XPK Start: Sun Apr 19 09:17:51 UTC 2026 2026-04-19 09:17:55.580492: 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 00:00:1776590275.593174 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1776590275.596883 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1776590275.608109 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776590275.608129 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776590275.608131 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776590275.608133 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-19 09:18:14.864417: 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) I0419 09:18:15.381211 138774187362112 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-19 09:18:24,422:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0419 09:18:24.422784 138774187362112 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-19 09:18:24,425:jax._src.distributed:157: Connecting to JAX distributed service on mt-08-checkpoint-async--5pdqr-slice-job-0-0.mt-08-checkpoint-async--5pdqr:8482 I0419 09:18:24.425141 138774187362112 distributed.py:157] Connecting to JAX distributed service on mt-08-checkpoint-async--5pdqr-slice-job-0-0.mt-08-checkpoint-async--5pdqr:8482 I0419 09:18:26.391144 138774187362112 max_utils.py:284] Jax distributed system initialized! I0419 09:18:33.301984 138774187362112 max_utils.py:800] System Information: Jax Version: 0.8.1 I0419 09:18:33.302090 138774187362112 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0419 09:18:33.302130 138774187362112 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 I0419 09:18:33.302166 138774187362112 train_utils.py:347] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0419 09:18:33.925184 138774187362112 maxtext_utils.py:1398] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0419 09:18:33.925485 138774187362112 checkpointing.py:677] Setting up checkpoint logger... I0419 09:18:33.925556 138774187362112 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True I0419 09:18:33.925601 138774187362112 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0419 09:18:33.925954 138774187362112 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 0x7e3658d74f50>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0419 09:18:37.282261 138774187362112 checkpointing.py:265] Enabling policy for fixed interval checkpointing. I0419 09:18:37.282695 138774187362112 checkpoint_manager.py:709] [process=7][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e22e7ae0140>}, handler_registry=None I0419 09:18:37.282935 138774187362112 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e22e7ae0140>` 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`. I0419 09:18:37.282985 138774187362112 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e22e7ae3b60>` 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`. I0419 09:18:37.283024 138774187362112 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 0x7e22e7ae0140>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e22e7ae0140>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e22e7ae3b60>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e22e7ae3b60>}). I0419 09:18:37.285730 138774187362112 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0419 09:18:37.285815 138774187362112 async_checkpointer.py:177] [process=7][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7e36419fb2e0> timeout: 600 secs and primary_host=0 for async checkpoint writes I0419 09:18:38.379028 138774187362112 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260419_085454/linen_xpk_test_pipeline_scan_nnx_20260419_085454_08_checkpoint_async_true/checkpoints I0419 09:18:39.280219 138774187362112 checkpoint_manager.py:929] [process=7][thread=MainThread] CheckpointManager created, primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=5), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260419_085454/linen_xpk_test_pipeline_scan_nnx_20260419_085454_08_checkpoint_async_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7e23b0a27470> I0419 09:18:39.280396 138774187362112 checkpointing.py:301] Checkpoint manager created! I0419 09:18:40.166607 138774187362112 nnx_wrappers.py:453] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0419 09:18:40.166716 138774187362112 nnx_wrappers.py:453] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0419 09:18:40.538551 138774187362112 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0419 09:18:40.538651 138774187362112 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0419 09:18:40.554294 138774187362112 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0419 09:18:40.554354 138774187362112 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0419 09:18:40.577518 138774187362112 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 09:18:40.577603 138774187362112 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 09:18:40.593388 138774187362112 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 09:18:40.593450 138774187362112 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 09:18:40.609139 138774187362112 