XPK Start: Fri Apr 24 15:30:39 UTC 2026 2026-04-24 15:30:43.393271: 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 00:00:1777044643.406196 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1777044643.409900 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1777044643.421477 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1777044643.421494 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1777044643.421497 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1777044643.421499 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-24 15:31:02.591350: 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) I0424 15:31:03.113609 133599900174144 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-24 15:31:12,155:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0424 15:31:12.155074 133599900174144 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-24 15:31:12,157:jax._src.distributed:157: Connecting to JAX distributed service on mt-09-pdb-lt-1-lgnmi-slice-job-0-0.mt-09-pdb-lt-1-lgnmi:8482 I0424 15:31:12.157374 133599900174144 distributed.py:157] Connecting to JAX distributed service on mt-09-pdb-lt-1-lgnmi-slice-job-0-0.mt-09-pdb-lt-1-lgnmi:8482 I0424 15:31:13.248705 133599900174144 max_utils.py:284] Jax distributed system initialized! I0424 15:31:20.171367 133599900174144 max_utils.py:800] System Information: Jax Version: 0.8.1 I0424 15:31:20.171474 133599900174144 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0424 15:31:20.171520 133599900174144 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 I0424 15:31:20.171557 133599900174144 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0424 15:31:20.789178 133599900174144 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 8, 1, 1, 1, 1, 4, 1, 1, 1, 1) I0424 15:31:20.789776 133599900174144 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 8, 1, 1, 1, 1, 4, 1, 1, 1, 1) I0424 15:31:20.789968 133599900174144 checkpointing.py:688] Setting up checkpoint logger... I0424 15:31:20.790022 133599900174144 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0424 15:31:20.790065 133599900174144 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0424 15:31:20.790447 133599900174144 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 0x7981a0379550>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0424 15:31:23.635846 133599900174144 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0424 15:31:23.636237 133599900174144 checkpoint_manager.py:709] [process=4][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x796cf44b1d30>}, handler_registry=None I0424 15:31:23.636491 133599900174144 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x796cf44b1d30>` 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`. I0424 15:31:23.636541 133599900174144 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x796cf44b6d80>` 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`. I0424 15:31:23.636577 133599900174144 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 0x796cf44b1d30>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x796cf44b1d30>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x796cf44b6d80>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x796cf44b6d80>}). I0424 15:31:23.637104 133599900174144 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0424 15:31:23.637183 133599900174144 async_checkpointer.py:177] [process=4][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x796c7c490180> timeout: 600 secs and primary_host=0 for async checkpoint writes I0424 15:31:24.456418 133599900174144 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/linen_xpk_feat_nnx_set_defaults_true_20260424_145744_09_pdb_lt_1/checkpoints I0424 15:31:24.867658 133599900174144 checkpoint_manager.py:929] [process=4][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_20260424_145744/linen_xpk_feat_nnx_set_defaults_true_20260424_145744_09_pdb_lt_1/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x796cf44b2a50> I0424 15:31:24.867843 133599900174144 checkpointing.py:302] Checkpoint manager created! I0424 15:31:25.779693 133599900174144 nnx_wrappers.py:455] Unknown Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0424 15:31:25.779800 133599900174144 nnx_wrappers.py:455] Unknown Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor'). I0424 15:31:26.150654 133599900174144 attentions.py:1084] attentions/inputs_q Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0424 15:31:26.150750 133599900174144 attentions.py:1084] attentions/inputs_q Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor'). I0424 15:31:26.166710 133599900174144 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0424 15:31:26.166773 133599900174144 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor'). I0424 15:31:26.189743 133599900174144 attentions.py:1150] attentions/query Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0424 15:31:26.189818 133599900174144 attentions.py:1150] attentions/query Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0424 15:31:26.205842 133599900174144 attentions.py:1151] attentions/key Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0424 15:31:26.205914 133599900174144 attentions.py:1151] attentions/key Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0424 15:31:26.221907 133599900174144 attentions.py:1152] attentions/value Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0424 15:31:26.221965 133599900174144 attentions.py:1152] attentions/value Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0424 15:31:26.246926 133599900174144 attentions.py:1193] attentions/out Logical: bfloat16[8,2048,16,128]..................