MaxView

← Back to run

Log Summary

XPK Start: Sun Apr 19 06:14:28 UTC 2026
2026-04-19 06:14:32.301683: 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 00:00:1776579272.314399      10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
E0000 00:00:1776579272.318103      10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
W0000 00:00:1776579272.329370      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776579272.329391      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776579272.329393      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776579272.329395      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
2026-04-19 06:14:51.604142: 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)
I0419 06:14:52.134233 136140079724352 max_utils.py:273] Attempting to initialize the jax distributed system...
INFO:2026-04-19 06:15:01,175:jax._src.distributed:140: Starting JAX distributed service on [::]:8482
I0419 06:15:01.175798 136140079724352 distributed.py:140] Starting JAX distributed service on [::]:8482
INFO:2026-04-19 06:15:01,178:jax._src.distributed:157: Connecting to JAX distributed service on mt-16-zarr3-true-sv-6k0v3-slice-job-0-0.mt-16-zarr3-true-sv-6k0v3:8482
I0419 06:15:01.178133 136140079724352 distributed.py:157] Connecting to JAX distributed service on mt-16-zarr3-true-sv-6k0v3-slice-job-0-0.mt-16-zarr3-true-sv-6k0v3:8482
I0419 06:15:03.182451 136140079724352 max_utils.py:284] Jax distributed system initialized!
I0419 06:15:09.994300 136140079724352 max_utils.py:800] System Information: Jax Version: 0.8.1
I0419 06:15:09.994408 136140079724352 max_utils.py:801] System Information: Jaxlib Version: 0.8.1
I0419 06:15:09.994449 136140079724352 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
I0419 06:15:09.994483 136140079724352 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0419 06:15:10.602777 136140079724352 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0419 06:15:10.719163 136140079724352 checkpointing.py:688] Setting up checkpoint logger...
I0419 06:15:10.719292 136140079724352 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0419 06:15:10.719337 136140079724352 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None
I0419 06:15:10.719558 136140079724352 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 0x7bd11bdcf980>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0419 06:15:13.687775 136140079724352 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0419 06:15:13.688178 136140079724352 checkpoint_manager.py:709] [process=0][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7bcfcfc7b380>}, handler_registry=None
I0419 06:15:13.688424 136140079724352 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7bcfcfc7b380>` 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`.
I0419 06:15:13.688474 136140079724352 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7bbcac68ae70>` 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`.
I0419 06:15:13.688511 136140079724352 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 0x7bcfcfc7b380>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7bcfcfc7b380>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7bbcac68ae70>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7bbcac68ae70>}).
I0419 06:15:13.688854 136140079724352 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33
I0419 06:15:13.688933 136140079724352 async_checkpointer.py:177] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7bbcac71cf40> timeout: 600 secs and primary_host=0 for async checkpoint writes
I0419 06:15:15.160387 136140079724352 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints
I0419 06:15:15.595031 136140079724352 checkpoint_manager.py:929] [process=0][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_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7bbdd7471100>
I0419 06:15:15.595203 136140079724352 checkpointing.py:302] Checkpoint manager created!
