XPK Start: Sat Apr 25 17:48:52 UTC 2026 2026-04-25 17:48:56.540997: 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 00:00:1777139336.554210 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1777139336.557939 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1777139336.569452 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1777139336.569471 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1777139336.569473 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1777139336.569475 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-25 17:49:15.784305: 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) I0425 17:49:16.316017 135755091928896 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-25 17:49:25,355:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0425 17:49:25.355261 135755091928896 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-25 17:49:25,357:jax._src.distributed:157: Connecting to JAX distributed service on mt-16-zarr3-true-rs-mxuru-slice-job-0-0.mt-16-zarr3-true-rs-mxuru:8482 I0425 17:49:25.357578 135755091928896 distributed.py:157] Connecting to JAX distributed service on mt-16-zarr3-true-rs-mxuru-slice-job-0-0.mt-16-zarr3-true-rs-mxuru:8482 I0425 17:49:26.183158 135755091928896 max_utils.py:284] Jax distributed system initialized! I0425 17:49:33.090716 135755091928896 max_utils.py:800] System Information: Jax Version: 0.8.1 I0425 17:49:33.090822 135755091928896 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0425 17:49:33.090864 135755091928896 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 I0425 17:49:33.090900 135755091928896 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0425 17:49:33.716403 135755091928896 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0425 17:49:33.834817 135755091928896 checkpointing.py:688] Setting up checkpoint logger... I0425 17:49:33.834954 135755091928896 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0425 17:49:33.835004 135755091928896 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0425 17:49:33.835216 135755091928896 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 0x7b776bb6d4f0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0425 17:49:35.534111 135755091928896 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0425 17:49:35.534462 135755091928896 checkpoint_manager.py:709] [process=3][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b62c0737650>}, handler_registry=None I0425 17:49:35.534695 135755091928896 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b62c0737650>` 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`. I0425 17:49:35.534744 135755091928896 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b63e87d3e90>` 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`. I0425 17:49:35.534782 135755091928896 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 0x7b62c0737650>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b62c0737650>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b63e87d3e90>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b63e87d3e90>}). I0425 17:49:35.535136 135755091928896 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0425 17:49:35.535213 135755091928896 async_checkpointer.py:177] [process=3][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7b63e811cf40> timeout: 600 secs and primary_host=0 for async checkpoint writes I0425 17:49:37.031000 135634010216192 checkpoint.py:228] Read Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1777139212305150741, 'commit_timestamp_nsecs': 1777139258989633717, 'custom_metadata': {}} from gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints/9/_CHECKPOINT_METADATA I0425 17:49:37.289872 135634010216192 checkpoint_manager.py:1877] Missing metrics for step 9 I0425 17:49:37.380528 135755091928896 checkpoint_manager.py:1818] Found 3 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints I0425 17:49:38.552829 135755091928896 checkpoint_manager.py:929] [process=3][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=5), 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_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7b63e81ada30> I0425 17:49:38.553038 135755091928896 checkpointing.py:302] Checkpoint manager created! I0425 17:49:39.482895 135755091928896 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0425 