XPK Start: Mon Apr 20 20:56:12 UTC 2026 2026-04-20 20:56:17.015282: 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 00:00:1776718577.027990 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1776718577.031622 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1776718577.042986 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776718577.043004 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776718577.043006 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776718577.043008 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-20 20:56:36.282123: 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 20:56:36.800614 134889579550528 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-20 20:56:45,842:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0420 20:56:45.842207 134889579550528 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-20 20:56:45,844:jax._src.distributed:157: Connecting to JAX distributed service on mt-02-synthetic-2qicu-slice-job-0-0.mt-02-synthetic-2qicu:8482 I0420 20:56:45.844600 134889579550528 distributed.py:157] Connecting to JAX distributed service on mt-02-synthetic-2qicu-slice-job-0-0.mt-02-synthetic-2qicu:8482 I0420 20:56:46.703754 134889579550528 max_utils.py:284] Jax distributed system initialized! I0420 20:56:53.797003 134889579550528 max_utils.py:800] System Information: Jax Version: 0.8.1 I0420 20:56:53.797129 134889579550528 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0420 20:56:53.797172 134889579550528 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 20:56:53.797208 134889579550528 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0420 20:56:54.408899 134889579550528 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0420 20:56:54.524259 134889579550528 checkpointing.py:688] Setting up checkpoint logger... I0420 20:56:54.524378 134889579550528 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0420 20:56:54.524423 134889579550528 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0420 20:56:54.524622 134889579550528 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 0x7aadeb9791c0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0420 20:56:57.337028 134889579550528 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0420 20:56:57.337382 134889579550528 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 0x7a99946c86b0>}, handler_registry=None I0420 20:56:57.337623 134889579550528 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7a99946c86b0>` 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 20:56:57.337672 134889579550528 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7a9994326450>` 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 20:56:57.337708 134889579550528 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 0x7a99946c86b0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7a99946c86b0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7a9994326450>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7a9994326450>}). I0420 20:56:57.338043 134889579550528 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0420 20:56:57.338138 134889579550528 async_checkpointer.py:177] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7a999425d260> timeout: 600 secs and primary_host=0 for async checkpoint writes I0420 20:56:58.790441 134889579550528 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_02_synthetic/checkpoints I0420 20:56:59.028646 134889579550528 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_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7a9964610590> I0420 20:56:59.028817 134889579550528 checkpointing.py:302] Checkpoint manager created! I0420 20:56:59.951154 134889579550528 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0420 20:56:59.951276 134889579550528 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0420 20:57:01.368337 134889579550528 maxtext_utils.py:1749] decoder/decoder_norm/scale/value Shape: float32[2048] Physical: (None,) I0420 20:57:01.368461 134889579550528 maxtext_utils.py:1749] decoder/layers/mlp/wi_0/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0420 20:57:01.368518 134889579550528 maxtext_utils.py:1749] decoder/layers/mlp/wi_1/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0420 20:57:01.368563 134889579550528 maxtext_utils.py:1749] decoder/layers/mlp/wo/kernel/value Shape: float32[7168,16,2048] Physical: (None, None, 'fsdp') I0420 20:57:01.368601 134889579550528 maxtext_utils.py:1749] decoder/layers/post_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0420 20:57:01.368638 134889579550528 maxtext_utils.py:1749] decoder/layers/pre_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0420 20:57:01.368677 134889579550528 maxtext_utils.py:1749] decoder/layers/self_attention/key/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0420 20:57:01.368714 134889579550528 maxtext_utils.py:1749] decoder/layers/self_attention/out/kernel/value Shape: float32[16,16,128,2048] Physical: (None, None, None, 'fsdp') I0420 20:57:01.368750 134889579550528 maxtext_utils.py:1749] decoder/layers/self_attention/query/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0420 20:57:01.368785 134889579550528 maxtext_utils.py:1749] decoder/layers/self_attention/value/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0420 20:57:01.368819 134889579550528 maxtext_utils.py:1749] decoder/logits_dense/kernel/value Shape: float32[2048,32000] Physical: ('fsdp', None) I0420 20:57:01.368859 134889579550528 maxtext_utils.py:1749] token_embedder/embedding/value Shape: float32[32000,2048] Physical: (None, 'fsdp') I0420 20:57:01.767141 134889579550528 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0420 20:57:01.767297 134889579550528 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 20:57:01.772625 134889579550528 nnx_decoders.py:616] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0420 20:57:01.772718 134889579550528 nnx_decoders.py:616] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 20:57:01.926416 134889579550528 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0420 20:57:01.926553 134889579550528 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 20:57:02.077652 134889579550528 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0420 20:57:02.077786 134889579550528 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 20:57:02.243396 134889579550528 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 20:57:02.243537 134889579550528 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 20:57:02.396150 134889579550528 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 20:57:02.396286 134889579550528 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 20:57:02.546493 134889579550528 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 20:57:02.546630 134889579550528 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 20:57:02.709378 134889579550528 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0420 20:57:02.709519 