MaxView

← Back to run

Log Summary

2026-04-23 03:45:00.316336: 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 00:00:1776915900.329583 2391308 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
E0000 00:00:1776915900.333635 2391308 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
W0000 00:00:1776915900.344924 2391308 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776915900.344939 2391308 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776915900.344941 2391308 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776915900.344942 2391308 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'rope_theta'}
`rope_scaling`'s factor field must be a float >= 1, got 40
`rope_scaling`'s beta_fast field must be a float, got 32
`rope_scaling`'s beta_slow field must be a float, got 1
Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'rope_theta'}
Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'rope_theta'}
Unrecognized keys in `rope_scaling` for 'rope_type'='yarn': {'rope_theta'}
2026-04-23 03:45:04.248944: 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)
I0423 03:45:04.384477 133498426390336 max_utils.py:238] Skipping jax distributed system due to skip_jax_distributed_system=True flag.
I0423 03:45:34.161542 133498426390336 max_utils.py:800] System Information: Jax Version: 0.8.1
I0423 03:45:34.161722 133498426390336 max_utils.py:801] System Information: Jaxlib Version: 0.8.1
I0423 03:45:34.161769 133498426390336 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
I0423 03:45:34.161800 133498426390336 train_utils.py:348] WARNING: 'dataset_path' might be pointing your local file system
I0423 03:45:34.161825 133498426390336 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0423 03:45:34.161963 133498426390336 train.py:672] [DECOUPLED NO-OP] skipping cloud diagnostics wrapper.
I0423 03:45:34.829594 133498426390336 maxtext_utils.py:1604] Num_devices: 8, shape (1, 1, 1, 8, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0423 03:45:34.829874 133498426390336 checkpointing.py:677] Setting up checkpoint logger...
I0423 03:45:34.829919 133498426390336 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0423 03:45:34.829956 133498426390336 pytree_checkpoint_handler.py:577] save_device_host_concurrent_bytes=None
I0423 03:45:34.830427 133498426390336 base_pytree_checkpoint_handler.py:411] 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 0x796a3cbbf0b0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0423 03:45:37.267793 133498426390336 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0423 03:45:37.268241 133498426390336 checkpoint_manager.py:702] [process=0][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7968653b9880>}, handler_registry=None
I0423 03:45:37.268483 133498426390336 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7968653b9880>` 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`.
I0423 03:45:37.268526 133498426390336 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7968653b9e80>` 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`.
I0423 03:45:37.268556 133498426390336 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 0x7968653b9880>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7968653b9880>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7968653b9e80>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7968653b9e80>}).
I0423 03:45:37.268860 133498426390336 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.28
I0423 03:45:37.268912 133498426390336 async_checkpointer.py:177] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>.<lambda> at 0x7968653cc4a0> timeout: 600 secs and primary_host=0 for async checkpoint writes
I0423 03:45:37.393508 133498426390336 checkpoint_manager.py:1788] Found 0 checkpoint steps in gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints
I0423 03:45:37.393762 133498426390336 checkpoint_manager.py:921] [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_hns=False, 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), root_directory=gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7968653b8950>
I0423 03:45:37.394040 133498426390336 checkpointing.py:301] Checkpoint manager created!
I0423 03:45:37.780669 133498426390336 nnx_wrappers.py:437] Unknown Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0423 03:45:37.780783 133498426390336 nnx_wrappers.py:437] Unknown Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, None).
I0423 03:45:38.063157 133498426390336 attentions.py:1088] attentions/inputs_q Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0423 03:45:38.063248 133498426390336 attentions.py:1088] attentions/inputs_q Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, None).
I0423 03:45:38.075453 133498426390336 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0423 03:45:38.075511 133498426390336 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, None).
I0423 03:45:38.094600 133498426390336 attentions.py:1154] attentions/query Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0423 03:45:38.094655 133498426390336 attentions.py:1154] attentions/query Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, None, None).
I0423 03:45:38.106850 133498426390336 attentions.py:1155] attentions/key Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0423 03:45:38.106904 133498426390336 attentions.py:1155] attentions/key Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, None, None).
I0423 03:45:38.119091 133498426390336 attentions.py:1156] attentions/value Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0423 03:45:38.119143 133498426390336 attentions.py:1156] attentions/value Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, None, None).