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0424 15:31:26.246999 133599900174144 attentions.py:1193] attentions/out Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0424 15:31:26.266767 133599900174144 linears.py:541] linears/x Logical: bfloat16[8,2048,7168]....................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0424 15:31:26.266838 133599900174144 linears.py:541] linears/x Physical: bfloat16[8,2048,7168]....................................... ('fsdp', None, 'tensor'). I0424 15:31:26.460309 133599900174144 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0424 15:31:26.460417 133599900174144 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 8 tensor: 4 I0424 15:31:27.904540 133599900174144 maxtext_utils.py:1740] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: PartitionSpec('norm',) Physical: ('tensor',) I0424 15:31:27.904682 133599900174144 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, 'tensor') I0424 15:31:27.904766 133599900174144 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, 'tensor') I0424 15:31:27.904851 133599900174144 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: PartitionSpec('mlp', 'layers', 'embed') Physical: ('tensor', None, 'fsdp') I0424 15:31:27.904950 133599900174144 maxtext_utils.py:1740] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: PartitionSpec('norm', 'layers') Physical: ('tensor', None) I0424 15:31:27.905018 133599900174144 maxtext_utils.py:1740] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: PartitionSpec('norm', 'layers') Physical: ('tensor', None) I0424 15:31:27.905112 133599900174144 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, 'tensor', None) I0424 15:31:27.905200 133599900174144 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: ('tensor', None, None, 'fsdp') I0424 15:31:27.905275 133599900174144 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, 'tensor', None) I0424 15:31:27.905340 133599900174144 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, 'tensor', None) I0424 15:31:27.905424 133599900174144 maxtext_utils.py:1740] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: PartitionSpec('embed', 'vocab') Physical: ('fsdp', 'tensor') I0424 15:31:27.905497 133599900174144 maxtext_utils.py:1740] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: PartitionSpec('vocab', 'embed') Physical: ('tensor', 'fsdp') I0424 15:31:28.412223 133599900174144 train.py:158] train/xent Logical: float32[8,2048]............................................. ('activation_embed_and_logits_batch', 'activation_length'). I0424 15:31:28.412319 133599900174144 train.py:158] train/xent Physical: float32[8,2048]............................................. ('fsdp', None). I0424 15:31:28.427682 133599900174144 train.py:165] train/z_loss Logical: float32[8,2048]............................................. ('activation_embed_and_logits_batch', 'activation_length'). I0424 15:31:28.427747 133599900174144 train.py:165] train/z_loss Physical: float32[8,2048]............................................. ('fsdp', None). I0424 15:31:38.486400 133599900174144 max_utils.py:791] Total memory size: 0.9 GB, Output size: 0.4 GB, Temp size: 0.5 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB. I0424 15:31:38.487280 133599900174144 metric_logger.py:289] number parameters: 1.104 billion I0424 15:31:40.037936 133599900174144 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0424 15:31:48.651520 133599900174144 checkpointing.py:798] Waited 8.613563299179077 seconds for step 0 to finish before starting checkpointing. I0424 15:31:48.654024 133599900174144 checkpoint_manager.py:2013] [process=4][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0424 15:31:48.655947 133599900174144 checkpoint_manager.py:1518] [process=4] Saving checkpoint at step 0 I0424 15:31:48.658677 133599900174144 async_checkpointer.py:452] [process=4] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/linen_xpk_feat_nnx_set_defaults_true_20260424_145744_09_pdb_lt_1/checkpoints/0. I0424 15:31:49.210633 133599900174144 signaling_client.py:364] Using JaxDistributedSignalingClient I0424 15:31:49.213165 133599900174144 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0424 15:31:49.213223 133599900174144 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 I0424 15:31:49.438800 133599900174144 base_pytree_checkpoint_handler.py:153] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.228244s I0424 15:31:49.438985 133599900174144 base_pytree_checkpoint_handler.py:129] [process=4] /jax/checkpoint/write/blocking_gbytes_per_sec: 3.081 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5007405281066895 s) (per-host) I0424 15:31:49.439041 133599900174144 base_pytree_checkpoint_handler.py:737] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.500808s (batch_requests_ready=0.268957s, total_serialization_initiated=0.231776s, others=0.000075s) I0424 15:31:49.439130 133599900174144 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.505224s (all_items=0.000019s, per_item={'items': '0.00001907'}, temp_paths=0.505205) I0424 15:31:49.440033 133476938393344 async_checkpointer.py:79] [process=4][thread=async_save] Background save thread started. I0424 15:31:49.440204 133599900174144 async_checkpointer.py:561] Finished blocking save. Time taken: 0.784187s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/linen_xpk_feat_nnx_set_defaults_true_20260424_145744_09_pdb_lt_1/checkpoints/0. I0424 15:31:49.470200 133599900174144 checkpoint_manager.py:1566] [process=4][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0424 15:31:49.470479 133473759131392 async_checkpointer.py:265] [process=4][thread=save_finalize] Waiting for background save thread=async_save. I0424 15:31:49.470624 133599900174144 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/linen_xpk_feat_nnx_set_defaults_true_20260424_145744_09_pdb_lt_1/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777044708.6540055, 'wait_for_prev_duration_secs': 6.341934204101562e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777044708.6559865, 'checkpointer_blocking_duration_secs': 0.7843301296234131, 'get_old_steps_start_time': 1777044709.440336, 'get_old_steps_duration_secs': 2.574920654296875e-05, 'checkpoint_manager_blocking_start_time': 1777044708.65202, 'checkpoint_manager_blocking_duration_secs': 0.8185653686523438} I0424 15:31:49.470728 133599900174144 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0424 15:31:49.470780 133599900174144 max_utils.py:750] Memstats: After params initialized: I0424 15:31:49.470831 133599900174144 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_16(process=4,(0,4,0,0)) I0424 15:31:49.470866 133599900174144 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_17(process=4,(1,4,0,0)) I0424 15:31:49.470906 133599900174144 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_20(process=4,(0,5,0,0)) I0424 15:31:49.470929 133599900174144 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_21(process=4,(1,5,0,0)) I0424 15:31:49.735946 133599900174144 metric_logger.py:185] completed step: 0, seconds: 1.551, TFLOP/s/device: 2.191, Tokens/s/device: 330.212, total_weights: 16384, loss: 10.862 I0424 15:31:49.816099 133599900174144 metric_logger.py:185] completed step: 1, seconds: 9.697, TFLOP/s/device: 0.350, Tokens/s/device: 52.801, total_weights: 16384, loss: 10.862 I0424 15:31:50.235496 133599900174144 metric_logger.py:185] completed step: 2, seconds: 0.022, TFLOP/s/device: 156.160, Tokens/s/device: 23538.065, total_weights: 16384, loss: 9.763 I0424 15:31:50.294010 133599900174144 metric_logger.py:185] completed step: 3, seconds: 0.420, TFLOP/s/device: 8.088, Tokens/s/device: 1219.170, total_weights: 16384, loss: 8.823 I0424 15:31:50.411715 133599900174144 metric_logger.py:185] completed step: 4, seconds: 0.063, TFLOP/s/device: 53.596, Tokens/s/device: 8078.513, total_weights: 16384, loss: 7.947 I0424 15:31:50.417104 133599900174144 metric_logger.py:185] completed step: 5, seconds: 0.058, TFLOP/s/device: 58.232, Tokens/s/device: 8777.343, total_weights: 16384, loss: 7.212 I0424 15:32:02.849044 133599900174144 metric_logger.py:185] completed step: 6, seconds: 0.118, TFLOP/s/device: 28.730, Tokens/s/device: 4330.506, total_weights: 16384, loss: 6.652 I0424 15:32:02.907628 133599900174144 metric_logger.py:185] completed step: 7, seconds: 12.374, TFLOP/s/device: 0.275, Tokens/s/device: 41.378, total_weights: 16384, loss: 6.269 I0424 15:32:02.966201 133599900174144 metric_logger.py:185] completed step: 8, seconds: 0.063, TFLOP/s/device: 53.609, Tokens/s/device: 8080.553, total_weights: 16384, loss: 6.031 I0424 15:32:03.024469 133599900174144 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0424 15:32:03.025074 133599900174144 checkpointing.py:798] Waited 0.0006270408630371094 seconds for step 9 to finish before starting checkpointing. I0424 15:32:03.027041 133599900174144 checkpoint_manager.py:2024] [process=4][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0424 15:32:04.742823 2478 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0424 15:32:06.660198 133476946786048 array_metadata_store.py:203] [process=4][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/linen_xpk_feat_nnx_set_defaults_true_20260424_145744_09_pdb_lt_1/checkpoints/0/items/array_metadatas/process_4 I0424 15:32:38.879971 133476938393344 base_pytree_checkpoint_handler.py:129] [process=4] /jax/checkpoint/write/gbytes_per_sec: 31.629 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 49.94168567657471 s) (per-host) I0424 15:32:38.880094 133476938393344 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 49.439949s. I0424 15:32:45.837342 133476938393344 async_checkpointer.py:144] [process=4][thread=async_save] Background save thread done. Time taken: 56.397182s. I0424 15:32:45.837564 133473759131392 async_checkpointer.py:273] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save. I0424 15:32:45.837685 133473759131392 async_checkpointer.py:283] [process=4][thread=save_finalize] No errors found