MaxView

XPK Start: Tue Apr 21 08:51:18 UTC 2026
PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.
2026-04-21 08:51:42.807351: 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)
I0421 08:51:43.345001 133123942152000 max_utils.py:273] Attempting to initialize the jax distributed system...
I0421 08:51:52.386570 133123942152000 distributed.py:149] Starting JAX distributed service on [::]:8482
I0421 08:51:52.389307 133123942152000 distributed.py:172] Connecting to JAX distributed service on mt-07-eval-y6al9-slice-job-0-0.mt-07-eval-y6al9:8482
I0421 08:51:53.772011 133123942152000 max_utils.py:284] Jax distributed system initialized!
I0421 08:51:59.877557 133123942152000 max_utils.py:800] System Information: Jax Version: 0.9.2
I0421 08:51:59.877677 133123942152000 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0421 08:51:59.877723 133123942152000 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
I0421 08:51:59.877758 133123942152000 train_utils.py:378] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0421 08:52:00.619624 133123942152000 maxtext_utils.py:1718] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0421 08:52:00.620220 133123942152000 maxtext_utils.py:1718] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0421 08:52:00.620400 133123942152000 checkpointing.py:688] Setting up checkpoint logger...
I0421 08:52:00.620448 133123942152000 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0421 08:52:00.620491 133123942152000 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0421 08:52:00.620838 133123942152000 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 0x7912b6d1b710>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0421 08:52:03.388461 133123942152000 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0421 08:52:03.388712 133123942152000 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 0x78fe346a9730>}, handler_registry=None
I0421 08:52:03.388947 133123942152000 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x78fe346a9730>` 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`.
I0421 08:52:03.388995 133123942152000 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x78fe346b1220>` 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`.
I0421 08:52:03.389029 133123942152000 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 0x78fe346a9730>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x78fe346a9730>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x78fe346b1220>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x78fe346b1220>}).
I0421 08:52:03.389353 133123942152000 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0421 08:52:03.389423 133123942152000 async_checkpointer.py:192] [process=4][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x78fe342c3ec0> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0421 08:52:05.042837 133123942152000 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238/linen_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238_07_eval/checkpoints
I0421 08:52:05.045011 133123942152000 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_20260421_083238/linen_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238_07_eval/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x78fe346b2180>
I0421 08:52:05.045123 133123942152000 checkpointing.py:302] Checkpoint manager created!
I0421 08:52:05.973081 133123942152000 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0421 08:52:05.973179 133123942152000 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 08:52:06.357975 133123942152000 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0421 08:52:06.358062 133123942152000 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 08:52:06.375366 133123942152000 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0421 08:52:06.375426 133123942152000 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 08:52:06.399205 133123942152000 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0421 08:52:06.399273 133123942152000 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 08:52:06.415713 133123942152000 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0421 08:52:06.415772 133123942152000 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 08:52:06.432402 133123942152000 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0421 08:52:06.432464 133123942152000 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 08:52:06.458382 133123942152000 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0421 08:52:06.458448 133123942152000 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 08:52:06.479768 133123942152000 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0421 08:52:06.479832 133123942152000 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0421 08:52:06.695342 133123942152000 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0421 08:52:06.695448 133123942152000 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0421 08:52:08.118409 133123942152000 maxtext_utils.py:1821]  params/params/decoder/decoder_norm/scale
    Shape:     float32[2048]
    Logical:   P('norm',)
    Physical:  (None,)
I0421 08:52:08.118530 133123942152000 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)
I0421 08:52:08.118589 133123942152000 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)
I0421 08:52:08.118682 133123942152000 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')
I0421 08:52:08.118745 133123942152000 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)
I0421 08:52:08.118787 133123942152000 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)
I0421 08:52:08.118847 133123942152000 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)
I0421 08:52:08.118929 133123942152000 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')
I0421 08:52:08.118980 133123942152000 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)
I0421 08:52:08.119019 133123942152000 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)
I0421 08:52:08.119069 133123942152000 maxtext_utils.py:1821]  params/params/decoder/logits_dense/kernel
    Shape:     float32[2048,32000]
    Logical:   P('embed_vocab', 'vocab')
    Physical:  ('fsdp', None)
I0421 08:52:08.119118 133123942152000 maxtext_utils.py:1821]  params/params/token_embedder/embedding
    Shape:     float32[32000,2048]
    Logical:   P('vocab', 'embed_vocab')
    Physical:  (None, 'fsdp')

I0421 08:52:08.608911 133123942152000 train.py:157] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0421 08:52:08.609009 133123942152000 train.py:157] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0421 08:52:08.624311 133123942152000 train.py:164] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0421 08:52:08.624370 133123942152000 train.py:164] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0421 08:52:19.587607 133123942152000 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.
I0421 08:52:19.588452 133123942152000 metric_logger.py:301] number parameters: 1.104 billion
I0421 08:52:30.811420 133123942152000 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0421 08:52:30.965694 133123942152000 checkpointing.py:798] Waited 0.15423917770385742 seconds for step 0 to finish before starting checkpointing.
