MaxView

← Back to run

Log Summary

XPK Start: Sun Apr 19 05:55:42 UTC 2026
2026-04-19 05:55:46.202401: 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 00:00:1776578146.214988      10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
E0000 00:00:1776578146.218677      10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
W0000 00:00:1776578146.230464      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776578146.230498      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776578146.230500      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776578146.230503      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
2026-04-19 05:56:05.315773: 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 05:56:05.843630 136235037185856 max_utils.py:273] Attempting to initialize the jax distributed system...
INFO:2026-04-19 05:56:14,885:jax._src.distributed:140: Starting JAX distributed service on [::]:8482
I0419 05:56:14.885450 136235037185856 distributed.py:140] Starting JAX distributed service on [::]:8482
INFO:2026-04-19 05:56:14,887:jax._src.distributed:157: Connecting to JAX distributed service on mt-15-ocdbt-true-sv-umhuz-slice-job-0-0.mt-15-ocdbt-true-sv-umhuz:8482
I0419 05:56:14.887748 136235037185856 distributed.py:157] Connecting to JAX distributed service on mt-15-ocdbt-true-sv-umhuz-slice-job-0-0.mt-15-ocdbt-true-sv-umhuz:8482
I0419 05:56:16.460692 136235037185856 max_utils.py:284] Jax distributed system initialized!
I0419 05:56:22.381482 136235037185856 max_utils.py:800] System Information: Jax Version: 0.8.1
I0419 05:56:22.381585 136235037185856 max_utils.py:801] System Information: Jaxlib Version: 0.8.1
I0419 05:56:22.381626 136235037185856 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 05:56:22.381662 136235037185856 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0419 05:56:22.996391 136235037185856 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0419 05:56:23.111173 136235037185856 checkpointing.py:688] Setting up checkpoint logger...
I0419 05:56:23.111289 136235037185856 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0419 05:56:23.111334 136235037185856 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None
I0419 05:56:23.111541 136235037185856 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 0x7be725179250>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0419 05:56:25.841507 136235037185856 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0419 05:56:25.841862 136235037185856 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 0x7bd4ba690440>}, handler_registry=None
I0419 05:56:25.842097 136235037185856 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7bd4ba690440>` 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 05:56:25.842150 136235037185856 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7bd4b88c58b0>` 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 05:56:25.842185 136235037185856 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 0x7bd4ba690440>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7bd4ba690440>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7bd4b88c58b0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7bd4b88c58b0>}).
I0419 05:56:25.842509 136235037185856 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33
I0419 05:56:25.842584 136235037185856 async_checkpointer.py:177] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7bd4b9258f40> timeout: 600 secs and primary_host=0 for async checkpoint writes
I0419 05:56:27.914935 136235037185856 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_15_ocdbt_true/checkpoints
I0419 05:56:29.206267 136235037185856 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_15_ocdbt_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7bd4b92dc560>
I0419 05:56:29.206436 136235037185856 checkpointing.py:302] Checkpoint manager created!
