XPK Start: Sat Apr 18 00:36:14 UTC 2026
PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.
2026-04-18 00:36:38.871777: 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)
I0418 00:36:39.051128 134828115330880 max_utils.py:273] Attempting to initialize the jax distributed system...
I0418 00:36:48.091327 134828115330880 distributed.py:149] Starting JAX distributed service on [::]:8482
I0418 00:36:48.093760 134828115330880 distributed.py:172] Connecting to JAX distributed service on mt-08-checkpoint-async--3d118-slice-job-0-0.mt-08-checkpoint-async--3d118:8482
I0418 00:36:49.238714 134828115330880 max_utils.py:284] Jax distributed system initialized!
I0418 00:36:54.310787 134828115330880 max_utils.py:800] System Information: Jax Version: 0.9.2
I0418 00:36:54.310917 134828115330880 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0418 00:36:54.310959 134828115330880 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
I0418 00:36:54.310994 134828115330880 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0418 00:36:55.004436 134828115330880 maxtext_utils.py:1551] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0418 00:36:55.004714 134828115330880 checkpointing.py:677] Setting up checkpoint logger...
I0418 00:36:55.004767 134828115330880 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0418 00:36:55.004814 134828115330880 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0418 00:36:55.005172 134828115330880 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 0x7a9f84914f50>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0418 00:36:56.695516 134828115330880 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0418 00:36:56.695863 134828115330880 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 0x7a8c3809a150>}, handler_registry=None
I0418 00:36:56.696105 134828115330880 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7a8c3809a150>` 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`.
I0418 00:36:56.696154 134828115330880 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7a8c182cc0b0>` 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`.
I0418 00:36:56.696191 134828115330880 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 0x7a8c3809a150>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7a8c3809a150>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7a8c182cc0b0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7a8c182cc0b0>}).
I0418 00:36:56.696506 134828115330880 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0418 00:36:56.696573 134828115330880 async_checkpointer.py:192] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7a8b687bbec0> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0418 00:36:59.074625 134710055376640 checkpoint.py:228] Read Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776472338542414329, 'commit_timestamp_nsecs': 1776472381789972436, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/0/_CHECKPOINT_METADATA
I0418 00:36:59.750432 134710055376640 checkpoint_manager.py:1871] Missing metrics for step 0
I0418 00:37:00.567810 134828115330880 checkpoint_manager.py:1812] Found 3 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints
I0418 00:37:02.352595 134828115330880 checkpoint_manager.py:929] [process=0][thread=MainThread] CheckpointManager created, primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=5), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7a8c182c8590>
I0418 00:37:02.352771 134828115330880 checkpointing.py:301] Checkpoint manager created!
I0418 00:37:03.260783 134828115330880 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0418 00:37:03.260913 134828115330880 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 00:37:03.638309 134828115330880 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 00:37:03.638402 134828115330880 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 00:37:03.654647 134828115330880 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 00:37:03.654707 134828115330880 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 00:37:03.678070 134828115330880 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0418 00:37:03.678138 134828115330880 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 00:37:03.694532 134828115330880 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0418 00:37:03.694592 134828115330880 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 00:37:03.710640 134828115330880 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0418 00:37:03.710700 134828115330880 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 00:37:03.735451 134828115330880 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0418 00:37:03.735528 134828115330880 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 00:37:03.757331 134828115330880 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0418 00:37:03.757398 134828115330880 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0418 00:37:03.979579 134828115330880 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0418 00:37:03.979696 134828115330880 checkpointing.py:581] restoring from this run's directory step 9
I0418 00:37:04.664639 134828115330880 event_tracking.py:70] [process=0] [sync] Started load checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/9.
I0418 00:37:06.037478 134828115330880 checkpointer.py:307] Restoring checkpoint from gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/9.
I0418 00:37:06.037928 134828115330880 event_tracking.py:125] [process=0] [sync] Finished blocking load in 1.37 seconds. Continuing load @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/9.
