XPK Start: Mon Apr 20 18:48:52 UTC 2026 PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used. 2026-04-20 18:49:16.214509: 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) I0420 18:49:16.392384 136025790359360 max_utils.py:273] Attempting to initialize the jax distributed system... I0420 18:49:25.432014 136025790359360 distributed.py:149] Starting JAX distributed service on [::]:8482 I0420 18:49:25.434314 136025790359360 distributed.py:172] Connecting to JAX distributed service on mt-16-zarr3-true-rs-h7man-slice-job-0-0.mt-16-zarr3-true-rs-h7man:8482 I0420 18:49:26.689961 136025790359360 max_utils.py:284] Jax distributed system initialized! I0420 18:49:31.958607 136025790359360 max_utils.py:800] System Information: Jax Version: 0.9.2 I0420 18:49:31.958708 136025790359360 max_utils.py:801] System Information: Jaxlib Version: 0.9.2 I0420 18:49:31.958746 136025790359360 max_utils.py:802] System Information: Jax Backend: PJRT C API TFRT TPU v6 lite Built on Mar 4 2026 11:32:08 (1772652728) cl/878335365 I0420 18:49:31.958778 136025790359360 train_utils.py:378] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing. I0420 18:49:33.000779 136025790359360 maxtext_utils.py:1718] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1) I0420 18:49:33.114513 136025790359360 checkpointing.py:688] Setting up checkpoint logger... I0420 18:49:33.114634 136025790359360 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True I0420 18:49:33.114680 136025790359360 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None I0420 18:49:33.114872 136025790359360 base_pytree_checkpoint_handler.py:441] 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 0x7bb641548080>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB) I0420 18:49:34.372605 136025790359360 checkpointing.py:266] Enabling policy for fixed interval checkpointing. I0420 18:49:34.372833 136025790359360 checkpoint_manager.py:708] [process=0][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ba1e4474410>}, handler_registry=None I0420 18:49:34.373068 136025790359360 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ba1e4474410>` 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`. I0420 18:49:34.373129 136025790359360 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7bb535886d50>` 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`. I0420 18:49:34.373163 136025790359360 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 0x7ba1e4474410>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ba1e4474410>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7bb535886d50>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7bb535886d50>}). I0420 18:49:34.373472 136025790359360 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34 I0420 18:49:34.373540 136025790359360 async_checkpointer.py:192] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7ba1e41e84a0> timeout: 1200 secs and primary_host=0 for async checkpoint writes I0420 18:49:36.352313 135907034253056 checkpoint.py:228] Read Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776710758820783479, 'commit_timestamp_nsecs': 1776710805730687132, 'custom_metadata': {}} from gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/5/_CHECKPOINT_METADATA I0420 18:49:36.607836 135907034253056 checkpoint_manager.py:1871] Missing metrics for step 5 I0420 18:49:36.639247 136025790359360 checkpoint_manager.py:1812] Found 3 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints I0420 18:49:37.047183 136025790359360 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/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7ba1e44ec680> I0420 18:49:37.047347 136025790359360 checkpointing.py:302] Checkpoint manager created! I0420 18:49:37.990639 136025790359360 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint I0420 18:49:37.990752 136025790359360 checkpointing.py:582] restoring from this run's directory step 9 I0420 18:49:38.237476 136025790359360 event_tracking.py:70] [process=0] [sync] Started load checkpoint @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/9. I0420 18:49:38.480617 136025790359360 checkpointer.py:307] Restoring checkpoint from gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/9. I0420 18:49:38.481032 136025790359360 event_tracking.py:125] [process=0] [sync] Finished blocking load in 0.24 seconds. Continuing load @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/9. I0420 18:49:39.929551 1415 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com I0420 18:49:46.756636 136025790359360 jax_array_handlers.py:843] [process=0] /jax/orbax/read/worker/io/requested throughput: 231.252 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 6.833574295043945 s) (per-host) I0420 18:49:46.765988 136025790359360 