XPK Start: Fri Apr 17 14:04:39 UTC 2026
PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.
2026-04-17 14:05:04.127048: 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)
I0417 14:05:04.304692 135646978193216 max_utils.py:273] Attempting to initialize the jax distributed system...
I0417 14:05:13.346127 135646978193216 distributed.py:149] Starting JAX distributed service on [::]:8482
I0417 14:05:13.348619 135646978193216 distributed.py:172] Connecting to JAX distributed service on mt-11-optimizer-offload-tq2z7-slice-job-0-0.mt-11-optimizer-offload-tq2z7:8482
I0417 14:05:14.410198 135646978193216 max_utils.py:284] Jax distributed system initialized!
I0417 14:05:19.491348 135646978193216 max_utils.py:800] System Information: Jax Version: 0.9.2
I0417 14:05:19.491452 135646978193216 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0417 14:05:19.491491 135646978193216 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
I0417 14:05:19.491527 135646978193216 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0417 14:05:20.178388 135646978193216 maxtext_utils.py:1548] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0417 14:05:20.178656 135646978193216 checkpointing.py:677] Setting up checkpoint logger...
I0417 14:05:20.178710 135646978193216 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0417 14:05:20.178753 135646978193216 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0417 14:05:20.179104 135646978193216 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 0x7b5e0e2500e0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0417 14:05:23.541715 135646978193216 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0417 14:05:23.541965 135646978193216 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 0x7b5d0269e300>}, handler_registry=None
I0417 14:05:23.542204 135646978193216 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b5d0269e300>` 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`.
I0417 14:05:23.542254 135646978193216 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b4a645f8920>` 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`.
I0417 14:05:23.542289 135646978193216 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 0x7b5d0269e300>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b5d0269e300>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b4a645f8920>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b4a645f8920>}).
I0417 14:05:23.542610 135646978193216 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0417 14:05:23.542681 135646978193216 async_checkpointer.py:192] [process=0][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7b4a641bc680> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0417 14:05:24.252272 135646978193216 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints
I0417 14:05:25.105290 135646978193216 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_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7b4a647f1f70>
I0417 14:05:25.105461 135646978193216 checkpointing.py:301] Checkpoint manager created!
I0417 14:05:26.031435 135646978193216 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0417 14:05:26.031562 135646978193216 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0417 14:05:26.411822 135646978193216 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0417 14:05:26.411964 135646978193216 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0417 14:05:26.427913 135646978193216 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0417 14:05:26.427969 135646978193216 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0417 14:05:26.451275 135646978193216 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0417 14:05:26.451343 135646978193216 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0417 14:05:26.467387 135646978193216 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0417 14:05:26.467450 135646978193216 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0417 14:05:26.483614 135646978193216 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0417 14:05:26.483673 135646978193216 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0417 14:05:26.507864 135646978193216 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0417 14:05:26.507936 135646978193216 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0417 14:05:26.529286 135646978193216 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0417 14:05:26.529351 135646978193216 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0417 14:05:26.740525 135646978193216 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0417 14:05:26.740628 135646978193216 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0417 14:05:28.355987 135646978193216 maxtext_utils.py:1651] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: P('norm',)
Physical: (None,)
I0417 14:05:28.356109 135646978193216 maxtext_utils.py:1651] params/params/decoder/layers/mlp/wi_0/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0417 14:05:28.356158 135646978193216 maxtext_utils.py:1651] params/params/decoder/layers/mlp/wi_1/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0417 14:05:28.356215 135646978193216 maxtext_utils.py:1651] params/params/decoder/layers/mlp/wo/kernel
Shape: float32[7168,16,2048]
Logical: P('mlp', 'layers', 'embed')
Physical: (None, None, 'fsdp')
I0417 14:05:28.356267 135646978193216 maxtext_utils.py:1651] params/params/decoder/layers/post_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0417 14:05:28.356313 135646978193216 maxtext_utils.py:1651] params/params/decoder/layers/pre_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0417 14:05:28.356366 135646978193216 maxtext_utils.py:1651] 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)
I0417 14:05:28.356418 135646978193216 maxtext_utils.py:1651] 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')
I0417 14:05:28.356457 135646978193216 maxtext_utils.py:1651] 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)
I0417 14:05:28.356494 135646978193216 maxtext_utils.py:1651] 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)
I0417 14:05:28.356540 135646978193216 maxtext_utils.py:1651] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: P('embed_vocab', 'vocab')
Physical: ('fsdp', None)
I0417 14:05:28.356586 135646978193216 maxtext_utils.py:1651] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: P('vocab', 'embed_vocab')
Physical: (None, 'fsdp')
I0417 14:05:28.838124 135646978193216 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0417 14:05:28.838220 135646978193216 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0417 14:05:28.858522 135646978193216 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0417 14:05:28.858621 135646978193216 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0417 14:05:39.518027 135646978193216 max_utils.py:791] Total memory size: 1.3 GB, Output size: 0.1 GB, Temp size: 1.1 GB, Argument size: 0.1 GB, Host temp size: 0.0 GB.
