MaxView

← Back to run

Log Summary

XPK Start: Mon Apr 20 06:27:17 UTC 2026
PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.
2026-04-20 06:27:41.435146: 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)
I0420 06:27:41.616963 134467927693120 max_utils.py:273] Attempting to initialize the jax distributed system...
I0420 06:27:50.656934 134467927693120 distributed.py:149] Starting JAX distributed service on [::]:8482
I0420 06:27:50.659268 134467927693120 distributed.py:172] Connecting to JAX distributed service on mt-08-checkpoint-async--qwx2v-slice-job-0-0.mt-08-checkpoint-async--qwx2v:8482
I0420 06:27:52.058523 134467927693120 max_utils.py:284] Jax distributed system initialized!
I0420 06:27:57.172904 134467927693120 max_utils.py:800] System Information: Jax Version: 0.9.2
I0420 06:27:57.173008 134467927693120 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0420 06:27:57.173069 134467927693120 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
I0420 06:27:57.173105 134467927693120 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0420 06:27:57.866281 134467927693120 maxtext_utils.py:1551] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0420 06:27:57.866563 134467927693120 checkpointing.py:677] Setting up checkpoint logger...
I0420 06:27:57.866619 134467927693120 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0420 06:27:57.866664 134467927693120 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0420 06:27:57.867008 134467927693120 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 0x7a4b89650050>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0420 06:27:59.428253 134467927693120 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0420 06:27:59.428498 134467927693120 checkpoint_manager.py:708] [process=2][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7a4a787221b0>}, handler_registry=None
I0420 06:27:59.428733 134467927693120 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7a4a787221b0>` 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`.
I0420 06:27:59.428781 134467927693120 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7a37f4786ff0>` 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`.
I0420 06:27:59.428814 134467927693120 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 0x7a4a787221b0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7a4a787221b0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7a37f4786ff0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7a37f4786ff0>}).
I0420 06:27:59.429167 134467927693120 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0420 06:27:59.429239 134467927693120 async_checkpointer.py:192] [process=2][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7a371439bb00> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0420 06:28:00.996079 134349294909184 checkpoint.py:228] Read Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776666318066035915, 'commit_timestamp_nsecs': 1776666364061223918, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_08_checkpoint_async_true/checkpoints/9/_CHECKPOINT_METADATA
I0420 06:28:01.240521 134349294909184 checkpoint_manager.py:1871] Missing metrics for step 9
I0420 06:28:01.288969 134467927693120 checkpoint_manager.py:1812] Found 3 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_08_checkpoint_async_true/checkpoints
I0420 06:28:01.948099 134467927693120 checkpoint_manager.py:929] [process=2][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_20260420_060028/linen_xpk_main_20260420_060028_08_checkpoint_async_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7a37f4785130>
I0420 06:28:01.948269 134467927693120 checkpointing.py:301] Checkpoint manager created!
I0420 06:28:02.871465 134467927693120 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0420 06:28:02.871581 134467927693120 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 06:28:03.248115 134467927693120 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0420 06:28:03.248204 134467927693120 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 06:28:03.264515 134467927693120 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0420 06:28:03.264576 134467927693120 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 06:28:03.287910 134467927693120 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 06:28:03.287977 134467927693120 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 06:28:03.304283 134467927693120 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 06:28:03.304344 134467927693120 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 06:28:03.320433 134467927693120 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 06:28:03.320492 134467927693120 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 06:28:03.345564 134467927693120 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0420 06:28:03.345643 134467927693120 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 06:28:03.372437 134467927693120 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0420 06:28:03.372545 134467927693120 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0420 06:28:03.596259 134467927693120 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0420 06:28:03.596369 134467927693120 checkpointing.py:581] restoring from this run's directory step 9
I0420 06:28:03.845498 134467927693120 event_tracking.py:70] [process=2] [sync] Started load checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_08_checkpoint_async_true/checkpoints/9.
I0420 06:28:04.398594 134467927693120 checkpointer.py:307] Restoring checkpoint from gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_08_checkpoint_async_true/checkpoints/9.