base_pytree_checkpoint_handler.py:130] [process=0] /jax/checkpoint/read/gbytes_per_sec: 217.761 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 7.267621755599976 s) (per-host) I0420 18:49:49.940411 136025790359360 event_tracking.py:138] [process=0] [sync] Finished load in 11.70 seconds @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/9 I0420 18:49:49.940567 136025790359360 standard_logger.py:34] {'step': 9, 'event_type': 'restore', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints', 'checkpointer_start_time': 1776710978.2373767, 'checkpointer_duration_secs': 11.70315933227539, 'checkpoint_manager_start_time': 1776710977.9913633, 'checkpoint_manager_duration_secs': 11.949174880981445} fsdp: 32 I0420 18:49:50.016103 136025790359360 maxtext_utils.py:1836] decoder/decoder_norm/scale/value Shape: float32[2048] Physical: (None,) I0420 18:49:50.016211 136025790359360 maxtext_utils.py:1836] decoder/layers/mlp/wi_0/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0420 18:49:50.016262 136025790359360 maxtext_utils.py:1836] decoder/layers/mlp/wi_1/kernel/value Shape: float32[2048,16,7168] Physical: ('fsdp', None, None) I0420 18:49:50.016304 136025790359360 maxtext_utils.py:1836] decoder/layers/mlp/wo/kernel/value Shape: float32[7168,16,2048] Physical: (None, None, 'fsdp') I0420 18:49:50.016341 136025790359360 maxtext_utils.py:1836] decoder/layers/post_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0420 18:49:50.016374 136025790359360 maxtext_utils.py:1836] decoder/layers/pre_self_attention_layer_norm/scale/value Shape: float32[2048,16] Physical: (None, None) I0420 18:49:50.016407 136025790359360 maxtext_utils.py:1836] decoder/layers/self_attention/key/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0420 18:49:50.016441 136025790359360 maxtext_utils.py:1836] decoder/layers/self_attention/out/kernel/value Shape: float32[16,16,128,2048] Physical: (None, None, None, 'fsdp') I0420 18:49:50.016474 136025790359360 maxtext_utils.py:1836] decoder/layers/self_attention/query/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0420 18:49:50.016505 136025790359360 maxtext_utils.py:1836] decoder/layers/self_attention/value/kernel/value Shape: float32[2048,16,16,128] Physical: ('fsdp', None, None, None) I0420 18:49:50.016537 136025790359360 maxtext_utils.py:1836] decoder/logits_dense/kernel/value Shape: float32[2048,32000] Physical: ('fsdp', None) I0420 18:49:50.016568 136025790359360 maxtext_utils.py:1836] token_embedder/embedding/value Shape: float32[32000,2048] Physical: (None, 'fsdp') I0420 18:49:50.419304 136025790359360 nnx_decoders.py:465] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0420 18:49:50.419396 136025790359360 nnx_decoders.py:465] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 18:49:50.424853 136025790359360 nnx_decoders.py:465] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed'). I0420 18:49:50.424918 136025790359360 nnx_decoders.py:465] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 18:49:50.441421 136025790359360 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0420 18:49:50.441484 136025790359360 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 18:49:50.457147 136025790359360 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed'). I0420 18:49:50.457202 136025790359360 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None). I0420 18:49:50.480979 136025790359360 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0420 18:49:50.481043 136025790359360 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 18:49:50.496840 136025790359360 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0420 18:49:50.496909 136025790359360 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 18:49:50.512615 136025790359360 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim'). I0420 18:49:50.512675 136025790359360 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 18:49:50.542332 136025790359360 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv'). I0420 18:49:50.542402 136025790359360 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None). I0420 18:49:50.562105 136025790359360 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp'). I0420 18:49:50.562166 136025790359360 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None). I0420 18:50:01.943192 136025790359360 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. I0420 18:50:01.952415 136025790359360 metric_logger.py:301] number parameters: 1.104 billion I0420 18:50:13.697458 136025790359360 checkpointing.py:794] Waiting for step 10 to finish before checkpoint... I0420 18:50:13.848510 136025790359360 checkpointing.py:798] Waited 0.15104007720947266 seconds for step 10 to finish before starting checkpointing. I0420 18:50:13.850791 136025790359360 