17:49:39.483029 135755091928896 checkpointing.py:582] restoring from this run's directory step 9 I0425 17:49:40.415292 135755091928896 checkpointer.py:304] Restoring checkpoint from gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints/9. I0425 17:49:41.857710 1432 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0425 17:49:48.487639 135755091928896 jax_array_handlers.py:837] [process=3] /jax/orbax/read/worker/io/requested throughput: 238.156 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 6.635480642318726 s) (per-host) I0425 17:49:48.495789 135755091928896 base_pytree_checkpoint_handler.py:129] [process=3] /jax/checkpoint/read/gbytes_per_sec: 224.239 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 7.057658910751343 s) (per-host) I0425 17:49:49.296191 135755091928896 checkpointer.py:318] Finished restoring checkpoint in 9.13 seconds from gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints/9. I0425 17:49:49.296352 135755091928896 standard_logger.py:34] {'step': 9, 'event_type': 'restore', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints', 'checkpointer_start_time': 1777139380.1659126, 'checkpointer_duration_secs': 9.130409717559814, 'checkpoint_manager_start_time': 1777139379.4836655, 'checkpoint_manager_duration_secs': 9.812658786773682} fsdp: 32 I0425 17:49:49.379920 135755091928896 maxtext_utils.py:1749] decoder/decoder_norm/scale/value Shape: float32[2048] Physical: (None,) I0425 17:49:49.380067 135755091928896 maxtext_utils.py:1749] decoder/layers/mlp/wi_0/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0425 17:49:49.380121 135755091928896 maxtext_utils.py:1749] decoder/layers/mlp/wi_1/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0425 17:49:49.380165 135755091928896 maxtext_utils.py:1749] decoder/layers/mlp/wo/kernel/value Shape: float32[7168,16,2048] Physical: (None, None, 'fsdp') I0425 17:49:49.380205 135755091928896 maxtext_utils.py:1749] decoder/layers/post_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0425 17:49:49.380241 135755091928896 maxtext_utils.py:1749] decoder/layers/pre_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0425 17:49:49.380280 135755091928896 maxtext_utils.py:1749] decoder/layers/self_attention/key/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0425 17:49:49.380317 135755091928896 maxtext_utils.py:1749] decoder/layers/self_attention/out/kernel/value Shape: float32[16,16,128,2048] Physical: (None, None, None, 'fsdp') I0425 17:49:49.380352 135755091928896 maxtext_utils.py:1749] decoder/layers/self_attention/query/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0425 17:49:49.380388 135755091928896 maxtext_utils.py:1749] decoder/layers/self_attention/value/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0425 17:49:49.380422 135755091928896 maxtext_utils.py:1749] decoder/logits_dense/kernel/value Shape: float32[2048,32000] Physical: ('fsdp', None) I0425 17:49:49.380455 135755091928896 maxtext_utils.py:1749] token_embedder/embedding/value Shape: float32[32000,2048] Physical: (None, 'fsdp') I0425 17:49:49.784081 135755091928896 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0425 17:49:49.784228 135755091928896 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0425 17:49:49.789682 135755091928896 nnx_decoders.py:616] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0425 17:49:49.789781 135755091928896 nnx_decoders.py:616] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0425 17:49:49.943893 135755091928896 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0425 17:49:49.944085 135755091928896 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0425 17:49:50.097389 135755091928896 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0425 17:49:50.097537 135755091928896 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0425 17:49:50.261781 135755091928896 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0425 17:49:50.262144 135755091928896 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0425 17:49:50.412630 135755091928896 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0425 17:49:50.412763 135755091928896 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0425 17:49:50.559759 135755091928896 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0425 17:49:50.559897 135755091928896 