XPK Start: Sun Apr 19 04:18:14 UTC 2026 2026-04-19 04:18:18.665191: 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 00:00:1776572298.677929 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered E0000 00:00:1776572298.681873 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered W0000 00:00:1776572298.693145 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776572298.693164 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776572298.693166 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. W0000 00:00:1776572298.693168 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. 2026-04-19 04:18:37.870929: 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 04:18:38.393681 137381576783680 max_utils.py:273] Attempting to initialize the jax distributed system... INFO:2026-04-19 04:18:47,434:jax._src.distributed:140: Starting JAX distributed service on [::]:8482 I0419 04:18:47.434505 137381576783680 distributed.py:140] Starting JAX distributed service on [::]:8482 INFO:2026-04-19 04:18:47,436:jax._src.distributed:157: Connecting to JAX distributed service on mt-15-ocdbt-true-sv-qjcc1-slice-job-0-0.mt-15-ocdbt-true-sv-qjcc1:8482 I0419 04:18:47.436945 137381576783680 distributed.py:157] Connecting to JAX distributed service on mt-15-ocdbt-true-sv-qjcc1-slice-job-0-0.mt-15-ocdbt-true-sv-qjcc1:8482 I0419 04:18:49.028578 137381576783680 max_utils.py:284] Jax distributed system initialized! I0419 04:18:56.078349 137381576783680 max_utils.py:800] System Information: Jax Version: 0.8.1 I0419 04:18:56.078456 137381576783680 max_utils.py:801] System Information: Jaxlib Version: 0.8.1 I0419 04:18:56.078498 137381576783680 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 04:18:56.078531 137381576783680 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0419 04:18:56.692232 137381576783680 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0419 04:18:56.692838 137381576783680 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0419 04:18:56.693025 137381576783680 checkpointing.py:688] Setting up checkpoint logger... I0419 04:18:56.693076 137381576783680 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0419 04:18:56.693120 137381576783680 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None I0419 04:18:56.693468 137381576783680 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 0x7cf2273797f0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0419 04:19:00.043436 137381576783680 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0419 04:19:00.043779 137381576783680 checkpoint_manager.py:709] [process=0][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7cde183e56a0>}, handler_registry=None I0419 04:19:00.044033 137381576783680 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7cde183e56a0>` 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 04:19:00.044089 137381576783680 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7cde181caa50>` 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 04:19:00.044141 137381576783680 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 0x7cde183e56a0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7cde183e56a0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7cde181caa50>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7cde181caa50>}). I0419 04:19:00.044634 137381576783680 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33 I0419 04:19:00.044715 137381576783680 async_checkpointer.py:177] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7cdde057c180> timeout: 600 secs and primary_host=0 for async checkpoint writes I0419 04:19:01.612835 137381576783680 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints I0419 04:19:02.481644 137381576783680 checkpoint_manager.py:929] [process=0][thread=MainThread] CheckpointManager created, primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=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_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7cde183e67b0> I0419 04:19:02.481830 137381576783680 checkpointing.py:302] Checkpoint manager created! I0419 04:19:03.375141 137381576783680 nnx_wrappers.py:455] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0419 04:19:03.375252 137381576783680 nnx_wrappers.py:455] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0419 04:19:03.744554 137381576783680 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0419 04:19:03.744647 137381576783680 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0419 04:19:03.760434 137381576783680 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0419 04:19:03.760497 137381576783680 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0419 04:19:03.783262 137381576783680 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 04:19:03.783329 137381576783680 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 04:19:03.799147 137381576783680 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 04:19:03.799211 137381576783680 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 04:19:03.815000 137381576783680 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 04:19:03.815062 137381576783680 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 04:19:03.839745 137381576783680 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0419 04:19:03.839827 137381576783680 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0419 04:19:03.859328 137381576783680 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0419 