XPK Start: Sat Apr 18 14:13:32 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 14:13:56.976397: 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 14:13:57.513772 136446713710400 max_utils.py:273] Attempting to initialize the jax distributed system...
I0418 14:14:06.554419 136446713710400 distributed.py:149] Starting JAX distributed service on [::]:8482
I0418 14:14:06.556769 136446713710400 distributed.py:172] Connecting to JAX distributed service on mt-15-ocdbt-true-rs-npvxf-slice-job-0-0.mt-15-ocdbt-true-rs-npvxf:8482
I0418 14:14:07.748320 136446713710400 max_utils.py:284] Jax distributed system initialized!
I0418 14:14:13.891182 136446713710400 max_utils.py:800] System Information: Jax Version: 0.9.2
I0418 14:14:13.891285 136446713710400 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0418 14:14:13.891325 136446713710400 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 14:14:13.891362 136446713710400 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0418 14:14:14.591131 136446713710400 maxtext_utils.py:1551] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0418 14:14:14.591403 136446713710400 checkpointing.py:677] Setting up checkpoint logger...
I0418 14:14:14.591457 136446713710400 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0418 14:14:14.591501 136446713710400 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0418 14:14:14.591834 136446713710400 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 0x7c18571181d0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0418 14:14:16.723466 136446713710400 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0418 14:14:16.723700 136446713710400 checkpoint_manager.py:708] [process=7][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c1735cb6330>}, handler_registry=None
I0418 14:14:16.723938 136446713710400 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c1735cb6330>` 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 14:14:16.723986 136446713710400 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c04446c5790>` 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 14:14:16.724022 136446713710400 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 0x7c1735cb6330>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c1735cb6330>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c04446c5790>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c04446c5790>}).
I0418 14:14:16.724343 136446713710400 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0418 14:14:16.724412 136446713710400 async_checkpointer.py:192] [process=7][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7c0354777c40> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0418 14:14:19.504303 136326523369216 checkpoint.py:228] Read Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776521436094543513, 'commit_timestamp_nsecs': 1776521484505042421, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_true/checkpoints/5/_CHECKPOINT_METADATA
I0418 14:14:20.161530 136326531761920 checkpoint_manager.py:1871] Missing metrics for step 0
I0418 14:14:20.601891 136446713710400 checkpoint_manager.py:1812] Found 3 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_true/checkpoints
I0418 14:14:21.535869 136446713710400 checkpoint_manager.py:929] [process=7][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_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7c035469e000>
I0418 14:14:21.536040 136446713710400 checkpointing.py:301] Checkpoint manager created!
I0418 14:14:22.460649 136446713710400 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0418 14:14:22.460764 136446713710400 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 14:14:22.841155 136446713710400 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 14:14:22.841244 136446713710400 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 14:14:22.857630 136446713710400 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 14:14:22.857690 136446713710400 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 14:14:22.882058 136446713710400 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 14:14:22.882142 136446713710400 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 14:14:22.898506 136446713710400 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 14:14:22.898564 136446713710400 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 14:14:22.914967 136446713710400 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 14:14:22.915025 136446713710400 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 14:14:22.939547 136446713710400 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0418 14:14:22.939618 136446713710400 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 14:14:22.960680 136446713710400 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0418 14:14:22.960745 136446713710400 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0418 14:14:23.175992 136446713710400 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0418 14:14:23.176110 136446713710400 checkpointing.py:581] restoring from this run's directory step 9
I0418 14:14:23.426484 136446713710400 event_tracking.py:70] [process=7] [sync] Started load checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_true/checkpoints/9.
I0418 14:14:23.670660 136446713710400 checkpointer.py:307] Restoring checkpoint from gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_true/checkpoints/9.
I0418 14:14:23.671105 136446713710400 event_tracking.py:125] [process=7] [sync] Finished blocking load in 0.24 seconds. Continuing load @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_true/checkpoints/9.
