XPK Start: Sun Apr 19 18:43:56 UTC 2026 PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used. 2026-04-19 18:44:20.637540: 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 18:44:21.180347 138769057781568 max_utils.py:273] Attempting to initialize the jax distributed system... I0419 18:44:30.221711 138769057781568 distributed.py:149] Starting JAX distributed service on [::]:8482 I0419 18:44:30.224092 138769057781568 distributed.py:172] Connecting to JAX distributed service on mt-09-pdb-lt-1-1eeo6-slice-job-0-0.mt-09-pdb-lt-1-1eeo6:8482 I0419 18:44:31.311464 138769057781568 max_utils.py:284] Jax distributed system initialized! I0419 18:44:37.423405 138769057781568 max_utils.py:800] System Information: Jax Version: 0.9.2 I0419 18:44:37.423521 138769057781568 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0419 18:44:37.423560 138769057781568 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 I0419 18:44:37.423593 138769057781568 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0419 18:44:38.130628 138769057781568 maxtext_utils.py:1551] Num_devices: 32, shape (1, 1, 1, 8, 1, 1, 1, 1, 4, 1, 1, 1, 1) I0419 18:44:38.130903 138769057781568 checkpointing.py:677] Setting up checkpoint logger... I0419 18:44:38.130961 138769057781568 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True I0419 18:44:38.131005 138769057781568 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None I0419 18:44:38.131343 138769057781568 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 0x7e351f1185c0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0419 18:44:42.112240 138769057781568 checkpointing.py:265] Enabling policy for fixed interval checkpointing. I0419 18:44:42.112500 138769057781568 checkpoint_manager.py:708] [process=6][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e21343f94f0>}, handler_registry=None I0419 18:44:42.112738 138769057781568 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e21343f94f0>` 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 18:44:42.112784 138769057781568 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e21340cc6e0>` 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 18:44:42.112819 138769057781568 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 0x7e21343f94f0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e21343f94f0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e21340cc6e0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e21340cc6e0>}). I0419 18:44:42.113124 138769057781568 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34 I0419 18:44:42.113198 138769057781568 async_checkpointer.py:192] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7e211c7c7920> timeout: 1200 secs and primary_host=0 for async checkpoint writes I0419 18:44:43.230743 138769057781568 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_09_pdb_lt_1/checkpoints I0419 18:44:43.682025 138769057781568 checkpoint_manager.py:929] [process=6][thread=MainThread] CheckpointManager created, primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=10), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_09_pdb_lt_1/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7e21343ee180> I0419 18:44:43.682195 138769057781568 checkpointing.py:301] Checkpoint manager created! I0419 18:44:44.606303 138769057781568 nnx_wrappers.py:437] Unknown Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0419 18:44:44.606427 138769057781568 nnx_wrappers.py:437] Unknown Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor'). I0419 18:44:44.983448 138769057781568 attentions.py:1088] attentions/inputs_q Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0419 18:44:44.983541 138769057781568 attentions.py:1088] attentions/inputs_q Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor'). I0419 18:44:45.000356 138769057781568 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0419 18:44:45.000427 138769057781568 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor'). I0419 18:44:45.024652 138769057781568 attentions.py:1154] attentions/query Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 18:44:45.024719 138769057781568 attentions.py:1154] attentions/query Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0419 18:44:45.041104 138769057781568 attentions.py:1155] attentions/key Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 18:44:45.041170 138769057781568 attentions.py:1155] attentions/key Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0419 18:44:45.057543 138769057781568 attentions.py:1156] attentions/value Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 18:44:45.057603 138769057781568 attentions.py:1156] attentions/value Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0419 18:44:45.082857 138769057781568 attentions.py:1197] attentions/out Logical: bfloat16[8,2048,16,128]..................