I0418 00:37:08.816950 1407 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0418 00:37:14.549191 134828115330880 jax_array_handlers.py:843] [process=0] /jax/orbax/read/worker/io/requested throughput: 275.395 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 5.738204479217529 s) (per-host)
I0418 00:37:14.561655 134828115330880 base_pytree_checkpoint_handler.py:130] [process=0] /jax/checkpoint/read/gbytes_per_sec: 240.748 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 6.573657274246216 s) (per-host)
I0418 00:37:15.343117 134828115330880 event_tracking.py:138] [process=0] [sync] Finished load in 10.68 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/9
I0418 00:37:15.343272 134828115330880 standard_logger.py:34] {'step': 9, 'event_type': 'restore', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints', 'checkpointer_start_time': 1776472624.6645417, 'checkpointer_duration_secs': 10.67870020866394, 'checkpoint_manager_start_time': 1776472623.979952, 'checkpoint_manager_duration_secs': 11.363292217254639}
fsdp: 32
I0418 00:37:15.756451 134828115330880 maxtext_utils.py:1654] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: P('norm',)
Physical: (None,)
I0418 00:37:15.756574 134828115330880 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wi_0/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0418 00:37:15.756626 134828115330880 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wi_1/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0418 00:37:15.756683 134828115330880 maxtext_utils.py:1654] params/params/decoder/layers/mlp/wo/kernel
Shape: float32[7168,16,2048]
Logical: P('mlp', 'layers', 'embed')
Physical: (None, None, 'fsdp')
I0418 00:37:15.756735 134828115330880 maxtext_utils.py:1654] params/params/decoder/layers/post_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0418 00:37:15.756771 134828115330880 maxtext_utils.py:1654] params/params/decoder/layers/pre_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0418 00:37:15.756820 134828115330880 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/key/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'kv_heads', 'kv_head_dim')
Physical: ('fsdp', None, None, None)
I0418 00:37:15.756899 134828115330880 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/out/kernel
Shape: float32[16,16,128,2048]
Logical: P('heads', 'layers', 'kv', 'embed')
Physical: (None, None, None, 'fsdp')
I0418 00:37:15.756943 134828115330880 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/query/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'q_heads', 'kv')
Physical: ('fsdp', None, None, None)
I0418 00:37:15.756979 134828115330880 maxtext_utils.py:1654] params/params/decoder/layers/self_attention/value/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'kv_heads', 'kv_head_dim')
Physical: ('fsdp', None, None, None)
I0418 00:37:15.757024 134828115330880 maxtext_utils.py:1654] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: P('embed_vocab', 'vocab')
Physical: ('fsdp', None)
I0418 00:37:15.757069 134828115330880 maxtext_utils.py:1654] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: P('vocab', 'embed_vocab')
Physical: (None, 'fsdp')
I0418 00:37:16.245829 134828115330880 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0418 00:37:16.245934 134828115330880 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0418 00:37:16.261088 134828115330880 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0418 00:37:16.261151 134828115330880 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0418 00:37:27.159744 134828115330880 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.
I0418 00:37:27.167701 134828115330880 metric_logger.py:301] number parameters: 1.104 billion
I0418 00:37:39.032785 134828115330880 checkpointing.py:772] Waiting for step 10 to finish before checkpoint...
I0418 00:37:39.186052 134828115330880 checkpointing.py:776] Waited 0.15325498580932617 seconds for step 10 to finish before starting checkpointing.
I0418 00:37:39.188507 134828115330880 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 00:37:39.190416 134828115330880 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 10
I0418 00:37:39.191827 134828115330880 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/10.
I0418 00:37:39.922309 134828115330880 signaling_client.py:364] Using JaxDistributedSignalingClient
I0418 00:37:39.934528 134828115330880 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 00:37:39.934591 134828115330880 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
I0418 00:37:40.184239 134709837297408 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/10
I0418 00:37:40.210833 134828115330880 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.276934s
I0418 00:37:40.211617 134828115330880 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 5.473 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2818334102630615 s) (per-host)
I0418 00:37:40.211696 134828115330880 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.281929s (batch_requests_ready=0.002276s, total_serialization_initiated=0.278965s, others=0.000689s)
I0418 00:37:40.211779 134828115330880 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.285806s (all_items=0.000017s, per_item={'items': '0.00001669'}, temp_paths=0.285789)
I0418 00:37:40.212547 134828115330880 event_tracking.py:125] [process=0] [async] Finished blocking save in 1.02 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/10.
