MaxView

XPK Start: Tue Apr 21 16:32:01 UTC 2026
PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.
2026-04-21 16:32:25.928833: 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)
I0421 16:32:26.108843 135478467512128 max_utils.py:273] Attempting to initialize the jax distributed system...
I0421 16:32:35.147510 135478467512128 distributed.py:149] Starting JAX distributed service on [::]:8482
I0421 16:32:35.149917 135478467512128 distributed.py:172] Connecting to JAX distributed service on mt-03-dropout-jeau6-slice-job-0-0.mt-03-dropout-jeau6:8482
I0421 16:32:36.747546 135478467512128 max_utils.py:284] Jax distributed system initialized!
I0421 16:32:42.799309 135478467512128 max_utils.py:800] System Information: Jax Version: 0.9.2
I0421 16:32:42.799409 135478467512128 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0421 16:32:42.799450 135478467512128 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
I0421 16:32:42.799489 135478467512128 train_utils.py:378] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0421 16:32:43.509258 135478467512128 maxtext_utils.py:1718] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0421 16:32:43.624270 135478467512128 checkpointing.py:688] Setting up checkpoint logger...
I0421 16:32:43.624385 135478467512128 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0421 16:32:43.624429 135478467512128 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0421 16:32:43.624627 135478467512128 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 0x7b36e9367860>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0421 16:32:46.538527 135478467512128 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0421 16:32:46.538816 135478467512128 checkpoint_manager.py:708] [process=3][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b21c8268d10>}, handler_registry=None
I0421 16:32:46.539068 135478467512128 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b21c8268d10>` 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`.
I0421 16:32:46.539118 135478467512128 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b21c8269f70>` 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`.
I0421 16:32:46.539152 135478467512128 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 0x7b21c8268d10>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b21c8268d10>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b21c8269f70>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b21c8269f70>}).
I0421 16:32:46.539462 135478467512128 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0421 16:32:46.539534 135478467512128 async_checkpointer.py:192] [process=3][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7b21c81b5b20> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0421 16:32:48.091537 135478467512128 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_161900/nnx_xpk_feat_nnx_trainstate_and_training_loop_20260421_161900_03_dropout/checkpoints
I0421 16:32:48.511124 135478467512128 checkpoint_manager.py:929] [process=3][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_20260421_161900/nnx_xpk_feat_nnx_trainstate_and_training_loop_20260421_161900_03_dropout/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7b21c82ab470>
I0421 16:32:48.511299 135478467512128 checkpointing.py:302] Checkpoint manager created!
I0421 16:32:48.889090 135478467512128 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0421 16:32:48.889207 135478467512128 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32

I0421 16:32:50.659828 135478467512128 nnx_decoders.py:465] nnx_decoders/carry Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0421 16:32:50.659939 135478467512128 nnx_decoders.py:465] nnx_decoders/carry Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0421 16:32:50.666055 135478467512128 nnx_decoders.py:465] Unknown Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0421 16:32:50.666110 135478467512128 nnx_decoders.py:465] Unknown Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0421 16:32:50.682392 135478467512128 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0421 16:32:50.682452 135478467512128 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0421 16:32:50.698119 135478467512128 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0421 16:32:50.698176 135478467512128 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0421 16:32:50.722233 135478467512128 attentions.py:1154] attentions/query Logical: bfloat16[32,128,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0421 16:32:50.722301 135478467512128 attentions.py:1154] attentions/query Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0421 16:32:50.738142 135478467512128 attentions.py:1155] attentions/key Logical: bfloat16[32,128,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0421 16:32:50.738202 135478467512128 attentions.py:1155] attentions/key Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0421 16:32:50.753896 135478467512128 attentions.py:1156] attentions/value Logical: bfloat16[32,128,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0421 16:32:50.753967 135478467512128 attentions.py:1156] attentions/value Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0421 16:32:50.784193 135478467512128 attentions.py:1197] attentions/out Logical: bfloat16[32,128,16,128]..................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0421 16:32:50.784268 135478467512128 attentions.py:1197] attentions/out Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0421 16:32:50.810128 135478467512128 linears.py:525] linears/x Logical: bfloat16[32,128,7168]....................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0421 16:32:50.810195 135478467512128 linears.py:525] linears/x Physical: bfloat16[32,128,7168]....................................... ('fsdp', None, None).