I0418 14:14:27.801373 1440 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0418 14:14:34.058824 136446713710400 jax_array_handlers.py:843] [process=7] /jax/orbax/read/worker/io/requested throughput: 252.311 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 6.263200044631958 s) (per-host)
I0418 14:14:34.064252 136446713710400 base_pytree_checkpoint_handler.py:130] [process=7] /jax/checkpoint/read/gbytes_per_sec: 222.944 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 7.098621368408203 s) (per-host)
I0418 14:14:34.221785 136446713710400 event_tracking.py:138] [process=7] [sync] Finished load in 10.80 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_true/checkpoints/9
I0418 14:14:34.221943 136446713710400 standard_logger.py:34] {'step': 9, 'event_type': 'restore', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_true/checkpoints', 'checkpointer_start_time': 1776521663.4263885, 'checkpointer_duration_secs': 10.795521974563599, 'checkpoint_manager_start_time': 1776521663.176356, 'checkpoint_manager_duration_secs': 11.045556545257568}
fsdp: 32
I0418 14:14:34.638776 136446713710400 maxtext_utils.py:1654] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: P('norm',)
Physical: (None,)
I0418 14:14:34.638898 136446713710400 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 14:14:34.638956 136446713710400 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 14:14:34.639014 136446713710400 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 14:14:34.639069 136446713710400 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 14:14:34.639123 136446713710400 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 14:14:34.639175 136446713710400 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 14:14:34.639226 136446713710400 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 14:14:34.639264 136446713710400 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 14:14:34.639300 136446713710400 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 14:14:34.639345 136446713710400 maxtext_utils.py:1654] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: P('embed_vocab', 'vocab')
Physical: ('fsdp', None)
I0418 14:14:34.639390 136446713710400 maxtext_utils.py:1654] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: P('vocab', 'embed_vocab')
Physical: (None, 'fsdp')
I0418 14:14:35.128635 136446713710400 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0418 14:14:35.128732 136446713710400 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0418 14:14:35.144284 136446713710400 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0418 14:14:35.144344 136446713710400 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0418 14:14:46.161429 136446713710400 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 14:14:46.162284 136446713710400 metric_logger.py:301] number parameters: 1.104 billion
I0418 14:14:57.346833 136446713710400 checkpointing.py:772] Waiting for step 10 to finish before checkpoint...
I0418 14:14:57.503490 136446713710400 checkpointing.py:776] Waited 0.15663409233093262 seconds for step 10 to finish before starting checkpointing.
I0418 14:14:57.506098 136446713710400 checkpoint_manager.py:2009] [process=7][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 14:14:57.507978 136446713710400 checkpoint_manager.py:1512] [process=7] Saving checkpoint at step 10
I0418 14:14:57.509894 136446713710400 event_tracking.py:70] [process=7] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_true/checkpoints/10.
I0418 14:14:58.235618 136446713710400 signaling_client.py:364] Using JaxDistributedSignalingClient
I0418 14:14:58.236604 136446713710400 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 14:14:58.236660 136446713710400 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 14:14:58.515199 136446713710400 base_pytree_checkpoint_handler.py:154] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.279658s
I0418 14:14:58.515380 136446713710400 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/blocking_gbytes_per_sec: 5.437 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.28372645378112793 s) (per-host)
I0418 14:14:58.515435 136446713710400 base_pytree_checkpoint_handler.py:768] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.283793s (batch_requests_ready=0.002134s, total_serialization_initiated=0.281582s, others=0.000077s)
I0418 14:14:58.515523 136446713710400 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.287924s (all_items=0.000018s, per_item={'items': '0.00001788'}, temp_paths=0.287906)
I0418 14:14:58.516261 136446713710400 event_tracking.py:125] [process=7] [async] Finished blocking save in 1.01 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_true/checkpoints/10.
I0418 14:14:58.516614 136325936174848 async_checkpointer.py:76] [process=7][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 14:34:58.516576
I0418 14:14:58.521193 136446713710400 checkpoint_manager.py:1560] [process=7][thread=MainThread][step=10] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 14:14:58.521447 136322146748160 async_checkpointer.py:280] [process=7][thread=save_finalize] Waiting for background save thread=async_save.