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0420 18:50:13.852651 136025790359360 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 10 I0420 18:50:13.854146 136025790359360 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/10. I0420 18:50:14.183366 136025790359360 signaling_client.py:364] Using JaxDistributedSignalingClient I0420 18:50:14.200121 136025790359360 jax_array_handlers.py:360] Scheduling D2H of 69 prioritized jax.Array. I0420 18:50:14.200239 136025790359360 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 I0420 18:50:14.466686 135906430273280 atomicity.py:140] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/10 I0420 18:50:14.478503 136025790359360 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.279907s I0420 18:50:14.479355 136025790359360 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 5.350 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.28832292556762695 s) (per-host) I0420 18:50:14.479425 136025790359360 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.288416s (batch_requests_ready=0.002561s, total_serialization_initiated=0.285084s, others=0.000772s) I0420 18:50:14.479526 136025790359360 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.292504s (all_items=0.000016s, per_item={'items': '0.00001621'}, temp_paths=0.292488) I0420 18:50:14.480324 136025790359360 event_tracking.py:125] [process=0] [async] Finished blocking save in 0.63 seconds. Continuing save @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/10. I0420 18:50:14.480675 135906438665984 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-20 19:10:14.480638 I0420 18:50:14.482592 136025790359360 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=10] Starting CheckpointManager Save Finalize thread=save_finalize I0420 18:50:14.482834 135901582137088 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0420 18:50:14.482988 136025790359360 standard_logger.py:34] {'step': 10, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776711013.8507667, 'wait_for_prev_duration_secs': 7.677078247070312e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776711013.8526964, 'checkpointer_blocking_duration_secs': 0.6281287670135498, 'get_old_steps_start_time': 1776711014.4808488, 'get_old_steps_duration_secs': 2.9325485229492188e-05, 'checkpoint_manager_blocking_start_time': 1776711013.8488836, 'checkpoint_manager_blocking_duration_secs': 0.6340699195861816} I0420 18:50:14.483107 136025790359360 checkpointing.py:409] Started an asynchronous checkpoint save for step 10 I0420 18:50:14.483179 136025790359360 max_utils.py:750] Memstats: After params initialized: I0420 18:50:14.483226 136025790359360 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0)) I0420 18:50:14.483259 136025790359360 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0)) I0420 18:50:14.483285 136025790359360 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0)) I0420 18:50:14.483309 136025790359360 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0)) I0420 18:50:14.847902 136025790359360 metric_logger.py:196] completed step: 10, seconds: 11.400, TFLOP/s/device: 1.192, Tokens/s/device: 179.642, total_weights: 65536, loss: 6.827, lm_loss: 6.827, perplexity: 922.537 I0420 18:50:14.989006 136025790359360 metric_logger.py:196] completed step: 11, seconds: 1.147, TFLOP/s/device: 11.847, Tokens/s/device: 1785.764, total_weights: 65536, loss: 6.522, lm_loss: 6.522, perplexity: 679.821 I0420 18:50:15.137316 136025790359360 metric_logger.py:196] completed step: 12, seconds: 0.013, TFLOP/s/device: 1016.012, Tokens/s/device: 153144.395, total_weights: 65536, loss: 6.260, lm_loss: 6.260, perplexity: 523.013 I0420 18:50:15.284705 136025790359360 metric_logger.py:196] completed step: 13, seconds: 0.142, TFLOP/s/device: 95.572, Tokens/s/device: 14405.593, total_weights: 65536, loss: 6.039, lm_loss: 6.039, perplexity: 419.520 I0420 18:50:15.297038 136025790359360 checkpointing.py:794] Waiting for step 15 to finish before checkpoint... I0420 18:50:15.578996 136025790359360 checkpointing.py:798] Waited 0.28192591667175293 seconds for step 15 to finish before starting checkpointing. I0420 18:50:15.581828 136025790359360 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=10][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0420 18:50:16.251952 135902152091392 atomicity.py:140] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/10/items I0420 18:50:16.856718 135901573744384 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776711015996271124, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776711015996271124, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/10/_CHECKPOINT_METADATA I0420 18:50:19.673565 135902123083520 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/10/items/array_metadatas/process_0 