MaxView

XPK Start: Tue Apr 21 15:59:00 UTC 2026
PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.
`rope_parameters`'s factor field must be a float >= 1, got 40
`rope_parameters`'s beta_fast field must be a float, got 32
`rope_parameters`'s beta_slow field must be a float, got 1
DeepseekV32Config got `key=rope_scaling` in kwargs but hasn't set it as attribute. For RoPE standardization you need to set `self.rope_parameters` in model's config. 
2026-04-21 15:59:25.280488: 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 15:59:25.491994 140196603328320 max_utils.py:273] Attempting to initialize the jax distributed system...
I0421 15:59:34.534081 140196603328320 distributed.py:149] Starting JAX distributed service on [::]:8482
I0421 15:59:34.536490 140196603328320 distributed.py:172] Connecting to JAX distributed service on mt-16-zarr3-true-sv-vduwz-slice-job-0-0.mt-16-zarr3-true-sv-vduwz:8482
I0421 15:59:35.641252 140196603328320 max_utils.py:284] Jax distributed system initialized!
I0421 15:59:41.718806 140196603328320 max_utils.py:800] System Information: Jax Version: 0.9.2
I0421 15:59:41.718914 140196603328320 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0421 15:59:41.718956 140196603328320 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 15:59:41.718992 140196603328320 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0421 15:59:42.416493 140196603328320 maxtext_utils.py:1565] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0421 15:59:42.416789 140196603328320 checkpointing.py:677] Setting up checkpoint logger...
I0421 15:59:42.416866 140196603328320 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0421 15:59:42.416925 140196603328320 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0421 15:59:42.417299 140196603328320 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 0x7f817751c1a0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0421 15:59:45.368211 140196603328320 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0421 15:59:45.368463 140196603328320 checkpoint_manager.py:708] [process=5][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7f6d106956d0>}, handler_registry=None
I0421 15:59:45.368722 140196603328320 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7f6d106956d0>` 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 15:59:45.368790 140196603328320 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7f6d10699130>` 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 15:59:45.368838 140196603328320 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 0x7f6d106956d0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7f6d106956d0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7f6d10699130>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7f6d10699130>}).
I0421 15:59:45.369188 140196603328320 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0421 15:59:45.369262 140196603328320 async_checkpointer.py:192] [process=5][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7f6d102d56c0> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0421 15:59:47.012104 140196603328320 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_16_zarr3_true/checkpoints
I0421 15:59:47.016305 140196603328320 checkpoint_manager.py:929] [process=5][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=5), 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_20260421_144122/linen_xpk_main_20260421_144122_16_zarr3_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7f6d10694920>
I0421 15:59:47.016434 140196603328320 checkpointing.py:301] Checkpoint manager created!
I0421 15:59:47.945929 140196603328320 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0421 15:59:47.946044 140196603328320 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 15:59:48.327101 140196603328320 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0421 15:59:48.327197 140196603328320 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 15:59:48.343797 140196603328320 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0421 15:59:48.343861 140196603328320 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 15:59:48.367911 140196603328320 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 15:59:48.367981 140196603328320 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 15:59:48.384967 140196603328320 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 15:59:48.385032 140196603328320 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 15:59:48.401910 140196603328320 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 15:59:48.401968 140196603328320 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 15:59:48.427506 140196603328320 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0421 15:59:48.427581 140196603328320 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 15:59:48.449147 140196603328320 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0421 15:59:48.449223 140196603328320 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0421 15:59:48.664006 140196603328320 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0421 15:59:48.664113 140196603328320 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0421 15:59:50.098008 140196603328320 maxtext_utils.py:1668]  params/params/decoder/decoder_norm/scale
