XPK Start: Sat Apr 25 17:03:35 UTC 2026 2026-04-25 17:03:39.689535: 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 00:00:1777136619.702076 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1777136619.705788 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1777136619.716905 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1777136619.716923 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1777136619.716926 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1777136619.716928 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-25 17:03:58.816442: 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) I0425 17:03:59.349364 136762627626816 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-25 17:04:08,390:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0425 17:04:08.390413 136762627626816 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-25 17:04:08,392:jax._src.distributed:157: Connecting to JAX distributed service on mt-10-shardy-false-9llv1-slice-job-0-0.mt-10-shardy-false-9llv1:8482 I0425 17:04:08.392697 136762627626816 distributed.py:157] Connecting to JAX distributed service on mt-10-shardy-false-9llv1-slice-job-0-0.mt-10-shardy-false-9llv1:8482 I0425 17:04:09.529534 136762627626816 max_utils.py:284] Jax distributed system initialized! I0425 17:04:16.570533 136762627626816 max_utils.py:800] System Information: Jax Version: 0.8.1 I0425 17:04:16.570639 136762627626816 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0425 17:04:16.570701 136762627626816 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 I0425 17:04:16.570737 136762627626816 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0425 17:04:17.192847 136762627626816 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0425 17:04:17.307352 136762627626816 checkpointing.py:688] Setting up checkpoint logger... I0425 17:04:17.307471 136762627626816 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0425 17:04:17.307517 136762627626816 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0425 17:04:17.307734 136762627626816 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 0x7c62037791f0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0425 17:04:20.222147 136762627626816 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0425 17:04:20.222451 136762627626816 checkpoint_manager.py:709] [process=5][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c4d7040dc70>}, handler_registry=None I0425 17:04:20.222703 136762627626816 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c4d7040dc70>` 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`. I0425 17:04:20.222754 136762627626816 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c4d7028ff50>` 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`. I0425 17:04:20.222790 136762627626816 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 0x7c4d7040dc70>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c4d7040dc70>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c4d7028ff50>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c4d7028ff50>}). I0425 17:04:20.223125 136762627626816 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0425 17:04:20.223207 136762627626816 async_checkpointer.py:177] [process=5][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7c4d70208f40> timeout: 600 secs and primary_host=0 for async checkpoint writes I0425 17:04:21.396381 136762627626816 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_10_shardy_false/checkpoints I0425 17:04:21.499697 136762627626816 checkpoint_manager.py:929] [process=5][thread=MainThread] CheckpointManager created, primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=10), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_10_shardy_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7c4da84a3d10> I0425 17:04:21.499870 136762627626816 checkpointing.py:302] Checkpoint manager created! I0425 17:04:22.426495 136762627626816 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0425 17:04:22.426618 136762627626816 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0425 17:04:23.789947 136762627626816 maxtext_utils.py:1749] decoder/decoder_norm/scale/value Shape: float32[2048] Physical: (None,) I0425 17:04:23.790070 136762627626816 maxtext_utils.py:1749] decoder/layers/mlp/wi_0/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0425 17:04:23.790124 136762627626816 maxtext_utils.py:1749] decoder/layers/mlp/wi_1/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0425 17:04:23.790168 136762627626816 maxtext_utils.py:1749] decoder/layers/mlp/wo/kernel/value Shape: float32[7168,16,2048] Physical: (None, None, 'fsdp') I0425 17:04:23.790210 136762627626816 maxtext_utils.py:1749] decoder/layers/post_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0425 