XPK Start: Sun Apr 19 04:33:00 UTC 2026 2026-04-19 04:33:04.782988: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:467] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered WARNING: All log messages before absl::InitializeLog() is called are written to STDERR E0000 00:00:1776573184.795904 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1776573184.799616 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1776573184.810879 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776573184.810897 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776573184.810900 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776573184.810902 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-19 04:33:23.972975: 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) I0419 04:33:24.504646 139555201431360 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-19 04:33:33,547:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0419 04:33:33.547116 139555201431360 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-19 04:33:33,549:jax._src.distributed:157: Connecting to JAX distributed service on mt-15-ocdbt-false-rs-vp9yn-slice-job-0-0.mt-15-ocdbt-false-rs-vp9yn:8482 I0419 04:33:33.549434 139555201431360 distributed.py:157] Connecting to JAX distributed service on mt-15-ocdbt-false-rs-vp9yn-slice-job-0-0.mt-15-ocdbt-false-rs-vp9yn:8482 I0419 04:33:35.473663 139555201431360 max_utils.py:284] Jax distributed system initialized! I0419 04:33:42.631080 139555201431360 max_utils.py:800] System Information: Jax Version: 0.8.1 I0419 04:33:42.631187 139555201431360 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0419 04:33:42.631228 139555201431360 max_utils.py:802] System Information: Jax Backend: PJRT C API TFRT TPU v6 lite Built on Nov 12 2025 14:16:36 (1762985796) cl/831091709 I0419 04:33:42.631263 139555201431360 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0419 04:33:43.242604 139555201431360 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0419 04:33:43.243210 139555201431360 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0419 04:33:43.243422 139555201431360 checkpointing.py:688] Setting up checkpoint logger... I0419 04:33:43.243478 139555201431360 checkpointing.py:234] Creating checkpoint manager with ocdbt=False and zarr3=True I0419 04:33:43.243537 139555201431360 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0419 04:33:43.243926 139555201431360 base_pytree_checkpoint_handler.py:415] Created BasePyTreeCheckpointHandler: use_ocdbt=False, use_zarr3=True, pytree_metadata_options=PyTreeMetadataOptions(support_rich_types=False), array_metadata_store=<orbax.checkpoint._src.metadata.array_metadata_store.Store object at 0x7eec29579520>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0419 04:33:44.611312 139555201431360 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0419 04:33:44.611653 139555201431360 checkpoint_manager.py:709] [process=0][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ed8b4095520>}, handler_registry=None I0419 04:33:44.611901 139555201431360 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ed8b4095520>` 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`. I0419 04:33:44.611952 139555201431360 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ed8b409a2a0>` 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`. I0419 04:33:44.611990 139555201431360 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 0x7ed8b4095520>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ed8b4095520>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ed8b409a2a0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ed8b409a2a0>}). I0419 04:33:44.612517 139555201431360 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0419 04:33:44.612595 139555201431360 async_checkpointer.py:177] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7ed7c0114180> timeout: 600 secs and primary_host=0 for async checkpoint writes I0419 04:33:47.331384 139434527794944 checkpoint.py:228] Read Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776573089794932079, 'commit_timestamp_nsecs': 1776573104663347136, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/9/_CHECKPOINT_METADATA I0419 04:33:47.581192 139434527794944 checkpoint_manager.py:1877] Missing metrics for step 9 I0419 04:33:48.446341 139555201431360 checkpoint_manager.py:1818] Found 3 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints I0419 04:33:49.310684 139555201431360 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_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7ed8bfe1d7f0> I0419 04:33:49.310884 139555201431360 checkpointing.py:302] Checkpoint manager created! I0419 04:33:50.209451 139555201431360 nnx_wrappers.py:455] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0419 04:33:50.209555 139555201431360 nnx_wrappers.py:455] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0419 04:33:50.588005 139555201431360 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0419 04:33:50.588131 139555201431360 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0419 04:33:50.604786 139555201431360 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0419 04:33:50.604859 139555201431360 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0419 04:33:50.628152 139555201431360 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 04:33:50.628218 139555201431360 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 04:33:50.644545 139555201431360 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 04:33:50.644608 139555201431360 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 04:33:50.660921 139555201431360 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 04:33:50.660982 139555201431360 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 04:33:50.686376 139555201431360 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0419 