I0419 06:15:16.510225 136140079724352 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0419 06:15:16.510348 136140079724352 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0419 06:15:17.923681 136140079724352 maxtext_utils.py:1749]  decoder/decoder_norm/scale/value
    Shape:     float32[2048]
    Physical:  (None,)
I0419 06:15:17.923803 136140079724352 maxtext_utils.py:1749]  decoder/layers/mlp/wi_0/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0419 06:15:17.923875 136140079724352 maxtext_utils.py:1749]  decoder/layers/mlp/wi_1/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0419 06:15:17.923922 136140079724352 maxtext_utils.py:1749]  decoder/layers/mlp/wo/kernel/value
    Shape:     float32[7168,16,2048]
    Physical:  (None, None, 'fsdp')
I0419 06:15:17.923962 136140079724352 maxtext_utils.py:1749]  decoder/layers/post_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0419 06:15:17.923997 136140079724352 maxtext_utils.py:1749]  decoder/layers/pre_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0419 06:15:17.924035 136140079724352 maxtext_utils.py:1749]  decoder/layers/self_attention/key/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0419 06:15:17.924072 136140079724352 maxtext_utils.py:1749]  decoder/layers/self_attention/out/kernel/value
    Shape:     float32[16,16,128,2048]
    Physical:  (None, None, None, 'fsdp')
I0419 06:15:17.924107 136140079724352 maxtext_utils.py:1749]  decoder/layers/self_attention/query/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0419 06:15:17.924141 136140079724352 maxtext_utils.py:1749]  decoder/layers/self_attention/value/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)

I0419 06:15:17.924175 136140079724352 maxtext_utils.py:1749]  decoder/logits_dense/kernel/value
    Shape:     float32[2048,32000]
    Physical:  ('fsdp', None)
I0419 06:15:17.924208 136140079724352 maxtext_utils.py:1749]  token_embedder/embedding/value
    Shape:     float32[32000,2048]
    Physical:  (None, 'fsdp')
I0419 06:15:18.322161 136140079724352 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0419 06:15:18.322311 136140079724352 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0419 06:15:18.327691 136140079724352 nnx_decoders.py:616] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0419 06:15:18.327786 136140079724352 nnx_decoders.py:616] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0419 06:15:18.481081 136140079724352 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0419 06:15:18.481218 136140079724352 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0419 06:15:18.633753 136140079724352 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0419 06:15:18.633906 136140079724352 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0419 06:15:18.799037 136140079724352 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0419 06:15:18.799173 136140079724352 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0419 06:15:18.948860 136140079724352 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0419 06:15:18.948993 136140079724352 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0419 06:15:19.099033 136140079724352 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0419 06:15:19.099166 136140079724352 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0419 06:15:19.260107 136140079724352 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0419 06:15:19.260248 136140079724352 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0419 06:15:19.284281 136140079724352 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0419 06:15:19.284380 136140079724352 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0419 06:15:30.154010 136140079724352 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.
I0419 06:15:30.161857 136140079724352 metric_logger.py:289] number parameters: 1.104 billion
I0419 06:15:41.458128 136140079724352 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0419 06:15:41.608737 136140079724352 checkpointing.py:798] Waited 0.15059804916381836 seconds for step 0 to finish before starting checkpointing.
I0419 06:15:41.611315 136140079724352 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0419 06:15:41.613274 136140079724352 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 0
I0419 06:15:41.616039 136140079724352 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/0.
I0419 06:15:42.352540 136140079724352 signaling_client.py:364] Using JaxDistributedSignalingClient
I0419 06:15:42.606913 136012705535744 atomicity.py:137] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/0
I0419 06:15:42.672172 136140079724352 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0419 06:15:42.672270 136140079724352 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
I0419 06:15:42.914391 136140079724352 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.244207s
I0419 06:15:42.915168 136140079724352 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.780 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5549483299255371 s) (per-host)
I0419 06:15:42.915245 136140079724352 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.555050s (batch_requests_ready=0.302427s, total_serialization_initiated=0.251923s, others=0.000700s)
I0419 06:15:42.915350 136140079724352 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.558976s (all_items=0.000020s, per_item={'items': '0.00001955'}, temp_paths=0.558957)
I0419 06:15:42.916269 136014341326592 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started.
I0419 06:15:42.916378 136140079724352 async_checkpointer.py:561] Finished blocking save. Time taken: 1.303030s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/0.