I0419 05:56:30.116472 136235037185856 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0419 05:56:30.116588 136235037185856 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0419 05:56:31.534336 136235037185856 maxtext_utils.py:1749]  decoder/decoder_norm/scale/value
    Shape:     float32[2048]
    Physical:  (None,)
I0419 05:56:31.534467 136235037185856 maxtext_utils.py:1749]  decoder/layers/mlp/wi_0/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0419 05:56:31.534524 136235037185856 maxtext_utils.py:1749]  decoder/layers/mlp/wi_1/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0419 05:56:31.534593 136235037185856 maxtext_utils.py:1749]  decoder/layers/mlp/wo/kernel/value
    Shape:     float32[7168,16,2048]
    Physical:  (None, None, 'fsdp')
I0419 05:56:31.534671 136235037185856 maxtext_utils.py:1749]  decoder/layers/post_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0419 05:56:31.534719 136235037185856 maxtext_utils.py:1749]  decoder/layers/pre_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0419 05:56:31.534760 136235037185856 maxtext_utils.py:1749]  decoder/layers/self_attention/key/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0419 05:56:31.534796 136235037185856 maxtext_utils.py:1749]  decoder/layers/self_attention/out/kernel/value
    Shape:     float32[16,16,128,2048]
    Physical:  (None, None, None, 'fsdp')
I0419 05:56:31.534849 136235037185856 maxtext_utils.py:1749]  decoder/layers/self_attention/query/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0419 05:56:31.534885 136235037185856 maxtext_utils.py:1749]  decoder/layers/self_attention/value/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0419 05:56:31.534919 136235037185856 maxtext_utils.py:1749]  decoder/logits_dense/kernel/value
    Shape:     float32[2048,32000]
    Physical:  ('fsdp', None)
I0419 05:56:31.534952 136235037185856 maxtext_utils.py:1749]  token_embedder/embedding/value
    Shape:     float32[32000,2048]
    Physical:  (None, 'fsdp')

I0419 05:56:31.930927 136235037185856 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0419 05:56:31.931073 136235037185856 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0419 05:56:31.936420 136235037185856 nnx_decoders.py:616] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0419 05:56:31.936515 136235037185856 nnx_decoders.py:616] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0419 05:56:32.088307 136235037185856 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0419 05:56:32.088438 136235037185856 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0419 05:56:32.245147 136235037185856 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0419 05:56:32.245288 136235037185856 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0419 05:56:32.413321 136235037185856 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 05:56:32.413456 136235037185856 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0419 05:56:32.563262 136235037185856 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 05:56:32.563398 136235037185856 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0419 05:56:32.712412 136235037185856 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 05:56:32.712544 136235037185856 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0419 05:56:32.871526 136235037185856 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0419 05:56:32.871663 136235037185856 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0419 05:56:32.895853 136235037185856 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0419 05:56:32.895951 136235037185856 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0419 05:56:44.009081 136235037185856 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 05:56:44.017532 136235037185856 metric_logger.py:289] number parameters: 1.104 billion
I0419 05:56:55.576044 136235037185856 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0419 05:56:55.728189 136235037185856 checkpointing.py:798] Waited 0.15213251113891602 seconds for step 0 to finish before starting checkpointing.
I0419 05:56:55.730584 136235037185856 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0419 05:56:55.732300 136235037185856 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 0
I0419 05:56:55.735272 136235037185856 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_15_ocdbt_true/checkpoints/0.
I0419 05:56:56.882113 136235037185856 signaling_client.py:364] Using JaxDistributedSignalingClient
I0419 05:56:57.134499 136110734833408 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_15_ocdbt_true/checkpoints/0
I0419 05:56:57.597650 136235037185856 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0419 05:56:57.597742 136235037185856 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 05:56:57.840423 136235037185856 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.244989s
I0419 05:56:57.841168 136235037185856 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 1.622 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.9511866569519043 s) (per-host)
I0419 05:56:57.841245 136235037185856 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.951282s (batch_requests_ready=0.698102s, total_serialization_initiated=0.252507s, others=0.000673s)
I0419 05:56:57.841332 136235037185856 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.955237s (all_items=0.000020s, per_item={'items': '0.00001955'}, temp_paths=0.955218)
I0419 05:56:57.842303 136110156805888 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started.
I0419 05:56:57.842482 136235037185856 async_checkpointer.py:561] Finished blocking save. Time taken: 2.110114s. 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_15_ocdbt_true/checkpoints/0.