I0417 14:05:39.525600 135646978193216 metric_logger.py:301] number parameters: 1.104 billion
I0417 14:05:50.775028 135646978193216 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0417 14:05:50.946120 135646978193216 checkpointing.py:776] Waited 0.17107105255126953 seconds for step 0 to finish before starting checkpointing.
I0417 14:05:50.948812 135646978193216 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0417 14:05:50.950419 135646978193216 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 0
I0417 14:05:50.952120 135646978193216 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/0.
I0417 14:05:51.675399 135646978193216 signaling_client.py:364] Using JaxDistributedSignalingClient
I0417 14:05:51.691224 135646978193216 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0417 14:05:51.691290 135646978193216 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
I0417 14:05:51.958409 135646978193216 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.267943s
I0417 14:05:51.959231 135646978193216 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 5.592 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2759518623352051 s) (per-host)
I0417 14:05:51.959314 135646978193216 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.276057s (batch_requests_ready=0.002214s, total_serialization_initiated=0.273094s, others=0.000749s)
I0417 14:05:51.959398 135646978193216 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.279998s (all_items=0.000018s, per_item={'items': '0.00001836'}, temp_paths=0.279980)
I0417 14:05:51.961189 135646978193216 event_tracking.py:125] [process=0] [async] Finished blocking save in 1.01 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/0.
I0417 14:05:51.961494 135512123766528 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-17 14:25:51.961455
I0417 14:05:51.966729 135646978193216 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0417 14:05:51.967015 135513423980288 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0417 14:05:51.967176 135646978193216 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776434750.9487944, 'wait_for_prev_duration_secs': 7.557868957519531e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776434750.950458, 'checkpointer_blocking_duration_secs': 1.0111806392669678, 'get_old_steps_start_time': 1776434751.9616623, 'get_old_steps_duration_secs': 2.765655517578125e-05, 'checkpoint_manager_blocking_start_time': 1776434750.9466093, 'checkpoint_manager_blocking_duration_secs': 1.020524024963379}
I0417 14:05:51.967281 135646978193216 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0417 14:05:51.967332 135646978193216 max_utils.py:750]
Memstats: After params initialized:
I0417 14:05:51.967380 135646978193216 max_utils.py:756] Using (GB) 0.17 / 31.25 (0.544000%) on TPU_0(process=0,(0,0,0,0))
I0417 14:05:51.967412 135646978193216 max_utils.py:756] Using (GB) 0.17 / 31.25 (0.544000%) on TPU_1(process=0,(1,0,0,0))
I0417 14:05:51.967438 135646978193216 max_utils.py:756] Using (GB) 0.17 / 31.25 (0.544000%) on TPU_4(process=0,(0,1,0,0))
I0417 14:05:51.967462 135646978193216 max_utils.py:756] Using (GB) 0.17 / 31.25 (0.544000%) on TPU_5(process=0,(1,1,0,0))
I0417 14:05:52.025738 135513415587584 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/0
I0417 14:05:52.307595 135646978193216 metric_logger.py:196] completed step: 0, seconds: 10.937, TFLOP/s/device: 1.242, Tokens/s/device: 187.248, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0417 14:05:52.310049 135646978193216 metric_logger.py:281] To see full metrics 'tensorboard --logdir=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/tensorboard/'
I0417 14:05:52.716111 135646978193216 metric_logger.py:196] completed step: 1, seconds: 1.531, TFLOP/s/device: 8.876, Tokens/s/device: 1337.886, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0417 14:05:52.878278 135646978193216 metric_logger.py:196] completed step: 2, seconds: 0.409, TFLOP/s/device: 33.205, Tokens/s/device: 5005.034, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28637.129
I0417 14:05:53.041186 135646978193216 metric_logger.py:196] completed step: 3, seconds: 0.009, TFLOP/s/device: 1586.540, Tokens/s/device: 239140.589, total_weights: 65536, loss: 9.734, lm_loss: 9.734, perplexity: 16889.047
I0417 14:05:54.712414 135512367023872 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/0/items
I0417 14:05:54.845168 135511813539584 checkpoint.py:188] Wrote Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776434754425889863, 'commit_timestamp_nsecs': None, 'custom_metadata': {}}, json={"item_handlers": null, "metrics": {}, "performance_metrics": {}, "init_timestamp_nsecs": 1776434754425889863, "commit_timestamp_nsecs": null, "custom_metadata": {}} to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/0/_CHECKPOINT_METADATA
I0417 14:05:56.243621 2801 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0417 14:05:58.443922 135512350238464 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_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/0/items/array_metadatas/process_0
I0417 14:06:15.583238 135646978193216 metric_logger.py:196] completed step: 4, seconds: 0.163, TFLOP/s/device: 83.303, Tokens/s/device: 12556.329, total_weights: 65536, loss: 9.278, lm_loss: 9.278, perplexity: 10694.968
