MaxView

← Back to run

Log Summary

XPK Start: Mon Apr 20 22:02:04 UTC 2026
2026-04-20 22:02:08.534875: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:467] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
WARNING: All log messages before absl::InitializeLog() is called are written to STDERR
E0000 00:00:1776722528.547823      10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
E0000 00:00:1776722528.551527      10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
W0000 00:00:1776722528.563075      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776722528.563093      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776722528.563095      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776722528.563097      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
2026-04-20 22:02:27.744845: 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 22:02:28.274795 139264286537536 max_utils.py:273] Attempting to initialize the jax distributed system...
INFO:2026-04-20 22:02:37,315:jax._src.distributed:140: Starting JAX distributed service on [::]:8482
I0420 22:02:37.315989 139264286537536 distributed.py:140] Starting JAX distributed service on [::]:8482
INFO:2026-04-20 22:02:37,318:jax._src.distributed:157: Connecting to JAX distributed service on mt-15-ocdbt-false-sv-h39kc-slice-job-0-0.mt-15-ocdbt-false-sv-h39kc:8482
I0420 22:02:37.318402 139264286537536 distributed.py:157] Connecting to JAX distributed service on mt-15-ocdbt-false-sv-h39kc-slice-job-0-0.mt-15-ocdbt-false-sv-h39kc:8482
I0420 22:02:38.517881 139264286537536 max_utils.py:284] Jax distributed system initialized!
I0420 22:02:44.578057 139264286537536 max_utils.py:800] System Information: Jax Version: 0.8.1
I0420 22:02:44.578165 139264286537536 max_utils.py:801] System Information: Jaxlib Version: 0.8.1
I0420 22:02:44.578205 139264286537536 max_utils.py:802] System Information: Jax Backend: PJRT C API
TFRT TPU v6 lite
Built on Nov 12 2025 14:16:36 (1762985796) cl/831091709
I0420 22:02:44.578243 139264286537536 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0420 22:02:45.216356 139264286537536 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0420 22:02:45.333534 139264286537536 checkpointing.py:688] Setting up checkpoint logger...
I0420 22:02:45.333665 139264286537536 checkpointing.py:234] Creating checkpoint manager with ocdbt=False and zarr3=True
I0420 22:02:45.333711 139264286537536 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None
I0420 22:02:45.333921 139264286537536 base_pytree_checkpoint_handler.py:415] Created BasePyTreeCheckpointHandler: use_ocdbt=False, use_zarr3=True, pytree_metadata_options=PyTreeMetadataOptions(support_rich_types=False), array_metadata_store=<orbax.checkpoint._src.metadata.array_metadata_store.Store object at 0x7ea872779520>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0420 22:02:48.298347 139264286537536 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0420 22:02:48.298603 139264286537536 checkpoint_manager.py:709] [process=7][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ea8b59d3e60>}, handler_registry=None
I0420 22:02:48.298885 139264286537536 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ea8b59d3e60>` 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 22:02:48.298944 139264286537536 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e9480154170>` 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 22:02:48.299000 139264286537536 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 0x7ea8b59d3e60>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ea8b59d3e60>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e9480154170>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e9480154170>}).
I0420 22:02:48.299374 139264286537536 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33
I0420 22:02:48.299467 139264286537536 async_checkpointer.py:177] [process=7][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7e9480120f40> timeout: 600 secs and primary_host=0 for async checkpoint writes
I0420 22:02:49.905853 139264286537536 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_15_ocdbt_false/checkpoints
I0420 22:02:50.358551 139264286537536 checkpoint_manager.py:929] [process=7][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/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_15_ocdbt_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7e94802f8a10>
I0420 22:02:50.358755 139264286537536 checkpointing.py:302] Checkpoint manager created!