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0419 18:44:45.082924 138769057781568 attentions.py:1197] attentions/out Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0419 18:44:45.103915 138769057781568 linears.py:525] linears/x Logical: bfloat16[8,2048,7168]....................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0419 18:44:45.103976 138769057781568 linears.py:525] linears/x Physical: bfloat16[8,2048,7168]....................................... ('fsdp', None, 'tensor'). I0419 18:44:45.319905 138769057781568 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint I0419 18:44:45.320018 138769057781568 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint. fsdp: 8 tensor: 4 I0419 18:44:46.721427 138769057781568 maxtext_utils.py:1654] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: P('norm',) Physical: ('tensor',) I0419 18:44:46.721560 138769057781568 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, 'tensor') I0419 18:44:46.721612 138769057781568 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, 'tensor') I0419 18:44:46.721667 138769057781568 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: P('mlp', 'layers', 'embed') Physical: ('tensor', None, 'fsdp') I0419 18:44:46.721718 138769057781568 maxtext_utils.py:1654] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: ('tensor', None) I0419 18:44:46.721755 138769057781568 maxtext_utils.py:1654] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: P('norm', 'layers') Physical: ('tensor', None) I0419 18:44:46.721803 138769057781568 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, 'tensor', None) I0419 18:44:46.721853 138769057781568 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: ('tensor', None, None, 'fsdp') I0419 18:44:46.721893 138769057781568 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, 'tensor', None) I0419 18:44:46.721929 138769057781568 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, 'tensor', None) I0419 18:44:46.721976 138769057781568 maxtext_utils.py:1654] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: P('embed_vocab', 'vocab') Physical: ('fsdp', 'tensor') I0419 18:44:46.722023 138769057781568 maxtext_utils.py:1654] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: P('vocab', 'embed_vocab') Physical: ('tensor', 'fsdp') I0419 18:44:47.205436 138769057781568 train.py:155] train/xent Logical: float32[8,2048]............................................. ('activation_embed_and_logits_batch', 'activation_length'). I0419 18:44:47.205529 138769057781568 train.py:155] train/xent Physical: float32[8,2048]............................................. ('fsdp', None). I0419 18:44:47.220938 138769057781568 train.py:162] train/z_loss Logical: float32[8,2048]............................................. ('activation_embed_and_logits_batch', 'activation_length'). I0419 18:44:47.220997 138769057781568 train.py:162] train/z_loss Physical: float32[8,2048]............................................. ('fsdp', None). I0419 18:45:02.345670 138769057781568 max_utils.py:791] Total memory size: 0.9 GB, Output size: 0.4 GB, Temp size: 0.5 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB. I0419 18:45:02.346377 138769057781568 metric_logger.py:301] number parameters: 1.104 billion I0419 18:45:03.497587 138769057781568 checkpointing.py:772] Waiting for step 0 to finish before checkpoint... I0419 18:45:18.044322 138769057781568 checkpointing.py:776] Waited 14.546715021133423 seconds for step 0 to finish before starting checkpointing. I0419 18:45:18.046687 138769057781568 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0419 18:45:18.048706 138769057781568 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 0 I0419 18:45:18.050307 138769057781568 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_09_pdb_lt_1/checkpoints/0. I0419 18:45:18.386682 138769057781568 signaling_client.py:364] Using JaxDistributedSignalingClient I0419 18:45:18.387741 138769057781568 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0419 18:45:18.387795 138769057781568 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 18:45:18.653948 138769057781568 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.267337s I0419 18:45:18.654124 138769057781568 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 5.654 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.27283263206481934 s) (per-host) I0419 18:45:18.654179 138769057781568 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.272900s (batch_requests_ready=0.002260s, total_serialization_initiated=0.270565s, others=0.000075s) I0419 18:45:18.654270 138769057781568 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.277383s (all_items=0.000020s, per_item={'items': '0.00002003'}, temp_paths=0.277363) I0419 18:45:18.655115 138769057781568 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.61 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_09_pdb_lt_1/checkpoints/0. I0419 18:45:18.655477 138642640144128 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-19 19:05:18.655439 I0419 18:45:18.663153 138769057781568 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0419 18:45:18.663446 138642139899648 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0419 18:45:18.663602 138769057781568 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_09_pdb_lt_1/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776624318.046668, 'wait_for_prev_duration_secs': 6.270408630371094e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776624318.0487452, 'checkpointer_blocking_duration_secs': 0.606891393661499, 'get_old_steps_start_time': 1776624318.6556642, 'get_old_steps_duration_secs': 2.86102294921875e-05, 'checkpoint_manager_blocking_start_time': 1776624318.0448768, 'checkpoint_manager_blocking_duration_secs': 0.6186869144439697} I0419 18:45:18.663712 138769057781568 checkpointing.py:408] Started an asynchronous checkpoint save for step 0 I0419 18:45:18.663762 138769057781568 max_utils.py:750] Memstats: After params initialized: I0419 18:45:18.663812 138769057781568 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_24(process=6,(0,6,0,0)) I0419 18:45:18.663846 138769057781568 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_25(process=6,(1,6,0,0)) I0419 18:45:18.663874 138769057781568 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_28(process=6,(0,7,0,0)) I0419 18:45:18.663901 138769057781568 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_29(process=6,(1,7,0,0)) I0419 18:45:18.979697 138769057781568 metric_logger.py:196] completed step: 0, seconds: 1.151, TFLOP/s/device: 2.951, Tokens/s/device: 444.800, total_weights: 16384, loss: 10.862, lm_loss: 10.862, perplexity: 52153.723 I0419 18:45:19.065922 138769057781568 metric_logger.py:196] completed step: 1, seconds: 15.481, TFLOP/s/device: 0.219, Tokens/s/device: 33.074, total_weights: 16384, loss: 10.862, lm_loss: 10.862, perplexity: 52153.723 I0419 18:45:19.471465 138769057781568 metric_logger.py:196] completed step: 2, seconds: 0.028, TFLOP/s/device: 121.214, Tokens/s/device: 18270.706, total_weights: 16384, loss: 9.763, lm_loss: 9.763, perplexity: 17373.129 I0419 18:45:19.529854 138769057781568 metric_logger.py:196] completed step: 3, seconds: 0.406, TFLOP/s/device: 8.371, Tokens/s/device: 1261.705, total_weights: 16384, loss: 8.823, lm_loss: 8.823, perplexity: 6791.817 I0419 18:45:19.647056 138769057781568 metric_logger.py:196] completed step: 4, seconds: 0.064, TFLOP/s/device: 52.931, Tokens/s/device: 7978.309, total_weights: 16384, loss: 7.947, lm_loss: 7.947, perplexity: 2828.417 I0419 18:45:19.653026 138769057781568 metric_logger.py:196] completed step: 5, seconds: 0.058, TFLOP/s/device: 58.520, Tokens/s/device: 8820.743, total_weights: 16384, loss: 7.212, lm_loss: 7.212, perplexity: 1355.339 I0419 18:45:22.496130 2502 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0419 18:45:24.536906 138642148292352 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_09_pdb_lt_1/checkpoints/0/items/array_metadatas/process_6 I0419 18:45:33.724074 138769057781568 metric_logger.py:196] completed step: 6, seconds: 0.118, TFLOP/s/device: 28.836, Tokens/s/device: 4346.497, total_weights: 16384, loss: 6.652, lm_loss: 6.652, perplexity: 774.659 I0419 18:45:33.782287 138769057781568 metric_logger.py:196] completed step: 7, seconds: 14.013, TFLOP/s/device: 0.242, Tokens/s/device: 36.537, total_weights: 16384, loss: 6.269, lm_loss: 6.269, perplexity: 527.696 I0419 18:45:33.840532 138769057781568 metric_logger.py:196] completed step: 8, seconds: 0.063, TFLOP/s/device: 53.754, Tokens/s/device: 8102.420, total_weights: 16384, loss: 6.031, lm_loss: 6.031, perplexity: 416.165 I0419 18:45:33.898342 138769057781568 checkpointing.py:772] Waiting for step 9 to finish before checkpoint... I0419 18:45:33.899002 138769057781568 checkpointing.py:776] Waited 0.0006768703460693359 seconds for step 9 to finish before starting checkpointing. I0419 18:45:33.901119 138769057781568 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 18:45:54.947663 138642640144128 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 43.198 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 36.566333532333374 s) (per-host) I0419 