XPK Start: Fri Apr 17 04:22:55 UTC 2026 PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used. 2026-04-17 04:23:18.959112: 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) I0417 04:23:19.137583 136215914518336 max_utils.py:273] Attempting to initialize the jax distributed system... I0417 04:23:28.178584 136215914518336 distributed.py:149] Starting JAX distributed service on [::]:8482 I0417 04:23:28.180982 136215914518336 distributed.py:172] Connecting to JAX distributed service on mt-16-zarr3-true-rs-fyobu-slice-job-0-0.mt-16-zarr3-true-rs-fyobu:8482 I0417 04:23:29.496318 136215914518336 max_utils.py:284] Jax distributed system initialized! I0417 04:23:35.552648 136215914518336 max_utils.py:800] System Information: Jax Version: 0.9.2 I0417 04:23:35.552748 136215914518336 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0417 04:23:35.552788 136215914518336 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 I0417 04:23:35.552822 136215914518336 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0417 04:23:36.222511 136215914518336 maxtext_utils.py:1687] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0417 04:23:36.223093 136215914518336 maxtext_utils.py:1687] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0417 04:23:36.223270 136215914518336 checkpointing.py:688] Setting up checkpoint logger... I0417 04:23:36.223320 136215914518336 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0417 04:23:36.223362 136215914518336 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None I0417 04:23:36.223693 136215914518336 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 0x7be29e31c6b0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0417 04:23:37.449227 136215914518336 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0417 04:23:37.449456 136215914518336 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 0x7bce384549e0>}, handler_registry=None I0417 04:23:37.449691 136215914518336 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7bce384549e0>` 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`. I0417 04:23:37.449740 136215914518336 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7bce384558e0>` 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`. I0417 04:23:37.449776 136215914518336 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 0x7bce384549e0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7bce384549e0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7bce384558e0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7bce384558e0>}). I0417 04:23:37.450104 136215914518336 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34 I0417 04:23:37.450180 136215914518336 async_checkpointer.py:192] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7bcd505b87c0> timeout: 1200 secs and primary_host=0 for async checkpoint writes I0417 04:23:39.831631 136094016333568 checkpoint.py:228] Read Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776399598195380619, 'commit_timestamp_nsecs': 1776399644637618619, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/5/_CHECKPOINT_METADATA I0417 04:23:40.074985 136094016333568 checkpoint_manager.py:1871] Missing metrics for step 5 I0417 04:23:40.490260 136215914518336 checkpoint_manager.py:1812] Found 3 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints I0417 04:23:41.923720 136215914518336 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_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7bce38453440> I0417 04:23:41.923909 136215914518336 checkpointing.py:302] Checkpoint manager created! I0417 04:23:42.834001 136215914518336 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0417 04:23:42.834104 136215914518336 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0417 04:23:43.211524 136215914518336 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0417 04:23:43.211612 136215914518336 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0417 04:23:43.227966 136215914518336 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0417 04:23:43.228019 136215914518336 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0417 04:23:43.251710 136215914518336 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0417 04:23:43.251776 136215914518336 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0417 04:23:43.268112 136215914518336 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0417 04:23:43.268177 136215914518336 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0417 04:23:43.284389 136215914518336 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0417 04:23:43.284449 136215914518336 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0417 04:23:43.309956 136215914518336 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0417 04:23:43.310021 136215914518336 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0417 04:23:43.330810 136215914518336 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0417 04:23:43.330885 136215914518336 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0417 04:23:43.543552 136215914518336 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0417 04:23:43.543660 136215914518336 checkpointing.py:582] restoring from this run's directory step 9 I0417 04:23:43.787710 136215914518336 event_tracking.py:70] [process=0] [sync] Started load checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/9. I0417 04:23:44.035127 136215914518336 checkpointer.py:307] Restoring checkpoint from gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/9. I0417 