134889579550528 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 20:57:02.732301 134889579550528 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0420 20:57:02.732400 134889579550528 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0420 20:57:13.749824 134889579550528 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 20:57:13.757432 134889579550528 metric_logger.py:289] number parameters: 1.104 billion I0420 20:57:25.078363 134889579550528 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0420 20:57:25.229789 134889579550528 checkpointing.py:798] Waited 0.1514120101928711 seconds for step 0 to finish before starting checkpointing. I0420 20:57:25.232455 134889579550528 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0420 20:57:25.234175 134889579550528 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 0 I0420 20:57:25.237173 134889579550528 async_checkpointer.py:452] [process=0] 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_02_synthetic/checkpoints/0. I0420 20:57:25.562783 134889579550528 signaling_client.py:364] Using JaxDistributedSignalingClient I0420 20:57:25.824264 134762475792128 atomicity.py:137] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0 I0420 20:57:25.843782 134889579550528 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0420 20:57:25.843851 134889579550528 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 20:57:26.087144 134889579550528 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.245415s I0420 20:57:26.088039 134889579550528 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.981 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5174770355224609 s) (per-host) I0420 20:57:26.088118 134889579550528 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.517574s (batch_requests_ready=0.263634s, total_serialization_initiated=0.253118s, others=0.000822s) I0420 20:57:26.088199 134889579550528 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.521549s (all_items=0.000019s, per_item={'items': '0.00001884'}, temp_paths=0.521531) I0420 20:57:26.089159 134763541141248 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0420 20:57:26.089318 134889579550528 async_checkpointer.py:561] Finished blocking save. Time taken: 0.855071s. 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_02_synthetic/checkpoints/0. I0420 20:57:26.129539 134889579550528 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0420 20:57:26.129828 134762467399424 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0420 20:57:26.129966 134889579550528 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_02_synthetic/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776718645.2324386, 'wait_for_prev_duration_secs': 6.413459777832031e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776718645.234215, 'checkpointer_blocking_duration_secs': 0.8552088737487793, 'get_old_steps_start_time': 1776718646.089444, 'get_old_steps_duration_secs': 2.5272369384765625e-05, 'checkpoint_manager_blocking_start_time': 1776718645.230206, 'checkpoint_manager_blocking_duration_secs': 0.8997211456298828} I0420 20:57:26.130075 134889579550528 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0420 20:57:26.130163 134889579550528 max_utils.py:750] Memstats: After params initialized: I0420 20:57:26.130216 134889579550528 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0)) I0420 20:57:26.130251 134889579550528 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0)) I0420 20:57:26.130279 134889579550528 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0)) I0420 20:57:26.130306 134889579550528 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0)) I0420 20:57:26.438293 134889579550528 metric_logger.py:185] completed step: 0, seconds: 11.051, TFLOP/s/device: 1.230, Tokens/s/device: 185.328, total_weights: 65536, loss: 10.874 I0420 20:57:26.441027 134889579550528 metric_logger.py:269] To see full metrics 'tensorboard --logdir=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/tensorboard/' I0420 20:57:26.845905 134889579550528 metric_logger.py:185] completed step: 1, seconds: 1.356, TFLOP/s/device: 10.017, Tokens/s/device: 1509.879, total_weights: 65536, loss: 10.874 I0420 20:57:26.988175 134889579550528 metric_logger.py:185] completed step: 2, seconds: 0.409, TFLOP/s/device: 33.232, Tokens/s/device: 5009.062, total_weights: 65536, loss: 10.262 I0420 20:57:27.135511 134889579550528 metric_logger.py:185] completed step: 3, seconds: 0.011, TFLOP/s/device: 1200.489, Tokens/s/device: 180950.698, total_weights: 65536, loss: 9.732 I0420 20:57:46.377367 134763010414336 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776718647396335570, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776718647396335570, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0/_CHECKPOINT_METADATA I0420 20:57:46.378721 134763591497472 atomicity.py:137] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0/items I0420 20:57:46.379140 134889579550528 metric_logger.py:185] completed step: 4, seconds: 0.144, TFLOP/s/device: 94.032, Tokens/s/device: 14173.599, total_weights: 65536, loss: 9.274 I0420 20:57:46.389426 134889579550528 metric_logger.py:185] completed step: 5, seconds: 0.146, TFLOP/s/device: 92.765, Tokens/s/device: 13982.481, total_weights: 65536, loss: 8.888 I0420 20:57:46.533900 134889579550528 metric_logger.py:185] completed step: 6, seconds: 19.242, TFLOP/s/device: 0.706, Tokens/s/device: 106.432, total_weights: 65536, loss: 8.589 I0420 20:57:46.681337 134889579550528 metric_logger.py:185] completed step: 7, seconds: 0.007, TFLOP/s/device: 1812.584, Tokens/s/device: 273212.380, total_weights: 65536, loss: 8.380 I0420 20:57:46.828697 134889579550528 metric_logger.py:185] completed step: 8, seconds: 0.146, TFLOP/s/device: 92.959, Tokens/s/device: 14011.850, total_weights: 65536, loss: 8.251 I0420 20:57:46.976959 134889579550528 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0420 20:57:46.977652 134889579550528 checkpointing.py:798] Waited 0.0006985664367675781 seconds for step 9 to finish before starting checkpointing. I0420 20:57:46.979618 134889579550528 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0420 20:57:47.642395 2653 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0420 20:57:49.310703 134763574712064 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_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0/items/array_metadatas/process_0 I0420 20:58:25.058860 134763549533952 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 37.429570s (commit=36.036742s, array_metadata_write=1.392828s) I0420 20:58:25.060010 134763541141248 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 26.553 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 59.489418268203735 s) (per-host) I0420 20:58:25.060149 134763541141248 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 58.970880s. I0420 20:58:25.747065 134763541141248 