XPK Start: Mon Apr 20 22:21:23 UTC 2026 2026-04-20 22:21:27.888761: 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 00:00:1776723687.901716 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1776723687.905375 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1776723687.916779 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776723687.916799 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776723687.916801 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776723687.916803 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-20 22:21:46.904361: 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) I0420 22:21:47.433940 139917764216640 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-20 22:21:56,473:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0420 22:21:56.473941 139917764216640 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-20 22:21:56,476:jax._src.distributed:157: Connecting to JAX distributed service on mt-16-zarr3-false-sv-iv9gw-slice-job-0-0.mt-16-zarr3-false-sv-iv9gw:8482 I0420 22:21:56.476345 139917764216640 distributed.py:157] Connecting to JAX distributed service on mt-16-zarr3-false-sv-iv9gw-slice-job-0-0.mt-16-zarr3-false-sv-iv9gw:8482 I0420 22:21:57.966090 139917764216640 max_utils.py:284] Jax distributed system initialized! I0420 22:22:05.121686 139917764216640 max_utils.py:800] System Information: Jax Version: 0.8.1 I0420 22:22:05.121792 139917764216640 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0420 22:22:05.121832 139917764216640 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 I0420 22:22:05.121870 139917764216640 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0420 22:22:05.744309 139917764216640 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0420 22:22:05.859011 139917764216640 checkpointing.py:688] Setting up checkpoint logger... I0420 22:22:05.859129 139917764216640 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=False I0420 22:22:05.859179 139917764216640 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0420 22:22:05.859393 139917764216640 base_pytree_checkpoint_handler.py:415] Created BasePyTreeCheckpointHandler: use_ocdbt=True, use_zarr3=False, pytree_metadata_options=PyTreeMetadataOptions(support_rich_types=False), array_metadata_store=<orbax.checkpoint._src.metadata.array_metadata_store.Store object at 0x7f40a2d752b0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0420 22:22:08.776420 139917764216640 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0420 22:22:08.776659 139917764216640 checkpoint_manager.py:709] [process=1][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7f2c382c93a0>}, handler_registry=None I0420 22:22:08.776894 139917764216640 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7f2c382c93a0>` 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`. I0420 22:22:08.776942 139917764216640 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7f2c380a5e20>` 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`. I0420 22:22:08.776979 139917764216640 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 0x7f2c382c93a0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7f2c382c93a0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7f2c380a5e20>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7f2c380a5e20>}). I0420 22:22:08.777334 139917764216640 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0420 22:22:08.777428 139917764216640 async_checkpointer.py:177] [process=1][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7f2c38030f40> timeout: 600 secs and primary_host=0 for async checkpoint writes I0420 22:22:10.317538 139917764216640 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_false/checkpoints I0420 22:22:10.798530 139917764216640 checkpoint_manager.py:929] [process=1][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_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7f2cabd78bf0> I0420 22:22:10.798710 139917764216640 checkpointing.py:302] Checkpoint manager created! I0420 22:22:11.719094 139917764216640 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0420 22:22:11.719219 139917764216640 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0420 22:22:13.116254 139917764216640 maxtext_utils.py:1749] decoder/decoder_norm/scale/value Shape: float32[2048] Physical: (None,) I0420 22:22:13.116412 139917764216640 maxtext_utils.py:1749] decoder/layers/mlp/wi_0/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0420 22:22:13.116486 139917764216640 maxtext_utils.py:1749] decoder/layers/mlp/wi_1/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0420 22:22:13.116552 139917764216640 maxtext_utils.py:1749] decoder/layers/mlp/wo/kernel/value Shape: float32[7168,16,2048] Physical: (None, None, 'fsdp') I0420 22:22:13.116609 139917764216640 