XPK Start: Wed Apr 22 15:47:30 UTC 2026
2026-04-22 15:47:34.547810: 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 00:00:1776872854.560464 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
E0000 00:00:1776872854.564149 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
W0000 00:00:1776872854.575346 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776872854.575365 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776872854.575367 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776872854.575369 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
2026-04-22 15:47:54.631396: 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)
I0422 15:47:55.158417 132131956553536 max_utils.py:273] Attempting to initialize the jax distributed system...
INFO:2026-04-22 15:48:04,200:jax._src.distributed:157: Connecting to JAX distributed service on mt-01-base-z5ly1-slice-job-0-0.mt-01-base-z5ly1:8482
I0422 15:48:04.200445 132131956553536 distributed.py:157] Connecting to JAX distributed service on mt-01-base-z5ly1-slice-job-0-0.mt-01-base-z5ly1:8482
I0422 15:49:20.963425 132131956553536 max_utils.py:284] Jax distributed system initialized!
I0422 15:49:27.816052 132131956553536 max_utils.py:800] System Information: Jax Version: 0.8.1
I0422 15:49:27.816174 132131956553536 max_utils.py:801] System Information: Jaxlib Version: 0.8.1
I0422 15:49:27.816217 132131956553536 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
I0422 15:49:27.816251 132131956553536 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0422 15:49:28.429248 132131956553536 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0422 15:49:28.429852 132131956553536 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0422 15:49:28.430040 132131956553536 checkpointing.py:688] Setting up checkpoint logger...
I0422 15:49:28.430097 132131956553536 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0422 15:49:28.430142 132131956553536 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None
I0422 15:49:28.430499 132131956553536 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 0x782be69000e0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0422 15:49:32.696817 132131956553536 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0422 15:49:32.697172 132131956553536 checkpoint_manager.py:709] [process=0][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7819003b5c40>}, handler_registry=None
I0422 15:49:32.697408 132131956553536 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7819003b5c40>` 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`.
I0422 15:49:32.697458 132131956553536 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7819003b6900>` 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`.
I0422 15:49:32.697496 132131956553536 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 0x7819003b5c40>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7819003b5c40>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7819003b6900>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7819003b6900>}).
I0422 15:49:32.698014 132131956553536 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33
I0422 15:49:32.698099 132131956553536 async_checkpointer.py:177] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x781900318180> timeout: 600 secs and primary_host=0 for async checkpoint writes
I0422 15:49:33.431244 132131956553536 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints
I0422 15:49:33.461622 132131956553536 checkpoint_manager.py:929] [process=0][thread=MainThread] CheckpointManager created, primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=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_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7819003b5790>
I0422 15:49:33.461763 132131956553536 checkpointing.py:302] Checkpoint manager created!
I0422 15:49:34.333373 132131956553536 nnx_wrappers.py:455] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0422 15:49:34.333483 132131956553536 nnx_wrappers.py:455] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0422 15:49:34.707772 132131956553536 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0422 15:49:34.707871 132131956553536 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0422 15:49:34.724215 132131956553536 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0422 15:49:34.724284 132131956553536 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0422 15:49:34.753035 132131956553536 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0422 15:49:34.753139 132131956553536 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 15:49:34.769542 132131956553536 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0422 15:49:34.769615 132131956553536 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 15:49:34.785867 132131956553536 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0422 15:49:34.785930 132131956553536 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 15:49:34.811188 132131956553536 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0422 15:49:34.811266 132131956553536 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 15:49:34.831191 132131956553536 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0422 15:49:34.831260 132131956553536 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0422 15:49:35.023569 132131956553536 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0422 15:49:35.023691 132131956553536 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0422 15:49:36.470442 132131956553536 maxtext_utils.py:1740] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: PartitionSpec('norm',)
Physical: (None,)
I0422 15:49:36.470592 132131956553536 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)
I0422 15:49:36.470847 132131956553536 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)
I0422 15:49:36.471066 132131956553536 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')
I0422 15:49:36.471144 132131956553536 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)
I0422 15:49:36.471194 132131956553536 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)
I0422 15:49:36.471261 132131956553536 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)
I0422 15:49:36.471335 132131956553536 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')
I0422 15:49:36.471384 132131956553536 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)
I0422 15:49:36.471427 132131956553536 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)
I0422 15:49:36.471481 132131956553536 maxtext_utils.py:1740] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: PartitionSpec('embed', 'vocab')
Physical: ('fsdp', None)
I0422 15:49:36.471533 132131956553536 maxtext_utils.py:1740] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: PartitionSpec('vocab', 'embed')
Physical: (None, 'fsdp')
I0422 15:49:36.981132 132131956553536 train.py:158] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0422 15:49:36.981231 132131956553536 train.py:158] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0422 15:49:36.997112 132131956553536 train.py:165] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0422 15:49:36.997176 132131956553536 train.py:165] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0422 15:49:47.666381 132131956553536 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.