I0418 14:14:58.521599 136446713710400 standard_logger.py:34] {'step': 10, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776521697.5060587, 'wait_for_prev_duration_secs': 0.00010371208190917969, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776521697.5080361, 'checkpointer_blocking_duration_secs': 1.008737325668335, 'get_old_steps_start_time': 1776521698.516797, 'get_old_steps_duration_secs': 3.170967102050781e-05, 'checkpoint_manager_blocking_start_time': 1776521697.5041199, 'checkpoint_manager_blocking_duration_secs': 1.0174458026885986}
I0418 14:14:58.521702 136446713710400 checkpointing.py:408] Started an asynchronous checkpoint save for step 10
I0418 14:14:58.521752 136446713710400 max_utils.py:750]
Memstats: After params initialized:
I0418 14:14:58.521799 136446713710400 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_26(process=7,(2,6,0,0))
I0418 14:14:58.521831 136446713710400 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_27(process=7,(3,6,0,0))
I0418 14:14:58.521857 136446713710400 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_30(process=7,(2,7,0,0))
I0418 14:14:58.521881 136446713710400 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_31(process=7,(3,7,0,0))
I0418 14:14:58.831081 136446713710400 metric_logger.py:196] completed step: 10, seconds: 11.184, TFLOP/s/device: 1.215, Tokens/s/device: 183.112, total_weights: 65536, loss: 6.833, lm_loss: 6.833, perplexity: 927.869
I0418 14:14:59.009990 136446713710400 metric_logger.py:196] completed step: 11, seconds: 1.483, TFLOP/s/device: 9.164, Tokens/s/device: 1381.324, total_weights: 65536, loss: 6.527, lm_loss: 6.527, perplexity: 683.659
I0418 14:14:59.157433 136446713710400 metric_logger.py:196] completed step: 12, seconds: 0.032, TFLOP/s/device: 430.095, Tokens/s/device: 64828.590, total_weights: 65536, loss: 6.265, lm_loss: 6.265, perplexity: 526.068
I0418 14:14:59.304825 136446713710400 metric_logger.py:196] completed step: 13, seconds: 0.154, TFLOP/s/device: 88.500, Tokens/s/device: 13339.673, total_weights: 65536, loss: 6.045, lm_loss: 6.045, perplexity: 422.001
I0418 14:14:59.309477 136446713710400 checkpointing.py:772] Waiting for step 15 to finish before checkpoint...
I0418 14:14:59.599565 136446713710400 checkpointing.py:776] Waited 0.2900557518005371 seconds for step 15 to finish before starting checkpointing.
I0418 14:14:59.602579 136446713710400 checkpoint_manager.py:2020] [process=7][thread=MainThread][step=10][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 14:15:04.504022 136322155140864 array_metadata_store.py:203] [process=7][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_true/checkpoints/10/items/array_metadatas/process_7
I0418 14:15:39.821180 136325936174848 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/gbytes_per_sec: 37.981 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.58948993682861 s) (per-host)
I0418 14:15:39.821301 136325936174848 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.304573s.
I0418 14:15:48.142177 136325936174848 async_checkpointer.py:160] [process=7][thread=async_save] Background save thread done. Time taken: 49.625432s.
I0418 14:15:48.142446 136322146748160 async_checkpointer.py:288] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 14:15:48.142570 136322146748160 async_checkpointer.py:298] [process=7][thread=save_finalize] No errors found in background save thread=async_save.
I0418 14:15:48.142620 136322146748160 checkpoint_manager.py:2137] [process=7][thread=save_finalize][step=10] CheckpointManager Save Finalize is syncing with other hosts...
I0418 14:15:48.145181 136322146748160 checkpoint_manager.py:2146] [process=7][thread=save_finalize][step=10] CheckpointManager Save Finalize is done on all hosts.
I0418 14:15:48.145439 136446713710400 checkpoint_manager.py:2032] [process=7][thread=MainThread][step=10][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=10.
W0418 14:15:48.145578 136446713710400 checkpoint_manager.py:1452] Waiting for previous save to complete took 48.543000 seconds. If this number is high, consider checkpointing less frequently.
I0418 14:15:48.147364 136446713710400 checkpoint_manager.py:1512] [process=7] Saving checkpoint at step 15
I0418 14:15:48.149730 136446713710400 event_tracking.py:70] [process=7] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_true/checkpoints/15.
I0418 14:15:48.854704 136446713710400 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 14:15:48.854801 136446713710400 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 14:15:48.895921 136446713710400 base_pytree_checkpoint_handler.py:154] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.042265s
I0418 14:15:48.896102 136446713710400 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/blocking_gbytes_per_sec: 33.732 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.0457301139831543 s) (per-host)
I0418 14:15:48.896152 136446713710400 base_pytree_checkpoint_handler.py:768] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.045802s (batch_requests_ready=0.001816s, total_serialization_initiated=0.043907s, others=0.000079s)
I0418 14:15:48.896239 136446713710400 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.050522s (all_items=0.000013s, per_item={'items': '0.00001335'}, temp_paths=0.050509)
I0418 14:15:48.896882 136446713710400 event_tracking.py:125] [process=7] [async] Finished blocking save in 0.75 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_true/checkpoints/15.