04:33:50.686448 139555201431360 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 04:33:50.706712 139555201431360 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0419 04:33:50.706778 139555201431360 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0419 04:33:50.899749 139555201431360 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0419 04:33:50.899873 139555201431360 checkpointing.py:582] restoring from this run's directory step 9 I0419 04:33:51.828332 139555201431360 checkpointer.py:304] Restoring checkpoint from gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/9. I0419 04:33:55.559559 1458 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0419 04:34:04.503068 139555201431360 jax_array_handlers.py:837] [process=0] /jax/orbax/read/worker/io/requested throughput: 156.236 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 10.114668846130371 s) (per-host) I0419 04:34:04.510553 139555201431360 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/read/gbytes_per_sec: 150.529 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 10.51351809501648 s) (per-host) I0419 04:34:07.589763 139555201431360 checkpointer.py:318] Finished restoring checkpoint in 16.00 seconds from gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/9. I0419 04:34:07.589988 139555201431360 standard_logger.py:34] {'step': 9, 'event_type': 'restore', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints', 'checkpointer_start_time': 1776573231.5851288, 'checkpointer_duration_secs': 16.004825592041016, 'checkpoint_manager_start_time': 1776573230.9001477, 'checkpoint_manager_duration_secs': 16.68980884552002} fsdp: 32 I0419 04:34:07.994026 139555201431360 maxtext_utils.py:1740] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: PartitionSpec('norm',) Physical: (None,) I0419 04:34:07.994189 139555201431360 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wi_0/kernel Shape: float32[2048,16,7168] Logical: PartitionSpec('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0419 04:34:07.994262 139555201431360 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wi_1/kernel Shape: float32[2048,16,7168] Logical: PartitionSpec('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0419 04:34:07.994429 139555201431360 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: PartitionSpec('mlp', 'layers', 'embed') Physical: (None, None, 'fsdp') I0419 04:34:07.994529 139555201431360 maxtext_utils.py:1740] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: PartitionSpec('norm', 'layers') Physical: (None, None) I0419 04:34:07.994597 139555201431360 maxtext_utils.py:1740] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: PartitionSpec('norm', 'layers') Physical: (None, None) I0419 04:34:07.994671 139555201431360 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/key/kernel Shape: float32[2048,16,16,128] Logical: PartitionSpec('embed', 'layers', 'kv_heads', 'kv_head_dim') Physical: ('fsdp', None, None, None) I0419 04:34:07.994751 139555201431360 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/out/kernel Shape: float32[16,16,128,2048] Logical: PartitionSpec('heads', 'layers', 'kv', 'embed') Physical: (None, None, None, 'fsdp') I0419 04:34:07.994820 139555201431360 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/query/kernel Shape: float32[2048,16,16,128] Logical: PartitionSpec('embed', 'layers', 'q_heads', 'kv') Physical: ('fsdp', None, None, None) I0419 04:34:07.994879 139555201431360 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/value/kernel Shape: float32[2048,16,16,128] Logical: PartitionSpec('embed', 'layers', 'kv_heads', 'kv_head_dim') Physical: ('fsdp', None, None, None) I0419 04:34:07.994943 139555201431360 maxtext_utils.py:1740] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: PartitionSpec('embed', 'vocab') Physical: ('fsdp', None) I0419 04:34:07.995005 139555201431360 maxtext_utils.py:1740] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: PartitionSpec('vocab', 'embed') Physical: (None, 'fsdp') I0419 04:34:08.490446 139555201431360 train.py:158] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0419 04:34:08.490539 139555201431360 train.py:158] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0419 04:34:08.510637 139555201431360 train.py:165] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0419 04:34:08.510735 139555201431360 train.py:165] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0419 04:34:19.309206 139555201431360 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. I0419 04:34:19.315269 139555201431360 metric_logger.py:289] number parameters: 1.104 billion I0419 04:34:30.852671 139555201431360 checkpointing.py:794] Waiting for step 10 to finish before checkpoint... I0419 04:34:31.006307 139555201431360 checkpointing.py:798] Waited 0.15361881256103516 seconds for step 10 to finish before starting checkpointing. I0419 04:34:31.009046 139555201431360 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0419 04:34:31.010766 139555201431360 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 10 I0419 04:34:31.013589 139555201431360 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/10. I0419 04:34:31.746111 139555201431360 signaling_client.py:364] Using JaxDistributedSignalingClient I0419 04:34:31.995368 139429112551168 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/10 I0419 04:34:32.037871 139555201431360 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0419 04:34:32.037972 139555201431360 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 I0419 04:34:32.279544 139555201431360 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.242742s I0419 04:34:32.280440 139555201431360 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.931 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5262212753295898 s) (per-host) I0419 04:34:32.280515 139555201431360 