I0419 06:15:43.329409 136140079724352 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0419 06:15:43.329779 136012219021056 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0419 06:15:43.329967 136140079724352 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776579341.6112964, 'wait_for_prev_duration_secs': 6.556510925292969e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776579341.6133134, 'checkpointer_blocking_duration_secs': 1.303168535232544, 'get_old_steps_start_time': 1776579342.9165032, 'get_old_steps_duration_secs': 2.3603439331054688e-05, 'checkpoint_manager_blocking_start_time': 1776579341.60918, 'checkpoint_manager_blocking_duration_secs': 1.7207458019256592}
I0419 06:15:43.330091 136140079724352 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0419 06:15:43.330173 136140079724352 max_utils.py:750] 
Memstats: After params initialized:
I0419 06:15:43.330232 136140079724352 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0))
I0419 06:15:43.330264 136140079724352 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0))
I0419 06:15:43.330291 136140079724352 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0))
I0419 06:15:43.330317 136140079724352 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0))
I0419 06:15:43.711967 136140079724352 metric_logger.py:185] completed step: 0, seconds: 11.022, TFLOP/s/device: 1.233, Tokens/s/device: 185.808, total_weights: 65536, loss: 10.874
I0419 06:15:43.714892 136140079724352 metric_logger.py:269] To see full metrics 'tensorboard --logdir=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/tensorboard/'
I0419 06:15:44.130721 136140079724352 metric_logger.py:185] completed step: 1, seconds: 2.250, TFLOP/s/device: 6.038, Tokens/s/device: 910.152, total_weights: 65536, loss: 10.874
I0419 06:15:44.272749 136140079724352 metric_logger.py:185] completed step: 2, seconds: 0.420, TFLOP/s/device: 32.340, Tokens/s/device: 4874.693, total_weights: 65536, loss: 10.560
I0419 06:15:44.420258 136140079724352 metric_logger.py:185] completed step: 3, seconds: 0.011, TFLOP/s/device: 1193.319, Tokens/s/device: 179870.016, total_weights: 65536, loss: 9.979
I0419 06:15:44.432385 136140079724352 checkpointing.py:794] Waiting for step 5 to finish before checkpoint...
I0419 06:15:44.715402 136140079724352 checkpointing.py:798] Waited 0.28297948837280273 seconds for step 5 to finish before starting checkpointing.
I0419 06:15:44.718210 136140079724352 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0419 06:15:45.707998 136013788477184 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776579345303057479, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776579345303057479, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/0/_CHECKPOINT_METADATA
I0419 06:15:45.987876 136014358112000 atomicity.py:137] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/0/items
I0419 06:15:47.126867    2614 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0419 06:15:48.774423 136017545766656 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/0/items/array_metadatas/process_0
I0419 06:16:24.349744 136017537373952 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 37.235497s (commit=36.086131s, array_metadata_write=1.149366s)
I0419 06:16:24.352743 136014341326592 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 37.616 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.99248790740967 s) (per-host)
I0419 06:16:24.352926 136014341326592 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.436591s.
I0419 06:16:25.432111 136014341326592 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776579345303057479, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/0/_CHECKPOINT_METADATA
I0419 06:16:26.194796 136013788477184 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776579345303057479, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/0/_CHECKPOINT_METADATA
I0419 06:16:26.616133 136014341326592 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000756s.
I0419 06:16:26.818642 136014341326592 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0419 06:16:27.779093 136014341326592 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.226439s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/0/items
I0419 06:16:27.781197 136014341326592 atomicity.py:608] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/0/items
I0419 06:16:28.225491 136014341326592 atomicity.py:608] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/0
I0419 06:16:29.970699 136014341326592 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/0`.
I0419 06:16:29.971716 136014341326592 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 48.358384s. directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/0
I0419 06:16:29.973224 136014341326592 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 47.056888s.
I0419 06:16:29.973478 136012219021056 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0419 06:16:29.973639 136012219021056 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0419 06:16:29.973741 136012219021056 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0419 06:16:29.975727 136012219021056 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0419 06:16:29.975945 136140079724352 checkpoint_manager.py:2036] [process=0][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0419 06:16:29.976066 136140079724352 checkpoint_manager.py:1458] Waiting for previous save to complete took 45.257872 seconds. If this number is high, consider checkpointing less frequently.
I0419 06:16:29.977996 136140079724352 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 5
I0419 06:16:29.982020 136140079724352 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/5.
