XPK Start: Thu Apr 23 17:34:33 UTC 2026 2026-04-23 17:34:37.166449: 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 00:00:1776965677.178925 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1776965677.182604 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1776965677.193764 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776965677.193784 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776965677.193786 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776965677.193789 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-23 17:34:56.277788: 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 17:34:56.800462 136325455697728 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-23 17:35:05,841:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0423 17:35:05.841244 136325455697728 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-23 17:35:05,843:jax._src.distributed:157: Connecting to JAX distributed service on mt-02-synthetic-9yu6o-slice-job-0-0.mt-02-synthetic-9yu6o:8482 I0423 17:35:05.843597 136325455697728 distributed.py:157] Connecting to JAX distributed service on mt-02-synthetic-9yu6o-slice-job-0-0.mt-02-synthetic-9yu6o:8482 I0423 17:35:07.052488 136325455697728 max_utils.py:284] Jax distributed system initialized! I0423 17:35:13.016023 136325455697728 max_utils.py:800] System Information: Jax Version: 0.8.1 I0423 17:35:13.016129 136325455697728 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0423 17:35:13.016170 136325455697728 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 17:35:13.016207 136325455697728 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0423 17:35:13.632300 136325455697728 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0423 17:35:13.747771 136325455697728 checkpointing.py:688] Setting up checkpoint logger... I0423 17:35:13.747888 136325455697728 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0423 17:35:13.747938 136325455697728 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0423 17:35:13.748142 136325455697728 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 0x7bfc327791c0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0423 17:35:16.492842 136325455697728 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0423 17:35:16.493193 136325455697728 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 0x7be9870143e0>}, handler_registry=None I0423 17:35:16.493431 136325455697728 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7be9870143e0>` 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 17:35:16.493479 136325455697728 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7be88c263cb0>` 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 17:35:16.493515 136325455697728 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 0x7be9870143e0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7be9870143e0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7be88c263cb0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7be88c263cb0>}). I0423 17:35:16.493868 136325455697728 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0423 17:35:16.493947 136325455697728 async_checkpointer.py:177] [process=5][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7be88c24cf40> timeout: 600 secs and primary_host=0 for async checkpoint writes I0423 17:35:18.215093 136325455697728 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_02_synthetic/checkpoints I0423 17:35:18.231618 136325455697728 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_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_02_synthetic/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7be88c2e8680> I0423 17:35:18.231763 136325455697728 checkpointing.py:302] Checkpoint manager created! I0423 17:35:19.155417 136325455697728 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0423 17:35:19.155538 136325455697728 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0423 17:35:20.561287 136325455697728 maxtext_utils.py:1749] decoder/decoder_norm/scale/value Shape: float32[2048] Physical: (None,) I0423 17:35:20.561409 136325455697728 maxtext_utils.py:1749] decoder/layers/mlp/wi_0/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0423 17:35:20.561465 136325455697728 maxtext_utils.py:1749] decoder/layers/mlp/wi_1/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0423 17:35:20.561510 136325455697728 maxtext_utils.py:1749] decoder/layers/mlp/wo/kernel/value Shape: float32[7168,16,2048] Physical: (None, None, 'fsdp') I0423 17:35:20.561549 136325455697728 maxtext_utils.py:1749] decoder/layers/post_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0423 17:35:20.561584 136325455697728 maxtext_utils.py:1749] decoder/layers/pre_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0423 17:35:20.561621 136325455697728 maxtext_utils.py:1749] decoder/layers/self_attention/key/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0423 17:35:20.561669 136325455697728 maxtext_utils.py:1749] decoder/layers/self_attention/out/kernel/value Shape: float32[16,16,128,2048] Physical: (None, None, None, 'fsdp') I0423 17:35:20.561710 136325455697728 maxtext_utils.py:1749] decoder/layers/self_attention/query/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0423 17:35:20.561747 136325455697728 maxtext_utils.py:1749] decoder/layers/self_attention/value/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0423 17:35:20.561783 136325455697728 maxtext_utils.py:1749] decoder/logits_dense/kernel/value Shape: float32[2048,32000] Physical: ('fsdp', None) I0423 17:35:20.561817 136325455697728 maxtext_utils.py:1749] token_embedder/embedding/value Shape: float32[32000,2048] Physical: (None, 'fsdp') I0423 17:35:20.958138 136325455697728 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0423 17:35:20.958282 136325455697728 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 17:35:20.963638 136325455697728 nnx_decoders.py:616] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0423 17:35:20.963750 136325455697728 nnx_decoders.py:616] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 17:35:21.119647 136325455697728 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0423 17:35:21.119802 136325455697728 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 