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.526317s (batch_requests_ready=0.280486s, total_serialization_initiated=0.245013s, others=0.000818s) I0419 04:34:32.280597 139555201431360 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.530637s (all_items=0.000019s, per_item={'items': '0.00001860'}, temp_paths=0.530619) I0419 04:34:32.281465 139430695073536 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0419 04:34:32.281621 139555201431360 async_checkpointer.py:561] Finished blocking save. Time taken: 1.270769s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/10. I0419 04:34:32.694303 139555201431360 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=10] Starting CheckpointManager Save Finalize thread=save_finalize I0419 04:34:32.694673 139429104158464 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0419 04:34:32.694852 139555201431360 standard_logger.py:34] {'step': 10, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776573271.0090148, 'wait_for_prev_duration_secs': 0.00012564659118652344, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776573271.0108235, 'checkpointer_blocking_duration_secs': 1.2709033489227295, 'get_old_steps_start_time': 1776573272.2817466, 'get_old_steps_duration_secs': 2.5987625122070312e-05, 'checkpoint_manager_blocking_start_time': 1776573271.0068822, 'checkpoint_manager_blocking_duration_secs': 1.6879181861877441} I0419 04:34:32.694976 139555201431360 checkpointing.py:409] Started an asynchronous checkpoint save for step 10 I0419 04:34:32.695028 139555201431360 max_utils.py:750] Memstats: After params initialized: I0419 04:34:32.695088 139555201431360 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0)) I0419 04:34:32.695121 139555201431360 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0)) I0419 04:34:32.695148 139555201431360 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0)) I0419 04:34:32.695172 139555201431360 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0)) I0419 04:34:33.001770 139555201431360 metric_logger.py:185] completed step: 10, seconds: 11.169, TFLOP/s/device: 1.217, Tokens/s/device: 183.368, total_weights: 65536, loss: 6.833 I0419 04:34:33.147138 139555201431360 metric_logger.py:185] completed step: 11, seconds: 2.147, TFLOP/s/device: 6.327, Tokens/s/device: 953.686, total_weights: 65536, loss: 6.528 I0419 04:34:33.294963 139555201431360 metric_logger.py:185] completed step: 12, seconds: 0.010, TFLOP/s/device: 1417.393, Tokens/s/device: 213644.899, total_weights: 65536, loss: 6.265 I0419 04:34:33.442530 139555201431360 metric_logger.py:185] completed step: 13, seconds: 0.146, TFLOP/s/device: 93.174, Tokens/s/device: 14044.135, total_weights: 65536, loss: 6.045 I0419 04:34:33.451101 139555201431360 checkpointing.py:794] Waiting for step 15 to finish before checkpoint... I0419 04:34:33.737325 139555201431360 checkpointing.py:798] Waited 0.2860901355743408 seconds for step 15 to finish before starting checkpointing. I0419 04:34:33.740195 139555201431360 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=10][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 04:34:35.427150 139431818295040 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/10/items I0419 04:34:35.604326 139430686680832 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776573275184466695, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776573275184466695, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/10/_CHECKPOINT_METADATA I0419 04:34:39.022186 139431256246016 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/10/items/array_metadatas/process_0 I0419 04:34:46.236379 139430703466240 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 9.267556s (commit=8.332950s, array_metadata_write=0.934607s) I0419 04:34:46.237905 139430695073536 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 109.061 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 14.48366093635559 s) (per-host) I0419 04:34:46.238015 139430695073536 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 13.956437s. I0419 04:34:47.627526 139430686680832 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776573275184466695, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/10/_CHECKPOINT_METADATA W0419 04:34:48.046294 139430695073536 ocdbt_utils.py:168] [process=0][thread=async_save] Skipping merge of OCDBT checkpoints: No per-process OCDBT checkpoint subdirs found in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/10/items, I0419 04:34:48.929479 139430695073536 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000393s. I0419 04:34:48.929723 139430695073536 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.926649s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/10/items I0419 04:34:48.931505 139430695073536 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/10/items I0419 04:34:49.324487 139430695073536 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/10 I0419 04:34:50.562623 139430695073536 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/10`. I0419 04:34:50.563679 139430695073536 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 19.552822s. directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/10 I0419 04:34:50.565245 139430695073536 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 18.283662s. I0419 04:34:50.565483 139429104158464 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 04:34:50.565613 139429104158464 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0419 04:34:50.565743 139429104158464 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=10] CheckpointManager Save Finalize is syncing with other hosts... I0419 04:34:50.567663 139429104158464 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=10] CheckpointManager Save Finalize is done on all hosts. I0419 04:34:50.567980 139555201431360 