XPK Start: Sat Apr 18 01:35:06 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 01:35:30.357912: 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 01:35:30.541157 135757838190400 max_utils.py:273] Attempting to initialize the jax distributed system...
I0418 01:35:39.588992 135757838190400 distributed.py:149] Starting JAX distributed service on [::]:8482
I0418 01:35:39.591358 135757838190400 distributed.py:172] Connecting to JAX distributed service on mt-15-ocdbt-false-sv-eevo0-slice-job-0-0.mt-15-ocdbt-false-sv-eevo0:8482
I0418 01:35:40.676484 135757838190400 max_utils.py:284] Jax distributed system initialized!
I0418 01:35:45.917345 135757838190400 max_utils.py:800] System Information: Jax Version: 0.9.2
I0418 01:35:45.917448 135757838190400 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0418 01:35:45.917489 135757838190400 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 01:35:45.917524 135757838190400 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0418 01:35:46.613302 135757838190400 maxtext_utils.py:1551] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0418 01:35:46.613576 135757838190400 checkpointing.py:677] Setting up checkpoint logger...
I0418 01:35:46.613630 135757838190400 checkpointing.py:233] Creating checkpoint manager with ocdbt=False and zarr3=True
I0418 01:35:46.613674 135757838190400 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0418 01:35:46.614036 135757838190400 base_pytree_checkpoint_handler.py:441] Created BasePyTreeCheckpointHandler: use_ocdbt=False, use_zarr3=True, pytree_metadata_options=PyTreeMetadataOptions(support_rich_types=False), array_metadata_store=<orbax.checkpoint._src.metadata.array_metadata_store.Store object at 0x7b77de148050>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0418 01:35:50.756921 135757838190400 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0418 01:35:50.757257 135757838190400 checkpoint_manager.py:708] [process=6][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b64884308c0>}, handler_registry=None
I0418 01:35:50.757495 135757838190400 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b64884308c0>` 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 01:35:50.757553 135757838190400 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b6488435af0>` 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 01:35:50.757592 135757838190400 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 0x7b64884308c0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b64884308c0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b6488435af0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b6488435af0>}).
I0418 01:35:50.757940 135757838190400 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0418 01:35:50.758013 135757838190400 async_checkpointer.py:192] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7b6388784180> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0418 01:35:52.996812 135757838190400 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_false/checkpoints
I0418 01:35:53.411982 135757838190400 checkpoint_manager.py:929] [process=6][thread=MainThread] CheckpointManager created, primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=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_15_ocdbt_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7b64a80f0b00>
I0418 01:35:53.412148 135757838190400 checkpointing.py:301] Checkpoint manager created!
I0418 01:35:54.342665 135757838190400 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0418 01:35:54.342778 135757838190400 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 01:35:54.723017 135757838190400 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 01:35:54.723108 135757838190400 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 01:35:54.739958 135757838190400 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 01:35:54.740017 135757838190400 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 01:35:54.764349 135757838190400 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 01:35:54.764418 135757838190400 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 01:35:54.781446 135757838190400 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 01:35:54.781509 135757838190400 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 01:35:54.798396 135757838190400 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 01:35:54.798454 135757838190400 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 01:35:54.823725 135757838190400 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0418 01:35:54.823797 135757838190400 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 01:35:54.845360 135757838190400 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0418 01:35:54.845423 135757838190400 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0418 01:35:55.067117 135757838190400 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0418 01:35:55.067222 135757838190400 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0418 01:35:56.503106 135757838190400 maxtext_utils.py:1654] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: P('norm',)
Physical: (None,)
I0418 01:35:56.503271 135757838190400 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 01:35:56.503340 135757838190400 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 01:35:56.503418 135757838190400 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 01:35:56.503493 135757838190400 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 01:35:56.503552 135757838190400 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 01:35:56.503621 135757838190400 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 01:35:56.503677 135757838190400 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 01:35:56.503718 135757838190400 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 01:35:56.503756 135757838190400 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 01:35:56.503805 135757838190400 maxtext_utils.py:1654] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: P('embed_vocab', 'vocab')
Physical: ('fsdp', None)
I0418 01:35:56.503868 135757838190400 maxtext_utils.py:1654] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: P('vocab', 'embed_vocab')
Physical: (None, 'fsdp')
I0418 01:35:57.004530 135757838190400 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0418 01:35:57.004630 135757838190400 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0418 01:35:57.020866 135757838190400 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0418 01:35:57.020931 135757838190400 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0418 01:36:08.026425 135757838190400 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 01:36:08.027283 135757838190400 metric_logger.py:301] number parameters: 1.104 billion
I0418 01:36:19.613327 135757838190400 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0418 01:36:19.767657 135757838190400 checkpointing.py:776] Waited 0.1543104648590088 seconds for step 0 to finish before starting checkpointing.