I0423 03:45:38.138583 133498426390336 attentions.py:1198] attentions/out Logical: bfloat16[8,2048,16,128]..................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0423 03:45:38.138646 133498426390336 attentions.py:1198] attentions/out Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, None, None).
I0423 03:45:38.154186 133498426390336 linears.py:525] linears/x Logical: bfloat16[8,2048,7168]....................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0423 03:45:38.154247 133498426390336 linears.py:525] linears/x Physical: bfloat16[8,2048,7168]....................................... ('fsdp', None, None).
I0423 03:45:38.304390 133498426390336 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0423 03:45:38.304492 133498426390336 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
[DECOUPLED NO-OP] gcs_storage: using stubs.
[DECOUPLED NO-OP] mldiagnostics: using stub.
[DECOUPLED NO-OP] mldiagnostics: using stub.
[DECOUPLED NO-OP] mldiagnostics: using stub.
[DECOUPLED NO-OP] workload_monitor: using stub.
[DECOUPLED NO-OP] vertex_tensorboard: using stub.
fsdp: 8
I0423 03:45:38.917290 133498426390336 maxtext_utils.py:1707]  params/params/decoder/decoder_norm/scale
    Shape:     float32[2048]
    Logical:   PartitionSpec('norm',)
    Physical:  (None,)
I0423 03:45:38.917404 133498426390336 maxtext_utils.py:1707]  params/params/decoder/layers/mlp/wi_0/kernel
    Shape:     float32[2048,16,7168]
    Logical:   PartitionSpec('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0423 03:45:38.917445 133498426390336 maxtext_utils.py:1707]  params/params/decoder/layers/mlp/wi_1/kernel
    Shape:     float32[2048,16,7168]
    Logical:   PartitionSpec('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0423 03:45:38.917490 133498426390336 maxtext_utils.py:1707]  params/params/decoder/layers/mlp/wo/kernel
    Shape:     float32[7168,16,2048]
    Logical:   PartitionSpec('mlp', 'layers', 'embed')
    Physical:  (None, None, 'fsdp')
I0423 03:45:38.917531 133498426390336 maxtext_utils.py:1707]  params/params/decoder/layers/post_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   PartitionSpec('norm', 'layers')
    Physical:  (None, None)
I0423 03:45:38.917560 133498426390336 maxtext_utils.py:1707]  params/params/decoder/layers/pre_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   PartitionSpec('norm', 'layers')
    Physical:  (None, None)
I0423 03:45:38.917610 133498426390336 maxtext_utils.py:1707]  params/params/decoder/layers/self_attention/key/kernel
    Shape:     float32[2048,16,16,128]
    Logical:   PartitionSpec('embed', 'layers', 'kv_heads', 'kv_head_dim')
    Physical:  ('fsdp', None, None, None)
I0423 03:45:38.917657 133498426390336 maxtext_utils.py:1707]  params/params/decoder/layers/self_attention/out/kernel
    Shape:     float32[16,16,128,2048]
    Logical:   PartitionSpec('heads', 'layers', 'kv', 'embed')
    Physical:  (None, None, None, 'fsdp')
I0423 03:45:38.917686 133498426390336 maxtext_utils.py:1707]  params/params/decoder/layers/self_attention/query/kernel
    Shape:     float32[2048,16,16,128]
    Logical:   PartitionSpec('embed', 'layers', 'q_heads', 'kv')
    Physical:  ('fsdp', None, None, None)
I0423 03:45:38.917712 133498426390336 maxtext_utils.py:1707]  params/params/decoder/layers/self_attention/value/kernel
    Shape:     float32[2048,16,16,128]
    Logical:   PartitionSpec('embed', 'layers', 'kv_heads', 'kv_head_dim')
    Physical:  ('fsdp', None, None, None)
I0423 03:45:38.917752 133498426390336 maxtext_utils.py:1707]  params/params/decoder/logits_dense/kernel
    Shape:     float32[2048,32000]
    Logical:   PartitionSpec('embed_vocab', 'vocab')
    Physical:  ('fsdp', None)
I0423 03:45:38.917789 133498426390336 maxtext_utils.py:1707]  params/params/token_embedder/embedding
    Shape:     float32[32000,2048]
    Logical:   PartitionSpec('vocab', 'embed_vocab')
    Physical:  (None, 'fsdp')

I0423 03:45:39.321647 133498426390336 train.py:155] train/xent Logical: float32[8,2048]............................................. ('activation_embed_and_logits_batch', 'activation_length').