17:04:23.790249 136762627626816 maxtext_utils.py:1749] decoder/layers/pre_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0425 17:04:23.790287 136762627626816 maxtext_utils.py:1749] decoder/layers/self_attention/key/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0425 17:04:23.790325 136762627626816 maxtext_utils.py:1749] decoder/layers/self_attention/out/kernel/value Shape: float32[16,16,128,2048] Physical: (None, None, None, 'fsdp') I0425 17:04:23.790360 136762627626816 maxtext_utils.py:1749] decoder/layers/self_attention/query/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0425 17:04:23.790394 136762627626816 maxtext_utils.py:1749] decoder/layers/self_attention/value/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0425 17:04:23.790428 136762627626816 maxtext_utils.py:1749] decoder/logits_dense/kernel/value Shape: float32[2048,32000] Physical: ('fsdp', None) I0425 17:04:23.790462 136762627626816 maxtext_utils.py:1749] token_embedder/embedding/value Shape: float32[32000,2048] Physical: (None, 'fsdp') I0425 17:04:24.188674 136762627626816 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0425 17:04:24.188821 136762627626816 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0425 17:04:24.194222 136762627626816 nnx_decoders.py:616] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0425 17:04:24.194319 136762627626816 nnx_decoders.py:616] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0425 17:04:24.345055 136762627626816 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0425 17:04:24.345187 136762627626816 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0425 17:04:24.494246 136762627626816 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0425 17:04:24.494378 136762627626816 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0425 17:04:24.652307 136762627626816 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0425 17:04:24.652444 136762627626816 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0425 17:04:24.795077 136762627626816 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0425 17:04:24.795214 136762627626816 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0425 17:04:24.935353 136762627626816 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0425 17:04:24.935485 136762627626816 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0425 17:04:25.086344 136762627626816 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0425 17:04:25.086481 136762627626816 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0425 17:04:25.109218 136762627626816 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0425 17:04:25.109313 136762627626816 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0425 17:04:36.314705 136762627626816 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. I0425 17:04:36.317127 136762627626816 metric_logger.py:289] number parameters: 1.104 billion I0425 17:04:47.437276 136762627626816 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0425 17:04:47.591905 136762627626816 checkpointing.py:798] Waited 0.1546187400817871 seconds for step 0 to finish before starting checkpointing. I0425 17:04:47.594542 136762627626816 checkpoint_manager.py:2013] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0425 17:04:47.596173 136762627626816 checkpoint_manager.py:1518] [process=5] Saving checkpoint at step 0 I0425 17:04:47.598865 136762627626816 async_checkpointer.py:452] [process=5] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_10_shardy_false/checkpoints/0. I0425 17:04:48.613964 136762627626816 signaling_client.py:364] Using JaxDistributedSignalingClient I0425 17:04:48.615095 136762627626816 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0425 17:04:48.615153 136762627626816 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 I0425 17:04:48.865868 136762627626816 base_pytree_checkpoint_handler.py:153] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.253016s I0425 17:04:48.866052 136762627626816 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.983 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5171868801116943 s) (per-host) I0425 17:04:48.866114 136762627626816 base_pytree_checkpoint_handler.py:737] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.517259s (batch_requests_ready=0.257130s, total_serialization_initiated=0.260048s, others=0.000081s) I0425 17:04:48.866213 136762627626816 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.521233s (all_items=0.000019s, per_item={'items': '0.00001931'}, temp_paths=0.521214) I0425 17:04:48.867229 136640441849600 async_checkpointer.py:79] [process=5][thread=async_save] Background save thread started. I0425 17:04:48.867403 136762627626816 