XPK Start: Mon Apr 20 19:53:47 UTC 2026 2026-04-20 19:53:51.383406: 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 00:00:1776714831.396053 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1776714831.399711 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1776714831.411180 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776714831.411197 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776714831.411200 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776714831.411201 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-20 19:54:10.619124: 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 19:54:11.143291 137487649789760 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-20 19:54:20,182:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0420 19:54:20.182798 137487649789760 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-20 19:54:20,185:jax._src.distributed:157: Connecting to JAX distributed service on mt-11-optimizer-offload-czj4o-slice-job-0-0.mt-11-optimizer-offload-czj4o:8482 I0420 19:54:20.185204 137487649789760 distributed.py:157] Connecting to JAX distributed service on mt-11-optimizer-offload-czj4o-slice-job-0-0.mt-11-optimizer-offload-czj4o:8482 I0420 19:54:21.387700 137487649789760 max_utils.py:284] Jax distributed system initialized! I0420 19:54:28.395809 137487649789760 max_utils.py:800] System Information: Jax Version: 0.8.1 I0420 19:54:28.395918 137487649789760 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0420 19:54:28.395959 137487649789760 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 19:54:28.395995 137487649789760 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0420 19:54:29.005348 137487649789760 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0420 19:54:29.005935 137487649789760 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0420 19:54:29.006129 137487649789760 checkpointing.py:688] Setting up checkpoint logger... I0420 19:54:29.006180 137487649789760 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0420 19:54:29.006221 137487649789760 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0420 19:54:29.006557 137487649789760 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 0x7d0adeb79160>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0420 19:54:31.958045 137487649789760 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0420 19:54:31.958427 137487649789760 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 0x7cf618625400>}, handler_registry=None I0420 19:54:31.958661 137487649789760 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7cf618625400>` 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 19:54:31.958710 137487649789760 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7cf61862a9f0>` 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 19:54:31.958745 137487649789760 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 0x7cf618625400>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7cf618625400>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7cf61862a9f0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7cf61862a9f0>}). I0420 19:54:31.959260 137487649789760 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0420 19:54:31.959340 137487649789760 async_checkpointer.py:177] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7cf588700180> timeout: 600 secs and primary_host=0 for async checkpoint writes I0420 19:54:32.729247 137487649789760 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints I0420 19:54:32.774643 137487649789760 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/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7cf618625e50> I0420 19:54:32.774789 137487649789760 checkpointing.py:302] Checkpoint manager created! I0420 19:54:33.655608 137487649789760 nnx_wrappers.py:455] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0420 19:54:33.655717 137487649789760 nnx_wrappers.py:455] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 19:54:34.030487 137487649789760 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0420 19:54:34.030581 137487649789760 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 19:54:34.046419 137487649789760 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0420 19:54:34.046482 137487649789760 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 19:54:34.069339 137487649789760 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 19:54:34.069411 137487649789760 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 19:54:34.085381 137487649789760 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 19:54:34.085444 137487649789760 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 19:54:34.101227 137487649789760 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 19:54:34.101288 137487649789760 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 19:54:34.125999 137487649789760 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0420 19:54:34.126073 137487649789760 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 19:54:34.145725 137487649789760 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0420 19:54:34.145788 137487649789760 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0420 19:54:34.334344 137487649789760 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0420 