MaxView

XPK Start: Wed Apr 22 11:13:40 UTC 2026
PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.
2026-04-22 11:14:04.559060: 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)
I0422 11:14:04.739418 139312059557696 max_utils.py:273] Attempting to initialize the jax distributed system...
I0422 11:14:13.781253 139312059557696 distributed.py:149] Starting JAX distributed service on [::]:8482
I0422 11:14:13.783600 139312059557696 distributed.py:172] Connecting to JAX distributed service on mt-03-dropout-7zy2m-slice-job-0-0.mt-03-dropout-7zy2m:8482
I0422 11:14:15.386078 139312059557696 max_utils.py:284] Jax distributed system initialized!
I0422 11:14:21.487522 139312059557696 max_utils.py:800] System Information: Jax Version: 0.9.2
I0422 11:14:21.487622 139312059557696 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0422 11:14:21.487662 139312059557696 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
I0422 11:14:21.487698 139312059557696 train_utils.py:378] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0422 11:14:22.184597 139312059557696 maxtext_utils.py:1718] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0422 11:14:22.296610 139312059557696 checkpointing.py:688] Setting up checkpoint logger...
I0422 11:14:22.296725 139312059557696 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0422 11:14:22.296769 139312059557696 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0422 11:14:22.296959 139312059557696 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 0x7eb37afe9fd0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0422 11:14:26.011065 139312059557696 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0422 11:14:26.011344 139312059557696 checkpoint_manager.py:708] [process=6][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e9f50320650>}, handler_registry=None
I0422 11:14:26.011580 139312059557696 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e9f50320650>` 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`.
I0422 11:14:26.011629 139312059557696 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e9f5031b320>` 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`.
I0422 11:14:26.011664 139312059557696 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 0x7e9f50320650>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e9f50320650>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e9f5031b320>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e9f5031b320>}).
I0422 11:14:26.011976 139312059557696 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0422 11:14:26.012047 139312059557696 async_checkpointer.py:192] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7e9f50789b20> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0422 11:14:27.167983 139312059557696 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260422_093106/nnx_xpk_feat_nnx_trainstate_and_training_loop_20260422_093106_03_dropout/checkpoints
I0422 11:14:27.170363 139312059557696 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_trainstate_and_training_loop_20260422_093106/nnx_xpk_feat_nnx_trainstate_and_training_loop_20260422_093106_03_dropout/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7eb2582806b0>
I0422 11:14:27.170484 139312059557696 checkpointing.py:302] Checkpoint manager created!
I0422 11:14:27.547897 139312059557696 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0422 11:14:27.548012 139312059557696 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32

I0422 11:14:29.325687 139312059557696 nnx_decoders.py:465] nnx_decoders/carry Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0422 11:14:29.325782 139312059557696 nnx_decoders.py:465] nnx_decoders/carry Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0422 11:14:29.331866 139312059557696 nnx_decoders.py:465] Unknown Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0422 11:14:29.331922 139312059557696 nnx_decoders.py:465] Unknown Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0422 11:14:29.348310 139312059557696 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0422 11:14:29.348369 139312059557696 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0422 11:14:29.363981 139312059557696 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0422 11:14:29.364041 139312059557696 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0422 11:14:29.393256 139312059557696 attentions.py:1154] attentions/query Logical: bfloat16[32,128,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0422 11:14:29.393352 139312059557696 attentions.py:1154] attentions/query Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0422 11:14:29.408969 139312059557696 attentions.py:1155] attentions/key Logical: bfloat16[32,128,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0422 11:14:29.409026 139312059557696 attentions.py:1155] attentions/key Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0422 11:14:29.424628 139312059557696 attentions.py:1156] attentions/value Logical: bfloat16[32,128,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0422 11:14:29.424687 139312059557696 attentions.py:1156] attentions/value Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0422 11:14:29.454416 139312059557696 attentions.py:1197] attentions/out Logical: bfloat16[32,128,16,128]..................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0422 11:14:29.454488 139312059557696 attentions.py:1197] attentions/out Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0422 11:14:29.480351 139312059557696 linears.py:525] linears/x Logical: bfloat16[32,128,7168]....................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0422 11:14:29.480417 139312059557696 linears.py:525] linears/x Physical: bfloat16[32,128,7168]....................................... ('fsdp', None, None).
I0422 11:14:33.226291 139312059557696 max_utils.py:791] Total memory size: 0.8 GB, Output size: 0.4 GB, Temp size: 0.4 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB.
I0422 11:14:33.228546 139312059557696 metric_logger.py:301] number parameters: 1.104 billion
I0422 11:14:37.226366 139312059557696 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0422 11:14:37.512302 139312059557696 checkpointing.py:798] Waited 0.28592562675476074 seconds for step 0 to finish before starting checkpointing.
I0422 11:14:37.514521 139312059557696 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0422 11:14:37.516489 139312059557696 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 0
I0422 11:14:37.518049 139312059557696 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260422_093106/nnx_xpk_feat_nnx_trainstate_and_training_loop_20260422_093106_03_dropout/checkpoints/0.
