XPK Start: Sat Apr 18 13:09:39 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 13:10:03.455919: 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 13:10:03.987610 136657209214784 max_utils.py:273] Attempting to initialize the jax distributed system...
I0418 13:10:13.028196 136657209214784 distributed.py:149] Starting JAX distributed service on [::]:8482
I0418 13:10:13.030473 136657209214784 distributed.py:172] Connecting to JAX distributed service on mt-04-int8-5uvht-slice-job-0-0.mt-04-int8-5uvht:8482
I0418 13:10:14.579398 136657209214784 max_utils.py:284] Jax distributed system initialized!
I0418 13:10:20.609148 136657209214784 max_utils.py:800] System Information: Jax Version: 0.9.2
I0418 13:10:20.609249 136657209214784 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0418 13:10:20.609288 136657209214784 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 13:10:20.609322 136657209214784 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0418 13:10:21.285883 136657209214784 maxtext_utils.py:1551] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0418 13:10:21.286267 136657209214784 checkpointing.py:677] Setting up checkpoint logger...
I0418 13:10:21.286321 136657209214784 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0418 13:10:21.286365 136657209214784 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0418 13:10:21.286689 136657209214784 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 0x7c495e714170>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0418 13:10:24.674901 136657209214784 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0418 13:10:24.675132 136657209214784 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 0x7c35005275f0>}, handler_registry=None
I0418 13:10:24.675364 136657209214784 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c35005275f0>` 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 13:10:24.675410 136657209214784 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c350052b0b0>` 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 13:10:24.675443 136657209214784 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 0x7c35005275f0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c35005275f0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c350052b0b0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c350052b0b0>}).
I0418 13:10:24.675753 136657209214784 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0418 13:10:24.675820 136657209214784 async_checkpointer.py:192] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7c350028bba0> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0418 13:10:26.719539 136657209214784 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints
I0418 13:10:27.177446 136657209214784 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_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7c35007cd280>
I0418 13:10:27.177617 136657209214784 checkpointing.py:301] Checkpoint manager created!
I0418 13:10:28.293597 136657209214784 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0418 13:10:28.293710 136657209214784 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 13:10:28.671377 136657209214784 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 13:10:28.671463 136657209214784 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 13:10:28.687797 136657209214784 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0418 13:10:28.687869 136657209214784 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0418 13:10:28.753494 136657209214784 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 13:10:28.753576 136657209214784 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 13:10:28.769867 136657209214784 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 13:10:28.769927 136657209214784 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 13:10:28.786179 136657209214784 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 13:10:28.786240 136657209214784 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 13:10:28.810973 136657209214784 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0418 13:10:28.811041 136657209214784 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0418 13:10:28.876961 136657209214784 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0418 13:10:28.877041 136657209214784 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0418 13:10:29.394679 136657209214784 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0418 13:10:29.394789 136657209214784 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0418 13:10:31.424888 136657209214784 maxtext_utils.py:1654] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: P('norm',)
Physical: (None,)
I0418 13:10:31.425014 136657209214784 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 13:10:31.425065 136657209214784 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 13:10:31.425118 136657209214784 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 13:10:31.425175 136657209214784 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 13:10:31.425213 136657209214784 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 13:10:31.425264 136657209214784 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 13:10:31.425315 136657209214784 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 13:10:31.425353 136657209214784 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 13:10:31.425388 136657209214784 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 13:10:31.425434 136657209214784 maxtext_utils.py:1654] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: P('embed_vocab', 'vocab')
Physical: ('fsdp', None)
I0418 13:10:31.425479 136657209214784 maxtext_utils.py:1654] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: P('vocab', 'embed_vocab')
Physical: (None, 'fsdp')
I0418 13:10:33.248856 136657209214784 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0418 13:10:33.248953 136657209214784 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0418 13:10:33.264315 136657209214784 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0418 13:10:33.264373 136657209214784 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0418 13:10:46.708235 136657209214784 max_utils.py:791] Total memory size: 1.8 GB, Output size: 0.4 GB, Temp size: 1.4 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB.
I0418 13:10:46.716158 136657209214784 metric_logger.py:301] number parameters: 1.104 billion
I0418 13:11:01.962823 136657209214784 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0418 13:11:02.202621 136657209214784 checkpointing.py:776] Waited 0.23977971076965332 seconds for step 0 to finish before starting checkpointing.
I0418 13:11:02.205252 136657209214784 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 13:11:02.207437 136657209214784 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 0
I0418 13:11:02.208902 136657209214784 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/0.