    Shape:     float32[2048]
    Logical:   P('norm',)
    Physical:  (None,)
I0421 15:59:50.098135 140196603328320 maxtext_utils.py:1668]  params/params/decoder/layers/mlp/wi_0/kernel
    Shape:     float32[2048,16,7168]
    Logical:   P('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0421 15:59:50.098187 140196603328320 maxtext_utils.py:1668]  params/params/decoder/layers/mlp/wi_1/kernel
    Shape:     float32[2048,16,7168]
    Logical:   P('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0421 15:59:50.098245 140196603328320 maxtext_utils.py:1668]  params/params/decoder/layers/mlp/wo/kernel
    Shape:     float32[7168,16,2048]
    Logical:   P('mlp', 'layers', 'embed')
    Physical:  (None, None, 'fsdp')
I0421 15:59:50.098293 140196603328320 maxtext_utils.py:1668]  params/params/decoder/layers/post_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   P('norm', 'layers')
    Physical:  (None, None)
I0421 15:59:50.098331 140196603328320 maxtext_utils.py:1668]  params/params/decoder/layers/pre_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   P('norm', 'layers')
    Physical:  (None, None)
I0421 15:59:50.098382 140196603328320 maxtext_utils.py:1668]  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 15:59:50.098450 140196603328320 maxtext_utils.py:1668]  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 15:59:50.098495 140196603328320 maxtext_utils.py:1668]  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 15:59:50.098533 140196603328320 maxtext_utils.py:1668]  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 15:59:50.098582 140196603328320 maxtext_utils.py:1668]  params/params/decoder/logits_dense/kernel
    Shape:     float32[2048,32000]
    Logical:   P('embed_vocab', 'vocab')
    Physical:  ('fsdp', None)
I0421 15:59:50.098657 140196603328320 maxtext_utils.py:1668]  params/params/token_embedder/embedding

    Shape:     float32[32000,2048]
    Logical:   P('vocab', 'embed_vocab')
    Physical:  (None, 'fsdp')
I0421 15:59:50.597124 140196603328320 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0421 15:59:50.597221 140196603328320 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0421 15:59:50.612853 140196603328320 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0421 15:59:50.612917 140196603328320 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0421 16:00:01.482537 140196603328320 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 16:00:01.483373 140196603328320 metric_logger.py:301] number parameters: 1.104 billion
I0421 16:00:12.509063 140196603328320 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0421 16:00:13.223732 140196603328320 checkpointing.py:776] Waited 0.7146353721618652 seconds for step 0 to finish before starting checkpointing.
I0421 16:00:13.226047 140196603328320 checkpoint_manager.py:2009] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0421 16:00:13.227669 140196603328320 checkpoint_manager.py:1512] [process=5] Saving checkpoint at step 0
I0421 16:00:13.229294 140196603328320 event_tracking.py:70] [process=5] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_16_zarr3_true/checkpoints/0.
I0421 16:00:13.561619 140196603328320 signaling_client.py:364] Using JaxDistributedSignalingClient
I0421 16:00:13.562558 140196603328320 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 16:00:13.562617 140196603328320 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 16:00:13.847578 140196603328320 base_pytree_checkpoint_handler.py:154] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.286032s
I0421 16:00:13.847749 140196603328320 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/blocking_gbytes_per_sec: 5.287 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2917649745941162 s) (per-host)
I0421 16:00:13.847800 140196603328320 base_pytree_checkpoint_handler.py:768] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.291826s (batch_requests_ready=0.002536s, total_serialization_initiated=0.289222s, others=0.000068s)
I0421 16:00:13.847889 140196603328320 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.296086s (all_items=0.000017s, per_item={'items': '0.00001693'}, temp_paths=0.296069)
I0421 16:00:13.848703 140196603328320 event_tracking.py:125] [process=5] [async] Finished blocking save in 0.62 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_16_zarr3_true/checkpoints/0.
I0421 16:00:13.848999 140068161218304 async_checkpointer.py:76] [process=5][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 16:20:13.848966
I0421 16:00:13.850988 140196603328320 checkpoint_manager.py:1560] [process=5][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 16:00:13.851271 140066601039616 async_checkpointer.py:280] [process=5][thread=save_finalize] Waiting for background save thread=async_save.