I0422 11:14:38.282492 139312059557696 signaling_client.py:364] Using JaxDistributedSignalingClient
I0422 11:14:38.283574 139312059557696 jax_array_handlers.py:360] Scheduling D2H of 69 prioritized jax.Array.
I0422 11:14:38.283632 139312059557696 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
I0422 11:14:38.562012 139312059557696 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.280289s
I0422 11:14:38.562210 139312059557696 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 5.330 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2894175052642822 s) (per-host)
I0422 11:14:38.562277 139312059557696 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.289496s (batch_requests_ready=0.003014s, total_serialization_initiated=0.286396s, others=0.000086s)
I0422 11:14:38.562382 139312059557696 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.293721s (all_items=0.000018s, per_item={'items': '0.00001764'}, temp_paths=0.293703)
I0422 11:14:38.563236 139312059557696 event_tracking.py:125] [process=6] [async] Finished blocking save in 1.05 seconds. Continuing save @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260422_093106/nnx_xpk_feat_nnx_trainstate_and_training_loop_20260422_093106_03_dropout/checkpoints/0.
I0422 11:14:38.563544 139184241473280 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-22 11:34:38.563508
I0422 11:14:38.576696 139312059557696 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0422 11:14:38.577003 139183687460608 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0422 11:14:38.577173 139312059557696 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260422_093106/nnx_xpk_feat_nnx_trainstate_and_training_loop_20260422_093106_03_dropout/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776856477.5145032, 'wait_for_prev_duration_secs': 6.699562072753906e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776856477.5165346, 'checkpointer_blocking_duration_secs': 1.0471606254577637, 'get_old_steps_start_time': 1776856478.5637207, 'get_old_steps_duration_secs': 3.0040740966796875e-05, 'checkpoint_manager_blocking_start_time': 1776856477.5126896, 'checkpoint_manager_blocking_duration_secs': 1.0644433498382568}
I0422 11:14:38.577284 139312059557696 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0422 11:14:38.577361 139312059557696 max_utils.py:750] 
Memstats: After params initialized:
I0422 11:14:38.577411 139312059557696 max_utils.py:756] 	Using (GB) 0.41 / 31.25 (1.312000%) on TPU_24(process=6,(0,6,0,0))
I0422 11:14:38.577443 139312059557696 max_utils.py:756] 	Using (GB) 0.41 / 31.25 (1.312000%) on TPU_25(process=6,(1,6,0,0))
I0422 11:14:38.577470 139312059557696 max_utils.py:756] 	Using (GB) 0.41 / 31.25 (1.312000%) on TPU_28(process=6,(0,7,0,0))
I0422 11:14:38.577494 139312059557696 max_utils.py:756] 	Using (GB) 0.41 / 31.25 (1.312000%) on TPU_29(process=6,(1,7,0,0))
I0422 11:14:38.892768 139312059557696 metric_logger.py:196] completed step: 0, seconds: 3.997, TFLOP/s/device: 0.200, Tokens/s/device: 32.026, total_weights: 4096, loss: 10.874, lm_loss: 10.874, perplexity: 52805.141
I0422 11:14:38.988635 139312059557696 metric_logger.py:196] completed step: 1, seconds: 1.664, TFLOP/s/device: 0.481, Tokens/s/device: 76.901, total_weights: 4096, loss: 10.892, lm_loss: 10.892, perplexity: 53764.488
I0422 11:14:39.410166 139312059557696 metric_logger.py:196] completed step: 2, seconds: 0.027, TFLOP/s/device: 30.148, Tokens/s/device: 4818.370, total_weights: 4096, loss: 9.969, lm_loss: 9.969, perplexity: 21350.100
I0422 11:14:39.480668 139312059557696 metric_logger.py:196] completed step: 3, seconds: 0.411, TFLOP/s/device: 1.947, Tokens/s/device: 311.206, total_weights: 4096, loss: 9.125, lm_loss: 9.125, perplexity: 9186.262
I0422 11:14:39.622272 139312059557696 metric_logger.py:196] completed step: 4, seconds: 0.088, TFLOP/s/device: 9.145, Tokens/s/device: 1461.604, total_weights: 4096, loss: 8.469, lm_loss: 8.469, perplexity: 4765.735
I0422 11:14:39.629700 139312059557696 metric_logger.py:196] completed step: 5, seconds: 0.070, TFLOP/s/device: 11.456, Tokens/s/device: 1831.030, total_weights: 4096, loss: 7.908, lm_loss: 7.908, perplexity: 2719.938
I0422 11:14:42.186694    2790 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0422 11:14:44.683358 139184195323648 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_trainstate_and_training_loop_20260422_093106/nnx_xpk_feat_nnx_trainstate_and_training_loop_20260422_093106_03_dropout/checkpoints/0/items/array_metadatas/process_6
I0422 11:14:49.222923 139312059557696 metric_logger.py:196] completed step: 6, seconds: 0.142, TFLOP/s/device: 5.629, Tokens/s/device: 899.647, total_weights: 4096, loss: 7.554, lm_loss: 7.554, perplexity: 1909.063
I0422 11:14:49.293432 139312059557696 metric_logger.py:196] completed step: 7, seconds: 9.524, TFLOP/s/device: 0.084, Tokens/s/device: 13.440, total_weights: 4096, loss: 7.313, lm_loss: 7.313, perplexity: 1500.206
I0422 11:14:49.364153 139312059557696 metric_logger.py:196] completed step: 8, seconds: 0.076, TFLOP/s/device: 10.470, Tokens/s/device: 1673.312, total_weights: 4096, loss: 7.091, lm_loss: 7.091, perplexity: 1200.912
I0422 11:14:49.435463 139312059557696 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0422 11:14:49.436140 139312059557696 checkpointing.py:798] Waited 0.0006883144378662109 seconds for step 9 to finish before starting checkpointing.
