XPK Start: Sat Apr 18 01:29:35 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 01:29:59.843087: 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 01:30:00.022786 135664379631424 max_utils.py:273] Attempting to initialize the jax distributed system... I0418 01:30:09.064145 135664379631424 distributed.py:149] Starting JAX distributed service on [::]:8482 I0418 01:30:09.066589 135664379631424 distributed.py:172] Connecting to JAX distributed service on mt-15-ocdbt-true-rs-ta6l0-slice-job-0-0.mt-15-ocdbt-true-rs-ta6l0:8482 I0418 01:30:10.638072 135664379631424 max_utils.py:284] Jax distributed system initialized! I0418 01:30:16.644756 135664379631424 max_utils.py:800] System Information: Jax Version: 0.9.2 I0418 01:30:16.644878 135664379631424 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0418 01:30:16.644930 135664379631424 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 01:30:16.644965 135664379631424 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0418 01:30:17.332572 135664379631424 maxtext_utils.py:1551] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0418 01:30:17.332857 135664379631424 checkpointing.py:677] Setting up checkpoint logger... I0418 01:30:17.332916 135664379631424 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True I0418 01:30:17.332960 135664379631424 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None I0418 01:30:17.333295 135664379631424 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 0x7b6237b1c230>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0418 01:30:18.979409 135664379631424 checkpointing.py:265] Enabling policy for fixed interval checkpointing. I0418 01:30:18.979641 135664379631424 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 0x7b4dd018be30>}, handler_registry=None I0418 01:30:18.979896 135664379631424 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b4dd018be30>` 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 01:30:18.979947 135664379631424 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b610fc9fef0>` 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 01:30:18.979982 135664379631424 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 0x7b4dd018be30>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b4dd018be30>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b610fc9fef0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b610fc9fef0>}). I0418 01:30:18.980315 135664379631424 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34 I0418 01:30:18.980385 135664379631424 async_checkpointer.py:192] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7b4cf84b8040> timeout: 1200 secs and primary_host=0 for async checkpoint writes I0418 01:30:22.199994 135542826055424 checkpoint.py:228] Read Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776475587034777843, 'commit_timestamp_nsecs': 1776475641086523582, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/5/_CHECKPOINT_METADATA I0418 01:30:22.454421 135542826055424 checkpoint_manager.py:1871] Missing metrics for step 5 I0418 01:30:23.417877 135664379631424 checkpoint_manager.py:1812] Found 3 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints I0418 01:30:24.756566 135664379631424 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=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_15_ocdbt_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7b4cf873b200> I0418 01:30:24.756739 135664379631424 checkpointing.py:301] Checkpoint manager created! I0418 01:30:25.668481 135664379631424 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0418 01:30:25.668600 135664379631424 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0418 01:30:26.046951 135664379631424 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0418 01:30:26.047041 135664379631424 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0418 01:30:26.063410 135664379631424 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0418 01:30:26.063469 135664379631424 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0418 01:30:26.086964 135664379631424 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 01:30:26.087032 135664379631424 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0418 01:30:26.103171 135664379631424 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 01:30:26.103233 135664379631424 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0418 01:30:26.119453 135664379631424 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 01:30:26.119513 135664379631424 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0418 01:30:26.144175 135664379631424 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0418 01:30:26.144243 135664379631424 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0418 01:30:26.165132 135664379631424 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0418 01:30:26.165198 135664379631424 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0418 01:30:26.380709 135664379631424 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint I0418 01:30:26.380813 135664379631424 checkpointing.py:581] restoring from this run's directory step 9 I0418 01:30:26.624586 135664379631424 event_tracking.py:70] [process=0] [sync] Started load checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/9. I0418 01:30:27.437401 135664379631424 checkpointer.py:307] Restoring checkpoint from gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/9. I0418 01:30:27.437823 135664379631424 