XPK Start: Mon Apr 20 19:15:08 UTC 2026
2026-04-20 19:15:12.260336: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:467] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
WARNING: All log messages before absl::InitializeLog() is called are written to STDERR
E0000 00:00:1776712512.273245 10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
E0000 00:00:1776712512.277090 10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
W0000 00:00:1776712512.288363 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776712512.288380 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776712512.288383 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1776712512.288385 10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
2026-04-20 19:15:31.485662: 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 19:15:32.012526 135861667362624 max_utils.py:273] Attempting to initialize the jax distributed system...
INFO:2026-04-20 19:15:41,053:jax._src.distributed:140: Starting JAX distributed service on [::]:8482
I0420 19:15:41.053012 135861667362624 distributed.py:140] Starting JAX distributed service on [::]:8482
INFO:2026-04-20 19:15:41,055:jax._src.distributed:157: Connecting to JAX distributed service on mt-02-synthetic-cxcmq-slice-job-0-0.mt-02-synthetic-cxcmq:8482
I0420 19:15:41.055556 135861667362624 distributed.py:157] Connecting to JAX distributed service on mt-02-synthetic-cxcmq-slice-job-0-0.mt-02-synthetic-cxcmq:8482
I0420 19:15:41.989252 135861667362624 max_utils.py:284] Jax distributed system initialized!
I0420 19:15:48.126265 135861667362624 max_utils.py:800] System Information: Jax Version: 0.8.1
I0420 19:15:48.126367 135861667362624 max_utils.py:801] System Information: Jaxlib Version: 0.8.1
I0420 19:15:48.126408 135861667362624 max_utils.py:802] System Information: Jax Backend: PJRT C API
TFRT TPU v6 lite
Built on Nov 12 2025 14:16:36 (1762985796) cl/831091709
I0420 19:15:48.126452 135861667362624 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0420 19:15:48.741401 135861667362624 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0420 19:15:48.741993 135861667362624 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0420 19:15:48.742184 135861667362624 checkpointing.py:688] Setting up checkpoint logger...
I0420 19:15:48.742236 135861667362624 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0420 19:15:48.742281 135861667362624 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None
I0420 19:15:48.742621 135861667362624 base_pytree_checkpoint_handler.py:415] Created BasePyTreeCheckpointHandler: use_ocdbt=True, use_zarr3=True, pytree_metadata_options=PyTreeMetadataOptions(support_rich_types=False), array_metadata_store=<orbax.checkpoint._src.metadata.array_metadata_store.Store object at 0x7b90417b7980>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0420 19:15:51.695160 135861667362624 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0420 19:15:51.695505 135861667362624 checkpoint_manager.py:709] [process=0][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b7e20471040>}, handler_registry=None
I0420 19:15:51.695744 135861667362624 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b7e20471040>` 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 19:15:51.695794 135861667362624 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b7e20475e50>` 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 19:15:51.695830 135861667362624 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 0x7b7e20471040>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b7e20471040>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b7e20475e50>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b7e20475e50>}).
I0420 19:15:51.696345 135861667362624 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33
I0420 19:15:51.696425 135861667362624 async_checkpointer.py:177] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7b7c904e4180> timeout: 600 secs and primary_host=0 for async checkpoint writes
I0420 19:15:52.821359 135861667362624 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints
I0420 19:15:53.259690 135861667362624 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=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/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7b7e20470470>
I0420 19:15:53.259869 135861667362624 checkpointing.py:302] Checkpoint manager created!