I0418 00:37:40.212921 134709921158912 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 00:57:40.212884
I0418 00:37:40.218261 134828115330880 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=10] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 00:37:40.218516 134704258868992 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0418 00:37:40.218666 134828115330880 standard_logger.py:34] {'step': 10, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776472659.1884894, 'wait_for_prev_duration_secs': 6.413459777832031e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776472659.1904535, 'checkpointer_blocking_duration_secs': 1.0226178169250488, 'get_old_steps_start_time': 1776472660.2130942, 'get_old_steps_duration_secs': 3.0517578125e-05, 'checkpoint_manager_blocking_start_time': 1776472659.1865308, 'checkpoint_manager_blocking_duration_secs': 1.032099723815918}
I0418 00:37:40.218770 134828115330880 checkpointing.py:408] Started an asynchronous checkpoint save for step 10
I0418 00:37:40.218821 134828115330880 max_utils.py:750]
Memstats: After params initialized:
I0418 00:37:40.218882 134828115330880 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0))
I0418 00:37:40.218916 134828115330880 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0))
I0418 00:37:40.218945 134828115330880 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0))
I0418 00:37:40.218970 134828115330880 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0))
I0418 00:37:40.555990 134828115330880 metric_logger.py:196] completed step: 10, seconds: 11.551, TFLOP/s/device: 1.176, Tokens/s/device: 177.298, total_weights: 65536, loss: 6.833, lm_loss: 6.833, perplexity: 927.869
I0418 00:37:40.701053 134828115330880 metric_logger.py:196] completed step: 11, seconds: 1.521, TFLOP/s/device: 8.931, Tokens/s/device: 1346.230, total_weights: 65536, loss: 6.527, lm_loss: 6.527, perplexity: 683.659
I0418 00:37:40.849029 134828115330880 metric_logger.py:196] completed step: 12, seconds: 0.011, TFLOP/s/device: 1248.587, Tokens/s/device: 188200.698, total_weights: 65536, loss: 6.265, lm_loss: 6.265, perplexity: 526.068
I0418 00:37:40.996282 134828115330880 metric_logger.py:196] completed step: 13, seconds: 0.145, TFLOP/s/device: 93.579, Tokens/s/device: 14105.266, total_weights: 65536, loss: 6.045, lm_loss: 6.045, perplexity: 422.001
I0418 00:37:41.005867 134828115330880 checkpointing.py:772] Waiting for step 15 to finish before checkpoint...
I0418 00:37:41.290495 134828115330880 checkpointing.py:776] Waited 0.28452420234680176 seconds for step 15 to finish before starting checkpointing.
I0418 00:37:41.293542 134828115330880 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=10][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 00:37:42.357238 134704762865408 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/10/items
I0418 00:37:42.539605 134704250476288 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776472662100553067, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776472662100553067, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/10/_CHECKPOINT_METADATA
I0418 00:37:46.923492 134704746080000 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/10/items/array_metadatas/process_0
I0418 00:38:25.386301 134704267261696 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 41.467807s (commit=39.429471s, array_metadata_write=2.038336s)
I0418 00:38:25.387779 134709921158912 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 34.748 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 45.45798969268799 s) (per-host)
I0418 00:38:25.387919 134709921158912 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 45.174883s.
I0418 00:38:27.892861 134709921158912 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0418 00:38:28.102257 134709921158912 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000418s.
I0418 00:38:28.565124 134704250476288 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776472662100553067, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/10/_CHECKPOINT_METADATA
I0418 00:38:28.695746 134709921158912 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.620801s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/10/items
I0418 00:38:28.698069 134709921158912 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/10/items
I0418 00:38:29.228608 134709921158912 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/10
I0418 00:38:32.338219 134709921158912 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/10`.
I0418 00:38:32.339396 134709921158912 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 53.15 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/10
I0418 00:38:32.341043 134709921158912 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 52.128005s.
I0418 00:38:32.341231 134704258868992 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 00:38:32.341343 134704258868992 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0418 00:38:32.341438 134704258868992 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=10] CheckpointManager Save Finalize is syncing with other hosts...
I0418 00:38:32.343479 134704258868992 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=10] CheckpointManager Save Finalize is done on all hosts.
I0418 00:38:32.343723 134828115330880 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.
W0418 00:38:32.343831 134828115330880 checkpoint_manager.py:1452] Waiting for previous save to complete took 51.050299 seconds. If this number is high, consider checkpointing less frequently.
I0418 00:38:32.345805 134828115330880 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 15
I0418 00:38:32.348316 134828115330880 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/15.
