XPK Start: Sat Apr 18 01:58:26 UTC 2026
PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.
2026-04-18 01:58:50.728012: E external/local_xla/xla/stream_executor/cuda/cuda_platform.cc:51] failed call to cuInit: INTERNAL: CUDA error: Failed call to cuInit: UNKNOWN ERROR (303)
I0418 01:58:51.262096 140406617786176 max_utils.py:273] Attempting to initialize the jax distributed system...
I0418 01:59:00.303195 140406617786176 distributed.py:149] Starting JAX distributed service on [::]:8482
I0418 01:59:00.305568 140406617786176 distributed.py:172] Connecting to JAX distributed service on mt-16-zarr3-false-rs-87fwx-slice-job-0-0.mt-16-zarr3-false-rs-87fwx:8482
I0418 01:59:00.970693 140406617786176 max_utils.py:284] Jax distributed system initialized!
I0418 01:59:07.151144 140406617786176 max_utils.py:800] System Information: Jax Version: 0.9.2
I0418 01:59:07.151249 140406617786176 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0418 01:59:07.151291 140406617786176 max_utils.py:802] System Information: Jax Backend: PJRT C API
TFRT TPU v6 lite
Built on Mar 4 2026 11:32:08 (1772652728) cl/878335365
I0418 01:59:07.151326 140406617786176 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0418 01:59:07.835043 140406617786176 maxtext_utils.py:1551] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0418 01:59:07.835320 140406617786176 checkpointing.py:677] Setting up checkpoint logger...
I0418 01:59:07.835373 140406617786176 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=False
I0418 01:59:07.835417 140406617786176 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0418 01:59:07.835748 140406617786176 base_pytree_checkpoint_handler.py:441] Created BasePyTreeCheckpointHandler: use_ocdbt=True, use_zarr3=False, pytree_metadata_options=PyTreeMetadataOptions(support_rich_types=False), array_metadata_store=<orbax.checkpoint._src.metadata.array_metadata_store.Store object at 0x7fb2603ef5f0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0418 01:59:09.114739 140406617786176 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0418 01:59:09.115002 140406617786176 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 0x7f9e703f4ef0>}, handler_registry=None
I0418 01:59:09.115242 140406617786176 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7f9e703f4ef0>` for item "items" and save args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>` to `_handler_registry`.
I0418 01:59:09.115290 140406617786176 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7f9e705c6cf0>` for item "metrics" and save args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>` and restore args `<class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>` to `_handler_registry`.
I0418 01:59:09.115324 140406617786176 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 0x7f9e703f4ef0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7f9e703f4ef0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7f9e705c6cf0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7f9e705c6cf0>}).
I0418 01:59:09.115642 140406617786176 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0418 01:59:09.115713 140406617786176 async_checkpointer.py:192] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7f9e3870fc40> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0418 01:59:12.388463 140286416107264 checkpoint.py:228] Read Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776477381500788402, 'commit_timestamp_nsecs': 1776477429277019669, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/9/_CHECKPOINT_METADATA
I0418 01:59:12.638750 140286416107264 checkpoint_manager.py:1871] Missing metrics for step 9
I0418 01:59:13.035545 140406617786176 checkpoint_manager.py:1812] Found 3 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints
I0418 01:59:14.781749 140406617786176 checkpoint_manager.py:929] [process=0][thread=MainThread] CheckpointManager created, primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=5), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7f9e70717680>
I0418 01:59:14.781957 140406617786176 checkpointing.py:301] Checkpoint manager created!
I0418 01:59:15.691169 140406617786176 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0418 01:59:15.691283 140406617786176 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 01:59:16.070097 140406617786176 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 01:59:16.070189 140406617786176 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 01:59:16.086444 140406617786176 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 01:59:16.086511 140406617786176 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 01:59:16.110008 140406617786176 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0418 01:59:16.110078 140406617786176 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 01:59:16.126210 140406617786176 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0418 01:59:16.126268 140406617786176 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 01:59:16.142457 140406617786176 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0418 01:59:16.142518 140406617786176 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 01:59:16.167690 140406617786176 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0418 01:59:16.167763 140406617786176 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 01:59:16.188676 140406617786176 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0418 01:59:16.188741 140406617786176 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0418 01:59:16.404910 140406617786176 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0418 01:59:16.405020 140406617786176 checkpointing.py:581] restoring from this run's directory step 9
I0418 01:59:16.644518 140406617786176 event_tracking.py:70] [process=0] [sync] Started load checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/9.