I0419 05:56:57.878751 136235037185856 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0419 05:56:57.879082 136109124220672 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0419 05:56:57.879241 136235037185856 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_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776578215.730567, 'wait_for_prev_duration_secs': 5.984306335449219e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776578215.7323384, 'checkpointer_blocking_duration_secs': 2.1102523803710938, 'get_old_steps_start_time': 1776578217.8426116, 'get_old_steps_duration_secs': 2.3603439331054688e-05, 'checkpoint_manager_blocking_start_time': 1776578215.7286003, 'checkpoint_manager_blocking_duration_secs': 2.150601387023926}
I0419 05:56:57.879348 136235037185856 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0419 05:56:57.879422 136235037185856 max_utils.py:750] 
Memstats: After params initialized:
I0419 05:56:57.879482 136235037185856 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0))
I0419 05:56:57.879517 136235037185856 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0))
I0419 05:56:57.879546 136235037185856 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0))
I0419 05:56:57.879571 136235037185856 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0))
I0419 05:56:58.178598 136235037185856 metric_logger.py:185] completed step: 0, seconds: 11.287, TFLOP/s/device: 1.204, Tokens/s/device: 181.444, total_weights: 65536, loss: 10.874
I0419 05:56:58.181047 136235037185856 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_15_ocdbt_true/tensorboard/'
I0419 05:56:58.595783 136235037185856 metric_logger.py:185] completed step: 1, seconds: 2.599, TFLOP/s/device: 5.228, Tokens/s/device: 787.986, total_weights: 65536, loss: 10.874
I0419 05:56:58.737906 136235037185856 metric_logger.py:185] completed step: 2, seconds: 0.418, TFLOP/s/device: 32.468, Tokens/s/device: 4893.890, total_weights: 65536, loss: 10.560
I0419 05:56:58.885450 136235037185856 metric_logger.py:185] completed step: 3, seconds: 0.012, TFLOP/s/device: 1167.881, Tokens/s/device: 176035.757, total_weights: 65536, loss: 9.979
I0419 05:56:58.897339 136235037185856 checkpointing.py:794] Waiting for step 5 to finish before checkpoint...
I0419 05:56:59.180511 136235037185856 checkpointing.py:798] Waited 0.2831394672393799 seconds for step 5 to finish before starting checkpointing.
I0419 05:56:59.183261 136235037185856 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0419 05:57:00.166972 136110718048000 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_15_ocdbt_true/checkpoints/0/items
I0419 05:57:00.315117 136109661091584 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776578219911332720, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776578219911332720, "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_15_ocdbt_true/checkpoints/0/_CHECKPOINT_METADATA
I0419 05:57:01.710954    2644 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0419 05:57:04.231145 136113947666176 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_15_ocdbt_true/checkpoints/0/items/array_metadatas/process_0
I0419 05:57:38.337919 136113939273472 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 36.630536s (commit=35.535065s, array_metadata_write=1.095471s)
I0419 05:57:38.339292 136110156805888 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 38.109 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.449297189712524 s) (per-host)
I0419 05:57:38.339392 136110156805888 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 40.496969s.
I0419 05:57:39.922656 136110156805888 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776578219911332720, '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_15_ocdbt_true/checkpoints/0/_CHECKPOINT_METADATA
I0419 05:57:41.091349 136110156805888 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000774s.
I0419 05:57:41.106141 136109661091584 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776578219911332720, '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_15_ocdbt_true/checkpoints/0/_CHECKPOINT_METADATA
I0419 05:57:41.320386 136110156805888 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0419 05:57:42.235273 136110156805888 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.172624s. 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_15_ocdbt_true/checkpoints/0/items
I0419 05:57:42.238224 136110156805888 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_15_ocdbt_true/checkpoints/0/items
I0419 05:57:42.647024 136110156805888 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_15_ocdbt_true/checkpoints/0
I0419 05:57:44.346441 136110156805888 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_15_ocdbt_true/checkpoints/0`.
I0419 05:57:44.347442 136110156805888 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 48.615086s. directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/0
I0419 05:57:44.349090 136110156805888 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 46.506665s.
I0419 05:57:44.349354 136109124220672 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0419 05:57:44.349521 136109124220672 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0419 05:57:44.349627 136109124220672 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0419 05:57:44.352382 136109124220672 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0419 05:57:44.352726 136235037185856 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 05:57:44.352912 136235037185856 checkpoint_manager.py:1458] Waiting for previous save to complete took 45.169658 seconds. If this number is high, consider checkpointing less frequently.
I0419 05:57:44.354884 136235037185856 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 5
I0419 05:57:44.358929 136235037185856 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_15_ocdbt_true/checkpoints/5.
