MaxView

← Back to run

Log Summary

XPK Start: Sun Apr 19 05:31:03 UTC 2026
2026-04-19 05:31:07.746390: 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 00:00:1776576667.759591      10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
E0000 00:00:1776576667.763330      10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
W0000 00:00:1776576667.774663      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776576667.774681      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776576667.774683      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776576667.774685      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
2026-04-19 05:31:26.888650: 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 05:31:27.417913 135160546879296 max_utils.py:273] Attempting to initialize the jax distributed system...
INFO:2026-04-19 05:31:36,459:jax._src.distributed:140: Starting JAX distributed service on [::]:8482
I0419 05:31:36.459800 135160546879296 distributed.py:140] Starting JAX distributed service on [::]:8482
INFO:2026-04-19 05:31:36,462:jax._src.distributed:157: Connecting to JAX distributed service on mt-10-shardy-false-cy4wg-slice-job-0-0.mt-10-shardy-false-cy4wg:8482
I0419 05:31:36.462196 135160546879296 distributed.py:157] Connecting to JAX distributed service on mt-10-shardy-false-cy4wg-slice-job-0-0.mt-10-shardy-false-cy4wg:8482
I0419 05:31:38.159269 135160546879296 max_utils.py:284] Jax distributed system initialized!
I0419 05:31:43.980872 135160546879296 max_utils.py:800] System Information: Jax Version: 0.8.1
I0419 05:31:43.980981 135160546879296 max_utils.py:801] System Information: Jaxlib Version: 0.8.1
I0419 05:31:43.981022 135160546879296 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 05:31:43.981058 135160546879296 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0419 05:31:44.587439 135160546879296 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0419 05:31:44.701460 135160546879296 checkpointing.py:688] Setting up checkpoint logger...
I0419 05:31:44.701582 135160546879296 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0419 05:31:44.701626 135160546879296 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None
I0419 05:31:44.701835 135160546879296 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 0x7aed16f75280>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0419 05:31:48.540795 135160546879296 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0419 05:31:48.541049 135160546879296 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 0x7adaa08b2a20>}, handler_registry=None
I0419 05:31:48.541291 135160546879296 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7adaa08b2a20>` 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 05:31:48.541340 135160546879296 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7adaa08b1af0>` 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 05:31:48.541375 135160546879296 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 0x7adaa08b2a20>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7adaa08b2a20>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7adaa08b1af0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7adaa08b1af0>}).
I0419 05:31:48.541718 135160546879296 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33
I0419 05:31:48.541796 135160546879296 async_checkpointer.py:177] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7ad9ebe38f40> timeout: 600 secs and primary_host=0 for async checkpoint writes
I0419 05:31:50.073566 135160546879296 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_10_shardy_false/checkpoints
I0419 05:31:50.215395 135160546879296 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=10), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_10_shardy_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7ad9ebe97200>
I0419 05:31:50.215572 135160546879296 checkpointing.py:302] Checkpoint manager created!