04:23:44.035560 136215914518336 event_tracking.py:125] [process=0] [sync] Finished blocking load in 0.25 seconds. Continuing load @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/9. I0417 04:23:45.737506 1458 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0417 04:23:53.196236 136215914518336 jax_array_handlers.py:843] [process=0] /jax/orbax/read/worker/io/requested throughput: 211.699 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 7.4647252559661865 s) (per-host) I0417 04:23:53.201577 136215914518336 base_pytree_checkpoint_handler.py:130] [process=0] /jax/checkpoint/read/gbytes_per_sec: 201.385 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 7.85855507850647 s) (per-host) I0417 04:23:54.187740 136215914518336 event_tracking.py:138] [process=0] [sync] Finished load in 10.40 seconds @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/9 I0417 04:23:54.188040 136215914518336 standard_logger.py:34] {'step': 9, 'event_type': 'restore', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints', 'checkpointer_start_time': 1776399823.787615, 'checkpointer_duration_secs': 10.400376796722412, 'checkpoint_manager_start_time': 1776399823.5439203, 'checkpoint_manager_duration_secs': 10.6440749168396} fsdp: 32 I0417 04:23:54.603944 136215914518336 maxtext_utils.py:1796] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: P('norm',) Physical: (None,) I0417 04:23:54.604073 136215914518336 maxtext_utils.py:1796] params/params/decoder/layers/mlp/wi_0/kernel Shape: float32[2048,16,7168] Logical: P('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0417 04:23:54.604126 136215914518336 maxtext_utils.py:1796] params/params/decoder/layers/mlp/wi_1/kernel Shape: float32[2048,16,7168] Logical: P('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0417 04:23:54.604185 136215914518336 maxtext_utils.py:1796] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: P('mlp', 'layers', 'embed') Physical: (None, None, 'fsdp') I0417 04:23:54.604239 136215914518336 maxtext_utils.py:1796] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: (None, None) I0417 04:23:54.604279 136215914518336 maxtext_utils.py:1796] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: (None, None) I0417 04:23:54.604334 136215914518336 maxtext_utils.py:1796] 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) I0417 04:23:54.604388 136215914518336 maxtext_utils.py:1796] 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') I0417 04:23:54.604430 136215914518336 maxtext_utils.py:1796] 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) I0417 04:23:54.604467 136215914518336 maxtext_utils.py:1796] 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) I0417 04:23:54.604515 136215914518336 maxtext_utils.py:1796] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: P('embed_vocab', 'vocab') Physical: ('fsdp', None) I0417 04:23:54.604561 136215914518336 maxtext_utils.py:1796] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: P('vocab', 'embed_vocab') Physical: (None, 'fsdp') I0417 04:23:55.093077 136215914518336 train.py:157] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0417 04:23:55.093171 136215914518336 train.py:157] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0417 04:23:55.108913 136215914518336 train.py:164] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0417 04:23:55.108978 136215914518336 train.py:164] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0417 04:24:06.176254 136215914518336 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. I0417 04:24:06.183989 136215914518336 metric_logger.py:289] number parameters: 1.104 billion I0417 04:24:17.791315 136215914518336 checkpointing.py:794] Waiting for step 10 to finish before checkpoint... I0417 04:24:17.946678 136215914518336 checkpointing.py:798] Waited 0.15533828735351562 seconds for step 10 to finish before starting checkpointing. I0417 04:24:17.949093 136215914518336 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0417 04:24:17.950919 136215914518336 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 10 I0417 04:24:17.952322 136215914518336 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/10. I0417 04:24:18.663178 136215914518336 signaling_client.py:364] Using JaxDistributedSignalingClient I0417 04:24:18.676176 136215914518336 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0417 04:24:18.676243 136215914518336 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 I0417 04:24:18.951353 136215914518336 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.275939s I0417 04:24:18.952105 136215914518336 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 5.497 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2805972099304199 s) (per-host) I0417 04:24:18.952174 136215914518336 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.280681s (batch_requests_ready=0.002065s, total_serialization_initiated=0.277958s, others=0.000658s) I0417 04:24:18.952272 136215914518336 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.285000s (all_items=0.000017s, per_item={'items': '0.00001693'}, temp_paths=0.284983) I0417 04:24:18.952901 136090184242944 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/10 I0417 04:24:18.953396 136215914518336 event_tracking.py:125] [process=0] [async] Finished blocking save in 1.00 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/10. I0417 04:24:18.953679 136090686035712 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-17 04:44:18.953647 I0417 04:24:18.963363 136215914518336 