XPK Start: Wed Apr 22 18:00:31 UTC 2026 2026-04-22 18:00:35.304328: 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 00:00:1776880835.317144 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1776880835.321130 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1776880835.332262 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776880835.332281 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776880835.332283 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776880835.332285 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-22 18:00:54.414857: 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 18:00:54.934822 140085558994752 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-22 18:01:03,974:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0422 18:01:03.974915 140085558994752 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-22 18:01:03,977:jax._src.distributed:157: Connecting to JAX distributed service on mt-10-shardy-true-umelp-slice-job-0-0.mt-10-shardy-true-umelp:8482 I0422 18:01:03.977197 140085558994752 distributed.py:157] Connecting to JAX distributed service on mt-10-shardy-true-umelp-slice-job-0-0.mt-10-shardy-true-umelp:8482 I0422 18:01:05.151810 140085558994752 max_utils.py:284] Jax distributed system initialized! I0422 18:01:11.954517 140085558994752 max_utils.py:800] System Information: Jax Version: 0.8.1 I0422 18:01:11.954620 140085558994752 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0422 18:01:11.954680 140085558994752 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 18:01:11.954727 140085558994752 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0422 18:01:12.577980 140085558994752 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0422 18:01:12.692499 140085558994752 checkpointing.py:688] Setting up checkpoint logger... I0422 18:01:12.692619 140085558994752 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0422 18:01:12.692677 140085558994752 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0422 18:01:12.692894 140085558994752 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 0x7f67be9752e0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0422 18:01:15.677894 140085558994752 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0422 18:01:15.678139 140085558994752 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 0x7f551c472ea0>}, handler_registry=None I0422 18:01:15.678381 140085558994752 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7f551c472ea0>` 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 18:01:15.678429 140085558994752 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7f5350147410>` 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 18:01:15.678464 140085558994752 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 0x7f551c472ea0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7f551c472ea0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7f5350147410>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7f5350147410>}). I0422 18:01:15.678807 140085558994752 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0422 18:01:15.678884 140085558994752 async_checkpointer.py:177] [process=5][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7f52c8480f40> timeout: 600 secs and primary_host=0 for async checkpoint writes I0422 18:01:18.073561 140085558994752 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/nnx_xpk_feat_nnx_set_defaults_true_20260422_154647_10_shardy_true/checkpoints I0422 18:01:18.490332 140085558994752 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/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/nnx_xpk_feat_nnx_set_defaults_true_20260422_154647_10_shardy_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7f52c82e8200> I0422 18:01:18.490506 140085558994752 checkpointing.py:302] Checkpoint manager created! I0422 18:01:19.418472 140085558994752 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0422 18:01:19.418592 140085558994752 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0422 18:01:20.808171 140085558994752 maxtext_utils.py:1749] decoder/decoder_norm/scale/value Shape: float32[2048] Physical: (None,) I0422 18:01:20.808294 140085558994752 maxtext_utils.py:1749] decoder/layers/mlp/wi_0/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0422 18:01:20.808348 140085558994752 maxtext_utils.py:1749] decoder/layers/mlp/wi_1/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0422 18:01:20.808393 140085558994752 maxtext_utils.py:1749] decoder/layers/mlp/wo/kernel/value Shape: float32[7168,16,2048] Physical: (None, None, 'fsdp') I0422 18:01:20.808433 140085558994752 maxtext_utils.py:1749] decoder/layers/post_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0422 18:01:20.808470 140085558994752 