I0420 22:02:51.319135 139264286537536 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0420 22:02:51.319260 139264286537536 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0420 22:02:52.723535 139264286537536 maxtext_utils.py:1749]  decoder/decoder_norm/scale/value
    Shape:     float32[2048]
    Physical:  (None,)
I0420 22:02:52.723666 139264286537536 maxtext_utils.py:1749]  decoder/layers/mlp/wi_0/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0420 22:02:52.723725 139264286537536 maxtext_utils.py:1749]  decoder/layers/mlp/wi_1/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0420 22:02:52.723772 139264286537536 maxtext_utils.py:1749]  decoder/layers/mlp/wo/kernel/value
    Shape:     float32[7168,16,2048]
    Physical:  (None, None, 'fsdp')
I0420 22:02:52.723811 139264286537536 maxtext_utils.py:1749]  decoder/layers/post_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0420 22:02:52.723847 139264286537536 maxtext_utils.py:1749]  decoder/layers/pre_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0420 22:02:52.723884 139264286537536 maxtext_utils.py:1749]  decoder/layers/self_attention/key/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0420 22:02:52.723921 139264286537536 maxtext_utils.py:1749]  decoder/layers/self_attention/out/kernel/value
    Shape:     float32[16,16,128,2048]
    Physical:  (None, None, None, 'fsdp')
I0420 22:02:52.723955 139264286537536 maxtext_utils.py:1749]  decoder/layers/self_attention/query/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0420 22:02:52.723989 139264286537536 maxtext_utils.py:1749]  decoder/layers/self_attention/value/kernel/value

    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0420 22:02:52.724021 139264286537536 maxtext_utils.py:1749]  decoder/logits_dense/kernel/value
    Shape:     float32[2048,32000]
    Physical:  ('fsdp', None)
I0420 22:02:52.724054 139264286537536 maxtext_utils.py:1749]  token_embedder/embedding/value
    Shape:     float32[32000,2048]
    Physical:  (None, 'fsdp')
I0420 22:02:53.117371 139264286537536 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0420 22:02:53.117521 139264286537536 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 22:02:53.123012 139264286537536 nnx_decoders.py:616] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0420 22:02:53.123113 139264286537536 nnx_decoders.py:616] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 22:02:53.279237 139264286537536 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0420 22:02:53.279377 139264286537536 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 22:02:53.433375 139264286537536 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0420 22:02:53.433506 139264286537536 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 22:02:53.598325 139264286537536 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 22:02:53.598466 139264286537536 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 22:02:53.758679 139264286537536 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 22:02:53.758816 139264286537536 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 22:02:53.918521 139264286537536 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 22:02:53.918693 139264286537536 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 22:02:54.086699 139264286537536 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0420 22:02:54.086838 139264286537536 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 22:02:54.110970 139264286537536 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0420 22:02:54.111074 139264286537536 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0420 22:03:05.465894 139264286537536 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 22:03:05.468544 139264286537536 metric_logger.py:289] number parameters: 1.104 billion
I0420 22:03:16.774601 139264286537536 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0420 22:03:16.925495 139264286537536 checkpointing.py:798] Waited 0.15088248252868652 seconds for step 0 to finish before starting checkpointing.
I0420 22:03:16.927838 139264286537536 checkpoint_manager.py:2013] [process=7][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0420 22:03:16.929786 139264286537536 checkpoint_manager.py:1518] [process=7] Saving checkpoint at step 0
I0420 22:03:16.932634 139264286537536 async_checkpointer.py:452] [process=7] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_15_ocdbt_false/checkpoints/0.
I0420 22:03:17.922561 139264286537536 signaling_client.py:364] Using JaxDistributedSignalingClient
I0420 22:03:17.923739 139264286537536 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0420 22:03:17.923801 139264286537536 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 22:03:18.171214 139264286537536 base_pytree_checkpoint_handler.py:153] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.249769s
I0420 22:03:18.171384 139264286537536 base_pytree_checkpoint_handler.py:129] [process=7] /jax/checkpoint/write/blocking_gbytes_per_sec: 3.073 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5020570755004883 s) (per-host)
I0420 22:03:18.171435 139264286537536 base_pytree_checkpoint_handler.py:737] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.502119s (batch_requests_ready=0.245751s, total_serialization_initiated=0.256299s, others=0.000069s)
I0420 22:03:18.171530 139264286537536 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.506362s (all_items=0.000020s, per_item={'items': '0.00001979'}, temp_paths=0.506342)
I0420 22:03:18.172394 139143619262208 async_checkpointer.py:79] [process=7][thread=async_save] Background save thread started.
I0420 22:03:18.172556 139264286537536 async_checkpointer.py:561] Finished blocking save. Time taken: 1.242708s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_15_ocdbt_false/checkpoints/0.
I0420 22:03:18.280554 139264286537536 checkpoint_manager.py:1566] [process=7][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0420 22:03:18.281030 139138284123904 async_checkpointer.py:265] [process=7][thread=save_finalize] Waiting for background save thread=async_save.