I0423 03:45:39.321740 133498426390336 train.py:155] train/xent Physical: float32[8,2048]............................................. ('fsdp', None).
I0423 03:45:39.333337 133498426390336 train.py:162] train/z_loss Logical: float32[8,2048]............................................. ('activation_embed_and_logits_batch', 'activation_length').
I0423 03:45:39.333390 133498426390336 train.py:162] train/z_loss Physical: float32[8,2048]............................................. ('fsdp', None).
I0423 03:45:40.090881 133498426390336 max_utils.py:791] Total memory size: 3.6 GB, Output size: 1.5 GB, Temp size: 2.0 GB, Argument size: 1.5 GB, Host temp size: 0.0 GB.
I0423 03:45:40.091422 133498426390336 max_utils.py:194] tensorboardX not available; using no-op SummaryWriter.
I0423 03:45:40.091576 133498426390336 metric_logger.py:301] number parameters: 1.104 billion
I0423 03:45:42.741011 133498426390336 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0423 03:45:42.858798 133498426390336 checkpointing.py:776] Waited 0.11777281761169434 seconds for step 0 to finish before starting checkpointing.
I0423 03:45:42.859463 133498426390336 checkpoint_manager.py:1983] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0423 03:45:42.859619 133498426390336 checkpoint_manager.py:1501] [process=0] Saving checkpoint at step 0
I0423 03:45:42.860090 133498426390336 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/0.
I0423 03:45:42.945730 133498426390336 signaling_client.py:373] Using ThreadSafeKeyValueSignalingClient
I0423 03:45:43.036458 133467327170112 atomicity.py:137] Creating tmp directory gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/0
I0423 03:45:43.042541 133498426390336 jax_array_handlers.py:347] Scheduling D2H of 39 prioritized jax.Array.
I0423 03:45:43.042636 133498426390336 replica_slices.py:410] 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
I0423 03:45:43.789695 133467180369472 atomicity.py:137] Creating tmp directory gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/0/items
I0423 03:45:43.942325 133466943391296 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776915943696701679, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776915943696701679, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/0/_CHECKPOINT_METADATA
I0423 03:45:44.295992 2393303 google_auth_provider.cc:149] Using credentials at ~/.config/gcloud/application_default_credentials.json
I0423 03:45:44.296073 2393303 google_auth_provider.cc:156] Using OAuth2 AuthProvider
I0423 03:45:45.545798 133498426390336 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 2.503683s
I0423 03:45:45.551109 133498426390336 base_pytree_checkpoint_handler.py:128] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 4.740 GiB/s (total gbytes: 12.3 GiB) (time elapsed: 2 seconds) (per-host)
I0423 03:45:45.551188 133498426390336 base_pytree_checkpoint_handler.py:732] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 2.603880s (batch_requests_ready=0.090950s, total_serialization_initiated=2.507818s, others=0.005111s)
I0423 03:45:45.551277 133498426390336 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 2.604984s (all_items=0.000023s, per_item={'items': '0.00002265'}, temp_paths=2.604961)
I0423 03:45:45.552893 133467530593856 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started.
I0423 03:45:45.552998 133498426390336 async_checkpointer.py:561] Finished blocking save. Time taken: 2.693336s. Continuing background save to gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/0.
