XPK Start: Thu Apr 23 16:32:56 UTC 2026 2026-04-23 16:33:00.418905: 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 00:00:1776961980.431827 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1776961980.435815 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1776961980.447276 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776961980.447296 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776961980.447299 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776961980.447301 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-23 16:33:19.658010: 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 16:33:20.185155 133694201706304 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-23 16:33:29,225:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0423 16:33:29.225794 133694201706304 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-23 16:33:29,228:jax._src.distributed:157: Connecting to JAX distributed service on mt-11-optimizer-offload-txkqz-slice-job-0-0.mt-11-optimizer-offload-txkqz:8482 I0423 16:33:29.228200 133694201706304 distributed.py:157] Connecting to JAX distributed service on mt-11-optimizer-offload-txkqz-slice-job-0-0.mt-11-optimizer-offload-txkqz:8482 I0423 16:33:30.310710 133694201706304 max_utils.py:284] Jax distributed system initialized! I0423 16:33:36.209433 133694201706304 max_utils.py:800] System Information: Jax Version: 0.8.1 I0423 16:33:36.209539 133694201706304 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0423 16:33:36.209579 133694201706304 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 16:33:36.209613 133694201706304 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0423 16:33:36.831671 133694201706304 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0423 16:33:36.832518 133694201706304 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0423 16:33:36.832751 133694201706304 checkpointing.py:688] Setting up checkpoint logger... I0423 16:33:36.832811 133694201706304 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0423 16:33:36.832865 133694201706304 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0423 16:33:36.833273 133694201706304 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 0x79979c975070>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0423 16:33:39.775850 133694201706304 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0423 16:33:39.776247 133694201706304 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 0x7984aef72000>}, handler_registry=None I0423 16:33:39.776504 133694201706304 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7984aef72000>` 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 16:33:39.776557 133694201706304 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7984af00aba0>` 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 16:33:39.776595 133694201706304 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 0x7984aef72000>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7984aef72000>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7984af00aba0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7984af00aba0>}). I0423 16:33:39.777108 133694201706304 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0423 16:33:39.777190 133694201706304 async_checkpointer.py:177] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7984504fc180> timeout: 600 secs and primary_host=0 for async checkpoint writes I0423 16:33:40.883853 133694201706304 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_false/checkpoints I0423 16:33:40.886636 133694201706304 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_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7984af00b4d0> I0423 16:33:40.886752 133694201706304 checkpointing.py:302] Checkpoint manager created! I0423 16:33:41.780565 133694201706304 nnx_wrappers.py:455] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0423 16:33:41.780672 133694201706304 nnx_wrappers.py:455] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 16:33:42.154793 133694201706304 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0423 16:33:42.154891 133694201706304 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 16:33:42.171125 133694201706304 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0423 16:33:42.171184 133694201706304 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 16:33:42.194365 133694201706304 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 16:33:42.194433 133694201706304 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 16:33:42.210687 133694201706304 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 16:33:42.210750 133694201706304 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 16:33:42.226804 133694201706304 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 16:33:42.226867 133694201706304 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 16:33:42.251857 133694201706304 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0423 16:33:42.251925 133694201706304 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 16:33:42.271814 133694201706304 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0423 16:33:42.271880 133694201706304 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0423 16:33:42.463402 133694201706304 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0423 16:33:42.463515 133694201706304 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0423 16:33:43.910197 133694201706304 maxtext_utils.py:1740] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: PartitionSpec('norm',) Physical: (None,) I0423 16:33:43.910332 133694201706304 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 16:33:43.910388 133694201706304 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 16:33:43.910448 133694201706304 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 16:33:43.910501 