XPK Start: Thu Apr 23 18:01:53 UTC 2026 2026-04-23 18:01:57.610083: 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 00:00:1776967317.622855 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1776967317.626496 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1776967317.637667 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776967317.637684 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776967317.637686 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776967317.637688 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-23 18:02:16.759630: 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 18:02:17.284413 138354608629568 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-23 18:02:26,324:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0423 18:02:26.324712 138354608629568 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-23 18:02:26,327:jax._src.distributed:157: Connecting to JAX distributed service on mt-10-shardy-true-9amlk-slice-job-0-0.mt-10-shardy-true-9amlk:8482 I0423 18:02:26.327025 138354608629568 distributed.py:157] Connecting to JAX distributed service on mt-10-shardy-true-9amlk-slice-job-0-0.mt-10-shardy-true-9amlk:8482 I0423 18:02:27.499909 138354608629568 max_utils.py:284] Jax distributed system initialized! I0423 18:02:34.592886 138354608629568 max_utils.py:800] System Information: Jax Version: 0.8.1 I0423 18:02:34.592992 138354608629568 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0423 18:02:34.593032 138354608629568 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 18:02:34.593065 138354608629568 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0423 18:02:35.262140 138354608629568 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0423 18:02:35.378590 138354608629568 checkpointing.py:688] Setting up checkpoint logger... I0423 18:02:35.378728 138354608629568 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0423 18:02:35.378773 138354608629568 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0423 18:02:35.378981 138354608629568 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 0x7dd4af771160>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0423 18:02:38.220399 138354608629568 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0423 18:02:38.220784 138354608629568 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 0x7dc22d9b41a0>}, handler_registry=None I0423 18:02:38.221074 138354608629568 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7dc22d9b41a0>` 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 18:02:38.221133 138354608629568 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7dc22f44c590>` 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 18:02:38.221178 138354608629568 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 0x7dc22d9b41a0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7dc22d9b41a0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7dc22f44c590>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7dc22f44c590>}). I0423 18:02:38.221628 138354608629568 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0423 18:02:38.221735 138354608629568 async_checkpointer.py:177] [process=5][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7dc22f3f0f40> timeout: 600 secs and primary_host=0 for async checkpoint writes I0423 18:02:39.821006 138354608629568 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_10_shardy_true/checkpoints I0423 18:02:39.836955 138354608629568 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_10_shardy_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7dc22f4752b0> I0423 18:02:39.837083 138354608629568 checkpointing.py:302] Checkpoint manager created! I0423 18:02:40.770992 138354608629568 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0423 18:02:40.771106 138354608629568 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0423 18:02:42.201271 138354608629568 maxtext_utils.py:1749] decoder/decoder_norm/scale/value Shape: float32[2048] Physical: (None,) I0423 18:02:42.201394 138354608629568 maxtext_utils.py:1749] decoder/layers/mlp/wi_0/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0423 18:02:42.201451 138354608629568 maxtext_utils.py:1749] decoder/layers/mlp/wi_1/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0423 18:02:42.201496 138354608629568 maxtext_utils.py:1749] decoder/layers/mlp/wo/kernel/value Shape: float32[7168,16,2048] Physical: (None, None, 'fsdp') I0423 18:02:42.201535 138354608629568 maxtext_utils.py:1749] decoder/layers/post_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0423 18:02:42.201571 138354608629568 maxtext_utils.py:1749] decoder/layers/pre_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0423 18:02:42.201610 138354608629568 maxtext_utils.py:1749] decoder/layers/self_attention/key/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0423 18:02:42.201647 138354608629568 maxtext_utils.py:1749] decoder/layers/self_attention/out/kernel/value Shape: float32[16,16,128,2048] Physical: (None, None, None, 'fsdp') I0423 18:02:42.201698 138354608629568 maxtext_utils.py:1749] decoder/layers/self_attention/query/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0423 18:02:42.201736 138354608629568 maxtext_utils.py:1749] decoder/layers/self_attention/value/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0423 18:02:42.201778 138354608629568 maxtext_utils.py:1749] decoder/logits_dense/kernel/value Shape: float32[2048,32000] Physical: ('fsdp', None) I0423 18:02:42.201814 138354608629568 maxtext_utils.py:1749] token_embedder/embedding/value Shape: float32[32000,2048] Physical: (None, 'fsdp') I0423 18:02:42.600353 138354608629568 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0423 18:02:42.600503 138354608629568 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 18:02:42.606010 138354608629568 nnx_decoders.py:616] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0423 18:02:42.606108 138354608629568 nnx_decoders.py:616] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 18:02:42.770606 138354608629568 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0423 18:02:42.770756 138354608629568 