XPK Start: Sat Apr 18 01:47:48 UTC 2026
PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.
2026-04-18 01:48:12.690653: 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)
I0418 01:48:13.221203 135675897775936 max_utils.py:273] Attempting to initialize the jax distributed system...
I0418 01:48:22.261164 135675897775936 distributed.py:149] Starting JAX distributed service on [::]:8482
I0418 01:48:22.263503 135675897775936 distributed.py:172] Connecting to JAX distributed service on mt-16-zarr3-true-rs-2j9l4-slice-job-0-0.mt-16-zarr3-true-rs-2j9l4:8482
I0418 01:48:23.329012 135675897775936 max_utils.py:284] Jax distributed system initialized!
I0418 01:48:29.524505 135675897775936 max_utils.py:800] System Information: Jax Version: 0.9.2
I0418 01:48:29.524610 135675897775936 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0418 01:48:29.524649 135675897775936 max_utils.py:802] System Information: Jax Backend: PJRT C API
TFRT TPU v6 lite
Built on Mar 4 2026 11:32:08 (1772652728) cl/878335365
I0418 01:48:29.524682 135675897775936 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0418 01:48:30.220242 135675897775936 maxtext_utils.py:1551] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0418 01:48:30.220535 135675897775936 checkpointing.py:677] Setting up checkpoint logger...
I0418 01:48:30.220603 135675897775936 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0418 01:48:30.220657 135675897775936 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0418 01:48:30.221070 135675897775936 base_pytree_checkpoint_handler.py:441] 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 0x7b64c9f4cef0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0418 01:48:31.525645 135675897775936 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0418 01:48:31.526013 135675897775936 checkpoint_manager.py:708] [process=6][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b509c3812b0>}, handler_registry=None
I0418 01:48:31.526294 135675897775936 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b509c3812b0>` 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`.
I0418 01:48:31.526355 135675897775936 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b509c1881a0>` 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`.
I0418 01:48:31.526403 135675897775936 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 0x7b509c3812b0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b509c3812b0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b509c1881a0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b509c1881a0>}).
I0418 01:48:31.526727 135675897775936 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0418 01:48:31.526799 135675897775936 async_checkpointer.py:192] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7b50644a7c40> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0418 01:48:33.777238 135553647367936 checkpoint.py:228] Read Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776476739199733022, 'commit_timestamp_nsecs': 1776476790925027794, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_true/checkpoints/9/_CHECKPOINT_METADATA
I0418 01:48:34.018961 135553647367936 checkpoint_manager.py:1871] Missing metrics for step 9
I0418 01:48:34.455005 135675897775936 checkpoint_manager.py:1812] Found 3 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_true/checkpoints
I0418 01:48:35.714508 135675897775936 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=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/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7b509c349340>
I0418 01:48:35.714670 135675897775936 checkpointing.py:301] Checkpoint manager created!
I0418 01:48:36.647233 135675897775936 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0418 01:48:36.647346 135675897775936 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 01:48:37.029450 135675897775936 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 01:48:37.029541 135675897775936 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 01:48:37.046227 135675897775936 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 01:48:37.046288 135675897775936 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 01:48:37.070633 135675897775936 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0418 01:48:37.070701 135675897775936 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 01:48:37.087352 135675897775936 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0418 01:48:37.087414 135675897775936 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 01:48:37.103940 135675897775936 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0418 01:48:37.104000 135675897775936 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 01:48:37.129503 135675897775936 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0418 01:48:37.129584 135675897775936 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 01:48:37.151210 135675897775936 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0418 01:48:37.151278 135675897775936 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0418 01:48:37.373721 135675897775936 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0418 01:48:37.373847 135675897775936 checkpointing.py:581] restoring from this run's directory step 9
I0418 01:48:38.050241 135675897775936 event_tracking.py:70] [process=6] [sync] Started load checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_true/checkpoints/9.
I0418 01:48:38.717788 135675897775936 checkpointer.py:307] Restoring checkpoint from gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_true/checkpoints/9.