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0419 09:18:40.609199 138774187362112 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 09:18:40.632775 138774187362112 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0419 09:18:40.632846 138774187362112 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 09:18:40.652203 138774187362112 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0419 09:18:40.652269 138774187362112 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0419 09:18:40.840752 138774187362112 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint I0419 09:18:40.840863 138774187362112 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0419 09:18:42.263072 138774187362112 maxtext_utils.py:1501] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: PartitionSpec('norm',) Physical: (None,) I0419 09:18:42.263211 138774187362112 maxtext_utils.py:1501] params/params/decoder/layers/mlp/wi_0/kernel Shape: float32[2048,16,7168] Logical: PartitionSpec('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0419 09:18:42.263267 138774187362112 maxtext_utils.py:1501] params/params/decoder/layers/mlp/wi_1/kernel Shape: float32[2048,16,7168] Logical: PartitionSpec('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0419 09:18:42.263327 138774187362112 maxtext_utils.py:1501] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: PartitionSpec('mlp', 'layers', 'embed') Physical: (None, None, 'fsdp') I0419 09:18:42.263379 138774187362112 maxtext_utils.py:1501] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: PartitionSpec('norm', 'layers') Physical: (None, None) I0419 09:18:42.263425 138774187362112 maxtext_utils.py:1501] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: PartitionSpec('norm', 'layers') Physical: (None, None) I0419 09:18:42.263480 138774187362112 maxtext_utils.py:1501] 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) I0419 09:18:42.263546 138774187362112 maxtext_utils.py:1501] 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') I0419 09:18:42.263590 138774187362112 maxtext_utils.py:1501] 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) I0419 09:18:42.263628 138774187362112 maxtext_utils.py:1501] 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) I0419 09:18:42.263676 138774187362112 maxtext_utils.py:1501] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: PartitionSpec('embed', 'vocab') Physical: ('fsdp', None) I0419 09:18:42.263723 138774187362112 maxtext_utils.py:1501] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: PartitionSpec('vocab', 'embed') Physical: (None, 'fsdp') I0419 09:18:42.760652 138774187362112 train.py:156] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0419 09:18:42.760757 138774187362112 train.py:156] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0419 09:18:42.775651 138774187362112 train.py:163] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0419 09:18:42.775715 138774187362112 train.py:163] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0419 09:18:53.406148 138774187362112 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. I0419 09:18:53.407013 138774187362112 metric_logger.py:289] number parameters: 1.104 billion I0419 09:19:04.432380 138774187362112 checkpointing.py:772] Waiting for step 0 to finish before checkpoint... I0419 09:19:04.625806 138774187362112 checkpointing.py:776] Waited 0.19340920448303223 seconds for step 0 to finish before starting checkpointing. I0419 09:19:04.628199 138774187362112 checkpoint_manager.py:2013] [process=7][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0419 09:19:04.629876 138774187362112 checkpoint_manager.py:1518] [process=7] Saving checkpoint at step 0 I0419 09:19:04.632600 138774187362112 async_checkpointer.py:452] [process=7] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260419_085454/linen_xpk_test_pipeline_scan_nnx_20260419_085454_08_checkpoint_async_true/checkpoints/0. I0419 09:19:05.351242 138774187362112 signaling_client.py:364] Using JaxDistributedSignalingClient I0419 09:19:05.352306 138774187362112 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0419 09:19:05.352364 138774187362112 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 I0419 09:19:05.601070 138774187362112 base_pytree_checkpoint_handler.py:153] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.249908s I0419 09:19:05.601238 138774187362112 base_pytree_checkpoint_handler.py:129] [process=7] /jax/checkpoint/write/blocking_gbytes_per_sec: 3.043 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.506939172744751 s) (per-host) I0419 09:19:05.601287 138774187362112 base_pytree_checkpoint_handler.py:737] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.506998s (batch_requests_ready=0.253549s, total_serialization_initiated=0.253381s, others=0.000067s) I0419 09:19:05.601374 138774187362112 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.511173s (all_items=0.000019s, per_item={'items': '0.00001884'}, temp_paths=0.511154) I0419 09:19:05.602393 138647827367680 async_checkpointer.py:79] [process=7][thread=async_save] Background save thread