I0420 19:15:54.131731 135861667362624 nnx_wrappers.py:455] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0420 19:15:54.131838 135861667362624 nnx_wrappers.py:455] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 19:15:54.501915 135861667362624 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0420 19:15:54.502007 135861667362624 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 19:15:54.517811 135861667362624 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0420 19:15:54.517868 135861667362624 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0420 19:15:54.540463 135861667362624 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 19:15:54.540532 135861667362624 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 19:15:54.556367 135861667362624 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 19:15:54.556429 135861667362624 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 19:15:54.572229 135861667362624 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0420 19:15:54.572289 135861667362624 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 19:15:54.597323 135861667362624 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0420 19:15:54.597393 135861667362624 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0420 19:15:54.616880 135861667362624 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0420 19:15:54.616950 135861667362624 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0420 19:15:54.806063 135861667362624 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0420 19:15:54.806180 135861667362624 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0420 19:15:56.270841 135861667362624 maxtext_utils.py:1740] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: PartitionSpec('norm',)
Physical: (None,)
I0420 19:15:56.270970 135861667362624 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wi_0/kernel
Shape: float32[2048,16,7168]
Logical: PartitionSpec('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0420 19:15:56.271025 135861667362624 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wi_1/kernel
Shape: float32[2048,16,7168]
Logical: PartitionSpec('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0420 19:15:56.271102 135861667362624 maxtext_utils.py:1740] params/params/decoder/layers/mlp/wo/kernel
Shape: float32[7168,16,2048]
Logical: PartitionSpec('mlp', 'layers', 'embed')
Physical: (None, None, 'fsdp')
I0420 19:15:56.271168 135861667362624 maxtext_utils.py:1740] params/params/decoder/layers/post_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: PartitionSpec('norm', 'layers')
Physical: (None, None)
I0420 19:15:56.271209 135861667362624 maxtext_utils.py:1740] params/params/decoder/layers/pre_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: PartitionSpec('norm', 'layers')
Physical: (None, None)
I0420 19:15:56.271264 135861667362624 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/key/kernel
Shape: float32[2048,16,16,128]
Logical: PartitionSpec('embed', 'layers', 'kv_heads', 'kv_head_dim')
Physical: ('fsdp', None, None, None)
I0420 19:15:56.271317 135861667362624 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/out/kernel
Shape: float32[16,16,128,2048]
Logical: PartitionSpec('heads', 'layers', 'kv', 'embed')
Physical: (None, None, None, 'fsdp')
I0420 19:15:56.271359 135861667362624 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/query/kernel
Shape: float32[2048,16,16,128]
Logical: PartitionSpec('embed', 'layers', 'q_heads', 'kv')
Physical: ('fsdp', None, None, None)
I0420 19:15:56.271399 135861667362624 maxtext_utils.py:1740] params/params/decoder/layers/self_attention/value/kernel
Shape: float32[2048,16,16,128]
Logical: PartitionSpec('embed', 'layers', 'kv_heads', 'kv_head_dim')
Physical: ('fsdp', None, None, None)
I0420 19:15:56.271447 135861667362624 maxtext_utils.py:1740] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: PartitionSpec('embed', 'vocab')
Physical: ('fsdp', None)
I0420 19:15:56.271494 135861667362624 maxtext_utils.py:1740] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: PartitionSpec('vocab', 'embed')
Physical: (None, 'fsdp')
I0420 19:15:56.770393 135861667362624 train.py:158] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0420 19:15:56.770484 135861667362624 train.py:158] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0420 19:15:56.785313 135861667362624 train.py:165] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0420 19:15:56.785370 135861667362624 train.py:165] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0420 19:16:07.326842 135861667362624 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 19:16:07.332804 135861667362624 metric_logger.py:289] number parameters: 1.104 billion
I0420 19:16:18.729779 135861667362624 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0420 19:16:18.882313 135861667362624 checkpointing.py:798] Waited 0.15251517295837402 seconds for step 0 to finish before starting checkpointing.
I0420 19:16:18.884947 135861667362624 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0420 19:16:18.886809 135861667362624 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 0
I0420 19:16:18.889811 135861667362624 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0.
I0420 19:16:19.192071 135861667362624 signaling_client.py:364] Using JaxDistributedSignalingClient
I0420 19:16:19.501838 135861667362624 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array.
I0420 19:16:19.501932 135861667362624 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 19:16:19.512341 135737534048000 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0
I0420 19:16:19.745672 135861667362624 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.244825s
I0420 19:16:19.746611 135861667362624 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.829 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.5452208518981934 s) (per-host)
I0420 19:16:19.746687 135861667362624 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.545316s (batch_requests_ready=0.295834s, total_serialization_initiated=0.248623s, others=0.000859s)
I0420 19:16:19.746762 135861667362624 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.549162s (all_items=0.000018s, per_item={'items': '0.00001788'}, temp_paths=0.549144)
I0420 19:16:19.747701 135738056914688 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started.
I0420 19:16:19.747864 135861667362624 async_checkpointer.py:561] Finished blocking save. Time taken: 0.860984s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0.