I0418 01:48:38.718230 135675897775936 event_tracking.py:125] [process=6] [sync] Finished blocking load in 0.67 seconds. Continuing load @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_true/checkpoints/9.
I0418 01:48:42.830661 1463 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0418 01:48:49.725716 135675897775936 jax_array_handlers.py:843] [process=6] /jax/orbax/read/worker/io/requested throughput: 228.975 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 6.901507139205933 s) (per-host)
I0418 01:48:49.740159 135675897775936 base_pytree_checkpoint_handler.py:130] [process=6] /jax/checkpoint/read/gbytes_per_sec: 203.526 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 7.775894403457642 s) (per-host)
I0418 01:48:51.219620 135675897775936 event_tracking.py:138] [process=6] [sync] Finished load in 13.17 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_true/checkpoints/9
I0418 01:48:51.219789 135675897775936 standard_logger.py:34] {'step': 9, 'event_type': 'restore', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_true/checkpoints', 'checkpointer_start_time': 1776476918.0501437, 'checkpointer_duration_secs': 13.169602870941162, 'checkpoint_manager_start_time': 1776476917.374092, 'checkpoint_manager_duration_secs': 13.845656871795654}
fsdp: 32
I0418 01:48:51.634935 135675897775936 maxtext_utils.py:1654] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: P('norm',)
Physical: (None,)
I0418 01:48:51.635062 135675897775936 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wi_0/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0418 01:48:51.635117 135675897775936 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wi_1/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0418 01:48:51.635174 135675897775936 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wo/kernel
Shape: float32[7168,16,2048]
Logical: P('mlp', 'layers', 'embed')
Physical: (None, None, 'fsdp')
I0418 01:48:51.635224 135675897775936 maxtext_utils.py:1654] params/params/decoder/layers/post_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0418 01:48:51.635261 135675897775936 maxtext_utils.py:1654] params/params/decoder/layers/pre_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0418 01:48:51.635311 135675897775936 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/key/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'kv_heads', 'kv_head_dim')
Physical: ('fsdp', None, None, None)
I0418 01:48:51.635363 135675897775936 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/out/kernel
Shape: float32[16,16,128,2048]
Logical: P('heads', 'layers', 'kv', 'embed')
Physical: (None, None, None, 'fsdp')
I0418 01:48:51.635402 135675897775936 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/query/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'q_heads', 'kv')
Physical: ('fsdp', None, None, None)
I0418 01:48:51.635437 135675897775936 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/value/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'kv_heads', 'kv_head_dim')
Physical: ('fsdp', None, None, None)
I0418 01:48:51.635483 135675897775936 maxtext_utils.py:1654] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: P('embed_vocab', 'vocab')
Physical: ('fsdp', None)
I0418 01:48:51.635526 135675897775936 maxtext_utils.py:1654] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: P('vocab', 'embed_vocab')
Physical: (None, 'fsdp')
I0418 01:48:52.123375 135675897775936 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0418 01:48:52.123470 135675897775936 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0418 01:48:52.138795 135675897775936 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0418 01:48:52.138871 135675897775936 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0418 01:49:03.141967 135675897775936 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.
I0418 01:49:03.142817 135675897775936 metric_logger.py:301] number parameters: 1.104 billion
I0418 01:49:14.222052 135675897775936 checkpointing.py:772] Waiting for step 10 to finish before checkpoint...
I0418 01:49:14.390367 135675897775936 checkpointing.py:776] Waited 0.16829466819763184 seconds for step 10 to finish before starting checkpointing.
I0418 01:49:14.392710 135675897775936 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 01:49:14.394330 135675897775936 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 10
I0418 01:49:14.395806 135675897775936 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_true/checkpoints/10.