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0425 17:49:50.721355 135755091928896 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0425 17:49:50.721488 135755091928896 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0425 17:49:50.744143 135755091928896 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0425 17:49:50.744235 135755091928896 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0425 17:50:01.969427 135755091928896 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. I0425 17:50:01.971850 135755091928896 metric_logger.py:289] number parameters: 1.104 billion I0425 17:50:13.338548 135755091928896 checkpointing.py:794] Waiting for step 10 to finish before checkpoint... I0425 17:50:13.490163 135755091928896 checkpointing.py:798] Waited 0.15160250663757324 seconds for step 10 to finish before starting checkpointing. I0425 17:50:13.492330 135755091928896 checkpoint_manager.py:2013] [process=3][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0425 17:50:13.494552 135755091928896 checkpoint_manager.py:1518] [process=3] Saving checkpoint at step 10 I0425 17:50:13.497565 135755091928896 async_checkpointer.py:452] [process=3] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints/10. I0425 17:50:14.079565 135755091928896 signaling_client.py:364] Using JaxDistributedSignalingClient I0425 17:50:14.080595 135755091928896 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0425 17:50:14.080655 135755091928896 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 I0425 17:50:14.332081 135755091928896 base_pytree_checkpoint_handler.py:153] [process=3][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.253693s I0425 17:50:14.332257 135755091928896 base_pytree_checkpoint_handler.py:129] [process=3] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.945 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5237588882446289 s) (per-host) I0425 17:50:14.332310 135755091928896 base_pytree_checkpoint_handler.py:737] [process=3][thread=MainThread] Initiated Pytree async_save. Time taken: 0.523822s (batch_requests_ready=0.265226s, total_serialization_initiated=0.258523s, others=0.000073s) I0425 17:50:14.332405 135755091928896 composite_checkpoint_handler.py:715] [process=3][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.527889s (all_items=0.000027s, per_item={'items': '0.00002694'}, temp_paths=0.527863) I0425 17:50:14.333338 135633423021824 async_checkpointer.py:79] [process=3][thread=async_save] Background save thread started. I0425 17:50:14.333518 135755091928896 async_checkpointer.py:561] Finished blocking save. Time taken: 0.838851s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints/10. I0425 17:50:14.348504 135755091928896 checkpoint_manager.py:1566] [process=3][thread=MainThread][step=10] Starting CheckpointManager Save Finalize thread=save_finalize I0425 17:50:14.348783 135629153232640 async_checkpointer.py:265] [process=3][thread=save_finalize] Waiting for background save thread=async_save. I0425 17:50:14.348953 135755091928896 standard_logger.py:34] {'step': 10, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777139413.4923115, 'wait_for_prev_duration_secs': 6.580352783203125e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777139413.4946337, 'checkpointer_blocking_duration_secs': 0.8389925956726074, 'get_old_steps_start_time': 1777139414.3336449, 'get_old_steps_duration_secs': 2.7894973754882812e-05, 'checkpoint_manager_blocking_start_time': 1777139413.490538, 'checkpoint_manager_blocking_duration_secs': 0.8583650588989258} I0425 17:50:14.349064 135755091928896 checkpointing.py:409] Started an asynchronous checkpoint save for step 10 I0425 17:50:14.349138 135755091928896 max_utils.py:750] Memstats: After params initialized: I0425 17:50:14.349190 135755091928896 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_10(process=3,(2,2,0,0)) I0425 17:50:14.349222 135755091928896 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_11(process=3,(3,2,0,0)) I0425 17:50:14.349250 135755091928896 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_14(process=3,(2,3,0,0)) I0425 17:50:14.349275 135755091928896 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_15(process=3,(3,3,0,0)) I0425 17:50:14.622061 135755091928896 metric_logger.py:185] completed step: 10, seconds: 11.366, TFLOP/s/device: 1.195, Tokens/s/device: 