I0419 06:16:30.929020 136014341326592 atomicity.py:137] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/5
I0419 06:16:30.967293 136140079724352 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0419 06:16:30.967441 136140079724352 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
I0419 06:16:30.999013 136140079724352 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.035770s
I0419 06:16:30.999929 136140079724352 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 4.922 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.3133881092071533 s) (per-host)
I0419 06:16:31.000032 136140079724352 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.313513s (batch_requests_ready=0.264127s, total_serialization_initiated=0.048544s, others=0.000842s)
I0419 06:16:31.000166 136140079724352 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.318237s (all_items=0.000024s, per_item={'items': '0.00002408'}, temp_paths=0.318213)
I0419 06:16:31.001280 136011660875520 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started.
I0419 06:16:31.001487 136140079724352 async_checkpointer.py:561] Finished blocking save. Time taken: 1.023390s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/5.
I0419 06:16:31.430530 136140079724352 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=5] Starting CheckpointManager Save Finalize thread=save_finalize
I0419 06:16:31.431054 136012219021056 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0419 06:16:31.431301 136140079724352 standard_logger.py:34] {'step': 5, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776579344.7181659, 'wait_for_prev_duration_secs': 45.257871866226196, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776579389.97806, 'checkpointer_blocking_duration_secs': 1.0235753059387207, 'get_old_steps_start_time': 1776579391.0016572, 'get_old_steps_duration_secs': 3.838539123535156e-05, 'checkpoint_manager_blocking_start_time': 1776579344.7158954, 'checkpoint_manager_blocking_duration_secs': 46.71535277366638}
I0419 06:16:31.431451 136140079724352 checkpointing.py:409] Started an asynchronous checkpoint save for step 5
I0419 06:16:51.760706 136140079724352 metric_logger.py:185] completed step: 4, seconds: 0.144, TFLOP/s/device: 94.461, Tokens/s/device: 14238.240, total_weights: 65536, loss: 9.457
I0419 06:16:51.771566 136140079724352 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.377, Tokens/s/device: 13924.016, total_weights: 65536, loss: 8.955
I0419 06:16:51.915489 136140079724352 metric_logger.py:185] completed step: 6, seconds: 67.339, TFLOP/s/device: 0.202, Tokens/s/device: 30.413, total_weights: 65536, loss: 8.465
I0419 06:16:52.063064 136140079724352 metric_logger.py:185] completed step: 7, seconds: 0.009, TFLOP/s/device: 1551.573, Tokens/s/device: 233870.047, total_weights: 65536, loss: 7.998
I0419 06:16:52.210489 136140079724352 metric_logger.py:185] completed step: 8, seconds: 0.145, TFLOP/s/device: 93.398, Tokens/s/device: 14077.924, total_weights: 65536, loss: 7.567
I0419 06:16:52.358848 136140079724352 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0419 06:16:52.359528 136140079724352 checkpointing.py:798] Waited 0.0006926059722900391 seconds for step 9 to finish before starting checkpointing.
I0419 06:16:52.361914 136140079724352 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=5][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0419 06:16:52.701865 136012755891968 atomicity.py:137] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/5/items
I0419 06:16:55.567170 136011677660928 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/5/items/array_metadatas/process_0
I0419 06:17:36.552084 136011669268224 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 42.660377s (commit=41.481391s, array_metadata_write=1.178986s)
I0419 06:17:36.553177 136011660875520 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 23.982 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 65.86661648750305 s) (per-host)
I0419 06:17:36.553297 136011660875520 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 65.551866s.
I0419 06:17:38.653097 136011660875520 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0419 06:17:38.837697 136011660875520 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000411s.
I0419 06:17:39.718719 136011660875520 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.916723s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/5/items
I0419 06:17:39.720625 136011660875520 atomicity.py:608] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/5/items
I0419 06:17:40.153937 136011660875520 atomicity.py:608] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/5
I0419 06:17:41.453428 136011660875520 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/5`.
I0419 06:17:41.454266 136011660875520 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 71.476180s. directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/5
I0419 06:17:41.455926 136011660875520 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 70.454493s.
