XPK Start: Sat Apr 18 00:26:44 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 00:27:08.984905: 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 00:27:09.512616 136020450203456 max_utils.py:273] Attempting to initialize the jax distributed system...
I0418 00:27:18.553223 136020450203456 distributed.py:149] Starting JAX distributed service on [::]:8482
I0418 00:27:18.555489 136020450203456 distributed.py:172] Connecting to JAX distributed service on mt-07-eval-0k7pc-slice-job-0-0.mt-07-eval-0k7pc:8482
I0418 00:27:19.592715 136020450203456 max_utils.py:284] Jax distributed system initialized!
I0418 00:27:25.749083 136020450203456 max_utils.py:800] System Information: Jax Version: 0.9.2
I0418 00:27:25.749185 136020450203456 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0418 00:27:25.749224 136020450203456 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 00:27:25.749258 136020450203456 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0418 00:27:26.434434 136020450203456 maxtext_utils.py:1551] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0418 00:27:26.434705 136020450203456 checkpointing.py:677] Setting up checkpoint logger...
I0418 00:27:26.434758 136020450203456 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0418 00:27:26.434804 136020450203456 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0418 00:27:26.435166 136020450203456 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 0x7bb51f11c170>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0418 00:27:30.211421 136020450203456 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0418 00:27:30.211667 136020450203456 checkpoint_manager.py:708] [process=0][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b9fd866aea0>}, handler_registry=None
I0418 00:27:30.212022 136020450203456 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b9fd866aea0>` 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 00:27:30.212090 136020450203456 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b9fd8670bc0>` 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 00:27:30.212136 136020450203456 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 0x7b9fd866aea0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b9fd866aea0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b9fd8670bc0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b9fd8670bc0>}).
I0418 00:27:30.212490 136020450203456 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0418 00:27:30.212582 136020450203456 async_checkpointer.py:192] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7b9fb8203ba0> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0418 00:27:32.590603 136020450203456 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints
I0418 00:27:33.007808 136020450203456 checkpoint_manager.py:929] [process=0][thread=MainThread] CheckpointManager created, primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=10), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7b9fd86727e0>
I0418 00:27:33.008015 136020450203456 checkpointing.py:301] Checkpoint manager created!
I0418 00:27:33.915928 136020450203456 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0418 00:27:33.916037 136020450203456 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 00:27:34.296506 136020450203456 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 00:27:34.296599 136020450203456 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 00:27:34.312747 136020450203456 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 00:27:34.312809 136020450203456 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 00:27:34.336233 136020450203456 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 00:27:34.336299 136020450203456 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 00:27:34.352524 136020450203456 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 00:27:34.352588 136020450203456 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 00:27:34.368741 136020450203456 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 00:27:34.368804 136020450203456 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 00:27:34.394088 136020450203456 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0418 00:27:34.394166 136020450203456 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 00:27:34.415168 136020450203456 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0418 00:27:34.415231 136020450203456 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0418 00:27:34.639939 136020450203456 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0418 00:27:34.640051 136020450203456 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0418 00:27:36.074983 136020450203456 maxtext_utils.py:1654] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: P('norm',)
Physical: (None,)
I0418 00:27:36.075106 136020450203456 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 00:27:36.075162 136020450203456 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 00:27:36.075216 136020450203456 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 00:27:36.075269 136020450203456 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 00:27:36.075309 136020450203456 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 00:27:36.075362 136020450203456 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 00:27:36.075415 136020450203456 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 00:27:36.075454 136020450203456 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 00:27:36.075492 136020450203456 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 00:27:36.075540 136020450203456 maxtext_utils.py:1654] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: P('embed_vocab', 'vocab')
Physical: ('fsdp', None)
I0418 00:27:36.075587 136020450203456 maxtext_utils.py:1654] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: P('vocab', 'embed_vocab')
Physical: (None, 'fsdp')
I0418 00:27:36.568213 136020450203456 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0418 00:27:36.568310 136020450203456 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0418 00:27:36.583701 136020450203456 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0418 00:27:36.583762 136020450203456 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0418 00:27:47.593159 136020450203456 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 00:27:47.600984 136020450203456 metric_logger.py:301] number parameters: 1.104 billion
I0418 00:27:59.413088 136020450203456 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0418 00:27:59.567157 136020450203456 checkpointing.py:776] Waited 0.15404844284057617 seconds for step 0 to finish before starting checkpointing.
I0418 00:27:59.569673 136020450203456 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 00:27:59.571304 136020450203456 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 0
I0418 00:27:59.572708 136020450203456 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/0.