I0419 05:57:45.323342 136110156805888 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_15_ocdbt_true/checkpoints/5
I0419 05:57:45.789014 136235037185856 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0419 05:57:45.789148 136235037185856 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 05:57:45.822156 136235037185856 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.037406s
I0419 05:57:45.823147 136235037185856 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.076 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.7430605888366699 s) (per-host)
I0419 05:57:45.823249 136235037185856 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.743185s (batch_requests_ready=0.692613s, total_serialization_initiated=0.049660s, others=0.000913s)
I0419 05:57:45.823381 136235037185856 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.748137s (all_items=0.000026s, per_item={'items': '0.00002623'}, temp_paths=0.748110)
I0419 05:57:45.824549 136109124220672 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started.
I0419 05:57:45.824757 136235037185856 async_checkpointer.py:561] Finished blocking save. Time taken: 1.469773s. 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_15_ocdbt_true/checkpoints/5.
I0419 05:57:45.827374 136235037185856 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=5] Starting CheckpointManager Save Finalize thread=save_finalize
I0419 05:57:45.827712 136108570564352 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0419 05:57:45.827938 136235037185856 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_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776578219.1832185, 'wait_for_prev_duration_secs': 45.16965842247009, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776578264.3549466, 'checkpointer_blocking_duration_secs': 1.4699764251708984, 'get_old_steps_start_time': 1776578265.8249493, 'get_old_steps_duration_secs': 3.719329833984375e-05, 'checkpoint_manager_blocking_start_time': 1776578219.1810663, 'checkpoint_manager_blocking_duration_secs': 46.64681649208069}
I0419 05:57:45.828086 136235037185856 checkpointing.py:409] Started an asynchronous checkpoint save for step 5
I0419 05:58:07.874847 136235037185856 metric_logger.py:185] completed step: 4, seconds: 0.144, TFLOP/s/device: 94.658, Tokens/s/device: 14267.899, total_weights: 65536, loss: 9.457
I0419 05:58:07.885250 136235037185856 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.448, Tokens/s/device: 13934.722, total_weights: 65536, loss: 8.955
I0419 05:58:08.029135 136235037185856 metric_logger.py:185] completed step: 6, seconds: 68.988, TFLOP/s/device: 0.197, Tokens/s/device: 29.686, total_weights: 65536, loss: 8.465
I0419 05:58:08.176727 136235037185856 metric_logger.py:185] completed step: 7, seconds: 0.008, TFLOP/s/device: 1625.060, Tokens/s/device: 244946.777, total_weights: 65536, loss: 7.998
I0419 05:58:08.324208 136235037185856 metric_logger.py:185] completed step: 8, seconds: 0.146, TFLOP/s/device: 93.133, Tokens/s/device: 14037.974, total_weights: 65536, loss: 7.567
I0419 05:58:08.375072 136110709655296 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_15_ocdbt_true/checkpoints/5/items
I0419 05:58:08.472871 136235037185856 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0419 05:58:08.473591 136235037185856 checkpointing.py:798] Waited 0.0007278919219970703 seconds for step 9 to finish before starting checkpointing.
I0419 05:58:08.476004 136235037185856 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=5][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0419 05:58:13.040985 136108587349760 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_15_ocdbt_true/checkpoints/5/items/array_metadatas/process_0
I0419 05:58:52.637247 136108578957056 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 42.538713s (commit=41.470398s, array_metadata_write=1.068315s)
I0419 05:58:52.638375 136109124220672 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 23.381 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 67.55827140808105 s) (per-host)
I0419 05:58:52.638502 136109124220672 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 66.813807s.
I0419 05:58:55.303503 136109124220672 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000391s.
I0419 05:58:55.705384 136109124220672 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0419 05:58:56.647459 136109124220672 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.330435s. 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_15_ocdbt_true/checkpoints/5/items
I0419 05:58:56.649367 136109124220672 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_15_ocdbt_true/checkpoints/5/items
I0419 05:58:57.482499 136109124220672 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_15_ocdbt_true/checkpoints/5
I0419 05:58:58.764124 136109124220672 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_15_ocdbt_true/checkpoints/5`.
I0419 05:58:58.764802 136109124220672 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 74.409841s. directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/5
I0419 05:58:58.766364 136109124220672 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 72.941667s.
