XPK Start: Thu Apr 23 18:01:15 UTC 2026 2026-04-23 18:01:19.600447: 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 00:00:1776967279.613760 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1776967279.617505 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1776967279.629227 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776967279.629247 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776967279.629249 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776967279.629251 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-23 18:01:38.806320: 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) I0423 18:01:39.332361 135590297855808 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-23 18:01:48,372:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0423 18:01:48.372411 135590297855808 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-23 18:01:48,374:jax._src.distributed:157: Connecting to JAX distributed service on mt-09-pdb-lt-1-uq1sm-slice-job-0-0.mt-09-pdb-lt-1-uq1sm:8482 I0423 18:01:48.374963 135590297855808 distributed.py:157] Connecting to JAX distributed service on mt-09-pdb-lt-1-uq1sm-slice-job-0-0.mt-09-pdb-lt-1-uq1sm:8482 I0423 18:01:49.450392 135590297855808 max_utils.py:284] Jax distributed system initialized! I0423 18:01:56.575542 135590297855808 max_utils.py:800] System Information: Jax Version: 0.8.1 I0423 18:01:56.575649 135590297855808 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0423 18:01:56.575689 135590297855808 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 I0423 18:01:56.575726 135590297855808 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0423 18:01:57.194618 135590297855808 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 8, 1, 1, 1, 1, 4, 1, 1, 1, 1) I0423 18:01:57.307089 135590297855808 checkpointing.py:688] Setting up checkpoint logger... I0423 18:01:57.307264 135590297855808 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0423 18:01:57.307310 135590297855808 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0423 18:01:57.307523 135590297855808 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 0x7b5111b71310>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0423 18:02:00.359699 135590297855808 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0423 18:02:00.359946 135590297855808 checkpoint_manager.py:709] [process=6][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b3cb07ac1d0>}, handler_registry=None I0423 18:02:00.360193 135590297855808 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b3cb07ac1d0>` 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`. I0423 18:02:00.360247 135590297855808 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b3cd8605940>` 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`. I0423 18:02:00.360284 135590297855808 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 0x7b3cb07ac1d0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b3cb07ac1d0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b3cd8605940>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b3cd8605940>}). I0423 18:02:00.360620 135590297855808 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0423 18:02:00.360698 135590297855808 async_checkpointer.py:177] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7b3cd81e4f40> timeout: 600 secs and primary_host=0 for async checkpoint writes I0423 18:02:01.776420 135590297855808 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_09_pdb_lt_1/checkpoints I0423 18:02:02.212146 135590297855808 checkpoint_manager.py:929] [process=6][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_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_09_pdb_lt_1/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7b3cd8605b80> I0423 18:02:02.212330 135590297855808 checkpointing.py:302] Checkpoint manager created! I0423 18:02:03.164885 135590297855808 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0423 18:02:03.165010 135590297855808 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 8 tensor: 4 I0423 18:02:04.654532 135590297855808 maxtext_utils.py:1749] decoder/decoder_norm/scale/value Shape: float32[2048] Physical: ('tensor',) I0423 18:02:04.654655 135590297855808 maxtext_utils.py:1749] decoder/layers/mlp/wi_0/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, 'tensor') I0423 18:02:04.654710 135590297855808 maxtext_utils.py:1749] decoder/layers/mlp/wi_1/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, 'tensor') I0423 18:02:04.654754 135590297855808 maxtext_utils.py:1749] decoder/layers/mlp/wo/kernel/value Shape: float32[7168,16,2048] Physical: ('tensor', None, 'fsdp') I0423 18:02:04.654791 135590297855808 maxtext_utils.py:1749] decoder/layers/post_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: ('tensor', None) I0423 