04:19:03.859395 137381576783680 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0419 04:19:04.047075 137381576783680 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0419 04:19:04.047192 137381576783680 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint. fsdp: 32 I0419 04:19:05.468433 137381576783680 maxtext_utils.py:1740] params/params/decoder/decoder_norm/scale Shape: float32[2048] Logical: PartitionSpec('norm',) Physical: (None,) I0419 04:19:05.468568 137381576783680 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wi_0/kernel Shape: float32[2048,16,7168] Logical: PartitionSpec('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0419 04:19:05.468637 137381576783680 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wi_1/kernel Shape: float32[2048,16,7168] Logical: PartitionSpec('embed', 'layers', 'mlp') Physical: ('fsdp', None, None) I0419 04:19:05.468706 137381576783680 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wo/kernel Shape: float32[7168,16,2048] Logical: PartitionSpec('mlp', 'layers', 'embed') Physical: (None, None, 'fsdp') I0419 04:19:05.468767 137381576783680 maxtext_utils.py:1740] params/params/decoder/layers/post_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: PartitionSpec('norm', 'layers') Physical: (None, None) I0419 04:19:05.468822 137381576783680 maxtext_utils.py:1740] params/params/decoder/layers/pre_self_attention_layer_norm/scale Shape: float32[2048,16] Logical: PartitionSpec('norm', 'layers') Physical: (None, None) I0419 04:19:05.468902 137381576783680 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/key/kernel Shape: float32[2048,16,16,128] Logical: PartitionSpec('embed', 'layers', 'kv_heads', 'kv_head_dim') Physical: ('fsdp', None, None, None) I0419 04:19:05.468961 137381576783680 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/out/kernel Shape: float32[16,16,128,2048] Logical: PartitionSpec('heads', 'layers', 'kv', 'embed') Physical: (None, None, None, 'fsdp') I0419 04:19:05.469002 137381576783680 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/query/kernel Shape: float32[2048,16,16,128] Logical: PartitionSpec('embed', 'layers', 'q_heads', 'kv') Physical: ('fsdp', None, None, None) I0419 04:19:05.469041 137381576783680 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/value/kernel Shape: float32[2048,16,16,128] Logical: PartitionSpec('embed', 'layers', 'kv_heads', 'kv_head_dim') Physical: ('fsdp', None, None, None) I0419 04:19:05.469091 137381576783680 maxtext_utils.py:1740] params/params/decoder/logits_dense/kernel Shape: float32[2048,32000] Logical: PartitionSpec('embed', 'vocab') Physical: ('fsdp', None) I0419 04:19:05.469145 137381576783680 maxtext_utils.py:1740] params/params/token_embedder/embedding Shape: float32[32000,2048] Logical: PartitionSpec('vocab', 'embed') Physical: (None, 'fsdp') I0419 04:19:05.967264 137381576783680 train.py:158] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0419 04:19:05.967358 137381576783680 train.py:158] train/xent Physical: float32[32,2048]............................................ ('fsdp', None). I0419 04:19:05.982222 137381576783680 train.py:165] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length'). I0419 04:19:05.982283 137381576783680 train.py:165] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None). I0419 04:19:16.670377 137381576783680 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 04:19:16.676185 137381576783680 metric_logger.py:289] number parameters: 1.104 billion I0419 04:19:28.200588 137381576783680 checkpointing.py:794] Waiting for step 0 to finish before checkpoint... I0419 04:19:28.354124 137381576783680 checkpointing.py:798] Waited 0.1535177230834961 seconds for step 0 to finish before starting checkpointing. I0419 04:19:28.356603 137381576783680 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0419 04:19:28.358217 137381576783680 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 0 I0419 04:19:28.361010 137381576783680 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/0. I0419 04:19:29.525381 137381576783680 signaling_client.py:364] Using JaxDistributedSignalingClient I0419 04:19:29.769354 137255502980864 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/0 I0419 04:19:29.790205 137381576783680 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0419 04:19:29.790304 137381576783680 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 04:19:30.040142 137381576783680 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.250831s I0419 04:19:30.040988 137381576783680 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 3.040 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5074882507324219 s) (per-host) I0419 04:19:30.041058 137381576783680 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.507578s (batch_requests_ready=0.252074s, total_serialization_initiated=0.254743s, others=0.000762s) I0419 04:19:30.041137 137381576783680 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.511600s (all_items=0.000019s, per_item={'items': '0.00001884'}, temp_paths=0.511581) I0419 04:19:30.042102 137256598533888 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0419 04:19:30.042266 137381576783680 async_checkpointer.py:561] Finished blocking save. Time taken: 1.683982s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/0. I0419 04:19:30.921614 137381576783680 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize I0419 04:19:30.922018 137255494588160 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0419 04:19:30.922161 137381576783680 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776572368.356585, 'wait_for_prev_duration_secs': 6.151199340820312e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776572368.3582537, 'checkpointer_blocking_duration_secs': 1.6841175556182861, 'get_old_steps_start_time': 1776572370.0423915, 'get_old_steps_duration_secs': 2.3603439331054688e-05, 'checkpoint_manager_blocking_start_time': 1776572368.354726, 'checkpoint_manager_blocking_duration_secs': 2.567396402359009} I0419 04:19:30.922327 137381576783680 checkpointing.py:409] Started an asynchronous checkpoint save for step 0 I0419 04:19:30.922379 137381576783680 max_utils.py:750] Memstats: After params initialized: I0419 04:19:30.922436 137381576783680 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0)) I0419 04:19:30.922469 137381576783680 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0)) I0419 04:19:30.922496 137381576783680 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0)) I0419 04:19:30.922521 137381576783680 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0)) I0419 04:19:31.220756 137381576783680 metric_logger.py:185] completed step: 0, seconds: 11.167, TFLOP/s/device: 1.217, Tokens/s/device: 183.397, total_weights: 65536, loss: 10.874 I0419 04:19:31.223632 137381576783680 metric_logger.py:269] To see full metrics 'tensorboard --logdir=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/tensorboard/' I0419 04:19:31.566119 137256618686208 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/0/items I0419 04:19:31.620662 137381576783680 metric_logger.py:185] completed step: 1, seconds: 3.019, TFLOP/s/device: 4.501, Tokens/s/device: 678.465, total_weights: 65536, loss: 10.874 I0419 04:19:31.724004 137256037672704 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776572371322019147, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776572371322019147, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/0/_CHECKPOINT_METADATA I0419 04:19:31.766880 137381576783680 metric_logger.py:185] completed step: 2, seconds: 0.401, TFLOP/s/device: 33.913, Tokens/s/device: 5111.745, total_weights: 65536, loss: 10.560 I0419 04:19:31.914535 137381576783680 metric_logger.py:185] completed step: 3, seconds: 0.008, TFLOP/s/device: 1704.570, Tokens/s/device: 256931.376, total_weights: 65536, loss: 9.980 I0419 04:19:31.923180 137381576783680 checkpointing.py:794] Waiting for step 5 to finish before checkpoint... I0419 04:19:32.209445 137381576783680 checkpointing.py:798] Waited 0.2861490249633789 seconds for step 5 to finish before starting checkpointing. I0419 04:19:32.212265 137381576783680 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 04:19:33.974802 2633 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0419 04:19:36.066168 137259286255360 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/0/items/array_metadatas/process_0 I0419 04:20:11.446522 137259277862656 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 37.475288s (commit=36.417819s, array_metadata_write=1.057468s) I0419 04:20:11.448180 137256598533888 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 37.686 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.91461968421936 s) (per-host) I0419 04:20:11.448344 137256598533888 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.406122s. I0419 04:20:12.224181 137256598533888 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776572371322019147, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/0/_CHECKPOINT_METADATA I0419 04:20:12.999511 137256037672704 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776572371322019147, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/0/_CHECKPOINT_METADATA I0419 04:20:13.726127 137256598533888 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0419 04:20:13.774837 137256598533888 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000387s. I0419 04:20:14.632083 137256598533888 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.277812s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/0/items I0419 04:20:14.635251 137256598533888 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/0/items I0419 04:20:15.058984 137256598533888 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/0 I0419 04:20:18.080351 137256598533888 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/0`. I0419 04:20:18.081504 137256598533888 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 49.723218s. directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/0 I0419 04:20:18.083249 137256598533888 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 48.041026s. I0419 04:20:18.083528 137255494588160 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 04:20:18.083684 137255494588160 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0419 04:20:18.083789 137255494588160 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts... I0419 04:20:18.085751 137255494588160 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts. I0419 04:20:18.086056 137381576783680 checkpoint_manager.py:2036] [process=0][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0. W0419 04:20:18.086244 137381576783680 checkpoint_manager.py:1458] Waiting for previous save to complete took 45.873982 seconds. If this number is high, consider checkpointing less frequently. I0419 