MaxView

← Back to run

Log Summary

XPK Start: Thu Apr 23 18:14:12 UTC 2026
2026-04-23 18:14:17.134281: 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 00:00:1776968057.147637      10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
E0000 00:00:1776968057.151406      10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
W0000 00:00:1776968057.163103      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776968057.163122      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776968057.163124      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776968057.163126      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
2026-04-23 18:14:36.388569: 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)
I0423 18:14:36.916830 134700917843776 max_utils.py:273] Attempting to initialize the jax distributed system...
INFO:2026-04-23 18:14:45,957:jax._src.distributed:140: Starting JAX distributed service on [::]:8482
I0423 18:14:45.957949 134700917843776 distributed.py:140] Starting JAX distributed service on [::]:8482
INFO:2026-04-23 18:14:45,960:jax._src.distributed:157: Connecting to JAX distributed service on mt-13-scan-layers-true-vm0ng-slice-job-0-0.mt-13-scan-layers-true-vm0ng:8482
I0423 18:14:45.960436 134700917843776 distributed.py:157] Connecting to JAX distributed service on mt-13-scan-layers-true-vm0ng-slice-job-0-0.mt-13-scan-layers-true-vm0ng:8482
I0423 18:14:46.830381 134700917843776 max_utils.py:284] Jax distributed system initialized!
I0423 18:14:53.684936 134700917843776 max_utils.py:800] System Information: Jax Version: 0.8.1
I0423 18:14:53.685042 134700917843776 max_utils.py:801] System Information: Jaxlib Version: 0.8.1
I0423 18:14:53.685083 134700917843776 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
I0423 18:14:53.685134 134700917843776 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0423 18:14:54.311207 134700917843776 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0423 18:14:54.425784 134700917843776 checkpointing.py:688] Setting up checkpoint logger...
I0423 18:14:54.425902 134700917843776 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0423 18:14:54.425947 134700917843776 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None
I0423 18:14:54.426169 134700917843776 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 0x7a81ecf70f20>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0423 18:14:57.256494 134700917843776 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0423 18:14:57.256906 134700917843776 checkpoint_manager.py:709] [process=6][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7a6d6404c260>}, handler_registry=None
I0423 18:14:57.257164 134700917843776 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7a6d6404c260>` 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`.
I0423 18:14:57.257217 134700917843776 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7a6d6442ffb0>` 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`.
I0423 18:14:57.257254 134700917843776 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 0x7a6d6404c260>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7a6d6404c260>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7a6d6442ffb0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7a6d6442ffb0>}).
I0423 18:14:57.257598 134700917843776 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33
I0423 18:14:57.257678 134700917843776 async_checkpointer.py:177] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7a6d4c0ecf40> timeout: 600 secs and primary_host=0 for async checkpoint writes
I0423 18:14:58.427401 134700917843776 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_13_scan_layers_true/checkpoints
I0423 18:14:58.429632 134700917843776 checkpoint_manager.py:929] [process=6][thread=MainThread] CheckpointManager created,  primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=10), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_13_scan_layers_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7a6d4c11ddf0>
I0423 18:14:58.429748 134700917843776 checkpointing.py:302] Checkpoint manager created!
I0423 18:14:59.384504 134700917843776 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0423 18:14:59.384628 134700917843776 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0423 18:15:00.851138 134700917843776 maxtext_utils.py:1749]  decoder/decoder_norm/scale/value
    Shape:     float32[2048]
    Physical:  (None,)
I0423 18:15:00.851264 134700917843776 maxtext_utils.py:1749]  decoder/layers/mlp/wi_0/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0423 18:15:00.851320 134700917843776 maxtext_utils.py:1749]  decoder/layers/mlp/wi_1/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0423 18:15:00.851365 134700917843776 maxtext_utils.py:1749]  decoder/layers/mlp/wo/kernel/value
    Shape:     float32[7168,16,2048]
    Physical:  (None, None, 'fsdp')
I0423 18:15:00.851402 134700917843776 maxtext_utils.py:1749]  decoder/layers/post_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0423 18:15:00.851440 134700917843776 maxtext_utils.py:1749]  decoder/layers/pre_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0423 18:15:00.851478 134700917843776 maxtext_utils.py:1749]  decoder/layers/self_attention/key/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0423 18:15:00.851514 134700917843776 maxtext_utils.py:1749]  decoder/layers/self_attention/out/kernel/value
    Shape:     float32[16,16,128,2048]
    Physical:  (None, None, None, 'fsdp')
I0423 18:15:00.851550 134700917843776 maxtext_utils.py:1749]  decoder/layers/self_attention/query/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0423 18:15:00.851584 134700917843776 maxtext_utils.py:1749]  decoder/layers/self_attention/value/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0423 18:15:00.851619 134700917843776 maxtext_utils.py:1749]  decoder/logits_dense/kernel/value
    Shape:     float32[2048,32000]
    Physical:  ('fsdp', None)

I0423 18:15:00.851654 134700917843776 maxtext_utils.py:1749]  token_embedder/embedding/value
    Shape:     float32[32000,2048]
    Physical:  (None, 'fsdp')
I0423 18:15:01.252937 134700917843776 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0423 18:15:01.253196 134700917843776 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0423 18:15:01.258804 134700917843776 nnx_decoders.py:616] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0423 18:15:01.258905 134700917843776 nnx_decoders.py:616] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0423 18:15:01.418943 134700917843776 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0423 18:15:01.419078 134700917843776 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0423 18:15:01.582207 134700917843776 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0423 18:15:01.582344 134700917843776 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0423 18:15:01.749222 134700917843776 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0423 18:15:01.749366 134700917843776 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 18:15:01.906565 134700917843776 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0423 18:15:01.906698 134700917843776 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 18:15:02.063860 134700917843776 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0423 18:15:02.063995 134700917843776 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 18:15:02.231712 134700917843776 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0423 18:15:02.231853 134700917843776 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0423 18:15:02.255007 134700917843776 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0423 18:15:02.255128 134700917843776 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0423 18:15:13.423973 134700917843776 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.
