MaxView

XPK Start: Sun Apr 19 22:35:22 UTC 2026
PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.
2026-04-19 22:35:46.914805: 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)
I0419 22:35:47.094222 137590998546240 max_utils.py:273] Attempting to initialize the jax distributed system...
I0419 22:35:56.136523 137590998546240 distributed.py:149] Starting JAX distributed service on [::]:8482
I0419 22:35:56.138857 137590998546240 distributed.py:172] Connecting to JAX distributed service on mt-10-shardy-false-98eco-slice-job-0-0.mt-10-shardy-false-98eco:8482
I0419 22:35:57.929102 137590998546240 max_utils.py:284] Jax distributed system initialized!
I0419 22:36:04.286981 137590998546240 max_utils.py:800] System Information: Jax Version: 0.9.2
I0419 22:36:04.287080 137590998546240 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0419 22:36:04.287120 137590998546240 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
I0419 22:36:04.287153 137590998546240 train_utils.py:378] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0419 22:36:04.990913 137590998546240 maxtext_utils.py:1718] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0419 22:36:04.991492 137590998546240 maxtext_utils.py:1718] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0419 22:36:04.991667 137590998546240 checkpointing.py:688] Setting up checkpoint logger...
I0419 22:36:04.991716 137590998546240 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0419 22:36:04.991757 137590998546240 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0419 22:36:04.992076 137590998546240 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 0x7d22aee50d70>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0419 22:36:09.209457 137590998546240 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0419 22:36:09.209791 137590998546240 checkpoint_manager.py:708] [process=4][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d0e540b4aa0>}, handler_registry=None
I0419 22:36:09.210078 137590998546240 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d0e540b4aa0>` 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`.
I0419 22:36:09.210126 137590998546240 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d0e540ba8a0>` 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`.
I0419 22:36:09.210161 137590998546240 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 0x7d0e540b4aa0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d0e540b4aa0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d0e540ba8a0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d0e540ba8a0>}).
I0419 22:36:09.210487 137590998546240 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0419 22:36:09.210556 137590998546240 async_checkpointer.py:192] [process=4][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7d0d604a42c0> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0419 22:36:10.786835 137590998546240 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260419_214525/linen_xpk_feat_nnx_trainstate_and_training_loop_20260419_214525_10_shardy_false/checkpoints
I0419 22:36:12.488942 137590998546240 checkpoint_manager.py:929] [process=4][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_feat_nnx_trainstate_and_training_loop_20260419_214525/linen_xpk_feat_nnx_trainstate_and_training_loop_20260419_214525_10_shardy_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7d0e540b8f80>
I0419 22:36:12.489116 137590998546240 checkpointing.py:302] Checkpoint manager created!
I0419 22:36:13.415304 137590998546240 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0419 22:36:13.415439 137590998546240 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0419 22:36:13.792679 137590998546240 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0419 22:36:13.792766 137590998546240 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0419 22:36:13.808857 137590998546240 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0419 22:36:13.808913 137590998546240 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0419 22:36:13.832402 137590998546240 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0419 22:36:13.832474 137590998546240 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0419 22:36:13.848821 137590998546240 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0419 22:36:13.848882 137590998546240 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0419 22:36:13.865146 137590998546240 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0419 22:36:13.865204 137590998546240 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0419 22:36:13.889549 137590998546240 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0419 22:36:13.889614 137590998546240 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0419 22:36:13.910218 137590998546240 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0419 22:36:13.910282 137590998546240 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0419 22:36:14.123300 137590998546240 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0419 22:36:14.123412 137590998546240 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0419 22:36:15.498232 137590998546240 maxtext_utils.py:1821]  params/params/decoder/decoder_norm/scale
