XPK Start: Thu Apr 23 16:34:11 UTC 2026 2026-04-23 16:34:15.313454: 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 00:00:1776962055.326039 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1776962055.329732 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1776962055.340997 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776962055.341019 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776962055.341022 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776962055.341024 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-23 16:34:34.475951: 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 16:34:35.001010 140510344148800 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-23 16:34:44,042:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0423 16:34:44.042166 140510344148800 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-23 16:34:44,044:jax._src.distributed:157: Connecting to JAX distributed service on mt-11-optimizer-offload-bheqe-slice-job-0-0.mt-11-optimizer-offload-bheqe:8482 I0423 16:34:44.044420 140510344148800 distributed.py:157] Connecting to JAX distributed service on mt-11-optimizer-offload-bheqe-slice-job-0-0.mt-11-optimizer-offload-bheqe:8482 I0423 16:34:45.243535 140510344148800 max_utils.py:284] Jax distributed system initialized! I0423 16:34:51.173487 140510344148800 max_utils.py:800] System Information: Jax Version: 0.8.1 I0423 16:34:51.173595 140510344148800 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0423 16:34:51.173636 140510344148800 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 16:34:51.173692 140510344148800 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0423 16:34:51.787637 140510344148800 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0423 16:34:51.788255 140510344148800 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0423 16:34:51.788437 140510344148800 checkpointing.py:688] Setting up checkpoint logger... I0423 16:34:51.788487 140510344148800 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0423 16:34:51.788529 140510344148800 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0423 16:34:51.788899 140510344148800 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 0x7fcaa85790d0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0423 16:34:54.755096 140510344148800 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0423 16:34:54.755438 140510344148800 checkpoint_manager.py:709] [process=5][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7fb82aeb9c10>}, handler_registry=None I0423 16:34:54.755689 140510344148800 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7fb82aeb9c10>` 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 16:34:54.755740 140510344148800 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7fb82aeb3aa0>` 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 16:34:54.755776 140510344148800 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 0x7fb82aeb9c10>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7fb82aeb9c10>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7fb82aeb3aa0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7fb82aeb3aa0>}). I0423 16:34:54.756340 140510344148800 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0423 16:34:54.756423 140510344148800 async_checkpointer.py:177] [process=5][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7fb82aa0c180> timeout: 600 secs and primary_host=0 for async checkpoint writes I0423 16:34:55.424704 140510344148800 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_true/checkpoints I0423 16:34:55.614706 140510344148800 checkpoint_manager.py:929] [process=5][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_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7fb82aeb9670> I0423 16:34:55.614883 140510344148800 checkpointing.py:302] Checkpoint manager created! I0423 16:34:56.512897 140510344148800 nnx_wrappers.py:455] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0423 16:34:56.513009 140510344148800 nnx_wrappers.py:455] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 16:34:56.881521 140510344148800 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0423 16:34:56.881614 140510344148800 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 16:34:56.897994 140510344148800 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0423 16:34:56.898054 140510344148800 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 16:34:56.921683 140510344148800 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0423 16:34:56.921752 140510344148800 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 16:34:56.937911 140510344148800 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0423 16:34:56.937984 140510344148800 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 16:34:56.953980 140510344148800 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0423 16:34:56.954042 140510344148800 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 16:34:56.979085 140510344148800 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0423 16:34:56.979151 140510344148800 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 16:34:56.998707 140510344148800 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0423 16:34:56.998778 140510344148800 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0423 16:34:57.190568 140510344148800 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0423 16:34:57.190692 140510344148800 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0423 16:34:59.021926 140510344148800 maxtext_utils.py:1740] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: PartitionSpec('norm',) Physical: (None,) I0423 16:34:59.022077 140510344148800 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) I0423 16:34:59.022135 140510344148800 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) I0423 16:34:59.022204 140510344148800 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') I0423 16:34:59.022261 140510344148800 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) I0423 16:34:59.022300 140510344148800 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) I0423 16:34:59.022356 140510344148800 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) I0423 