started. I0419 09:19:05.602583 138774187362112 async_checkpointer.py:561] Finished blocking save. Time taken: 0.972632s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260419_085454/linen_xpk_test_pipeline_scan_nnx_20260419_085454_08_checkpoint_async_true/checkpoints/0. I0419 09:19:05.654357 138774187362112 checkpoint_manager.py:1566] [process=7][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0419 09:19:05.654715 138645705062144 async_checkpointer.py:265] [process=7][thread=save_finalize] Waiting for background save thread=async_save. I0419 09:19:05.654866 138774187362112 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260419_085454/linen_xpk_test_pipeline_scan_nnx_20260419_085454_08_checkpoint_async_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776590344.6281803, 'wait_for_prev_duration_secs': 6.365776062011719e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776590344.6299162, 'checkpointer_blocking_duration_secs': 0.9727756977081299, 'get_old_steps_start_time': 1776590345.6027172, 'get_old_steps_duration_secs': 2.6226043701171875e-05, 'checkpoint_manager_blocking_start_time': 1776590344.6263824, 'checkpoint_manager_blocking_duration_secs': 1.028442621231079} I0419 09:19:05.654974 138774187362112 checkpointing.py:408] Started an asynchronous checkpoint save for step 0 I0419 09:19:05.655028 138774187362112 max_utils.py:750] Memstats: After params initialized: I0419 09:19:05.655081 138774187362112 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_26(process=7,(2,6,0,0)) I0419 09:19:05.655116 138774187362112 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_27(process=7,(3,6,0,0)) I0419 09:19:05.655145 138774187362112 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_30(process=7,(2,7,0,0)) I0419 09:19:05.655172 138774187362112 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_31(process=7,(3,7,0,0)) I0419 09:19:05.921227 138774187362112 metric_logger.py:185] completed step: 0, seconds: 11.025, TFLOP/s/device: 1.232, Tokens/s/device: 185.756, total_weights: 65536, loss: 10.874 I0419 09:19:06.086421 138774187362112 metric_logger.py:185] completed step: 1, seconds: 1.488, TFLOP/s/device: 9.134, Tokens/s/device: 1376.742, total_weights: 65536, loss: 10.874 I0419 09:19:06.489344 138774187362112 metric_logger.py:185] completed step: 2, seconds: 0.018, TFLOP/s/device: 757.618, Tokens/s/device: 114196.498, total_weights: 65536, loss: 10.560 I0419 09:19:06.636853 138774187362112 metric_logger.py:185] completed step: 3, seconds: 0.403, TFLOP/s/device: 33.687, Tokens/s/device: 5077.652, total_weights: 65536, loss: 9.980 I0419 09:19:06.641601 138774187362112 checkpointing.py:772] Waiting for step 5 to finish before checkpoint... I0419 09:19:06.931609 138774187362112 checkpointing.py:776] Waited 0.28993749618530273 seconds for step 5 to finish before starting checkpointing. I0419 09:19:06.934385 138774187362112 checkpoint_manager.py:2024] [process=7][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 09:19:08.655205 2606 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0419 09:19:11.424631 138647835760384 array_metadata_store.py:203] [process=7][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260419_085454/linen_xpk_test_pipeline_scan_nnx_20260419_085454_08_checkpoint_async_true/checkpoints/0/items/array_metadatas/process_7 I0419 09:19:41.856361 138647827367680 base_pytree_checkpoint_handler.py:129] [process=7] /jax/checkpoint/write/gbytes_per_sec: 42.968 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 36.762017011642456 s) (per-host) I0419 09:19:41.856466 138647827367680 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 36.253959s. I0419 09:19:50.162202 138647827367680 async_checkpointer.py:144] [process=7][thread=async_save] Background save thread done. Time taken: 44.559680s. I0419 09:19:50.162499 138645705062144 async_checkpointer.py:273] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 09:19:50.162636 138645705062144 async_checkpointer.py:283] [process=7][thread=save_finalize] No errors found in background save thread=async_save. I0419 09:19:50.162687 138645705062144 checkpoint_manager.py:2133] [process=7][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0419 09:19:50.164484 138645705062144 checkpoint_manager.py:2142] [process=7][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0419 09:19:50.164674 138774187362112 checkpoint_manager.py:2036] [process=7][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0419 09:19:50.164822 138774187362112 checkpoint_manager.py:1458] Waiting for previous save to complete took 43.230440 seconds. If this number is high, consider checkpointing less frequently. I0419 09:19:50.166935 138774187362112 checkpoint_manager.py:1518] [process=7] Saving checkpoint at step 5 I0419 09:19:50.170993 138774187362112 async_checkpointer.py:452] [process=7] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260419_085454/linen_xpk_test_pipeline_scan_nnx_20260419_085454_08_checkpoint_async_true/checkpoints/5. I0419 09:19:51.139210 138774187362112 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0419 