18:02:04.654826 135590297855808 maxtext_utils.py:1749] decoder/layers/pre_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: ('tensor', None) I0423 18:02:04.654862 135590297855808 maxtext_utils.py:1749] decoder/layers/self_attention/key/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, 'tensor', None) I0423 18:02:04.654899 135590297855808 maxtext_utils.py:1749] decoder/layers/self_attention/out/kernel/value Shape: float32[16,16,128,2048] Physical: ('tensor', None, None, 'fsdp') I0423 18:02:04.654935 135590297855808 maxtext_utils.py:1749] decoder/layers/self_attention/query/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, 'tensor', None) I0423 18:02:04.654969 135590297855808 maxtext_utils.py:1749] decoder/layers/self_attention/value/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, 'tensor', None) I0423 18:02:04.655004 135590297855808 maxtext_utils.py:1749] decoder/logits_dense/kernel/value Shape: float32[2048,32000] Physical: ('fsdp', 'tensor') I0423 18:02:04.655039 135590297855808 maxtext_utils.py:1749] token_embedder/embedding/value Shape: float32[32000,2048] Physical: ('tensor', 'fsdp') I0423 18:02:05.052855 135590297855808 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0423 18:02:05.053007 135590297855808 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor'). I0423 18:02:05.058673 135590297855808 nnx_decoders.py:616] Unknown Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0423 18:02:05.058766 135590297855808 nnx_decoders.py:616] Unknown Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor'). I0423 18:02:05.202630 135590297855808 attentions.py:1084] attentions/inputs_q Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0423 18:02:05.202763 135590297855808 attentions.py:1084] attentions/inputs_q Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor'). I0423 18:02:05.347451 135590297855808 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0423 18:02:05.347589 135590297855808 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor'). I0423 18:02:05.500482 135590297855808 attentions.py:1150] attentions/query Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0423 18:02:05.500615 135590297855808 attentions.py:1150] attentions/query Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0423 18:02:05.644455 135590297855808 attentions.py:1151] attentions/key Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0423 18:02:05.644594 135590297855808 attentions.py:1151] attentions/key Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0423 18:02:05.787899 135590297855808 attentions.py:1152] attentions/value Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0423 18:02:05.788033 135590297855808 attentions.py:1152] attentions/value Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0423 18:02:05.942829 135590297855808 attentions.py:1193] attentions/out Logical: bfloat16[8,2048,16,128]..................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0423 18:02:05.942965 135590297855808 attentions.py:1193] attentions/out Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None). I0423 18:02:05.965388 135590297855808 linears.py:541] linears/x Logical: bfloat16[8,2048,7168]....................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0423 18:02:05.965485 135590297855808 linears.py:541] linears/x Physical: bfloat16[8,2048,7168]....................................... ('fsdp', None, 'tensor'). I0423 18:02:16.096737 135590297855808 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. I0423 18:02:16.099115 135590297855808 metric_logger.py:289] number parameters: 1.104 billion I0423 18:02:17.402045 135590297855808 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0423 18:02:26.162087 135590297855808 checkpointing.py:798] Waited 8.760025024414062 seconds for step 0 to finish before starting checkpointing. I0423 18:02:26.164255 135590297855808 checkpoint_manager.py:2013] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0423 18:02:26.166056 135590297855808 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 0 I0423 18:02:26.168830 135590297855808 async_checkpointer.py:452] [process=6] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_09_pdb_lt_1/checkpoints/0. I0423 18:02:27.139395 135590297855808 signaling_client.py:364] Using JaxDistributedSignalingClient I0423 18:02:27.140554 135590297855808 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0423 18:02:27.140613 135590297855808 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 I0423 18:02:27.370172 135590297855808 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.231908s I0423 18:02:27.370353 135590297855808 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 3.142 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.49100184440612793 s) (per-host) I0423 18:02:27.370407 135590297855808 