I0418 00:38:33.091517 134828115330880 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 00:38:33.091651 134828115330880 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
I0418 00:38:33.119476 134828115330880 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.028989s
I0418 00:38:33.120400 134828115330880 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 40.506 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.03808236122131348 s) (per-host)
I0418 00:38:33.120521 134828115330880 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.038217s (batch_requests_ready=0.004608s, total_serialization_initiated=0.032786s, others=0.000823s)
I0418 00:38:33.120671 134828115330880 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.042962s (all_items=0.000021s, per_item={'items': '0.00002074'}, temp_paths=0.042941)
I0418 00:38:33.121762 134828115330880 event_tracking.py:125] [process=0] [async] Finished blocking save in 0.78 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/15.
I0418 00:38:33.122140 134704258868992 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 00:58:33.122114
I0418 00:38:33.124567 134828115330880 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=15] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 00:38:33.124917 134704267261696 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0418 00:38:33.125145 134828115330880 standard_logger.py:34] {'step': 15, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776472661.2934978, 'wait_for_prev_duration_secs': 51.05029892921448, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776472712.3458824, 'checkpointer_blocking_duration_secs': 0.7763912677764893, 'get_old_steps_start_time': 1776472713.1223023, 'get_old_steps_duration_secs': 4.57763671875e-05, 'checkpoint_manager_blocking_start_time': 1776472661.2911785, 'checkpoint_manager_blocking_duration_secs': 51.833914041519165}
I0418 00:38:33.125297 134828115330880 checkpointing.py:408] Started an asynchronous checkpoint save for step 15
I0418 00:38:33.765433 134709921158912 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/15
I0418 00:38:36.827738 134704762865408 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/15/items
I0418 00:38:43.181817 134708578981632 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/15/items/array_metadatas/process_0
I0418 00:38:55.207620 134828115330880 metric_logger.py:196] completed step: 14, seconds: 0.148, TFLOP/s/device: 91.722, Tokens/s/device: 13825.320, total_weights: 65536, loss: 5.864, lm_loss: 5.864, perplexity: 352.193
I0418 00:38:55.217045 134828115330880 metric_logger.py:196] completed step: 15, seconds: 0.147, TFLOP/s/device: 92.205, Tokens/s/device: 13898.126, total_weights: 65536, loss: 5.719, lm_loss: 5.719, perplexity: 304.661
I0418 00:38:55.362573 134828115330880 metric_logger.py:196] completed step: 16, seconds: 74.210, TFLOP/s/device: 0.183, Tokens/s/device: 27.597, total_weights: 65536, loss: 5.605, lm_loss: 5.605, perplexity: 271.872
I0418 00:38:55.509952 134828115330880 metric_logger.py:196] completed step: 17, seconds: 0.007, TFLOP/s/device: 1829.177, Tokens/s/device: 275713.516, total_weights: 65536, loss: 5.517, lm_loss: 5.517, perplexity: 248.975
I0418 00:38:55.657162 134828115330880 metric_logger.py:196] completed step: 18, seconds: 0.146, TFLOP/s/device: 92.898, Tokens/s/device: 14002.557, total_weights: 65536, loss: 5.448, lm_loss: 5.448, perplexity: 232.333
I0418 00:38:55.803891 134828115330880 checkpointing.py:772] Waiting for step 19 to finish before checkpoint...
I0418 00:38:55.804603 134828115330880 checkpointing.py:776] Waited 0.0007638931274414062 seconds for step 19 to finish before starting checkpointing.
I0418 00:38:55.806665 134828115330880 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=15][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 00:39:21.433813 134704754472704 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 42.619786s (commit=40.542144s, array_metadata_write=2.077642s)
I0418 00:39:21.434971 134704258868992 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 32.668 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 48.352635622024536 s) (per-host)
I0418 00:39:21.435108 134704258868992 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 48.312855s.
I0418 00:39:23.610673 134704258868992 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0418 00:39:23.741365 134704258868992 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000264s.
I0418 00:39:24.624976 134704258868992 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.898858s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/15/items
I0418 00:39:24.626735 134704258868992 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/15/items
I0418 00:39:26.000452 134704258868992 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/15
I0418 00:39:29.125879 134704258868992 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/15`.
I0418 00:39:29.126741 134704258868992 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 56.78 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/15
I0418 00:39:29.129413 134704258868992 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 56.007168s.