I0418 01:59:16.894152 140406617786176 checkpointer.py:307] Restoring checkpoint from gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/9.
I0418 01:59:16.894574 140406617786176 event_tracking.py:125] [process=0] [sync] Finished blocking load in 0.25 seconds. Continuing load @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/9.
I0418 01:59:18.454940 1437 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0418 01:59:24.506656 140406617786176 jax_array_handlers.py:843] [process=0] /jax/orbax/read/worker/io/requested throughput: 260.879 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 6.0574867725372314 s) (per-host)
I0418 01:59:24.520987 140406617786176 base_pytree_checkpoint_handler.py:130] [process=0] /jax/checkpoint/read/gbytes_per_sec: 242.988 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 6.513057231903076 s) (per-host)
I0418 01:59:28.819436 140406617786176 event_tracking.py:138] [process=0] [sync] Finished load in 12.17 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/9
I0418 01:59:28.819592 140406617786176 standard_logger.py:34] {'step': 9, 'event_type': 'restore', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints', 'checkpointer_start_time': 1776477556.6444185, 'checkpointer_duration_secs': 12.17514181137085, 'checkpoint_manager_start_time': 1776477556.4052658, 'checkpoint_manager_duration_secs': 12.414296865463257}
fsdp: 32
I0418 01:59:29.233758 140406617786176 maxtext_utils.py:1654] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: P('norm',)
Physical: (None,)
I0418 01:59:29.233894 140406617786176 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wi_0/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0418 01:59:29.233949 140406617786176 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wi_1/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0418 01:59:29.234007 140406617786176 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wo/kernel
Shape: float32[7168,16,2048]
Logical: P('mlp', 'layers', 'embed')
Physical: (None, None, 'fsdp')
I0418 01:59:29.234058 140406617786176 maxtext_utils.py:1654] params/params/decoder/layers/post_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0418 01:59:29.234096 140406617786176 maxtext_utils.py:1654] params/params/decoder/layers/pre_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0418 01:59:29.234153 140406617786176 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/key/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'kv_heads', 'kv_head_dim')
Physical: ('fsdp', None, None, None)
I0418 01:59:29.234206 140406617786176 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/out/kernel
Shape: float32[16,16,128,2048]
Logical: P('heads', 'layers', 'kv', 'embed')
Physical: (None, None, None, 'fsdp')
I0418 01:59:29.234246 140406617786176 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/query/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'q_heads', 'kv')
Physical: ('fsdp', None, None, None)
I0418 01:59:29.234281 140406617786176 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/value/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'kv_heads', 'kv_head_dim')
Physical: ('fsdp', None, None, None)
I0418 01:59:29.234326 140406617786176 maxtext_utils.py:1654] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: P('embed_vocab', 'vocab')
Physical: ('fsdp', None)
I0418 01:59:29.234372 140406617786176 maxtext_utils.py:1654] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: P('vocab', 'embed_vocab')
Physical: (None, 'fsdp')
I0418 01:59:29.721757 140406617786176 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0418 01:59:29.721863 140406617786176 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0418 01:59:29.736958 140406617786176 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0418 01:59:29.737020 140406617786176 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0418 01:59:40.712643 140406617786176 max_utils.py:791] Total memory size: 1.5 GB, Output size: 0.4 GB, Temp size: 1.1 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB.
I0418 01:59:40.720287 140406617786176 metric_logger.py:301] number parameters: 1.104 billion
I0418 01:59:52.235217 140406617786176 checkpointing.py:772] Waiting for step 10 to finish before checkpoint...
I0418 01:59:52.391002 140406617786176 checkpointing.py:776] Waited 0.15576457977294922 seconds for step 10 to finish before starting checkpointing.
I0418 01:59:52.393419 140406617786176 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 01:59:52.395261 140406617786176 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 10
I0418 01:59:52.397398 140406617786176 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/10.
I0418 01:59:53.553915 140406617786176 signaling_client.py:364] Using JaxDistributedSignalingClient
I0418 01:59:53.566580 140406617786176 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 01:59:53.566644 140406617786176 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False
I0418 01:59:53.837038 140406617786176 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.271176s
I0418 01:59:53.837903 140406617786176 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 5.589 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2760138511657715 s) (per-host)
I0418 01:59:53.837985 140406617786176 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.276112s (batch_requests_ready=0.002162s, total_serialization_initiated=0.273162s, others=0.000788s)
I0418 01:59:53.838067 140406617786176 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.280236s (all_items=0.000017s, per_item={'items': '0.00001693'}, temp_paths=0.280219)
I0418 01:59:53.838832 140406617786176 event_tracking.py:125] [process=0] [async] Finished blocking save in 1.44 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/10.