I0418 00:28:00.730792 136020450203456 signaling_client.py:364] Using JaxDistributedSignalingClient
I0418 00:28:00.745063 136020450203456 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 00:28:00.745133 136020450203456 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 00:28:00.981953 135890106730240 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/0
I0418 00:28:01.015189 136020450203456 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.270848s
I0418 00:28:01.016089 136020450203456 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 5.564 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2772216796875 s) (per-host)
I0418 00:28:01.016170 136020450203456 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.277320s (batch_requests_ready=0.002192s, total_serialization_initiated=0.274313s, others=0.000815s)
I0418 00:28:01.016253 136020450203456 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.281364s (all_items=0.000017s, per_item={'items': '0.00001717'}, temp_paths=0.281347)
I0418 00:28:01.017013 136020450203456 event_tracking.py:125] [process=0] [async] Finished blocking save in 1.45 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/0.
I0418 00:28:01.017418 135890632963840 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 00:48:01.017377
I0418 00:28:01.028736 136020450203456 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 00:28:01.029009 135889563784960 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0418 00:28:01.029140 136020450203456 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776472079.5696564, 'wait_for_prev_duration_secs': 5.841255187988281e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776472079.5713427, 'checkpointer_blocking_duration_secs': 1.4462227821350098, 'get_old_steps_start_time': 1776472081.017592, 'get_old_steps_duration_secs': 2.8133392333984375e-05, 'checkpoint_manager_blocking_start_time': 1776472079.5676866, 'checkpoint_manager_blocking_duration_secs': 1.4614133834838867}
I0418 00:28:01.029257 136020450203456 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0418 00:28:01.029307 136020450203456 max_utils.py:750]
Memstats: After params initialized:
I0418 00:28:01.029355 136020450203456 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0))
I0418 00:28:01.029387 136020450203456 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0))
I0418 00:28:01.029413 136020450203456 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0))
I0418 00:28:01.029438 136020450203456 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0))
I0418 00:28:01.382339 136020450203456 metric_logger.py:196] completed step: 0, seconds: 11.512, TFLOP/s/device: 1.180, Tokens/s/device: 177.904, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0418 00:28:01.385562 136020450203456 metric_logger.py:281] To see full metrics 'tensorboard --logdir=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/tensorboard/'
I0418 00:28:02.216859 136020450203456 metric_logger.py:196] completed step: 1, seconds: 1.967, TFLOP/s/device: 6.907, Tokens/s/device: 1041.029, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0418 00:28:02.363135 136020450203456 metric_logger.py:196] completed step: 2, seconds: 0.835, TFLOP/s/device: 16.268, Tokens/s/device: 2452.043, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28637.129
I0418 00:28:02.510236 136020450203456 metric_logger.py:196] completed step: 3, seconds: 0.010, TFLOP/s/device: 1336.921, Tokens/s/device: 201515.301, total_weights: 65536, loss: 9.734, lm_loss: 9.734, perplexity: 16889.203
I0418 00:28:02.719121 135890089944832 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/0/items
I0418 00:28:02.897258 135888000513792 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776472082464778740, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776472082464778740, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/0/_CHECKPOINT_METADATA
I0418 00:28:04.284812 2953 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0418 00:28:08.564437 135894961501952 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/0/items/array_metadatas/process_0
I0418 00:28:26.077030 136020450203456 metric_logger.py:196] completed step: 4, seconds: 0.147, TFLOP/s/device: 92.343, Tokens/s/device: 13919.001, total_weights: 65536, loss: 9.277, lm_loss: 9.277, perplexity: 10694.614
I0418 00:28:26.085594 136020450203456 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.476, Tokens/s/device: 13938.990, total_weights: 65536, loss: 8.892, lm_loss: 8.892, perplexity: 7270.717
I0418 00:28:26.231621 136020450203456 metric_logger.py:196] completed step: 6, seconds: 23.565, TFLOP/s/device: 0.577, Tokens/s/device: 86.909, total_weights: 65536, loss: 8.592, lm_loss: 8.592, perplexity: 5390.759
I0418 00:28:26.379050 136020450203456 metric_logger.py:196] completed step: 7, seconds: 0.006, TFLOP/s/device: 2124.317, Tokens/s/device: 320200.125, total_weights: 65536, loss: 8.384, lm_loss: 8.384, perplexity: 4376.905
I0418 00:28:26.526258 136020450203456 metric_logger.py:196] completed step: 8, seconds: 0.147, TFLOP/s/device: 92.377, Tokens/s/device: 13924.016, total_weights: 65536, loss: 8.255, lm_loss: 8.255, perplexity: 3846.406
I0418 00:28:26.672936 136020450203456 checkpointing.py:772] Waiting for step 9 to finish before checkpoint...
I0418 00:28:26.673660 136020450203456 checkpointing.py:776] Waited 0.0007443428039550781 seconds for step 9 to finish before starting checkpointing.
