XPK Start: Sat Apr 25 15:04:49 UTC 2026
2026-04-25 15:04:53.563024: 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 00:00:1777129493.575894 9 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
E0000 00:00:1777129493.579617 9 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
W0000 00:00:1777129493.591290 9 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1777129493.591308 9 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1777129493.591311 9 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1777129493.591313 9 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
2026-04-25 15:05:12.712889: 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)
I0425 15:05:13.236313 139980942305088 max_utils.py:273] Attempting to initialize the jax distributed system...
INFO:2026-04-25 15:05:22,276:jax._src.distributed:140: Starting JAX distributed service on [::]:8482
I0425 15:05:22.276894 139980942305088 distributed.py:140] Starting JAX distributed service on [::]:8482
INFO:2026-04-25 15:05:22,279:jax._src.distributed:157: Connecting to JAX distributed service on mt-01-base-a1diu-slice-job-0-0.mt-01-base-a1diu:8482
I0425 15:05:22.279316 139980942305088 distributed.py:157] Connecting to JAX distributed service on mt-01-base-a1diu-slice-job-0-0.mt-01-base-a1diu:8482
I0425 15:05:23.804972 139980942305088 max_utils.py:284] Jax distributed system initialized!
I0425 15:05:30.694481 139980942305088 max_utils.py:800] System Information: Jax Version: 0.8.1
I0425 15:05:30.694586 139980942305088 max_utils.py:801] System Information: Jaxlib Version: 0.8.1
I0425 15:05:30.694626 139980942305088 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
I0425 15:05:30.694661 139980942305088 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0425 15:05:31.315828 139980942305088 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0425 15:05:31.316439 139980942305088 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0425 15:05:31.316626 139980942305088 checkpointing.py:688] Setting up checkpoint logger...
I0425 15:05:31.316679 139980942305088 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0425 15:05:31.316723 139980942305088 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None
I0425 15:05:31.317068 139980942305088 base_pytree_checkpoint_handler.py:415] 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 0x7f4f58771430>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0425 15:05:34.989872 139980942305088 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0425 15:05:34.990127 139980942305088 checkpoint_manager.py:709] [process=6][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7f3ac8161cd0>}, handler_registry=None
I0425 15:05:34.990421 139980942305088 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7f3ac8161cd0>` 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`.
I0425 15:05:34.990475 139980942305088 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7f3ac8166cf0>` 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`.
I0425 15:05:34.990512 139980942305088 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 0x7f3ac8161cd0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7f3ac8161cd0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7f3ac8166cf0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7f3ac8166cf0>}).
I0425 15:05:34.991012 139980942305088 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33
I0425 15:05:34.991106 139980942305088 async_checkpointer.py:177] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7f3a882f4180> timeout: 600 secs and primary_host=0 for async checkpoint writes
I0425 15:05:35.769617 139980942305088 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_01_base/checkpoints
I0425 15:05:35.890469 139980942305088 checkpoint_manager.py:929] [process=6][thread=MainThread] CheckpointManager created, primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=10), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_01_base/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7f3ac8166c00>
I0425 15:05:35.890647 139980942305088 checkpointing.py:302] Checkpoint manager created!
I0425 15:05:36.782845 139980942305088 nnx_wrappers.py:455] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0425 15:05:36.782951 139980942305088 nnx_wrappers.py:455] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0425 15:05:37.156429 139980942305088 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0425 15:05:37.156526 139980942305088 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0425 15:05:37.172463 139980942305088 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0425 15:05:37.172523 139980942305088 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0425 15:05:37.195659 139980942305088 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0425 15:05:37.195727 139980942305088 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0425 15:05:37.211806 139980942305088 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0425 15:05:37.211870 139980942305088 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0425 15:05:37.227837 139980942305088 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0425 15:05:37.227901 139980942305088 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0425 15:05:37.252739 139980942305088 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0425 15:05:37.252810 139980942305088 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0425 15:05:37.272632 139980942305088 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0425 15:05:37.272700 139980942305088 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0425 15:05:37.463207 139980942305088 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0425 15:05:37.463322 139980942305088 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0425 15:05:38.904912 139980942305088 maxtext_utils.py:1740] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: PartitionSpec('norm',)
Physical: (None,)
I0425 15:05:38.905045 139980942305088 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wi_0/kernel
Shape: float32[2048,16,7168]
Logical: PartitionSpec('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0425 15:05:38.905113 139980942305088 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wi_1/kernel
Shape: float32[2048,16,7168]
Logical: PartitionSpec('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0425 15:05:38.905177 139980942305088 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wo/kernel
Shape: float32[7168,16,2048]
Logical: PartitionSpec('mlp', 'layers', 'embed')
Physical: (None, None, 'fsdp')
I0425 15:05:38.905231 139980942305088 maxtext_utils.py:1740] params/params/decoder/layers/post_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: PartitionSpec('norm', 'layers')
Physical: (None, None)
I0425 15:05:38.905271 139980942305088 maxtext_utils.py:1740] params/params/decoder/layers/pre_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: PartitionSpec('norm', 'layers')
Physical: (None, None)
I0425 15:05:38.905323 139980942305088 maxtext_utils.py:1740] 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)
I0425 15:05:38.905376 139980942305088 maxtext_utils.py:1740] 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')
I0425 15:05:38.905418 139980942305088 maxtext_utils.py:1740] 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)
I0425 15:05:38.905457 139980942305088 maxtext_utils.py:1740] 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)
I0425 15:05:38.905512 139980942305088 maxtext_utils.py:1740] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: PartitionSpec('embed', 'vocab')
Physical: ('fsdp', None)
I0425 15:05:38.905561 139980942305088 maxtext_utils.py:1740] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: PartitionSpec('vocab', 'embed')
Physical: (None, 'fsdp')
I0425 15:05:39.417440 139980942305088 train.py:158] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0425 15:05:39.417537 139980942305088 train.py:158] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0425 15:05:39.432763 139980942305088 train.py:165] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0425 15:05:39.432824 139980942305088 train.py:165] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0425 15:05:50.186779 139980942305088 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.