I0418 14:15:48.897243 136322146748160 async_checkpointer.py:76] [process=7][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 14:35:48.897207
I0418 14:15:48.899202 136446713710400 checkpoint_manager.py:1560] [process=7][thread=MainThread][step=15] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 14:15:48.899505 136322155140864 async_checkpointer.py:280] [process=7][thread=save_finalize] Waiting for background save thread=async_save.
I0418 14:15:48.899666 136446713710400 standard_logger.py:34] {'step': 15, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776521699.60255, 'wait_for_prev_duration_secs': 48.54299974441528, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776521748.1474051, 'checkpointer_blocking_duration_secs': 0.7499816417694092, 'get_old_steps_start_time': 1776521748.8974113, 'get_old_steps_duration_secs': 3.123283386230469e-05, 'checkpoint_manager_blocking_start_time': 1776521699.6002307, 'checkpoint_manager_blocking_duration_secs': 49.29940319061279}
I0418 14:15:48.899778 136446713710400 checkpointing.py:408] Started an asynchronous checkpoint save for step 15
I0418 14:15:48.900574 136446713710400 metric_logger.py:196] completed step: 14, seconds: 0.147, TFLOP/s/device: 92.239, Tokens/s/device: 13903.221, total_weights: 65536, loss: 5.864, lm_loss: 5.864, perplexity: 352.193
I0418 14:15:48.906492 136446713710400 metric_logger.py:196] completed step: 15, seconds: 0.147, TFLOP/s/device: 92.130, Tokens/s/device: 13886.911, total_weights: 65536, loss: 5.719, lm_loss: 5.719, perplexity: 304.661
I0418 14:15:56.043177 136321112729344 array_metadata_store.py:203] [process=7][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_true/checkpoints/15/items/array_metadatas/process_7
I0418 14:16:09.968958 136446713710400 metric_logger.py:196] completed step: 16, seconds: 49.596, TFLOP/s/device: 0.274, Tokens/s/device: 41.294, total_weights: 65536, loss: 5.605, lm_loss: 5.605, perplexity: 271.872
I0418 14:16:10.116525 136446713710400 metric_logger.py:196] completed step: 17, seconds: 20.916, TFLOP/s/device: 0.650, Tokens/s/device: 97.913, total_weights: 65536, loss: 5.517, lm_loss: 5.517, perplexity: 248.975
I0418 14:16:10.263847 136446713710400 metric_logger.py:196] completed step: 18, seconds: 0.152, TFLOP/s/device: 89.289, Tokens/s/device: 13458.632, total_weights: 65536, loss: 5.448, lm_loss: 5.448, perplexity: 232.333
I0418 14:16:10.410619 136446713710400 checkpointing.py:772] Waiting for step 19 to finish before checkpoint...
I0418 14:16:10.411308 136446713710400 checkpointing.py:776] Waited 0.0007092952728271484 seconds for step 19 to finish before starting checkpointing.
I0418 14:16:10.413433 136446713710400 checkpoint_manager.py:2020] [process=7][thread=MainThread][step=15][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 14:16:33.203422 136322146748160 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/gbytes_per_sec: 35.614 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 44.3530216217041 s) (per-host)
I0418 14:16:33.203544 136322146748160 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 44.306185s.
I0418 14:16:39.995486 136322146748160 async_checkpointer.py:160] [process=7][thread=async_save] Background save thread done. Time taken: 51.098113s.
I0418 14:16:39.995779 136322155140864 async_checkpointer.py:288] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 14:16:39.995899 136322155140864 async_checkpointer.py:298] [process=7][thread=save_finalize] No errors found in background save thread=async_save.
I0418 14:16:39.995948 136322155140864 checkpoint_manager.py:2137] [process=7][thread=save_finalize][step=15] CheckpointManager Save Finalize is syncing with other hosts...
I0418 14:16:39.998983 136322155140864 checkpoint_manager.py:2146] [process=7][thread=save_finalize][step=15] CheckpointManager Save Finalize is done on all hosts.
I0418 14:16:39.999184 136446713710400 checkpoint_manager.py:2032] [process=7][thread=MainThread][step=15][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=15.