I0420 18:51:00.136161 135902114690816 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 42.295789s (commit=40.875292s, array_metadata_write=1.420497s) I0420 18:51:00.137482 135906438665984 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 34.379 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 45.94641613960266 s) (per-host) I0420 18:51:00.137650 135906438665984 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 45.656860s. I0420 18:51:01.642376 135906438665984 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000767s. I0420 18:51:01.699654 135901573744384 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776711015996271124, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/10/_CHECKPOINT_METADATA I0420 18:51:02.273639 135906438665984 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0420 18:51:03.230280 135906438665984 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.205064s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/10/items I0420 18:51:03.232589 135906438665984 atomicity.py:666] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/10/items I0420 18:51:03.851551 135906438665984 atomicity.py:666] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/10 I0420 18:51:05.280371 135906438665984 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/10`. I0420 18:51:05.281495 135906438665984 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 51.43 seconds @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/10 I0420 18:51:05.283187 135906438665984 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 50.802394s. I0420 18:51:05.283412 135901582137088 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0420 18:51:05.283522 135901582137088 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0420 18:51:05.283635 135901582137088 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=10] CheckpointManager Save Finalize is syncing with other hosts... I0420 18:51:05.286228 135901582137088 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=10] CheckpointManager Save Finalize is done on all hosts. I0420 18:51:05.286567 136025790359360 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=10][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=10. W0420 18:51:05.286754 136025790359360 checkpoint_manager.py:1452] Waiting for previous save to complete took 49.704930 seconds. If this number is high, consider checkpointing less frequently. I0420 18:51:05.288637 136025790359360 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 15 I0420 18:51:05.291134 136025790359360 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/15. I0420 18:51:05.954409 136025790359360 jax_array_handlers.py:360] Scheduling D2H of 69 prioritized jax.Array. I0420 18:51:05.954540 136025790359360 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 I0420 18:51:05.987149 136025790359360 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.035807s I0420 18:51:05.988150 136025790359360 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 28.282 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.05454301834106445 s) (per-host) I0420 18:51:05.988250 136025790359360 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.054665s (batch_requests_ready=0.006184s, total_serialization_initiated=0.047585s, others=0.000896s) I0420 18:51:05.988368 136025790359360 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.059810s (all_items=0.000021s, per_item={'items': '0.00002146'}, temp_paths=0.059788) I0420 18:51:05.989315 136025790359360 event_tracking.py:125] [process=0] [async] Finished blocking save in 0.70 seconds. Continuing save @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/15. I0420 18:51:05.989710 135901582137088 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-20 19:11:05.989673 I0420 18:51:05.992343 136025790359360 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=15] Starting CheckpointManager Save Finalize thread=save_finalize I0420 18:51:05.992661 135902114690816 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0420 18:51:05.992879 136025790359360 standard_logger.py:34] {'step': 15, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776711015.5817828, 'wait_for_prev_duration_secs': 49.70492959022522, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776711065.288702, 'checkpointer_blocking_duration_secs': 0.7011923789978027, 'get_old_steps_start_time': 1776711065.9899216, 'get_old_steps_duration_secs': 4.2438507080078125e-05, 'checkpoint_manager_blocking_start_time': 1776711015.5795083, 'checkpoint_manager_blocking_duration_secs': 50.41332149505615} I0420 18:51:05.993028 136025790359360 checkpointing.py:409] Started an asynchronous checkpoint save for step 15 I0420 18:51:06.186930 135906438665984 