I0417 14:06:15.591902 135646978193216 metric_logger.py:196] completed step: 5, seconds: 0.163, TFLOP/s/device: 83.553, Tokens/s/device: 12594.009, total_weights: 65536, loss: 8.892, lm_loss: 8.892, perplexity: 7270.520
I0417 14:06:15.753657 135646978193216 metric_logger.py:196] completed step: 6, seconds: 22.542, TFLOP/s/device: 0.603, Tokens/s/device: 90.855, total_weights: 65536, loss: 8.592, lm_loss: 8.592, perplexity: 5390.688
I0417 14:06:15.916242 135646978193216 metric_logger.py:196] completed step: 7, seconds: 0.006, TFLOP/s/device: 2213.609, Tokens/s/device: 333659.172, total_weights: 65536, loss: 8.384, lm_loss: 8.384, perplexity: 4377.010
I0417 14:06:16.079171 135646978193216 metric_logger.py:196] completed step: 8, seconds: 0.163, TFLOP/s/device: 83.473, Tokens/s/device: 12582.017, total_weights: 65536, loss: 8.255, lm_loss: 8.255, perplexity: 3845.943
I0417 14:06:16.241200 135646978193216 checkpointing.py:772] Waiting for step 9 to finish before checkpoint...
I0417 14:06:16.241916 135646978193216 checkpointing.py:776] Waited 0.0007395744323730469 seconds for step 9 to finish before starting checkpointing.
I0417 14:06:16.244049 135646978193216 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0417 14:06:41.343411 135512341845760 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 45.104272s (commit=43.882683s, array_metadata_write=1.221589s)
I0417 14:06:41.344606 135512123766528 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 31.816 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 49.661296129226685 s) (per-host)
I0417 14:06:41.344732 135512123766528 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 49.383127s.
I0417 14:06:42.132381 135512123766528 checkpoint.py:228] Read Metadata={'item_handlers': None, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776434754425889863, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} from gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/0/_CHECKPOINT_METADATA
I0417 14:06:42.949213 135511813539584 checkpoint.py:247] Updated Metadata={'item_handlers': {'items': 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler'}, 'metrics': {}, 'performance_metrics': {}, 'init_timestamp_nsecs': 1776434754425889863, 'commit_timestamp_nsecs': None, 'custom_metadata': {}} to gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/0/_CHECKPOINT_METADATA
I0417 14:06:43.301436 135512123766528 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000251s.
I0417 14:06:43.551659 135512123766528 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0417 14:06:44.493063 135512123766528 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.239672s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/0/items
I0417 14:06:44.494875 135512123766528 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/0/items
I0417 14:06:45.058436 135512123766528 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/0
I0417 14:06:46.815724 135512123766528 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/0`.
I0417 14:06:46.816585 135512123766528 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 55.87 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/0
I0417 14:06:46.818004 135512123766528 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 54.856397s.
I0417 14:06:46.818222 135513423980288 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0417 14:06:46.818359 135513423980288 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0417 14:06:46.818429 135513423980288 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0417 14:06:46.820491 135513423980288 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0417 14:06:46.820677 135646978193216 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.
W0417 14:06:46.820826 135646978193216 checkpoint_manager.py:1452] Waiting for previous save to complete took 30.576766 seconds. If this number is high, consider checkpointing less frequently.
I0417 14:06:46.822651 135646978193216 checkpoint_manager.py:1512] [process=0] Saving checkpoint at step 9
I0417 14:06:46.824780 135646978193216 event_tracking.py:70] [process=0] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/9.
I0417 14:06:47.108787 135646978193216 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0417 14:06:47.108893 135646978193216 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
I0417 14:06:47.128854 135646978193216 base_pytree_checkpoint_handler.py:154] [process=0][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.020808s
I0417 14:06:47.129719 135646978193216 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/blocking_gbytes_per_sec: 61.150 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.02523326873779297 s) (per-host)
I0417 14:06:47.129800 135646978193216 base_pytree_checkpoint_handler.py:768] [process=0][thread=MainThread] Initiated Pytree async_save. Time taken: 0.025333s (batch_requests_ready=0.001804s, total_serialization_initiated=0.022738s, others=0.000791s)
I0417 14:06:47.129899 135646978193216 composite_checkpoint_handler.py:715] [process=0][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.029371s (all_items=0.000017s, per_item={'items': '0.00001693'}, temp_paths=0.029354)
I0417 14:06:47.130601 135646978193216 event_tracking.py:125] [process=0] [async] Finished blocking save in 0.31 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/9.