I0423 03:45:45.553239 133498426390336 checkpoint_manager.py:1549] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0423 03:45:45.553473 133466138084928 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0423 03:45:45.553597 133498426390336 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776915942.8594499, 'wait_for_prev_duration_secs': 4.363059997558594e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776915942.8596408, 'checkpointer_blocking_duration_secs': 2.6934945583343506, 'get_old_steps_start_time': 1776915945.5531676, 'get_old_steps_duration_secs': 2.9802322387695312e-05, 'checkpoint_manager_blocking_start_time': 1776915942.859363, 'checkpoint_manager_blocking_duration_secs': 2.6942059993743896}
I0423 03:45:45.553732 133498426390336 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0423 03:45:45.553786 133498426390336 max_utils.py:750] 
Memstats: After params initialized:
I0423 03:45:45.553842 133498426390336 max_utils.py:756] 	Using (GB) 1.59 / 31.25 (5.088000%) on TPU_0(process=0,(0,0,0,0))
I0423 03:45:45.553871 133498426390336 max_utils.py:756] 	Using (GB) 1.59 / 31.25 (5.088000%) on TPU_1(process=0,(1,0,0,0))
I0423 03:45:45.553892 133498426390336 max_utils.py:756] 	Using (GB) 1.59 / 31.25 (5.088000%) on TPU_2(process=0,(0,1,0,0))
I0423 03:45:45.553913 133498426390336 max_utils.py:756] 	Using (GB) 1.59 / 31.25 (5.088000%) on TPU_3(process=0,(1,1,0,0))
I0423 03:45:45.553934 133498426390336 max_utils.py:756] 	Using (GB) 1.59 / 31.25 (5.088000%) on TPU_4(process=0,(0,2,0,0))
I0423 03:45:45.553951 133498426390336 max_utils.py:756] 	Using (GB) 1.59 / 31.25 (5.088000%) on TPU_5(process=0,(1,2,0,0))
I0423 03:45:45.553972 133498426390336 max_utils.py:756] 	Using (GB) 1.59 / 31.25 (5.088000%) on TPU_6(process=0,(0,3,0,0))
I0423 03:45:45.553989 133498426390336 max_utils.py:756] 	Using (GB) 1.59 / 31.25 (5.088000%) on TPU_7(process=0,(1,3,0,0))
I0423 03:45:45.805800 133498426390336 metric_logger.py:196] completed step: 0, seconds: 2.649, TFLOP/s/device: 5.129, Tokens/s/device: 773.148, total_weights: 16384, loss: 10.874, lm_loss: 10.874, perplexity: 52770.941
I0423 03:45:45.806743 133498426390336 metric_logger.py:281] To see full metrics 'tensorboard --logdir=gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/tensorboard/'
I0423 03:45:45.921815 133498426390336 metric_logger.py:196] completed step: 1, seconds: 3.062, TFLOP/s/device: 4.437, Tokens/s/device: 668.749, total_weights: 16384, loss: 10.874, lm_loss: 10.874, perplexity: 52770.941
I0423 03:45:46.040398 133498426390336 metric_logger.py:196] completed step: 2, seconds: 0.007, TFLOP/s/device: 2020.691, Tokens/s/device: 304580.607, total_weights: 16384, loss: 10.276, lm_loss: 10.276, perplexity: 29019.973
I0423 03:45:46.159397 133498426390336 metric_logger.py:196] completed step: 3, seconds: 0.117, TFLOP/s/device: 116.290, Tokens/s/device: 17528.544, total_weights: 16384, loss: 9.282, lm_loss: 9.282, perplexity: 10739.224
I0423 03:45:46.164579 133498426390336 checkpointing.py:772] Waiting for step 5 to finish before checkpoint...
I0423 03:45:46.323942 133467159397952 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/0/items/array_metadatas/process_0
I0423 03:45:46.396573 133498426390336 checkpointing.py:776] Waited 0.23195338249206543 seconds for step 5 to finish before starting checkpointing.
I0423 03:45:46.397569 133498426390336 checkpoint_manager.py:1994] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0423 03:46:18.984339 133466339411520 base_pytree_checkpoint_handler.py:1217] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 33.432315s (commit=32.955236s, array_metadata_write=0.477078s)
I0423 03:46:18.985488 133467530593856 base_pytree_checkpoint_handler.py:128] [process=0] /jax/checkpoint/write/gbytes_per_sec: 350.650 MiB/s (total gbytes: 12.3 GiB) (time elapsed: 36 seconds) (per-host)
I0423 03:46:18.985548 133467530593856 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 33.432486s.
I0423 03:46:19.324218 133467530593856 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776915943696701679, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/0/_CHECKPOINT_METADATA
I0423 03:46:19.533442 133467530593856 array_metadata_store.py:367] [process=0][thread=async_save] Skipped cross-host ArrayMetadata validation because only one process is found: process_index=0.
I0423 03:46:19.750715 133466943391296 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776915943696701679, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/0/_CHECKPOINT_METADATA
I0423 03:46:19.977600 133467530593856 ocdbt_utils.py:56] Param validation support for Zarr3 will be added later (b/362328389).