I0423 18:15:13.426464 134700917843776 metric_logger.py:289] number parameters: 1.104 billion
I0423 18:15:24.458902 134700917843776 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0423 18:15:24.612141 134700917843776 checkpointing.py:798] Waited 0.15320205688476562 seconds for step 0 to finish before starting checkpointing.
I0423 18:15:24.614410 134700917843776 checkpoint_manager.py:2013] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0423 18:15:24.616199 134700917843776 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 0
I0423 18:15:24.619186 134700917843776 async_checkpointer.py:452] [process=6] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_13_scan_layers_true/checkpoints/0.
I0423 18:15:25.243721 134700917843776 signaling_client.py:364] Using JaxDistributedSignalingClient
I0423 18:15:25.244796 134700917843776 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0423 18:15:25.244856 134700917843776 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
I0423 18:15:25.492253 134700917843776 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.249663s
I0423 18:15:25.492428 134700917843776 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 3.020 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5107777118682861 s) (per-host)
I0423 18:15:25.492482 134700917843776 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.510842s (batch_requests_ready=0.254587s, total_serialization_initiated=0.256182s, others=0.000073s)
I0423 18:15:25.492587 134700917843776 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.516163s (all_items=0.000026s, per_item={'items': '0.00002599'}, temp_paths=0.516137)
I0423 18:15:25.493561 134575057520384 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started.
I0423 18:15:25.493723 134700917843776 async_checkpointer.py:561] Finished blocking save. Time taken: 0.877453s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_13_scan_layers_true/checkpoints/0.
I0423 18:15:25.571800 134700917843776 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0423 18:15:25.572201 134573472085760 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0423 18:15:25.572368 134700917843776 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_13_scan_layers_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776968124.6143909, 'wait_for_prev_duration_secs': 7.295608520507812e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776968124.6162376, 'checkpointer_blocking_duration_secs': 0.8775932788848877, 'get_old_steps_start_time': 1776968125.4938526, 'get_old_steps_duration_secs': 2.5033950805664062e-05, 'checkpoint_manager_blocking_start_time': 1776968124.6125512, 'checkpoint_manager_blocking_duration_secs': 0.9597744941711426}
I0423 18:15:25.572476 134700917843776 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0423 18:15:25.572551 134700917843776 max_utils.py:750] 
Memstats: After params initialized:
I0423 18:15:25.572610 134700917843776 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_24(process=6,(0,6,0,0))
I0423 18:15:25.572645 134700917843776 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_25(process=6,(1,6,0,0))
I0423 18:15:25.572674 134700917843776 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_28(process=6,(0,7,0,0))
I0423 18:15:25.572700 134700917843776 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_29(process=6,(1,7,0,0))
I0423 18:15:25.842266 134700917843776 metric_logger.py:185] completed step: 0, seconds: 11.031, TFLOP/s/device: 1.232, Tokens/s/device: 185.652, total_weights: 65536, loss: 10.874
I0423 18:15:26.014051 134700917843776 metric_logger.py:185] completed step: 1, seconds: 1.382, TFLOP/s/device: 9.834, Tokens/s/device: 1482.293, total_weights: 65536, loss: 10.874
I0423 18:15:26.413988 134700917843776 metric_logger.py:185] completed step: 2, seconds: 0.026, TFLOP/s/device: 521.179, Tokens/s/device: 78557.729, total_weights: 65536, loss: 10.262
I0423 18:15:26.561539 134700917843776 metric_logger.py:185] completed step: 3, seconds: 0.391, TFLOP/s/device: 34.780, Tokens/s/device: 5242.437, total_weights: 65536, loss: 9.732
I0423 18:15:26.857260 134700917843776 metric_logger.py:185] completed step: 4, seconds: 0.163, TFLOP/s/device: 83.376, Tokens/s/device: 12567.347, total_weights: 65536, loss: 9.274
I0423 18:15:26.864430 134700917843776 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.357, Tokens/s/device: 13921.082, total_weights: 65536, loss: 8.888
I0423 18:15:45.617841 134700917843776 metric_logger.py:185] completed step: 6, seconds: 0.296, TFLOP/s/device: 45.880, Tokens/s/device: 6915.461, total_weights: 65536, loss: 8.589
I0423 18:15:45.765054 134700917843776 metric_logger.py:185] completed step: 7, seconds: 18.607, TFLOP/s/device: 0.730, Tokens/s/device: 110.064, total_weights: 65536, loss: 8.380
I0423 18:15:45.912502 134700917843776 metric_logger.py:185] completed step: 8, seconds: 0.153, TFLOP/s/device: 88.662, Tokens/s/device: 13364.133, total_weights: 65536, loss: 8.251
I0423 18:15:46.060920 134700917843776 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0423 18:15:46.061541 134700917843776 checkpointing.py:798] Waited 0.0006310939788818359 seconds for step 9 to finish before starting checkpointing.