I0420 06:28:04.399038 134467927693120 event_tracking.py:125] [process=2] [sync] Finished blocking load in 0.55 seconds. Continuing load @ gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_08_checkpoint_async_true/checkpoints/9.
I0420 06:28:05.794677    1409 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0420 06:28:13.461568 134467927693120 jax_array_handlers.py:843] [process=2] /jax/orbax/read/worker/io/requested throughput: 205.955 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 7.672889947891235 s) (per-host)
I0420 06:28:13.466765 134467927693120 base_pytree_checkpoint_handler.py:130] [process=2] /jax/checkpoint/read/gbytes_per_sec: 196.141 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 8.068634271621704 s) (per-host)
I0420 06:28:13.469078 134467927693120 event_tracking.py:138] [process=2] [sync] Finished load in 9.62 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_08_checkpoint_async_true/checkpoints/9
I0420 06:28:13.469171 134467927693120 standard_logger.py:34] {'step': 9, 'event_type': 'restore', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_08_checkpoint_async_true/checkpoints', 'checkpointer_start_time': 1776666483.845397, 'checkpointer_duration_secs': 9.623747110366821, 'checkpoint_manager_start_time': 1776666483.59661, 'checkpoint_manager_duration_secs': 9.872535943984985}
fsdp: 32
I0420 06:28:13.888524 134467927693120 maxtext_utils.py:1654]  params/params/decoder/decoder_norm/scale
    Shape:     float32[2048]
    Logical:   P('norm',)
    Physical:  (None,)
I0420 06:28:13.888645 134467927693120 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)
I0420 06:28:13.888698 134467927693120 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)
I0420 06:28:13.888756 134467927693120 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')
I0420 06:28:13.888811 134467927693120 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)
I0420 06:28:13.888849 134467927693120 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)
I0420 06:28:13.888901 134467927693120 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)
I0420 06:28:13.888955 134467927693120 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')
I0420 06:28:13.888996 134467927693120 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)
I0420 06:28:13.889049 134467927693120 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)
I0420 06:28:13.889122 134467927693120 maxtext_utils.py:1654]  params/params/decoder/logits_dense/kernel
    Shape:     float32[2048,32000]
    Logical:   P('embed_vocab', 'vocab')
    Physical:  ('fsdp', None)
I0420 06:28:13.889178 134467927693120 maxtext_utils.py:1654]  params/params/token_embedder/embedding
    Shape:     float32[32000,2048]
    Logical:   P('vocab', 'embed_vocab')
    Physical:  (None, 'fsdp')

I0420 06:28:14.371991 134467927693120 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0420 06:28:14.372104 134467927693120 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0420 06:28:14.387274 134467927693120 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0420 06:28:14.387334 134467927693120 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0420 06:28:25.480213 134467927693120 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.
I0420 06:28:25.481108 134467927693120 metric_logger.py:301] number parameters: 1.104 billion
I0420 06:28:36.807101 134467927693120 checkpointing.py:772] Waiting for step 10 to finish before checkpoint...
I0420 06:28:36.960575 134467927693120 checkpointing.py:776] Waited 0.15345525741577148 seconds for step 10 to finish before starting checkpointing.
I0420 06:28:36.962988 134467927693120 checkpoint_manager.py:2009] [process=2][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0420 06:28:36.964949 134467927693120 checkpoint_manager.py:1512] [process=2] Saving checkpoint at step 10
I0420 06:28:36.966535 134467927693120 event_tracking.py:70] [process=2] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_08_checkpoint_async_true/checkpoints/10.