async_checkpointer.py:561] Finished blocking save. Time taken: 1.271152s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_10_shardy_false/checkpoints/0. I0425 17:04:49.250418 136762627626816 checkpoint_manager.py:1566] [process=5][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0425 17:04:49.250841 136635635189504 async_checkpointer.py:265] [process=5][thread=save_finalize] Waiting for background save thread=async_save. I0425 17:04:49.251009 136762627626816 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_10_shardy_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777136687.5945244, 'wait_for_prev_duration_secs': 6.29425048828125e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777136687.5962179, 'checkpointer_blocking_duration_secs': 1.2712953090667725, 'get_old_steps_start_time': 1777136688.8675323, 'get_old_steps_duration_secs': 2.6702880859375e-05, 'checkpoint_manager_blocking_start_time': 1777136687.592359, 'checkpoint_manager_blocking_duration_secs': 1.6586089134216309} I0425 17:04:49.251123 136762627626816 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0425 17:04:49.251200 136762627626816 max_utils.py:750] Memstats: After params initialized: I0425 17:04:49.251255 136762627626816 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_18(process=5,(2,4,0,0)) I0425 17:04:49.251295 136762627626816 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_19(process=5,(3,4,0,0)) I0425 17:04:49.251323 136762627626816 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_22(process=5,(2,5,0,0)) I0425 17:04:49.251348 136762627626816 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_23(process=5,(3,5,0,0)) I0425 17:04:49.517400 136762627626816 metric_logger.py:185] completed step: 0, seconds: 11.119, TFLOP/s/device: 1.222, Tokens/s/device: 184.188, total_weights: 65536, loss: 10.874 I0425 17:04:49.698184 136762627626816 metric_logger.py:185] completed step: 1, seconds: 2.078, TFLOP/s/device: 6.538, Tokens/s/device: 985.411, total_weights: 65536, loss: 10.874 I0425 17:04:50.157299 136762627626816 metric_logger.py:185] completed step: 2, seconds: 0.035, TFLOP/s/device: 383.546, Tokens/s/device: 57812.279, total_weights: 65536, loss: 10.262 I0425 17:04:50.304844 136762627626816 metric_logger.py:185] completed step: 3, seconds: 0.450, TFLOP/s/device: 30.186, Tokens/s/device: 4549.918, total_weights: 65536, loss: 9.732 I0425 17:04:50.600497 136762627626816 metric_logger.py:185] completed step: 4, seconds: 0.162, TFLOP/s/device: 83.678, Tokens/s/device: 12612.857, total_weights: 65536, loss: 9.274 I0425 17:04:50.608733 136762627626816 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.281, Tokens/s/device: 13909.547, total_weights: 65536, loss: 8.888 I0425 17:05:10.118607 136762627626816 metric_logger.py:185] completed step: 6, seconds: 0.297, TFLOP/s/device: 45.773, Tokens/s/device: 6899.456, total_weights: 65536, loss: 8.589 I0425 17:05:10.265855 136762627626816 metric_logger.py:185] completed step: 7, seconds: 19.364, TFLOP/s/device: 0.702, Tokens/s/device: 105.764, total_weights: 65536, loss: 8.380 I0425 17:05:10.413322 136762627626816 metric_logger.py:185] completed step: 8, seconds: 0.154, TFLOP/s/device: 88.066, Tokens/s/device: 13274.308, total_weights: 65536, loss: 8.251 I0425 17:05:10.561825 136762627626816 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0425 17:05:10.562432 136762627626816 checkpointing.py:798] Waited 0.0006201267242431641 seconds for step 9 to finish before starting checkpointing. I0425 17:05:10.564341 136762627626816 checkpoint_manager.py:2024] [process=5][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0425 17:05:11.027344 2617 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0425 17:05:12.972167 136640450242304 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_10_shardy_false/checkpoints/0/items/array_metadatas/process_5 I0425 17:05:45.309754 136640441849600 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/gbytes_per_sec: 27.731 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 56.960835695266724 s) (per-host) I0425 17:05:45.309863 136640441849600 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 56.442521s. I0425 17:05:53.911672 136640441849600 async_checkpointer.py:144] [process=5][thread=async_save] Background save thread done. Time taken: 65.044301s. I0425 17:05:53.911965 136635635189504 async_checkpointer.py:273] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save. I0425 17:05:53.912115 136635635189504 async_checkpointer.py:283] [process=5][thread=save_finalize] No errors found in background save thread=async_save. I0425 17:05:53.912167 136635635189504 checkpoint_manager.py:2133] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0425 17:05:53.914418 