133694201706304 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 16:33:43.910540 133694201706304 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 16:33:43.910600 133694201706304 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 16:33:43.910658 133694201706304 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 16:33:43.910702 133694201706304 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 16:33:43.910742 133694201706304 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 16:33:43.910792 133694201706304 maxtext_utils.py:1740] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: PartitionSpec('embed', 'vocab') Physical: ('fsdp', None) I0423 16:33:43.910840 133694201706304 maxtext_utils.py:1740] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: PartitionSpec('vocab', 'embed') Physical: (None, 'fsdp') I0423 16:33:44.418781 133694201706304 train.py:158] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0423 16:33:44.418879 133694201706304 train.py:158] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0423 16:33:44.434003 133694201706304 train.py:165] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0423 16:33:44.434064 133694201706304 train.py:165] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0423 16:33:55.291489 133694201706304 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. I0423 16:33:55.292349 133694201706304 metric_logger.py:289] number parameters: 1.104 billion I0423 16:34:06.492503 133694201706304 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0423 16:34:06.648331 133694201706304 checkpointing.py:798] Waited 0.15581464767456055 seconds for step 0 to finish before starting checkpointing. I0423 16:34:06.651029 133694201706304 checkpoint_manager.py:2013] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0423 16:34:06.652812 133694201706304 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 0 I0423 16:34:06.655539 133694201706304 async_checkpointer.py:452] [process=6] 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_false/checkpoints/0. I0423 16:34:07.235621 133694201706304 signaling_client.py:364] Using JaxDistributedSignalingClient I0423 16:34:07.236654 133694201706304 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0423 16:34:07.236709 133694201706304 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 16:34:07.483089 133694201706304 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.247553s I0423 16:34:07.483282 133694201706304 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.905 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5310413837432861 s) (per-host) I0423 16:34:07.483336 133694201706304 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.531106s (batch_requests_ready=0.279947s, total_serialization_initiated=0.251086s, others=0.000073s) I0423 16:34:07.483427 133694201706304 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.535203s (all_items=0.000019s, per_item={'items': '0.00001884'}, temp_paths=0.535184) I0423 16:34:07.484387 133572049516288 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started. I0423 16:34:07.484557 133694201706304 async_checkpointer.py:561] Finished blocking save. Time taken: 0.831676s. 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_false/checkpoints/0. I0423 16:34:07.604732 133694201706304 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0423 16:34:07.605133 133570949519104 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0423 16:34:07.605306 133694201706304 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_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776962046.6510105, 'wait_for_prev_duration_secs': 6.389617919921875e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776962046.6528502, 'checkpointer_blocking_duration_secs': 0.8318166732788086, 'get_old_steps_start_time': 1776962047.4846869, 'get_old_steps_duration_secs': 2.5272369384765625e-05, 'checkpoint_manager_blocking_start_time': 1776962046.6488943, 'checkpoint_manager_blocking_duration_secs': 0.956369161605835} I0423 16:34:07.605414 133694201706304 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0423 16:34:07.605467 133694201706304 max_utils.py:750] Memstats: After params initialized: I0423 16:34:07.605525 133694201706304 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_24(process=6,(0,6,0,0)) I0423 16:34:07.605559 133694201706304 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_25(process=6,(1,6,0,0)) I0423 16:34:07.605595 133694201706304 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_28(process=6,(0,7,0,0)) I0423 16:34:07.605622 133694201706304 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_29(process=6,(1,7,0,0)) I0423 16:34:07.866454 133694201706304 metric_logger.py:185] completed step: 0, seconds: 11.200, TFLOP/s/device: 1.213, Tokens/s/device: 182.856, total_weights: 65536, loss: 10.874 I0423 16:34:08.042018 133694201706304 metric_logger.py:185] completed step: 1, seconds: 1.373, TFLOP/s/device: 9.898, Tokens/s/device: 1491.982, total_weights: 65536, loss: 10.874 I0423 16:34:08.486516 133694201706304 metric_logger.py:185] completed step: 2, seconds: 0.028, TFLOP/s/device: 478.471, Tokens/s/device: 72120.294, total_weights: 65536, loss: 10.263 I0423 16:34:08.633933 133694201706304 metric_logger.py:185] completed step: 3, seconds: 0.445, TFLOP/s/device: 30.521, Tokens/s/device: 4600.428, total_weights: 65536, loss: 9.734 I0423 16:34:08.929221 133694201706304 metric_logger.py:185] completed step: 4, seconds: 0.152, TFLOP/s/device: 89.223, Tokens/s/device: 13448.645, total_weights: 65536, loss: 9.278 I0423 16:34:08.934689 133694201706304 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.323, Tokens/s/device: 13915.974, total_weights: 65536, loss: 8.892 I0423 16:34:27.916116 133694201706304 metric_logger.py:185] completed step: 6, seconds: 