I0420 06:28:37.322454 134467927693120 signaling_client.py:364] Using JaxDistributedSignalingClient
I0420 06:28:37.323401 134467927693120 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0420 06:28:37.323459 134467927693120 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
I0420 06:28:37.598825 134467927693120 base_pytree_checkpoint_handler.py:154] [process=2][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.276450s
I0420 06:28:37.598992 134467927693120 base_pytree_checkpoint_handler.py:130] [process=2] /jax/orbax/write/blocking_gbytes_per_sec: 5.501 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2804090976715088 s) (per-host)
I0420 06:28:37.599058 134467927693120 base_pytree_checkpoint_handler.py:768] [process=2][thread=MainThread] Initiated Pytree async_save. Time taken: 0.280485s (batch_requests_ready=0.002263s, total_serialization_initiated=0.278138s, others=0.000084s)
I0420 06:28:37.599136 134467927693120 composite_checkpoint_handler.py:715] [process=2][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.284601s (all_items=0.000018s, per_item={'items': '0.00001788'}, temp_paths=0.284583)
I0420 06:28:37.599937 134467927693120 event_tracking.py:125] [process=2] [async] Finished blocking save in 0.63 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_08_checkpoint_async_true/checkpoints/10.
I0420 06:28:37.600296 134348707714816 async_checkpointer.py:76] [process=2][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-20 06:48:37.600257
I0420 06:28:37.613380 134467927693120 checkpoint_manager.py:1560] [process=2][thread=MainThread][step=10] Starting CheckpointManager Save Finalize thread=save_finalize
I0420 06:28:37.613647 134344922892032 async_checkpointer.py:280] [process=2][thread=save_finalize] Waiting for background save thread=async_save.
I0420 06:28:37.613800 134467927693120 standard_logger.py:34] {'step': 10, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_08_checkpoint_async_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776666516.9629707, 'wait_for_prev_duration_secs': 7.271766662597656e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776666516.9649868, 'checkpointer_blocking_duration_secs': 0.6354753971099854, 'get_old_steps_start_time': 1776666517.6004848, 'get_old_steps_duration_secs': 3.0994415283203125e-05, 'checkpoint_manager_blocking_start_time': 1776666516.961109, 'checkpoint_manager_blocking_duration_secs': 0.6526544094085693}
I0420 06:28:37.613909 134467927693120 checkpointing.py:408] Started an asynchronous checkpoint save for step 10
I0420 06:28:37.613958 134467927693120 max_utils.py:750] 
Memstats: After params initialized:
I0420 06:28:37.614005 134467927693120 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_8(process=2,(0,2,0,0))
I0420 06:28:37.614053 134467927693120 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_9(process=2,(1,2,0,0))
I0420 06:28:37.614080 134467927693120 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_12(process=2,(0,3,0,0))
I0420 06:28:37.614104 134467927693120 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_13(process=2,(1,3,0,0))
I0420 06:28:37.921892 134467927693120 metric_logger.py:196] completed step: 10, seconds: 11.326, TFLOP/s/device: 1.200, Tokens/s/device: 180.825, total_weights: 65536, loss: 6.833, lm_loss: 6.833, perplexity: 927.869
I0420 06:28:38.107733 134467927693120 metric_logger.py:196] completed step: 11, seconds: 1.113, TFLOP/s/device: 12.205, Tokens/s/device: 1839.683, total_weights: 65536, loss: 6.527, lm_loss: 6.527, perplexity: 683.659
I0420 06:28:38.255122 134467927693120 metric_logger.py:196] completed step: 12, seconds: 0.038, TFLOP/s/device: 355.117, Tokens/s/device: 53527.090, total_weights: 65536, loss: 6.265, lm_loss: 6.265, perplexity: 526.068
I0420 06:28:38.402592 134467927693120 metric_logger.py:196] completed step: 13, seconds: 0.153, TFLOP/s/device: 88.590, Tokens/s/device: 13353.241, total_weights: 65536, loss: 6.045, lm_loss: 6.045, perplexity: 422.001
I0420 06:28:38.407066 134467927693120 checkpointing.py:772] Waiting for step 15 to finish before checkpoint...
I0420 06:28:38.696967 134467927693120 checkpointing.py:776] Waited 0.28986597061157227 seconds for step 15 to finish before starting checkpointing.
I0420 06:28:38.699918 134467927693120 checkpoint_manager.py:2020] [process=2][thread=MainThread][step=10][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0420 06:28:42.814857 134344931284736 array_metadata_store.py:203] [process=2][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_08_checkpoint_async_true/checkpoints/10/items/array_metadatas/process_2
I0420 06:29:19.536436 134348707714816 base_pytree_checkpoint_handler.py:130] [process=2] /jax/orbax/write/gbytes_per_sec: 37.415 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.217806577682495 s) (per-host)
I0420 06:29:19.536559 134348707714816 async_checkpointer.py:90] [process=2][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.936150s.