17:35:21.271995 136325455697728 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0423 17:35:21.272128 136325455697728 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 17:35:21.435351 136325455697728 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 17:35:21.435491 136325455697728 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 17:35:21.584977 136325455697728 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 17:35:21.585108 136325455697728 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 17:35:21.733573 136325455697728 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 17:35:21.733731 136325455697728 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 17:35:21.894061 136325455697728 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0423 17:35:21.894204 136325455697728 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 17:35:21.917548 136325455697728 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0423 17:35:21.917645 136325455697728 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0423 17:35:33.086724 136325455697728 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 17:35:33.089137 136325455697728 metric_logger.py:289] number parameters: 1.104 billion I0423 17:35:44.226170 136325455697728 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0423 17:35:44.376593 136325455697728 checkpointing.py:798] Waited 0.15041232109069824 seconds for step 0 to finish before starting checkpointing. I0423 17:35:44.379240 136325455697728 checkpoint_manager.py:2013] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0423 17:35:44.381074 136325455697728 checkpoint_manager.py:1518] [process=5] Saving checkpoint at step 0 I0423 17:35:44.383916 136325455697728 async_checkpointer.py:452] [process=5] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_02_synthetic/checkpoints/0. I0423 17:35:44.986621 136325455697728 signaling_client.py:364] Using JaxDistributedSignalingClient I0423 17:35:44.987516 136325455697728 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0423 17:35:44.987574 136325455697728 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 17:35:45.235482 136325455697728 base_pytree_checkpoint_handler.py:153] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.250013s I0423 17:35:45.235664 136325455697728 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/blocking_gbytes_per_sec: 3.000 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5142250061035156 s) (per-host) I0423 17:35:45.235719 136325455697728 base_pytree_checkpoint_handler.py:737] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.514302s (batch_requests_ready=0.257684s, total_serialization_initiated=0.256534s, others=0.000084s) I0423 17:35:45.235817 136325455697728 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.518310s (all_items=0.000019s, per_item={'items': '0.00001907'}, temp_paths=0.518291) I0423 17:35:45.236819 136202892748544 async_checkpointer.py:79] [process=5][thread=async_save] Background save thread started. I0423 17:35:45.236992 136325455697728 async_checkpointer.py:561] Finished blocking save. Time taken: 0.855846s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_02_synthetic/checkpoints/0. I0423 17:35:45.711464 136325455697728 checkpoint_manager.py:1566] [process=5][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0423 17:35:45.711838 136201315014400 async_checkpointer.py:265] [process=5][thread=save_finalize] Waiting for background save thread=async_save. I0423 17:35:45.712003 136325455697728 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_02_synthetic/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776965744.3792226, 'wait_for_prev_duration_secs': 6.389617919921875e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776965744.3811128, 'checkpointer_blocking_duration_secs': 0.855987548828125, 'get_old_steps_start_time': 1776965745.237123, 'get_old_steps_duration_secs': 2.4080276489257812e-05, 'checkpoint_manager_blocking_start_time': 1776965744.3770287, 'checkpoint_manager_blocking_duration_secs': 1.3349347114562988} I0423 17:35:45.712115 136325455697728 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0423 17:35:45.712197 136325455697728 max_utils.py:750] Memstats: After params initialized: I0423 17:35:45.712255 136325455697728 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_18(process=5,(2,4,0,0)) I0423 17:35:45.712291 136325455697728 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_19(process=5,(3,4,0,0)) I0423 17:35:45.712320 136325455697728 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_22(process=5,(2,5,0,0)) I0423 17:35:45.712347 136325455697728 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_23(process=5,(3,5,0,0)) I0423 17:35:45.982918 136325455697728 metric_logger.py:185] completed step: 0, seconds: 11.136, TFLOP/s/device: 1.220, Tokens/s/device: 183.908, total_weights: 65536, loss: 10.874 I0423 17:35:46.147113 136325455697728 metric_logger.py:185] completed step: 1, seconds: 1.755, TFLOP/s/device: 7.742, Tokens/s/device: 1166.939, total_weights: 65536, loss: 10.874 I0423 17:35:46.548895 136325455697728 metric_logger.py:185] completed step: 2, seconds: 0.018, TFLOP/s/device: 746.013, Tokens/s/device: 112447.153, total_weights: 65536, loss: 10.262 I0423 17:35:46.696496 136325455697728 metric_logger.py:185] completed step: 3, seconds: 0.392, TFLOP/s/device: 34.678, Tokens/s/device: 5227.077, total_weights: 65536, loss: 9.732 I0423 17:35:46.991873 136325455697728 metric_logger.py:185] completed step: 4, seconds: 0.164, TFLOP/s/device: 82.803, Tokens/s/device: 12480.956, total_weights: 65536, loss: 9.274 I0423 17:35:46.999391 136325455697728 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.380, Tokens/s/device: 13924.490, total_weights: 65536, loss: 8.888 I0423 17:36:05.861538 136325455697728 metric_logger.py:185] completed step: 6, seconds: 0.296, TFLOP/s/device: 45.901, Tokens/s/device: 6918.732, total_weights: 65536, loss: 8.589 I0423 17:36:06.008776 136325455697728 metric_logger.py:185] completed step: 7, seconds: 18.716, TFLOP/s/device: 0.726, Tokens/s/device: 109.426, total_weights: 65536, loss: 8.380 I0423 17:36:06.156507 136325455697728 metric_logger.py:185] completed step: 