I0421 08:52:30.968023 133123942152000 checkpoint_manager.py:2009] [process=4][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0421 08:52:30.969710 133123942152000 checkpoint_manager.py:1512] [process=4] Saving checkpoint at step 0
I0421 08:52:30.971240 133123942152000 event_tracking.py:70] [process=4] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238/linen_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238_07_eval/checkpoints/0.
I0421 08:52:31.724465 133123942152000 signaling_client.py:364] Using JaxDistributedSignalingClient
I0421 08:52:31.725445 133123942152000 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 08:52:31.725502 133123942152000 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
I0421 08:52:32.004651 133123942152000 base_pytree_checkpoint_handler.py:154] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.280247s
I0421 08:52:32.004821 133123942152000 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/blocking_gbytes_per_sec: 5.398 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2857789993286133 s) (per-host)
I0421 08:52:32.004872 133123942152000 base_pytree_checkpoint_handler.py:768] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.285840s (batch_requests_ready=0.002300s, total_serialization_initiated=0.283472s, others=0.000068s)
I0421 08:52:32.004961 133123942152000 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.290198s (all_items=0.000018s, per_item={'items': '0.00001812'}, temp_paths=0.290180)
I0421 08:52:32.005807 133123942152000 event_tracking.py:125] [process=4] [async] Finished blocking save in 1.04 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238/linen_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238_07_eval/checkpoints/0.
I0421 08:52:32.006108 132995085223680 async_checkpointer.py:76] [process=4][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 09:12:32.006074
I0421 08:52:32.010731 133123942152000 checkpoint_manager.py:1560] [process=4][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 08:52:32.010962 132993470043904 async_checkpointer.py:280] [process=4][thread=save_finalize] Waiting for background save thread=async_save.
I0421 08:52:32.011081 133123942152000 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238/linen_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238_07_eval/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776761550.9680057, 'wait_for_prev_duration_secs': 6.198883056640625e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776761550.9697475, 'checkpointer_blocking_duration_secs': 1.0365288257598877, 'get_old_steps_start_time': 1776761552.0063016, 'get_old_steps_duration_secs': 3.0040740966796875e-05, 'checkpoint_manager_blocking_start_time': 1776761550.966164, 'checkpoint_manager_blocking_duration_secs': 1.0448768138885498}
I0421 08:52:32.011192 133123942152000 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0421 08:52:32.011244 133123942152000 max_utils.py:750] 
Memstats: After params initialized:
I0421 08:52:32.011292 133123942152000 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_16(process=4,(0,4,0,0))
I0421 08:52:32.011330 133123942152000 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_17(process=4,(1,4,0,0))
I0421 08:52:32.011357 133123942152000 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_20(process=4,(0,5,0,0))
I0421 08:52:32.011382 133123942152000 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_21(process=4,(1,5,0,0))
I0421 08:52:32.330050 133123942152000 metric_logger.py:196] completed step: 0, seconds: 11.223, TFLOP/s/device: 1.211, Tokens/s/device: 182.485, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0421 08:52:32.498340 133123942152000 metric_logger.py:196] completed step: 1, seconds: 1.517, TFLOP/s/device: 8.956, Tokens/s/device: 1349.895, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0421 08:52:32.914096 133123942152000 metric_logger.py:196] completed step: 2, seconds: 0.021, TFLOP/s/device: 652.381, Tokens/s/device: 98333.894, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28637.129
I0421 08:52:33.061295 133123942152000 metric_logger.py:196] completed step: 3, seconds: 0.416, TFLOP/s/device: 32.675, Tokens/s/device: 4925.184, total_weights: 65536, loss: 9.734, lm_loss: 9.734, perplexity: 16889.203
I0421 08:52:33.356580 133123942152000 metric_logger.py:196] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.710, Tokens/s/device: 13371.288, total_weights: 65536, loss: 9.277, lm_loss: 9.277, perplexity: 10694.614
I0421 08:52:33.362370 133123942152000 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.274, Tokens/s/device: 13908.508, total_weights: 65536, loss: 8.892, lm_loss: 8.892, perplexity: 7270.717
I0421 08:52:35.103444    2813 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0421 08:52:38.397565 132993526978304 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_20260421_083238/linen_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238_07_eval/checkpoints/0/items/array_metadatas/process_4
I0421 08:52:54.997097 133123942152000 metric_logger.py:196] completed step: 6, seconds: 0.296, TFLOP/s/device: 45.943, Tokens/s/device: 6925.002, total_weights: 65536, loss: 8.592, lm_loss: 8.592, perplexity: 5390.759
I0421 08:52:55.144663 133123942152000 metric_logger.py:196] completed step: 7, seconds: 21.488, TFLOP/s/device: 0.632, Tokens/s/device: 95.310, total_weights: 65536, loss: 8.384, lm_loss: 8.384, perplexity: 4376.905
I0421 08:52:55.291962 133123942152000 metric_logger.py:196] completed step: 8, seconds: 0.152, TFLOP/s/device: 89.184, Tokens/s/device: 13442.819, total_weights: 65536, loss: 8.255, lm_loss: 8.255, perplexity: 3846.406
I0421 08:52:55.438593 133123942152000 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0421 08:52:55.439282 133123942152000 checkpointing.py:798] Waited 0.0007090568542480469 seconds for step 9 to finish before starting checkpointing.