maxtext_utils.py:1749] decoder/layers/pre_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0422 18:01:20.808509 140085558994752 maxtext_utils.py:1749] decoder/layers/self_attention/key/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0422 18:01:20.808546 140085558994752 maxtext_utils.py:1749] decoder/layers/self_attention/out/kernel/value Shape: float32[16,16,128,2048] Physical: (None, None, None, 'fsdp') I0422 18:01:20.808581 140085558994752 maxtext_utils.py:1749] decoder/layers/self_attention/query/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0422 18:01:20.808616 140085558994752 maxtext_utils.py:1749] decoder/layers/self_attention/value/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0422 18:01:20.808664 140085558994752 maxtext_utils.py:1749] decoder/logits_dense/kernel/value Shape: float32[2048,32000] Physical: ('fsdp', None) I0422 18:01:20.808701 140085558994752 maxtext_utils.py:1749] token_embedder/embedding/value Shape: float32[32000,2048] Physical: (None, 'fsdp') I0422 18:01:21.206475 140085558994752 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0422 18:01:21.206619 140085558994752 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0422 18:01:21.212052 140085558994752 nnx_decoders.py:616] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0422 18:01:21.212147 140085558994752 nnx_decoders.py:616] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0422 18:01:21.366747 140085558994752 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0422 18:01:21.366877 140085558994752 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0422 18:01:21.519011 140085558994752 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0422 18:01:21.519149 140085558994752 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0422 18:01:21.686724 140085558994752 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 18:01:21.686889 140085558994752 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0422 18:01:21.840057 140085558994752 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 18:01:21.840191 140085558994752 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0422 18:01:21.988088 140085558994752 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 18:01:21.988220 140085558994752 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0422 18:01:22.145391 140085558994752 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0422 18:01:22.145530 140085558994752 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0422 18:01:22.169260 140085558994752 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0422 18:01:22.169357 140085558994752 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0422 18:01:33.313329 140085558994752 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 18:01:33.315679 140085558994752 metric_logger.py:289] number parameters: 1.104 billion I0422 18:01:44.434989 140085558994752 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0422 18:01:44.585411 140085558994752 checkpointing.py:798] Waited 0.15040922164916992 seconds for step 0 to finish before starting checkpointing. I0422 18:01:44.587737 140085558994752 checkpoint_manager.py:2013] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0422 18:01:44.589383 140085558994752 checkpoint_manager.py:1518] [process=5] Saving checkpoint at step 0 I0422 18:01:44.592174 140085558994752 async_checkpointer.py:452] [process=5] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/nnx_xpk_feat_nnx_set_defaults_true_20260422_154647_10_shardy_true/checkpoints/0. I0422 18:01:45.193437 140085558994752 signaling_client.py:364] Using JaxDistributedSignalingClient I0422 18:01:45.194693 140085558994752 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0422 18:01:45.194751 140085558994752 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 18:01:45.441522 140085558994752 base_pytree_checkpoint_handler.py:153] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.249211s I0422 18:01:45.441733 140085558994752 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.966 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5200188159942627 s) (per-host) I0422 18:01:45.441786 140085558994752 base_pytree_checkpoint_handler.py:737] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.520082s (batch_requests_ready=0.264287s, total_serialization_initiated=0.255723s, others=0.000072s) I0422 18:01:45.441881 140085558994752 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.524260s (all_items=0.000018s, per_item={'items': '0.00001812'}, temp_paths=0.524242) I0422 18:01:45.442804 139959888430848 async_checkpointer.py:79] [process=5][thread=async_save] Background save thread started. I0422 18:01:45.442970 140085558994752 async_checkpointer.py:561] Finished blocking save. Time taken: 