I0420 19:16:20.146936 135861667362624 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0420 19:16:20.147313 135737525655296 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0420 19:16:20.147481 135861667362624 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776712578.884929, 'wait_for_prev_duration_secs': 6.508827209472656e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776712578.8868482, 'checkpointer_blocking_duration_secs': 0.8611218929290771, 'get_old_steps_start_time': 1776712579.7479894, 'get_old_steps_duration_secs': 2.4318695068359375e-05, 'checkpoint_manager_blocking_start_time': 1776712578.8829043, 'checkpoint_manager_blocking_duration_secs': 1.2645363807678223}
I0420 19:16:20.147587 135861667362624 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0420 19:16:20.147639 135861667362624 max_utils.py:750]
Memstats: After params initialized:
I0420 19:16:20.147694 135861667362624 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0))
I0420 19:16:20.147728 135861667362624 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0))
I0420 19:16:20.147756 135861667362624 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0))
I0420 19:16:20.147780 135861667362624 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0))
I0420 19:16:20.439305 135861667362624 metric_logger.py:185] completed step: 0, seconds: 11.052, TFLOP/s/device: 1.229, Tokens/s/device: 185.299, total_weights: 65536, loss: 10.874
I0420 19:16:20.442017 135861667362624 metric_logger.py:269] To see full metrics 'tensorboard --logdir=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/tensorboard/'
I0420 19:16:20.851906 135861667362624 metric_logger.py:185] completed step: 1, seconds: 1.708, TFLOP/s/device: 7.955, Tokens/s/device: 1199.129, total_weights: 65536, loss: 10.874
I0420 19:16:20.998116 135861667362624 metric_logger.py:185] completed step: 2, seconds: 0.413, TFLOP/s/device: 32.868, Tokens/s/device: 4954.267, total_weights: 65536, loss: 10.263
I0420 19:16:21.145579 135861667362624 metric_logger.py:185] completed step: 3, seconds: 0.009, TFLOP/s/device: 1510.688, Tokens/s/device: 227707.360, total_weights: 65536, loss: 9.734
I0420 19:16:40.229999 135738031949568 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776712581110762905, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776712581110762905, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0/_CHECKPOINT_METADATA
I0420 19:16:40.231158 135739664746240 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0/items
I0420 19:16:40.231590 135861667362624 metric_logger.py:185] completed step: 4, seconds: 0.147, TFLOP/s/device: 92.709, Tokens/s/device: 13974.085, total_weights: 65536, loss: 9.278
I0420 19:16:40.239772 135861667362624 metric_logger.py:185] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.146, Tokens/s/device: 13889.266, total_weights: 65536, loss: 8.892
I0420 19:16:40.386106 135861667362624 metric_logger.py:185] completed step: 6, seconds: 19.085, TFLOP/s/device: 0.712, Tokens/s/device: 107.310, total_weights: 65536, loss: 8.592
I0420 19:16:40.533452 135861667362624 metric_logger.py:185] completed step: 7, seconds: 0.006, TFLOP/s/device: 2267.167, Tokens/s/device: 341732.021, total_weights: 65536, loss: 8.384
I0420 19:16:40.680804 135861667362624 metric_logger.py:185] completed step: 8, seconds: 0.148, TFLOP/s/device: 92.065, Tokens/s/device: 13877.031, total_weights: 65536, loss: 8.255
I0420 19:16:40.827856 135861667362624 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0420 19:16:40.828566 135861667362624 checkpointing.py:798] Waited 0.0007302761077880859 seconds for step 9 to finish before starting checkpointing.
I0420 19:16:40.830541 135861667362624 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0420 19:16:41.516419 2592 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0420 19:16:43.218200 135738073700096 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0/items/array_metadatas/process_0
I0420 19:17:18.497225 135738065307392 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 36.981545s (commit=35.875933s, array_metadata_write=1.105613s)
I0420 19:17:18.498238 135738056914688 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 26.639 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 59.29681634902954 s) (per-host)
I0420 19:17:18.498357 135738056914688 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 58.750545s.
I0420 19:17:19.713732 135738056914688 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776712581110762905, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0/_CHECKPOINT_METADATA
I0420 19:17:21.342306 135738031949568 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776712581110762905, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0/_CHECKPOINT_METADATA
I0420 19:17:21.568346 135738056914688 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0420 19:17:21.672617 135738056914688 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000248s.
I0420 19:17:22.373958 135738056914688 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.086657s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0/items
I0420 19:17:22.375945 135738056914688 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0/items
I0420 19:17:22.840034 135738056914688 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0
I0420 19:17:25.426391 135738056914688 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0`.
I0420 19:17:25.427319 135738056914688 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 66.540458s. directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/0
I0420 19:17:25.428526 135738056914688 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 65.680713s.
I0420 19:17:25.428653 135737525655296 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0420 19:17:25.428725 135737525655296 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0420 19:17:25.428791 135737525655296 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0420 19:17:25.430969 135737525655296 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0420 19:17:25.431171 135861667362624 checkpoint_manager.py:2036] [process=0][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0420 19:17:25.431311 135861667362624 checkpoint_manager.py:1458] Waiting for previous save to complete took 44.600775 seconds. If this number is high, consider checkpointing less frequently.
