XPK Start: Tue Apr 21 07:15:40 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 07:16:05.178106: 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 07:16:05.388440 133151012456256 max_utils.py:273] Attempting to initialize the jax distributed system...
I0421 07:16:14.428519 133151012456256 distributed.py:149] Starting JAX distributed service on [::]:8482
I0421 07:16:14.431216 133151012456256 distributed.py:172] Connecting to JAX distributed service on mt-15-ocdbt-true-sv-0tkqd-slice-job-0-0.mt-15-ocdbt-true-sv-0tkqd:8482
I0421 07:16:15.762737 133151012456256 max_utils.py:284] Jax distributed system initialized!
I0421 07:16:21.847621 133151012456256 max_utils.py:800] System Information: Jax Version: 0.9.2
I0421 07:16:21.847748 133151012456256 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0421 07:16:21.847791 133151012456256 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 07:16:21.847827 133151012456256 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0421 07:16:22.543474 133151012456256 maxtext_utils.py:1565] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0421 07:16:22.543772 133151012456256 checkpointing.py:677] Setting up checkpoint logger...
I0421 07:16:22.543827 133151012456256 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0421 07:16:22.543871 133151012456256 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0421 07:16:22.544220 133151012456256 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 0x7918eb334200>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0421 07:16:25.391769 133151012456256 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0421 07:16:25.392009 133151012456256 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 0x7917de7c7e00>}, handler_registry=None
I0421 07:16:25.392249 133151012456256 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7917de7c7e00>` 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 07:16:25.392297 133151012456256 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x790488486f90>` 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 07:16:25.392334 133151012456256 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 0x7917de7c7e00>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7917de7c7e00>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x790488486f90>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x790488486f90>}).
I0421 07:16:25.392657 133151012456256 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0421 07:16:25.392737 133151012456256 async_checkpointer.py:192] [process=4][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7904505c5300> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0421 07:16:26.530891 133151012456256 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints
I0421 07:16:26.963369 133151012456256 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=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_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x79048848b080>
I0421 07:16:26.963546 133151012456256 checkpointing.py:301] Checkpoint manager created!
I0421 07:16:27.906657 133151012456256 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0421 07:16:27.906794 133151012456256 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 07:16:28.293647 133151012456256 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0421 07:16:28.293756 133151012456256 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 07:16:28.310142 133151012456256 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0421 07:16:28.310205 133151012456256 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 07:16:28.333882 133151012456256 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 07:16:28.333954 133151012456256 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 07:16:28.350430 133151012456256 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 07:16:28.350490 133151012456256 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 07:16:28.366726 133151012456256 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 07:16:28.366783 133151012456256 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 07:16:28.391343 133151012456256 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0421 07:16:28.391414 133151012456256 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 07:16:28.413104 133151012456256 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0421 07:16:28.413183 133151012456256 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0421 07:16:28.621543 133151012456256 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0421 07:16:28.621647 133151012456256 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0421 07:16:30.050272 133151012456256 maxtext_utils.py:1668] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: P('norm',)
Physical: (None,)
I0421 07:16:30.050396 133151012456256 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 07:16:30.050449 133151012456256 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 07:16:30.050508 133151012456256 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 07:16:30.050560 133151012456256 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 07:16:30.050601 133151012456256 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 07:16:30.050653 133151012456256 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 07:16:30.050717 133151012456256 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 07:16:30.050762 133151012456256 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 07:16:30.050799 133151012456256 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 07:16:30.050848 133151012456256 maxtext_utils.py:1668] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: P('embed_vocab', 'vocab')
Physical: ('fsdp', None)
I0421 07:16:30.050897 133151012456256 maxtext_utils.py:1668] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: P('vocab', 'embed_vocab')
Physical: (None, 'fsdp')
I0421 07:16:30.538074 133151012456256 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0421 07:16:30.538167 133151012456256 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0421 07:16:30.553582 133151012456256 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0421 07:16:30.553640 133151012456256 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0421 07:16:41.319516 133151012456256 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 07:16:41.320739 133151012456256 metric_logger.py:301] number parameters: 1.104 billion
I0421 07:16:52.453248 133151012456256 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0421 07:16:53.071473 133151012456256 checkpointing.py:776] Waited 0.6181991100311279 seconds for step 0 to finish before starting checkpointing.