09:19:51.139314 138774187362112 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 I0419 09:19:51.183823 138774187362112 base_pytree_checkpoint_handler.py:153] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.046571s I0419 09:19:51.183995 138774187362112 base_pytree_checkpoint_handler.py:129] [process=7] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.095 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.30274200439453125 s) (per-host) I0419 09:19:51.184049 138774187362112 base_pytree_checkpoint_handler.py:737] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.302806s (batch_requests_ready=0.254353s, total_serialization_initiated=0.048382s, others=0.000071s) I0419 09:19:51.184144 138774187362112 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.307840s (all_items=0.000017s, per_item={'items': '0.00001740'}, temp_paths=0.307822) I0419 09:19:51.185102 138646729324288 async_checkpointer.py:79] [process=7][thread=async_save] Background save thread started. I0419 09:19:51.185266 138774187362112 async_checkpointer.py:561] Finished blocking save. Time taken: 1.018256s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260419_085454/linen_xpk_test_pipeline_scan_nnx_20260419_085454_08_checkpoint_async_true/checkpoints/5. I0419 09:19:51.191448 138774187362112 checkpoint_manager.py:1566] [process=7][thread=MainThread][step=5] Starting CheckpointManager Save Finalize thread=save_finalize I0419 09:19:51.191739 138645705062144 async_checkpointer.py:265] [process=7][thread=save_finalize] Waiting for background save thread=async_save. I0419 09:19:51.191901 138774187362112 standard_logger.py:34] {'step': 5, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260419_085454/linen_xpk_test_pipeline_scan_nnx_20260419_085454_08_checkpoint_async_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776590346.934357, 'wait_for_prev_duration_secs': 43.23043966293335, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776590390.166976, 'checkpointer_blocking_duration_secs': 1.0183982849121094, 'get_old_steps_start_time': 1776590391.1853926, 'get_old_steps_duration_secs': 2.7179718017578125e-05, 'checkpoint_manager_blocking_start_time': 1776590346.9322252, 'checkpoint_manager_blocking_duration_secs': 44.25964283943176} I0419 09:19:51.192006 138774187362112 checkpointing.py:408] Started an asynchronous checkpoint save for step 5 I0419 09:19:51.192580 138774187362112 metric_logger.py:185] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.838, Tokens/s/device: 13390.697, total_weights: 65536, loss: 9.460 I0419 09:19:51.198751 138774187362112 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.227, Tokens/s/device: 13901.522, total_weights: 65536, loss: 8.959 I0419 09:20:10.539910 138774187362112 metric_logger.py:185] completed step: 6, seconds: 44.556, TFLOP/s/device: 0.305, Tokens/s/device: 45.965, total_weights: 65536, loss: 8.469 I0419 09:20:10.687325 138774187362112 metric_logger.py:185] completed step: 7, seconds: 19.195, TFLOP/s/device: 0.708, Tokens/s/device: 106.695, total_weights: 65536, loss: 8.003 I0419 09:20:10.834826 138774187362112 metric_logger.py:185] completed step: 8, seconds: 0.152, TFLOP/s/device: 89.228, Tokens/s/device: 13449.352, total_weights: 65536, loss: 7.572 I0419 09:20:10.981855 138774187362112 checkpointing.py:772] Waiting for step 9 to finish before checkpoint... I0419 09:20:10.982508 138774187362112 checkpointing.py:776] Waited 0.0006730556488037109 seconds for step 9 to finish before starting checkpointing. I0419 09:20:10.984487 138774187362112 checkpoint_manager.py:2024] [process=7][thread=MainThread][step=5][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 09:20:15.846295 138647835760384 array_metadata_store.py:203] [process=7][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260419_085454/linen_xpk_test_pipeline_scan_nnx_20260419_085454_08_checkpoint_async_true/checkpoints/5/items/array_metadatas/process_7 I0419 09:20:51.597167 138646729324288 base_pytree_checkpoint_handler.py:129] [process=7] /jax/checkpoint/write/gbytes_per_sec: 26.016 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 60.71587371826172 s) (per-host) I0419 09:20:51.597285 138646729324288 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 60.412072s. I0419 09:20:59.919985 138646729324288 async_checkpointer.py:144] [process=7][thread=async_save] Background save thread done. Time taken: 68.734751s. I0419 09:20:59.920287 138645705062144 async_checkpointer.py:273] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 09:20:59.920424 138645705062144 async_checkpointer.py:283] [process=7][thread=save_finalize] No errors found in background save thread=async_save. I0419 09:20:59.920500 138645705062144 checkpoint_manager.py:2133] [process=7][thread=save_finalize][step=5] CheckpointManager Save Finalize is syncing with other hosts... I0419 09:20:59.922063 138645705062144 checkpoint_manager.py:2142] [process=7][thread=save_finalize][step=5] CheckpointManager Save Finalize is done on all hosts. I0419 09:20:59.922274 138774187362112 