180.193, total_weights: 65536, loss: 6.827 I0425 17:50:14.797382 135755091928896 metric_logger.py:185] completed step: 11, seconds: 1.282, TFLOP/s/device: 10.600, Tokens/s/device: 1597.822, total_weights: 65536, loss: 6.522 I0425 17:50:14.944822 135755091928896 metric_logger.py:185] completed step: 12, seconds: 0.020, TFLOP/s/device: 681.366, Tokens/s/device: 102702.974, total_weights: 65536, loss: 6.260 I0425 17:50:15.092536 135755091928896 metric_logger.py:185] completed step: 13, seconds: 0.164, TFLOP/s/device: 83.070, Tokens/s/device: 12521.246, total_weights: 65536, loss: 6.039 I0425 17:50:15.100113 135755091928896 checkpointing.py:794] Waiting for step 15 to finish before checkpoint... I0425 17:50:15.387181 135755091928896 checkpointing.py:798] Waited 0.28704047203063965 seconds for step 15 to finish before starting checkpointing. I0425 17:50:15.389861 135755091928896 checkpoint_manager.py:2024] [process=3][thread=MainThread][step=10][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0425 17:50:18.930714 135629648140032 array_metadata_store.py:203] [process=3][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints/10/items/array_metadatas/process_3 I0425 17:50:55.763422 135633423021824 base_pytree_checkpoint_handler.py:129] [process=3] /jax/checkpoint/write/gbytes_per_sec: 37.650 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.954882860183716 s) (per-host) I0425 17:50:55.763555 135633423021824 async_checkpointer.py:90] [process=3][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.430094s. I0425 17:51:03.400611 135633423021824 async_checkpointer.py:144] [process=3][thread=async_save] Background save thread done. Time taken: 49.067133s. I0425 17:51:03.400911 135629153232640 async_checkpointer.py:273] [process=3][thread=save_finalize] Done with waiting for background save thread=async_save. I0425 17:51:03.401050 135629153232640 async_checkpointer.py:283] [process=3][thread=save_finalize] No errors found in background save thread=async_save. I0425 17:51:03.401099 135629153232640 checkpoint_manager.py:2133] [process=3][thread=save_finalize][step=10] CheckpointManager Save Finalize is syncing with other hosts... I0425 17:51:03.403511 135629153232640 checkpoint_manager.py:2142] [process=3][thread=save_finalize][step=10] CheckpointManager Save Finalize is done on all hosts. I0425 17:51:03.403697 135755091928896 checkpoint_manager.py:2036] [process=3][thread=MainThread][step=10][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=10. W0425 17:51:03.403828 135755091928896 checkpoint_manager.py:1458] Waiting for previous save to complete took 48.013972 seconds. If this number is high, consider checkpointing less frequently. I0425 17:51:03.406036 135755091928896 checkpoint_manager.py:1518] [process=3] Saving checkpoint at step 15 I0425 17:51:03.410234 135755091928896 async_checkpointer.py:452] [process=3] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints/15. I0425 17:51:03.989469 135755091928896 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0425 17:51:03.989571 135755091928896 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 I0425 17:51:04.035797 135755091928896 base_pytree_checkpoint_handler.py:153] [process=3][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.048395s I0425 17:51:04.035990 135755091928896 base_pytree_checkpoint_handler.py:129] [process=3] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.005 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.30820751190185547 s) (per-host) I0425 17:51:04.036048 135755091928896 base_pytree_checkpoint_handler.py:737] [process=3][thread=MainThread] Initiated Pytree async_save. Time taken: 0.308276s (batch_requests_ready=0.255053s, total_serialization_initiated=0.053145s, others=0.000077s) I0425 17:51:04.036150 135755091928896 composite_checkpoint_handler.py:715] [process=3][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.313437s (all_items=0.000014s, per_item={'items': '0.00001407'}, temp_paths=0.313423) I0425 17:51:04.037044 135629648140032 async_checkpointer.py:79] [process=3][thread=async_save] Background save thread started. I0425 17:51:04.037201 135755091928896 async_checkpointer.py:561] Finished blocking save. Time taken: 0.631096s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints/15. I0425 17:51:04.053946 135755091928896 