04:20:18.088150 137381576783680 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 5 I0419 04:20:18.092215 137381576783680 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/5. I0419 04:20:19.056863 137381576783680 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0419 04:20:19.057003 137381576783680 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 04:20:19.105664 137381576783680 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.050027s I0419 04:20:19.106618 137381576783680 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.030 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.30669116973876953 s) (per-host) I0419 04:20:19.106726 137381576783680 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.306824s (batch_requests_ready=0.251975s, total_serialization_initiated=0.054016s, others=0.000833s) I0419 04:20:19.106865 137381576783680 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.311673s (all_items=0.000025s, per_item={'items': '0.00002480'}, temp_paths=0.311648) I0419 04:20:19.108107 137255528158976 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0419 04:20:19.108272 137381576783680 async_checkpointer.py:561] Finished blocking save. Time taken: 1.020021s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/5. I0419 04:20:19.484910 137256598533888 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/5 I0419 04:20:19.530051 137381576783680 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=5] Starting CheckpointManager Save Finalize thread=save_finalize I0419 04:20:19.530546 137255494588160 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0419 04:20:19.530792 137381576783680 standard_logger.py:34] {'step': 5, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776572372.2122247, 'wait_for_prev_duration_secs': 45.873982429504395, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776572418.0882142, 'checkpointer_blocking_duration_secs': 1.0202925205230713, 'get_old_steps_start_time': 1776572419.108532, 'get_old_steps_duration_secs': 3.743171691894531e-05, 'checkpoint_manager_blocking_start_time': 1776572372.2100587, 'checkpoint_manager_blocking_duration_secs': 47.32067799568176} I0419 04:20:19.530957 137381576783680 checkpointing.py:409] Started an asynchronous checkpoint save for step 5 I0419 04:20:39.250993 137381576783680 metric_logger.py:185] completed step: 4, seconds: 0.148, TFLOP/s/device: 92.062, Tokens/s/device: 13876.655, total_weights: 65536, loss: 9.460 I0419 04:20:39.259863 137381576783680 metric_logger.py:185] completed step: 5, seconds: 0.148, TFLOP/s/device: 92.001, Tokens/s/device: 13867.353, total_weights: 65536, loss: 8.959 I0419 04:20:39.405569 137381576783680 metric_logger.py:185] completed step: 6, seconds: 67.335, TFLOP/s/device: 0.202, Tokens/s/device: 30.415, total_weights: 65536, loss: 8.469 I0419 04:20:39.553405 137381576783680 metric_logger.py:185] completed step: 7, seconds: 0.007, TFLOP/s/device: 1951.893, Tokens/s/device: 294210.602, total_weights: 65536, loss: 8.003 I0419 04:20:39.700904 137381576783680 metric_logger.py:185] completed step: 8, seconds: 0.147, TFLOP/s/device: 92.677, Tokens/s/device: 13969.224, total_weights: 65536, loss: 7.572 I0419 04:20:39.847903 137381576783680 checkpointing.py:794] Waiting for step 9 to finish before checkpoint... I0419 04:20:39.848537 137381576783680 checkpointing.py:798] Waited 0.0006556510925292969 seconds for step 9 to finish before starting checkpointing. I0419 04:20:39.850673 137381576783680 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=5][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 04:20:41.187703 137254454417152 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/5/items I0419 04:20:44.842939 137256618686208 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/5/items/array_metadatas/process_0 I0419 04:21:25.365185 137255536551680 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 42.606684s (commit=41.524634s, array_metadata_write=1.082050s) I0419 04:21:25.366209 137255528158976 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 23.730 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 66.56626343727112 s) (per-host) I0419 04:21:25.366331 137255528158976 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 66.257951s. I0419 04:21:26.980197 137255528158976 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0419 04:21:27.210178 137255528158976 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000246s. I0419 04:21:28.078766 137255528158976 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.916822s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/5/items I0419 04:21:28.080765 137255528158976 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/5/items I0419 04:21:28.491101 137255528158976 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/5 I0419 04:21:29.869905 137255528158976 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/5`. I0419 04:21:29.870723 137255528158976 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 71.782481s. directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/5 I0419 04:21:29.872369 137255528158976 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 70.763988s. I0419 04:21:29.872590 137255494588160 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 04:21:29.872725 137255494588160 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0419 04:21:29.872791 137255494588160 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=5] CheckpointManager