I0422 11:14:49.438048 139312059557696 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0422 11:15:11.507495 139184241473280 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 47.529 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 33.23466610908508 s) (per-host)
I0422 11:15:11.507615 139184241473280 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 32.943961s.
I0422 11:15:21.039735 139184241473280 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 42.476063s.
I0422 11:15:21.040026 139183687460608 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0422 11:15:21.040161 139183687460608 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0422 11:15:21.040213 139183687460608 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0422 11:15:21.041819 139183687460608 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0422 11:15:21.041985 139312059557696 checkpoint_manager.py:2032] [process=6][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0422 11:15:21.042125 139312059557696 checkpoint_manager.py:1452] Waiting for previous save to complete took 31.604065 seconds. If this number is high, consider checkpointing less frequently.
I0422 11:15:21.044165 139312059557696 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 9
I0422 11:15:21.046249 139312059557696 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260422_093106/nnx_xpk_feat_nnx_trainstate_and_training_loop_20260422_093106_03_dropout/checkpoints/9.
I0422 11:15:21.345882 139312059557696 jax_array_handlers.py:360] Scheduling D2H of 69 prioritized jax.Array.
I0422 11:15:21.345974 139312059557696 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
I0422 11:15:21.381757 139312059557696 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.037674s
I0422 11:15:21.381927 139312059557696 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 34.527 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.0446772575378418 s) (per-host)
I0422 11:15:21.381983 139312059557696 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.044740s (batch_requests_ready=0.002663s, total_serialization_initiated=0.042006s, others=0.000071s)
I0422 11:15:21.382122 139312059557696 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.048850s (all_items=0.000016s, per_item={'items': '0.00001645'}, temp_paths=0.048834)
I0422 11:15:21.382872 139312059557696 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.34 seconds. Continuing save @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260422_093106/nnx_xpk_feat_nnx_trainstate_and_training_loop_20260422_093106_03_dropout/checkpoints/9.
I0422 11:15:21.383179 139183687460608 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-22 11:35:21.383147
I0422 11:15:21.385187 139312059557696 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0422 11:15:21.385425 139159305918208 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0422 11:15:21.385555 139312059557696 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260422_093106/nnx_xpk_feat_nnx_trainstate_and_training_loop_20260422_093106_03_dropout/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776856489.4380198, 'wait_for_prev_duration_secs': 31.60406517982483, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776856521.0442023, 'checkpointer_blocking_duration_secs': 0.3391242027282715, 'get_old_steps_start_time': 1776856521.3833532, 'get_old_steps_duration_secs': 3.266334533691406e-05, 'checkpoint_manager_blocking_start_time': 1776856489.436371, 'checkpoint_manager_blocking_duration_secs': 31.94914746284485}
I0422 11:15:21.385668 139312059557696 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0422 11:15:21.385716 139312059557696 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0422 11:15:27.467939 139184195323648 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_trainstate_and_training_loop_20260422_093106/nnx_xpk_feat_nnx_trainstate_and_training_loop_20260422_093106_03_dropout/checkpoints/9/items/array_metadatas/process_6
I0422 11:16:02.102790 139183687460608 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 38.748 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 40.765507221221924 s) (per-host)
I0422 11:16:02.102912 139183687460608 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 40.719621s.
I0422 11:16:11.613693 139183687460608 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 50.230386s.
I0422 11:16:11.613925 139159305918208 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0422 11:16:11.613983 139159305918208 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0422 11:16:11.614033 139159305918208 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0422 11:16:11.616843 139159305918208 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0422 11:16:11.617009 139312059557696 checkpoint_manager.py:2032] [process=6][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0422 11:16:11.617194 139312059557696 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0422 11:16:11.618189 139312059557696 metric_logger.py:196] completed step: 9, seconds: 0.072, TFLOP/s/device: 11.194, Tokens/s/device: 1789.134, total_weights: 4096, loss: 7.016, lm_loss: 7.016, perplexity: 1113.843
Per train step:
 Total TFLOPs: 0.80 
 split as 99.60% learnable weight flops and 0.40% attention flops
XPK End: Wed Apr 22 11:16:20 UTC 2026
EXIT_CODE=0