in background save thread=async_save. I0424 15:32:45.837739 133473759131392 checkpoint_manager.py:2133] [process=4][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0424 15:32:45.839279 133473759131392 checkpoint_manager.py:2142] [process=4][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0424 15:32:45.839505 133599900174144 checkpoint_manager.py:2036] [process=4][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0424 15:32:45.839661 133599900174144 checkpoint_manager.py:1458] Waiting for previous save to complete took 42.812620 seconds. If this number is high, consider checkpointing less frequently. I0424 15:32:45.841666 133599900174144 checkpoint_manager.py:1518] [process=4] Saving checkpoint at step 9 I0424 15:32:45.845061 133599900174144 async_checkpointer.py:452] [process=4] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/linen_xpk_feat_nnx_set_defaults_true_20260424_145744_09_pdb_lt_1/checkpoints/9. I0424 15:32:46.642066 133599900174144 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0424 15:32:46.642166 133599900174144 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 I0424 15:32:46.684822 133599900174144 base_pytree_checkpoint_handler.py:153] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.043889s I0424 15:32:46.685003 133599900174144 base_pytree_checkpoint_handler.py:129] [process=4] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.099 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.3025362491607666 s) (per-host) I0424 15:32:46.685054 133599900174144 base_pytree_checkpoint_handler.py:737] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.302598s (batch_requests_ready=0.256836s, total_serialization_initiated=0.045693s, others=0.000069s) I0424 15:32:46.685142 133599900174144 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.306696s (all_items=0.000014s, per_item={'items': '0.00001431'}, temp_paths=0.306682) I0424 15:32:46.686056 133473742345984 async_checkpointer.py:79] [process=4][thread=async_save] Background save thread started. I0424 15:32:46.686219 133599900174144 async_checkpointer.py:561] Finished blocking save. Time taken: 0.844481s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/linen_xpk_feat_nnx_set_defaults_true_20260424_145744_09_pdb_lt_1/checkpoints/9. I0424 15:32:46.691596 133599900174144 checkpoint_manager.py:1566] [process=4][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0424 15:32:46.691863 133473759131392 async_checkpointer.py:265] [process=4][thread=save_finalize] Waiting for background save thread=async_save. I0424 15:32:46.692040 133599900174144 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/linen_xpk_feat_nnx_set_defaults_true_20260424_145744_09_pdb_lt_1/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777044723.0270126, 'wait_for_prev_duration_secs': 42.81261968612671, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777044765.8417056, 'checkpointer_blocking_duration_secs': 0.8446230888366699, 'get_old_steps_start_time': 1777044766.6863463, 'get_old_steps_duration_secs': 2.8133392333984375e-05, 'checkpoint_manager_blocking_start_time': 1777044723.025293, 'checkpoint_manager_blocking_duration_secs': 43.666709899902344} I0424 15:32:46.692147 133599900174144 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0424 15:32:46.692193 133599900174144 checkpoint_manager.py:2024] [process=4][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0424 15:32:51.440577 133476946786048 array_metadata_store.py:203] [process=4][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/linen_xpk_feat_nnx_set_defaults_true_20260424_145744_09_pdb_lt_1/checkpoints/9/items/array_metadatas/process_4 I0424 15:33:28.473754 133473742345984 base_pytree_checkpoint_handler.py:129] [process=4] /jax/checkpoint/write/gbytes_per_sec: 37.528 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.09125590324402 s) (per-host) I0424 15:33:28.473858 133473742345984 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.787688s. I0424 15:33:34.584684 133473742345984 async_checkpointer.py:144] [process=4][thread=async_save] Background save thread done. Time taken: 47.898499s. I0424 15:33:34.584923 133473759131392 async_checkpointer.py:273] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save. I0424 15:33:34.584990 133473759131392 async_checkpointer.py:283] [process=4][thread=save_finalize] No errors found in background save thread=async_save. I0424 15:33:34.585043 133473759131392 checkpoint_manager.py:2133] [process=4][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0424 15:33:34.586557 133473759131392 checkpoint_manager.py:2142] [process=4][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0424 15:33:34.586715 133599900174144 checkpoint_manager.py:2036] [process=4][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0424 15:33:34.586862 133599900174144 checkpoint_manager.py:2013] [process=4][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0424 15:33:34.587544 133599900174144 metric_logger.py:185] completed step: 9, seconds: 0.058, TFLOP/s/device: 58.171, Tokens/s/device: 8768.174, total_weights: 16384, loss: 5.892 Per train step: Total TFLOPs: 3.40 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Fri Apr 24 15:33:44 UTC 2026 EXIT_CODE=0