I0418 00:39:29.129616 134704267261696 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 00:39:29.129735 134704267261696 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0418 00:39:29.129799 134704267261696 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=15] CheckpointManager Save Finalize is syncing with other hosts...
I0418 00:39:29.132057 134704267261696 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=15] CheckpointManager Save Finalize is done on all hosts.
I0418 00:39:29.132185 134828115330880 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.
W0418 00:39:29.132270 134828115330880 checkpoint_manager.py:1452] Waiting for previous save to complete took 33.325617 seconds. If this number is high, consider checkpointing less frequently.
I0418 00:39:29.133899 134828115330880 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 19
I0418 00:39:29.135960 134828115330880 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/19.
I0418 00:39:30.287161 134828115330880 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 00:39:30.287253 134828115330880 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
I0418 00:39:30.312086 134828115330880 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.025672s
I0418 00:39:30.312886 134828115330880 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 50.752 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.030394554138183594 s) (per-host)
I0418 00:39:30.312948 134828115330880 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.030477s (batch_requests_ready=0.001967s, total_serialization_initiated=0.027805s, others=0.000705s)
I0418 00:39:30.313023 134828115330880 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.034444s (all_items=0.000015s, per_item={'items': '0.00001478'}, temp_paths=0.034429)
I0418 00:39:30.313610 134828115330880 event_tracking.py:125] [process=0] [async] Finished blocking save in 1.18 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/19.
I0418 00:39:30.313889 134707463309056 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 00:59:30.313850
I0418 00:39:30.317152 134828115330880 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=19] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 00:39:30.317406 134704267261696 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0418 00:39:30.317561 134828115330880 standard_logger.py:34] {'step': 19, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776472735.806631, 'wait_for_prev_duration_secs': 33.32561659812927, 'time_between_consecutive_saves_sec': 23.463083267211914, 'checkpointer_blocking_start_time': 1776472769.1339374, 'checkpointer_blocking_duration_secs': 1.1800882816314697, 'get_old_steps_start_time': 1776472770.3140419, 'get_old_steps_duration_secs': 2.574920654296875e-05, 'checkpoint_manager_blocking_start_time': 1776472735.8048494, 'checkpoint_manager_blocking_duration_secs': 34.512680530548096}
I0418 00:39:30.317652 134828115330880 checkpointing.py:408] Started an asynchronous checkpoint save for step 19
I0418 00:39:30.317695 134828115330880 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=19][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 00:39:30.530394 134704258868992 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/19
I0418 00:39:32.672337 134705309714176 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/19/items
I0418 00:39:38.961784 134709837297408 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/19/items/array_metadatas/process_0
I0418 00:40:16.801321 134708578981632 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 41.737649s (commit=40.543485s, array_metadata_write=1.194164s)
I0418 00:40:16.802388 134707463309056 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 33.955 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 46.5198700428009 s) (per-host)
I0418 00:40:16.802496 134707463309056 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 46.488494s.
I0418 00:40:19.381254 134707463309056 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0418 00:40:19.618108 134707463309056 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000222s.
I0418 00:40:20.331063 134707463309056 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.201629s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/19/items
I0418 00:40:20.332781 134707463309056 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/19/items
I0418 00:40:20.848047 134707463309056 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/19
I0418 00:40:22.673324 134707463309056 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/19`.
I0418 00:40:22.674137 134707463309056 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 53.54 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_08_checkpoint_async_true/checkpoints/19
I0418 00:40:22.675432 134707463309056 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 52.361427s.
I0418 00:40:22.675570 134704267261696 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 00:40:22.675650 134704267261696 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0418 00:40:22.675715 134704267261696 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=19] CheckpointManager Save Finalize is syncing with other hosts...
I0418 00:40:22.678002 134704267261696 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=19] CheckpointManager Save Finalize is done on all hosts.
I0418 00:40:22.678175 134828115330880 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.
I0418 00:40:22.678328 134828115330880 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 00:40:22.679354 134828115330880 metric_logger.py:196] completed step: 19, seconds: 0.147, TFLOP/s/device: 92.151, Tokens/s/device: 13890.019, total_weights: 65536, loss: 5.391, lm_loss: 5.391, perplexity: 219.349
Per train step:
Total TFLOPs: 13.59
split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Sat Apr 18 00:40:35 UTC 2026
EXIT_CODE=0