I0420 06:29:28.262918 134348707714816 async_checkpointer.py:160] [process=2][thread=async_save] Background save thread done. Time taken: 50.662492s.
I0420 06:29:28.263222 134344922892032 async_checkpointer.py:288] [process=2][thread=save_finalize] Done with waiting for background save thread=async_save.
I0420 06:29:28.263339 134344922892032 async_checkpointer.py:298] [process=2][thread=save_finalize] No errors found in background save thread=async_save.
I0420 06:29:28.263388 134344922892032 checkpoint_manager.py:2137] [process=2][thread=save_finalize][step=10] CheckpointManager Save Finalize is syncing with other hosts...
I0420 06:29:28.264873 134344922892032 checkpoint_manager.py:2146] [process=2][thread=save_finalize][step=10] CheckpointManager Save Finalize is done on all hosts.
I0420 06:29:28.265099 134467927693120 checkpoint_manager.py:2032] [process=2][thread=MainThread][step=10][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=10.
W0420 06:29:28.265241 134467927693120 checkpoint_manager.py:1452] Waiting for previous save to complete took 49.565321 seconds. If this number is high, consider checkpointing less frequently.
I0420 06:29:28.267277 134467927693120 checkpoint_manager.py:1512] [process=2] Saving checkpoint at step 15
I0420 06:29:28.269714 134467927693120 event_tracking.py:70] [process=2] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_08_checkpoint_async_true/checkpoints/15.
I0420 06:29:28.567116 134467927693120 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0420 06:29:28.567215 134467927693120 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
I0420 06:29:28.593279 134467927693120 base_pytree_checkpoint_handler.py:154] [process=2][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.027256s
I0420 06:29:28.593433 134467927693120 base_pytree_checkpoint_handler.py:130] [process=2] /jax/orbax/write/blocking_gbytes_per_sec: 50.135 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.030768394470214844 s) (per-host)
I0420 06:29:28.593500 134467927693120 base_pytree_checkpoint_handler.py:768] [process=2][thread=MainThread] Initiated Pytree async_save. Time taken: 0.030843s (batch_requests_ready=0.001891s, total_serialization_initiated=0.028870s, others=0.000082s)
I0420 06:29:28.593602 134467927693120 composite_checkpoint_handler.py:715] [process=2][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.035667s (all_items=0.000013s, per_item={'items': '0.00001287'}, temp_paths=0.035654)
I0420 06:29:28.594328 134467927693120 event_tracking.py:125] [process=2] [async] Finished blocking save in 0.33 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_08_checkpoint_async_true/checkpoints/15.
I0420 06:29:28.594598 134344922892032 async_checkpointer.py:76] [process=2][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-20 06:49:28.594576
I0420 06:29:28.606658 134467927693120 checkpoint_manager.py:1560] [process=2][thread=MainThread][step=15] Starting CheckpointManager Save Finalize thread=save_finalize
I0420 06:29:28.606945 134344914499328 async_checkpointer.py:280] [process=2][thread=save_finalize] Waiting for background save thread=async_save.