maxtext_utils.py:1749] decoder/layers/post_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0420 22:22:13.116664 139917764216640 maxtext_utils.py:1749] decoder/layers/pre_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0420 22:22:13.116717 139917764216640 maxtext_utils.py:1749] decoder/layers/self_attention/key/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0420 22:22:13.116779 139917764216640 maxtext_utils.py:1749] decoder/layers/self_attention/out/kernel/value Shape: float32[16,16,128,2048] Physical: (None, None, None, 'fsdp') I0420 22:22:13.116829 139917764216640 maxtext_utils.py:1749] decoder/layers/self_attention/query/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0420 22:22:13.116886 139917764216640 maxtext_utils.py:1749] decoder/layers/self_attention/value/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0420 22:22:13.116937 139917764216640 maxtext_utils.py:1749] decoder/logits_dense/kernel/value Shape: float32[2048,32000] Physical: ('fsdp', None) I0420 22:22:13.116995 139917764216640 maxtext_utils.py:1749] token_embedder/embedding/value Shape: float32[32000,2048] Physical: (None, 'fsdp') I0420 22:22:13.508123 139917764216640 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0420 22:22:13.508280 139917764216640 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 22:22:13.513729 139917764216640 nnx_decoders.py:616] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0420 22:22:13.513825 139917764216640 nnx_decoders.py:616] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 22:22:13.673489 139917764216640 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0420 22:22:13.673618 139917764216640 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 22:22:13.824106 139917764216640 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0420 22:22:13.824239 139917764216640 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 22:22:13.985487 139917764216640 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0420 22:22:13.985619 139917764216640 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 22:22:14.134870 139917764216640 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0420 22:22:14.135002 139917764216640 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 22:22:14.284226 139917764216640 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0420 22:22:14.284358 139917764216640 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 22:22:14.445753 139917764216640 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0420 22:22:14.445905 139917764216640 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 22:22:14.470113 139917764216640 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0420 22:22:14.470209 139917764216640 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0420 22:22:25.436548 139917764216640 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. I0420 22:22:25.439067 139917764216640 metric_logger.py:289] number parameters: 1.104 billion I0420 22:22:36.781550 139917764216640 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0420 22:22:36.932471 139917764216640 checkpointing.py:798] Waited 0.1509110927581787 seconds for step 0 to finish before starting checkpointing. I0420 22:22:36.934971 139917764216640 checkpoint_manager.py:2013] [process=1][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0420 22:22:36.936611 139917764216640 checkpoint_manager.py:1518] [process=1] Saving checkpoint at step 0 I0420 22:22:36.939529 139917764216640 async_checkpointer.py:452] [process=1] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_false/checkpoints/0. I0420 22:22:38.340618 139917764216640 signaling_client.py:364] Using JaxDistributedSignalingClient I0420 22:22:38.341719 139917764216640 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0420 22:22:38.341776 139917764216640 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 I0420 22:22:38.589479 139917764216640 base_pytree_checkpoint_handler.py:153] [process=1][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.249984s I0420 22:22:38.589649 139917764216640 base_pytree_checkpoint_handler.py:129] [process=1] /jax/checkpoint/write/blocking_gbytes_per_sec: 1.644 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.938255786895752 s) (per-host) I0420 22:22:38.589699 139917764216640 base_pytree_checkpoint_handler.py:737] [process=1][thread=MainThread] Initiated Pytree async_save. Time taken: 0.938315s (batch_requests_ready=0.681753s, total_serialization_initiated=0.256495s, others=0.000067s) I0420 22:22:38.589794 139917764216640 composite_checkpoint_handler.py:715] [process=1][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.942295s (all_items=0.000019s, per_item={'items': '0.00001884'}, temp_paths=0.942276) I0420 22:22:38.590796 139794675922688 async_checkpointer.py:79] [process=1][thread=async_save] Background save thread started. I0420 