MaxView

← Back to run

Log Summary

XPK Start: Fri Apr 17 13:07:38 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 13:08:02.911866: 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 13:08:03.089581 133020070946624 max_utils.py:273] Attempting to initialize the jax distributed system...
I0417 13:08:12.130510 133020070946624 distributed.py:149] Starting JAX distributed service on [::]:8482
I0417 13:08:12.132855 133020070946624 distributed.py:172] Connecting to JAX distributed service on mt-02-synthetic-w8snw-slice-job-0-0.mt-02-synthetic-w8snw:8482
I0417 13:08:13.863147 133020070946624 max_utils.py:284] Jax distributed system initialized!
I0417 13:08:20.044211 133020070946624 max_utils.py:800] System Information: Jax Version: 0.9.2
I0417 13:08:20.044316 133020070946624 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0417 13:08:20.044356 133020070946624 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 13:08:20.044390 133020070946624 train_utils.py:348] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0417 13:08:20.749747 133020070946624 maxtext_utils.py:1548] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0417 13:08:20.750040 133020070946624 checkpointing.py:677] Setting up checkpoint logger...
I0417 13:08:20.750092 133020070946624 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0417 13:08:20.750138 133020070946624 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0417 13:08:20.750473 133020070946624 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 0x78fa85117320>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0417 13:08:25.164951 133020070946624 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0417 13:08:25.165182 133020070946624 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 0x78e650108590>}, handler_registry=None
I0417 13:08:25.165423 133020070946624 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x78e650108590>` 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 13:08:25.165471 133020070946624 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x78e65010be60>` 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 13:08:25.165506 133020070946624 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 0x78e650108590>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x78e650108590>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x78e65010be60>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x78e65010be60>}).
I0417 13:08:25.165838 133020070946624 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0417 13:08:25.165912 133020070946624 async_checkpointer.py:192] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x78e618444a40> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0417 13:08:28.050039 133020070946624 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_02_synthetic/checkpoints
I0417 13:08:29.266054 133020070946624 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=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_02_synthetic/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x78e650109070>
I0417 13:08:29.266225 133020070946624 checkpointing.py:301] Checkpoint manager created!
I0417 13:08:30.242523 133020070946624 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0417 13:08:30.242634 133020070946624 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0417 13:08:30.621661 133020070946624 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0417 13:08:30.621754 133020070946624 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0417 13:08:30.638026 133020070946624 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0417 13:08:30.638086 133020070946624 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0417 13:08:30.661746 133020070946624 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 13:08:30.661829 133020070946624 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0417 13:08:30.678200 133020070946624 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 13:08:30.678259 133020070946624 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0417 13:08:30.694553 133020070946624 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 13:08:30.694614 133020070946624 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0417 13:08:30.719218 133020070946624 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0417 13:08:30.719288 133020070946624 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0417 13:08:30.741209 133020070946624 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0417 13:08:30.741277 133020070946624 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0417 13:08:30.965328 133020070946624 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0417 13:08:30.965438 133020070946624 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0417 13:08:32.391809 133020070946624 maxtext_utils.py:1651]  params/params/decoder/decoder_norm/scale
    Shape:     float32[2048]
    Logical:   P('norm',)
    Physical:  (None,)
I0417 13:08:32.391946 133020070946624 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 13:08:32.392000 133020070946624 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 13:08:32.392066 133020070946624 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 13:08:32.392121 133020070946624 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 13:08:32.392159 133020070946624 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 13:08:32.392211 133020070946624 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 13:08:32.392263 133020070946624 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 13:08:32.392306 133020070946624 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 13:08:32.392345 133020070946624 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 13:08:32.392392 133020070946624 maxtext_utils.py:1651]  params/params/decoder/logits_dense/kernel
    Shape:     float32[2048,32000]
    Logical:   P('embed_vocab', 'vocab')
    Physical:  ('fsdp', None)

I0417 13:08:32.392440 133020070946624 maxtext_utils.py:1651]  params/params/token_embedder/embedding
    Shape:     float32[32000,2048]
    Logical:   P('vocab', 'embed_vocab')
    Physical:  (None, 'fsdp')
I0417 13:08:32.881386 133020070946624 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0417 13:08:32.881480 133020070946624 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0417 13:08:32.896925 133020070946624 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0417 13:08:32.896991 133020070946624 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0417 13:08:43.858491 133020070946624 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.