checkpoint_manager.py:2036] [process=7][thread=MainThread][step=5][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=5. W0419 09:20:59.922413 138774187362112 checkpoint_manager.py:1458] Waiting for previous save to complete took 48.937943 seconds. If this number is high, consider checkpointing less frequently. I0419 09:20:59.924171 138774187362112 checkpoint_manager.py:1518] [process=7] Saving checkpoint at step 9 I0419 09:20:59.927694 138774187362112 async_checkpointer.py:452] [process=7] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260419_085454/linen_xpk_test_pipeline_scan_nnx_20260419_085454_08_checkpoint_async_true/checkpoints/9. I0419 09:21:00.926526 138774187362112 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0419 09:21:00.926632 138774187362112 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 I0419 09:21:00.974574 138774187362112 base_pytree_checkpoint_handler.py:153] [process=7][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.049221s I0419 09:21:00.974743 138774187362112 base_pytree_checkpoint_handler.py:129] [process=7] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.084 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.30343127250671387 s) (per-host) I0419 09:21:00.974792 138774187362112 base_pytree_checkpoint_handler.py:737] [process=7][thread=MainThread] Initiated Pytree async_save. Time taken: 0.303491s (batch_requests_ready=0.252215s, total_serialization_initiated=0.051211s, others=0.000065s) I0419 09:21:00.974886 138774187362112 composite_checkpoint_handler.py:715] [process=7][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.307670s (all_items=0.000011s, per_item={'items': '0.00001144'}, temp_paths=0.307658) I0419 09:21:00.975740 138647827367680 async_checkpointer.py:79] [process=7][thread=async_save] Background save thread started. I0419 09:21:00.975906 138774187362112 async_checkpointer.py:561] Finished blocking save. Time taken: 1.051664s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260419_085454/linen_xpk_test_pipeline_scan_nnx_20260419_085454_08_checkpoint_async_true/checkpoints/9. I0419 09:21:01.006783 138774187362112 checkpoint_manager.py:1566] [process=7][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0419 09:21:01.007081 138645705062144 async_checkpointer.py:265] [process=7][thread=save_finalize] Waiting for background save thread=async_save. I0419 09:21:01.007242 138774187362112 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260419_085454/linen_xpk_test_pipeline_scan_nnx_20260419_085454_08_checkpoint_async_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776590410.984453, 'wait_for_prev_duration_secs': 48.93794345855713, 'time_between_consecutive_saves_sec': 20.819929361343384, 'checkpointer_blocking_start_time': 1776590459.9242117, 'checkpointer_blocking_duration_secs': 1.0518016815185547, 'get_old_steps_start_time': 1776590460.976025, 'get_old_steps_duration_secs': 2.0265579223632812e-05, 'checkpoint_manager_blocking_start_time': 1776590410.9827461, 'checkpoint_manager_blocking_duration_secs': 50.024465799331665} I0419 09:21:01.007346 138774187362112 checkpointing.py:408] Started an asynchronous checkpoint save for step 9 I0419 09:21:01.007388 138774187362112 checkpoint_manager.py:2024] [process=7][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 09:21:06.198568 138647835760384 array_metadata_store.py:203] [process=7][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260419_085454/linen_xpk_test_pipeline_scan_nnx_20260419_085454_08_checkpoint_async_true/checkpoints/9/items/array_metadatas/process_7 I0419 09:21:42.361043 138647827367680 base_pytree_checkpoint_handler.py:129] [process=7] /jax/checkpoint/write/gbytes_per_sec: 37.889 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.689697265625 s) (per-host) I0419 09:21:42.361147 138647827367680 async_checkpointer.py:90] [process=7][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.385288s. I0419 09:21:51.361100 138647827367680 async_checkpointer.py:144] [process=7][thread=async_save] Background save thread done. Time taken: 50.385225s. I0419 09:21:51.361377 138645705062144 async_checkpointer.py:273] [process=7][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 09:21:51.361501 138645705062144 async_checkpointer.py:283] [process=7][thread=save_finalize] No errors found in background save thread=async_save. I0419 09:21:51.361568 138645705062144 checkpoint_manager.py:2133] [process=7][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0419 09:21:51.362972 138645705062144 checkpoint_manager.py:2142] [process=7][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0419 09:21:51.363132 138774187362112 checkpoint_manager.py:2036] [process=7][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0419 09:21:51.363278 138774187362112 checkpoint_manager.py:2013] [process=7][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0419 09:21:51.364021 138774187362112 metric_logger.py:185] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.136, Tokens/s/device: 13887.759, total_weights: 65536, loss: 7.181 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Sun Apr 19 09:22:01 UTC 2026 EXIT_CODE=0