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.491067s (batch_requests_ready=0.252726s, total_serialization_initiated=0.238267s, others=0.000073s) I0423 18:02:27.370506 135590297855808 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.495162s (all_items=0.000021s, per_item={'items': '0.00002074'}, temp_paths=0.495142) I0423 18:02:27.371485 135468855301888 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started. I0423 18:02:27.371653 135590297855808 async_checkpointer.py:561] Finished blocking save. Time taken: 1.205514s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_09_pdb_lt_1/checkpoints/0. I0423 18:02:27.396258 135590297855808 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0423 18:02:27.396533 135465667647232 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0423 18:02:27.396670 135590297855808 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_09_pdb_lt_1/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776967346.164237, 'wait_for_prev_duration_secs': 6.985664367675781e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776967346.1661103, 'checkpointer_blocking_duration_secs': 1.2056572437286377, 'get_old_steps_start_time': 1776967347.3717892, 'get_old_steps_duration_secs': 2.5033950805664062e-05, 'checkpoint_manager_blocking_start_time': 1776967346.1625488, 'checkpoint_manager_blocking_duration_secs': 1.2340826988220215} I0423 18:02:27.396775 135590297855808 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0423 18:02:27.396845 135590297855808 max_utils.py:750] Memstats: After params initialized: I0423 18:02:27.396896 135590297855808 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_24(process=6,(0,6,0,0)) I0423 18:02:27.396928 135590297855808 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_25(process=6,(1,6,0,0)) I0423 18:02:27.396954 135590297855808 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_28(process=6,(0,7,0,0)) I0423 18:02:27.396979 135590297855808 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_29(process=6,(1,7,0,0)) I0423 18:02:27.668346 135590297855808 metric_logger.py:185] completed step: 0, seconds: 1.302, TFLOP/s/device: 2.609, Tokens/s/device: 393.288, total_weights: 16384, loss: 10.864 I0423 18:02:27.749657 135590297855808 metric_logger.py:185] completed step: 1, seconds: 10.265, TFLOP/s/device: 0.331, Tokens/s/device: 49.880, total_weights: 16384, loss: 10.864 I0423 18:02:28.212440 135590297855808 metric_logger.py:185] completed step: 2, seconds: 0.025, TFLOP/s/device: 135.093, Tokens/s/device: 20362.711, total_weights: 16384, loss: 9.760 I0423 18:02:28.271219 135590297855808 metric_logger.py:185] completed step: 3, seconds: 0.453, TFLOP/s/device: 7.501, Tokens/s/device: 1130.575, total_weights: 16384, loss: 8.820 I0423 18:02:28.390794 135590297855808 metric_logger.py:185] completed step: 4, seconds: 0.074, TFLOP/s/device: 45.949, Tokens/s/device: 6925.938, total_weights: 16384, loss: 7.947 I0423 18:02:28.397457 135590297855808 metric_logger.py:185] completed step: 5, seconds: 0.058, TFLOP/s/device: 58.831, Tokens/s/device: 8867.643, total_weights: 16384, loss: 7.209 I0423 18:02:40.878636 135590297855808 metric_logger.py:185] completed step: 6, seconds: 0.120, TFLOP/s/device: 28.268, Tokens/s/device: 4260.808, total_weights: 16384, loss: 6.651 I0423 18:02:40.937246 135590297855808 metric_logger.py:185] completed step: 7, seconds: 12.424, TFLOP/s/device: 0.273, Tokens/s/device: 41.211, total_weights: 16384, loss: 6.267 I0423 18:02:40.995738 135590297855808 metric_logger.py:185] completed step: 8, seconds: 0.064, TFLOP/s/device: 53.331, Tokens/s/device: 8038.686, total_weights: 16384, loss: 6.030 I0423 18:02:41.055302 135590297855808 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0423 18:02:41.055860 135590297855808 checkpointing.py:798] Waited 0.0005681514739990234 seconds for step 9 to finish before starting checkpointing. I0423 18:02:41.057788 135590297855808 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0423 18:02:42.465269 2424 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0423 18:02:44.869358 135468863694592 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_09_pdb_lt_1/checkpoints/0/items/array_metadatas/process_6 I0423 18:03:16.475091 135468855301888 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 31.849 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 49.59569978713989 s) (per-host) I0423 18:03:16.475216 135468855301888 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 49.103617s. I0423 18:03:23.648070 135468855301888 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 56.276456s. I0423 18:03:23.648399 135465667647232 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0423 18:03:23.648529 135465667647232 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0423 18:03:23.648582 