checkpoint_manager.py:1566] [process=3][thread=MainThread][step=15] Starting CheckpointManager Save Finalize thread=save_finalize I0425 17:51:04.054184 135629153232640 async_checkpointer.py:265] [process=3][thread=save_finalize] Waiting for background save thread=async_save. I0425 17:51:04.054316 135755091928896 standard_logger.py:34] {'step': 15, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777139415.389831, 'wait_for_prev_duration_secs': 48.01397228240967, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777139463.406074, 'checkpointer_blocking_duration_secs': 0.6312377452850342, 'get_old_steps_start_time': 1777139464.0373309, 'get_old_steps_duration_secs': 2.6702880859375e-05, 'checkpoint_manager_blocking_start_time': 1777139415.3875406, 'checkpoint_manager_blocking_duration_secs': 48.66674518585205} I0425 17:51:04.054424 135755091928896 checkpointing.py:409] Started an asynchronous checkpoint save for step 15 I0425 17:51:04.055183 135755091928896 metric_logger.py:185] completed step: 14, seconds: 0.147, TFLOP/s/device: 92.271, Tokens/s/device: 13908.130, total_weights: 65536, loss: 5.858 I0425 17:51:04.062715 135755091928896 metric_logger.py:185] completed step: 15, seconds: 0.149, TFLOP/s/device: 91.335, Tokens/s/device: 13766.957, total_weights: 65536, loss: 5.713 I0425 17:51:22.738851 135755091928896 metric_logger.py:185] completed step: 16, seconds: 48.961, TFLOP/s/device: 0.278, Tokens/s/device: 41.829, total_weights: 65536, loss: 5.599 I0425 17:51:22.888035 135755091928896 metric_logger.py:185] completed step: 17, seconds: 18.531, TFLOP/s/device: 0.733, Tokens/s/device: 110.517, total_weights: 65536, loss: 5.511 I0425 17:51:23.035336 135755091928896 metric_logger.py:185] completed step: 18, seconds: 0.153, TFLOP/s/device: 88.822, Tokens/s/device: 13388.246, total_weights: 65536, loss: 5.442 I0425 17:51:23.183676 135755091928896 checkpointing.py:794] Waiting for step 19 to finish before checkpoint... I0425 17:51:23.184275 135755091928896 checkpointing.py:798] Waited 0.0006129741668701172 seconds for step 19 to finish before starting checkpointing. I0425 17:51:23.186396 135755091928896 checkpoint_manager.py:2024] [process=3][thread=MainThread][step=15][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0425 17:51:27.373454 135630177425152 array_metadata_store.py:203] [process=3][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints/15/items/array_metadatas/process_3 I0425 17:52:04.735201 135629648140032 base_pytree_checkpoint_handler.py:129] [process=3] /jax/checkpoint/write/gbytes_per_sec: 25.892 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 61.00738072395325 s) (per-host) I0425 17:52:04.735323 135629648140032 async_checkpointer.py:90] [process=3][thread=async_save] 3 Handler Commit operations completed. Time taken: 60.698168s. I0425 17:52:12.717011 135629648140032 async_checkpointer.py:144] [process=3][thread=async_save] Background save thread done. Time taken: 68.679842s. I0425 17:52:12.717289 135629153232640 async_checkpointer.py:273] [process=3][thread=save_finalize] Done with waiting for background save thread=async_save. I0425 17:52:12.717442 135629153232640 async_checkpointer.py:283] [process=3][thread=save_finalize] No errors found in background save thread=async_save. I0425 17:52:12.717495 135629153232640 checkpoint_manager.py:2133] [process=3][thread=save_finalize][step=15] CheckpointManager Save Finalize is syncing with other hosts... I0425 17:52:12.718983 135629153232640 checkpoint_manager.py:2142] [process=3][thread=save_finalize][step=15] CheckpointManager Save Finalize is done on all hosts. I0425 17:52:12.719174 135755091928896 checkpoint_manager.py:2036] [process=3][thread=MainThread][step=15][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=15. W0425 17:52:12.719308 135755091928896 checkpoint_manager.py:1458] Waiting for previous save to complete took 49.532922 seconds. If this number is high, consider checkpointing less frequently. I0425 17:52:12.720904 135755091928896 checkpoint_manager.py:1518] [process=3] Saving checkpoint at step 19 I0425 17:52:12.724318 135755091928896 async_checkpointer.py:452] [process=3] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints/19. I0425 17:52:13.560602 135755091928896 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0425 17:52:13.560700 