I0421 16:00:13.851438 140196603328320 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776787213.2260299, 'wait_for_prev_duration_secs': 5.888938903808594e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776787213.2277138, 'checkpointer_blocking_duration_secs': 0.6214382648468018, 'get_old_steps_start_time': 1776787213.8491764, 'get_old_steps_duration_secs': 2.956390380859375e-05, 'checkpoint_manager_blocking_start_time': 1776787213.2241561, 'checkpoint_manager_blocking_duration_secs': 0.6272318363189697}
I0421 16:00:13.851540 140196603328320 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0421 16:00:13.851588 140196603328320 max_utils.py:750] 
Memstats: After params initialized:
I0421 16:00:13.851641 140196603328320 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_18(process=5,(2,4,0,0))
I0421 16:00:13.851674 140196603328320 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_19(process=5,(3,4,0,0))
I0421 16:00:13.851701 140196603328320 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_22(process=5,(2,5,0,0))
I0421 16:00:13.851728 140196603328320 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_23(process=5,(3,5,0,0))
I0421 16:00:14.162618 140196603328320 metric_logger.py:196] completed step: 0, seconds: 11.026, TFLOP/s/device: 1.232, Tokens/s/device: 185.750, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0421 16:00:14.340831 140196603328320 metric_logger.py:196] completed step: 1, seconds: 1.652, TFLOP/s/device: 8.225, Tokens/s/device: 1239.707, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0421 16:00:14.723242 140196603328320 metric_logger.py:196] completed step: 2, seconds: 0.031, TFLOP/s/device: 440.112, Tokens/s/device: 66338.430, total_weights: 65536, loss: 10.560, lm_loss: 10.560, perplexity: 38547.719
I0421 16:00:14.870543 140196603328320 metric_logger.py:196] completed step: 3, seconds: 0.383, TFLOP/s/device: 35.500, Tokens/s/device: 5350.989, total_weights: 65536, loss: 9.980, lm_loss: 9.980, perplexity: 21590.506
I0421 16:00:14.875180 140196603328320 checkpointing.py:772] Waiting for step 5 to finish before checkpoint...
I0421 16:00:15.164780 140196603328320 checkpointing.py:776] Waited 0.2895689010620117 seconds for step 5 to finish before starting checkpointing.
I0421 16:00:15.167768 140196603328320 checkpoint_manager.py:2020] [process=5][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 16:00:17.151083    2808 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0421 16:00:19.355130 140067620239104 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_16_zarr3_true/checkpoints/0/items/array_metadatas/process_5
I0421 16:00:50.771458 140068161218304 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/gbytes_per_sec: 42.445 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 37.21542453765869 s) (per-host)
I0421 16:00:50.771579 140068161218304 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 36.922469s.
I0421 16:00:59.680778 140068161218304 async_checkpointer.py:160] [process=5][thread=async_save] Background save thread done. Time taken: 45.831650s.
I0421 16:00:59.681073 140066601039616 async_checkpointer.py:288] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 16:00:59.681197 140066601039616 async_checkpointer.py:298] [process=5][thread=save_finalize] No errors found in background save thread=async_save.
I0421 16:00:59.681248 140066601039616 checkpoint_manager.py:2137] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0421 16:00:59.683678 140066601039616 checkpoint_manager.py:2146] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0421 16:00:59.683840 140196603328320 checkpoint_manager.py:2032] [process=5][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0421 16:00:59.683975 140196603328320 checkpoint_manager.py:1452] Waiting for previous save to complete took 44.516207 seconds. If this number is high, consider checkpointing less frequently.
I0421 16:00:59.686025 140196603328320 checkpoint_manager.py:1512] [process=5] Saving checkpoint at step 5
I0421 16:00:59.688555 140196603328320 event_tracking.py:70] [process=5] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_16_zarr3_true/checkpoints/5.
I0421 16:00:59.990334 140196603328320 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 16:00:59.990443 140196603328320 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 16:01:00.023909 140196603328320 base_pytree_checkpoint_handler.py:154] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.034391s
I0421 16:01:00.024081 140196603328320 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/blocking_gbytes_per_sec: 40.717 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.03788495063781738 s) (per-host)
I0421 16:01:00.024133 140196603328320 base_pytree_checkpoint_handler.py:768] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.037946s (batch_requests_ready=0.001774s, total_serialization_initiated=0.036103s, others=0.000069s)
I0421 16:01:00.024221 140196603328320 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.042814s (all_items=0.000015s, per_item={'items': '0.00001526'}, temp_paths=0.042799)
I0421 16:01:00.024947 140196603328320 event_tracking.py:125] [process=5] [async] Finished blocking save in 0.34 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_16_zarr3_true/checkpoints/5.