I0421 07:16:53.073858 133151012456256 checkpoint_manager.py:2009] [process=4][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0421 07:16:53.075654 133151012456256 checkpoint_manager.py:1512] [process=4] Saving checkpoint at step 0
I0421 07:16:53.077066 133151012456256 event_tracking.py:70] [process=4] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/0.
I0421 07:16:53.418819 133151012456256 signaling_client.py:364] Using JaxDistributedSignalingClient
I0421 07:16:53.419784 133151012456256 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 07:16:53.419842 133151012456256 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 07:16:53.696777 133151012456256 base_pytree_checkpoint_handler.py:154] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.278031s
I0421 07:16:53.696953 133151012456256 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/blocking_gbytes_per_sec: 5.442 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.28344011306762695 s) (per-host)
I0421 07:16:53.697005 133151012456256 base_pytree_checkpoint_handler.py:768] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.283502s (batch_requests_ready=0.002213s, total_serialization_initiated=0.281218s, others=0.000070s)
I0421 07:16:53.697098 133151012456256 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.287735s (all_items=0.000019s, per_item={'items': '0.00001907'}, temp_paths=0.287715)
I0421 07:16:53.697897 133151012456256 event_tracking.py:125] [process=4] [async] Finished blocking save in 0.62 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/0.
I0421 07:16:53.698178 133021809866496 async_checkpointer.py:76] [process=4][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 07:36:53.698145
I0421 07:16:54.116470 133151012456256 checkpoint_manager.py:1560] [process=4][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 07:16:54.116865 133021793081088 async_checkpointer.py:280] [process=4][thread=save_finalize] Waiting for background save thread=async_save.
I0421 07:16:54.117032 133151012456256 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776755813.07384, 'wait_for_prev_duration_secs': 6.127357482910156e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776755813.075707, 'checkpointer_blocking_duration_secs': 0.6226274967193604, 'get_old_steps_start_time': 1776755813.6983597, 'get_old_steps_duration_secs': 3.0279159545898438e-05, 'checkpoint_manager_blocking_start_time': 1776755813.0720234, 'checkpoint_manager_blocking_duration_secs': 1.0449671745300293}
I0421 07:16:54.117148 133151012456256 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0421 07:16:54.117207 133151012456256 max_utils.py:750]
Memstats: After params initialized:
I0421 07:16:54.117265 133151012456256 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_16(process=4,(0,4,0,0))
I0421 07:16:54.117299 133151012456256 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_17(process=4,(1,4,0,0))
I0421 07:16:54.117327 133151012456256 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_20(process=4,(0,5,0,0))
I0421 07:16:54.117352 133151012456256 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_21(process=4,(1,5,0,0))
I0421 07:16:54.424628 133151012456256 metric_logger.py:196] completed step: 0, seconds: 11.132, TFLOP/s/device: 1.221, Tokens/s/device: 183.968, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0421 07:16:54.619237 133151012456256 metric_logger.py:196] completed step: 1, seconds: 1.970, TFLOP/s/device: 6.897, Tokens/s/device: 1039.647, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0421 07:16:55.027374 133151012456256 metric_logger.py:196] completed step: 2, seconds: 0.047, TFLOP/s/device: 286.171, Tokens/s/device: 43134.860, total_weights: 65536, loss: 10.560, lm_loss: 10.560, perplexity: 38547.719
I0421 07:16:55.174573 133151012456256 metric_logger.py:196] completed step: 3, seconds: 0.408, TFLOP/s/device: 33.264, Tokens/s/device: 5013.930, total_weights: 65536, loss: 9.980, lm_loss: 9.980, perplexity: 21590.506
I0421 07:16:55.179184 133151012456256 checkpointing.py:772] Waiting for step 5 to finish before checkpoint...
I0421 07:16:55.468752 133151012456256 checkpointing.py:776] Waited 0.28949761390686035 seconds for step 5 to finish before starting checkpointing.