22:22:38.590976 139917764216640 async_checkpointer.py:561] Finished blocking save. Time taken: 1.654294s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_false/checkpoints/0. I0420 22:22:38.631427 139917764216640 checkpoint_manager.py:1566] [process=1][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0420 22:22:38.631716 139790875862784 async_checkpointer.py:265] [process=1][thread=save_finalize] Waiting for background save thread=async_save. I0420 22:22:38.631863 139917764216640 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776723756.9349532, 'wait_for_prev_duration_secs': 6.246566772460938e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776723756.936651, 'checkpointer_blocking_duration_secs': 1.6544320583343506, 'get_old_steps_start_time': 1776723758.5911024, 'get_old_steps_duration_secs': 2.956390380859375e-05, 'checkpoint_manager_blocking_start_time': 1776723756.9328601, 'checkpoint_manager_blocking_duration_secs': 1.6989636421203613} I0420 22:22:38.631971 139917764216640 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0420 22:22:38.632042 139917764216640 max_utils.py:750] Memstats: After params initialized: I0420 22:22:38.632095 139917764216640 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_2(process=1,(2,0,0,0)) I0420 22:22:38.632134 139917764216640 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_3(process=1,(3,0,0,0)) I0420 22:22:38.632160 139917764216640 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_6(process=1,(2,1,0,0)) I0420 22:22:38.632183 139917764216640 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_7(process=1,(3,1,0,0)) I0420 22:22:38.905068 139917764216640 metric_logger.py:185] completed step: 0, seconds: 11.341, TFLOP/s/device: 1.198, Tokens/s/device: 180.577, total_weights: 65536, loss: 10.874 I0420 22:22:39.074360 139917764216640 metric_logger.py:185] completed step: 1, seconds: 2.122, TFLOP/s/device: 6.404, Tokens/s/device: 965.258, total_weights: 65536, loss: 10.874 I0420 22:22:39.504032 139917764216640 metric_logger.py:185] completed step: 2, seconds: 0.023, TFLOP/s/device: 580.052, Tokens/s/device: 87431.694, total_weights: 65536, loss: 10.560 I0420 22:22:39.651513 139917764216640 metric_logger.py:185] completed step: 3, seconds: 0.421, TFLOP/s/device: 32.274, Tokens/s/device: 4864.677, total_weights: 65536, loss: 9.979 I0420 22:22:39.657311 139917764216640 checkpointing.py:794] Waiting for step 5 to finish before checkpoint... I0420 22:22:39.946521 139917764216640 checkpointing.py:798] Waited 0.2891812324523926 seconds for step 5 to finish before starting checkpointing. I0420 22:22:39.949104 139917764216640 checkpoint_manager.py:2024] [process=1][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0420 22:22:41.679216 2605 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0420 22:22:44.069458 139795162437376 array_metadata_store.py:203] [process=1][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_false/checkpoints/0/items/array_metadatas/process_1 I0420 22:23:19.342252 139794675922688 base_pytree_checkpoint_handler.py:129] [process=1] /jax/checkpoint/write/gbytes_per_sec: 37.888 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.690815925598145 s) (per-host) I0420 22:23:19.342373 139794675922688 async_checkpointer.py:90] [process=1][thread=async_save] 3 Handler Commit operations completed. Time taken: 40.751464s. I0420 22:23:27.164493 139794675922688 async_checkpointer.py:144] [process=1][thread=async_save] Background save thread done. Time taken: 48.573570s. I0420 22:23:27.164809 139790875862784 async_checkpointer.py:273] [process=1][thread=save_finalize] Done with waiting for background save thread=async_save. I0420 22:23:27.164940 139790875862784 async_checkpointer.py:283] [process=1][thread=save_finalize] No errors found in background save thread=async_save. I0420 22:23:27.164992 139790875862784 checkpoint_manager.py:2133] [process=1][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0420 22:23:27.166754 139790875862784 checkpoint_manager.py:2142] [process=1][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0420 22:23:27.166928 139917764216640 checkpoint_manager.py:2036] [process=1][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0420 22:23:27.167055 139917764216640 checkpoint_manager.py:1458] Waiting for previous save to complete took 47.217955 seconds. If this number is high, consider checkpointing less frequently. I0420 22:23:27.168999 139917764216640 checkpoint_manager.py:1518] [process=1] Saving checkpoint at step 5 I0420 22:23:27.173203 139917764216640 async_checkpointer.py:452] [process=1] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_false/checkpoints/5. I0420 22:23:28.565578 139917764216640 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0420 22:23:28.565725 139917764216640 