I0418 01:59:53.839170 140285820520192 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 02:19:53.839131
I0418 01:59:53.852179 140406617786176 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=10] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 01:59:53.852459 140281501071104 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0418 01:59:53.852608 140406617786176 standard_logger.py:34] {'step': 10, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776477592.3934007, 'wait_for_prev_duration_secs': 5.984306335449219e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776477592.395299, 'checkpointer_blocking_duration_secs': 1.4440150260925293, 'get_old_steps_start_time': 1776477593.8393369, 'get_old_steps_duration_secs': 3.147125244140625e-05, 'checkpoint_manager_blocking_start_time': 1776477592.3915818, 'checkpoint_manager_blocking_duration_secs': 1.4609897136688232}
I0418 01:59:53.852707 140406617786176 checkpointing.py:408] Started an asynchronous checkpoint save for step 10
I0418 01:59:53.852757 140406617786176 max_utils.py:750]
Memstats: After params initialized:
I0418 01:59:53.852806 140406617786176 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0))
I0418 01:59:53.852855 140406617786176 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0))
I0418 01:59:53.852885 140406617786176 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0))
I0418 01:59:53.852911 140406617786176 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0))
I0418 01:59:54.191153 140406617786176 metric_logger.py:196] completed step: 10, seconds: 11.206, TFLOP/s/device: 1.212, Tokens/s/device: 182.759, total_weights: 65536, loss: 6.833, lm_loss: 6.833, perplexity: 927.869
I0418 01:59:54.253287 140284763563776 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/10
I0418 01:59:54.336409 140406617786176 metric_logger.py:196] completed step: 11, seconds: 1.954, TFLOP/s/device: 6.953, Tokens/s/device: 1048.075, total_weights: 65536, loss: 6.527, lm_loss: 6.527, perplexity: 683.659
I0418 01:59:54.484534 140406617786176 metric_logger.py:196] completed step: 12, seconds: 0.010, TFLOP/s/device: 1312.766, Tokens/s/device: 197874.396, total_weights: 65536, loss: 6.265, lm_loss: 6.265, perplexity: 526.068
I0418 01:59:54.632121 140406617786176 metric_logger.py:196] completed step: 13, seconds: 0.145, TFLOP/s/device: 93.554, Tokens/s/device: 14101.478, total_weights: 65536, loss: 6.045, lm_loss: 6.045, perplexity: 422.001
I0418 01:59:54.641037 140406617786176 checkpointing.py:772] Waiting for step 15 to finish before checkpoint...
I0418 01:59:54.926046 140406617786176 checkpointing.py:776] Waited 0.2848656177520752 seconds for step 15 to finish before starting checkpointing.
I0418 01:59:54.928975 140406617786176 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=10][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 01:59:56.902359 140282063156992 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/10/items
I0418 01:59:57.061052 140281492678400 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776477596656911978, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776477596656911978, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/10/_CHECKPOINT_METADATA
I0418 01:59:59.798520 140281550731008 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/10/items/array_metadatas/process_0
I0418 02:00:40.930908 140282046371584 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 42.880976s (commit=41.651881s, array_metadata_write=1.229095s)
I0418 02:00:40.932418 140285820520192 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 33.346 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 47.37048840522766 s) (per-host)
I0418 02:00:40.932587 140285820520192 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 47.093304s.
I0418 02:00:42.883612 140285820520192 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000401s.
I0418 02:00:42.927659 140281492678400 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776477596656911978, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/10/_CHECKPOINT_METADATA
I0418 02:00:43.899592 140285820520192 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.041301s. use_zarr3=False, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/10/items
I0418 02:00:43.901885 140285820520192 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/10/items
I0418 02:00:44.456087 140285820520192 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/10
I0418 02:00:46.306542 140285820520192 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/10`.
I0418 02:00:46.307686 140285820520192 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 53.91 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/10
I0418 02:00:46.309105 140285820520192 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 52.469819s.
I0418 02:00:46.309329 140281501071104 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 02:00:46.309444 140281501071104 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0418 02:00:46.309552 140281501071104 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=10] CheckpointManager Save Finalize is syncing with other hosts...
I0418 02:00:46.311255 140281501071104 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=10] CheckpointManager Save Finalize is done on all hosts.
I0418 02:00:46.311572 140406617786176 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=10][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=10.
W0418 02:00:46.311751 140406617786176 checkpoint_manager.py:1452] Waiting for previous save to complete took 51.382777 seconds. If this number is high, consider checkpointing less frequently.