I0423 18:15:46.063571 134700917843776 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0423 18:15:46.777473    2592 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0423 18:15:48.738393 134575065913088 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_13_scan_layers_true/checkpoints/0/items/array_metadatas/process_6
I0423 18:16:21.198687 134575057520384 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 28.098 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 56.21701383590698 s) (per-host)
I0423 18:16:21.198790 134575057520384 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 55.705121s.
I0423 18:16:30.032023 134575057520384 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 64.538336s.
I0423 18:16:30.032361 134573472085760 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0423 18:16:30.032498 134573472085760 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0423 18:16:30.032547 134573472085760 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0423 18:16:30.034118 134573472085760 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0423 18:16:30.034330 134700917843776 checkpoint_manager.py:2036] [process=6][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0423 18:16:30.034469 134700917843776 checkpoint_manager.py:1458] Waiting for previous save to complete took 43.970905 seconds. If this number is high, consider checkpointing less frequently.
I0423 18:16:30.036383 134700917843776 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 9
I0423 18:16:30.039986 134700917843776 async_checkpointer.py:452] [process=6] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_13_scan_layers_true/checkpoints/9.
I0423 18:16:30.697887 134700917843776 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0423 18:16:30.698044 134700917843776 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
I0423 18:16:30.739315 134700917843776 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.043432s
I0423 18:16:30.739481 134700917843776 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.023 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.3071174621582031 s) (per-host)
I0423 18:16:30.739532 134700917843776 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.307177s (batch_requests_ready=0.259006s, total_serialization_initiated=0.048105s, others=0.000067s)
I0423 18:16:30.739632 134700917843776 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.311183s (all_items=0.000015s, per_item={'items': '0.00001502'}, temp_paths=0.311168)
I0423 18:16:30.740512 134574504670976 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started.
I0423 18:16:30.740666 134700917843776 async_checkpointer.py:561] Finished blocking save. Time taken: 0.704212s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_13_scan_layers_true/checkpoints/9.
I0423 18:16:30.751767 134700917843776 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0423 18:16:30.752007 134573472085760 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0423 18:16:30.752149 134700917843776 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_13_scan_layers_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776968146.0635407, 'wait_for_prev_duration_secs': 43.97090458869934, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776968190.0364227, 'checkpointer_blocking_duration_secs': 0.7043488025665283, 'get_old_steps_start_time': 1776968190.7407906, 'get_old_steps_duration_secs': 2.5987625122070312e-05, 'checkpoint_manager_blocking_start_time': 1776968146.0617635, 'checkpoint_manager_blocking_duration_secs': 44.690351247787476}
I0423 18:16:30.752261 134700917843776 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0423 18:16:30.752307 134700917843776 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0423 18:16:35.735971 134575065913088 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_13_scan_layers_true/checkpoints/9/items/array_metadatas/process_6
I0423 18:17:12.829076 134574504670976 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 37.258 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.39667224884033 s) (per-host)
I0423 18:17:12.829228 134574504670976 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.088607s.
I0423 18:17:21.302185 134574504670976 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 50.561549s.
I0423 18:17:21.302397 134573472085760 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0423 18:17:21.302461 134573472085760 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0423 18:17:21.302512 134573472085760 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0423 18:17:21.303968 134573472085760 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0423 18:17:21.304137 134700917843776 checkpoint_manager.py:2036] [process=6][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0423 18:17:21.304305 134700917843776 checkpoint_manager.py:2013] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0423 18:17:21.304903 134700917843776 metric_logger.py:185] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.486, Tokens/s/device: 13940.508, 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: Thu Apr 23 18:17:33 UTC 2026
EXIT_CODE=0