I0422 15:49:47.672540 132131956553536 metric_logger.py:289] number parameters: 1.104 billion
I0422 15:49:59.196230 132131956553536 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0422 15:49:59.352881 132131956553536 checkpointing.py:798] Waited 0.15662765502929688 seconds for step 0 to finish before starting checkpointing.
I0422 15:49:59.355621 132131956553536 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0422 15:49:59.357365 132131956553536 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 0
I0422 15:49:59.360203 132131956553536 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/0.
I0422 15:49:59.685648 132131956553536 signaling_client.py:364] Using JaxDistributedSignalingClient
I0422 15:49:59.977876 132131956553536 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array.
I0422 15:49:59.977971 132131956553536 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
I0422 15:49:59.997423 131996416206592 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/0
I0422 15:50:00.219441 132131956553536 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.242811s
I0422 15:50:00.220380 132131956553536 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.928 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5268592834472656 s) (per-host)
I0422 15:50:00.220451 132131956553536 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.526949s (batch_requests_ready=0.279419s, total_serialization_initiated=0.246723s, others=0.000806s)
I0422 15:50:00.220523 132131956553536 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.530964s (all_items=0.000019s, per_item={'items': '0.00001884'}, temp_paths=0.530946)
I0422 15:50:00.221498 131997900994304 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started.
I0422 15:50:00.221632 132131956553536 async_checkpointer.py:561] Finished blocking save. Time taken: 0.864196s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/0.
I0422 15:50:00.268890 132131956553536 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0422 15:50:00.269230 131996407813888 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0422 15:50:00.269371 132131956553536 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776872999.355601, 'wait_for_prev_duration_secs': 7.033348083496094e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776872999.3574054, 'checkpointer_blocking_duration_secs': 0.8643996715545654, 'get_old_steps_start_time': 1776873000.2218258, 'get_old_steps_duration_secs': 2.5510787963867188e-05, 'checkpoint_manager_blocking_start_time': 1776872999.353414, 'checkpoint_manager_blocking_duration_secs': 0.9159178733825684}
I0422 15:50:00.269529 132131956553536 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0422 15:50:00.269603 132131956553536 max_utils.py:750]
Memstats: After params initialized:
I0422 15:50:00.269658 132131956553536 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0))
I0422 15:50:00.269692 132131956553536 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0))
I0422 15:50:00.269719 132131956553536 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0))
I0422 15:50:00.269743 132131956553536 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0))
I0422 15:50:00.558158 132131956553536 metric_logger.py:185] completed step: 0, seconds: 11.151, TFLOP/s/device: 1.218, Tokens/s/device: 183.657, total_weights: 65536, loss: 10.874
I0422 15:50:00.561023 132131956553536 metric_logger.py:269] To see full metrics 'tensorboard --logdir=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/tensorboard/'
I0422 15:50:01.005211 132131956553536 metric_logger.py:185] completed step: 1, seconds: 1.360, TFLOP/s/device: 9.988, Tokens/s/device: 1505.540, total_weights: 65536, loss: 10.874
I0422 15:50:01.151328 132131956553536 metric_logger.py:185] completed step: 2, seconds: 0.448, TFLOP/s/device: 30.340, Tokens/s/device: 4573.205, total_weights: 65536, loss: 10.263
I0422 15:50:01.298684 132131956553536 metric_logger.py:185] completed step: 3, seconds: 0.008, TFLOP/s/device: 1657.776, Tokens/s/device: 249877.989, total_weights: 65536, loss: 9.734
I0422 15:50:20.464821 132131956553536 metric_logger.py:185] completed step: 4, seconds: 0.147, TFLOP/s/device: 92.296, Tokens/s/device: 13911.815, total_weights: 65536, loss: 9.278
I0422 15:50:20.473638 132131956553536 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.215, Tokens/s/device: 13899.635, total_weights: 65536, loss: 8.892
I0422 15:50:20.619664 132131956553536 metric_logger.py:185] completed step: 6, seconds: 19.166, TFLOP/s/device: 0.709, Tokens/s/device: 106.858, total_weights: 65536, loss: 8.592
I0422 15:50:20.767208 132131956553536 metric_logger.py:185] completed step: 7, seconds: 0.006, TFLOP/s/device: 2222.298, Tokens/s/device: 334968.924, total_weights: 65536, loss: 8.384
I0422 15:50:20.882307 132000828606208 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/0/items
I0422 15:50:20.914775 132131956553536 metric_logger.py:185] completed step: 8, seconds: 0.147, TFLOP/s/device: 92.241, Tokens/s/device: 13903.504, total_weights: 65536, loss: 8.255
I0422 15:50:21.055277 131997883434752 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776873020596979739, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776873020596979739, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/0/_CHECKPOINT_METADATA
I0422 15:50:21.062241 132131956553536 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0422 15:50:21.062965 132131956553536 checkpointing.py:798] Waited 0.0007469654083251953 seconds for step 9 to finish before starting checkpointing.