I0420 19:17:25.433361 135861667362624 checkpoint_manager.py:1518] [process=0] Saving checkpoint at step 9
I0420 19:17:25.436925 135861667362624 async_checkpointer.py:452] [process=0] Started async saving checkpoint to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/9.
I0420 19:17:26.814646 135738056914688 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/9
I0420 19:17:26.838353 135861667362624 jax_array_handlers.py:358] Scheduling D2H of 39 prioritized jax.Array.
I0420 19:17:26.838456 135861667362624 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 19:17:26.865704 135861667362624 base_pytree_checkpoint_handler.py:153] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.028436s
I0420 19:17:26.866623 135861667362624 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.280 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.29215383529663086 s) (per-host)
I0420 19:17:26.866705 135861667362624 base_pytree_checkpoint_handler.py:737] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.292253s (batch_requests_ready=0.260701s, total_serialization_initiated=0.030705s, others=0.000847s)
I0420 19:17:26.866787 135861667362624 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.296221s (all_items=0.000017s, per_item={'items': '0.00001693'}, temp_paths=0.296204)
I0420 19:17:26.867690 135736956090112 async_checkpointer.py:79] [process=0][thread=async_save] Background save thread started.
I0420 19:17:26.867846 135861667362624 async_checkpointer.py:561] Finished blocking save. Time taken: 1.434412s. Continuing background save to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/9.
I0420 19:17:27.728231 135861667362624 checkpoint_manager.py:1566] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0420 19:17:27.728591 135737525655296 async_checkpointer.py:265] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0420 19:17:27.728770 135861667362624 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776712600.8305085, 'wait_for_prev_duration_secs': 44.60077524185181, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776712645.4334004, 'checkpointer_blocking_duration_secs': 1.43455171585083, 'get_old_steps_start_time': 1776712646.8679783, 'get_old_steps_duration_secs': 2.8133392333984375e-05, 'checkpoint_manager_blocking_start_time': 1776712600.828799, 'checkpoint_manager_blocking_duration_secs': 46.89993453025818}
I0420 19:17:27.728888 135861667362624 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0420 19:17:27.728935 135861667362624 checkpoint_manager.py:2024] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0420 19:17:28.977768 135738073700096 atomicity.py:137] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/9/items
I0420 19:17:35.458965 135737489504000 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/9/items/array_metadatas/process_0
I0420 19:18:13.953367 135736964482816 base_pytree_checkpoint_handler.py:1232] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 42.538029s (commit=41.463476s, array_metadata_write=1.074553s)
I0420 19:18:13.954478 135736956090112 base_pytree_checkpoint_handler.py:129] [process=0] /jax/checkpoint/write/gbytes_per_sec: 33.339 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 47.379976749420166 s) (per-host)
I0420 19:18:13.954598 135736956090112 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 47.086798s.
I0420 19:18:15.713245 135736956090112 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000285s.
I0420 19:18:15.983403 135736956090112 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0420 19:18:17.065771 135736956090112 base_pytree_checkpoint_handler.py:1356] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.942122s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/9/items
I0420 19:18:17.068482 135736956090112 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/9/items
I0420 19:18:17.498046 135736956090112 atomicity.py:608] Finalizing gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/9
I0420 19:18:18.828011 135736956090112 atomicity.py:790] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/9`.
I0420 19:18:18.828854 135736956090112 async_checkpointer.py:420] Finished async_save (blocking + background). Time taken: 53.395439s. directory=gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_set_defaults_true_20260420_190404/linen_xpk_feat_nnx_set_defaults_true_20260420_190404_02_synthetic/checkpoints/9
I0420 19:18:18.830198 135736956090112 async_checkpointer.py:144] [process=0][thread=async_save] Background save thread done. Time taken: 51.962396s.
I0420 19:18:18.830344 135737525655296 async_checkpointer.py:273] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0420 19:18:18.830404 135737525655296 async_checkpointer.py:283] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0420 19:18:18.830461 135737525655296 checkpoint_manager.py:2133] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0420 19:18:18.832368 135737525655296 checkpoint_manager.py:2142] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0420 19:18:18.832522 135861667362624 checkpoint_manager.py:2036] [process=0][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0420 19:18:18.832665 135861667362624 checkpoint_manager.py:2013] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0420 19:18:18.833404 135861667362624 metric_logger.py:185] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.201, Tokens/s/device: 13897.560, total_weights: 65536, loss: 8.179
Per train step:
Total TFLOPs: 13.59
split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Mon Apr 20 19:18:33 UTC 2026
EXIT_CODE=0