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 I0420 22:23:28.608726 139917764216640 base_pytree_checkpoint_handler.py:153] [process=1][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.045337s I0420 22:23:28.608895 139917764216640 base_pytree_checkpoint_handler.py:129] [process=1] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.336 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2890913486480713 s) (per-host) I0420 22:23:28.608948 139917764216640 base_pytree_checkpoint_handler.py:737] [process=1][thread=MainThread] Initiated Pytree async_save. Time taken: 0.289153s (batch_requests_ready=0.238635s, total_serialization_initiated=0.050448s, others=0.000070s) I0420 22:23:28.609044 139917764216640 composite_checkpoint_handler.py:715] [process=1][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.293931s (all_items=0.000015s, per_item={'items': '0.00001526'}, temp_paths=0.293916) I0420 22:23:28.611294 139795162437376 async_checkpointer.py:79] [process=1][thread=async_save] Background save thread started. I0420 22:23:28.611470 139917764216640 async_checkpointer.py:561] Finished blocking save. Time taken: 1.442400s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_false/checkpoints/5. I0420 22:23:29.050872 139917764216640 checkpoint_manager.py:1566] [process=1][thread=MainThread][step=5] Starting CheckpointManager Save Finalize thread=save_finalize I0420 22:23:29.051258 139790875862784 async_checkpointer.py:265] [process=1][thread=save_finalize] Waiting for background save thread=async_save. I0420 22:23:29.051440 139917764216640 standard_logger.py:34] {'step': 5, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776723759.949077, 'wait_for_prev_duration_secs': 47.217955350875854, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776723807.1690383, 'checkpointer_blocking_duration_secs': 1.4425387382507324, 'get_old_steps_start_time': 1776723808.611596, 'get_old_steps_duration_secs': 2.47955322265625e-05, 'checkpoint_manager_blocking_start_time': 1776723759.9468627, 'checkpoint_manager_blocking_duration_secs': 49.10452890396118} I0420 22:23:29.051552 139917764216640 checkpointing.py:409] Started an asynchronous checkpoint save for step 5 I0420 22:23:29.052323 139917764216640 metric_logger.py:185] completed step: 4, seconds: 0.163, TFLOP/s/device: 83.601, Tokens/s/device: 12601.216, total_weights: 65536, loss: 9.457 I0420 22:23:29.059869 139917764216640 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.394, Tokens/s/device: 13926.573, total_weights: 65536, loss: 8.955 I0420 22:23:50.088641 139917764216640 metric_logger.py:185] completed step: 6, seconds: 49.401, TFLOP/s/device: 0.275, Tokens/s/device: 41.457, total_weights: 65536, loss: 8.465 I0420 22:23:50.236135 139917764216640 metric_logger.py:185] completed step: 7, seconds: 20.883, TFLOP/s/device: 0.651, Tokens/s/device: 98.070, total_weights: 65536, loss: 7.998 I0420 22:23:50.383530 139917764216640 metric_logger.py:185] completed step: 8, seconds: 0.154, TFLOP/s/device: 88.433, Tokens/s/device: 13329.515, total_weights: 65536, loss: 7.567 I0420 22:23:50.532028 139917764216640 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0420 22:23:50.532617 139917764216640 checkpointing.py:798] Waited 0.0005993843078613281 seconds for step 9 to finish before starting checkpointing. I0420 22:23:50.535187 139917764216640 checkpoint_manager.py:2024] [process=1][thread=MainThread][step=5][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0420 22:23:55.254415 139790867470080 array_metadata_store.py:203] [process=1][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_false/checkpoints/5/items/array_metadatas/process_1 I0420 22:24:31.689656 139795162437376 base_pytree_checkpoint_handler.py:129] [process=1] /jax/checkpoint/write/gbytes_per_sec: 24.927 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 63.36978840827942 s) (per-host) I0420 22:24:31.689787 139795162437376 async_checkpointer.py:90] [process=1][thread=async_save] 3 Handler Commit operations completed. Time taken: 63.078382s. I0420 22:24:39.566844 139795162437376 async_checkpointer.py:144] [process=1][thread=async_save] Background save thread done. Time taken: 70.955425s. I0420 22:24:39.567124 139790875862784 async_checkpointer.py:273] [process=1][thread=save_finalize] Done with waiting for background save thread=async_save. I0420 22:24:39.567264 139790875862784 async_checkpointer.py:283] [process=1][thread=save_finalize] No errors found in background save thread=async_save. I0420 22:24:39.567314 139790875862784 checkpoint_manager.py:2133] [process=1][thread=save_finalize][step=5] CheckpointManager Save Finalize is syncing with other hosts... I0420 22:24:39.571052 139790875862784 checkpoint_manager.py:2142] [process=1][thread=save_finalize][step=5] CheckpointManager Save Finalize is done on all hosts. I0420 