I0418 01:49:15.573046 135675897775936 signaling_client.py:364] Using JaxDistributedSignalingClient
I0418 01:49:15.573975 135675897775936 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 01:49:15.574031 135675897775936 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
I0418 01:49:15.865772 135675897775936 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.292800s
I0418 01:49:15.865961 135675897775936 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 5.198 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.29677343368530273 s) (per-host)
I0418 01:49:15.866016 135675897775936 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.296839s (batch_requests_ready=0.002266s, total_serialization_initiated=0.294498s, others=0.000075s)
I0418 01:49:15.866109 135675897775936 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.300912s (all_items=0.000018s, per_item={'items': '0.00001788'}, temp_paths=0.300894)
I0418 01:49:15.866790 135675897775936 event_tracking.py:125] [process=6] [async] Finished blocking save in 1.47 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_true/checkpoints/10.
I0418 01:49:15.867149 135549788591872 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 02:09:15.867108
I0418 01:49:15.869127 135675897775936 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=10] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 01:49:15.869374 135553622189824 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0418 01:49:15.869475 135675897775936 standard_logger.py:34] {'step': 10, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776476954.3926914, 'wait_for_prev_duration_secs': 6.270408630371094e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776476954.394373, 'checkpointer_blocking_duration_secs': 1.4729316234588623, 'get_old_steps_start_time': 1776476955.8673296, 'get_old_steps_duration_secs': 3.24249267578125e-05, 'checkpoint_manager_blocking_start_time': 1776476954.3909109, 'checkpoint_manager_blocking_duration_secs': 1.4785287380218506}
I0418 01:49:15.869622 135675897775936 checkpointing.py:408] Started an asynchronous checkpoint save for step 10
I0418 01:49:15.869674 135675897775936 max_utils.py:750]
Memstats: After params initialized:
I0418 01:49:15.869728 135675897775936 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_24(process=6,(0,6,0,0))
I0418 01:49:15.869776 135675897775936 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_25(process=6,(1,6,0,0))
I0418 01:49:15.869812 135675897775936 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_28(process=6,(0,7,0,0))
I0418 01:49:15.869855 135675897775936 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_29(process=6,(1,7,0,0))
I0418 01:49:16.179268 135675897775936 metric_logger.py:196] completed step: 10, seconds: 11.079, TFLOP/s/device: 1.226, Tokens/s/device: 184.852, total_weights: 65536, loss: 6.833, lm_loss: 6.833, perplexity: 927.869
I0418 01:49:16.351576 135675897775936 metric_logger.py:196] completed step: 11, seconds: 1.956, TFLOP/s/device: 6.948, Tokens/s/device: 1047.211, total_weights: 65536, loss: 6.527, lm_loss: 6.527, perplexity: 683.659
I0418 01:49:16.499124 135675897775936 metric_logger.py:196] completed step: 12, seconds: 0.025, TFLOP/s/device: 546.260, Tokens/s/device: 82338.278, total_weights: 65536, loss: 6.265, lm_loss: 6.265, perplexity: 526.068
I0418 01:49:16.646515 135675897775936 metric_logger.py:196] completed step: 13, seconds: 0.154, TFLOP/s/device: 88.506, Tokens/s/device: 13340.629, total_weights: 65536, loss: 6.045, lm_loss: 6.045, perplexity: 422.001
I0418 01:49:16.650986 135675897775936 checkpointing.py:772] Waiting for step 15 to finish before checkpoint...
I0418 01:49:16.940900 135675897775936 checkpointing.py:776] Waited 0.28987908363342285 seconds for step 15 to finish before starting checkpointing.
I0418 01:49:16.943929 135675897775936 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=10][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 01:49:23.695484 135549796984576 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_true/checkpoints/10/items/array_metadatas/process_6
I0418 01:49:59.954588 135549788591872 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 35.588 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 44.3853554725647 s) (per-host)
I0418 01:49:59.954713 135549788591872 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 44.087446s.
I0418 01:50:08.420297 135549788591872 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 52.553012s.
I0418 01:50:08.420533 135553622189824 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 01:50:08.420601 135553622189824 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0418 01:50:08.420661 135553622189824 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=10] CheckpointManager Save Finalize is syncing with other hosts...
I0418 01:50:08.422115 135553622189824 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=10] CheckpointManager Save Finalize is done on all hosts.