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=10] Starting CheckpointManager Save Finalize thread=save_finalize I0417 04:24:18.963685 136089624893184 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0417 04:24:18.963862 136215914518336 standard_logger.py:34] {'step': 10, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776399857.9490755, 'wait_for_prev_duration_secs': 6.270408630371094e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776399857.9509573, 'checkpointer_blocking_duration_secs': 1.0028736591339111, 'get_old_steps_start_time': 1776399858.953871, 'get_old_steps_duration_secs': 3.0994415283203125e-05, 'checkpoint_manager_blocking_start_time': 1776399857.9471712, 'checkpoint_manager_blocking_duration_secs': 1.0166356563568115} I0417 04:24:18.963971 136215914518336 checkpointing.py:409] Started an asynchronous checkpoint save for step 10 I0417 04:24:18.964022 136215914518336 max_utils.py:750] Memstats: After params initialized: I0417 04:24:18.964072 136215914518336 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0)) I0417 04:24:18.964107 136215914518336 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0)) I0417 04:24:18.964134 136215914518336 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0)) I0417 04:24:18.964159 136215914518336 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0)) I0417 04:24:19.300633 136215914518336 metric_logger.py:185] completed step: 10, seconds: 11.278, TFLOP/s/device: 1.205, Tokens/s/device: 181.596, total_weights: 65536, loss: 6.833 I0417 04:24:19.446006 136215914518336 metric_logger.py:185] completed step: 11, seconds: 1.508, TFLOP/s/device: 9.012, Tokens/s/device: 1358.324, total_weights: 65536, loss: 6.527 I0417 04:24:19.593615 136215914518336 metric_logger.py:185] completed step: 12, seconds: 0.009, TFLOP/s/device: 1440.077, Tokens/s/device: 217064.123, total_weights: 65536, loss: 6.265 I0417 04:24:19.740819 136215914518336 metric_logger.py:185] completed step: 13, seconds: 0.146, TFLOP/s/device: 92.936, Tokens/s/device: 14008.400, total_weights: 65536, loss: 6.045 I0417 04:24:19.749305 136215914518336 checkpointing.py:794] Waiting for step 15 to finish before checkpoint... I0417 04:24:20.035084 136215914518336 checkpointing.py:798] Waited 0.28562426567077637 seconds for step 15 to finish before starting checkpointing. I0417 04:24:20.038055 136215914518336 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=10][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0417 04:24:20.800686 136090167457536 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/10/items I0417 04:24:21.009058 136089608107776 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776399860542188123, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776399860542188123, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/10/_CHECKPOINT_METADATA I0417 04:24:24.126429 136089651201792 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_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/10/items/array_metadatas/process_0 I0417 04:25:03.884933 136089642809088 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 41.542358s (commit=40.242238s, array_metadata_write=1.300121s) I0417 04:25:03.886497 136090686035712 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 34.935 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 45.21494674682617 s) (per-host) I0417 04:25:03.886669 136090686035712 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 44.932878s. I0417 04:25:06.606731 136090686035712 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000399s. I0417 04:25:06.678916 136089608107776 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776399860542188123, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/10/_CHECKPOINT_METADATA I0417 04:25:07.391714 136090686035712 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0417 04:25:08.317025 136090686035712 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.292507s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/10/items I0417 04:25:08.326430 136090686035712 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/10/items I0417 04:25:08.833360 136090686035712 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/10 I0417 04:25:10.294629 136090686035712 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/10`. I0417 04:25:10.295736 136090686035712 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 52.34 seconds @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/10 I0417 04:25:10.297406 136090686035712 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 51.343613s. I0417 04:25:10.297680 136089624893184 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0417 04:25:10.297857 136089624893184 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0417 04:25:10.297971 136089624893184 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=10] CheckpointManager Save Finalize is syncing with other hosts... I0417 04:25:10.299536 136089624893184 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=10] CheckpointManager Save Finalize is done on all hosts. I0417 04:25:10.299849 136215914518336 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. W0417 04:25:10.299981 136215914518336 checkpoint_manager.py:1452] Waiting for previous save to complete took 50.261930 seconds. If this number is high, consider checkpointing less frequently. I0417 04:25:10.301898 136215914518336 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 15 I0417 04:25:10.304277 136215914518336 