checkpoint_manager.py:2036] [process=0][thread=MainThread][step=10][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=10. W0419 04:34:50.568152 139555201431360 checkpoint_manager.py:1458] Waiting for previous save to complete took 16.827985 seconds. If this number is high, consider checkpointing less frequently. I0419 04:34:50.570041 139555201431360 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 15 I0419 04:34:50.574138 139555201431360 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/15. I0419 04:34:51.092355 139430695073536 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/15 I0419 04:34:51.104830 139555201431360 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0419 04:34:51.104975 139555201431360 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 I0419 04:34:51.148023 139555201431360 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.044400s I0419 04:34:51.149095 139555201431360 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.032 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.3065519332885742 s) (per-host) I0419 04:34:51.149209 139555201431360 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.306691s (batch_requests_ready=0.257125s, total_serialization_initiated=0.048610s, others=0.000957s) I0419 04:34:51.149344 139555201431360 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.311382s (all_items=0.000023s, per_item={'items': '0.00002265'}, temp_paths=0.311359) I0419 04:34:51.150542 139430703466240 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0419 04:34:51.150749 139555201431360 async_checkpointer.py:561] Finished blocking save. Time taken: 0.580603s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/15. I0419 04:34:51.586883 139555201431360 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=15] Starting CheckpointManager Save Finalize thread=save_finalize I0419 04:34:51.587369 139429104158464 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0419 04:34:51.587616 139555201431360 standard_logger.py:34] {'step': 15, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776573273.7401462, 'wait_for_prev_duration_secs': 16.82798457145691, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776573290.5701075, 'checkpointer_blocking_duration_secs': 0.5808131694793701, 'get_old_steps_start_time': 1776573291.1509447, 'get_old_steps_duration_secs': 3.981590270996094e-05, 'checkpoint_manager_blocking_start_time': 1776573273.7379699, 'checkpoint_manager_blocking_duration_secs': 17.849590063095093} I0419 04:34:51.587767 139555201431360 checkpointing.py:409] Started an asynchronous checkpoint save for step 15 I0419 04:35:09.802328 139555201431360 metric_logger.py:185] completed step: 14, seconds: 0.148, TFLOP/s/device: 91.611, Tokens/s/device: 13808.541, total_weights: 65536, loss: 5.864 I0419 04:35:09.810820 139555201431360 metric_logger.py:185] completed step: 15, seconds: 0.147, TFLOP/s/device: 92.354, Tokens/s/device: 13920.609, total_weights: 65536, loss: 5.719 I0419 04:35:09.956611 139555201431360 metric_logger.py:185] completed step: 16, seconds: 36.358, TFLOP/s/device: 0.374, Tokens/s/device: 56.328, total_weights: 65536, loss: 5.605 I0419 04:35:10.104284 139555201431360 metric_logger.py:185] completed step: 17, seconds: 0.006, TFLOP/s/device: 2097.750, Tokens/s/device: 316195.770, total_weights: 65536, loss: 5.517 I0419 04:35:10.251456 139555201431360 metric_logger.py:185] completed step: 18, seconds: 0.147, TFLOP/s/device: 92.475, Tokens/s/device: 13938.895, total_weights: 65536, loss: 5.448 I0419 04:35:10.398804 139555201431360 checkpointing.py:794] Waiting for step 19 to finish before checkpoint... I0419 04:35:10.399533 139555201431360 checkpointing.py:798] Waited 0.0007483959197998047 seconds for step 19 to finish before starting checkpointing. I0419 04:35:10.401495 139555201431360 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=15][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 04:35:10.804893 139428043413248 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/15/items I0419 04:35:14.089455 139431264638720 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/15/items/array_metadatas/process_0 I0419 04:35:22.726755 139431256246016 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 10.714341s (commit=8.892491s, array_metadata_write=1.821850s) I0419 04:35:22.727904 139430703466240 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 49.540 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 31.8853440284729 s) (per-host) I0419 04:35:22.728026 139430703466240 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 31.577338s. W0419 04:35:24.157145 139430703466240 ocdbt_utils.py:168] [process=0][thread=async_save] Skipping merge of OCDBT checkpoints: No per-process OCDBT checkpoint subdirs found in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/15/items, I0419 04:35:25.003672 139430703466240 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000225s. I0419 04:35:25.003865 139430703466240 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.497787s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/15/items I0419 04:35:25.005259 139430703466240 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/15/items I0419 04:35:25.404561 139430703466240 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/15 I0419 04:35:26.657989 139430703466240 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/15`. I0419 04:35:26.658793 139430703466240 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 36.088658s. directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/15 I0419 04:35:26.660116 139430703466240 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 35.509422s. I0419 04:35:26.660290 139429104158464 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 04:35:26.660361 139429104158464 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0419 04:35:26.660426 139429104158464 