I0418 02:00:46.313621 140406617786176 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 15
I0418 02:00:46.316169 140406617786176 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/15.
I0418 02:00:47.450640 140406617786176 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 02:00:47.450770 140406617786176 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False
I0418 02:00:47.478739 140406617786176 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.029016s
I0418 02:00:47.479772 140406617786176 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 40.974 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.03764748573303223 s) (per-host)
I0418 02:00:47.479901 140406617786176 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.037796s (batch_requests_ready=0.004006s, total_serialization_initiated=0.032841s, others=0.000950s)
I0418 02:00:47.480075 140406617786176 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.042881s (all_items=0.000022s, per_item={'items': '0.00002217'}, temp_paths=0.042859)
I0418 02:00:47.481088 140406617786176 event_tracking.py:125] [process=0] [async] Finished blocking save in 1.17 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/15.
I0418 02:00:47.481436 140281501071104 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 02:20:47.481408
I0418 02:00:47.483867 140406617786176 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=15] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 02:00:47.484177 140282054764288 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0418 02:00:47.484361 140406617786176 standard_logger.py:34] {'step': 15, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776477594.9289303, 'wait_for_prev_duration_secs': 51.382776975631714, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776477646.3136892, 'checkpointer_blocking_duration_secs': 1.1678848266601562, 'get_old_steps_start_time': 1776477647.481603, 'get_old_steps_duration_secs': 4.506111145019531e-05, 'checkpoint_manager_blocking_start_time': 1776477594.9266908, 'checkpoint_manager_blocking_duration_secs': 52.55761432647705}
I0418 02:00:47.484507 140406617786176 checkpointing.py:408] Started an asynchronous checkpoint save for step 15
I0418 02:00:47.688318 140285820520192 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/15
I0418 02:00:49.449022 140280425817856 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/15/items
I0418 02:00:54.502995 140282574116608 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/15/items/array_metadatas/process_0
I0418 02:01:08.858469 140406617786176 metric_logger.py:196] completed step: 14, seconds: 0.148, TFLOP/s/device: 91.658, Tokens/s/device: 13815.714, total_weights: 65536, loss: 5.864, lm_loss: 5.864, perplexity: 352.193
I0418 02:01:08.867595 140406617786176 metric_logger.py:196] completed step: 15, seconds: 0.148, TFLOP/s/device: 91.914, Tokens/s/device: 13854.313, total_weights: 65536, loss: 5.719, lm_loss: 5.719, perplexity: 304.661
I0418 02:01:09.013737 140406617786176 metric_logger.py:196] completed step: 16, seconds: 74.225, TFLOP/s/device: 0.183, Tokens/s/device: 27.592, total_weights: 65536, loss: 5.605, lm_loss: 5.605, perplexity: 271.872
I0418 02:01:09.160852 140406617786176 metric_logger.py:196] completed step: 17, seconds: 0.007, TFLOP/s/device: 2004.001, Tokens/s/device: 302064.897, total_weights: 65536, loss: 5.517, lm_loss: 5.517, perplexity: 248.975
I0418 02:01:09.308147 140406617786176 metric_logger.py:196] completed step: 18, seconds: 0.148, TFLOP/s/device: 92.114, Tokens/s/device: 13884.369, total_weights: 65536, loss: 5.448, lm_loss: 5.448, perplexity: 232.333
I0418 02:01:09.454864 140406617786176 checkpointing.py:772] Waiting for step 19 to finish before checkpoint...
I0418 02:01:09.455588 140406617786176 checkpointing.py:776] Waited 0.0007696151733398438 seconds for step 19 to finish before starting checkpointing.
I0418 02:01:09.457635 140406617786176 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=15][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 02:01:33.570650 140282063156992 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 42.132051s (commit=40.457256s, array_metadata_write=1.674795s)
I0418 02:01:33.571682 140281501071104 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 34.243 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 46.12953329086304 s) (per-host)
I0418 02:01:33.571808 140281501071104 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 46.090270s.
I0418 02:01:35.862790 140281501071104 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000230s.
I0418 02:01:36.745403 140281501071104 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.929708s. use_zarr3=False, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/15/items
I0418 02:01:36.747148 140281501071104 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/15/items
I0418 02:01:37.909206 140281501071104 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/15
I0418 02:01:39.717221 140281501071104 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/15`.
I0418 02:01:39.718101 140281501071104 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 53.40 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/15
I0418 02:01:39.719825 140281501071104 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 52.238283s.