I0423 03:46:19.977843 133467530593856 base_pytree_checkpoint_handler.py:1342] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 0.618454s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/0/items
I0423 03:46:19.978615 133467530593856 atomicity.py:608] Finalizing gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/0/items
I0423 03:46:20.246746 133467530593856 atomicity.py:608] Finalizing gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/0
I0423 03:46:21.014370 133467530593856 atomicity.py:794] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/0`.
I0423 03:46:21.015190 133467530593856 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 38.155537s. directory=gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/0
I0423 03:46:21.015272 133467530593856 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 35.462211s.
I0423 03:46:21.015529 133466138084928 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0423 03:46:21.015661 133466138084928 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0423 03:46:21.015726 133466138084928 checkpoint_manager.py:2103] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0423 03:46:21.015775 133466138084928 checkpoint_manager.py:2112] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0423 03:46:21.015952 133498426390336 checkpoint_manager.py:2006] [process=0][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0423 03:46:21.016116 133498426390336 checkpoint_manager.py:1441] Waiting for previous save to complete took 34.618553 seconds. If this number is high, consider checkpointing less frequently.
I0423 03:46:21.017546 133498426390336 checkpoint_manager.py:1501] [process=0] Saving checkpoint at step 5
I0423 03:46:21.017884 133498426390336 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/5.
I0423 03:46:21.182620 133467530593856 atomicity.py:137] Creating tmp directory gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/5
I0423 03:46:21.206014 133498426390336 jax_array_handlers.py:347] Scheduling D2H of 39 prioritized jax.Array.
I0423 03:46:21.206131 133498426390336 replica_slices.py:410] 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
I0423 03:46:21.944894 133467159397952 atomicity.py:137] Creating tmp directory gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/5/items
I0423 03:46:26.267844 133498426390336 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 5.062309s
I0423 03:46:26.272256 133498426390336 base_pytree_checkpoint_handler.py:128] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.388 GiB/s (total gbytes: 12.3 GiB) (time elapsed: 5 seconds) (per-host)
I0423 03:46:26.272324 133498426390336 base_pytree_checkpoint_handler.py:732] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 5.167528s (batch_requests_ready=0.097428s, total_serialization_initiated=5.065807s, others=0.004293s)
I0423 03:46:26.272412 133498426390336 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 5.168123s (all_items=0.000016s, per_item={'items': '0.00001574'}, temp_paths=5.168108)
I0423 03:46:26.274273 133466138084928 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started.
I0423 03:46:26.274363 133498426390336 async_checkpointer.py:561] Finished blocking save. Time taken: 5.256770s. Continuing background save to gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/5.
I0423 03:46:26.274560 133498426390336 checkpoint_manager.py:1549] [process=0][thread=MainThread][step=5] Starting CheckpointManager Save Finalize thread=save_finalize
I0423 03:46:26.274756 133467530593856 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0423 03:46:26.274882 133498426390336 standard_logger.py:34] {'step': 5, 'event_type': 'save', 'directory': 'gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776915946.397537, 'wait_for_prev_duration_secs': 34.61855340003967, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776915981.0175712, 'checkpointer_blocking_duration_secs': 5.2569005489349365, 'get_old_steps_start_time': 1776915986.2744918, 'get_old_steps_duration_secs': 2.86102294921875e-05, 'checkpoint_manager_blocking_start_time': 1776915946.3974693, 'checkpoint_manager_blocking_duration_secs': 39.87738299369812}
I0423 03:46:26.275030 133498426390336 checkpointing.py:408] Started an asynchronous checkpoint save for step 5
I0423 03:46:26.275797 133498426390336 metric_logger.py:196] completed step: 4, seconds: 0.119, TFLOP/s/device: 114.463, Tokens/s/device: 17253.144, total_weights: 16384, loss: 8.327, lm_loss: 8.327, perplexity: 4134.812
I0423 03:46:26.284734 133498426390336 metric_logger.py:196] completed step: 5, seconds: 0.119, TFLOP/s/device: 114.490, Tokens/s/device: 17257.215, total_weights: 16384, loss: 7.367, lm_loss: 7.367, perplexity: 1583.522
I0423 03:46:26.400912 133498426390336 metric_logger.py:196] completed step: 6, seconds: 40.118, TFLOP/s/device: 0.339, Tokens/s/device: 51.049, total_weights: 16384, loss: 6.459, lm_loss: 6.459, perplexity: 638.363
I0423 03:46:26.519414 133498426390336 metric_logger.py:196] completed step: 7, seconds: 0.007, TFLOP/s/device: 2087.117, Tokens/s/device: 314592.934, total_weights: 16384, loss: 5.622, lm_loss: 5.622, perplexity: 276.577
I0423 03:46:26.638210 133498426390336 metric_logger.py:196] completed step: 8, seconds: 0.117, TFLOP/s/device: 116.135, Tokens/s/device: 17505.171, total_weights: 16384, loss: 4.864, lm_loss: 4.864, perplexity: 129.523
I0423 03:46:26.756193 133498426390336 checkpointing.py:772] Waiting for step 9 to finish before checkpoint...