I0419 05:58:58.766575 136108570564352 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0419 05:58:58.766714 136108570564352 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0419 05:58:58.766782 136108570564352 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=5] CheckpointManager Save Finalize is syncing with other hosts...
I0419 05:58:58.768754 136108570564352 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=5] CheckpointManager Save Finalize is done on all hosts.
I0419 05:58:58.768883 136235037185856 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 05:58:58.768949 136235037185856 checkpoint_manager.py:1458] Waiting for previous save to complete took 50.292959 seconds. If this number is high, consider checkpointing less frequently.
I0419 05:58:58.775474 136235037185856 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 9
I0419 05:58:58.778885 136235037185856 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_15_ocdbt_true/checkpoints/9.
I0419 05:58:59.329144 136235037185856 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0419 05:58:59.329245 136235037185856 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 05:58:59.355702 136235037185856 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.028675s
I0419 05:58:59.356430 136235037185856 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.415 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2848834991455078 s) (per-host)
I0419 05:58:59.356493 136235037185856 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.284962s (batch_requests_ready=0.250161s, total_serialization_initiated=0.034150s, others=0.000650s)
I0419 05:58:59.356569 136235037185856 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.288861s (all_items=0.000017s, per_item={'items': '0.00001717'}, temp_paths=0.288844)
I0419 05:58:59.357356 136109066278656 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started.
I0419 05:58:59.357506 136235037185856 async_checkpointer.py:561] Finished blocking save. Time taken: 0.581958s. 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_15_ocdbt_true/checkpoints/9.
I0419 05:58:59.739210 136109124220672 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_15_ocdbt_true/checkpoints/9
I0419 05:58:59.819526 136235037185856 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0419 05:58:59.819901 136108570564352 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0419 05:58:59.820075 136235037185856 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_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776578288.4759734, 'wait_for_prev_duration_secs': 50.29295897483826, 'time_between_consecutive_saves_sec': 24.12352204322815, 'checkpointer_blocking_start_time': 1776578338.7755194, 'checkpointer_blocking_duration_secs': 0.5820903778076172, 'get_old_steps_start_time': 1776578339.3576243, 'get_old_steps_duration_secs': 2.384185791015625e-05, 'checkpoint_manager_blocking_start_time': 1776578288.4738374, 'checkpoint_manager_blocking_duration_secs': 51.346203565597534}
I0419 05:58:59.820202 136235037185856 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0419 05:58:59.820245 136235037185856 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0419 05:59:02.302932 136108562171648 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_15_ocdbt_true/checkpoints/9/items
I0419 05:59:05.451191 136110156805888 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_15_ocdbt_true/checkpoints/9/items/array_metadatas/process_0
I0419 05:59:45.316395 136109621860096 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 41.920198s (commit=40.832738s, array_metadata_write=1.087461s)
I0419 05:59:45.317452 136109066278656 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 34.157 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 46.24587297439575 s) (per-host)
I0419 05:59:45.317569 136109066278656 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 45.960104s.
I0419 05:59:46.971194 136109066278656 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000427s.
I0419 05:59:47.023553 136109066278656 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0419 05:59:47.734877 136109066278656 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.610531s. 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_15_ocdbt_true/checkpoints/9/items
I0419 05:59:47.740362 136109066278656 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_15_ocdbt_true/checkpoints/9/items
I0419 05:59:48.563077 136109066278656 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_15_ocdbt_true/checkpoints/9
I0419 05:59:49.869773 136109066278656 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_15_ocdbt_true/checkpoints/9`.
I0419 05:59:49.870580 136109066278656 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 51.095035s. directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/9
I0419 05:59:49.872213 136109066278656 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 50.514747s.
I0419 05:59:49.872437 136108570564352 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0419 05:59:49.872554 136108570564352 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0419 05:59:49.872608 136108570564352 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0419 05:59:49.874235 136108570564352 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0419 05:59:49.874414 136235037185856 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 05:59:49.874570 136235037185856 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0419 05:59:49.875304 136235037185856 metric_logger.py:185] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.514, Tokens/s/device: 13944.779, 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 06:00:02 UTC 2026
EXIT_CODE=0