I0419 05:31:51.134054 135160546879296 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0419 05:31:51.134181 135160546879296 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0419 05:31:52.501378 135160546879296 maxtext_utils.py:1749]  decoder/decoder_norm/scale/value
    Shape:     float32[2048]
    Physical:  (None,)
I0419 05:31:52.501499 135160546879296 maxtext_utils.py:1749]  decoder/layers/mlp/wi_0/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0419 05:31:52.501554 135160546879296 maxtext_utils.py:1749]  decoder/layers/mlp/wi_1/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0419 05:31:52.501599 135160546879296 maxtext_utils.py:1749]  decoder/layers/mlp/wo/kernel/value
    Shape:     float32[7168,16,2048]
    Physical:  (None, None, 'fsdp')
I0419 05:31:52.501636 135160546879296 maxtext_utils.py:1749]  decoder/layers/post_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0419 05:31:52.501671 135160546879296 maxtext_utils.py:1749]  decoder/layers/pre_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0419 05:31:52.501709 135160546879296 maxtext_utils.py:1749]  decoder/layers/self_attention/key/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0419 05:31:52.501745 135160546879296 maxtext_utils.py:1749]  decoder/layers/self_attention/out/kernel/value
    Shape:     float32[16,16,128,2048]
    Physical:  (None, None, None, 'fsdp')
I0419 05:31:52.501781 135160546879296 maxtext_utils.py:1749]  decoder/layers/self_attention/query/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0419 05:31:52.501827 135160546879296 maxtext_utils.py:1749]  decoder/layers/self_attention/value/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0419 05:31:52.501863 135160546879296 maxtext_utils.py:1749]  decoder/logits_dense/kernel/value
    Shape:     float32[2048,32000]
    Physical:  ('fsdp', None)
I0419 05:31:52.501897 135160546879296 maxtext_utils.py:1749]  token_embedder/embedding/value
    Shape:     float32[32000,2048]
    Physical:  (None, 'fsdp')

I0419 05:31:52.899756 135160546879296 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0419 05:31:52.899924 135160546879296 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0419 05:31:52.905285 135160546879296 nnx_decoders.py:616] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0419 05:31:52.905380 135160546879296 nnx_decoders.py:616] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0419 05:31:53.053324 135160546879296 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0419 05:31:53.053457 135160546879296 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0419 05:31:53.196856 135160546879296 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0419 05:31:53.197003 135160546879296 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0419 05:31:53.350483 135160546879296 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 05:31:53.350615 135160546879296 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0419 05:31:53.491254 135160546879296 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 05:31:53.491382 135160546879296 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0419 05:31:53.631158 135160546879296 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 05:31:53.631297 135160546879296 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0419 05:31:53.779993 135160546879296 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0419 05:31:53.780127 135160546879296 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0419 05:31:53.808610 135160546879296 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0419 05:31:53.808743 135160546879296 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0419 05:32:04.774650 135160546879296 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 05:32:04.782870 135160546879296 metric_logger.py:289] number parameters: 1.104 billion
I0419 05:32:15.898404 135160546879296 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0419 05:32:16.054386 135160546879296 checkpointing.py:798] Waited 0.15596914291381836 seconds for step 0 to finish before starting checkpointing.
I0419 05:32:16.057473 135160546879296 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0419 05:32:16.059905 135160546879296 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 0
I0419 05:32:16.062559 135160546879296 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_10_shardy_false/checkpoints/0.
I0419 05:32:16.790275 135160546879296 signaling_client.py:364] Using JaxDistributedSignalingClient
I0419 05:32:17.488141 135034937779968 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_10_shardy_false/checkpoints/0
I0419 05:32:17.489329 135160546879296 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0419 05:32:17.489436 135160546879296 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 05:32:17.732007 135160546879296 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.245139s
I0419 05:32:17.732781 135160546879296 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 1.650 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.9346358776092529 s) (per-host)
I0419 05:32:17.732868 135160546879296 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.934741s (batch_requests_ready=0.679284s, total_serialization_initiated=0.254754s, others=0.000702s)
I0419 05:32:17.732952 135160546879296 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.939262s (all_items=0.000023s, per_item={'items': '0.00002265'}, temp_paths=0.939239)
I0419 05:32:17.733886 135036011521792 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started.
I0419 05:32:17.734045 135160546879296 async_checkpointer.py:561] Finished blocking save. Time taken: 1.674073s. 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_10_shardy_false/checkpoints/0.