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/15. I0417 04:25:11.027462 136215914518336 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0417 04:25:11.027601 136215914518336 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 I0417 04:25:11.077038 136215914518336 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.050598s I0417 04:25:11.077955 136215914518336 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 25.712 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.05999326705932617 s) (per-host) I0417 04:25:11.078070 136215914518336 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.060134s (batch_requests_ready=0.004765s, total_serialization_initiated=0.054549s, others=0.000820s) I0417 04:25:11.078208 136215914518336 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.065183s (all_items=0.000024s, per_item={'items': '0.00002384'}, temp_paths=0.065159) I0417 04:25:11.079223 136215914518336 event_tracking.py:125] [process=0] [async] Finished blocking save in 0.78 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/15. I0417 04:25:11.079685 136089624893184 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-17 04:45:11.079642 I0417 04:25:11.082330 136215914518336 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=15] Starting CheckpointManager Save Finalize thread=save_finalize I0417 04:25:11.082699 136089659594496 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0417 04:25:11.082969 136215914518336 standard_logger.py:34] {'step': 15, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776399860.0380104, 'wait_for_prev_duration_secs': 50.26193046569824, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776399910.3019633, 'checkpointer_blocking_duration_secs': 0.7779369354248047, 'get_old_steps_start_time': 1776399911.0799332, 'get_old_steps_duration_secs': 5.4836273193359375e-05, 'checkpoint_manager_blocking_start_time': 1776399860.0357246, 'checkpoint_manager_blocking_duration_secs': 51.0471727848053} I0417 04:25:11.083162 136215914518336 checkpointing.py:409] Started an asynchronous checkpoint save for step 15 I0417 04:25:11.275035 136090686035712 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/15 I0417 04:25:13.544690 136089651201792 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/15/items I0417 04:25:17.320147 136090175850240 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_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/15/items/array_metadatas/process_0 I0417 04:25:31.614134 136215914518336 metric_logger.py:185] completed step: 14, seconds: 0.148, TFLOP/s/device: 91.612, Tokens/s/device: 13808.728, total_weights: 65536, loss: 5.864 I0417 04:25:31.623144 136215914518336 metric_logger.py:185] completed step: 15, seconds: 0.146, TFLOP/s/device: 92.948, Tokens/s/device: 14010.125, total_weights: 65536, loss: 5.719 I0417 04:25:31.768535 136215914518336 metric_logger.py:185] completed step: 16, seconds: 71.873, TFLOP/s/device: 0.189, Tokens/s/device: 28.495, total_weights: 65536, loss: 5.605 I0417 04:25:31.916029 136215914518336 metric_logger.py:185] completed step: 17, seconds: 0.007, TFLOP/s/device: 2003.410, Tokens/s/device: 301975.818, total_weights: 65536, loss: 5.517 I0417 04:25:32.063333 136215914518336 metric_logger.py:185] completed step: 18, seconds: 0.147, TFLOP/s/device: 92.668, Tokens/s/device: 13967.890, total_weights: 65536, loss: 5.448 I0417 04:25:32.210283 136215914518336 checkpointing.py:794] Waiting for step 19 to finish before checkpoint... I0417 04:25:32.210981 136215914518336 checkpointing.py:798] Waited 0.0007178783416748047 seconds for step 19 to finish before starting checkpointing. I0417 04:25:32.213105 136215914518336 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=15][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0417 04:25:56.467821 136090167457536 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 41.757367s (commit=40.456991s, array_metadata_write=1.300376s) I0417 04:25:56.468953 136089624893184 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 34.754 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 45.45097494125366 s) (per-host) I0417 04:25:56.469083 136089624893184 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 45.389246s. I0417 04:25:58.302136 136089624893184 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000248s. I0417 04:25:58.670048 136089624893184 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0417 04:25:59.602268 136089624893184 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.299248s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/15/items I0417 04:25:59.604099 136089624893184 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/15/items I0417 04:26:00.124135 136089624893184 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/15 I0417 04:26:01.525030 136089624893184 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/15`. I0417 04:26:01.525947 136089624893184 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 51.22 seconds @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/15 I0417 04:26:01.527511 136089624893184 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 50.447673s. I0417 04:26:01.527732 136089659594496 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0417 04:26:01.527872 136089659594496 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0417 04:26:01.527942 136089659594496 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=15] CheckpointManager Save Finalize is syncing with other hosts... I0417 04:26:01.530519 