I0420 06:29:28.607097 134467927693120 standard_logger.py:34] {'step': 15, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_08_checkpoint_async_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776666518.699889, 'wait_for_prev_duration_secs': 49.56532096862793, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776666568.267315, 'checkpointer_blocking_duration_secs': 0.32738447189331055, 'get_old_steps_start_time': 1776666568.5947237, 'get_old_steps_duration_secs': 3.147125244140625e-05, 'checkpoint_manager_blocking_start_time': 1776666518.697606, 'checkpoint_manager_blocking_duration_secs': 49.90945839881897}
I0420 06:29:28.607259 134467927693120 checkpointing.py:408] Started an asynchronous checkpoint save for step 15
I0420 06:29:28.608218 134467927693120 metric_logger.py:196] completed step: 14, seconds: 0.148, TFLOP/s/device: 92.099, Tokens/s/device: 13882.205, total_weights: 65536, loss: 5.864, lm_loss: 5.864, perplexity: 352.193
I0420 06:29:28.614467 134467927693120 metric_logger.py:196] completed step: 15, seconds: 0.147, TFLOP/s/device: 92.149, Tokens/s/device: 13889.737, total_weights: 65536, loss: 5.719, lm_loss: 5.719, perplexity: 304.661
I0420 06:29:33.713379 134342802134784 array_metadata_store.py:203] [process=2][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_08_checkpoint_async_true/checkpoints/15/items/array_metadatas/process_2
I0420 06:29:49.633197 134467927693120 metric_logger.py:196] completed step: 16, seconds: 50.206, TFLOP/s/device: 0.271, Tokens/s/device: 40.792, total_weights: 65536, loss: 5.605, lm_loss: 5.605, perplexity: 271.872
I0420 06:29:49.780679 134467927693120 metric_logger.py:196] completed step: 17, seconds: 20.873, TFLOP/s/device: 0.651, Tokens/s/device: 98.119, total_weights: 65536, loss: 5.517, lm_loss: 5.517, perplexity: 248.975
I0420 06:29:49.928202 134467927693120 metric_logger.py:196] completed step: 18, seconds: 0.152, TFLOP/s/device: 89.151, Tokens/s/device: 13437.791, total_weights: 65536, loss: 5.448, lm_loss: 5.448, perplexity: 232.333
I0420 06:29:50.074974 134467927693120 checkpointing.py:772] Waiting for step 19 to finish before checkpoint...
I0420 06:29:50.075705 134467927693120 checkpointing.py:776] Waited 0.0007500648498535156 seconds for step 19 to finish before starting checkpointing.
I0420 06:29:50.077797 134467927693120 checkpoint_manager.py:2020] [process=2][thread=MainThread][step=15][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0420 06:30:10.728660 134344922892032 base_pytree_checkpoint_handler.py:130] [process=2] /jax/orbax/write/gbytes_per_sec: 37.461 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.165969371795654 s) (per-host)
I0420 06:30:10.728768 134344922892032 async_checkpointer.py:90] [process=2][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.134093s.
I0420 06:30:17.384518 134344922892032 async_checkpointer.py:160] [process=2][thread=async_save] Background save thread done. Time taken: 48.789826s.
I0420 06:30:17.384805 134344914499328 async_checkpointer.py:288] [process=2][thread=save_finalize] Done with waiting for background save thread=async_save.
I0420 06:30:17.384922 134344914499328 async_checkpointer.py:298] [process=2][thread=save_finalize] No errors found in background save thread=async_save.
I0420 06:30:17.384970 134344914499328 checkpoint_manager.py:2137] [process=2][thread=save_finalize][step=15] CheckpointManager Save Finalize is syncing with other hosts...
I0420 06:30:17.386344 134344914499328 checkpoint_manager.py:2146] [process=2][thread=save_finalize][step=15] CheckpointManager Save Finalize is done on all hosts.
I0420 06:30:17.386533 134467927693120 checkpoint_manager.py:2032] [process=2][thread=MainThread][step=15][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=15.
W0420 06:30:17.386661 134467927693120 checkpoint_manager.py:1452] Waiting for previous save to complete took 27.308870 seconds. If this number is high, consider checkpointing less frequently.
I0420 06:30:17.388183 134467927693120 checkpoint_manager.py:1512] [process=2] Saving checkpoint at step 19
I0420 06:30:17.390326 134467927693120 event_tracking.py:70] [process=2] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_08_checkpoint_async_true/checkpoints/19.