I0419 05:32:17.781542 135160546879296 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0419 05:32:17.781895 135034929387264 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0419 05:32:17.782056 135160546879296 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_10_shardy_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776576736.0574548, 'wait_for_prev_duration_secs': 6.580352783203125e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776576736.0599422, 'checkpointer_blocking_duration_secs': 1.674210548400879, 'get_old_steps_start_time': 1776576737.7341723, 'get_old_steps_duration_secs': 2.6941299438476562e-05, 'checkpoint_manager_blocking_start_time': 1776576736.0547857, 'checkpoint_manager_blocking_duration_secs': 1.7272298336029053}
I0419 05:32:17.782168 135160546879296 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0419 05:32:17.782245 135160546879296 max_utils.py:750] 
Memstats: After params initialized:
I0419 05:32:17.782299 135160546879296 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0))
I0419 05:32:17.782333 135160546879296 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0))
I0419 05:32:17.782361 135160546879296 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0))
I0419 05:32:17.782386 135160546879296 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0))
I0419 05:32:18.070623 135160546879296 metric_logger.py:185] completed step: 0, seconds: 10.846, TFLOP/s/device: 1.253, Tokens/s/device: 188.834, total_weights: 65536, loss: 10.874
I0419 05:32:18.073492 135160546879296 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_10_shardy_false/tensorboard/'
I0419 05:32:18.473240 135160546879296 metric_logger.py:185] completed step: 1, seconds: 2.169, TFLOP/s/device: 6.265, Tokens/s/device: 944.352, total_weights: 65536, loss: 10.874
I0419 05:32:18.615335 135160546879296 metric_logger.py:185] completed step: 2, seconds: 0.404, TFLOP/s/device: 33.639, Tokens/s/device: 5070.386, total_weights: 65536, loss: 10.262
I0419 05:32:18.762946 135160546879296 metric_logger.py:185] completed step: 3, seconds: 0.011, TFLOP/s/device: 1229.048, Tokens/s/device: 185255.540, total_weights: 65536, loss: 9.732
I0419 05:32:39.455000 135035469350656 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776576738982680130, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776576738982680130, "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_10_shardy_false/checkpoints/0/_CHECKPOINT_METADATA
I0419 05:32:39.456332 135036045092608 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_10_shardy_false/checkpoints/0/items
I0419 05:32:39.456593 135160546879296 metric_logger.py:185] completed step: 4, seconds: 0.144, TFLOP/s/device: 94.309, Tokens/s/device: 14215.312, total_weights: 65536, loss: 9.274
I0419 05:32:39.467681 135160546879296 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.475, Tokens/s/device: 13938.895, total_weights: 65536, loss: 8.888
I0419 05:32:39.611011 135160546879296 metric_logger.py:185] completed step: 6, seconds: 20.694, TFLOP/s/device: 0.657, Tokens/s/device: 98.968, total_weights: 65536, loss: 8.589
I0419 05:32:39.758430 135160546879296 metric_logger.py:185] completed step: 7, seconds: 0.008, TFLOP/s/device: 1761.588, Tokens/s/device: 265525.736, total_weights: 65536, loss: 8.380
I0419 05:32:39.905985 135160546879296 metric_logger.py:185] completed step: 8, seconds: 0.145, TFLOP/s/device: 93.674, Tokens/s/device: 14119.561, total_weights: 65536, loss: 8.251
I0419 05:32:40.054371 135160546879296 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0419 05:32:40.054971 135160546879296 checkpointing.py:798] Waited 0.0006158351898193359 seconds for step 9 to finish before starting checkpointing.
I0419 05:32:40.056948 135160546879296 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0419 05:32:41.144272    2640 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0419 05:32:43.148681 135039803188992 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_10_shardy_false/checkpoints/0/items/array_metadatas/process_0
I0419 05:33:18.400890 135039794796288 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 37.275425s (commit=36.272003s, array_metadata_write=1.003422s)
I0419 05:33:18.402063 135036011521792 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 25.641 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 61.60388684272766 s) (per-host)
I0419 05:33:18.402214 135036011521792 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 60.668218s.
I0419 05:33:19.528573 135036011521792 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776576738982680130, '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_10_shardy_false/checkpoints/0/_CHECKPOINT_METADATA
I0419 05:33:20.263338 135035469350656 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776576738982680130, '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_10_shardy_false/checkpoints/0/_CHECKPOINT_METADATA
I0419 05:33:20.910912 135036011521792 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0419 05:33:21.078786 135036011521792 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000426s.
I0419 05:33:21.852969 135036011521792 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.200955s. 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_10_shardy_false/checkpoints/0/items
I0419 05:33:21.855066 135036011521792 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_10_shardy_false/checkpoints/0/items
I0419 05:33:22.270551 135036011521792 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_10_shardy_false/checkpoints/0
I0419 05:33:24.866799 135036011521792 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_10_shardy_false/checkpoints/0`.
