XPK Start: Sun Apr 19 04:58:35 UTC 2026 2026-04-19 04:58:39.995784: 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 00:00:1776574720.008906 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1776574720.012673 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1776574720.024057 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776574720.024076 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776574720.024079 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776574720.024081 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-19 04:58:59.105151: 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) I0419 04:58:59.623174 132218940716864 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-19 04:59:08,663:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0419 04:59:08.663762 132218940716864 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-19 04:59:08,666:jax._src.distributed:157: Connecting to JAX distributed service on mt-01-base-lq6h7-slice-job-0-0.mt-01-base-lq6h7:8482 I0419 04:59:08.666062 132218940716864 distributed.py:157] Connecting to JAX distributed service on mt-01-base-lq6h7-slice-job-0-0.mt-01-base-lq6h7:8482 I0419 04:59:10.235116 132218940716864 max_utils.py:284] Jax distributed system initialized! I0419 04:59:17.004281 132218940716864 max_utils.py:800] System Information: Jax Version: 0.8.1 I0419 04:59:17.004388 132218940716864 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0419 04:59:17.004432 132218940716864 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 I0419 04:59:17.004466 132218940716864 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0419 04:59:17.617793 132218940716864 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0419 04:59:17.731495 132218940716864 checkpointing.py:688] Setting up checkpoint logger... I0419 04:59:17.731610 132218940716864 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0419 04:59:17.731657 132218940716864 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0419 04:59:17.731870 132218940716864 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 0x78400e179460>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0419 04:59:21.530128 132218940716864 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0419 04:59:21.530469 132218940716864 checkpoint_manager.py:709] [process=0][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x782d4c129fd0>}, handler_registry=None I0419 04:59:21.530707 132218940716864 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x782d4c129fd0>` 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`. I0419 04:59:21.530758 132218940716864 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x782d4c128170>` 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`. I0419 04:59:21.530792 132218940716864 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 0x782d4c129fd0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x782d4c129fd0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x782d4c128170>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x782d4c128170>}). I0419 04:59:21.531137 132218940716864 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0419 04:59:21.531216 132218940716864 async_checkpointer.py:177] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x782cc2c78f40> timeout: 600 secs and primary_host=0 for async checkpoint writes I0419 04:59:23.403952 132218940716864 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints I0419 04:59:24.265519 132218940716864 checkpoint_manager.py:929] [process=0][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_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x782cc364bfb0> I0419 04:59:24.265693 132218940716864 checkpointing.py:302] Checkpoint manager created! I0419 04:59:25.182564 132218940716864 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0419 04:59:25.182690 132218940716864 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0419 04:59:26.629654 132218940716864 maxtext_utils.py:1749] decoder/decoder_norm/scale/value Shape: float32[2048] Physical: (None,) I0419 04:59:26.629778 132218940716864 maxtext_utils.py:1749] decoder/layers/mlp/wi_0/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0419 04:59:26.629846 132218940716864 maxtext_utils.py:1749] decoder/layers/mlp/wi_1/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0419 04:59:26.629894 132218940716864 maxtext_utils.py:1749] decoder/layers/mlp/wo/kernel/value Shape: float32[7168,16,2048] Physical: (None, None, 'fsdp') I0419 04:59:26.629934 132218940716864 maxtext_utils.py:1749] decoder/layers/post_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0419 04:59:26.629969 132218940716864 maxtext_utils.py:1749] decoder/layers/pre_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0419 04:59:26.630006 132218940716864 maxtext_utils.py:1749] decoder/layers/self_attention/key/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0419 04:59:26.630046 132218940716864 maxtext_utils.py:1749] decoder/layers/self_attention/out/kernel/value Shape: float32[16,16,128,2048] Physical: (None, None, None, 'fsdp') I0419 04:59:26.630085 132218940716864 maxtext_utils.py:1749] decoder/layers/self_attention/query/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0419 04:59:26.630121 132218940716864 maxtext_utils.py:1749] decoder/layers/self_attention/value/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0419 04:59:26.630157 132218940716864 maxtext_utils.py:1749] decoder/logits_dense/kernel/value Shape: float32[2048,32000] Physical: ('fsdp', None) I0419 04:59:26.630191 132218940716864 maxtext_utils.py:1749] token_embedder/embedding/value Shape: float32[32000,2048] Physical: (None, 'fsdp') I0419 04:59:27.026441 132218940716864 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0419 04:59:27.026596 132218940716864 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0419 04:59:27.031973 132218940716864 nnx_decoders.py:616] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0419 04:59:27.032066 132218940716864 nnx_decoders.py:616] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0419 04:59:27.184802 132218940716864 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0419 04:59:27.184946 132218940716864 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0419 04:59:27.334472 132218940716864 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0419 04:59:27.334607 132218940716864 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0419 04:59:27.492528 132218940716864 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 04:59:27.492667 132218940716864 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 04:59:27.638861 132218940716864 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 04:59:27.638995 132218940716864 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 04:59:27.782630 132218940716864 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 