event_tracking.py:125] [process=0] [sync] Finished blocking load in 0.81 seconds. Continuing load @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/9. I0418 01:30:29.731295 1458 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0418 01:30:36.890213 135664379631424 jax_array_handlers.py:843] [process=0] /jax/orbax/read/worker/io/requested throughput: 220.562 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 7.164757490158081 s) (per-host) I0418 01:30:36.895100 135664379631424 base_pytree_checkpoint_handler.py:130] [process=0] /jax/checkpoint/read/gbytes_per_sec: 196.680 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 8.046546459197998 s) (per-host) I0418 01:30:38.428454 135664379631424 event_tracking.py:138] [process=0] [sync] Finished load in 11.80 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/9 I0418 01:30:38.428610 135664379631424 standard_logger.py:34] {'step': 9, 'event_type': 'restore', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints', 'checkpointer_start_time': 1776475826.6244884, 'checkpointer_duration_secs': 11.804089069366455, 'checkpoint_manager_start_time': 1776475826.3810713, 'checkpoint_manager_duration_secs': 12.047508478164673} fsdp: 32 I0418 01:30:38.839548 135664379631424 maxtext_utils.py:1654] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: P('norm',) Physical: (None,) I0418 01:30:38.839670 135664379631424 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 01:30:38.839728 135664379631424 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 01:30:38.839783 135664379631424 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 01:30:38.839845 135664379631424 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 01:30:38.839887 135664379631424 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 01:30:38.839942 135664379631424 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 01:30:38.839997 135664379631424 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 01:30:38.840044 135664379631424 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 01:30:38.840083 135664379631424 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 01:30:38.840138 135664379631424 maxtext_utils.py:1654] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: P('embed_vocab', 'vocab') Physical: ('fsdp', None) I0418 01:30:38.840188 135664379631424 maxtext_utils.py:1654] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: P('vocab', 'embed_vocab') Physical: (None, 'fsdp') I0418 01:30:39.320728 135664379631424 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0418 01:30:39.320822 135664379631424 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0418 01:30:39.336019 135664379631424 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0418 01:30:39.336081 135664379631424 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0418 01:30:50.293250 135664379631424 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 01:30:50.301008 135664379631424 metric_logger.py:301] number parameters: 1.104 billion I0418 01:31:02.169251 135664379631424 checkpointing.py:772] Waiting for step 10 to finish before checkpoint... I0418 01:31:02.321156 135664379631424 checkpointing.py:776] Waited 0.1518843173980713 seconds for step 10 to finish before starting checkpointing. I0418 01:31:02.324061 135664379631424 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0418 01:31:02.325708 135664379631424 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 10 I0418 01:31:02.327067 135664379631424 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/10. I0418 01:31:03.085284 135664379631424 signaling_client.py:364] Using JaxDistributedSignalingClient I0418 01:31:03.097357 135664379631424 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0418 01:31:03.097424 135664379631424 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 01:31:03.370735 135664379631424 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.274196s I0418 01:31:03.371713 135664379631424 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 5.526 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2791459560394287 s) (per-host) I0418 01:31:03.371798 135664379631424 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.279247s (batch_requests_ready=0.002137s, total_serialization_initiated=0.276211s, others=0.000899s) I0418 01:31:03.371903 135664379631424 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.283203s (all_items=0.000016s, per_item={'items': '0.00001574'}, temp_paths=0.283187) I0418 01:31:03.372653 135664379631424 event_tracking.py:125] [process=0] [async] Finished blocking save in 1.05 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/10. I0418 01:31:03.372977 135541135755008 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 01:51:03.372941 I0418 01:31:03.382191 135664379631424 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=10] Starting CheckpointManager Save Finalize thread=save_finalize I0418 01:31:03.382477 135537415415552 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0418 01:31:03.382628 135664379631424 standard_logger.py:34] {'step': 10, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776475862.3240423, 'wait_for_prev_duration_secs': 6.246566772460938e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776475862.3257477, 'checkpointer_blocking_duration_secs': 1.0473737716674805, 'get_old_steps_start_time': 1776475863.3731458, 'get_old_steps_duration_secs': 3.3855438232421875e-05, 'checkpoint_manager_blocking_start_time': 