atomicity.py:140] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/15 I0420 18:51:08.021606 135906430273280 atomicity.py:140] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/15/items I0420 18:51:11.588958 135902131476224 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/15/items/array_metadatas/process_0 I0420 18:51:28.074351 136025790359360 metric_logger.py:196] completed step: 14, seconds: 0.148, TFLOP/s/device: 91.935, Tokens/s/device: 13857.501, total_weights: 65536, loss: 5.858, lm_loss: 5.858, perplexity: 350.067 I0420 18:51:28.085655 136025790359360 metric_logger.py:196] completed step: 15, seconds: 0.148, TFLOP/s/device: 91.915, Tokens/s/device: 13854.501, total_weights: 65536, loss: 5.713, lm_loss: 5.713, perplexity: 302.780 I0420 18:51:28.229680 136025790359360 metric_logger.py:196] completed step: 16, seconds: 72.788, TFLOP/s/device: 0.187, Tokens/s/device: 28.136, total_weights: 65536, loss: 5.599, lm_loss: 5.599, perplexity: 270.201 I0420 18:51:28.377008 136025790359360 metric_logger.py:196] completed step: 17, seconds: 0.009, TFLOP/s/device: 1491.780, Tokens/s/device: 224857.268, total_weights: 65536, loss: 5.511, lm_loss: 5.511, perplexity: 247.443 I0420 18:51:28.524863 136025790359360 metric_logger.py:196] completed step: 18, seconds: 0.145, TFLOP/s/device: 93.622, Tokens/s/device: 14111.778, total_weights: 65536, loss: 5.442, lm_loss: 5.442, perplexity: 230.909 I0420 18:51:28.672800 136025790359360 checkpointing.py:794] Waiting for step 19 to finish before checkpoint... I0420 18:51:28.673564 136025790359360 checkpointing.py:798] Waited 0.0007765293121337891 seconds for step 19 to finish before starting checkpointing. I0420 18:51:28.676194 136025790359360 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=15][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0420 18:51:51.040981 135902123083520 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 41.862355s (commit=40.636119s, array_metadata_write=1.226236s) I0420 18:51:51.042030 135901582137088 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 35.018 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 45.108402729034424 s) (per-host) I0420 18:51:51.042170 135901582137088 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 45.052316s. I0420 18:51:52.613860 135901582137088 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000359s. I0420 18:51:52.814482 135901582137088 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0420 18:51:53.469942 135901582137088 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.561593s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/15/items I0420 18:51:53.471845 135901582137088 atomicity.py:666] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/15/items I0420 18:51:53.996278 135901582137088 atomicity.py:666] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/15 I0420 18:51:55.510795 135901582137088 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/15`. I0420 18:51:55.511585 135901582137088 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 50.22 seconds @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/15 I0420 18:51:55.512976 135901582137088 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 49.523119s. I0420 18:51:55.513205 135902114690816 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0420 18:51:55.513323 135902114690816 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0420 18:51:55.513388 135902114690816 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=15] CheckpointManager Save Finalize is syncing with other hosts... I0420 18:51:55.515388 135902114690816 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=15] CheckpointManager Save Finalize is done on all hosts. I0420 18:51:55.515559 136025790359360 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=15][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=15. W0420 18:51:55.515689 136025790359360 checkpoint_manager.py:1452] Waiting for previous save to complete took 26.839502 seconds. If this number is high, consider checkpointing less frequently. I0420 18:51:55.517240 136025790359360 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 19 I0420 18:51:55.519189 136025790359360 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/19. I0420 18:51:55.828963 136025790359360 jax_array_handlers.py:360] Scheduling D2H of 69 prioritized jax.Array. I0420 18:51:55.829060 136025790359360 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 I0420 18:51:55.856470 136025790359360 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.029036s I0420 18:51:55.857368 136025790359360 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 41.161 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.03747701644897461 s) (per-host) I0420 18:51:55.857441 136025790359360 