19:54:34.334447 137487649789760 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0420 19:54:35.761177 137487649789760 maxtext_utils.py:1740] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: PartitionSpec('norm',) Physical: (None,) I0420 19:54:35.761305 137487649789760 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wi_0/kernel Shape: float32[2048,16,7168] Logical: PartitionSpec('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0420 19:54:35.761360 137487649789760 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wi_1/kernel Shape: float32[2048,16,7168] Logical: PartitionSpec('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0420 19:54:35.761419 137487649789760 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: PartitionSpec('mlp', 'layers', 'embed') Physical: (None, None, 'fsdp') I0420 19:54:35.761471 137487649789760 maxtext_utils.py:1740] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: PartitionSpec('norm', 'layers') Physical: (None, None) I0420 19:54:35.761511 137487649789760 maxtext_utils.py:1740] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: PartitionSpec('norm', 'layers') Physical: (None, None) I0420 19:54:35.761563 137487649789760 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/key/kernel Shape: float32[2048,16,16,128] Logical: PartitionSpec('embed', 'layers', 'kv_heads', 'kv_head_dim') Physical: ('fsdp', None, None, None) I0420 19:54:35.761622 137487649789760 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/out/kernel Shape: float32[16,16,128,2048] Logical: PartitionSpec('heads', 'layers', 'kv', 'embed') Physical: (None, None, None, 'fsdp') I0420 19:54:35.761665 137487649789760 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/query/kernel Shape: float32[2048,16,16,128] Logical: PartitionSpec('embed', 'layers', 'q_heads', 'kv') Physical: ('fsdp', None, None, None) I0420 19:54:35.761703 137487649789760 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/value/kernel Shape: float32[2048,16,16,128] Logical: PartitionSpec('embed', 'layers', 'kv_heads', 'kv_head_dim') Physical: ('fsdp', None, None, None) I0420 19:54:35.761775 137487649789760 maxtext_utils.py:1740] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: PartitionSpec('embed', 'vocab') Physical: ('fsdp', None) I0420 19:54:35.761831 137487649789760 maxtext_utils.py:1740] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: PartitionSpec('vocab', 'embed') Physical: (None, 'fsdp') I0420 19:54:36.259642 137487649789760 train.py:158] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0420 19:54:36.259735 137487649789760 train.py:158] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0420 19:54:36.274657 137487649789760 train.py:165] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0420 19:54:36.274715 137487649789760 train.py:165] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0420 19:54:47.045382 137487649789760 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 19:54:47.051637 137487649789760 metric_logger.py:289] number parameters: 1.104 billion I0420 19:54:58.442919 137487649789760 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0420 19:54:58.596248 137487649789760 checkpointing.py:798] Waited 0.15331459045410156 seconds for step 0 to finish before starting checkpointing. I0420 19:54:58.598623 137487649789760 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0420 19:54:58.600416 137487649789760 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 0 I0420 19:54:58.603531 137487649789760 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/0. I0420 19:54:58.874965 137487649789760 signaling_client.py:364] Using JaxDistributedSignalingClient I0420 19:54:59.127190 137359739832064 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/0 I0420 19:54:59.141261 137487649789760 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0420 19:54:59.141354 137487649789760 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 19:54:59.384154 137487649789760 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.243804s I0420 19:54:59.384957 137487649789760 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 3.071 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.502305269241333 s) (per-host) I0420 19:54:59.385022 137487649789760 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.502393s (batch_requests_ready=0.253818s, total_serialization_initiated=0.247865s, others=0.000710s) I0420 19:54:59.385110 137487649789760 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.506389s (all_items=0.000019s, per_item={'items': '0.00001931'}, temp_paths=0.506370) I0420 19:54:59.386122 137360293488384 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0420 19:54:59.386284 137487649789760 async_checkpointer.py:561] Finished blocking save. Time taken: 0.785796s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/0. I0420 19:54:59.848049 137487649789760 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0420 19:54:59.848440 137359731439360 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0420 19:54:59.848619 137487649789760 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776714898.598606, 'wait_for_prev_duration_secs': 6.365776062011719e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776714898.6004546, 'checkpointer_blocking_duration_secs': 