I0418 13:11:03.343286 136657209214784 signaling_client.py:364] Using JaxDistributedSignalingClient
I0418 13:11:03.357574 136657209214784 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 13:11:03.357644 136657209214784 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 13:11:03.600844 136522664978176 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/0
I0418 13:11:03.645763 136657209214784 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.288932s
I0418 13:11:03.646551 136657209214784 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 5.224 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.29531288146972656 s) (per-host)
I0418 13:11:03.646611 136657209214784 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.295392s (batch_requests_ready=0.002240s, total_serialization_initiated=0.292486s, others=0.000666s)
I0418 13:11:03.646686 136657209214784 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.299276s (all_items=0.000016s, per_item={'items': '0.00001574'}, temp_paths=0.299261)
I0418 13:11:03.647505 136657209214784 event_tracking.py:125] [process=0] [async] Finished blocking save in 1.44 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/0.
I0418 13:11:03.647800 136522673370880 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 13:31:03.647767
I0418 13:11:03.649719 136657209214784 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 13:11:03.649981 136521201018624 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0418 13:11:03.650131 136657209214784 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776517862.2052343, 'wait_for_prev_duration_secs': 6.127357482910156e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776517862.2074757, 'checkpointer_blocking_duration_secs': 1.4404795169830322, 'get_old_steps_start_time': 1776517863.6479805, 'get_old_steps_duration_secs': 2.7418136596679688e-05, 'checkpoint_manager_blocking_start_time': 1776517862.2031796, 'checkpoint_manager_blocking_duration_secs': 1.446913242340088}
I0418 13:11:03.650232 136657209214784 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0418 13:11:03.650280 136657209214784 max_utils.py:750]
Memstats: After params initialized:
I0418 13:11:03.650326 136657209214784 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_0(process=0,(0,0,0,0))
I0418 13:11:03.650358 136657209214784 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_1(process=0,(1,0,0,0))
I0418 13:11:03.650383 136657209214784 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_4(process=0,(0,1,0,0))
I0418 13:11:03.650406 136657209214784 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_5(process=0,(1,1,0,0))
I0418 13:11:04.013500 136657209214784 metric_logger.py:196] completed step: 0, seconds: 14.896, TFLOP/s/device: 0.912, Tokens/s/device: 137.485, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52796.277
I0418 13:11:04.016608 136657209214784 metric_logger.py:281] To see full metrics 'tensorboard --logdir=gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/tensorboard/'
I0418 13:11:04.860795 136657209214784 metric_logger.py:196] completed step: 1, seconds: 2.049, TFLOP/s/device: 6.632, Tokens/s/device: 999.621, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52796.277
I0418 13:11:05.088390 136657209214784 metric_logger.py:196] completed step: 2, seconds: 0.848, TFLOP/s/device: 16.022, Tokens/s/device: 2415.037, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28634.820
I0418 13:11:05.318191 136657209214784 metric_logger.py:196] completed step: 3, seconds: 0.009, TFLOP/s/device: 1434.149, Tokens/s/device: 216170.572, total_weights: 65536, loss: 9.731, lm_loss: 9.731, perplexity: 16827.199
I0418 13:11:06.010635 136521617229568 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/0/items
I0418 13:11:06.167173 136504298956544 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776517865757576899, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776517865757576899, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/0/_CHECKPOINT_METADATA
I0418 13:11:08.031888 3013 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0418 13:11:10.670659 136521600444160 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/0/items/array_metadatas/process_0
I0418 13:11:37.123267 136657209214784 metric_logger.py:196] completed step: 4, seconds: 0.229, TFLOP/s/device: 59.431, Tokens/s/device: 8958.018, total_weights: 65536, loss: 9.272, lm_loss: 9.272, perplexity: 10638.531
I0418 13:11:37.132153 136657209214784 metric_logger.py:196] completed step: 5, seconds: 0.230, TFLOP/s/device: 59.155, Tokens/s/device: 8916.560, total_weights: 65536, loss: 8.887, lm_loss: 8.887, perplexity: 7233.941
I0418 13:11:37.359759 136657209214784 metric_logger.py:196] completed step: 6, seconds: 31.804, TFLOP/s/device: 0.427, Tokens/s/device: 64.395, total_weights: 65536, loss: 8.588, lm_loss: 8.588, perplexity: 5365.468
I0418 13:11:37.590895 136657209214784 metric_logger.py:196] completed step: 7, seconds: 0.007, TFLOP/s/device: 2068.056, Tokens/s/device: 311719.939, total_weights: 65536, loss: 8.380, lm_loss: 8.380, perplexity: 4358.779
I0418 13:11:37.820547 136657209214784 metric_logger.py:196] completed step: 8, seconds: 0.228, TFLOP/s/device: 59.470, Tokens/s/device: 8963.978, total_weights: 65536, loss: 8.251, lm_loss: 8.251, perplexity: 3831.453
I0418 13:11:38.049179 136657209214784 checkpointing.py:772] Waiting for step 9 to finish before checkpoint...
I0418 13:11:38.049882 136657209214784 checkpointing.py:776] Waited 0.0007224082946777344 seconds for step 9 to finish before starting checkpointing.
I0418 13:11:38.052704 136657209214784 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 13:11:41.928790 136521592051456 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 33.895897s (commit=32.284219s, array_metadata_write=1.611679s)
I0418 13:11:41.929933 136522673370880 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 40.945 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 38.57864761352539 s) (per-host)
I0418 13:11:41.930062 136522673370880 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 38.282140s.