    Shape:     float32[2048]
    Logical:   P('norm',)
    Physical:  (None,)
I0419 22:36:15.498369 137590998546240 maxtext_utils.py:1821]  params/params/decoder/layers/mlp/wi_0/kernel
    Shape:     float32[2048,16,7168]
    Logical:   P('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0419 22:36:15.498426 137590998546240 maxtext_utils.py:1821]  params/params/decoder/layers/mlp/wi_1/kernel
    Shape:     float32[2048,16,7168]
    Logical:   P('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0419 22:36:15.498507 137590998546240 maxtext_utils.py:1821]  params/params/decoder/layers/mlp/wo/kernel
    Shape:     float32[7168,16,2048]
    Logical:   P('mlp', 'layers', 'embed')
    Physical:  (None, None, 'fsdp')
I0419 22:36:15.498565 137590998546240 maxtext_utils.py:1821]  params/params/decoder/layers/post_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   P('norm', 'layers')
    Physical:  (None, None)
I0419 22:36:15.498625 137590998546240 maxtext_utils.py:1821]  params/params/decoder/layers/pre_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   P('norm', 'layers')
    Physical:  (None, None)
I0419 22:36:15.498702 137590998546240 maxtext_utils.py:1821]  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)
I0419 22:36:15.498765 137590998546240 maxtext_utils.py:1821]  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')
I0419 22:36:15.498806 137590998546240 maxtext_utils.py:1821]  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)
I0419 22:36:15.498843 137590998546240 maxtext_utils.py:1821]  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)
I0419 22:36:15.498890 137590998546240 maxtext_utils.py:1821]  params/params/decoder/logits_dense/kernel
    Shape:     float32[2048,32000]
    Logical:   P('embed_vocab', 'vocab')
    Physical:  ('fsdp', None)
I0419 22:36:15.498937 137590998546240 maxtext_utils.py:1821]  params/params/token_embedder/embedding
    Shape:     float32[32000,2048]
    Logical:   P('vocab', 'embed_vocab')
    Physical:  (None, 'fsdp')

I0419 22:36:15.987807 137590998546240 train.py:157] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0419 22:36:15.987900 137590998546240 train.py:157] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0419 22:36:16.003028 137590998546240 train.py:164] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0419 22:36:16.003089 137590998546240 train.py:164] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0419 22:36:26.800654 137590998546240 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.
I0419 22:36:26.801473 137590998546240 metric_logger.py:301] number parameters: 1.104 billion
I0419 22:36:37.850661 137590998546240 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0419 22:36:38.195582 137590998546240 checkpointing.py:798] Waited 0.34489941596984863 seconds for step 0 to finish before starting checkpointing.
I0419 22:36:38.198206 137590998546240 checkpoint_manager.py:2009] [process=4][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0419 22:36:38.199738 137590998546240 checkpoint_manager.py:1512] [process=4] Saving checkpoint at step 0
I0419 22:36:38.201204 137590998546240 event_tracking.py:70] [process=4] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260419_214525/linen_xpk_feat_nnx_trainstate_and_training_loop_20260419_214525_10_shardy_false/checkpoints/0.
I0419 22:36:38.948437 137590998546240 signaling_client.py:364] Using JaxDistributedSignalingClient
I0419 22:36:38.949379 137590998546240 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0419 22:36:38.949435 137590998546240 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
I0419 22:36:39.219778 137590998546240 base_pytree_checkpoint_handler.py:154] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.271435s
I0419 22:36:39.219961 137590998546240 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/blocking_gbytes_per_sec: 5.566 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.27713632583618164 s) (per-host)
I0419 22:36:39.220017 137590998546240 base_pytree_checkpoint_handler.py:768] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.277203s (batch_requests_ready=0.002148s, total_serialization_initiated=0.274980s, others=0.000076s)
I0419 22:36:39.220110 137590998546240 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.281334s (all_items=0.000018s, per_item={'items': '0.00001788'}, temp_paths=0.281316)
I0419 22:36:39.220957 137590998546240 event_tracking.py:125] [process=4] [async] Finished blocking save in 1.02 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260419_214525/linen_xpk_feat_nnx_trainstate_and_training_loop_20260419_214525_10_shardy_false/checkpoints/0.
I0419 22:36:39.221233 137461151225600 async_checkpointer.py:76] [process=4][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-19 22:56:39.221201
I0419 22:36:39.232872 137590998546240 checkpoint_manager.py:1560] [process=4][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0419 22:36:39.233109 137459019020032 async_checkpointer.py:280] [process=4][thread=save_finalize] Waiting for background save thread=async_save.