135755091928896 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 I0425 17:52:13.607892 135755091928896 base_pytree_checkpoint_handler.py:153] [process=3][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.049340s I0425 17:52:13.608068 135755091928896 base_pytree_checkpoint_handler.py:129] [process=3] /jax/checkpoint/write/blocking_gbytes_per_sec: 4.930 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.31292080879211426 s) (per-host) I0425 17:52:13.608117 135755091928896 base_pytree_checkpoint_handler.py:737] [process=3][thread=MainThread] Initiated Pytree async_save. Time taken: 0.312979s (batch_requests_ready=0.258824s, total_serialization_initiated=0.054092s, others=0.000063s) I0425 17:52:13.608208 135755091928896 composite_checkpoint_handler.py:715] [process=3][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.317143s (all_items=0.000012s, per_item={'items': '0.00001192'}, temp_paths=0.317131) I0425 17:52:13.609046 135632382850816 async_checkpointer.py:79] [process=3][thread=async_save] Background save thread started. I0425 17:52:13.609201 135755091928896 async_checkpointer.py:561] Finished blocking save. Time taken: 0.888215s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints/19. I0425 17:52:13.617877 135755091928896 checkpoint_manager.py:1566] [process=3][thread=MainThread][step=19] Starting CheckpointManager Save Finalize thread=save_finalize I0425 17:52:13.618188 135629153232640 async_checkpointer.py:265] [process=3][thread=save_finalize] Waiting for background save thread=async_save. I0425 17:52:13.618346 135755091928896 standard_logger.py:34] {'step': 19, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777139483.186368, 'wait_for_prev_duration_secs': 49.53292179107666, 'time_between_consecutive_saves_sec': 19.782808542251587, 'checkpointer_blocking_start_time': 1777139532.7209594, 'checkpointer_blocking_duration_secs': 0.8883490562438965, 'get_old_steps_start_time': 1777139533.609322, 'get_old_steps_duration_secs': 2.193450927734375e-05, 'checkpoint_manager_blocking_start_time': 1777139483.1844988, 'checkpoint_manager_blocking_duration_secs': 50.43381452560425} I0425 17:52:13.618456 135755091928896 checkpointing.py:409] Started an asynchronous checkpoint save for step 19 I0425 17:52:13.618499 135755091928896 checkpoint_manager.py:2024] [process=3][thread=MainThread][step=19][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0425 17:52:18.544675 135633423021824 array_metadata_store.py:203] [process=3][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_16_zarr3_true/checkpoints/19/items/array_metadatas/process_3 I0425 17:52:55.227560 135632382850816 base_pytree_checkpoint_handler.py:129] [process=3] /jax/checkpoint/write/gbytes_per_sec: 37.670 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.93235516548157 s) (per-host) I0425 17:52:55.227680 135632382850816 async_checkpointer.py:90] [process=3][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.618522s. I0425 17:53:01.555027 135632382850816 async_checkpointer.py:144] [process=3][thread=async_save] Background save thread done. Time taken: 47.945856s. I0425 17:53:01.555306 135629153232640 async_checkpointer.py:273] [process=3][thread=save_finalize] Done with waiting for background save thread=async_save. I0425 17:53:01.555429 135629153232640 async_checkpointer.py:283] [process=3][thread=save_finalize] No errors found in background save thread=async_save. I0425 17:53:01.555472 135629153232640 checkpoint_manager.py:2133] [process=3][thread=save_finalize][step=19] CheckpointManager Save Finalize is syncing with other hosts... I0425 17:53:01.558134 135629153232640 checkpoint_manager.py:2142] [process=3][thread=save_finalize][step=19] CheckpointManager Save Finalize is done on all hosts. I0425 17:53:01.558302 135755091928896 checkpoint_manager.py:2036] [process=3][thread=MainThread][step=19][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=19. I0425 17:53:01.558476 135755091928896 checkpoint_manager.py:2013] [process=3][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0425 17:53:01.560090 135755091928896 metric_logger.py:185] completed step: 19, seconds: 0.149, TFLOP/s/device: 91.323, Tokens/s/device: 13765.199, total_weights: 65536, loss: 5.385 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Sat Apr 25 17:53:13 UTC 2026 EXIT_CODE=0