I0420 22:03:18.281204 139264286537536 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776722596.9278204, 'wait_for_prev_duration_secs': 6.389617919921875e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776722596.9298248, 'checkpointer_blocking_duration_secs': 1.2428035736083984, 'get_old_steps_start_time': 1776722598.1726484, 'get_old_steps_duration_secs': 3.218650817871094e-05, 'checkpoint_manager_blocking_start_time': 1776722596.9259405, 'checkpoint_manager_blocking_duration_secs': 1.355222225189209}
I0420 22:03:18.281324 139264286537536 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0420 22:03:18.281399 139264286537536 max_utils.py:750] 
Memstats: After params initialized:
I0420 22:03:18.281457 139264286537536 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_26(process=7,(2,6,0,0))
I0420 22:03:18.281490 139264286537536 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_27(process=7,(3,6,0,0))
I0420 22:03:18.281516 139264286537536 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_30(process=7,(2,7,0,0))
I0420 22:03:18.281539 139264286537536 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_31(process=7,(3,7,0,0))
I0420 22:03:18.553381 139264286537536 metric_logger.py:185] completed step: 0, seconds: 11.305, TFLOP/s/device: 1.202, Tokens/s/device: 181.159, total_weights: 65536, loss: 10.874
I0420 22:03:18.719850 139264286537536 metric_logger.py:185] completed step: 1, seconds: 1.777, TFLOP/s/device: 7.646, Tokens/s/device: 1152.496, total_weights: 65536, loss: 10.874
I0420 22:03:19.593994 139264286537536 metric_logger.py:185] completed step: 2, seconds: 0.021, TFLOP/s/device: 655.939, Tokens/s/device: 98870.329, total_weights: 65536, loss: 10.560
I0420 22:03:19.741670 139264286537536 metric_logger.py:185] completed step: 3, seconds: 0.865, TFLOP/s/device: 15.701, Tokens/s/device: 2366.615, total_weights: 65536, loss: 9.979
I0420 22:03:19.747774 139264286537536 checkpointing.py:794] Waiting for step 5 to finish before checkpoint...
I0420 22:03:20.036702 139264286537536 checkpointing.py:798] Waited 0.288865327835083 seconds for step 5 to finish before starting checkpointing.
I0420 22:03:20.039672 139264286537536 checkpoint_manager.py:2024] [process=7][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0420 22:03:21.825884    2576 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0420 22:03:23.764298 139143627654912 array_metadata_store.py:203] [process=7][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_15_ocdbt_false/checkpoints/0/items/array_metadatas/process_7
I0420 22:03:29.853110 139143619262208 base_pytree_checkpoint_handler.py:129] [process=7] /jax/checkpoint/write/gbytes_per_sec: 129.648 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 12.183743238449097 s) (per-host)
I0420 22:03:29.853228 139143619262208 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 11.680719s.
I0420 22:03:35.991220 139143619262208 async_checkpointer.py:144] [process=7][thread=async_save] Background save thread done. Time taken: 17.818696s.
I0420 22:03:35.991519 139138284123904 async_checkpointer.py:273] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save.
I0420 22:03:35.991638 139138284123904 async_checkpointer.py:283] [process=7][thread=save_finalize] No errors found in background save thread=async_save.
I0420 22:03:35.991717 139138284123904 checkpoint_manager.py:2133] [process=7][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0420 22:03:35.994686 139138284123904 checkpoint_manager.py:2142] [process=7][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0420 22:03:35.994849 139264286537536 checkpoint_manager.py:2036] [process=7][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0420 22:03:35.994985 139264286537536 checkpoint_manager.py:1458] Waiting for previous save to complete took 15.955328 seconds. If this number is high, consider checkpointing less frequently.
I0420 22:03:35.996955 139264286537536 checkpoint_manager.py:1518] [process=7] Saving checkpoint at step 5
I0420 22:03:36.001249 139264286537536 async_checkpointer.py:452] [process=7] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_15_ocdbt_false/checkpoints/5.