I0419 22:36:39.233239 137590998546240 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260419_214525/linen_xpk_feat_nnx_trainstate_and_training_loop_20260419_214525_10_shardy_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776638198.1981885, 'wait_for_prev_duration_secs': 5.841255187988281e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776638198.1997745, 'checkpointer_blocking_duration_secs': 1.0216388702392578, 'get_old_steps_start_time': 1776638199.2214413, 'get_old_steps_duration_secs': 2.956390380859375e-05, 'checkpoint_manager_blocking_start_time': 1776638198.1961236, 'checkpoint_manager_blocking_duration_secs': 1.037076473236084}
I0419 22:36:39.233361 137590998546240 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0419 22:36:39.233411 137590998546240 max_utils.py:750] 
Memstats: After params initialized:
I0419 22:36:39.233458 137590998546240 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_16(process=4,(0,4,0,0))
I0419 22:36:39.233489 137590998546240 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_17(process=4,(1,4,0,0))
I0419 22:36:39.233515 137590998546240 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_20(process=4,(0,5,0,0))
I0419 22:36:39.233539 137590998546240 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_21(process=4,(1,5,0,0))
I0419 22:36:39.544694 137590998546240 metric_logger.py:196] completed step: 0, seconds: 11.049, TFLOP/s/device: 1.230, Tokens/s/device: 185.355, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0419 22:36:39.719383 137590998546240 metric_logger.py:196] completed step: 1, seconds: 1.693, TFLOP/s/device: 8.028, Tokens/s/device: 1210.027, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0419 22:36:40.565836 137590998546240 metric_logger.py:196] completed step: 2, seconds: 0.027, TFLOP/s/device: 501.592, Tokens/s/device: 75605.434, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28637.129
I0419 22:36:40.713423 137590998546240 metric_logger.py:196] completed step: 3, seconds: 0.847, TFLOP/s/device: 16.045, Tokens/s/device: 2418.474, total_weights: 65536, loss: 9.734, lm_loss: 9.734, perplexity: 16889.203
I0419 22:36:41.009029 137590998546240 metric_logger.py:196] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.671, Tokens/s/device: 13365.442, total_weights: 65536, loss: 9.277, lm_loss: 9.277, perplexity: 10694.614
I0419 22:36:41.015044 137590998546240 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.119, Tokens/s/device: 13885.216, total_weights: 65536, loss: 8.892, lm_loss: 8.892, perplexity: 7270.717
I0419 22:36:48.793280    2899 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0419 22:36:52.563232 137460042405632 array_metadata_store.py:203] [process=4][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260419_214525/linen_xpk_feat_nnx_trainstate_and_training_loop_20260419_214525_10_shardy_false/checkpoints/0/items/array_metadatas/process_4
I0419 22:37:05.057099 137590998546240 metric_logger.py:196] completed step: 6, seconds: 0.296, TFLOP/s/device: 45.873, Tokens/s/device: 6914.434, total_weights: 65536, loss: 8.592, lm_loss: 8.592, perplexity: 5390.759
I0419 22:37:05.204835 137590998546240 metric_logger.py:196] completed step: 7, seconds: 23.895, TFLOP/s/device: 0.569, Tokens/s/device: 85.708, total_weights: 65536, loss: 8.384, lm_loss: 8.384, perplexity: 4376.905
I0419 22:37:05.352077 137590998546240 metric_logger.py:196] completed step: 8, seconds: 0.153, TFLOP/s/device: 89.088, Tokens/s/device: 13428.363, total_weights: 65536, loss: 8.255, lm_loss: 8.255, perplexity: 3846.406
I0419 22:37:05.498730 137590998546240 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0419 22:37:05.499390 137590998546240 checkpointing.py:798] Waited 0.0006797313690185547 seconds for step 9 to finish before starting checkpointing.
I0419 22:37:05.501540 137590998546240 checkpoint_manager.py:2020] [process=4][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0419 22:37:22.216694 137461151225600 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/gbytes_per_sec: 36.502 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 43.27383375167847 s) (per-host)
I0419 22:37:22.216814 137461151225600 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.995451s.
I0419 22:37:31.869831 137461151225600 async_checkpointer.py:160] [process=4][thread=async_save] Background save thread done. Time taken: 52.648452s.
I0419 22:37:31.870138 137459019020032 async_checkpointer.py:288] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save.
I0419 22:37:31.870263 137459019020032 async_checkpointer.py:298] [process=4][thread=save_finalize] No errors found in background save thread=async_save.