I0418 13:11:43.074365 136522673370880 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776517865757576899, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/0/_CHECKPOINT_METADATA
I0418 13:11:43.826192 136504298956544 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776517865757576899, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/0/_CHECKPOINT_METADATA
I0418 13:11:44.460266 136522673370880 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0418 13:11:44.590228 136522673370880 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000262s.
I0418 13:11:45.397917 136522673370880 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.193343s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/0/items
I0418 13:11:45.400529 136522673370880 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/0/items
I0418 13:11:45.921695 136522673370880 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/0
I0418 13:11:49.084513 136522673370880 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/0`.
I0418 13:11:49.086733 136522673370880 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 46.88 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/0
I0418 13:11:49.089188 136522673370880 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 45.441261s.
I0418 13:11:49.089381 136521201018624 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 13:11:49.089494 136521201018624 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0418 13:11:49.089557 136521201018624 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0418 13:11:49.092708 136521201018624 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0418 13:11:49.092849 136657209214784 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0418 13:11:49.092928 136657209214784 checkpoint_manager.py:1452] Waiting for previous save to complete took 11.040233 seconds. If this number is high, consider checkpointing less frequently.
I0418 13:11:49.094885 136657209214784 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 9
I0418 13:11:49.096858 136657209214784 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/9.
I0418 13:11:49.812253 136657209214784 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0418 13:11:49.812343 136657209214784 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 13:11:49.838857 136657209214784 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.027243s
I0418 13:11:49.839643 136657209214784 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 48.715 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.03166508674621582 s) (per-host)
I0418 13:11:49.839706 136657209214784 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.031745s (batch_requests_ready=0.001806s, total_serialization_initiated=0.029209s, others=0.000729s)
I0418 13:11:49.839774 136657209214784 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.035742s (all_items=0.000016s, per_item={'items': '0.00001621'}, temp_paths=0.035725)
I0418 13:11:49.840444 136657209214784 event_tracking.py:125] [process=0] [async] Finished blocking save in 0.75 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/9.
I0418 13:11:49.840698 136521201018624 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-18 13:31:49.840675
I0418 13:11:49.849300 136657209214784 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0418 13:11:49.849520 136505779537664 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0418 13:11:49.849643 136657209214784 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776517898.0526676, 'wait_for_prev_duration_secs': 11.040233373641968, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776517909.094923, 'checkpointer_blocking_duration_secs': 0.745868444442749, 'get_old_steps_start_time': 1776517909.8408148, 'get_old_steps_duration_secs': 3.719329833984375e-05, 'checkpoint_manager_blocking_start_time': 1776517898.0501108, 'checkpoint_manager_blocking_duration_secs': 11.799495935440063}
I0418 13:11:49.849746 136657209214784 checkpointing.py:408] Started an asynchronous checkpoint save for step 9
I0418 13:11:49.849789 136657209214784 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0418 13:11:50.051647 136522673370880 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/9
I0418 13:11:52.207581 136521617229568 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/9/items
I0418 13:11:55.516487 136505804715776 array_metadata_store.py:203] [process=0][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/9/items/array_metadatas/process_0
I0418 13:12:35.721566 136505796323072 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 41.958524s (commit=40.302333s, array_metadata_write=1.656190s)
I0418 13:12:35.722691 136521201018624 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 34.403 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 45.91468262672424 s) (per-host)
I0418 13:12:35.722825 136521201018624 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 45.882052s.
I0418 13:12:38.071887 136521201018624 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0418 13:12:38.507679 136521201018624 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000238s.
I0418 13:12:39.035476 136521201018624 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 1.967400s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/9/items
I0418 13:12:39.038609 136521201018624 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/9/items
I0418 13:12:39.992887 136521201018624 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/9
I0418 13:12:41.400475 136521201018624 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/9`.
I0418 13:12:41.401187 136521201018624 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 52.31 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260418_125719/linen_xpk_main_20260418_125719_04_int8/checkpoints/9
I0418 13:12:41.403386 136521201018624 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 51.562612s.
I0418 13:12:41.403596 136505779537664 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0418 13:12:41.403718 136505779537664 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0418 13:12:41.403780 136505779537664 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0418 13:12:41.406518 136505779537664 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0418 13:12:41.406704 136657209214784 checkpoint_manager.py:2032] [process=0][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0418 13:12:41.406871 136657209214784 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0418 13:12:41.407874 136657209214784 metric_logger.py:196] completed step: 9, seconds: 0.231, TFLOP/s/device: 58.726, Tokens/s/device: 8851.776, total_weights: 65536, loss: 8.175, lm_loss: 8.175, perplexity: 3552.333
Per train step:
Total TFLOPs: 13.59
split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Sat Apr 18 13:12:54 UTC 2026
EXIT_CODE=0