I0425 15:05:50.187626 139980942305088 metric_logger.py:289] number parameters: 1.104 billion
I0425 15:06:01.600446 139980942305088 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0425 15:06:01.755718 139980942305088 checkpointing.py:798] Waited 0.15525269508361816 seconds for step 0 to finish before starting checkpointing.
I0425 15:06:01.758258 139980942305088 checkpoint_manager.py:2013] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0425 15:06:01.759984 139980942305088 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 0
I0425 15:06:01.762677 139980942305088 async_checkpointer.py:452] [process=6] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_01_base/checkpoints/0.
I0425 15:06:02.350571 139980942305088 signaling_client.py:364] Using JaxDistributedSignalingClient
I0425 15:06:02.351638 139980942305088 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array.
I0425 15:06:02.351695 139980942305088 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
I0425 15:06:02.602811 139980942305088 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.252319s
I0425 15:06:02.602984 139980942305088 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.970 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5194516181945801 s) (per-host)
I0425 15:06:02.603038 139980942305088 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.519514s (batch_requests_ready=0.263535s, total_serialization_initiated=0.255908s, others=0.000070s)
I0425 15:06:02.603143 139980942305088 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.523411s (all_items=0.000020s, per_item={'items': '0.00001979'}, temp_paths=0.523391)
I0425 15:06:02.604151 139854503458560 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started.
I0425 15:06:02.604336 139980942305088 async_checkpointer.py:561] Finished blocking save. Time taken: 0.844281s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_01_base/checkpoints/0.
I0425 15:06:02.615304 139980942305088 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0425 15:06:02.615561 139854528636672 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0425 15:06:02.615692 139980942305088 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_01_base/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777129561.7582383, 'wait_for_prev_duration_secs': 6.961822509765625e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777129561.7600214, 'checkpointer_blocking_duration_secs': 0.8444223403930664, 'get_old_steps_start_time': 1777129562.6044629, 'get_old_steps_duration_secs': 2.765655517578125e-05, 'checkpoint_manager_blocking_start_time': 1777129561.756231, 'checkpoint_manager_blocking_duration_secs': 0.8594217300415039}
I0425 15:06:02.615798 139980942305088 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0425 15:06:02.615847 139980942305088 max_utils.py:750]
Memstats: After params initialized:
I0425 15:06:02.615897 139980942305088 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_24(process=6,(0,6,0,0))
I0425 15:06:02.615931 139980942305088 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_25(process=6,(1,6,0,0))
I0425 15:06:02.615958 139980942305088 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_28(process=6,(0,7,0,0))
I0425 15:06:02.615982 139980942305088 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_29(process=6,(1,7,0,0))
I0425 15:06:02.890787 139980942305088 metric_logger.py:185] completed step: 0, seconds: 11.413, TFLOP/s/device: 1.191, Tokens/s/device: 179.449, total_weights: 65536, loss: 10.874
I0425 15:06:03.071499 139980942305088 metric_logger.py:185] completed step: 1, seconds: 1.289, TFLOP/s/device: 10.541, Tokens/s/device: 1588.832, total_weights: 65536, loss: 10.874
I0425 15:06:03.496483 139980942305088 metric_logger.py:185] completed step: 2, seconds: 0.033, TFLOP/s/device: 406.642, Tokens/s/device: 61293.509, total_weights: 65536, loss: 10.263
I0425 15:06:03.643722 139980942305088 metric_logger.py:185] completed step: 3, seconds: 0.426, TFLOP/s/device: 31.927, Tokens/s/device: 4812.358, total_weights: 65536, loss: 9.734
I0425 15:06:03.938882 139980942305088 metric_logger.py:185] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.876, Tokens/s/device: 13396.303, total_weights: 65536, loss: 9.278
I0425 15:06:03.944700 139980942305088 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.470, Tokens/s/device: 13938.136, total_weights: 65536, loss: 8.892
I0425 15:06:22.400398 139980942305088 metric_logger.py:185] completed step: 6, seconds: 0.296, TFLOP/s/device: 45.954, Tokens/s/device: 6926.665, total_weights: 65536, loss: 8.592
I0425 15:06:22.547810 139980942305088 metric_logger.py:185] completed step: 7, seconds: 18.309, TFLOP/s/device: 0.742, Tokens/s/device: 111.858, total_weights: 65536, loss: 8.384
I0425 15:06:22.695075 139980942305088 metric_logger.py:185] completed step: 8, seconds: 0.152, TFLOP/s/device: 89.227, Tokens/s/device: 13449.263, total_weights: 65536, loss: 8.255
I0425 15:06:22.842382 139980942305088 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0425 15:06:22.843035 139980942305088 checkpointing.py:798] Waited 0.0006740093231201172 seconds for step 9 to finish before starting checkpointing.