I0421 16:32:54.527944 135478467512128 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.
I0421 16:32:54.530188 135478467512128 metric_logger.py:301] number parameters: 1.104 billion
I0421 16:32:58.457819 135478467512128 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0421 16:32:58.650248 135478467512128 checkpointing.py:798] Waited 0.1924149990081787 seconds for step 0 to finish before starting checkpointing.
I0421 16:32:58.652966 135478467512128 checkpoint_manager.py:2009] [process=3][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0421 16:32:58.654988 135478467512128 checkpoint_manager.py:1512] [process=3] Saving checkpoint at step 0
I0421 16:32:58.656447 135478467512128 event_tracking.py:70] [process=3] [async] Started save checkpoint @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_161900/nnx_xpk_feat_nnx_trainstate_and_training_loop_20260421_161900_03_dropout/checkpoints/0.
I0421 16:32:59.368131 135478467512128 signaling_client.py:364] Using JaxDistributedSignalingClient
I0421 16:32:59.369226 135478467512128 jax_array_handlers.py:360] Scheduling D2H of 69 prioritized jax.Array.
I0421 16:32:59.369287 135478467512128 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
I0421 16:32:59.648163 135478467512128 base_pytree_checkpoint_handler.py:154] [process=3][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.280806s
I0421 16:32:59.648346 135478467512128 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/blocking_gbytes_per_sec: 5.320 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.28995776176452637 s) (per-host)
I0421 16:32:59.648405 135478467512128 base_pytree_checkpoint_handler.py:768] [process=3][thread=MainThread] Initiated Pytree async_save. Time taken: 0.290029s (batch_requests_ready=0.003028s, total_serialization_initiated=0.286921s, others=0.000080s)
I0421 16:32:59.648515 135478467512128 composite_checkpoint_handler.py:715] [process=3][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.294664s (all_items=0.000019s, per_item={'items': '0.00001907'}, temp_paths=0.294645)
I0421 16:32:59.649283 135478467512128 event_tracking.py:125] [process=3] [async] Finished blocking save in 0.99 seconds. Continuing save @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_161900/nnx_xpk_feat_nnx_trainstate_and_training_loop_20260421_161900_03_dropout/checkpoints/0.
I0421 16:32:59.649561 135348629788416 async_checkpointer.py:76] [process=3][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 16:52:59.649537
I0421 16:32:59.666823 135478467512128 checkpoint_manager.py:1560] [process=3][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 16:32:59.667136 135347046442752 async_checkpointer.py:280] [process=3][thread=save_finalize] Waiting for background save thread=async_save.
I0421 16:32:59.667294 135478467512128 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_161900/nnx_xpk_feat_nnx_trainstate_and_training_loop_20260421_161900_03_dropout/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776789178.652948, 'wait_for_prev_duration_secs': 5.984306335449219e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776789178.6550267, 'checkpointer_blocking_duration_secs': 0.9946444034576416, 'get_old_steps_start_time': 1776789179.6496952, 'get_old_steps_duration_secs': 2.9325485229492188e-05, 'checkpoint_manager_blocking_start_time': 1776789178.65064, 'checkpoint_manager_blocking_duration_secs': 1.0166163444519043}
I0421 16:32:59.667401 135478467512128 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0421 16:32:59.667474 135478467512128 max_utils.py:750] 
Memstats: After params initialized:
I0421 16:32:59.667520 135478467512128 max_utils.py:756] 	Using (GB) 0.41 / 31.25 (1.312000%) on TPU_10(process=3,(2,2,0,0))
I0421 16:32:59.667551 135478467512128 max_utils.py:756] 	Using (GB) 0.41 / 31.25 (1.312000%) on TPU_11(process=3,(3,2,0,0))
I0421 16:32:59.667576 135478467512128 max_utils.py:756] 	Using (GB) 0.41 / 31.25 (1.312000%) on TPU_14(process=3,(2,3,0,0))
I0421 16:32:59.667599 135478467512128 max_utils.py:756] 	Using (GB) 0.41 / 31.25 (1.312000%) on TPU_15(process=3,(3,3,0,0))