I0422 15:50:21.064951 132131956553536 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0422 15:50:22.506250 2370 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0422 15:50:24.270824 132000677357312 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/0/items/array_metadatas/process_0
I0422 15:50:59.994927 132000668964608 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 37.495797s (commit=36.337769s, array_metadata_write=1.158028s)
I0422 15:50:59.996013 131997900994304 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 26.195 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 60.30245780944824 s) (per-host)
I0422 15:50:59.996140 131997900994304 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 59.774426s.
I0422 15:51:00.783414 131997900994304 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776873020596979739, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/0/_CHECKPOINT_METADATA
I0422 15:51:01.541631 131997900994304 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000235s.
I0422 15:51:01.588850 131997883434752 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776873020596979739, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/0/_CHECKPOINT_METADATA
I0422 15:51:02.162750 131997900994304 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0422 15:51:03.122496 131997900994304 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.205428s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/0/items
I0422 15:51:03.124711 131997900994304 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/0/items
I0422 15:51:03.594697 131997900994304 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/0
I0422 15:51:04.950222 131997900994304 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/0`.
I0422 15:51:04.950996 131997900994304 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 65.593577s. directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/0
I0422 15:51:04.952846 131997900994304 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 64.731132s.
I0422 15:51:04.953053 131996407813888 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0422 15:51:04.953189 131996407813888 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0422 15:51:04.953268 131996407813888 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0422 15:51:04.955929 131996407813888 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0422 15:51:04.956052 132131956553536 checkpoint_manager.py:2036] [process=0][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0422 15:51:04.956131 132131956553536 checkpoint_manager.py:1458] Waiting for previous save to complete took 43.891190 seconds. If this number is high, consider checkpointing less frequently.
I0422 15:51:04.958121 132131956553536 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 9
I0422 15:51:04.961529 132131956553536 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/9.
I0422 15:51:05.556365 131997900994304 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/9
I0422 15:51:05.560114 132131956553536 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array.
I0422 15:51:05.560219 132131956553536 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
I0422 15:51:05.588320 132131956553536 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.029294s
I0422 15:51:05.589534 132131956553536 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.054 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.30523109436035156 s) (per-host)
I0422 15:51:05.589622 132131956553536 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.305335s (batch_requests_ready=0.272624s, total_serialization_initiated=0.031539s, others=0.001173s)
I0422 15:51:05.589704 132131956553536 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.309356s (all_items=0.000015s, per_item={'items': '0.00001502'}, temp_paths=0.309341)
I0422 15:51:05.590536 131993453901568 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started.
I0422 15:51:05.590660 132131956553536 async_checkpointer.py:561] Finished blocking save. Time taken: 0.632471s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/9.
I0422 15:51:05.618613 132131956553536 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0422 15:51:05.618896 131996407813888 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0422 15:51:05.619071 132131956553536 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776873021.0649197, 'wait_for_prev_duration_secs': 43.89119029045105, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776873064.9581614, 'checkpointer_blocking_duration_secs': 0.6326544284820557, 'get_old_steps_start_time': 1776873065.5908356, 'get_old_steps_duration_secs': 2.8848648071289062e-05, 'checkpoint_manager_blocking_start_time': 1776873021.063199, 'checkpoint_manager_blocking_duration_secs': 44.55583071708679}
I0422 15:51:05.619178 132131956553536 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0422 15:51:05.619222 132131956553536 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0422 15:51:07.496369 132000668964608 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/9/items
I0422 15:51:10.616417 131993470686976 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/9/items/array_metadatas/process_0
I0422 15:51:50.235525 131993462294272 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 41.344952s (commit=40.249971s, array_metadata_write=1.094981s)
I0422 15:51:50.236662 131993453901568 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 35.139 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 44.952327489852905 s) (per-host)
I0422 15:51:50.236785 131993453901568 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 44.646054s.
I0422 15:51:51.789938 131993453901568 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000200s.
I0422 15:51:51.995360 131993453901568 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0422 15:51:53.059291 131993453901568 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.893857s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/9/items
I0422 15:51:53.061083 131993453901568 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/9/items
I0422 15:51:53.504332 131993453901568 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/9
I0422 15:51:54.918562 131993453901568 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/9`.
I0422 15:51:54.919380 131993453901568 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 49.961191s. directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/linen_xpk_feat_nnx_set_defaults_true_20260422_154647_01_base/checkpoints/9
I0422 15:51:54.921193 131993453901568 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 49.330460s.
I0422 15:51:54.921442 131996407813888 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0422 15:51:54.921513 131996407813888 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0422 15:51:54.921573 131996407813888 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0422 15:51:54.923248 131996407813888 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0422 15:51:54.923422 132131956553536 checkpoint_manager.py:2036] [process=0][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0422 15:51:54.923564 132131956553536 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0422 15:51:54.924210 132131956553536 metric_logger.py:185] completed step: 9, seconds: 0.148, TFLOP/s/device: 92.017, Tokens/s/device: 13869.795, 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: Wed Apr 22 15:52:05 UTC 2026
EXIT_CODE=0