0.853515s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/nnx_xpk_feat_nnx_set_defaults_true_20260422_154647_10_shardy_true/checkpoints/0. I0422 18:01:45.480846 140085558994752 checkpoint_manager.py:1566] [process=5][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0422 18:01:45.481155 139958715131648 async_checkpointer.py:265] [process=5][thread=save_finalize] Waiting for background save thread=async_save. I0422 18:01:45.481297 140085558994752 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/nnx_xpk_feat_nnx_set_defaults_true_20260422_154647_10_shardy_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776880904.5877187, 'wait_for_prev_duration_secs': 6.461143493652344e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776880904.5894227, 'checkpointer_blocking_duration_secs': 0.8536550998687744, 'get_old_steps_start_time': 1776880905.443097, 'get_old_steps_duration_secs': 2.6702880859375e-05, 'checkpoint_manager_blocking_start_time': 1776880904.5858324, 'checkpoint_manager_blocking_duration_secs': 0.8954257965087891} I0422 18:01:45.481405 140085558994752 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0422 18:01:45.481476 140085558994752 max_utils.py:750] Memstats: After params initialized: I0422 18:01:45.481525 140085558994752 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_18(process=5,(2,4,0,0)) I0422 18:01:45.481559 140085558994752 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_19(process=5,(3,4,0,0)) I0422 18:01:45.481586 140085558994752 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_22(process=5,(2,5,0,0)) I0422 18:01:45.481609 140085558994752 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_23(process=5,(3,5,0,0)) I0422 18:01:45.752644 140085558994752 metric_logger.py:185] completed step: 0, seconds: 11.118, TFLOP/s/device: 1.222, Tokens/s/device: 184.202, total_weights: 65536, loss: 10.874 I0422 18:01:45.914304 140085558994752 metric_logger.py:185] completed step: 1, seconds: 1.316, TFLOP/s/device: 10.325, Tokens/s/device: 1556.296, total_weights: 65536, loss: 10.874 I0422 18:01:46.322171 140085558994752 metric_logger.py:185] completed step: 2, seconds: 0.016, TFLOP/s/device: 844.236, Tokens/s/device: 127252.392, total_weights: 65536, loss: 10.262 I0422 18:01:46.469520 140085558994752 metric_logger.py:185] completed step: 3, seconds: 0.399, TFLOP/s/device: 34.081, Tokens/s/device: 5137.119, total_weights: 65536, loss: 9.732 I0422 18:01:46.765053 140085558994752 metric_logger.py:185] completed step: 4, seconds: 0.162, TFLOP/s/device: 83.630, Tokens/s/device: 12605.559, total_weights: 65536, loss: 9.274 I0422 18:01:46.771843 140085558994752 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.426, Tokens/s/device: 13931.404, total_weights: 65536, loss: 8.888 I0422 18:02:05.500339 140085558994752 metric_logger.py:185] completed step: 6, seconds: 0.296, TFLOP/s/device: 45.935, Tokens/s/device: 6923.855, total_weights: 65536, loss: 8.589 I0422 18:02:05.647708 140085558994752 metric_logger.py:185] completed step: 7, seconds: 18.582, TFLOP/s/device: 0.731, Tokens/s/device: 110.214, total_weights: 65536, loss: 8.380 I0422 18:02:05.795261 140085558994752 metric_logger.py:185] completed step: 8, seconds: 0.153, TFLOP/s/device: 88.581, Tokens/s/device: 13351.936, total_weights: 65536, loss: 8.251 I0422 18:02:05.943905 140085558994752 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0422 18:02:05.944499 140085558994752 checkpointing.py:798] Waited 0.0006067752838134766 seconds for step 9 to finish before starting checkpointing. I0422 18:02:05.946482 140085558994752 checkpoint_manager.py:2024] [process=5][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0422 18:02:07.461499 2614 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0422 18:02:09.368133 139959896823552 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/nnx_xpk_feat_nnx_set_defaults_true_20260422_154647_10_shardy_true/checkpoints/0/items/array_metadatas/process_5 I0422 18:02:41.047597 139959888430848 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/gbytes_per_sec: 28.144 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 56.12585735321045 s) (per-host) I0422 18:02:41.047717 139959888430848 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 55.604804s. I0422 18:02:49.655129 139959888430848 async_checkpointer.py:144] [process=5][thread=async_save] Background save thread done. Time taken: 64.212199s. I0422 18:02:49.655425 139958715131648 async_checkpointer.py:273] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save. I0422 18:02:49.655550 139958715131648 async_checkpointer.py:283] [process=5][thread=save_finalize] No errors found in background save thread=async_save. I0422 18:02:49.655598 139958715131648 checkpoint_manager.py:2133] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0422 18:02:49.656844 139958715131648 