I0421 16:32:59.980684 135478467512128 metric_logger.py:196] completed step: 0, seconds: 3.927, TFLOP/s/device: 0.204, Tokens/s/device: 32.598, total_weights: 4096, loss: 10.874, lm_loss: 10.874, perplexity: 52805.141
I0421 16:33:00.081633 135478467512128 metric_logger.py:196] completed step: 1, seconds: 1.521, TFLOP/s/device: 0.527, Tokens/s/device: 84.156, total_weights: 4096, loss: 10.892, lm_loss: 10.892, perplexity: 53764.488
I0421 16:33:00.522689 135478467512128 metric_logger.py:196] completed step: 2, seconds: 0.032, TFLOP/s/device: 25.133, Tokens/s/device: 4016.820, total_weights: 4096, loss: 9.969, lm_loss: 9.969, perplexity: 21350.100
I0421 16:33:00.593362 135478467512128 metric_logger.py:196] completed step: 3, seconds: 0.431, TFLOP/s/device: 1.858, Tokens/s/device: 296.905, total_weights: 4096, loss: 9.125, lm_loss: 9.125, perplexity: 9186.262
I0421 16:33:00.735314 135478467512128 metric_logger.py:196] completed step: 4, seconds: 0.087, TFLOP/s/device: 9.203, Tokens/s/device: 1470.808, total_weights: 4096, loss: 8.469, lm_loss: 8.469, perplexity: 4765.735
I0421 16:33:00.742992 135478467512128 metric_logger.py:196] completed step: 5, seconds: 0.070, TFLOP/s/device: 11.412, Tokens/s/device: 1823.959, total_weights: 4096, loss: 7.908, lm_loss: 7.908, perplexity: 2719.938
I0421 16:33:02.703124    2817 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0421 16:33:04.712940 135347054835456 array_metadata_store.py:203] [process=3][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_161900/nnx_xpk_feat_nnx_trainstate_and_training_loop_20260421_161900_03_dropout/checkpoints/0/items/array_metadatas/process_3
I0421 16:33:10.968189 135478467512128 metric_logger.py:196] completed step: 6, seconds: 0.143, TFLOP/s/device: 5.606, Tokens/s/device: 896.013, total_weights: 4096, loss: 7.554, lm_loss: 7.554, perplexity: 1909.063
I0421 16:33:11.039015 135478467512128 metric_logger.py:196] completed step: 7, seconds: 10.156, TFLOP/s/device: 0.079, Tokens/s/device: 12.603, total_weights: 4096, loss: 7.313, lm_loss: 7.313, perplexity: 1500.206
I0421 16:33:11.109417 135478467512128 metric_logger.py:196] completed step: 8, seconds: 0.076, TFLOP/s/device: 10.542, Tokens/s/device: 1684.831, total_weights: 4096, loss: 7.091, lm_loss: 7.091, perplexity: 1200.912
I0421 16:33:11.478205 135478467512128 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0421 16:33:11.478892 135478467512128 checkpointing.py:798] Waited 0.0006999969482421875 seconds for step 9 to finish before starting checkpointing.
I0421 16:33:11.480900 135478467512128 checkpoint_manager.py:2020] [process=3][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 16:33:31.661408 135348629788416 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/gbytes_per_sec: 48.899 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 32.30299735069275 s) (per-host)
I0421 16:33:31.661512 135348629788416 async_checkpointer.py:90] [process=3][thread=async_save] 3 Handler Commit operations completed. Time taken: 32.011872s.
I0421 16:33:40.846852 135348629788416 async_checkpointer.py:160] [process=3][thread=async_save] Background save thread done. Time taken: 41.197194s.
I0421 16:33:40.847159 135347046442752 async_checkpointer.py:288] [process=3][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 16:33:40.847276 135347046442752 async_checkpointer.py:298] [process=3][thread=save_finalize] No errors found in background save thread=async_save.
I0421 16:33:40.847327 135347046442752 checkpoint_manager.py:2137] [process=3][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0421 16:33:40.848943 135347046442752 checkpoint_manager.py:2146] [process=3][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0421 16:33:40.849111 135478467512128 checkpoint_manager.py:2032] [process=3][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0421 16:33:40.849245 135478467512128 checkpoint_manager.py:1452] Waiting for previous save to complete took 29.368343 seconds. If this number is high, consider checkpointing less frequently.