I0418 02:01:39.720107 140282054764288 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 02:01:39.720227 140282054764288 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0418 02:01:39.720297 140282054764288 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=15] CheckpointManager Save Finalize is syncing with other hosts...
I0418 02:01:39.725161 140282054764288 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=15] CheckpointManager Save Finalize is done on all hosts.
I0418 02:01:39.725342 140406617786176 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=15][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=15.
W0418 02:01:39.725463 140406617786176 checkpoint_manager.py:1452] Waiting for previous save to complete took 30.267836 seconds. If this number is high, consider checkpointing less frequently.
I0418 02:01:39.726881 140406617786176 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 19
I0418 02:01:39.728941 140406617786176 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/19.
I0418 02:01:40.434459 140406617786176 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 02:01:40.434556 140406617786176 replica_slices.py:424] Transferring arrays to host memory with options: use_replica_parallel=True, min_slice_bytes_for_replica_parallel=None, max_replicas_for_replica_parallel=None, enable_pinned_host_transfer=False
I0418 02:01:40.460304 140406617786176 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.026822s
I0418 02:01:40.461034 140406617786176 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 49.546 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.03113412857055664 s) (per-host)
I0418 02:01:40.461112 140406617786176 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.031232s (batch_requests_ready=0.001763s, total_serialization_initiated=0.028777s, others=0.000692s)
I0418 02:01:40.461189 140406617786176 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.035201s (all_items=0.000015s, per_item={'items': '0.00001526'}, temp_paths=0.035186)
I0418 02:01:40.461861 140406617786176 event_tracking.py:125] [process=0] [async] Finished blocking save in 0.73 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/19.
I0418 02:01:40.462149 140282054764288 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 02:21:40.462116
I0418 02:01:40.464851 140406617786176 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=19] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 02:01:40.465155 140282574116608 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0418 02:01:40.465318 140406617786176 standard_logger.py:34] {'step': 19, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776477669.4576, 'wait_for_prev_duration_secs': 30.267836332321167, 'time_between_consecutive_saves_sec': 23.146275520324707, 'checkpointer_blocking_start_time': 1776477699.7269201, 'checkpointer_blocking_duration_secs': 0.7353577613830566, 'get_old_steps_start_time': 1776477700.4622958, 'get_old_steps_duration_secs': 2.8848648071289062e-05, 'checkpoint_manager_blocking_start_time': 1776477669.455829, 'checkpoint_manager_blocking_duration_secs': 31.009455680847168}
I0418 02:01:40.465421 140406617786176 checkpointing.py:408] Started an asynchronous checkpoint save for step 19
I0418 02:01:40.465464 140406617786176 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=19][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 02:01:40.686616 140281501071104 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/19
I0418 02:01:43.716986 140282046371584 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/19/items
I0418 02:01:48.290513 140285812127488 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/19/items/array_metadatas/process_0
I0418 02:02:30.180702 140284763563776 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 44.492877s (commit=40.726691s, array_metadata_write=3.766186s)
I0418 02:02:30.181653 140282054764288 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 31.750 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 49.75172710418701 s) (per-host)
I0418 02:02:30.181768 140282054764288 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 49.719513s.
I0418 02:02:34.662205 140282054764288 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000207s.
I0418 02:02:35.026374 140282054764288 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.207108s. use_zarr3=False, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/19/items
I0418 02:02:35.028000 140282054764288 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/19/items
I0418 02:02:36.020214 140282054764288 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/19
I0418 02:02:38.726869 140282054764288 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/19`.
I0418 02:02:38.727695 140282054764288 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 59.00 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_16_zarr3_false/checkpoints/19
I0418 02:02:38.729150 140282054764288 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 58.266892s.
I0418 02:02:38.729357 140282574116608 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 02:02:38.729476 140282574116608 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0418 02:02:38.729534 140282574116608 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=19] CheckpointManager Save Finalize is syncing with other hosts...
I0418 02:02:38.731291 140282574116608 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=19] CheckpointManager Save Finalize is done on all hosts.
I0418 02:02:38.731456 140406617786176 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=19][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=19.
I0418 02:02:38.731606 140406617786176 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 02:02:38.732530 140406617786176 metric_logger.py:196] completed step: 19, seconds: 0.147, TFLOP/s/device: 92.356, Tokens/s/device: 13920.893, total_weights: 65536, loss: 5.391, lm_loss: 5.391, perplexity: 219.349
Per train step:
Total TFLOPs: 13.59
split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Sat Apr 18 02:02:47 UTC 2026
EXIT_CODE=0