136635635189504 checkpoint_manager.py:2142] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0425 17:05:53.914615 136762627626816 checkpoint_manager.py:2036] [process=5][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0425 17:05:53.914786 136762627626816 checkpoint_manager.py:1458] Waiting for previous save to complete took 43.350449 seconds. If this number is high, consider checkpointing less frequently. I0425 17:05:53.916532 136762627626816 checkpoint_manager.py:1518] [process=5] Saving checkpoint at step 9 I0425 17:05:53.920085 136762627626816 async_checkpointer.py:452] [process=5] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_10_shardy_false/checkpoints/9. I0425 17:05:54.476629 136762627626816 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0425 17:05:54.476806 136762627626816 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 I0425 17:05:54.526833 136762627626816 base_pytree_checkpoint_handler.py:153] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.052276s I0425 17:05:54.526998 136762627626816 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/blocking_gbytes_per_sec: 4.973 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.31020331382751465 s) (per-host) I0425 17:05:54.527054 136762627626816 base_pytree_checkpoint_handler.py:737] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.310267s (batch_requests_ready=0.253230s, total_serialization_initiated=0.056966s, others=0.000071s) I0425 17:05:54.527147 136762627626816 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.314176s (all_items=0.000015s, per_item={'items': '0.00001526'}, temp_paths=0.314160) I0425 17:05:54.527983 136636725716736 async_checkpointer.py:79] [process=5][thread=async_save] Background save thread started. I0425 17:05:54.528141 136762627626816 async_checkpointer.py:561] Finished blocking save. Time taken: 0.611537s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_10_shardy_false/checkpoints/9. I0425 17:05:54.537451 136762627626816 checkpoint_manager.py:1566] [process=5][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0425 17:05:54.537700 136635635189504 async_checkpointer.py:265] [process=5][thread=save_finalize] Waiting for background save thread=async_save. I0425 17:05:54.537826 136762627626816 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_10_shardy_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777136710.564314, 'wait_for_prev_duration_secs': 43.35044884681702, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777136753.9165707, 'checkpointer_blocking_duration_secs': 0.6116831302642822, 'get_old_steps_start_time': 1777136754.5282724, 'get_old_steps_duration_secs': 3.218650817871094e-05, 'checkpoint_manager_blocking_start_time': 1777136710.5626693, 'checkpoint_manager_blocking_duration_secs': 43.97512125968933} I0425 17:05:54.537937 136762627626816 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0425 17:05:54.537982 136762627626816 checkpoint_manager.py:2024] [process=5][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0425 17:05:59.455670 136640450242304 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260425_150412/nnx_xpk_feat_nnx_set_defaults_true_20260425_150412_10_shardy_false/checkpoints/9/items/array_metadatas/process_5 I0425 17:06:36.679471 136636725716736 base_pytree_checkpoint_handler.py:129] [process=5] /jax/checkpoint/write/gbytes_per_sec: 37.200 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.46263670921326 s) (per-host) I0425 17:06:36.679588 136636725716736 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.151495s. I0425 17:06:43.665688 136636725716736 async_checkpointer.py:144] [process=5][thread=async_save] Background save thread done. Time taken: 49.137578s. I0425 17:06:43.665907 136635635189504 async_checkpointer.py:273] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save. I0425 17:06:43.665964 136635635189504 async_checkpointer.py:283] [process=5][thread=save_finalize] No errors found in background save thread=async_save. I0425 17:06:43.666018 136635635189504 checkpoint_manager.py:2133] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0425 17:06:43.668874 136635635189504 checkpoint_manager.py:2142] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0425 17:06:43.669106 136762627626816 checkpoint_manager.py:2036] [process=5][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0425 17:06:43.669276 136762627626816 checkpoint_manager.py:2013] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0425 17:06:43.669920 136762627626816 metric_logger.py:185] completed step: 9, seconds: 0.146, TFLOP/s/device: 92.974, Tokens/s/device: 14014.055, total_weights: 65536, loss: 8.175 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Sat Apr 25 17:06:56 UTC 2026 EXIT_CODE=0