16:34:59.022411 140510344148800 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') I0423 16:34:59.022453 140510344148800 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) I0423 16:34:59.022492 140510344148800 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) I0423 16:34:59.022541 140510344148800 maxtext_utils.py:1740] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: PartitionSpec('embed', 'vocab') Physical: ('fsdp', None) I0423 16:34:59.022590 140510344148800 maxtext_utils.py:1740] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: PartitionSpec('vocab', 'embed') Physical: (None, 'fsdp') I0423 16:34:59.536114 140510344148800 train.py:158] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0423 16:34:59.536213 140510344148800 train.py:158] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0423 16:34:59.551340 140510344148800 train.py:165] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0423 16:34:59.551398 140510344148800 train.py:165] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0423 16:35:10.289678 140510344148800 max_utils.py:791] Total memory size: 1.3 GB, Output size: 0.1 GB, Temp size: 1.1 GB, Argument size: 0.1 GB, Host temp size: 0.0 GB. I0423 16:35:10.290559 140510344148800 metric_logger.py:289] number parameters: 1.104 billion I0423 16:35:21.700846 140510344148800 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0423 16:35:21.871273 140510344148800 checkpointing.py:798] Waited 0.1704103946685791 seconds for step 0 to finish before starting checkpointing. I0423 16:35:21.873891 140510344148800 checkpoint_manager.py:2013] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0423 16:35:21.875473 140510344148800 checkpoint_manager.py:1518] [process=5] Saving checkpoint at step 0 I0423 16:35:21.878241 140510344148800 async_checkpointer.py:452] [process=5] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_true/checkpoints/0. I0423 16:35:22.480765 140510344148800 signaling_client.py:364] Using JaxDistributedSignalingClient I0423 16:35:22.481746 140510344148800 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0423 16:35:22.481804 140510344148800 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 I0423 16:35:22.729709 140510344148800 base_pytree_checkpoint_handler.py:153] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.249025s I0423 16:35:22.729883 140510344148800 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.876 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.536290168762207 s) (per-host) I0423 16:35:22.729945 140510344148800 base_pytree_checkpoint_handler.py:737] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.536363s (batch_requests_ready=0.282390s, total_serialization_initiated=0.253891s, others=0.000081s) I0423 16:35:22.730063 140510344148800 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.540608s (all_items=0.000020s, per_item={'items': '0.00001979'}, temp_paths=0.540589) I0423 16:35:22.731083 140381291255552 async_checkpointer.py:79] [process=5][thread=async_save] Background save thread started. I0423 16:35:22.731213 140510344148800 async_checkpointer.py:561] Finished blocking save. Time taken: 0.855671s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_true/checkpoints/0. I0423 16:35:22.733288 140510344148800 checkpoint_manager.py:1566] [process=5][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0423 16:35:22.733536 140379311560448 async_checkpointer.py:265] [process=5][thread=save_finalize] Waiting for background save thread=async_save. I0423 16:35:22.733702 140510344148800 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776962121.8738735, 'wait_for_prev_duration_secs': 6.198883056640625e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776962121.8755097, 'checkpointer_blocking_duration_secs': 0.8558883666992188, 'get_old_steps_start_time': 1776962122.7314172, 'get_old_steps_duration_secs': 2.765655517578125e-05, 'checkpoint_manager_blocking_start_time': 1776962121.871876, 'checkpoint_manager_blocking_duration_secs': 0.8617870807647705} I0423 16:35:22.733807 140510344148800 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0423 16:35:22.733857 140510344148800 max_utils.py:750] Memstats: After params initialized: I0423 16:35:22.733907 140510344148800 max_utils.py:756] Using (GB) 0.17 / 31.25 (0.544000%) on TPU_18(process=5,(2,4,0,0)) I0423 16:35:22.733941 140510344148800 max_utils.py:756] Using (GB) 0.17 / 31.25 (0.544000%) on TPU_19(process=5,(3,4,0,0)) I0423 16:35:22.733976 140510344148800 max_utils.py:756] Using (GB) 0.17 / 31.25 (0.544000%) on TPU_22(process=5,(2,5,0,0)) I0423 16:35:22.734002 140510344148800 max_utils.py:756] Using (GB) 0.17 / 31.25 (0.544000%) on TPU_23(process=5,(3,5,0,0)) I0423 16:35:22.993869 140510344148800 metric_logger.py:185] completed step: 0, seconds: 11.410, TFLOP/s/device: 1.191, Tokens/s/device: 179.489, total_weights: 65536, loss: 10.874 I0423 16:35:23.190193 140510344148800 metric_logger.py:185] completed step: 1, seconds: 1.292, TFLOP/s/device: 10.519, Tokens/s/device: 1585.509, total_weights: 65536, loss: 10.874 I0423 16:35:23.619082 140510344148800 metric_logger.py:185] completed step: 2, seconds: 0.033, TFLOP/s/device: 411.096, Tokens/s/device: 61964.842, total_weights: 65536, loss: 10.263 I0423 16:35:23.782721 140510344148800 metric_logger.py:185] completed step: 3, seconds: 0.429, TFLOP/s/device: 31.649, Tokens/s/device: 4770.434, total_weights: 65536, loss: 9.735 I0423 16:35:24.110382 140510344148800 metric_logger.py:185] completed step: 4, seconds: 0.169, TFLOP/s/device: 80.542, Tokens/s/device: 12140.109, total_weights: 65536, loss: 9.278 I0423 16:35:24.115765 140510344148800 metric_logger.py:185] completed step: 5, seconds: 0.164, TFLOP/s/device: 83.031, Tokens/s/device: 12515.277, total_weights: 65536, loss: 8.892 I0423 16:35:43.080582 140510344148800 metric_logger.py:185] completed step: 6, seconds: 0.328, TFLOP/s/device: 41.452, Tokens/s/device: 6248.036, total_weights: 65536, loss: 8.592 I0423 16:35:43.244188 140510344148800 metric_logger.py:185] completed step: 7, seconds: 18.802, TFLOP/s/device: 0.723, Tokens/s/device: 108.925, total_weights: 65536, loss: 8.384 I0423 