I0417 13:08:43.859322 133020070946624 metric_logger.py:301] number parameters: 1.104 billion
I0417 13:08:55.217696 133020070946624 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0417 13:08:55.371027 133020070946624 checkpointing.py:776] Waited 0.1533069610595703 seconds for step 0 to finish before starting checkpointing.
I0417 13:08:55.373511 133020070946624 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0417 13:08:55.375383 133020070946624 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 0
I0417 13:08:55.376849 133020070946624 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_02_synthetic/checkpoints/0.
I0417 13:08:56.558300 133020070946624 signaling_client.py:364] Using JaxDistributedSignalingClient
I0417 13:08:56.559336 133020070946624 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0417 13:08:56.559396 133020070946624 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 13:08:56.833487 133020070946624 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.275297s
I0417 13:08:56.833669 133020070946624 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 5.492 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2808685302734375 s) (per-host)
I0417 13:08:56.833722 133020070946624 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.280932s (batch_requests_ready=0.002364s, total_serialization_initiated=0.278495s, others=0.000074s)
I0417 13:08:56.833839 133020070946624 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.285000s (all_items=0.000018s, per_item={'items': '0.00001764'}, temp_paths=0.284982)
I0417 13:08:56.834699 133020070946624 event_tracking.py:125] [process=6] [async] Finished blocking save in 1.46 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_02_synthetic/checkpoints/0.
I0417 13:08:56.835055 132891665938176 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-17 13:28:56.835018
I0417 13:08:56.848848 133020070946624 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0417 13:08:56.849140 132889536747264 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0417 13:08:56.849296 133020070946624 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_02_synthetic/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776431335.3734927, 'wait_for_prev_duration_secs': 6.842613220214844e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776431335.3754234, 'checkpointer_blocking_duration_secs': 1.4597811698913574, 'get_old_steps_start_time': 1776431336.835231, 'get_old_steps_duration_secs': 3.0040740966796875e-05, 'checkpoint_manager_blocking_start_time': 1776431335.3715482, 'checkpoint_manager_blocking_duration_secs': 1.4777066707611084}
I0417 13:08:56.849407 133020070946624 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0417 13:08:56.849456 133020070946624 max_utils.py:750] 
Memstats: After params initialized:
I0417 13:08:56.849505 133020070946624 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_24(process=6,(0,6,0,0))
I0417 13:08:56.849538 133020070946624 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_25(process=6,(1,6,0,0))
I0417 13:08:56.849565 133020070946624 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_28(process=6,(0,7,0,0))
I0417 13:08:56.849589 133020070946624 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_29(process=6,(1,7,0,0))
I0417 13:08:57.161862 133020070946624 metric_logger.py:196] completed step: 0, seconds: 11.358, TFLOP/s/device: 1.196, Tokens/s/device: 180.309, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0417 13:08:57.335764 133020070946624 metric_logger.py:196] completed step: 1, seconds: 1.943, TFLOP/s/device: 6.995, Tokens/s/device: 1054.289, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0417 13:08:57.741944 133020070946624 metric_logger.py:196] completed step: 2, seconds: 0.027, TFLOP/s/device: 510.967, Tokens/s/device: 77018.540, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28637.129
I0417 13:08:57.889351 133020070946624 metric_logger.py:196] completed step: 3, seconds: 0.407, TFLOP/s/device: 33.415, Tokens/s/device: 5036.668, total_weights: 65536, loss: 9.734, lm_loss: 9.734, perplexity: 16889.203
I0417 13:08:58.184469 133020070946624 metric_logger.py:196] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.899, Tokens/s/device: 13399.897, total_weights: 65536, loss: 9.277, lm_loss: 9.277, perplexity: 10694.614
I0417 13:08:58.190222 133020070946624 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.124, Tokens/s/device: 13885.970, total_weights: 65536, loss: 8.892, lm_loss: 8.892, perplexity: 7270.717
I0417 13:09:04.216525    2848 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0417 13:09:06.742870 132889576470272 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_125726/linen_xpk_main_20260417_125726_02_synthetic/checkpoints/0/items/array_metadatas/process_6
I0417 13:09:25.718862 133020070946624 metric_logger.py:196] completed step: 6, seconds: 0.295, TFLOP/s/device: 45.989, Tokens/s/device: 6931.987, total_weights: 65536, loss: 8.592, lm_loss: 8.592, perplexity: 5390.759
I0417 13:09:25.866131 133020070946624 metric_logger.py:196] completed step: 7, seconds: 27.382, TFLOP/s/device: 0.496, Tokens/s/device: 74.795, total_weights: 65536, loss: 8.384, lm_loss: 8.384, perplexity: 4376.905
I0417 13:09:26.013403 133020070946624 metric_logger.py:196] completed step: 8, seconds: 0.152, TFLOP/s/device: 89.130, Tokens/s/device: 13434.618, total_weights: 65536, loss: 8.255, lm_loss: 8.255, perplexity: 3846.406
I0417 13:09:26.160096 133020070946624 checkpointing.py:772] Waiting for step 9 to finish before checkpoint...