I0419 05:33:24.867765 135036011521792 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 68.807797s. directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_10_shardy_false/checkpoints/0
I0419 05:33:24.869361 135036011521792 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 67.135365s.
I0419 05:33:24.869497 135034929387264 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0419 05:33:24.869594 135034929387264 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0419 05:33:24.869674 135034929387264 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0419 05:33:24.871652 135034929387264 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0419 05:33:24.871761 135160546879296 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 05:33:24.871845 135160546879296 checkpoint_manager.py:1458] Waiting for previous save to complete took 44.814907 seconds. If this number is high, consider checkpointing less frequently.
I0419 05:33:24.873868 135160546879296 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 9
I0419 05:33:24.877259 135160546879296 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_10_shardy_false/checkpoints/9.
I0419 05:33:26.253599 135036011521792 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_10_shardy_false/checkpoints/9
I0419 05:33:26.266890 135160546879296 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0419 05:33:26.266988 135160546879296 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 05:33:26.295535 135160546879296 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.030670s
I0419 05:33:26.296353 135160546879296 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.380 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2867393493652344 s) (per-host)
I0419 05:33:26.296429 135160546879296 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.286832s (batch_requests_ready=0.249669s, total_serialization_initiated=0.036429s, others=0.000734s)
I0419 05:33:26.296514 135160546879296 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.290593s (all_items=0.000016s, per_item={'items': '0.00001597'}, temp_paths=0.290577)
I0419 05:33:26.297377 135034946172672 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started.
I0419 05:33:26.297544 135160546879296 async_checkpointer.py:561] Finished blocking save. Time taken: 1.423609s. 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_10_shardy_false/checkpoints/9.
I0419 05:33:26.753503 135160546879296 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0419 05:33:26.753873 135034929387264 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0419 05:33:26.754060 135160546879296 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_10_shardy_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776576760.0569165, 'wait_for_prev_duration_secs': 44.81490707397461, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776576804.8739066, 'checkpointer_blocking_duration_secs': 1.4237444400787354, 'get_old_steps_start_time': 1776576806.2976718, 'get_old_steps_duration_secs': 2.5033950805664062e-05, 'checkpoint_manager_blocking_start_time': 1776576760.0552046, 'checkpoint_manager_blocking_duration_secs': 46.69881772994995}
I0419 05:33:26.754206 135160546879296 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0419 05:33:26.754255 135160546879296 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0419 05:33:29.297635 135039794796288 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_10_shardy_false/checkpoints/9/items
I0419 05:33:33.010008 135034974131968 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_10_shardy_false/checkpoints/9/items/array_metadatas/process_0
I0419 05:34:13.357138 135034965739264 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 42.472919s (commit=40.549330s, array_metadata_write=1.923590s)
I0419 05:34:13.358314 135034946172672 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 33.361 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 47.348668336868286 s) (per-host)
I0419 05:34:13.358442 135034946172672 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 47.060939s.
I0419 05:34:15.461117 135034946172672 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0419 05:34:15.682039 135034946172672 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000409s.
I0419 05:34:16.563605 135034946172672 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.923902s. 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_10_shardy_false/checkpoints/9/items
I0419 05:34:16.568586 135034946172672 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_10_shardy_false/checkpoints/9/items
I0419 05:34:16.987072 135034946172672 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_10_shardy_false/checkpoints/9
I0419 05:34:18.254913 135034946172672 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_10_shardy_false/checkpoints/9`.
I0419 05:34:18.255702 135034946172672 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 53.381771s. directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_10_shardy_false/checkpoints/9
I0419 05:34:18.257191 135034946172672 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 51.959688s.
I0419 05:34:18.257340 135034929387264 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0419 05:34:18.257399 135034929387264 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0419 05:34:18.257458 135034929387264 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0419 05:34:18.259848 135034929387264 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0419 05:34:18.260000 135160546879296 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 05:34:18.260159 135160546879296 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0419 05:34:18.260790 135160546879296 metric_logger.py:185] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.380, Tokens/s/device: 13924.584, total_weights: 65536, loss: 8.175
Per train step:
 Total TFLOPs: 13.59 
 split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Sun Apr 19 05:34:29 UTC 2026
EXIT_CODE=0