16:35:43.407907 140510344148800 metric_logger.py:185] completed step: 8, seconds: 0.168, TFLOP/s/device: 80.839, Tokens/s/device: 12184.964, total_weights: 65536, loss: 8.255 I0423 16:35:43.571215 140510344148800 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0423 16:35:43.571806 140510344148800 checkpointing.py:798] Waited 0.0006148815155029297 seconds for step 9 to finish before starting checkpointing. I0423 16:35:43.573987 140510344148800 checkpoint_manager.py:2024] [process=5][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0423 16:35:44.232264 2568 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0423 16:35:46.186680 140381299648256 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_true/checkpoints/0/items/array_metadatas/process_5 I0423 16:36:18.175355 140381291255552 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/gbytes_per_sec: 28.215 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 55.98172473907471 s) (per-host) I0423 16:36:18.175471 140381291255552 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 55.444174s. I0423 16:36:26.292115 140381291255552 async_checkpointer.py:144] [process=5][thread=async_save] Background save thread done. Time taken: 63.560802s. I0423 16:36:26.292411 140379311560448 async_checkpointer.py:273] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save. I0423 16:36:26.292538 140379311560448 async_checkpointer.py:283] [process=5][thread=save_finalize] No errors found in background save thread=async_save. I0423 16:36:26.292590 140379311560448 checkpoint_manager.py:2133] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0423 16:36:26.294515 140379311560448 checkpoint_manager.py:2142] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0423 16:36:26.294703 140510344148800 checkpoint_manager.py:2036] [process=5][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0423 16:36:26.294831 140510344148800 checkpoint_manager.py:1458] Waiting for previous save to complete took 42.720850 seconds. If this number is high, consider checkpointing less frequently. I0423 16:36:26.296754 140510344148800 checkpoint_manager.py:1518] [process=5] Saving checkpoint at step 9 I0423 16:36:26.300129 140510344148800 async_checkpointer.py:452] [process=5] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_true/checkpoints/9. I0423 16:36:27.082789 140510344148800 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0423 16:36:27.082891 140510344148800 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 I0423 16:36:27.126131 140510344148800 base_pytree_checkpoint_handler.py:153] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.044362s I0423 16:36:27.126304 140510344148800 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/blocking_gbytes_per_sec: 4.972 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.31025099754333496 s) (per-host) I0423 16:36:27.126357 140510344148800 base_pytree_checkpoint_handler.py:737] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.310313s (batch_requests_ready=0.264056s, total_serialization_initiated=0.046187s, others=0.000071s) I0423 16:36:27.126442 140510344148800 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.314585s (all_items=0.000015s, per_item={'items': '0.00001478'}, temp_paths=0.314570) I0423 16:36:27.127327 140380771170048 async_checkpointer.py:79] [process=5][thread=async_save] Background save thread started. I0423 16:36:27.127493 140510344148800 async_checkpointer.py:561] Finished blocking save. Time taken: 0.830669s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_true/checkpoints/9. I0423 16:36:27.134976 140510344148800 checkpoint_manager.py:1566] [process=5][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0423 16:36:27.135279 140379311560448 async_checkpointer.py:265] [process=5][thread=save_finalize] Waiting for background save thread=async_save. I0423 16:36:27.135442 140510344148800 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776962143.5739572, 'wait_for_prev_duration_secs': 42.72084975242615, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776962186.296792, 'checkpointer_blocking_duration_secs': 0.8308079242706299, 'get_old_steps_start_time': 1776962187.1276171, 'get_old_steps_duration_secs': 2.6464462280273438e-05, 'checkpoint_manager_blocking_start_time': 1776962143.5720358, 'checkpoint_manager_blocking_duration_secs': 43.56336975097656} I0423 16:36:27.135550 140510344148800 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0423 16:36:27.135595 140510344148800 checkpoint_manager.py:2024] [process=5][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0423 16:36:33.038238 140381299648256 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_true/checkpoints/9/items/array_metadatas/process_5 I0423 16:37:10.176601 140380771170048 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/gbytes_per_sec: 36.428 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 43.36050772666931 s) (per-host) I0423 16:37:10.176733 140380771170048 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 43.049292s. I0423 16:37:17.579600 140380771170048 async_checkpointer.py:144] [process=5][thread=async_save] Background save thread done. Time taken: 50.452144s. I0423 16:37:17.579935 140379311560448 async_checkpointer.py:273] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save. I0423 16:37:17.580060 140379311560448 async_checkpointer.py:283] [process=5][thread=save_finalize] No errors found in background save thread=async_save. I0423 16:37:17.580107 140379311560448 checkpoint_manager.py:2133] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0423 16:37:17.582296 140379311560448 checkpoint_manager.py:2142] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0423 16:37:17.582465 140510344148800 checkpoint_manager.py:2036] [process=5][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0423 16:37:17.582623 140510344148800 checkpoint_manager.py:2013] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0423 16:37:17.583361 140510344148800 metric_logger.py:185] completed step: 9, seconds: 0.164, TFLOP/s/device: 83.033, Tokens/s/device: 12515.660, 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: Thu Apr 23 16:37:27 UTC 2026 EXIT_CODE=0