04:59:27.782763 132218940716864 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 04:59:27.934453 132218940716864 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0419 04:59:27.934588 132218940716864 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 04:59:27.957588 132218940716864 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0419 04:59:27.957688 132218940716864 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0419 04:59:38.838712 132218940716864 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. I0419 04:59:38.846410 132218940716864 metric_logger.py:289] number parameters: 1.104 billion I0419 04:59:50.422878 132218940716864 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0419 04:59:50.574939 132218940716864 checkpointing.py:798] Waited 0.15204501152038574 seconds for step 0 to finish before starting checkpointing. I0419 04:59:50.577256 132218940716864 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0419 04:59:50.579221 132218940716864 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 0 I0419 04:59:50.582339 132218940716864 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/0. I0419 04:59:51.307901 132218940716864 signaling_client.py:364] Using JaxDistributedSignalingClient I0419 04:59:51.586626 132090821592832 atomicity.py:137] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/0 I0419 04:59:51.604545 132218940716864 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0419 04:59:51.604642 132218940716864 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 I0419 04:59:51.844673 132218940716864 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.242192s I0419 04:59:51.845526 132218940716864 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.913 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5294609069824219 s) (per-host) I0419 04:59:51.845587 132218940716864 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.529537s (batch_requests_ready=0.278935s, total_serialization_initiated=0.249852s, others=0.000750s) I0419 04:59:51.845665 132218940716864 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.533665s (all_items=0.000020s, per_item={'items': '0.00001979'}, temp_paths=0.533645) I0419 04:59:51.846599 132091920512768 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0419 04:59:51.846756 132218940716864 async_checkpointer.py:561] Finished blocking save. Time taken: 1.267461s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/0. I0419 04:59:52.262120 132218940716864 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0419 04:59:52.262515 132090813200128 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0419 04:59:52.262683 132218940716864 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776574790.5772378, 'wait_for_prev_duration_secs': 6.365776062011719e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776574790.5792615, 'checkpointer_blocking_duration_secs': 1.26761794090271, 'get_old_steps_start_time': 1776574791.8468988, 'get_old_steps_duration_secs': 2.5510787963867188e-05, 'checkpoint_manager_blocking_start_time': 1776574790.5753434, 'checkpoint_manager_blocking_duration_secs': 1.6872992515563965} I0419 04:59:52.262793 132218940716864 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0419 04:59:52.262882 132218940716864 max_utils.py:750] Memstats: After params initialized: I0419 04:59:52.262939 132218940716864 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0)) I0419 04:59:52.262976 132218940716864 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0)) I0419 04:59:52.263002 132218940716864 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0)) I0419 04:59:52.263029 132218940716864 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0)) I0419 04:59:52.554795 132218940716864 metric_logger.py:185] completed step: 0, seconds: 11.222, TFLOP/s/device: 1.211, Tokens/s/device: 182.496, total_weights: 65536, loss: 10.874 I0419 04:59:52.557450 132218940716864 metric_logger.py:269] To see full metrics 'tensorboard --logdir=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/tensorboard/' I0419 04:59:53.389624 132218940716864 metric_logger.py:185] completed step: 1, seconds: 2.128, TFLOP/s/device: 6.384, Tokens/s/device: 962.218, total_weights: 65536, loss: 10.874 I0419 04:59:53.531691 132218940716864 metric_logger.py:185] completed step: 2, seconds: 0.836, TFLOP/s/device: 16.252, Tokens/s/device: 2449.699, total_weights: 65536, loss: 10.262 I0419 04:59:53.679251 132218940716864 metric_logger.py:185] completed step: 3, seconds: 0.011, TFLOP/s/device: 1252.039, Tokens/s/device: 188720.973, total_weights: 65536, loss: 9.732 I0419 05:00:14.446132 132218940716864 metric_logger.py:185] completed step: 4, seconds: 0.144, TFLOP/s/device: 94.337, Tokens/s/device: 14219.457, total_weights: 65536, loss: 9.274 I0419 05:00:14.456308 132218940716864 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.318, Tokens/s/device: 13915.123, total_weights: 65536, loss: 8.888 I0419 05:00:14.600658 132218940716864 metric_logger.py:185] completed step: 6, seconds: 20.766, TFLOP/s/device: 0.654, Tokens/s/device: 98.623, total_weights: 65536, loss: 8.589 I0419 05:00:14.748076 132218940716864 metric_logger.py:185] completed step: 7, seconds: 0.008, TFLOP/s/device: 1782.152, Tokens/s/device: 268625.393, total_weights: 65536, loss: 8.380 I0419 05:00:14.895566 132218940716864 metric_logger.py:185] completed step: 8, seconds: 0.146, TFLOP/s/device: 93.188, Tokens/s/device: 14046.350, total_weights: 65536, loss: 8.251 I0419 05:00:15.044198 132218940716864 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0419 05:00:15.044828 132218940716864 checkpointing.py:798] Waited 0.0006268024444580078 seconds for step 9 to finish before starting checkpointing. I0419 05:00:15.047181 132218940716864 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 05:00:15.267629 132093354944256 atomicity.py:137] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/0/items I0419 05:00:15.415398 132091906008832 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776574815006909052, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776574815006909052, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/0/_CHECKPOINT_METADATA I0419 05:00:16.373530 2606 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0419 05:00:17.995088 132095661823744 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/0/items/array_metadatas/process_0 I0419 05:00:52.984383 132095653431040 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 36.609863s (commit=35.652953s, array_metadata_write=0.956910s) I0419 05:00:52.985406 132091920512768 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 25.614 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 61.669339179992676 s) (per-host) I0419 05:00:52.985469 132091920512768 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 