I0418 01:50:08.422277 135675897775936 checkpoint_manager.py:2032] [process=6][thread=MainThread][step=10][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=10.
W0418 01:50:08.422411 135675897775936 checkpoint_manager.py:1452] Waiting for previous save to complete took 51.478480 seconds. If this number is high, consider checkpointing less frequently.
I0418 01:50:08.424332 135675897775936 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 15
I0418 01:50:08.426784 135675897775936 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_true/checkpoints/15.
I0418 01:50:09.560005 135675897775936 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 01:50:09.560105 135675897775936 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
I0418 01:50:09.586717 135675897775936 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.028716s
I0418 01:50:09.586870 135675897775936 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 47.889 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.0322115421295166 s) (per-host)
I0418 01:50:09.586924 135675897775936 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.032274s (batch_requests_ready=0.001878s, total_serialization_initiated=0.030325s, others=0.000071s)
I0418 01:50:09.587014 135675897775936 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.037196s (all_items=0.000012s, per_item={'items': '0.00001240'}, temp_paths=0.037184)
I0418 01:50:09.587730 135675897775936 event_tracking.py:125] [process=6] [async] Finished blocking save in 1.16 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_true/checkpoints/15.
I0418 01:50:09.588064 135553622189824 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 02:10:09.588027
I0418 01:50:09.596453 135675897775936 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=15] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 01:50:09.596756 135549796984576 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0418 01:50:09.596911 135675897775936 standard_logger.py:34] {'step': 15, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776476956.9439006, 'wait_for_prev_duration_secs': 51.478479623794556, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776477008.4243724, 'checkpointer_blocking_duration_secs': 1.1638274192810059, 'get_old_steps_start_time': 1776477009.588222, 'get_old_steps_duration_secs': 3.147125244140625e-05, 'checkpoint_manager_blocking_start_time': 1776476956.9414966, 'checkpoint_manager_blocking_duration_secs': 52.65537977218628}
I0418 01:50:09.597084 135675897775936 checkpointing.py:408] Started an asynchronous checkpoint save for step 15
I0418 01:50:09.597946 135675897775936 metric_logger.py:196] completed step: 14, seconds: 0.147, TFLOP/s/device: 92.135, Tokens/s/device: 13887.665, total_weights: 65536, loss: 5.864, lm_loss: 5.864, perplexity: 352.193
I0418 01:50:09.603744 135675897775936 metric_logger.py:196] completed step: 15, seconds: 0.147, TFLOP/s/device: 92.243, Tokens/s/device: 13903.881, total_weights: 65536, loss: 5.719, lm_loss: 5.719, perplexity: 304.661
I0418 01:50:17.332816 135547666286336 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_true/checkpoints/15/items/array_metadatas/process_6
I0418 01:50:31.000965 135675897775936 metric_logger.py:196] completed step: 16, seconds: 52.951, TFLOP/s/device: 0.257, Tokens/s/device: 38.677, total_weights: 65536, loss: 5.605, lm_loss: 5.605, perplexity: 271.872
I0418 01:50:31.148300 135675897775936 metric_logger.py:196] completed step: 17, seconds: 21.251, TFLOP/s/device: 0.639, Tokens/s/device: 96.372, total_weights: 65536, loss: 5.517, lm_loss: 5.517, perplexity: 248.975
I0418 01:50:31.295497 135675897775936 metric_logger.py:196] completed step: 18, seconds: 0.152, TFLOP/s/device: 89.212, Tokens/s/device: 13446.967, total_weights: 65536, loss: 5.448, lm_loss: 5.448, perplexity: 232.333
I0418 01:50:31.442163 135675897775936 checkpointing.py:772] Waiting for step 19 to finish before checkpoint...
I0418 01:50:31.442865 135675897775936 checkpointing.py:776] Waited 0.0007207393646240234 seconds for step 19 to finish before starting checkpointing.
I0418 01:50:31.445136 135675897775936 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=15][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 01:50:52.992868 135553622189824 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 36.364 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 43.43815565109253 s) (per-host)
I0418 01:50:52.992986 135553622189824 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 43.404812s.