I0423 03:46:26.757498 133498426390336 checkpointing.py:776] Waited 0.0013184547424316406 seconds for step 9 to finish before starting checkpointing.
I0423 03:46:26.757911 133498426390336 checkpoint_manager.py:1994] [process=0][thread=MainThread][step=5][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0423 03:46:27.045186 133467148912192 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/5/items/array_metadatas/process_0
I0423 03:47:04.383325 133467159397952 base_pytree_checkpoint_handler.py:1217] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 38.110363s (commit=37.628434s, array_metadata_write=0.481929s)
I0423 03:47:04.384562 133466138084928 base_pytree_checkpoint_handler.py:128] [process=0] /jax/checkpoint/write/gbytes_per_sec: 291.979 MiB/s (total gbytes: 12.3 GiB) (time elapsed: 43 seconds) (per-host)
I0423 03:47:04.384623 133466138084928 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 38.110214s.
I0423 03:47:04.871628 133466138084928 array_metadata_store.py:367] [process=0][thread=async_save] Skipped cross-host ArrayMetadata validation because only one process is found: process_index=0.
I0423 03:47:05.279235 133466138084928 ocdbt_utils.py:56] Param validation support for Zarr3 will be added later (b/362328389).
I0423 03:47:05.279502 133466138084928 base_pytree_checkpoint_handler.py:1342] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 0.578790s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/5/items
I0423 03:47:05.280435 133466138084928 atomicity.py:608] Finalizing gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/5/items
I0423 03:47:05.532895 133466138084928 atomicity.py:608] Finalizing gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/5
I0423 03:47:06.279335 133466138084928 atomicity.py:794] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/5`.
I0423 03:47:06.280097 133466138084928 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 45.262511s. directory=gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/5
I0423 03:47:06.280180 133466138084928 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 40.005774s.
I0423 03:47:06.280443 133467530593856 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0423 03:47:06.280576 133467530593856 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0423 03:47:06.280652 133467530593856 checkpoint_manager.py:2103] [process=0][thread=save_finalize][step=5] CheckpointManager Save Finalize is syncing with other hosts...
I0423 03:47:06.280696 133467530593856 checkpoint_manager.py:2112] [process=0][thread=save_finalize][step=5] CheckpointManager Save Finalize is done on all hosts.
I0423 03:47:06.280850 133498426390336 checkpoint_manager.py:2006] [process=0][thread=MainThread][step=5][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=5.
W0423 03:47:06.280985 133498426390336 checkpoint_manager.py:1441] Waiting for previous save to complete took 39.523077 seconds. If this number is high, consider checkpointing less frequently.
I0423 03:47:06.281601 133498426390336 checkpoint_manager.py:1501] [process=0] Saving checkpoint at step 9
I0423 03:47:06.281906 133498426390336 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/9.
I0423 03:47:06.471757 133466138084928 atomicity.py:137] Creating tmp directory gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/9
I0423 03:47:06.478150 133498426390336 jax_array_handlers.py:347] Scheduling D2H of 39 prioritized jax.Array.