I0417 13:09:26.160749 133020070946624 checkpointing.py:776] Waited 0.0006735324859619141 seconds for step 9 to finish before starting checkpointing.
I0417 13:09:26.162869 133020070946624 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0417 13:09:35.505866 132891665938176 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 40.551 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 38.953012228012085 s) (per-host)
I0417 13:09:35.505989 132891665938176 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 38.670823s.
I0417 13:09:53.839925 132891665938176 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 57.004743s.
I0417 13:09:53.840214 132889536747264 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0417 13:09:53.840345 132889536747264 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0417 13:09:53.840398 132889536747264 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0417 13:09:53.842787 132889536747264 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0417 13:09:53.842991 133020070946624 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.
W0417 13:09:53.843118 133020070946624 checkpoint_manager.py:1452] Waiting for previous save to complete took 27.680273 seconds. If this number is high, consider checkpointing less frequently.
I0417 13:09:53.845006 133020070946624 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 9
I0417 13:09:53.847074 133020070946624 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_02_synthetic/checkpoints/9.
I0417 13:09:55.048770 133020070946624 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0417 13:09:55.048882 133020070946624 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 13:09:55.085261 133020070946624 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.037529s
I0417 13:09:55.085435 133020070946624 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 37.502 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04113292694091797 s) (per-host)
I0417 13:09:55.085488 133020070946624 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.041196s (batch_requests_ready=0.001932s, total_serialization_initiated=0.039192s, others=0.000072s)
I0417 13:09:55.085580 133020070946624 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.046565s (all_items=0.000017s, per_item={'items': '0.00001669'}, temp_paths=0.046548)
I0417 13:09:55.086330 133020070946624 event_tracking.py:125] [process=6] [async] Finished blocking save in 1.24 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_02_synthetic/checkpoints/9.
I0417 13:09:55.086684 132889536747264 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-17 13:29:55.086648
I0417 13:09:55.091490 133020070946624 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0417 13:09:55.091742 132890650212096 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0417 13:09:55.091902 133020070946624 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260417_125726/linen_xpk_main_20260417_125726_02_synthetic/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776431366.1628153, 'wait_for_prev_duration_secs': 27.680273056030273, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776431393.8450906, 'checkpointer_blocking_duration_secs': 1.2417597770690918, 'get_old_steps_start_time': 1776431395.086875, 'get_old_steps_duration_secs': 3.075599670410156e-05, 'checkpoint_manager_blocking_start_time': 1776431366.1610124, 'checkpoint_manager_blocking_duration_secs': 28.930854320526123}
I0417 13:09:55.092017 133020070946624 checkpointing.py:408] Started an asynchronous checkpoint save for step 9
I0417 13:09:55.092064 133020070946624 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0417 13:10:08.402031 132891146020608 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_125726/linen_xpk_main_20260417_125726_02_synthetic/checkpoints/9/items/array_metadatas/process_6
I0417 13:10:42.215966 132889536747264 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 33.486 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 47.171640157699585 s) (per-host)
I0417 13:10:42.216074 132889536747264 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 47.129275s.
I0417 13:10:53.347755 132889536747264 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 58.260939s.
I0417 13:10:53.348049 132890650212096 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0417 13:10:53.348171 132890650212096 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0417 13:10:53.348222 132890650212096 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0417 13:10:53.350668 132890650212096 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0417 13:10:53.350843 133020070946624 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.
I0417 13:10:53.350991 133020070946624 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0417 13:10:53.351953 133020070946624 metric_logger.py:196] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.321, Tokens/s/device: 13915.690, total_weights: 65536, loss: 8.179, lm_loss: 8.179, perplexity: 3564.635
Per train step:
 Total TFLOPs: 13.59 
 split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Fri Apr 17 13:11:04 UTC 2026
EXIT_CODE=0