I0420 22:03:36.968402 139264286537536 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0420 22:03:36.968556 139264286537536 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 22:03:37.011578 139264286537536 base_pytree_checkpoint_handler.py:153] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.045295s
I0420 22:03:37.011775 139264286537536 base_pytree_checkpoint_handler.py:129] [process=7] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.057 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.305020809173584 s) (per-host)
I0420 22:03:37.011833 139264286537536 base_pytree_checkpoint_handler.py:737] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.305087s (batch_requests_ready=0.254611s, total_serialization_initiated=0.050402s, others=0.000074s)
I0420 22:03:37.011926 139264286537536 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.310076s (all_items=0.000033s, per_item={'items': '0.00003266'}, temp_paths=0.310044)
I0420 22:03:37.012863 139134492456704 async_checkpointer.py:79] [process=7][thread=async_save] Background save thread started.
I0420 22:03:37.013025 139264286537536 async_checkpointer.py:561] Finished blocking save. Time taken: 1.015989s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_15_ocdbt_false/checkpoints/5.
I0420 22:03:37.032940 139264286537536 checkpoint_manager.py:1566] [process=7][thread=MainThread][step=5] Starting CheckpointManager Save Finalize thread=save_finalize
I0420 22:03:37.033238 139138284123904 async_checkpointer.py:265] [process=7][thread=save_finalize] Waiting for background save thread=async_save.
I0420 22:03:37.033421 139264286537536 standard_logger.py:34] {'step': 5, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776722600.0396285, 'wait_for_prev_duration_secs': 15.95532751083374, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776722615.9970014, 'checkpointer_blocking_duration_secs': 1.0161375999450684, 'get_old_steps_start_time': 1776722617.0131571, 'get_old_steps_duration_secs': 2.6702880859375e-05, 'checkpoint_manager_blocking_start_time': 1776722600.03705, 'checkpoint_manager_blocking_duration_secs': 16.996336698532104}
I0420 22:03:37.033530 139264286537536 checkpointing.py:409] Started an asynchronous checkpoint save for step 5
I0420 22:03:37.034357 139264286537536 metric_logger.py:185] completed step: 4, seconds: 0.163, TFLOP/s/device: 83.531, Tokens/s/device: 12590.757, total_weights: 65536, loss: 9.457
I0420 22:03:37.041624 139264286537536 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.187, Tokens/s/device: 13895.485, total_weights: 65536, loss: 8.955
I0420 22:03:58.080969 139264286537536 metric_logger.py:185] completed step: 6, seconds: 17.292, TFLOP/s/device: 0.786, Tokens/s/device: 118.435, total_weights: 65536, loss: 8.465
I0420 22:03:58.228434 139264286537536 metric_logger.py:185] completed step: 7, seconds: 20.894, TFLOP/s/device: 0.650, Tokens/s/device: 98.019, total_weights: 65536, loss: 7.998
I0420 22:03:58.375859 139264286537536 metric_logger.py:185] completed step: 8, seconds: 0.153, TFLOP/s/device: 88.571, Tokens/s/device: 13350.456, total_weights: 65536, loss: 7.567
I0420 22:03:58.524405 139264286537536 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0420 22:03:58.524987 139264286537536 checkpointing.py:798] Waited 0.0005946159362792969 seconds for step 9 to finish before starting checkpointing.
I0420 22:03:58.527244 139264286537536 checkpoint_manager.py:2024] [process=7][thread=MainThread][step=5][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0420 22:04:14.323973 139143627654912 array_metadata_store.py:203] [process=7][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_15_ocdbt_false/checkpoints/5/items/array_metadatas/process_7
I0420 22:04:15.195099 139134492456704 base_pytree_checkpoint_handler.py:129] [process=7] /jax/checkpoint/write/gbytes_per_sec: 41.041 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 38.48829531669617 s) (per-host)
I0420 22:04:15.195223 139134492456704 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 38.182250s.
I0420 22:04:28.033413 139134492456704 async_checkpointer.py:144] [process=7][thread=async_save] Background save thread done. Time taken: 51.020423s.
I0420 22:04:28.033723 139138284123904 async_checkpointer.py:273] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save.
I0420 22:04:28.033855 139138284123904 async_checkpointer.py:283] [process=7][thread=save_finalize] No errors found in background save thread=async_save.
I0420 22:04:28.033904 139138284123904 checkpoint_manager.py:2133] [process=7][thread=save_finalize][step=5] CheckpointManager Save Finalize is syncing with other hosts...