0.296, TFLOP/s/device: 45.939, Tokens/s/device: 6924.393, total_weights: 65536, loss: 8.592 I0423 16:34:28.063771 133694201706304 metric_logger.py:185] completed step: 7, seconds: 18.835, TFLOP/s/device: 0.721, Tokens/s/device: 108.735, total_weights: 65536, loss: 8.384 I0423 16:34:28.211123 133694201706304 metric_logger.py:185] completed step: 8, seconds: 0.152, TFLOP/s/device: 89.476, Tokens/s/device: 13486.727, total_weights: 65536, loss: 8.255 I0423 16:34:28.358212 133694201706304 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0423 16:34:28.358864 133694201706304 checkpointing.py:798] Waited 0.0006709098815917969 seconds for step 9 to finish before starting checkpointing. I0423 16:34:28.360897 133694201706304 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0423 16:34:29.554565 2558 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0423 16:34:31.481290 133572057908992 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_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_false/checkpoints/0/items/array_metadatas/process_6 I0423 16:35:01.568536 133572049516288 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 28.922 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 54.61625647544861 s) (per-host) I0423 16:35:01.568655 133572049516288 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 54.084156s. I0423 16:35:10.999173 133572049516288 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 63.514660s. I0423 16:35:10.999467 133570949519104 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0423 16:35:10.999583 133570949519104 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0423 16:35:10.999633 133570949519104 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0423 16:35:11.001835 133570949519104 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0423 16:35:11.001995 133694201706304 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. W0423 16:35:11.002076 133694201706304 checkpoint_manager.py:1458] Waiting for previous save to complete took 42.641186 seconds. If this number is high, consider checkpointing less frequently. I0423 16:35:11.004069 133694201706304 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 9 I0423 16:35:11.007445 133694201706304 async_checkpointer.py:452] [process=6] 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_false/checkpoints/9. I0423 16:35:11.549868 133694201706304 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0423 16:35:11.549966 133694201706304 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 16:35:11.594777 133694201706304 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.045807s I0423 16:35:11.594926 133694201706304 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.015 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.3075985908508301 s) (per-host) I0423 16:35:11.594976 133694201706304 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.307658s (batch_requests_ready=0.259910s, total_serialization_initiated=0.047683s, others=0.000065s) I0423 16:35:11.595058 133694201706304 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.311574s (all_items=0.000015s, per_item={'items': '0.00001550'}, temp_paths=0.311559) I0423 16:35:11.595961 133571478873856 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started. I0423 16:35:11.596130 133694201706304 async_checkpointer.py:561] Finished blocking save. Time taken: 0.591965s. 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_false/checkpoints/9. I0423 16:35:11.608383 133694201706304 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0423 16:35:11.608650 133570949519104 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0423 16:35:11.608813 133694201706304 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_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776962068.3608692, 'wait_for_prev_duration_secs': 42.641186475753784, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776962111.0041227, 'checkpointer_blocking_duration_secs': 0.5921149253845215, 'get_old_steps_start_time': 1776962111.596257, 'get_old_steps_duration_secs': 2.4318695068359375e-05, 'checkpoint_manager_blocking_start_time': 1776962068.3590822, 'checkpoint_manager_blocking_duration_secs': 43.24969482421875} I0423 16:35:11.608921 133694201706304 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0423 16:35:11.608967 133694201706304 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0423 16:35:16.486967 133572057908992 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_20260423_155242/linen_xpk_feat_nnx_set_defaults_true_20260423_155242_11_optimizer_offload_false/checkpoints/9/items/array_metadatas/process_6 I0423 16:35:53.665942 133571478873856 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 37.273 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.37857151031494 s) (per-host) I0423 16:35:53.666060 133571478873856 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.069990s. I0423 16:36:03.217746 133571478873856 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 51.621661s. I0423 16:36:03.218037 133570949519104 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0423 16:36:03.218192 133570949519104 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0423 16:36:03.218239 133570949519104 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0423 16:36:03.219860 133570949519104 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0423 16:36:03.220027 133694201706304 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. I0423 16:36:03.220180 133694201706304 checkpoint_manager.py:2013] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0423 16:36:03.220809 133694201706304 metric_logger.py:185] completed step: 9, seconds: 0.148, TFLOP/s/device: 91.985, Tokens/s/device: 13864.912, 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 16:36:16 UTC 2026 EXIT_CODE=0