0.7859351634979248, 'get_old_steps_start_time': 1776714899.3864105, 'get_old_steps_duration_secs': 2.4318695068359375e-05, 'checkpoint_manager_blocking_start_time': 1776714898.5968344, 'checkpoint_manager_blocking_duration_secs': 1.2517273426055908} I0420 19:54:59.848724 137487649789760 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0420 19:54:59.848791 137487649789760 max_utils.py:750] Memstats: After params initialized: I0420 19:54:59.848866 137487649789760 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0)) I0420 19:54:59.848923 137487649789760 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0)) I0420 19:54:59.848967 137487649789760 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0)) I0420 19:54:59.849010 137487649789760 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0)) I0420 19:55:00.147045 137487649789760 metric_logger.py:185] completed step: 0, seconds: 11.123, TFLOP/s/device: 1.222, Tokens/s/device: 184.131, total_weights: 65536, loss: 10.874 I0420 19:55:00.149927 137487649789760 metric_logger.py:269] To see full metrics 'tensorboard --logdir=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/tensorboard/' I0420 19:55:00.568663 137487649789760 metric_logger.py:185] completed step: 1, seconds: 1.703, TFLOP/s/device: 7.981, Tokens/s/device: 1202.925, total_weights: 65536, loss: 10.874 I0420 19:55:00.714777 137487649789760 metric_logger.py:185] completed step: 2, seconds: 0.422, TFLOP/s/device: 32.168, Tokens/s/device: 4848.726, total_weights: 65536, loss: 10.263 I0420 19:55:00.862273 137487649789760 metric_logger.py:185] completed step: 3, seconds: 0.008, TFLOP/s/device: 1632.088, Tokens/s/device: 246006.006, total_weights: 65536, loss: 9.734 I0420 19:55:19.507291 137487649789760 metric_logger.py:185] completed step: 4, seconds: 0.147, TFLOP/s/device: 92.385, Tokens/s/device: 13925.342, total_weights: 65536, loss: 9.278 I0420 19:55:19.515893 137487649789760 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.142, Tokens/s/device: 13888.606, total_weights: 65536, loss: 8.892 I0420 19:55:19.662137 137487649789760 metric_logger.py:185] completed step: 6, seconds: 18.644, TFLOP/s/device: 0.729, Tokens/s/device: 109.845, total_weights: 65536, loss: 8.592 I0420 19:55:19.809697 137487649789760 metric_logger.py:185] completed step: 7, seconds: 0.006, TFLOP/s/device: 2243.211, Tokens/s/device: 338121.182, total_weights: 65536, loss: 8.384 I0420 19:55:19.892648 137361870530304 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/0/items I0420 19:55:19.956951 137487649789760 metric_logger.py:185] completed step: 8, seconds: 0.147, TFLOP/s/device: 92.200, Tokens/s/device: 13897.371, total_weights: 65536, loss: 8.255 I0420 19:55:20.054059 137360268596992 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776714919636215820, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776714919636215820, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/0/_CHECKPOINT_METADATA I0420 19:55:20.104001 137487649789760 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0420 19:55:20.104706 137487649789760 checkpointing.py:798] Waited 0.0007250308990478516 seconds for step 9 to finish before starting checkpointing. I0420 19:55:20.106646 137487649789760 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0420 19:55:21.028390 2611 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0420 19:55:22.692790 137360838752000 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/0/items/array_metadatas/process_0 I0420 19:55:58.840956 137360301881088 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 37.816241s (commit=36.401275s, array_metadata_write=1.414966s) I0420 19:55:58.842137 137360293488384 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 26.344 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 59.95944428443909 s) (per-host) I0420 19:55:58.842252 137360293488384 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 59.456020s. I0420 19:55:59.618770 137360293488384 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776714919636215820, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/0/_CHECKPOINT_METADATA I0420 19:56:00.380409 137360268596992 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776714919636215820, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/0/_CHECKPOINT_METADATA I0420 19:56:00.679406 137360293488384 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000250s. I0420 19:56:01.003234 137360293488384 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0420 19:56:01.958966 137360293488384 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.216154s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/0/items I0420 19:56:01.960998 137360293488384 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/0/items I0420 19:56:02.378463 137360293488384 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/0 I0420 19:56:03.709216 137360293488384 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/0`. I0420 19:56:03.710041 137360293488384 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 65.109560s. directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/0 I0420 19:56:03.711438 137360293488384 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 64.325205s. I0420 19:56:03.711560 137359731439360 