I0420 06:30:17.674628 134467927693120 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0420 06:30:17.674720 134467927693120 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
I0420 06:30:17.700879 134467927693120 base_pytree_checkpoint_handler.py:154] [process=2][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.027205s
I0420 06:30:17.701005 134467927693120 base_pytree_checkpoint_handler.py:130] [process=2] /jax/orbax/write/blocking_gbytes_per_sec: 50.097 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.030791759490966797 s) (per-host)
I0420 06:30:17.701071 134467927693120 base_pytree_checkpoint_handler.py:768] [process=2][thread=MainThread] Initiated Pytree async_save. Time taken: 0.030866s (batch_requests_ready=0.001787s, total_serialization_initiated=0.029000s, others=0.000079s)
I0420 06:30:17.701146 134467927693120 composite_checkpoint_handler.py:715] [process=2][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.034941s (all_items=0.000012s, per_item={'items': '0.00001168'}, temp_paths=0.034930)
I0420 06:30:17.701795 134467927693120 event_tracking.py:125] [process=2] [async] Finished blocking save in 0.31 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_08_checkpoint_async_true/checkpoints/19.
I0420 06:30:17.702055 134344914499328 async_checkpointer.py:76] [process=2][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-20 06:50:17.702032
I0420 06:30:18.124735 134467927693120 checkpoint_manager.py:1560] [process=2][thread=MainThread][step=19] Starting CheckpointManager Save Finalize thread=save_finalize
I0420 06:30:18.125118 134342802134784 async_checkpointer.py:280] [process=2][thread=save_finalize] Waiting for background save thread=async_save.
I0420 06:30:18.125290 134467927693120 standard_logger.py:34] {'step': 19, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_08_checkpoint_async_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776666590.0777671, 'wait_for_prev_duration_secs': 27.30886960029602, 'time_between_consecutive_saves_sec': 21.8128559589386, 'checkpointer_blocking_start_time': 1776666617.3882203, 'checkpointer_blocking_duration_secs': 0.31394410133361816, 'get_old_steps_start_time': 1776666617.7021828, 'get_old_steps_duration_secs': 2.5272369384765625e-05, 'checkpoint_manager_blocking_start_time': 1776666590.0759318, 'checkpoint_manager_blocking_duration_secs': 28.04932403564453}
I0420 06:30:18.125397 134467927693120 checkpointing.py:408] Started an asynchronous checkpoint save for step 19
I0420 06:30:18.125441 134467927693120 checkpoint_manager.py:2020] [process=2][thread=MainThread][step=19][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0420 06:30:23.726528 134343322953472 array_metadata_store.py:203] [process=2][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260420_060028/linen_xpk_main_20260420_060028_08_checkpoint_async_true/checkpoints/19/items/array_metadatas/process_2
I0420 06:31:00.838668 134344914499328 base_pytree_checkpoint_handler.py:130] [process=2] /jax/orbax/write/gbytes_per_sec: 36.591 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 43.16841244697571 s) (per-host)
I0420 06:31:00.838796 134344914499328 async_checkpointer.py:90] [process=2][thread=async_save] 3 Handler Commit operations completed. Time taken: 43.136660s.
I0420 06:31:08.120726 134344914499328 async_checkpointer.py:160] [process=2][thread=async_save] Background save thread done. Time taken: 50.418574s.
I0420 06:31:08.121032 134342802134784 async_checkpointer.py:288] [process=2][thread=save_finalize] Done with waiting for background save thread=async_save.
I0420 06:31:08.121160 134342802134784 async_checkpointer.py:298] [process=2][thread=save_finalize] No errors found in background save thread=async_save.
I0420 06:31:08.121203 134342802134784 checkpoint_manager.py:2137] [process=2][thread=save_finalize][step=19] CheckpointManager Save Finalize is syncing with other hosts...
I0420 06:31:08.122786 134342802134784 checkpoint_manager.py:2146] [process=2][thread=save_finalize][step=19] CheckpointManager Save Finalize is done on all hosts.
I0420 06:31:08.122950 134467927693120 checkpoint_manager.py:2032] [process=2][thread=MainThread][step=19][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=19.
I0420 06:31:08.123104 134467927693120 checkpoint_manager.py:2009] [process=2][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0420 06:31:08.124091 134467927693120 metric_logger.py:196] completed step: 19, seconds: 0.147, TFLOP/s/device: 92.139, Tokens/s/device: 13888.230, 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: Mon Apr 20 06:31:17 UTC 2026
EXIT_CODE=0