I0419 06:17:41.456083 136012219021056 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0419 06:17:41.456144 136012219021056 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0419 06:17:41.456254 136012219021056 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=5] CheckpointManager Save Finalize is syncing with other hosts...
I0419 06:17:41.458188 136012219021056 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=5] CheckpointManager Save Finalize is done on all hosts.
I0419 06:17:41.458325 136140079724352 checkpoint_manager.py:2036] [process=0][thread=MainThread][step=5][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=5.
W0419 06:17:41.458410 136140079724352 checkpoint_manager.py:1458] Waiting for previous save to complete took 49.096509 seconds. If this number is high, consider checkpointing less frequently.
I0419 06:17:41.459983 136140079724352 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 9
I0419 06:17:41.463412 136140079724352 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/9.
I0419 06:17:42.409621 136011660875520 atomicity.py:137] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/9
I0419 06:17:42.437731 136140079724352 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0419 06:17:42.437848 136140079724352 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
I0419 06:17:42.464279 136140079724352 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.028592s
I0419 06:17:42.465166 136140079724352 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.215 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.29581713676452637 s) (per-host)
I0419 06:17:42.465235 136140079724352 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.295904s (batch_requests_ready=0.260819s, total_serialization_initiated=0.034286s, others=0.000798s)
I0419 06:17:42.465318 136140079724352 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.300163s (all_items=0.000016s, per_item={'items': '0.00001645'}, temp_paths=0.300146)
I0419 06:17:42.466184 136011669268224 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started.
I0419 06:17:42.466336 136140079724352 async_checkpointer.py:561] Finished blocking save. Time taken: 1.006285s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/9.
I0419 06:17:42.909499 136140079724352 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0419 06:17:42.909877 136012219021056 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0419 06:17:42.910050 136140079724352 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776579412.361883, 'wait_for_prev_duration_secs': 49.09650897979736, 'time_between_consecutive_saves_sec': 22.3860867023468, 'checkpointer_blocking_start_time': 1776579461.4600213, 'checkpointer_blocking_duration_secs': 1.0064244270324707, 'get_old_steps_start_time': 1776579462.4664633, 'get_old_steps_duration_secs': 2.5033950805664062e-05, 'checkpoint_manager_blocking_start_time': 1776579412.3597445, 'checkpoint_manager_blocking_duration_secs': 50.5502712726593}
I0419 06:17:42.910176 136140079724352 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0419 06:17:42.910220 136140079724352 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0419 06:17:44.158287 136014341351168 atomicity.py:137] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/9/items
I0419 06:17:47.395451 136012169471744 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/9/items/array_metadatas/process_0
I0419 06:18:27.448962 136011677660928 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 42.158577s (commit=40.636953s, array_metadata_write=1.521625s)
I0419 06:18:27.450098 136011669268224 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 34.885 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 45.28071737289429 s) (per-host)
I0419 06:18:27.450220 136011669268224 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 44.983925s.
I0419 06:18:29.185559 136011669268224 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0419 06:18:29.359740 136011669268224 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000584s.
I0419 06:18:30.264121 136011669268224 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.920541s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/9/items
I0419 06:18:30.266756 136011669268224 atomicity.py:608] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/9/items
I0419 06:18:30.665843 136011669268224 atomicity.py:608] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/9
I0419 06:18:32.835111 136011669268224 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/9`.
I0419 06:18:32.835937 136011669268224 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 51.375891s. directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_16_zarr3_true/checkpoints/9
I0419 06:18:32.837571 136011669268224 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 50.371275s.
I0419 06:18:32.837776 136012219021056 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0419 06:18:32.837899 136012219021056 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0419 06:18:32.837974 136012219021056 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0419 06:18:32.839947 136012219021056 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0419 06:18:32.840072 136140079724352 checkpoint_manager.py:2036] [process=0][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0419 06:18:32.840189 136140079724352 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0419 06:18:32.840886 136140079724352 metric_logger.py:185] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.241, Tokens/s/device: 13903.504, 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: Sun Apr 19 06:18:43 UTC 2026
EXIT_CODE=0