135465667647232 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0423 18:03:23.650150 135465667647232 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0423 18:03:23.650304 135590297855808 checkpoint_manager.py:2036] [process=6][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0423 18:03:23.650387 135590297855808 checkpoint_manager.py:1458] Waiting for previous save to complete took 42.592606 seconds. If this number is high, consider checkpointing less frequently. I0423 18:03:23.652526 135590297855808 checkpoint_manager.py:1518] [process=6] Saving checkpoint at step 9 I0423 18:03:23.656420 135590297855808 async_checkpointer.py:452] [process=6] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_09_pdb_lt_1/checkpoints/9. I0423 18:03:24.310917 135590297855808 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array. I0423 18:03:24.311071 135590297855808 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 I0423 18:03:24.353132 135590297855808 base_pytree_checkpoint_handler.py:153] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.044393s I0423 18:03:24.353300 135590297855808 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.096 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.30269694328308105 s) (per-host) I0423 18:03:24.353353 135590297855808 base_pytree_checkpoint_handler.py:737] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.302759s (batch_requests_ready=0.253726s, total_serialization_initiated=0.048964s, others=0.000070s) I0423 18:03:24.353447 135590297855808 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.306901s (all_items=0.000016s, per_item={'items': '0.00001574'}, temp_paths=0.306885) I0423 18:03:24.354321 135465644128000 async_checkpointer.py:79] [process=6][thread=async_save] Background save thread started. I0423 18:03:24.354445 135590297855808 async_checkpointer.py:561] Finished blocking save. Time taken: 0.701852s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_09_pdb_lt_1/checkpoints/9. I0423 18:03:24.365908 135590297855808 checkpoint_manager.py:1566] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0423 18:03:24.366220 135465667647232 async_checkpointer.py:265] [process=6][thread=save_finalize] Waiting for background save thread=async_save. I0423 18:03:24.366388 135590297855808 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_09_pdb_lt_1/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776967361.0577595, 'wait_for_prev_duration_secs': 42.59260606765747, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776967403.6525657, 'checkpointer_blocking_duration_secs': 0.7020444869995117, 'get_old_steps_start_time': 1776967404.3546293, 'get_old_steps_duration_secs': 2.6464462280273438e-05, 'checkpoint_manager_blocking_start_time': 1776967361.0560846, 'checkpoint_manager_blocking_duration_secs': 43.31026649475098} I0423 18:03:24.366501 135590297855808 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0423 18:03:24.366546 135590297855808 checkpoint_manager.py:2024] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0423 18:03:29.541812 135458117883648 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260423_155242/nnx_xpk_feat_nnx_set_defaults_true_20260423_155242_09_pdb_lt_1/checkpoints/9/items/array_metadatas/process_6 I0423 18:04:05.236687 135465644128000 base_pytree_checkpoint_handler.py:129] [process=6] /jax/checkpoint/write/gbytes_per_sec: 38.353 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.18604850769043 s) (per-host) I0423 18:04:05.236795 135465644128000 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 40.882275s. I0423 18:04:12.603606 135465644128000 async_checkpointer.py:144] [process=6][thread=async_save] Background save thread done. Time taken: 48.249069s. I0423 18:04:12.603906 135465667647232 async_checkpointer.py:273] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save. I0423 18:04:12.604020 135465667647232 async_checkpointer.py:283] [process=6][thread=save_finalize] No errors found in background save thread=async_save. I0423 18:04:12.604070 135465667647232 checkpoint_manager.py:2133] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0423 18:04:12.605412 135465667647232 checkpoint_manager.py:2142] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0423 18:04:12.605589 135590297855808 checkpoint_manager.py:2036] [process=6][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0423 18:04:12.605748 135590297855808 checkpoint_manager.py:2013] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0423 18:04:12.606335 135590297855808 metric_logger.py:185] completed step: 9, seconds: 0.058, TFLOP/s/device: 58.410, Tokens/s/device: 8804.210, total_weights: 16384, loss: 5.890 Per train step: Total TFLOPs: 3.40 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Thu Apr 23 18:04:25 UTC 2026 EXIT_CODE=0