136089659594496 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=15] CheckpointManager Save Finalize is done on all hosts. I0417 04:26:01.530667 136215914518336 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. W0417 04:26:01.530745 136215914518336 checkpoint_manager.py:1452] Waiting for previous save to complete took 29.317659 seconds. If this number is high, consider checkpointing less frequently. I0417 04:26:01.532469 136215914518336 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 19 I0417 04:26:01.534571 136215914518336 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/19. I0417 04:26:02.256684 136215914518336 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0417 04:26:02.256777 136215914518336 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 I0417 04:26:02.282713 136215914518336 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.026949s I0417 04:26:02.283465 136215914518336 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 49.057 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.031444549560546875 s) (per-host) I0417 04:26:02.283536 136215914518336 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.031534s (batch_requests_ready=0.001948s, total_serialization_initiated=0.028882s, others=0.000704s) I0417 04:26:02.283624 136215914518336 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.035580s (all_items=0.000015s, per_item={'items': '0.00001550'}, temp_paths=0.035564) I0417 04:26:02.284344 136215914518336 event_tracking.py:125] [process=0] [async] Finished blocking save in 0.75 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/19. I0417 04:26:02.284625 136090167457536 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-17 04:46:02.284589 I0417 04:26:02.286559 136215914518336 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=19] Starting CheckpointManager Save Finalize thread=save_finalize I0417 04:26:02.286865 136089659594496 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0417 04:26:02.287001 136215914518336 standard_logger.py:34] {'step': 19, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776399932.2130644, 'wait_for_prev_duration_secs': 29.3176589012146, 'time_between_consecutive_saves_sec': 21.913447856903076, 'checkpointer_blocking_start_time': 1776399961.5325086, 'checkpointer_blocking_duration_secs': 0.7522461414337158, 'get_old_steps_start_time': 1776399962.284774, 'get_old_steps_duration_secs': 2.8848648071289062e-05, 'checkpoint_manager_blocking_start_time': 1776399932.2112327, 'checkpoint_manager_blocking_duration_secs': 30.075737476348877} I0417 04:26:02.287167 136215914518336 checkpointing.py:409] Started an asynchronous checkpoint save for step 19 I0417 04:26:02.287211 136215914518336 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=19][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0417 04:26:02.500657 136089624893184 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/19 I0417 04:26:05.115598 136090184242944 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/19/items I0417 04:26:08.745947 136090686035712 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_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/19/items/array_metadatas/process_0 I0417 04:26:47.429188 136090175850240 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 41.182863s (commit=39.958361s, array_metadata_write=1.224502s) I0417 04:26:47.430236 136090167457536 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 34.964 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 45.17818284034729 s) (per-host) I0417 04:26:47.430358 136090167457536 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 45.145626s. I0417 04:26:50.147279 136090167457536 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000304s. I0417 04:26:50.339797 136090167457536 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0417 04:26:51.286054 136090167457536 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.196712s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/19/items I0417 04:26:51.287807 136090167457536 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/19/items I0417 04:26:51.798782 136090167457536 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/19 I0417 04:26:53.203135 136090167457536 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/19`. I0417 04:26:53.204018 136090167457536 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 51.67 seconds @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_16_zarr3_true/checkpoints/19 I0417 04:26:53.205571 136090167457536 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 50.920836s. I0417 04:26:53.205717 136089659594496 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0417 04:26:53.205783 136089659594496 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0417 04:26:53.205856 136089659594496 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=19] CheckpointManager Save Finalize is syncing with other hosts... I0417 04:26:53.208423 136089659594496 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=19] CheckpointManager Save Finalize is done on all hosts. I0417 04:26:53.208545 136215914518336 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. I0417 04:26:53.208655 136215914518336 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0417 04:26:53.209419 136215914518336 metric_logger.py:185] completed step: 19, seconds: 0.148, TFLOP/s/device: 92.072, Tokens/s/device: 13878.160, 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: Fri Apr 17 04:27:05 UTC 2026 EXIT_CODE=0