I0417 14:06:47.130922 135513423980288 async_checkpointer.py:76] [process=0][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-17 14:26:47.130884
I0417 14:06:47.148050 135646978193216 checkpoint_manager.py:1560] [process=0][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0417 14:06:47.148342 135512081803008 async_checkpointer.py:280] [process=0][thread=save_finalize] Waiting for background save thread=async_save.
I0417 14:06:47.148509 135646978193216 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776434776.2440178, 'wait_for_prev_duration_secs': 30.5767662525177, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776434806.8226902, 'checkpointer_blocking_duration_secs': 0.3083662986755371, 'get_old_steps_start_time': 1776434807.1310804, 'get_old_steps_duration_secs': 3.0994415283203125e-05, 'checkpoint_manager_blocking_start_time': 1776434776.2421455, 'checkpoint_manager_blocking_duration_secs': 30.906328201293945}
I0417 14:06:47.148614 135646978193216 checkpointing.py:408] Started an asynchronous checkpoint save for step 9
I0417 14:06:47.148659 135646978193216 checkpoint_manager.py:2020] [process=0][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0417 14:06:47.765457 135512123766528 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/9
I0417 14:06:49.490124 135512090195712 atomicity.py:140] Creating tmp directory gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/9/items
I0417 14:06:53.039425 135512115373824 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_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/9/items/array_metadatas/process_0
I0417 14:07:32.854472 135512106981120 base_pytree_checkpoint_handler.py:1282] [process=0][thread=write_metadata_after_commits] Commit + Array metadata written. Time taken: 41.647053s (commit=40.429449s, array_metadata_write=1.217604s)
I0417 14:07:32.856922 135513423980288 base_pytree_checkpoint_handler.py:130] [process=0] /jax/orbax/write/gbytes_per_sec: 34.535 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 45.75240731239319 s) (per-host)
I0417 14:07:32.857046 135513423980288 async_checkpointer.py:90] [process=0][thread=async_save] 3 Handler Commit operations completed. Time taken: 45.726013s.
I0417 14:07:34.373817 135513423980288 array_metadata_store.py:411] [process=0][thread=async_save] Validated ArrayMetadata from all 8 hosts. Time taken: 0.000139s.
I0417 14:07:35.165194 135513423980288 ocdbt_utils.py:49] Param validation support for Zarr3 will be added later (b/362328389).
I0417 14:07:36.097119 135513423980288 base_pytree_checkpoint_handler.py:1406] [process=0][thread=async_save] Pytree save finalize (merge_ocdbt + ArrayMetadata validation) completed. Time taken: 2.325733s. use_zarr3=True, enable_post_merge_validation=True, directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/9/items
I0417 14:07:36.099551 135513423980288 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/9/items
I0417 14:07:36.623754 135513423980288 atomicity.py:666] Finalizing gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/9
I0417 14:07:38.033200 135513423980288 atomicity.py:847] [process=0][thread=async_save] Finished saving checkpoint (finalized tmp dir) to `gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/9`.
I0417 14:07:38.035333 135513423980288 event_tracking.py:138] [process=0] [async] Finished save (blocking + background) in 51.21 seconds @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_11_optimizer_offload_true/checkpoints/9
I0417 14:07:38.036847 135513423980288 async_checkpointer.py:160] [process=0][thread=async_save] Background save thread done. Time taken: 50.905799s.
I0417 14:07:38.036979 135512081803008 async_checkpointer.py:288] [process=0][thread=save_finalize] Done with waiting for background save thread=async_save.
I0417 14:07:38.037039 135512081803008 async_checkpointer.py:298] [process=0][thread=save_finalize] No errors found in background save thread=async_save.
I0417 14:07:38.037103 135512081803008 checkpoint_manager.py:2137] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0417 14:07:38.039595 135512081803008 checkpoint_manager.py:2146] [process=0][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0417 14:07:38.039729 135646978193216 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.
I0417 14:07:38.039822 135646978193216 checkpoint_manager.py:2009] [process=0][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0417 14:07:38.040783 135646978193216 metric_logger.py:196] completed step: 9, seconds: 0.162, TFLOP/s/device: 83.658, Tokens/s/device: 12609.828, total_weights: 65536, loss: 8.179, lm_loss: 8.179, perplexity: 3564.739
Per train step:
Total TFLOPs: 13.59
split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Fri Apr 17 14:07:50 UTC 2026
EXIT_CODE=0