I0419 22:37:31.870315 137459019020032 checkpoint_manager.py:2137] [process=4][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0419 22:37:31.872036 137459019020032 checkpoint_manager.py:2146] [process=4][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0419 22:37:31.872202 137590998546240 checkpoint_manager.py:2032] [process=4][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0419 22:37:31.872345 137590998546240 checkpoint_manager.py:1452] Waiting for previous save to complete took 26.370791 seconds. If this number is high, consider checkpointing less frequently.
I0419 22:37:31.874235 137590998546240 checkpoint_manager.py:1512] [process=4] Saving checkpoint at step 9
I0419 22:37:31.876329 137590998546240 event_tracking.py:70] [process=4] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260419_214525/linen_xpk_feat_nnx_trainstate_and_training_loop_20260419_214525_10_shardy_false/checkpoints/9.
I0419 22:37:32.588002 137590998546240 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0419 22:37:32.588095 137590998546240 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
I0419 22:37:32.624210 137590998546240 base_pytree_checkpoint_handler.py:154] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.037330s
I0419 22:37:32.624371 137590998546240 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/blocking_gbytes_per_sec: 37.865 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04073834419250488 s) (per-host)
I0419 22:37:32.624424 137590998546240 base_pytree_checkpoint_handler.py:768] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.040801s (batch_requests_ready=0.001778s, total_serialization_initiated=0.038938s, others=0.000084s)
I0419 22:37:32.624512 137590998546240 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.045096s (all_items=0.000014s, per_item={'items': '0.00001431'}, temp_paths=0.045081)
I0419 22:37:32.625120 137590998546240 event_tracking.py:125] [process=4] [async] Finished blocking save in 0.75 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260419_214525/linen_xpk_feat_nnx_trainstate_and_training_loop_20260419_214525_10_shardy_false/checkpoints/9.
I0419 22:37:32.625461 137440052356864 async_checkpointer.py:76] [process=4][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-19 22:57:32.625424
I0419 22:37:32.627174 137590998546240 checkpoint_manager.py:1560] [process=4][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0419 22:37:32.627414 137459019020032 async_checkpointer.py:280] [process=4][thread=save_finalize] Waiting for background save thread=async_save.
I0419 22:37:32.627573 137590998546240 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260419_214525/linen_xpk_feat_nnx_trainstate_and_training_loop_20260419_214525_10_shardy_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776638225.5015113, 'wait_for_prev_duration_secs': 26.3707914352417, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776638251.8742735, 'checkpointer_blocking_duration_secs': 0.7512974739074707, 'get_old_steps_start_time': 1776638252.6255898, 'get_old_steps_duration_secs': 2.765655517578125e-05, 'checkpoint_manager_blocking_start_time': 1776638225.4996278, 'checkpoint_manager_blocking_duration_secs': 27.12791156768799}
I0419 22:37:32.627680 137590998546240 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0419 22:37:32.627724 137590998546240 checkpoint_manager.py:2020] [process=4][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0419 22:37:42.347556 137460042405632 array_metadata_store.py:203] [process=4][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260419_214525/linen_xpk_feat_nnx_trainstate_and_training_loop_20260419_214525_10_shardy_false/checkpoints/9/items/array_metadatas/process_4
I0419 22:38:16.347539 137440052356864 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/gbytes_per_sec: 36.094 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 43.76386904716492 s) (per-host)
I0419 22:38:16.347673 137440052356864 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 43.721692s.
I0419 22:38:23.632182 137440052356864 async_checkpointer.py:160] [process=4][thread=async_save] Background save thread done. Time taken: 51.006185s.
I0419 22:38:23.632502 137459019020032 async_checkpointer.py:288] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save.
I0419 22:38:23.632625 137459019020032 async_checkpointer.py:298] [process=4][thread=save_finalize] No errors found in background save thread=async_save.
I0419 22:38:23.632673 137459019020032 checkpoint_manager.py:2137] [process=4][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0419 22:38:23.634196 137459019020032 checkpoint_manager.py:2146] [process=4][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0419 22:38:23.634409 137590998546240 checkpoint_manager.py:2032] [process=4][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0419 22:38:23.634565 137590998546240 checkpoint_manager.py:2009] [process=4][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0419 22:38:23.635500 137590998546240 metric_logger.py:196] completed step: 9, seconds: 0.148, TFLOP/s/device: 92.110, Tokens/s/device: 13883.804, 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: Sun Apr 19 22:38:32 UTC 2026
EXIT_CODE=0