I0421 08:52:55.441345 133123942152000 checkpoint_manager.py:2020] [process=4][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 08:53:09.751631 132995085223680 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/gbytes_per_sec: 41.533 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 38.03255653381348 s) (per-host)
I0421 08:53:09.751754 132995085223680 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 37.745533s.
I0421 08:53:17.513273 132995085223680 async_checkpointer.py:160] [process=4][thread=async_save] Background save thread done. Time taken: 45.507035s.
I0421 08:53:17.513512 132993470043904 async_checkpointer.py:288] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 08:53:17.513580 132993470043904 async_checkpointer.py:298] [process=4][thread=save_finalize] No errors found in background save thread=async_save.
I0421 08:53:17.513656 132993470043904 checkpoint_manager.py:2137] [process=4][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0421 08:53:17.516309 132993470043904 checkpoint_manager.py:2146] [process=4][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0421 08:53:17.516484 133123942152000 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.
W0421 08:53:17.516624 133123942152000 checkpoint_manager.py:1452] Waiting for previous save to complete took 22.075284 seconds. If this number is high, consider checkpointing less frequently.
I0421 08:53:17.518718 133123942152000 checkpoint_manager.py:1512] [process=4] Saving checkpoint at step 9
I0421 08:53:17.520858 133123942152000 event_tracking.py:70] [process=4] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238/linen_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238_07_eval/checkpoints/9.
I0421 08:53:17.832531 133123942152000 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 08:53:17.832623 133123942152000 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
I0421 08:53:17.864713 133123942152000 base_pytree_checkpoint_handler.py:154] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.033284s
I0421 08:53:17.864882 133123942152000 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/blocking_gbytes_per_sec: 41.648 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.03703808784484863 s) (per-host)
I0421 08:53:17.864933 133123942152000 base_pytree_checkpoint_handler.py:768] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.037098s (batch_requests_ready=0.001884s, total_serialization_initiated=0.035146s, others=0.000068s)
I0421 08:53:17.865026 133123942152000 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.041432s (all_items=0.000015s, per_item={'items': '0.00001550'}, temp_paths=0.041416)
I0421 08:53:17.865782 133123942152000 event_tracking.py:125] [process=4] [async] Finished blocking save in 0.35 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238/linen_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238_07_eval/checkpoints/9.
I0421 08:53:17.866111 132993470043904 async_checkpointer.py:76] [process=4][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 09:13:17.866075
I0421 08:53:17.869792 133123942152000 checkpoint_manager.py:1560] [process=4][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 08:53:17.870079 132994054870784 async_checkpointer.py:280] [process=4][thread=save_finalize] Waiting for background save thread=async_save.
I0421 08:53:17.871373 133123942152000 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238/linen_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238_07_eval/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776761575.4413085, 'wait_for_prev_duration_secs': 22.075283527374268, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776761597.5187578, 'checkpointer_blocking_duration_secs': 0.34751248359680176, 'get_old_steps_start_time': 1776761597.8662932, 'get_old_steps_duration_secs': 3.0517578125e-05, 'checkpoint_manager_blocking_start_time': 1776761575.4395146, 'checkpoint_manager_blocking_duration_secs': 22.431819677352905}
I0421 08:53:17.871506 133123942152000 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0421 08:53:17.871556 133123942152000 checkpoint_manager.py:2020] [process=4][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 08:53:22.912359 132993526978304 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_20260421_083238/linen_xpk_feat_nnx_trainstate_and_training_loop_20260421_083238_07_eval/checkpoints/9/items/array_metadatas/process_4
I0421 08:54:01.194995 132993470043904 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/gbytes_per_sec: 36.424 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 43.36711025238037 s) (per-host)
I0421 08:54:01.195123 132993470043904 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 43.328900s.
I0421 08:54:08.148922 132993470043904 async_checkpointer.py:160] [process=4][thread=async_save] Background save thread done. Time taken: 50.282682s.
I0421 08:54:08.149137 132994054870784 async_checkpointer.py:288] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 08:54:08.149203 132994054870784 async_checkpointer.py:298] [process=4][thread=save_finalize] No errors found in background save thread=async_save.
I0421 08:54:08.149255 132994054870784 checkpoint_manager.py:2137] [process=4][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0421 08:54:08.151043 132994054870784 checkpoint_manager.py:2146] [process=4][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0421 08:54:08.151231 133123942152000 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.
I0421 08:54:08.151413 133123942152000 checkpoint_manager.py:2009] [process=4][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0421 08:54:08.152340 133123942152000 metric_logger.py:196] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.169, Tokens/s/device: 13892.752, 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: Tue Apr 21 08:54:20 UTC 2026
EXIT_CODE=0