I0418 01:36:19.770023 135757838190400 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 01:36:19.771629 135757838190400 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 0
I0418 01:36:19.773164 135757838190400 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_false/checkpoints/0.
I0418 01:36:20.539227 135757838190400 signaling_client.py:364] Using JaxDistributedSignalingClient
I0418 01:36:20.540192 135757838190400 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 01:36:20.540249 135757838190400 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 01:36:20.818926 135757838190400 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.279772s
I0418 01:36:20.819104 135757838190400 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 5.403 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2855050563812256 s) (per-host)
I0418 01:36:20.819170 135757838190400 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.285573s (batch_requests_ready=0.002526s, total_serialization_initiated=0.282972s, others=0.000075s)
I0418 01:36:20.819295 135757838190400 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.289744s (all_items=0.000019s, per_item={'items': '0.00001884'}, temp_paths=0.289725)
I0418 01:36:20.820112 135757838190400 event_tracking.py:125] [process=6] [async] Finished blocking save in 1.05 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_false/checkpoints/0.
I0418 01:36:20.820431 135632099137280 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 01:56:20.820394
I0418 01:36:20.822485 135757838190400 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 01:36:20.822729 135626745706240 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0418 01:36:20.822880 135757838190400 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776476179.7700062, 'wait_for_prev_duration_secs': 6.0558319091796875e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776476179.771668, 'checkpointer_blocking_duration_secs': 1.0489206314086914, 'get_old_steps_start_time': 1776476180.8206203, 'get_old_steps_duration_secs': 3.170967102050781e-05, 'checkpoint_manager_blocking_start_time': 1776476179.7680824, 'checkpoint_manager_blocking_duration_secs': 1.0547573566436768}
I0418 01:36:20.822986 135757838190400 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0418 01:36:20.823035 135757838190400 max_utils.py:750]
Memstats: After params initialized:
I0418 01:36:20.823084 135757838190400 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_24(process=6,(0,6,0,0))
I0418 01:36:20.823116 135757838190400 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_25(process=6,(1,6,0,0))
I0418 01:36:20.823143 135757838190400 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_28(process=6,(0,7,0,0))
I0418 01:36:20.823168 135757838190400 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_29(process=6,(1,7,0,0))
I0418 01:36:21.136672 135757838190400 metric_logger.py:196] completed step: 0, seconds: 11.586, TFLOP/s/device: 1.173, Tokens/s/device: 176.766, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0418 01:36:21.312668 135757838190400 metric_logger.py:196] completed step: 1, seconds: 1.522, TFLOP/s/device: 8.928, Tokens/s/device: 1345.783, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0418 01:36:21.718782 135757838190400 metric_logger.py:196] completed step: 2, seconds: 0.029, TFLOP/s/device: 469.818, Tokens/s/device: 70816.044, total_weights: 65536, loss: 10.560, lm_loss: 10.560, perplexity: 38547.719
I0418 01:36:21.866376 135757838190400 metric_logger.py:196] completed step: 3, seconds: 0.407, TFLOP/s/device: 33.416, Tokens/s/device: 5036.829, total_weights: 65536, loss: 9.980, lm_loss: 9.980, perplexity: 21590.506
I0418 01:36:21.870931 135757838190400 checkpointing.py:772] Waiting for step 5 to finish before checkpoint...