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 18:02:42.932450 138354608629568 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0423 18:02:42.932586 138354608629568 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0423 18:02:43.107523 138354608629568 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 18:02:43.107673 138354608629568 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 18:02:43.261230 138354608629568 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 18:02:43.261362 138354608629568 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 18:02:43.413852 138354608629568 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 18:02:43.413985 138354608629568 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 18:02:43.577172 138354608629568 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0423 18:02:43.577308 138354608629568 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0423 18:02:43.599954 138354608629568 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0423 18:02:43.600050 138354608629568 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0423 18:02:54.509407 138354608629568 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 18:02:54.511777 138354608629568 metric_logger.py:289] number parameters: 1.104 billion I0423 18:03:05.588273 138354608629568 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0423 18:03:05.943687 138354608629568 checkpointing.py:798] Waited 0.3553638458251953 seconds for step 0 to finish before starting checkpointing. I0423 18:03:05.945887 138354608629568 checkpoint_manager.py:2013] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0423 18:03:05.947877 138354608629568 checkpoint_manager.py:1518] [process=5] Saving checkpoint at step 0 I0423 18:03:05.950474 138354608629568 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_10_shardy_true/checkpoints/0. I0423 18:03:07.038684 138354608629568 signaling_client.py:364] Using JaxDistributedSignalingClient I0423 18:03:07.039864 138354608629568 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0423 18:03:07.039925 138354608629568 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 18:03:07.289146 138354608629568 base_pytree_checkpoint_handler.py:153] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.251584s I0423 18:03:07.289324 138354608629568 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/blocking_gbytes_per_sec: 1.559 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.9892098903656006 s) (per-host) I0423 18:03:07.289377 138354608629568 base_pytree_checkpoint_handler.py:737] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.989274s (batch_requests_ready=0.731069s, total_serialization_initiated=0.258132s, others=0.000073s) I0423 18:03:07.289476 138354608629568 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.993255s (all_items=0.000019s, per_item={'items': '0.00001907'}, temp_paths=0.993236) I0423 18:03:07.290425 138232616433408 async_checkpointer.py:79] [process=5][thread=async_save] Background save thread started. I0423 18:03:07.290593 138354608629568 async_checkpointer.py:561] Finished blocking save. Time taken: 1.342648s. 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_10_shardy_true/checkpoints/0. I0423 18:03:07.292437 138354608629568 checkpoint_manager.py:1566] [process=5][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0423 18:03:07.292744 138229932078848 async_checkpointer.py:265] [process=5][thread=save_finalize] Waiting for background save thread=async_save. I0423 18:03:07.292871 138354608629568 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_10_shardy_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776967385.9458683, 'wait_for_prev_duration_secs': 6.4849853515625e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776967385.9479165, 'checkpointer_blocking_duration_secs': 1.3427984714508057, 'get_old_steps_start_time': 1776967387.2907345, 'get_old_steps_duration_secs': 2.5987625122070312e-05, 'checkpoint_manager_blocking_start_time': 1776967385.944102, 'checkpoint_manager_blocking_duration_secs': 1.3487331867218018} I0423 18:03:07.293033 138354608629568 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0423 18:03:07.293105 138354608629568 max_utils.py:750] Memstats: After params initialized: I0423 18:03:07.293154 138354608629568 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_18(process=5,(2,4,0,0)) I0423 18:03:07.293188 138354608629568 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_19(process=5,(3,4,0,0)) I0423 18:03:07.293220 138354608629568 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_22(process=5,(2,5,0,0)) I0423 18:03:07.293246 138354608629568 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_23(process=5,(3,5,0,0)) I0423 18:03:07.559811 138354608629568 metric_logger.py:185] completed step: 0, seconds: 11.075, TFLOP/s/device: 1.227, Tokens/s/device: 184.914, total_weights: 65536, loss: 10.874 I0423 18:03:07.729501 138354608629568 metric_logger.py:185] completed step: 1, seconds: 1.970, TFLOP/s/device: 6.897, Tokens/s/device: 1039.628, total_weights: 65536, loss: 10.874 I0423 18:03:08.169687 138354608629568 metric_logger.py:185] completed step: 2, seconds: 0.024, TFLOP/s/device: 566.248, Tokens/s/device: 85351.115, total_weights: 65536, loss: 10.262 I0423 18:03:08.317066 138354608629568 metric_logger.py:185] completed step: 3, seconds: 0.431, TFLOP/s/device: 31.529, Tokens/s/device: 4752.424, total_weights: 65536, loss: 9.732 I0423 18:03:08.612823 138354608629568 metric_logger.py:185] completed step: 4, seconds: 0.163, TFLOP/s/device: 83.255, Tokens/s/device: 12549.173, total_weights: 65536, loss: 9.274 I0423 18:03:08.620161 138354608629568 metric_logger.py:185] completed step: 5, seconds: 0.146, TFLOP/s/device: 92.789, Tokens/s/device: 13986.205, total_weights: 65536, loss: 8.888 I0423 18:03:27.387397 138354608629568 metric_logger.py:185] completed step: 6, seconds: 0.297, TFLOP/s/device: 45.824, Tokens/s/device: 6907.158, total_weights: 65536, loss: 8.589 I0423 18:03:27.534767 138354608629568 metric_logger.py:185] completed step: 7, seconds: 18.621, TFLOP/s/device: 0.730, Tokens/s/device: 109.982, total_weights: 