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776718647396335570, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0/_CHECKPOINT_METADATA I0420 20:58:26.531431 134763010414336 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776718647396335570, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0/_CHECKPOINT_METADATA I0420 20:58:27.623914 134763541141248 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0420 20:58:27.737030 134763541141248 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000426s. I0420 20:58:28.551740 134763541141248 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.663340s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0/items I0420 20:58:28.553894 134763541141248 atomicity.py:608] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0/items I0420 20:58:28.980463 134763541141248 atomicity.py:608] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0 I0420 20:58:30.268975 134763541141248 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_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0`. I0420 20:58:30.269712 134763541141248 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 65.035484s. directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0 I0420 20:58:30.271120 134763541141248 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 64.181850s. I0420 20:58:30.271316 134762467399424 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0420 20:58:30.271435 134762467399424 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0420 20:58:30.271500 134762467399424 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0420 20:58:30.273176 134762467399424 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0420 20:58:30.273348 134889579550528 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. W0420 20:58:30.273487 134889579550528 checkpoint_manager.py:1458] Waiting for previous save to complete took 43.293872 seconds. If this number is high, consider checkpointing less frequently. I0420 20:58:30.275792 134889579550528 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 9 I0420 20:58:30.279269 134889579550528 async_checkpointer.py:452] [process=0] 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_02_synthetic/checkpoints/9. I0420 20:58:30.805762 134763541141248 atomicity.py:137] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/9 I0420 20:58:30.830261 134889579550528 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0420 20:58:30.830360 134889579550528 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 20:58:30.861293 134889579550528 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.033201s I0420 20:58:30.862166 134889579550528 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.133 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.3005211353302002 s) (per-host) I0420 20:58:30.862229 134889579550528 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.300598s (batch_requests_ready=0.260794s, total_serialization_initiated=0.039024s, others=0.000779s) I0420 20:58:30.862307 134889579550528 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.304418s (all_items=0.000017s, per_item={'items': '0.00001740'}, temp_paths=0.304401) I0420 20:58:30.863147 134762509362944 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0420 20:58:30.863290 134889579550528 async_checkpointer.py:561] Finished blocking save. Time taken: 0.587436s. 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_02_synthetic/checkpoints/9. I0420 20:58:30.888006 134889579550528 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0420 20:58:30.888282 134762467399424 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0420 20:58:30.888444 134889579550528 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_02_synthetic/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776718666.9795866, 'wait_for_prev_duration_secs': 43.29387164115906, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776718710.275831, 'checkpointer_blocking_duration_secs': 0.5875298976898193, 'get_old_steps_start_time': 1776718710.8633814, 'get_old_steps_duration_secs': 2.3603439331054688e-05, 'checkpoint_manager_blocking_start_time': 1776718666.977886, 'checkpoint_manager_blocking_duration_secs': 43.910526752471924} I0420 20:58:30.888553 134889579550528 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0420 20:58:30.888603 134889579550528 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0420 20:58:33.010354 134761939728128 atomicity.py:137] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/9/items I0420 20:58:37.381594 134763574712064 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_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/9/items/array_metadatas/process_0 I0420 20:59:16.522429 134763549533952 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 41.212279s (commit=40.103329s, array_metadata_write=1.108950s) I0420 20:59:16.523697 134762509362944 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 34.367 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 45.96202278137207 s) (per-host) I0420 20:59:16.523803 134762509362944 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 45.660547s. I0420 20:59:18.379411 134762509362944 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0420 20:59:18.541481 134762509362944 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000408s. I0420 20:59:19.441215 134762509362944 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.902832s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/9/items I0420 20:59:19.442912 134762509362944 atomicity.py:608] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/9/items I0420 20:59:19.889666 134762509362944 atomicity.py:608] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/9 I0420 20:59:21.181010 134762509362944 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_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/9`. I0420 20:59:21.181753 134762509362944 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 50.905908s. directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/nnx_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/9 I0420 20:59:21.183100 134762509362944 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 50.319843s. I0420 20:59:21.183292 134762467399424 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0420 20:59:21.183415 134762467399424 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0420 20:59:21.183478 134762467399424 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0420 20:59:21.185058 134762467399424 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0420 20:59:21.185253 134889579550528 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. I0420 20:59:21.185422 134889579550528 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0420 20:59:21.186173 134889579550528 metric_logger.py:185] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.230, Tokens/s/device: 13901.899, 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: Mon Apr 20 20:59:34 UTC 2026 EXIT_CODE=0