I0420 22:04:28.035622 139138284123904 checkpoint_manager.py:2142] [process=7][thread=save_finalize][step=5] CheckpointManager Save Finalize is done on all hosts.
I0420 22:04:28.035853 139264286537536 checkpoint_manager.py:2036] [process=7][thread=MainThread][step=5][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=5.
W0420 22:04:28.035985 139264286537536 checkpoint_manager.py:1458] Waiting for previous save to complete took 29.508753 seconds. If this number is high, consider checkpointing less frequently.
I0420 22:04:28.037912 139264286537536 checkpoint_manager.py:1518] [process=7] Saving checkpoint at step 9
I0420 22:04:28.041420 139264286537536 async_checkpointer.py:452] [process=7] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_15_ocdbt_false/checkpoints/9.
I0420 22:04:29.861574 139264286537536 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0420 22:04:29.861688 139264286537536 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 22:04:29.904756 139264286537536 base_pytree_checkpoint_handler.py:153] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.045199s
I0420 22:04:29.904885 139264286537536 base_pytree_checkpoint_handler.py:129] [process=7] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.103 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.7333767414093018 s) (per-host)
I0420 22:04:29.904933 139264286537536 base_pytree_checkpoint_handler.py:737] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.733433s (batch_requests_ready=0.683506s, total_serialization_initiated=0.049866s, others=0.000061s)
I0420 22:04:29.905019 139264286537536 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.737371s (all_items=0.000011s, per_item={'items': '0.00001121'}, temp_paths=0.737360)
I0420 22:04:29.905919 139137193596672 async_checkpointer.py:79] [process=7][thread=async_save] Background save thread started.
I0420 22:04:29.906082 139264286537536 async_checkpointer.py:561] Finished blocking save. Time taken: 1.868099s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_15_ocdbt_false/checkpoints/9.
I0420 22:04:30.338065 139264286537536 checkpoint_manager.py:1566] [process=7][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0420 22:04:30.338435 139138284123904 async_checkpointer.py:265] [process=7][thread=save_finalize] Waiting for background save thread=async_save.
I0420 22:04:30.338607 139264286537536 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776722638.5272164, 'wait_for_prev_duration_secs': 29.508752822875977, 'time_between_consecutive_saves_sec': 22.53248953819275, 'checkpointer_blocking_start_time': 1776722668.0379515, 'checkpointer_blocking_duration_secs': 1.868238925933838, 'get_old_steps_start_time': 1776722669.9062047, 'get_old_steps_duration_secs': 2.0742416381835938e-05, 'checkpoint_manager_blocking_start_time': 1776722638.5252116, 'checkpoint_manager_blocking_duration_secs': 31.813360929489136}
I0420 22:04:30.338734 139264286537536 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0420 22:04:30.338780 139264286537536 checkpoint_manager.py:2024] [process=7][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0420 22:04:38.603768 139136673511168 array_metadata_store.py:203] [process=7][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_15_ocdbt_false/checkpoints/9/items/array_metadatas/process_7
I0420 22:04:43.458807 139137193596672 base_pytree_checkpoint_handler.py:129] [process=7] /jax/checkpoint/write/gbytes_per_sec: 110.560 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 14.28725790977478 s) (per-host)
I0420 22:04:43.458930 139137193596672 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 13.552900s.
I0420 22:04:55.244137 139137193596672 async_checkpointer.py:144] [process=7][thread=async_save] Background save thread done. Time taken: 25.338092s.
I0420 22:04:55.244458 139138284123904 async_checkpointer.py:273] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save.
I0420 22:04:55.244589 139138284123904 async_checkpointer.py:283] [process=7][thread=save_finalize] No errors found in background save thread=async_save.
I0420 22:04:55.244632 139138284123904 checkpoint_manager.py:2133] [process=7][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0420 22:04:55.246182 139138284123904 checkpoint_manager.py:2142] [process=7][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0420 22:04:55.246359 139264286537536 checkpoint_manager.py:2036] [process=7][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0420 22:04:55.246519 139264286537536 checkpoint_manager.py:2013] [process=7][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0420 22:04:55.247235 139264286537536 metric_logger.py:185] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.435, Tokens/s/device: 13932.826, total_weights: 65536, loss: 7.176
Per train step:
 Total TFLOPs: 13.59 
 split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Mon Apr 20 22:05:04 UTC 2026
EXIT_CODE=0