1776475862.3216333, 'checkpoint_manager_blocking_duration_secs': 1.060960054397583} I0418 01:31:03.382729 135664379631424 checkpointing.py:408] Started an asynchronous checkpoint save for step 10 I0418 01:31:03.382777 135664379631424 max_utils.py:750] Memstats: After params initialized: I0418 01:31:03.382861 135664379631424 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0)) I0418 01:31:03.382902 135664379631424 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0)) I0418 01:31:03.382936 135664379631424 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0)) I0418 01:31:03.382961 135664379631424 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0)) I0418 01:31:03.728698 135664379631424 metric_logger.py:196] completed step: 10, seconds: 11.551, TFLOP/s/device: 1.176, Tokens/s/device: 177.305, total_weights: 65536, loss: 6.833, lm_loss: 6.833, perplexity: 927.869 I0418 01:31:03.874106 135664379631424 metric_logger.py:196] completed step: 11, seconds: 1.558, TFLOP/s/device: 8.723, Tokens/s/device: 1314.842, total_weights: 65536, loss: 6.527, lm_loss: 6.527, perplexity: 683.659 I0418 01:31:04.021537 135664379631424 metric_logger.py:196] completed step: 12, seconds: 0.010, TFLOP/s/device: 1326.350, Tokens/s/device: 199921.906, total_weights: 65536, loss: 6.265, lm_loss: 6.265, perplexity: 526.068 I0418 01:31:04.168848 135664379631424 metric_logger.py:196] completed step: 13, seconds: 0.146, TFLOP/s/device: 93.282, Tokens/s/device: 14060.526, total_weights: 65536, loss: 6.045, lm_loss: 6.045, perplexity: 422.001 I0418 01:31:04.177728 135664379631424 checkpointing.py:772] Waiting for step 15 to finish before checkpoint... I0418 01:31:04.205020 135538482272000 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/10 I0418 01:31:04.462967 135664379631424 checkpointing.py:776] Waited 0.28503942489624023 seconds for step 15 to finish before starting checkpointing. I0418 01:31:04.466118 135664379631424 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=10][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0418 01:31:06.393020 135538465486592 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/10/items I0418 01:31:06.589448 135537407022848 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776475866137999245, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776475866137999245, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/10/_CHECKPOINT_METADATA I0418 01:31:09.685307 135537907963648 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_15_ocdbt_true/checkpoints/10/items/array_metadatas/process_0 I0418 01:31:50.441854 135537423808256 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 42.927353s (commit=40.760776s, array_metadata_write=2.166578s) I0418 01:31:50.443371 135541135755008 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 33.359 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 47.35076355934143 s) (per-host) I0418 01:31:50.443542 135541135755008 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 47.070450s. I0418 01:31:52.604579 135541135755008 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000368s. I0418 01:31:52.675701 135541135755008 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0418 01:31:52.886716 135537407022848 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776475866137999245, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/10/_CHECKPOINT_METADATA I0418 01:31:53.781521 135541135755008 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.934361s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/10/items I0418 01:31:53.783723 135541135755008 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/10/items I0418 01:31:54.798943 135541135755008 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/10 I0418 01:31:56.269180 135541135755008 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_15_ocdbt_true/checkpoints/10`. I0418 01:31:56.270339 135541135755008 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 53.94 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/10 I0418 01:31:56.271831 135541135755008 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 52.898736s. I0418 01:31:56.272139 135537415415552 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0418 01:31:56.272327 135537415415552 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0418 01:31:56.272441 135537415415552 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=10] CheckpointManager Save Finalize is syncing with other hosts... I0418 01:31:56.274277 135537415415552 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=10] CheckpointManager Save Finalize is done on all hosts. I0418 01:31:56.274599 135664379631424 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=10][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=10. W0418 01:31:56.274777 135664379631424 checkpoint_manager.py:1452] Waiting for previous save to complete took 51.808661 seconds. If this number is high, consider checkpointing less frequently. I0418 01:31:56.276754 135664379631424 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 15 I0418 01:31:56.279322 135664379631424 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/15. I0418 01:31:57.021276 135664379631424 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0418 01:31:57.021411 135664379631424 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 01:31:57.050506 135664379631424 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.030236s I0418 01:31:57.051424 