I0423 03:47:06.478257 133498426390336 replica_slices.py:410] 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
I0423 03:47:07.204816 133467148912192 atomicity.py:137] Creating tmp directory gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/9/items
I0423 03:47:09.813280 133498426390336 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 3.335572s
I0423 03:47:09.813886 133498426390336 base_pytree_checkpoint_handler.py:128] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 3.599 GiB/s (total gbytes: 12.3 GiB) (time elapsed: 3 seconds) (per-host)
I0423 03:47:09.813940 133498426390336 base_pytree_checkpoint_handler.py:732] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 3.428789s (batch_requests_ready=0.088250s, total_serialization_initiated=3.340068s, others=0.000471s)
I0423 03:47:09.814009 133498426390336 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 3.429387s (all_items=0.000012s, per_item={'items': '0.00001192'}, temp_paths=3.429375)
I0423 03:47:09.819481 133467127940672 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started.
I0423 03:47:09.819588 133498426390336 async_checkpointer.py:561] Finished blocking save. Time taken: 3.537938s. Continuing background save to gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/9.
I0423 03:47:09.819785 133498426390336 checkpoint_manager.py:1549] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0423 03:47:09.819986 133467530593856 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0423 03:47:09.820119 133498426390336 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776915986.7578826, 'wait_for_prev_duration_secs': 39.52307724952698, 'time_between_consecutive_saves_sec': 5.742092609405518, 'checkpointer_blocking_start_time': 1776916026.28163, 'checkpointer_blocking_duration_secs': 3.5380706787109375, 'get_old_steps_start_time': 1776916029.8197186, 'get_old_steps_duration_secs': 2.5510787963867188e-05, 'checkpoint_manager_blocking_start_time': 1776915986.7578444, 'checkpoint_manager_blocking_duration_secs': 43.062248945236206}
I0423 03:47:09.820248 133498426390336 checkpointing.py:408] Started an asynchronous checkpoint save for step 9
I0423 03:47:09.820280 133498426390336 checkpoint_manager.py:1994] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0423 03:47:10.600878 133467138426432 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/9/items/array_metadatas/process_0
I0423 03:47:50.607409 133467148912192 base_pytree_checkpoint_handler.py:1217] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 40.789015s (commit=40.272099s, array_metadata_write=0.516915s)
I0423 03:47:50.608654 133467127940672 base_pytree_checkpoint_handler.py:128] [process=0] /jax/checkpoint/write/gbytes_per_sec: 285.749 MiB/s (total gbytes: 12.3 GiB) (time elapsed: 44 seconds) (per-host)
I0423 03:47:50.608777 133467127940672 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 40.789144s.
I0423 03:47:51.051031 133467127940672 array_metadata_store.py:367] [process=0][thread=async_save] Skipped cross-host ArrayMetadata validation because only one process is found: process_index=0.
I0423 03:47:51.481980 133467127940672 ocdbt_utils.py:56] Param validation support for Zarr3 will be added later (b/362328389).
I0423 03:47:51.482273 133467127940672 base_pytree_checkpoint_handler.py:1342] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 0.588524s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/9/items
I0423 03:47:51.483162 133467127940672 atomicity.py:608] Finalizing gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/9/items
I0423 03:47:51.748721 133467127940672 atomicity.py:608] Finalizing gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/9
I0423 03:47:52.480957 133467127940672 atomicity.py:794] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/9`.
I0423 03:47:52.481717 133467127940672 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 46.200075s. directory=gs://wanglance-maxtext/linen_ckpt_fix_gate_fp8_nan_sanitization_20260423_024158/linen_fix_gate_fp8_nan_sanitization_20260423_024158_15_ocdbt_true/checkpoints/9
I0423 03:47:52.481781 133467127940672 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 42.662150s.
I0423 03:47:52.481994 133467530593856 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0423 03:47:52.482127 133467530593856 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0423 03:47:52.482181 133467530593856 checkpoint_manager.py:2103] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0423 03:47:52.482222 133467530593856 checkpoint_manager.py:2112] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0423 03:47:52.482396 133498426390336 checkpoint_manager.py:2006] [process=0][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0423 03:47:52.482551 133498426390336 checkpoint_manager.py:1983] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0423 03:47:52.483415 133498426390336 metric_logger.py:196] completed step: 9, seconds: 0.119, TFLOP/s/device: 114.580, Tokens/s/device: 17270.749, total_weights: 16384, loss: 4.189, lm_loss: 4.189, perplexity: 65.979
Per train step:
 Total TFLOPs: 13.59 
 split as 93.93% learnable weight flops and 6.07% attention flops