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.037570s (batch_requests_ready=0.002556s, total_serialization_initiated=0.034200s, others=0.000814s) I0420 18:51:55.857525 136025790359360 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.041776s (all_items=0.000017s, per_item={'items': '0.00001669'}, temp_paths=0.041760) I0420 18:51:55.858255 136025790359360 event_tracking.py:125] [process=0] [async] Finished blocking save in 0.34 seconds. Continuing save @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/19. I0420 18:51:55.858551 135902114690816 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-20 19:11:55.858517 I0420 18:51:55.860672 136025790359360 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=19] Starting CheckpointManager Save Finalize thread=save_finalize I0420 18:51:55.860921 135902663784192 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save. I0420 18:51:55.861049 136025790359360 standard_logger.py:34] {'step': 19, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776711088.6761618, 'wait_for_prev_duration_secs': 26.839502096176147, 'time_between_consecutive_saves_sec': 23.389867305755615, 'checkpointer_blocking_start_time': 1776711115.5172787, 'checkpointer_blocking_duration_secs': 0.341418981552124, 'get_old_steps_start_time': 1776711115.8587196, 'get_old_steps_duration_secs': 3.361701965332031e-05, 'checkpoint_manager_blocking_start_time': 1776711088.6737964, 'checkpoint_manager_blocking_duration_secs': 27.187220335006714} I0420 18:51:55.861176 136025790359360 checkpointing.py:409] Started an asynchronous checkpoint save for step 19 I0420 18:51:55.861219 136025790359360 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=19][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete. I0420 18:51:56.074684 135901582137088 atomicity.py:140] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/19 I0420 18:51:57.891779 135905381709568 atomicity.py:140] Creating tmp directory gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/19/items I0420 18:52:00.838746 135906430273280 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/19/items/array_metadatas/process_0 I0420 18:52:40.324063 135905868224256 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 41.281676s (commit=39.616073s, array_metadata_write=1.665604s) I0420 18:52:40.326598 135902114690816 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 35.491 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 44.506678342819214 s) (per-host) I0420 18:52:40.326708 135902114690816 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 44.468037s. I0420 18:52:42.053159 135902114690816 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389). I0420 18:52:42.095716 135902114690816 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000400s. I0420 18:52:42.695042 135902114690816 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.483847s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/19/items I0420 18:52:42.698050 135902114690816 atomicity.py:666] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/19/items I0420 18:52:43.241968 135902114690816 atomicity.py:666] Finalizing gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/19 I0420 18:52:44.697394 135902114690816 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/19`. I0420 18:52:44.698137 135902114690816 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 49.18 seconds @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_post_train_fixes_20260420_153543/nnx_xpk_feat_nnx_post_train_fixes_20260420_153543_16_zarr3_true/checkpoints/19 I0420 18:52:44.699558 135902114690816 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 48.840884s. I0420 18:52:44.699794 135902663784192 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save. I0420 18:52:44.699937 135902663784192 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save. I0420 18:52:44.700004 135902663784192 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=19] CheckpointManager Save Finalize is syncing with other hosts... I0420 18:52:44.702023 135902663784192 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=19] CheckpointManager Save Finalize is done on all hosts. I0420 18:52:44.702169 136025790359360 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=19][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=19. I0420 18:52:44.702279 136025790359360 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning. I0420 18:52:44.703270 136025790359360 metric_logger.py:196] completed step: 19, seconds: 0.147, TFLOP/s/device: 92.216, Tokens/s/device: 13899.824, total_weights: 65536, loss: 5.385, lm_loss: 5.385, perplexity: 218.045 Per train step: Total TFLOPs: 13.59 split as 93.93% learnable weight flops and 6.07% attention flops XPK End: Mon Apr 20 18:52:57 UTC 2026 EXIT_CODE=0