I0418 01:51:03.179229 135553622189824 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 53.591041s.
I0418 01:51:03.179549 135549796984576 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 01:51:03.179691 135549796984576 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0418 01:51:03.179740 135549796984576 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=15] CheckpointManager Save Finalize is syncing with other hosts...
I0418 01:51:03.181170 135549796984576 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=15] CheckpointManager Save Finalize is done on all hosts.
I0418 01:51:03.181363 135675897775936 checkpoint_manager.py:2032] [process=6][thread=MainThread][step=15][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=15.
W0418 01:51:03.181486 135675897775936 checkpoint_manager.py:1452] Waiting for previous save to complete took 31.736355 seconds. If this number is high, consider checkpointing less frequently.
I0418 01:51:03.182879 135675897775936 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 19
I0418 01:51:03.184940 135675897775936 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_true/checkpoints/19.
I0418 01:51:04.315107 135675897775936 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 01:51:04.315203 135675897775936 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
I0418 01:51:04.341017 135675897775936 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.026745s
I0418 01:51:04.341202 135675897775936 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 50.914 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.030297517776489258 s) (per-host)
I0418 01:51:04.341262 135675897775936 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.030370s (batch_requests_ready=0.001754s, total_serialization_initiated=0.028537s, others=0.000080s)
I0418 01:51:04.341359 135675897775936 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.034436s (all_items=0.000015s, per_item={'items': '0.00001526'}, temp_paths=0.034420)
I0418 01:51:04.341997 135675897775936 event_tracking.py:125] [process=6] [async] Finished blocking save in 1.16 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_true/checkpoints/19.
I0418 01:51:04.342290 135549796984576 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 02:11:04.342253
I0418 01:51:04.345120 135675897775936 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=19] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 01:51:04.345358 135547666286336 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0418 01:51:04.345482 135675897775936 standard_logger.py:34] {'step': 19, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776477031.4451075, 'wait_for_prev_duration_secs': 31.73635482788086, 'time_between_consecutive_saves_sec': 23.022947788238525, 'checkpointer_blocking_start_time': 1776477063.1829195, 'checkpointer_blocking_duration_secs': 1.1595194339752197, 'get_old_steps_start_time': 1776477064.342453, 'get_old_steps_duration_secs': 2.1457672119140625e-05, 'checkpoint_manager_blocking_start_time': 1776477031.4430988, 'checkpoint_manager_blocking_duration_secs': 32.90235376358032}
I0418 01:51:04.345587 135675897775936 checkpointing.py:408] Started an asynchronous checkpoint save for step 19
I0418 01:51:04.345628 135675897775936 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=19][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 01:51:12.252116 135547674679040 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_true/checkpoints/19/items/array_metadatas/process_6
I0418 01:51:47.768512 135549796984576 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 36.348 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 43.45757055282593 s) (per-host)
I0418 01:51:47.768631 135549796984576 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 43.426226s.
I0418 01:51:58.757496 135549796984576 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 54.415076s.
I0418 01:51:58.757716 135547666286336 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 01:51:58.757781 135547666286336 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0418 01:51:58.757867 135547666286336 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=19] CheckpointManager Save Finalize is syncing with other hosts...
I0418 01:51:58.759723 135547666286336 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=19] CheckpointManager Save Finalize is done on all hosts.
I0418 01:51:58.759910 135675897775936 checkpoint_manager.py:2032] [process=6][thread=MainThread][step=19][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=19.
I0418 01:51:58.760068 135675897775936 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 01:51:58.760960 135675897775936 metric_logger.py:196] completed step: 19, seconds: 0.147, TFLOP/s/device: 92.251, Tokens/s/device: 13905.108, total_weights: 65536, loss: 5.391, lm_loss: 5.391, perplexity: 219.349
Per train step:
Total TFLOPs: 13.59
split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Sat Apr 18 01:52:10 UTC 2026
EXIT_CODE=0