61.138764s. I0419 05:00:53.691317 132091920512768 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776574815006909052, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/0/_CHECKPOINT_METADATA I0419 05:00:55.220892 132091920512768 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000461s. I0419 05:00:55.639446 132091920512768 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0419 05:00:55.729855 132091906008832 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776574815006909052, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/0/_CHECKPOINT_METADATA I0419 05:00:56.564378 132091920512768 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.314822s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/0/items I0419 05:00:56.566256 132091920512768 atomicity.py:608] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/0/items I0419 05:00:56.960589 132091920512768 atomicity.py:608] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/0 I0419 05:00:59.079519 132091920512768 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/0`. I0419 05:00:59.080281 132091920512768 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 68.501006s. directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/0 I0419 05:00:59.081671 132091920512768 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 67.234963s. I0419 05:00:59.081884 132090813200128 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 05:00:59.082030 132090813200128 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0419 05:00:59.082100 132090813200128 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0419 05:00:59.085361 132090813200128 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0419 05:00:59.085539 132218940716864 checkpoint_manager.py:2036] [process=0][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0419 05:00:59.085687 132218940716864 checkpoint_manager.py:1458] Waiting for previous save to complete took 44.038515 seconds. If this number is high, consider checkpointing less frequently. I0419 05:00:59.087542 132218940716864 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 9 I0419 05:00:59.090928 132218940716864 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/9. I0419 05:01:00.030988 132091920512768 atomicity.py:137] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/9 I0419 05:01:00.042133 132218940716864 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0419 05:01:00.042237 132218940716864 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 I0419 05:01:00.070919 132218940716864 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.031111s I0419 05:01:00.071732 132218940716864 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.490 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2809560298919678 s) (per-host) I0419 05:01:00.071803 132218940716864 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.281045s (batch_requests_ready=0.243324s, total_serialization_initiated=0.036981s, others=0.000740s) I0419 05:01:00.071899 132218940716864 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.285048s (all_items=0.000017s, per_item={'items': '0.00001693'}, temp_paths=0.285031) I0419 05:01:00.072750 132090804807424 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0419 05:01:00.072881 132218940716864 async_checkpointer.py:561] Finished blocking save. Time taken: 0.985268s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/9. I0419 05:01:00.548369 132218940716864 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0419 05:01:00.548759 132090813200128 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0419 05:01:00.548958 132218940716864 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776574815.0471454, 'wait_for_prev_duration_secs': 44.038514852523804, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776574859.087581, 'checkpointer_blocking_duration_secs': 0.985464334487915, 'get_old_steps_start_time': 1776574860.0730646, 'get_old_steps_duration_secs': 2.8371810913085938e-05, 'checkpoint_manager_blocking_start_time': 1776574815.045055, 'checkpoint_manager_blocking_duration_secs': 45.50386595726013} I0419 05:01:00.549071 132218940716864 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0419 05:01:00.549116 132218940716864 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 05:01:02.303736 132091342485248 atomicity.py:137] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/9/items I0419 05:01:05.916193 132090838378240 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/9/items/array_metadatas/process_0 I0419 05:01:46.149899 132090829985536 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 42.739106s (commit=41.801007s, array_metadata_write=0.938098s) I0419 05:01:46.151013 132090804807424 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 34.072 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 46.360204458236694 s) (per-host) I0419 05:01:46.151119 132090804807424 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 46.078158s. I0419 05:01:47.773111 132090804807424 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0419 05:01:47.949631 132090804807424 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000419s. I0419 05:01:48.846506 132090804807424 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.909791s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/9/items I0419 05:01:48.849220 132090804807424 atomicity.py:608] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/9/items I0419 05:01:49.254201 132090804807424 atomicity.py:608] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/9 I0419 05:01:50.809339 132090804807424 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/9`. I0419 05:01:50.810186 132090804807424 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 51.722578s. directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/nnx_xpk_feat_nnx_set_defaults_true_20260419_025828_01_base/checkpoints/9 I0419 05:01:50.811685 132090804807424 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 50.738723s. I0419 05:01:50.811924 132090813200128 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 05:01:50.812065 132090813200128 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0419 05:01:50.812155 132090813200128 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0419 05:01:50.813878 132090813200128 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0419 05:01:50.814054 132218940716864 checkpoint_manager.py:2036] [process=0][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0419 05:01:50.814227 132218940716864 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0419 05:01:50.814866 132218940716864 metric_logger.py:185] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.289, Tokens/s/device: 13910.775, 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: Sun Apr 19 05:02:01 UTC 2026 EXIT_CODE=0