65536, loss: 8.380 I0423 18:03:27.682299 138354608629568 metric_logger.py:185] completed step: 8, seconds: 0.153, TFLOP/s/device: 88.815, Tokens/s/device: 13387.108, total_weights: 65536, loss: 8.251 I0423 18:03:27.830795 138354608629568 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0423 18:03:27.831389 138354608629568 checkpointing.py:798] Waited 0.0006055831909179688 seconds for step 9 to finish before starting checkpointing. I0423 18:03:27.833303 138354608629568 checkpoint_manager.py:2024] [process=5][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0423 18:03:28.949392 2614 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0423 18:03:31.314344 138232624826112 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_10_shardy_true/checkpoints/0/items/array_metadatas/process_5 I0423 18:04:03.910163 138232616433408 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/gbytes_per_sec: 27.419 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 57.61000895500183 s) (per-host) I0423 18:04:03.910290 138232616433408 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 56.619752s. I0423 18:04:11.247356 138232616433408 async_checkpointer.py:144] [process=5][thread=async_save] Background save thread done. Time taken: 63.956801s. I0423 18:04:11.247646 138229932078848 async_checkpointer.py:273] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save. I0423 18:04:11.247804 138229932078848 async_checkpointer.py:283] [process=5][thread=save_finalize] No errors found in background save thread=async_save. I0423 18:04:11.247857 138229932078848 checkpoint_manager.py:2133] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0423 18:04:11.249615 138229932078848 checkpoint_manager.py:2142] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0423 18:04:11.249812 138354608629568 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 18:04:11.249948 138354608629568 checkpoint_manager.py:1458] Waiting for previous save to complete took 43.416648 seconds. If this number is high, consider checkpointing less frequently. I0423 18:04:11.251844 138354608629568 checkpoint_manager.py:1518] [process=5] Saving checkpoint at step 9 I0423 18:04:11.255146 138354608629568 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_10_shardy_true/checkpoints/9. I0423 18:04:11.787197 138354608629568 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0423 18:04:11.787361 138354608629568 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 18:04:11.832445 138354608629568 base_pytree_checkpoint_handler.py:153] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.047222s I0423 18:04:11.832615 138354608629568 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.161 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2989068031311035 s) (per-host) I0423 18:04:11.832681 138354608629568 base_pytree_checkpoint_handler.py:737] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.298981s (batch_requests_ready=0.247042s, total_serialization_initiated=0.051857s, others=0.000082s) I0423 18:04:11.832776 138354608629568 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.302997s (all_items=0.000015s, per_item={'items': '0.00001502'}, temp_paths=0.302982) I0423 18:04:11.833698 138229411993344 async_checkpointer.py:79] [process=5][thread=async_save] Background save thread started. I0423 18:04:11.833862 138354608629568 async_checkpointer.py:561] Finished blocking save. Time taken: 0.581948s. 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_10_shardy_true/checkpoints/9. I0423 18:04:11.856706 138354608629568 checkpoint_manager.py:1566] [process=5][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0423 18:04:11.856971 138229932078848 async_checkpointer.py:265] [process=5][thread=save_finalize] Waiting for background save thread=async_save. I0423 18:04:11.857131 138354608629568 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_10_shardy_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776967407.8332746, 'wait_for_prev_duration_secs': 43.416648387908936, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776967451.2518811, 'checkpointer_blocking_duration_secs': 0.582089900970459, 'get_old_steps_start_time': 1776967451.8339965, 'get_old_steps_duration_secs': 2.574920654296875e-05, 'checkpoint_manager_blocking_start_time': 1776967407.8316095, 'checkpoint_manager_blocking_duration_secs': 44.02548670768738} I0423 18:04:11.857245 138354608629568 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0423 18:04:11.857289 138354608629568 checkpoint_manager.py:2024] [process=5][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0423 18:04:17.166563 138232624826112 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_10_shardy_true/checkpoints/9/items/array_metadatas/process_5 I0423 18:04:56.093543 138229411993344 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/gbytes_per_sec: 35.449 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 44.55979609489441 s) (per-host) I0423 18:04:56.093673 138229411993344 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 44.259861s. I0423 18:05:02.680533 138229411993344 async_checkpointer.py:144] [process=5][thread=async_save] Background save thread done. Time taken: 50.846705s. I0423 18:05:02.680838 138229932078848 async_checkpointer.py:273] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save. I0423 18:05:02.680956 138229932078848 async_checkpointer.py:283] [process=5][thread=save_finalize] No errors found in background save thread=async_save. I0423 18:05:02.681016 138229932078848 checkpoint_manager.py:2133] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0423 18:05:02.682634 138229932078848 checkpoint_manager.py:2142] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0423 18:05:02.682817 138354608629568 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 18:05:02.682980 138354608629568 checkpoint_manager.py:2013] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0423 18:05:02.683600 138354608629568 metric_logger.py:185] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.402, Tokens/s/device: 13927.804, 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 18:05:14 UTC 2026 EXIT_CODE=0