135664379631424 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 39.814 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.03874468803405762 s) (per-host) I0418 01:31:57.051527 135664379631424 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.038870s (batch_requests_ready=0.004007s, total_serialization_initiated=0.034014s, others=0.000849s) I0418 01:31:57.051646 135664379631424 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.044013s (all_items=0.000024s, per_item={'items': '0.00002360'}, temp_paths=0.043989) I0418 01:31:57.052618 135664379631424 event_tracking.py:125] [process=0] [async] Finished blocking save in 0.78 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/15. I0418 01:31:57.053058 135537415415552 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 01:51:57.053016 I0418 01:31:57.055554 135664379631424 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=15] Starting CheckpointManager Save Finalize thread=save_finalize I0418 01:31:57.055895 135537907963648 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0418 01:31:57.056090 135664379631424 standard_logger.py:34] {'step': 15, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776475864.466073, 'wait_for_prev_duration_secs': 51.80866074562073, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776475916.2768202, 'checkpointer_blocking_duration_secs': 0.7764270305633545, 'get_old_steps_start_time': 1776475917.053274, 'get_old_steps_duration_secs': 4.601478576660156e-05, 'checkpoint_manager_blocking_start_time': 1776475864.463638, 'checkpoint_manager_blocking_duration_secs': 52.592392921447754} I0418 01:31:57.056241 135664379631424 checkpointing.py:408] Started an asynchronous checkpoint save for step 15 I0418 01:31:57.262260 135541135755008 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/15 I0418 01:32:00.431594 135537916356352 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/15/items I0418 01:32:03.435166 135538473879296 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_15_ocdbt_true/checkpoints/15/items/array_metadatas/process_0 I0418 01:32:20.551736 135664379631424 metric_logger.py:196] completed step: 14, seconds: 0.148, TFLOP/s/device: 91.998, Tokens/s/device: 13866.883, total_weights: 65536, loss: 5.864, lm_loss: 5.864, perplexity: 352.193 I0418 01:32:20.560810 135664379631424 metric_logger.py:196] completed step: 15, seconds: 0.147, TFLOP/s/device: 92.327, Tokens/s/device: 13916.542, total_weights: 65536, loss: 5.719, lm_loss: 5.719, perplexity: 304.661 I0418 01:32:20.706361 135664379631424 metric_logger.py:196] completed step: 16, seconds: 76.382, TFLOP/s/device: 0.178, Tokens/s/device: 26.813, total_weights: 65536, loss: 5.605, lm_loss: 5.605, perplexity: 271.872 I0418 01:32:20.853636 135664379631424 metric_logger.py:196] completed step: 17, seconds: 0.007, TFLOP/s/device: 1942.406, Tokens/s/device: 292780.558, total_weights: 65536, loss: 5.517, lm_loss: 5.517, perplexity: 248.975 I0418 01:32:21.001077 135664379631424 metric_logger.py:196] completed step: 18, seconds: 0.147, TFLOP/s/device: 92.651, Tokens/s/device: 13965.319, total_weights: 65536, loss: 5.448, lm_loss: 5.448, perplexity: 232.333 I0418 01:32:21.147692 135664379631424 checkpointing.py:772] Waiting for step 19 to finish before checkpoint... I0418 01:32:21.148402 135664379631424 checkpointing.py:776] Waited 0.0007297992706298828 seconds for step 19 to finish before starting checkpointing. I0418 01:32:21.150392 135664379631424 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=15][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0418 01:32:45.012409 135538465486592 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 43.407406s (commit=40.422932s, array_metadata_write=2.984474s) I0418 01:32:45.013521 135537415415552 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 32.908 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 48.000818490982056 s) (per-host) I0418 01:32:45.013652 135537415415552 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 47.960444s. I0418 01:32:48.042417 135537415415552 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0418 01:32:49.127931 135537415415552 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000252s. I0418 01:32:49.128123 135537415415552 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.359860s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/15/items I0418 01:32:49.129508 135537415415552 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/15/items I0418 01:32:50.964473 135537415415552 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/15 I0418 01:32:53.412606 135537415415552 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_15_ocdbt_true/checkpoints/15`. I0418 01:32:53.413424 135537415415552 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 57.14 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/15 I0418 01:32:53.414970 135537415415552 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 56.361760s. I0418 01:32:53.415193 135537907963648 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0418 01:32:53.415318 135537907963648 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0418 01:32:53.415389 135537907963648 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=15] CheckpointManager Save Finalize is syncing with other hosts... I0418 01:32:53.417400 135537907963648 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=15] CheckpointManager Save Finalize is done on all hosts. I0418 