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=15] CheckpointManager Save Finalize is syncing with other hosts... I0419 04:35:26.662200 139429104158464 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=15] CheckpointManager Save Finalize is done on all hosts. I0419 04:35:26.662392 139555201431360 checkpoint_manager.py:2036] [process=0][thread=MainThread][step=15][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=15. W0419 04:35:26.662543 139555201431360 checkpoint_manager.py:1458] Waiting for previous save to complete took 16.261061 seconds. If this number is high, consider checkpointing less frequently. I0419 04:35:26.664029 139555201431360 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 19 I0419 04:35:26.667646 139555201431360 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/19. I0419 04:35:27.616418 139430703466240 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/19 I0419 04:35:27.628336 139555201431360 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0419 04:35:27.628437 139555201431360 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 I0419 04:35:27.655274 139555201431360 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.027932s I0419 04:35:27.656087 139555201431360 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.534 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2787332534790039 s) (per-host) I0419 04:35:27.656158 139555201431360 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.278822s (batch_requests_ready=0.247846s, total_serialization_initiated=0.030251s, others=0.000725s) I0419 04:35:27.656244 139555201431360 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.282612s (all_items=0.000017s, per_item={'items': '0.00001740'}, temp_paths=0.282595) I0419 04:35:27.656982 139429078980352 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0419 04:35:27.657140 139555201431360 async_checkpointer.py:561] Finished blocking save. Time taken: 0.993038s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/19. I0419 04:35:28.112483 139555201431360 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=19] Starting CheckpointManager Save Finalize thread=save_finalize I0419 04:35:28.112907 139429104158464 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0419 04:35:28.113110 139555201431360 standard_logger.py:34] {'step': 19, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776573310.401464, 'wait_for_prev_duration_secs': 16.26106095314026, 'time_between_consecutive_saves_sec': 19.83373475074768, 'checkpointer_blocking_start_time': 1776573326.6640747, 'checkpointer_blocking_duration_secs': 0.9931621551513672, 'get_old_steps_start_time': 1776573327.6572526, 'get_old_steps_duration_secs': 2.384185791015625e-05, 'checkpoint_manager_blocking_start_time': 1776573310.3997629, 'checkpoint_manager_blocking_duration_secs': 17.713311672210693} I0419 04:35:28.113224 139555201431360 checkpointing.py:409] Started an asynchronous checkpoint save for step 19 I0419 04:35:28.113269 139555201431360 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=19][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 04:35:29.343483 139428060198656 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/19/items I0419 04:35:32.514455 139431264638720 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/19/items/array_metadatas/process_0 I0419 04:35:40.112830 139431256246016 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 9.700243s (commit=7.838425s, array_metadata_write=1.861818s) I0419 04:35:40.113896 139429078980352 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 124.021 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 12.736527681350708 s) (per-host) I0419 04:35:40.113976 139429078980352 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 12.456867s. W0419 04:35:41.958134 139429078980352 ocdbt_utils.py:168] [process=0][thread=async_save] Skipping merge of OCDBT checkpoints: No per-process OCDBT checkpoint subdirs found in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/19/items, I0419 04:35:42.398280 139429078980352 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000253s. I0419 04:35:42.398468 139429078980352 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.502185s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/19/items I0419 04:35:42.399862 139429078980352 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/19/items I0419 04:35:43.226196 139429078980352 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/19 I0419 04:35:44.527254 139429078980352 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/19`. I0419 04:35:44.528042 139429078980352 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 17.863940s. directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_false/checkpoints/19 I0419 04:35:44.529402 139429078980352 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 16.872298s. I0419 04:35:44.529636 139429104158464 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 04:35:44.529757 139429104158464 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0419 04:35:44.529829 139429104158464 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=19] CheckpointManager Save Finalize is syncing with other hosts... I0419 04:35:44.531791 139429104158464 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=19] CheckpointManager Save Finalize is done on all hosts. I0419 04:35:44.531951 139555201431360 checkpoint_manager.py:2036] [process=0][thread=MainThread][step=19][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=19. I0419 04:35:44.532048 139555201431360 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0419 04:35:44.532736 139555201431360 metric_logger.py:185] completed step: 19, seconds: 0.148, TFLOP/s/device: 92.053, Tokens/s/device: 13875.245, total_weights: 65536, loss: 5.391 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Sun Apr 19 04:35:56 UTC 2026 EXIT_CODE=0