W0418 14:16:39.999311 136446713710400 checkpoint_manager.py:1452] Waiting for previous save to complete took 29.585885 seconds. If this number is high, consider checkpointing less frequently.
I0418 14:16:40.000562 136446713710400 checkpoint_manager.py:1512] [process=7] Saving checkpoint at step 19
I0418 14:16:40.002606 136446713710400 event_tracking.py:70] [process=7] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_true/checkpoints/19.
I0418 14:16:40.722794 136446713710400 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 14:16:40.722890 136446713710400 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 14:16:40.765092 136446713710400 base_pytree_checkpoint_handler.py:154] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.043266s
I0418 14:16:40.765261 136446713710400 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/blocking_gbytes_per_sec: 33.023 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04671216011047363 s) (per-host)
I0418 14:16:40.765312 136446713710400 base_pytree_checkpoint_handler.py:768] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.046776s (batch_requests_ready=0.001654s, total_serialization_initiated=0.045053s, others=0.000068s)
I0418 14:16:40.765390 136446713710400 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.050964s (all_items=0.000010s, per_item={'items': '0.00001049'}, temp_paths=0.050953)
I0418 14:16:40.766134 136446713710400 event_tracking.py:125] [process=7] [async] Finished blocking save in 0.77 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_true/checkpoints/19.
I0418 14:16:40.766426 136321112729344 async_checkpointer.py:76] [process=7][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 14:36:40.766393
I0418 14:16:40.768415 136446713710400 checkpoint_manager.py:1560] [process=7][thread=MainThread][step=19] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 14:16:40.768647 136322155140864 async_checkpointer.py:280] [process=7][thread=save_finalize] Waiting for background save thread=async_save.
I0418 14:16:40.768766 136446713710400 standard_logger.py:34] {'step': 19, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776521770.4134047, 'wait_for_prev_duration_secs': 29.58588480949402, 'time_between_consecutive_saves_sec': 22.26812171936035, 'checkpointer_blocking_start_time': 1776521800.0006, 'checkpointer_blocking_duration_secs': 0.765984058380127, 'get_old_steps_start_time': 1776521800.7665992, 'get_old_steps_duration_secs': 2.765655517578125e-05, 'checkpoint_manager_blocking_start_time': 1776521770.4115372, 'checkpoint_manager_blocking_duration_secs': 30.357197284698486}
I0418 14:16:40.768876 136446713710400 checkpointing.py:408] Started an asynchronous checkpoint save for step 19
I0418 14:16:40.768921 136446713710400 checkpoint_manager.py:2020] [process=7][thread=MainThread][step=19][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 14:16:46.820328 136319460153088 array_metadata_store.py:203] [process=7][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_15_ocdbt_true/checkpoints/19/items/array_metadatas/process_7
I0418 14:17:23.765324 136321112729344 base_pytree_checkpoint_handler.py:130] [process=7] /jax/orbax/write/gbytes_per_sec: 36.695 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 43.046735763549805 s) (per-host)
I0418 14:17:23.765453 136321112729344 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.998913s.
I0418 14:17:32.881632 136321112729344 async_checkpointer.py:160] [process=7][thread=async_save] Background save thread done. Time taken: 52.115075s.
I0418 14:17:32.881827 136322155140864 async_checkpointer.py:288] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 14:17:32.881896 136322155140864 async_checkpointer.py:298] [process=7][thread=save_finalize] No errors found in background save thread=async_save.
I0418 14:17:32.881937 136322155140864 checkpoint_manager.py:2137] [process=7][thread=save_finalize][step=19] CheckpointManager Save Finalize is syncing with other hosts...
I0418 14:17:32.883410 136322155140864 checkpoint_manager.py:2146] [process=7][thread=save_finalize][step=19] CheckpointManager Save Finalize is done on all hosts.
I0418 14:17:32.883570 136446713710400 checkpoint_manager.py:2032] [process=7][thread=MainThread][step=19][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=19.
I0418 14:17:32.883723 136446713710400 checkpoint_manager.py:2009] [process=7][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 14:17:32.884622 136446713710400 metric_logger.py:196] completed step: 19, seconds: 0.147, TFLOP/s/device: 92.134, Tokens/s/device: 13887.382, 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 14:17:44 UTC 2026
EXIT_CODE=0