Save Finalize is syncing with other hosts... I0419 04:21:29.874443 137255494588160 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=5] CheckpointManager Save Finalize is done on all hosts. I0419 04:21:29.874618 137381576783680 checkpoint_manager.py:2036] [process=0][thread=MainThread][step=5][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=5. W0419 04:21:29.874755 137381576783680 checkpoint_manager.py:1458] Waiting for previous save to complete took 50.024093 seconds. If this number is high, consider checkpointing less frequently. I0419 04:21:29.876331 137381576783680 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 9 I0419 04:21:29.880002 137381576783680 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/9. I0419 04:21:30.825893 137255528158976 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/9 I0419 04:21:30.836250 137381576783680 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array. I0419 04:21:30.836351 137381576783680 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 04:21:30.862201 137381576783680 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.026879s I0419 04:21:30.862953 137381576783680 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.441 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.28351569175720215 s) (per-host) I0419 04:21:30.863027 137381576783680 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.283610s (batch_requests_ready=0.253831s, total_serialization_initiated=0.029111s, others=0.000668s) I0419 04:21:30.863106 137381576783680 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.287401s (all_items=0.000017s, per_item={'items': '0.00001740'}, temp_paths=0.287383) I0419 04:21:30.863798 137254979602176 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started. I0419 04:21:30.863904 137381576783680 async_checkpointer.py:561] Finished blocking save. Time taken: 0.987498s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/9. I0419 04:21:31.315500 137381576783680 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize I0419 04:21:31.315876 137255494588160 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0419 04:21:31.316052 137381576783680 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776572439.850643, 'wait_for_prev_duration_secs': 50.02409267425537, 'time_between_consecutive_saves_sec': 21.764812231063843, 'checkpointer_blocking_start_time': 1776572489.8763716, 'checkpointer_blocking_duration_secs': 0.9876291751861572, 'get_old_steps_start_time': 1776572490.864017, 'get_old_steps_duration_secs': 2.1457672119140625e-05, 'checkpoint_manager_blocking_start_time': 1776572439.8487587, 'checkpoint_manager_blocking_duration_secs': 51.4672589302063} I0419 04:21:31.316180 137381576783680 checkpointing.py:409] Started an asynchronous checkpoint save for step 9 I0419 04:21:31.316229 137381576783680 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0419 04:21:32.985007 137256598533888 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/9/items I0419 04:21:36.549878 137255513032448 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/9/items/array_metadatas/process_0 I0419 04:22:17.120373 137254996387584 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 42.606477s (commit=41.058194s, array_metadata_write=1.548283s) I0419 04:22:17.121410 137254979602176 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 33.939 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 46.54194259643555 s) (per-host) I0419 04:22:17.121529 137254979602176 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 46.257662s. I0419 04:22:19.255487 137254979602176 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0419 04:22:19.381537 137254979602176 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000246s. I0419 04:22:20.340214 137254979602176 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.925543s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/9/items I0419 04:22:20.342558 137254979602176 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/9/items I0419 04:22:20.761499 137254979602176 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/9 I0419 04:22:22.455050 137254979602176 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/9`. I0419 04:22:22.455864 137254979602176 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 52.579459s. directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260419_025828/linen_xpk_feat_nnx_set_defaults_true_20260419_025828_15_ocdbt_true/checkpoints/9 I0419 04:22:22.458104 137254979602176 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 51.594231s. I0419 04:22:22.458314 137255494588160 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0419 04:22:22.458430 137255494588160 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0419 04:22:22.458483 137255494588160 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts... I0419 04:22:22.461050 137255494588160 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts. I0419 04:22:22.461201 137381576783680 checkpoint_manager.py:2036] [process=0][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9. I0419 04:22:22.461294 137381576783680 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0419 04:22:22.462057 137381576783680 metric_logger.py:185] completed step: 9, seconds: 0.148, TFLOP/s/device: 91.750, Tokens/s/device: 13829.521, 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 04:22:35 UTC 2026 EXIT_CODE=0