checkpoint_manager.py:2142] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0422 18:02:49.657042 140085558994752 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. W0422 18:02:49.657170 140085558994752 checkpoint_manager.py:1458] Waiting for previous save to complete took 43.710692 seconds. If this number is high, consider checkpointing less frequently. I0422 18:02:49.659102 140085558994752 checkpoint_manager.py:1518] [process=5] Saving checkpoint at step 9 I0422 18:02:49.662538 140085558994752 async_checkpointer.py:452] [process=5] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/nnx_xpk_feat_nnx_set_defaults_true_20260422_154647_10_shardy_true/checkpoints/9. I0422 18:02:50.617442 140085558994752 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0422 18:02:50.617595 140085558994752 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 18:02:50.662431 140085558994752 base_pytree_checkpoint_handler.py:153] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.047009s I0422 18:02:50.662601 140085558994752 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/blocking_gbytes_per_sec: 4.942 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.31215906143188477 s) (per-host) I0422 18:02:50.662668 140085558994752 base_pytree_checkpoint_handler.py:737] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.312220s (batch_requests_ready=0.260460s, total_serialization_initiated=0.051692s, others=0.000068s) I0422 18:02:50.662766 140085558994752 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.316319s (all_items=0.000016s, per_item={'items': '0.00001621'}, temp_paths=0.316303) I0422 18:02:50.663614 139959850510080 async_checkpointer.py:79] [process=5][thread=async_save] Background save thread started. I0422 18:02:50.663785 140085558994752 async_checkpointer.py:561] Finished blocking save. Time taken: 1.004613s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/nnx_xpk_feat_nnx_set_defaults_true_20260422_154647_10_shardy_true/checkpoints/9. I0422 18:02:51.092780 140085558994752 checkpoint_manager.py:1566] [process=5][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0422 18:02:51.093131 139958715131648 async_checkpointer.py:265] [process=5][thread=save_finalize] Waiting for background save thread=async_save. I0422 18:02:51.093301 140085558994752 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/nnx_xpk_feat_nnx_set_defaults_true_20260422_154647_10_shardy_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776880925.9464526, 'wait_for_prev_duration_secs': 43.710691928863525, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776880969.6591394, 'checkpointer_blocking_duration_secs': 1.004753589630127, 'get_old_steps_start_time': 1776880970.6639123, 'get_old_steps_duration_secs': 2.47955322265625e-05, 'checkpoint_manager_blocking_start_time': 1776880925.9447336, 'checkpoint_manager_blocking_duration_secs': 45.148531436920166} I0422 18:02:51.093411 140085558994752 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0422 18:02:51.093456 140085558994752 checkpoint_manager.py:2024] [process=5][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0422 18:02:55.458278 139959896823552 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260422_154647/nnx_xpk_feat_nnx_set_defaults_true_20260422_154647_10_shardy_true/checkpoints/9/items/array_metadatas/process_5 I0422 18:03:32.250000 139959850510080 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/gbytes_per_sec: 37.700 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.899516582489014 s) (per-host) I0422 18:03:32.250131 139959850510080 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.586394s. I0422 18:03:39.965279 139959850510080 async_checkpointer.py:144] [process=5][thread=async_save] Background save thread done. Time taken: 49.301527s. I0422 18:03:39.965556 139958715131648 async_checkpointer.py:273] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save. I0422 18:03:39.965687 139958715131648 async_checkpointer.py:283] [process=5][thread=save_finalize] No errors found in background save thread=async_save. I0422 18:03:39.965742 139958715131648 checkpoint_manager.py:2133] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0422 18:03:39.967377 139958715131648 checkpoint_manager.py:2142] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0422 18:03:39.967546 140085558994752 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. I0422 18:03:39.967716 140085558994752 checkpoint_manager.py:2013] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0422 18:03:39.968283 140085558994752 metric_logger.py:185] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.304, Tokens/s/device: 13913.138, total_weights: 65536, loss: 8.175 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Wed Apr 22 18:03:52 UTC 2026 EXIT_CODE=0