I0421 16:33:40.851665 135478467512128 checkpoint_manager.py:1512] [process=3] Saving checkpoint at step 9
I0421 16:33:40.853809 135478467512128 event_tracking.py:70] [process=3] [async] Started save checkpoint @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_161900/nnx_xpk_feat_nnx_trainstate_and_training_loop_20260421_161900_03_dropout/checkpoints/9.
I0421 16:33:41.149095 135478467512128 jax_array_handlers.py:360] Scheduling D2H of 69 prioritized jax.Array.
I0421 16:33:41.149191 135478467512128 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
I0421 16:33:41.189958 135478467512128 base_pytree_checkpoint_handler.py:154] [process=3][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.042479s
I0421 16:33:41.190120 135478467512128 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/blocking_gbytes_per_sec: 31.105 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04959297180175781 s) (per-host)
I0421 16:33:41.190171 135478467512128 base_pytree_checkpoint_handler.py:768] [process=3][thread=MainThread] Initiated Pytree async_save. Time taken: 0.049653s (batch_requests_ready=0.002723s, total_serialization_initiated=0.046864s, others=0.000066s)
I0421 16:33:41.190270 135478467512128 composite_checkpoint_handler.py:715] [process=3][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.053644s (all_items=0.000016s, per_item={'items': '0.00001574'}, temp_paths=0.053628)
I0421 16:33:41.191024 135478467512128 event_tracking.py:125] [process=3] [async] Finished blocking save in 0.34 seconds. Continuing save @ gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_161900/nnx_xpk_feat_nnx_trainstate_and_training_loop_20260421_161900_03_dropout/checkpoints/9.
I0421 16:33:41.191361 135347046442752 async_checkpointer.py:76] [process=3][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 16:53:41.191325
I0421 16:33:41.193950 135478467512128 checkpoint_manager.py:1560] [process=3][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 16:33:41.194179 135347054835456 async_checkpointer.py:280] [process=3][thread=save_finalize] Waiting for background save thread=async_save.
I0421 16:33:41.194337 135478467512128 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_161900/nnx_xpk_feat_nnx_trainstate_and_training_loop_20260421_161900_03_dropout/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776789191.4808712, 'wait_for_prev_duration_secs': 29.368343114852905, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776789220.851704, 'checkpointer_blocking_duration_secs': 0.3397984504699707, 'get_old_steps_start_time': 1776789221.191527, 'get_old_steps_duration_secs': 2.956390380859375e-05, 'checkpoint_manager_blocking_start_time': 1776789191.4791427, 'checkpoint_manager_blocking_duration_secs': 29.715161323547363}
I0421 16:33:41.194454 135478467512128 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0421 16:33:41.194499 135478467512128 checkpoint_manager.py:2020] [process=3][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 16:33:47.328243 135325967865600 array_metadata_store.py:203] [process=3][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_161900/nnx_xpk_feat_nnx_trainstate_and_training_loop_20260421_161900_03_dropout/checkpoints/9/items/array_metadatas/process_3
I0421 16:34:22.832850 135347046442752 base_pytree_checkpoint_handler.py:130] [process=3] /jax/orbax/write/gbytes_per_sec: 37.887 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.69228482246399 s) (per-host)
I0421 16:34:22.832972 135347046442752 async_checkpointer.py:90] [process=3][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.641499s.
I0421 16:34:31.202517 135347046442752 async_checkpointer.py:160] [process=3][thread=async_save] Background save thread done. Time taken: 50.011026s.
I0421 16:34:31.202812 135347054835456 async_checkpointer.py:288] [process=3][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 16:34:31.202944 135347054835456 async_checkpointer.py:298] [process=3][thread=save_finalize] No errors found in background save thread=async_save.
I0421 16:34:31.202995 135347054835456 checkpoint_manager.py:2137] [process=3][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0421 16:34:31.204437 135347054835456 checkpoint_manager.py:2146] [process=3][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0421 16:34:31.204613 135478467512128 checkpoint_manager.py:2032] [process=3][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0421 16:34:31.204792 135478467512128 checkpoint_manager.py:2009] [process=3][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0421 16:34:31.205757 135478467512128 metric_logger.py:196] completed step: 9, seconds: 0.072, TFLOP/s/device: 11.177, Tokens/s/device: 1786.412, 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: Tue Apr 21 16:34:41 UTC 2026
EXIT_CODE=0