 13.59 
 split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Mon Apr 20 22:14:24 UTC 2026
EXIT_CODE=0