I0421 16:01:00.025240 140066601039616 async_checkpointer.py:76] [process=5][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 16:21:00.025207
I0421 16:01:00.035990 140196603328320 checkpoint_manager.py:1560] [process=5][thread=MainThread][step=5] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 16:01:00.036230 140067645114112 async_checkpointer.py:280] [process=5][thread=save_finalize] Waiting for background save thread=async_save.
I0421 16:01:00.036361 140196603328320 standard_logger.py:34] {'step': 5, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776787215.1677399, 'wait_for_prev_duration_secs': 44.51620697975159, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776787259.6860645, 'checkpointer_blocking_duration_secs': 0.3393216133117676, 'get_old_steps_start_time': 1776787260.0254073, 'get_old_steps_duration_secs': 4.220008850097656e-05, 'checkpoint_manager_blocking_start_time': 1776787215.165411, 'checkpoint_manager_blocking_duration_secs': 44.87091588973999}
I0421 16:01:00.036485 140196603328320 checkpointing.py:408] Started an asynchronous checkpoint save for step 5
I0421 16:01:00.037313 140196603328320 metric_logger.py:196] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.743, Tokens/s/device: 13376.354, total_weights: 65536, loss: 9.460, lm_loss: 9.460, perplexity: 12840.355
I0421 16:01:00.043340 140196603328320 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.250, Tokens/s/device: 13904.920, total_weights: 65536, loss: 8.959, lm_loss: 8.959, perplexity: 7776.467
I0421 16:01:05.403952 140067620239104 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_16_zarr3_true/checkpoints/5/items/array_metadatas/process_5
I0421 16:01:22.889485 140196603328320 metric_logger.py:196] completed step: 6, seconds: 45.167, TFLOP/s/device: 0.301, Tokens/s/device: 45.343, total_weights: 65536, loss: 8.469, lm_loss: 8.469, perplexity: 4762.740
I0421 16:01:23.036915 140196603328320 metric_logger.py:196] completed step: 7, seconds: 22.700, TFLOP/s/device: 0.599, Tokens/s/device: 90.219, total_weights: 65536, loss: 8.003, lm_loss: 8.003, perplexity: 2989.151
I0421 16:01:23.184178 140196603328320 metric_logger.py:196] completed step: 8, seconds: 0.152, TFLOP/s/device: 89.303, Tokens/s/device: 13460.755, total_weights: 65536, loss: 7.572, lm_loss: 7.572, perplexity: 1943.098
I0421 16:01:23.330871 140196603328320 checkpointing.py:772] Waiting for step 9 to finish before checkpoint...
I0421 16:01:23.331585 140196603328320 checkpointing.py:776] Waited 0.0007345676422119141 seconds for step 9 to finish before starting checkpointing.
I0421 16:01:23.333713 140196603328320 checkpoint_manager.py:2020] [process=5][thread=MainThread][step=5][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 16:01:42.554201 140066601039616 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/gbytes_per_sec: 37.108 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.56796479225159 s) (per-host)
I0421 16:01:42.554327 140066601039616 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.528976s.
I0421 16:01:50.015580 140066601039616 async_checkpointer.py:160] [process=5][thread=async_save] Background save thread done. Time taken: 49.990211s.
I0421 16:01:50.015805 140067645114112 async_checkpointer.py:288] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 16:01:50.015867 140067645114112 async_checkpointer.py:298] [process=5][thread=save_finalize] No errors found in background save thread=async_save.
I0421 16:01:50.015915 140067645114112 checkpoint_manager.py:2137] [process=5][thread=save_finalize][step=5] CheckpointManager Save Finalize is syncing with other hosts...
I0421 16:01:50.018342 140067645114112 checkpoint_manager.py:2146] [process=5][thread=save_finalize][step=5] CheckpointManager Save Finalize is done on all hosts.