I0418 01:36:22.160927 135757838190400 checkpointing.py:776] Waited 0.2899587154388428 seconds for step 5 to finish before starting checkpointing.
I0418 01:36:22.163934 135757838190400 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 01:36:25.031993 2796 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0418 01:36:28.810796 135627765638912 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_false/checkpoints/0/items/array_metadatas/process_6
I0418 01:36:32.336255 135632099137280 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 133.834 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 11.802619934082031 s) (per-host)
I0418 01:36:32.336365 135632099137280 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 11.515824s.
I0418 01:36:41.554071 135632099137280 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 20.733515s.
I0418 01:36:41.554293 135626745706240 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 01:36:41.554343 135626745706240 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0418 01:36:41.554389 135626745706240 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0418 01:36:41.557110 135626745706240 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0418 01:36:41.557277 135757838190400 checkpoint_manager.py:2032] [process=6][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0418 01:36:41.557413 135757838190400 checkpoint_manager.py:1452] Waiting for previous save to complete took 19.393490 seconds. If this number is high, consider checkpointing less frequently.
I0418 01:36:41.559422 135757838190400 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 5
I0418 01:36:41.562054 135757838190400 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_false/checkpoints/5.
I0418 01:36:42.263325 135757838190400 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 01:36:42.263420 135757838190400 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 01:36:42.306483 135757838190400 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.044271s
I0418 01:36:42.306652 135757838190400 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 32.359 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04767107963562012 s) (per-host)
I0418 01:36:42.306702 135757838190400 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.047731s (batch_requests_ready=0.001755s, total_serialization_initiated=0.045910s, others=0.000066s)
I0418 01:36:42.306790 135757838190400 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.052712s (all_items=0.000014s, per_item={'items': '0.00001407'}, temp_paths=0.052698)
I0418 01:36:42.307538 135757838190400 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.75 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_false/checkpoints/5.
I0418 01:36:42.307845 135626745706240 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 01:56:42.307809
I0418 01:36:42.309744 135757838190400 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=5] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 01:36:42.310073 135632082351872 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0418 01:36:42.310233 135757838190400 standard_logger.py:34] {'step': 5, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776476182.163895, 'wait_for_prev_duration_secs': 19.393489599227905, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776476201.55946, 'checkpointer_blocking_duration_secs': 0.7484867572784424, 'get_old_steps_start_time': 1776476202.3079703, 'get_old_steps_duration_secs': 2.8133392333984375e-05, 'checkpoint_manager_blocking_start_time': 1776476182.1615596, 'checkpoint_manager_blocking_duration_secs': 20.1486394405365}
I0418 01:36:42.310347 135757838190400 checkpointing.py:408] Started an asynchronous checkpoint save for step 5
I0418 01:36:42.311212 135757838190400 metric_logger.py:196] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.965, Tokens/s/device: 13409.723, total_weights: 65536, loss: 9.460, lm_loss: 9.460, perplexity: 12840.355
I0418 01:36:42.317379 135757838190400 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.156, Tokens/s/device: 13890.773, total_weights: 65536, loss: 8.959, lm_loss: 8.959, perplexity: 7776.467
I0418 01:36:49.176741 135627765638912 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_false/checkpoints/5/items/array_metadatas/process_6
I0418 01:36:53.058737 135626745706240 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 146.263 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 10.79971718788147 s) (per-host)
I0418 01:36:53.058893 135626745706240 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 10.750968s.
I0418 01:37:01.462863 135626745706240 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 19.154923s.
I0418 01:37:01.463140 135632082351872 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 01:37:01.463257 135632082351872 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0418 01:37:01.463304 135632082351872 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=5] CheckpointManager Save Finalize is syncing with other hosts...
I0418 01:37:01.465717 135632082351872 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=5] CheckpointManager Save Finalize is done on all hosts.