I0425 15:06:22.845031 139980942305088 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0425 15:06:23.812329 2646 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0425 15:06:25.759641 139854511851264 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_01_base/checkpoints/0/items/array_metadatas/process_6
I0425 15:06:57.928781 139854503458560 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 28.285 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 55.84519386291504 s) (per-host)
I0425 15:06:57.928889 139854503458560 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 55.324622s.
I0425 15:07:03.900567 139854503458560 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 61.296285s.
I0425 15:07:03.900851 139854528636672 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0425 15:07:03.900986 139854528636672 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0425 15:07:03.901033 139854528636672 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0425 15:07:03.902561 139854528636672 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0425 15:07:03.902755 139980942305088 checkpoint_manager.py:2036] [process=6][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0425 15:07:03.902890 139980942305088 checkpoint_manager.py:1458] Waiting for previous save to complete took 41.057863 seconds. If this number is high, consider checkpointing less frequently.
I0425 15:07:03.904800 139980942305088 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 9
I0425 15:07:03.908238 139980942305088 async_checkpointer.py:452] [process=6] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_01_base/checkpoints/9.
I0425 15:07:04.541306 139980942305088 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array.
I0425 15:07:04.541403 139980942305088 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
I0425 15:07:04.586147 139980942305088 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.046073s
I0425 15:07:04.586307 139980942305088 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.168 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.29846858978271484 s) (per-host)
I0425 15:07:04.586356 139980942305088 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.298527s (batch_requests_ready=0.250572s, total_serialization_initiated=0.047891s, others=0.000064s)
I0425 15:07:04.586439 139980942305088 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.302636s (all_items=0.000016s, per_item={'items': '0.00001597'}, temp_paths=0.302620)
I0425 15:07:04.587394 139854511851264 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started.
I0425 15:07:04.587559 139980942305088 async_checkpointer.py:561] Finished blocking save. Time taken: 0.682685s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_01_base/checkpoints/9.
I0425 15:07:04.602711 139980942305088 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0425 15:07:04.602951 139854528636672 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0425 15:07:04.603118 139980942305088 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_01_base/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777129582.8450027, 'wait_for_prev_duration_secs': 41.057862997055054, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777129623.90484, 'checkpointer_blocking_duration_secs': 0.6828272342681885, 'get_old_steps_start_time': 1777129624.5876887, 'get_old_steps_duration_secs': 2.4080276489257812e-05, 'checkpoint_manager_blocking_start_time': 1777129582.8433127, 'checkpoint_manager_blocking_duration_secs': 41.75976228713989}
I0425 15:07:04.603222 139980942305088 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0425 15:07:04.603265 139980942305088 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0425 15:07:09.335577 139854520243968 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/linen_xpk_feat_nnx_set_defaults_true_20260425_150412_01_base/checkpoints/9/items/array_metadatas/process_6
I0425 15:07:47.401880 139854511851264 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 36.638 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 43.11400651931763 s) (per-host)
I0425 15:07:47.401987 139854511851264 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.814478s.
I0425 15:07:52.465299 139854511851264 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 47.877772s.
I0425 15:07:52.465609 139854528636672 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0425 15:07:52.465746 139854528636672 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0425 15:07:52.465811 139854528636672 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0425 15:07:52.467356 139854528636672 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0425 15:07:52.467533 139980942305088 checkpoint_manager.py:2036] [process=6][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0425 15:07:52.467674 139980942305088 checkpoint_manager.py:2013] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0425 15:07:52.468357 139980942305088 metric_logger.py:185] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.165, Tokens/s/device: 13892.186, total_weights: 65536, loss: 8.179
Per train step:
Total TFLOPs: 13.59
split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Sat Apr 25 15:08:05 UTC 2026
EXIT_CODE=0