I0418 00:28:26.675741 136020450203456 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 00:28:41.626452 135889572177664 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 37.349134s (commit=34.871838s, array_metadata_write=2.477296s)
I0418 00:28:41.628652 135890632963840 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 38.631 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 40.88975930213928 s) (per-host)
I0418 00:28:41.628779 135890632963840 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 40.611247s.
I0418 00:28:42.445275 135890632963840 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776472082464778740, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/0/_CHECKPOINT_METADATA
I0418 00:28:43.659945 135888000513792 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776472082464778740, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/0/_CHECKPOINT_METADATA
I0418 00:28:43.855392 135890632963840 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0418 00:28:44.043693 135890632963840 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000234s.
I0418 00:28:44.784104 135890632963840 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.207993s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/0/items
I0418 00:28:44.785792 135890632963840 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/0/items
I0418 00:28:45.772110 135890632963840 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/0
I0418 00:28:47.573945 135890632963840 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/0`.
I0418 00:28:47.574781 135890632963840 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 48.00 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/0
I0418 00:28:47.576140 135890632963840 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 46.558605s.
I0418 00:28:47.576328 135889563784960 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 00:28:47.576448 135889563784960 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0418 00:28:47.576510 135889563784960 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0418 00:28:47.579397 135889563784960 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0418 00:28:47.579580 136020450203456 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0418 00:28:47.579720 136020450203456 checkpoint_manager.py:1452] Waiting for previous save to complete took 20.903976 seconds. If this number is high, consider checkpointing less frequently.
I0418 00:28:47.581423 136020450203456 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 9
I0418 00:28:47.583649 136020450203456 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/9.
I0418 00:28:48.723610 136020450203456 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 00:28:48.723700 136020450203456 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 00:28:48.748226 136020450203456 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.025328s
I0418 00:28:48.749082 136020450203456 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 51.139 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.030164480209350586 s) (per-host)
I0418 00:28:48.749152 136020450203456 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.030253s (batch_requests_ready=0.001813s, total_serialization_initiated=0.027623s, others=0.000818s)
I0418 00:28:48.749222 136020450203456 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.034183s (all_items=0.000017s, per_item={'items': '0.00001693'}, temp_paths=0.034166)
I0418 00:28:48.749924 136020450203456 event_tracking.py:125] [process=0] [async] Finished blocking save in 1.17 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/9.
I0418 00:28:48.750234 135889563784960 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 00:48:48.750201
I0418 00:28:48.761177 136020450203456 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 00:28:48.761475 135889057470208 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0418 00:28:48.761638 136020450203456 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776472106.6757114, 'wait_for_prev_duration_secs': 20.903976440429688, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776472127.581461, 'checkpointer_blocking_duration_secs': 1.1689152717590332, 'get_old_steps_start_time': 1776472128.7503996, 'get_old_steps_duration_secs': 3.337860107421875e-05, 'checkpoint_manager_blocking_start_time': 1776472106.6739073, 'checkpoint_manager_blocking_duration_secs': 22.087695837020874}
I0418 00:28:48.761742 136020450203456 checkpointing.py:408] Started an asynchronous checkpoint save for step 9
I0418 00:28:48.761786 136020450203456 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 00:28:49.383972 135890632963840 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/9
I0418 00:28:52.403363 135887992121088 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/9/items
I0418 00:28:57.031956 135889074255616 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/9/items/array_metadatas/process_0
I0418 00:29:36.726480 135889065862912 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 42.767514s (commit=40.896499s, array_metadata_write=1.871015s)
I0418 00:29:36.727585 135889563784960 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 32.902 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 48.0086395740509 s) (per-host)
I0418 00:29:36.727714 135889563784960 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 47.977363s.
I0418 00:29:39.277105 135889563784960 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0418 00:29:39.872182 135889563784960 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000227s.
I0418 00:29:40.379748 135889563784960 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.933920s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/9/items
I0418 00:29:40.381570 135889563784960 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/9/items
I0418 00:29:41.326788 135889563784960 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/9
I0418 00:29:43.615458 135889563784960 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/9`.
I0418 00:29:43.616305 135889563784960 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 56.03 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_07_eval/checkpoints/9
I0418 00:29:43.617625 135889563784960 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 54.867272s.
I0418 00:29:43.617822 135889057470208 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 00:29:43.617953 135889057470208 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0418 00:29:43.618015 135889057470208 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0418 00:29:43.620726 135889057470208 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0418 00:29:43.620919 136020450203456 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0418 00:29:43.621062 136020450203456 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 00:29:43.621994 136020450203456 metric_logger.py:196] completed step: 9, seconds: 0.148, TFLOP/s/device: 91.891, Tokens/s/device: 13850.846, total_weights: 65536, loss: 8.179, lm_loss: 8.179, perplexity: 3564.635
Per train step:
Total TFLOPs: 13.59
split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Sat Apr 18 00:29:54 UTC 2026
EXIT_CODE=0