I0418 01:37:05.122101 135757838190400 metric_logger.py:196] completed step: 6, seconds: 20.445, TFLOP/s/device: 0.665, Tokens/s/device: 100.172, total_weights: 65536, loss: 8.469, lm_loss: 8.469, perplexity: 4762.740
I0418 01:37:05.269495 135757838190400 metric_logger.py:196] completed step: 7, seconds: 22.659, TFLOP/s/device: 0.600, Tokens/s/device: 90.385, total_weights: 65536, loss: 8.003, lm_loss: 8.003, perplexity: 2989.151
I0418 01:37:05.417111 135757838190400 metric_logger.py:196] completed step: 8, seconds: 0.152, TFLOP/s/device: 89.221, Tokens/s/device: 13448.292, total_weights: 65536, loss: 7.572, lm_loss: 7.572, perplexity: 1943.098
I0418 01:37:05.563741 135757838190400 checkpointing.py:772] Waiting for step 9 to finish before checkpoint...
I0418 01:37:05.564431 135757838190400 checkpointing.py:776] Waited 0.0007104873657226562 seconds for step 9 to finish before starting checkpointing.
I0418 01:37:05.566552 135757838190400 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 01:37:05.568124 135757838190400 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 9
I0418 01:37:05.569549 135757838190400 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_false/checkpoints/9.
I0418 01:37:06.702136 135757838190400 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 01:37:06.702229 135757838190400 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 01:37:06.744413 135757838190400 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.043262s
I0418 01:37:06.744579 135757838190400 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 33.057 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04666447639465332 s) (per-host)
I0418 01:37:06.744626 135757838190400 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.046724s (batch_requests_ready=0.001646s, total_serialization_initiated=0.045013s, others=0.000064s)
I0418 01:37:06.744713 135757838190400 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.050707s (all_items=0.000009s, per_item={'items': '0.00000930'}, temp_paths=0.050698)
I0418 01:37:06.745388 135757838190400 event_tracking.py:125] [process=6] [async] Finished blocking save in 1.18 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_false/checkpoints/9.
I0418 01:37:06.745730 135604415207168 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 01:57:06.745693
I0418 01:37:06.747622 135757838190400 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 01:37:06.747870 135632082351872 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0418 01:37:06.748014 135757838190400 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776476225.5665233, 'wait_for_prev_duration_secs': 6.914138793945312e-05, 'time_between_consecutive_saves_sec': 4.100764513015747, 'checkpointer_blocking_start_time': 1776476225.5681632, 'checkpointer_blocking_duration_secs': 1.17769193649292, 'get_old_steps_start_time': 1776476226.7458727, 'get_old_steps_duration_secs': 2.3126602172851562e-05, 'checkpoint_manager_blocking_start_time': 1776476225.5646546, 'checkpoint_manager_blocking_duration_secs': 1.1833281517028809}
I0418 01:37:06.748113 135757838190400 checkpointing.py:408] Started an asynchronous checkpoint save for step 9
I0418 01:37:06.748155 135757838190400 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 01:37:15.197099 135604440385280 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_235819/linen_xpk_main_20260417_235819_15_ocdbt_false/checkpoints/9/items/array_metadatas/process_6
I0418 01:37:19.608958 135604415207168 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 122.345 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 12.911007165908813 s) (per-host)
I0418 01:37:19.609083 135604415207168 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 12.862736s.
I0418 01:37:30.123261 135604415207168 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 23.376899s.
I0418 01:37:30.123542 135632082351872 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 01:37:30.123680 135632082351872 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0418 01:37:30.123724 135632082351872 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0418 01:37:30.125688 135632082351872 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0418 01:37:30.125906 135757838190400 checkpoint_manager.py:2032] [process=6][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0418 01:37:30.126051 135757838190400 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 01:37:30.127060 135757838190400 metric_logger.py:196] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.204, Tokens/s/device: 13897.937, total_weights: 65536, loss: 7.181, lm_loss: 7.181, perplexity: 1314.495
Per train step:
Total TFLOPs: 13.59
split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Sat Apr 18 01:37:41 UTC 2026
EXIT_CODE=0