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0420 19:56:03.711622 137359731439360 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0420 19:56:03.711698 137359731439360 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0420 19:56:03.713533 137359731439360 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0420 19:56:03.713669 137487649789760 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 19:56:03.713763 137487649789760 checkpoint_manager.py:1458] Waiting for previous save to complete took 43.607124 seconds. If this number is high, consider checkpointing less frequently. I0420 19:56:03.715783 137487649789760 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 9 I0420 19:56:03.719318 137487649789760 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/9. I0420 19:56:04.255934 137487649789760 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0420 19:56:04.256031 137487649789760 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 19:56:04.282503 137487649789760 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.027795s I0420 19:56:04.283421 137487649789760 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.574 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.27672457695007324 s) (per-host) I0420 19:56:04.283484 137487649789760 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.276803s (batch_requests_ready=0.245947s, total_serialization_initiated=0.030033s, others=0.000823s) I0420 19:56:04.283573 137487649789760 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.280802s (all_items=0.000016s, per_item={'items': '0.00001621'}, temp_paths=0.280786) I0420 19:56:04.284405 137360251811584 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0420 19:56:04.284556 137487649789760 async_checkpointer.py:561] Finished blocking save. Time taken: 0.568703s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/9. I0420 19:56:04.330550 137487649789760 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0420 19:56:04.330905 137359731439360 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0420 19:56:04.331073 137487649789760 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776714920.1066167, 'wait_for_prev_duration_secs': 43.60712385177612, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776714963.7158234, 'checkpointer_blocking_duration_secs': 0.5688371658325195, 'get_old_steps_start_time': 1776714964.2846794, 'get_old_steps_duration_secs': 2.5272369384765625e-05, 'checkpoint_manager_blocking_start_time': 1776714920.104936, 'checkpoint_manager_blocking_duration_secs': 44.22610330581665} I0420 19:56:04.331193 137487649789760 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0420 19:56:04.331237 137487649789760 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0420 19:56:04.697614 137360293488384 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/9 I0420 19:56:06.482230 137360243418880 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/9/items I0420 19:56:09.734064 137360838752000 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/9/items/array_metadatas/process_0 I0420 19:56:49.479119 137360301881088 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 41.466320s (commit=40.218220s, array_metadata_write=1.248100s) I0420 19:56:49.480234 137360251811584 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 34.737 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 45.47350454330444 s) (per-host) I0420 19:56:49.480356 137360251811584 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 45.195841s. I0420 19:56:51.269362 137360251811584 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000214s. I0420 19:56:51.685949 137360251811584 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0420 19:56:52.660792 137360251811584 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.232693s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/9/items I0420 19:56:52.662798 137360251811584 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/9/items I0420 19:56:53.090700 137360251811584 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/9 I0420 19:56:54.412226 137360251811584 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/9`. I0420 19:56:54.413002 137360251811584 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 50.697152s. directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_11_optimizer_offload_false/checkpoints/9 I0420 19:56:54.414633 137360251811584 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 50.130117s. I0420 19:56:54.414860 137359731439360 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0420 19:56:54.414972 137359731439360 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0420 19:56:54.415036 137359731439360 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0420 19:56:54.416705 137359731439360 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0420 19:56:54.416827 137487649789760 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 19:56:54.416917 137487649789760 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0420 19:56:54.417595 137487649789760 metric_logger.py:185] completed step: 9, seconds: 0.148, TFLOP/s/device: 92.029, Tokens/s/device: 13871.580, total_weights: 65536, loss: 8.179 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Mon Apr 20 19:57:07 UTC 2026 EXIT_CODE=0