8, seconds: 0.154, TFLOP/s/device: 88.462, Tokens/s/device: 13333.941, total_weights: 65536, loss: 8.251 I0423 17:36:06.304973 136325455697728 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0423 17:36:06.305573 136325455697728 checkpointing.py:798] Waited 0.0006122589111328125 seconds for step 9 to finish before starting checkpointing. I0423 17:36:06.307472 136325455697728 checkpoint_manager.py:2024] [process=5][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0423 17:36:07.055563 2573 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0423 17:36:08.913586 136202901141248 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_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_02_synthetic/checkpoints/0/items/array_metadatas/process_5 I0423 17:36:41.563347 136202892748544 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/gbytes_per_sec: 27.789 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 56.84188270568848 s) (per-host) I0423 17:36:41.563464 136202892748544 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 56.326535s. I0423 17:36:49.076035 136202892748544 async_checkpointer.py:144] [process=5][thread=async_save] Background save thread done. Time taken: 63.839077s. I0423 17:36:49.076326 136201315014400 async_checkpointer.py:273] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save. I0423 17:36:49.076462 136201315014400 async_checkpointer.py:283] [process=5][thread=save_finalize] No errors found in background save thread=async_save. I0423 17:36:49.076514 136201315014400 checkpoint_manager.py:2133] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0423 17:36:49.077963 136201315014400 checkpoint_manager.py:2142] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0423 17:36:49.078134 136325455697728 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 17:36:49.078280 136325455697728 checkpoint_manager.py:1458] Waiting for previous save to complete took 42.770811 seconds. If this number is high, consider checkpointing less frequently. I0423 17:36:49.080171 136325455697728 checkpoint_manager.py:1518] [process=5] Saving checkpoint at step 9 I0423 17:36:49.083770 136325455697728 async_checkpointer.py:452] [process=5] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_02_synthetic/checkpoints/9. I0423 17:36:49.666820 136325455697728 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0423 17:36:49.666981 136325455697728 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 17:36:49.708681 136325455697728 base_pytree_checkpoint_handler.py:153] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.043833s I0423 17:36:49.708849 136325455697728 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/blocking_gbytes_per_sec: 4.741 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.325364351272583 s) (per-host) I0423 17:36:49.708900 136325455697728 base_pytree_checkpoint_handler.py:737] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.325424s (batch_requests_ready=0.276833s, total_serialization_initiated=0.048523s, others=0.000068s) I0423 17:36:49.708995 136325455697728 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.329515s (all_items=0.000015s, per_item={'items': '0.00001550'}, temp_paths=0.329499) I0423 17:36:49.709915 136201860278016 async_checkpointer.py:79] [process=5][thread=async_save] Background save thread started. I0423 17:36:49.710048 136325455697728 async_checkpointer.py:561] Finished blocking save. Time taken: 0.629800s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_02_synthetic/checkpoints/9. I0423 17:36:50.128585 136325455697728 checkpoint_manager.py:1566] [process=5][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0423 17:36:50.128967 136201315014400 async_checkpointer.py:265] [process=5][thread=save_finalize] Waiting for background save thread=async_save. I0423 17:36:50.129138 136325455697728 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_02_synthetic/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776965766.3074436, 'wait_for_prev_duration_secs': 42.77081108093262, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776965809.0802147, 'checkpointer_blocking_duration_secs': 0.6300041675567627, 'get_old_steps_start_time': 1776965809.7102373, 'get_old_steps_duration_secs': 2.5987625122070312e-05, 'checkpoint_manager_blocking_start_time': 1776965766.305806, 'checkpoint_manager_blocking_duration_secs': 43.8232958316803} I0423 17:36:50.129250 136325455697728 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0423 17:36:50.129294 136325455697728 checkpoint_manager.py:2024] [process=5][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0423 17:36:54.656135 136202901141248 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_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_02_synthetic/checkpoints/9/items/array_metadatas/process_5 I0423 17:37:31.705159 136201860278016 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/gbytes_per_sec: 37.324 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.32164812088013 s) (per-host) I0423 17:37:31.705268 136201860278016 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.995140s. I0423 17:37:40.802918 136201860278016 async_checkpointer.py:144] [process=5][thread=async_save] Background save thread done. Time taken: 51.092774s. I0423 17:37:40.803255 136201315014400 async_checkpointer.py:273] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save. I0423 17:37:40.803372 136201315014400 async_checkpointer.py:283] [process=5][thread=save_finalize] No errors found in background save thread=async_save. I0423 17:37:40.803420 136201315014400 checkpoint_manager.py:2133] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0423 17:37:40.804852 136201315014400 checkpoint_manager.py:2142] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0423 17:37:40.805029 136325455697728 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 17:37:40.805214 136325455697728 checkpoint_manager.py:2013] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0423 17:37:40.805862 136325455697728 metric_logger.py:185] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.512, Tokens/s/device: 13944.399, 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: Thu Apr 23 17:37:53 UTC 2026 EXIT_CODE=0