18:45:54.947788 138642640144128 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 36.292197s. I0419 18:46:04.113462 138642640144128 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 45.457856s. I0419 18:46:04.113742 138642139899648 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 18:46:04.113863 138642139899648 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0419 18:46:04.113915 138642139899648 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0419 18:46:04.115732 138642139899648 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0419 18:46:04.115891 138769057781568 checkpoint_manager.py:2032] [process=6][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0419 18:46:04.116020 138769057781568 checkpoint_manager.py:1452] Waiting for previous save to complete took 30.214900 seconds. If this number is high, consider checkpointing less frequently. I0419 18:46:04.117747 138769057781568 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 9 I0419 18:46:04.119875 138769057781568 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_09_pdb_lt_1/checkpoints/9. I0419 18:46:04.424122 138769057781568 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array. I0419 18:46:04.424214 138769057781568 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 18:46:04.451041 138769057781568 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.027939s I0419 18:46:04.451172 138769057781568 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 49.293 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.03129386901855469 s) (per-host) I0419 18:46:04.451221 138769057781568 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.031352s (batch_requests_ready=0.001767s, total_serialization_initiated=0.029521s, others=0.000063s) I0419 18:46:04.451299 138769057781568 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.035372s (all_items=0.000016s, per_item={'items': '0.00001574'}, temp_paths=0.035356) I0419 18:46:04.451909 138769057781568 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.33 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_09_pdb_lt_1/checkpoints/9. I0419 18:46:04.452195 138641066694400 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-19 19:06:04.452155 I0419 18:46:04.454601 138769057781568 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0419 18:46:04.454828 138642139899648 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0419 18:46:04.454984 138769057781568 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_09_pdb_lt_1/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776624333.9010885, 'wait_for_prev_duration_secs': 30.214900255203247, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776624364.1177864, 'checkpointer_blocking_duration_secs': 0.33455896377563477, 'get_old_steps_start_time': 1776624364.4523644, 'get_old_steps_duration_secs': 2.6702880859375e-05, 'checkpoint_manager_blocking_start_time': 1776624333.8992267, 'checkpoint_manager_blocking_duration_secs': 30.55572533607483} I0419 18:46:04.455089 138769057781568 checkpointing.py:408] Started an asynchronous checkpoint save for step 9 I0419 18:46:04.455133 138769057781568 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 18:46:10.467768 138645906388736 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260419_180002/linen_xpk_main_20260419_180002_09_pdb_lt_1/checkpoints/9/items/array_metadatas/process_6 I0419 18:46:46.179295 138641066694400 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 37.826 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.75937533378601 s) (per-host) I0419 18:46:46.179445 138641066694400 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.727128s. I0419 18:46:55.706450 138641066694400 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 51.254119s. I0419 18:46:55.706735 138642139899648 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 18:46:55.706862 138642139899648 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0419 18:46:55.706908 138642139899648 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0419 18:46:55.708376 138642139899648 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0419 18:46:55.708567 138769057781568 checkpoint_manager.py:2032] [process=6][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0419 18:46:55.708714 138769057781568 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0419 18:46:55.709583 138769057781568 metric_logger.py:196] completed step: 9, seconds: 0.058, TFLOP/s/device: 58.408, Tokens/s/device: 8803.907, total_weights: 16384, loss: 5.892, lm_loss: 5.892, perplexity: 361.955 Per train step: Total TFLOPs: 3.40 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Sun Apr 19 18:47:07 UTC 2026 EXIT_CODE=0