01:32:53.417574 135664379631424 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=15][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=15. W0418 01:32:53.417699 135664379631424 checkpoint_manager.py:1452] Waiting for previous save to complete took 32.267312 seconds. If this number is high, consider checkpointing less frequently. I0418 01:32:53.419004 135664379631424 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 19 I0418 01:32:53.421062 135664379631424 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/19. I0418 01:32:54.166707 135664379631424 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0418 01:32:54.166803 135664379631424 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 01:32:54.191343 135664379631424 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.025329s I0418 01:32:54.192059 135664379631424 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 51.842 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.029755353927612305 s) (per-host) I0418 01:32:54.192126 135664379631424 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.029840s (batch_requests_ready=0.001845s, total_serialization_initiated=0.027323s, others=0.000672s) I0418 01:32:54.192197 135664379631424 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.033781s (all_items=0.000015s, per_item={'items': '0.00001550'}, temp_paths=0.033765) I0418 01:32:54.192819 135664379631424 event_tracking.py:125] [process=0] [async] Finished blocking save in 0.77 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/19. I0418 01:32:54.193070 135538482272000 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 01:52:54.193027 I0418 01:32:54.196526 135664379631424 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=19] Starting CheckpointManager Save Finalize thread=save_finalize I0418 01:32:54.196774 135537907963648 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0418 01:32:54.196922 135664379631424 standard_logger.py:34] {'step': 19, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776475941.1503592, 'wait_for_prev_duration_secs': 32.26731204986572, 'time_between_consecutive_saves_sec': 24.876012325286865, 'checkpointer_blocking_start_time': 1776475973.4190435, 'checkpointer_blocking_duration_secs': 0.7741658687591553, 'get_old_steps_start_time': 1776475974.1932244, 'get_old_steps_duration_secs': 2.5987625122070312e-05, 'checkpoint_manager_blocking_start_time': 1776475941.148638, 'checkpoint_manager_blocking_duration_secs': 33.048250675201416} I0418 01:32:54.197069 135664379631424 checkpointing.py:408] Started an asynchronous checkpoint save for step 19 I0418 01:32:54.197114 135664379631424 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=19][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0418 01:32:54.408436 135537415415552 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/19 I0418 01:32:57.907176 135538473879296 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/19/items I0418 01:33:00.984293 135541139941120 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_15_ocdbt_true/checkpoints/19/items/array_metadatas/process_0 I0418 01:33:41.405052 135541131548416 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 42.314785s (commit=40.623327s, array_metadata_write=1.691458s) I0418 01:33:41.407721 135538482272000 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 33.434 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 47.24538969993591 s) (per-host) I0418 01:33:41.407854 135538482272000 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 47.214654s. I0418 01:33:45.005331 135538482272000 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0418 01:33:45.095293 135538482272000 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000236s. I0418 01:33:46.092123 135538482272000 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.954157s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/19/items I0418 01:33:46.093826 135538482272000 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/19/items I0418 01:33:46.630498 135538482272000 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/19 I0418 01:33:48.078036 135538482272000 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_15_ocdbt_true/checkpoints/19`. I0418 01:33:48.078855 135538482272000 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 54.66 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_true/checkpoints/19 I0418 01:33:48.080309 135538482272000 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 53.887119s. I0418 01:33:48.080508 135537907963648 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0418 01:33:48.080625 135537907963648 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0418 01:33:48.080679 135537907963648 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=19] CheckpointManager Save Finalize is syncing with other hosts... I0418 01:33:48.082684 135537907963648 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=19] CheckpointManager Save Finalize is done on all hosts. I0418 01:33:48.082828 135664379631424 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=19][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=19. I0418 01:33:48.082937 135664379631424 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0418 01:33:48.083986 135664379631424 metric_logger.py:196] completed step: 19, seconds: 0.147, TFLOP/s/device: 92.402, Tokens/s/device: 13927.804, 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 01:34:00 UTC 2026 EXIT_CODE=0