22:24:39.571227 139917764216640 checkpoint_manager.py:2036] [process=1][thread=MainThread][step=5][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=5. W0420 22:24:39.571357 139917764216640 checkpoint_manager.py:1458] Waiting for previous save to complete took 49.036180 seconds. If this number is high, consider checkpointing less frequently. I0420 22:24:39.572855 139917764216640 checkpoint_manager.py:1518] [process=1] Saving checkpoint at step 9 I0420 22:24:39.576265 139917764216640 async_checkpointer.py:452] [process=1] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_false/checkpoints/9. I0420 22:24:40.556147 139917764216640 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0420 22:24:40.556247 139917764216640 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 I0420 22:24:40.601763 139917764216640 base_pytree_checkpoint_handler.py:153] [process=1][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.047603s I0420 22:24:40.601932 139917764216640 base_pytree_checkpoint_handler.py:129] [process=1] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.000 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.30849623680114746 s) (per-host) I0420 22:24:40.601990 139917764216640 base_pytree_checkpoint_handler.py:737] [process=1][thread=MainThread] Initiated Pytree async_save. Time taken: 0.308560s (batch_requests_ready=0.256471s, total_serialization_initiated=0.052021s, others=0.000068s) I0420 22:24:40.602102 139917764216640 composite_checkpoint_handler.py:715] [process=1][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.312560s (all_items=0.000012s, per_item={'items': '0.00001216'}, temp_paths=0.312547) I0420 22:24:40.603046 139790894524160 async_checkpointer.py:79] [process=1][thread=async_save] Background save thread started. I0420 22:24:40.603201 139917764216640 async_checkpointer.py:561] Finished blocking save. Time taken: 1.030278s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_false/checkpoints/9. I0420 22:24:41.040264 139917764216640 checkpoint_manager.py:1566] [process=1][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0420 22:24:41.040641 139790875862784 async_checkpointer.py:265] [process=1][thread=save_finalize] Waiting for background save thread=async_save. I0420 22:24:41.040807 139917764216640 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776723830.5351608, 'wait_for_prev_duration_secs': 49.03618001937866, 'time_between_consecutive_saves_sec': 23.368367671966553, 'checkpointer_blocking_start_time': 1776723879.5728931, 'checkpointer_blocking_duration_secs': 1.0304129123687744, 'get_old_steps_start_time': 1776723880.6033187, 'get_old_steps_duration_secs': 1.9311904907226562e-05, 'checkpoint_manager_blocking_start_time': 1776723830.53283, 'checkpoint_manager_blocking_duration_secs': 50.50794339179993} I0420 22:24:41.040926 139917764216640 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0420 22:24:41.040969 139917764216640 checkpoint_manager.py:2024] [process=1][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0420 22:24:46.709598 139794675922688 array_metadata_store.py:203] [process=1][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_16_zarr3_false/checkpoints/9/items/array_metadatas/process_1 I0420 22:25:21.932480 139790894524160 base_pytree_checkpoint_handler.py:129] [process=1] /jax/checkpoint/write/gbytes_per_sec: 37.936 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.63900637626648 s) (per-host) I0420 22:25:21.932595 139790894524160 async_checkpointer.py:90] [process=1][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.329440s. I0420 22:25:30.154834 139790894524160 async_checkpointer.py:144] [process=1][thread=async_save] Background save thread done. Time taken: 49.551664s. I0420 22:25:30.155121 139790875862784 async_checkpointer.py:273] [process=1][thread=save_finalize] Done with waiting for background save thread=async_save. I0420 22:25:30.155241 139790875862784 async_checkpointer.py:283] [process=1][thread=save_finalize] No errors found in background save thread=async_save. I0420 22:25:30.155282 139790875862784 checkpoint_manager.py:2133] [process=1][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0420 22:25:30.156880 139790875862784 checkpoint_manager.py:2142] [process=1][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0420 22:25:30.157056 139917764216640 checkpoint_manager.py:2036] [process=1][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0420 22:25:30.157222 139917764216640 checkpoint_manager.py:2013] [process=1][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0420 22:25:30.157912 139917764216640 metric_logger.py:185] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.445, Tokens/s/device: 13934.248, total_weights: 65536, loss: 7.176 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Mon Apr 20 22:25:39 UTC 2026 EXIT_CODE=0