I0421 16:01:50.018519 140196603328320 checkpoint_manager.py:2032] [process=5][thread=MainThread][step=5][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=5.
W0421 16:01:50.018650 140196603328320 checkpoint_manager.py:1452] Waiting for previous save to complete took 26.684942 seconds. If this number is high, consider checkpointing less frequently.
I0421 16:01:50.020115 140196603328320 checkpoint_manager.py:1512] [process=5] Saving checkpoint at step 9
I0421 16:01:50.022500 140196603328320 event_tracking.py:70] [process=5] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_16_zarr3_true/checkpoints/9.
I0421 16:01:50.749404 140196603328320 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 16:01:50.749511 140196603328320 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 16:01:50.783932 140196603328320 base_pytree_checkpoint_handler.py:154] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.035515s
I0421 16:01:50.784101 140196603328320 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/blocking_gbytes_per_sec: 39.474 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.0390779972076416 s) (per-host)
I0421 16:01:50.784153 140196603328320 base_pytree_checkpoint_handler.py:768] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.039141s (batch_requests_ready=0.001679s, total_serialization_initiated=0.037394s, others=0.000068s)
I0421 16:01:50.784243 140196603328320 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.043297s (all_items=0.000010s, per_item={'items': '0.00001001'}, temp_paths=0.043287)
I0421 16:01:50.784918 140196603328320 event_tracking.py:125] [process=5] [async] Finished blocking save in 0.76 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_16_zarr3_true/checkpoints/9.
I0421 16:01:50.785258 140067645114112 async_checkpointer.py:76] [process=5][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 16:21:50.785220
I0421 16:01:50.787235 140196603328320 checkpoint_manager.py:1560] [process=5][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 16:01:50.787515 140064403199744 async_checkpointer.py:280] [process=5][thread=save_finalize] Waiting for background save thread=async_save.
I0421 16:01:50.787674 140196603328320 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_16_zarr3_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776787283.3336825, 'wait_for_prev_duration_secs': 26.68494153022766, 'time_between_consecutive_saves_sec': 23.649964094161987, 'checkpointer_blocking_start_time': 1776787310.0201552, 'checkpointer_blocking_duration_secs': 0.765249490737915, 'get_old_steps_start_time': 1776787310.7854395, 'get_old_steps_duration_secs': 2.5272369384765625e-05, 'checkpoint_manager_blocking_start_time': 1776787283.331813, 'checkpoint_manager_blocking_duration_secs': 27.45582866668701}
I0421 16:01:50.787777 140196603328320 checkpointing.py:408] Started an asynchronous checkpoint save for step 9
I0421 16:01:50.787820 140196603328320 checkpoint_manager.py:2020] [process=5][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 16:01:56.395456 140067620239104 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_16_zarr3_true/checkpoints/9/items/array_metadatas/process_5
I0421 16:02:33.155307 140067645114112 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/gbytes_per_sec: 37.246 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.41025972366333 s) (per-host)
I0421 16:02:33.155409 140067645114112 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.370040s.
I0421 16:02:41.789876 140067645114112 async_checkpointer.py:160] [process=5][thread=async_save] Background save thread done. Time taken: 51.004488s.
I0421 16:02:41.790187 140064403199744 async_checkpointer.py:288] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 16:02:41.790310 140064403199744 async_checkpointer.py:298] [process=5][thread=save_finalize] No errors found in background save thread=async_save.
I0421 16:02:41.790352 140064403199744 checkpoint_manager.py:2137] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0421 16:02:41.791693 140064403199744 checkpoint_manager.py:2146] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0421 16:02:41.791971 140196603328320 checkpoint_manager.py:2032] [process=5][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0421 16:02:41.792139 140196603328320 checkpoint_manager.py:2009] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0421 16:02:41.793244 140196603328320 metric_logger.py:196] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.202, Tokens/s/device: 13897.748, total_weights: 65536, loss: 7.181, lm_loss: 7.181, perplexity: 1314.495
Per train step:
 Total TFLOPs: 13.59 
 split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Tue Apr 21 16:02:50 UTC 2026
EXIT_CODE=0