I0421 07:16:55.471754 133151012456256 checkpoint_manager.py:2020] [process=4][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 07:16:56.792659 2775 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0421 07:16:58.799742 133025975179008 array_metadata_store.py:203] [process=4][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/0/items/array_metadatas/process_4
I0421 07:17:29.940793 133021809866496 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/gbytes_per_sec: 43.244 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 36.52724003791809 s) (per-host)
I0421 07:17:29.940919 133021809866496 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 36.242626s.
I0421 07:17:39.845522 133021809866496 async_checkpointer.py:160] [process=4][thread=async_save] Background save thread done. Time taken: 46.147213s.
I0421 07:17:39.845845 133021793081088 async_checkpointer.py:288] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 07:17:39.845979 133021793081088 async_checkpointer.py:298] [process=4][thread=save_finalize] No errors found in background save thread=async_save.
I0421 07:17:39.846032 133021793081088 checkpoint_manager.py:2137] [process=4][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0421 07:17:39.848025 133021793081088 checkpoint_manager.py:2146] [process=4][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0421 07:17:39.848208 133151012456256 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 07:17:39.848349 133151012456256 checkpoint_manager.py:1452] Waiting for previous save to complete took 44.376594 seconds. If this number is high, consider checkpointing less frequently.
I0421 07:17:39.850401 133151012456256 checkpoint_manager.py:1512] [process=4] Saving checkpoint at step 5
I0421 07:17:39.852882 133151012456256 event_tracking.py:70] [process=4] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/5.
I0421 07:17:40.197928 133151012456256 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 07:17:40.198021 133151012456256 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 07:17:40.243896 133151012456256 base_pytree_checkpoint_handler.py:154] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.047221s
I0421 07:17:40.244060 133151012456256 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/blocking_gbytes_per_sec: 30.481 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.050608158111572266 s) (per-host)
I0421 07:17:40.244111 133151012456256 base_pytree_checkpoint_handler.py:768] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.050669s (batch_requests_ready=0.001770s, total_serialization_initiated=0.048832s, others=0.000067s)
I0421 07:17:40.244196 133151012456256 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.055590s (all_items=0.000016s, per_item={'items': '0.00001621'}, temp_paths=0.055574)
I0421 07:17:40.244973 133151012456256 event_tracking.py:125] [process=4] [async] Finished blocking save in 0.39 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/5.
I0421 07:17:40.245261 133021793081088 async_checkpointer.py:76] [process=4][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 07:37:40.245229
I0421 07:17:40.247240 133151012456256 checkpoint_manager.py:1560] [process=4][thread=MainThread][step=5] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 07:17:40.247516 133025975179008 async_checkpointer.py:280] [process=4][thread=save_finalize] Waiting for background save thread=async_save.
I0421 07:17:40.247679 133151012456256 standard_logger.py:34] {'step': 5, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776755815.4717245, 'wait_for_prev_duration_secs': 44.376593589782715, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776755859.8504407, 'checkpointer_blocking_duration_secs': 0.3949565887451172, 'get_old_steps_start_time': 1776755860.2454216, 'get_old_steps_duration_secs': 3.4809112548828125e-05, 'checkpoint_manager_blocking_start_time': 1776755815.4693446, 'checkpoint_manager_blocking_duration_secs': 44.778300762176514}
I0421 07:17:40.247804 133151012456256 checkpointing.py:408] Started an asynchronous checkpoint save for step 5
I0421 07:17:40.248534 133151012456256 metric_logger.py:196] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.952, Tokens/s/device: 13407.880, total_weights: 65536, loss: 9.460, lm_loss: 9.460, perplexity: 12840.355
I0421 07:17:40.255112 133151012456256 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.283, Tokens/s/device: 13909.831, total_weights: 65536, loss: 8.959, lm_loss: 8.959, perplexity: 7776.467
I0421 07:17:46.037682 133015224243968 array_metadata_store.py:203] [process=4][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/5/items/array_metadatas/process_4
I0421 07:18:02.444452 133151012456256 metric_logger.py:196] completed step: 6, seconds: 45.074, TFLOP/s/device: 0.301, Tokens/s/device: 45.436, total_weights: 65536, loss: 8.469, lm_loss: 8.469, perplexity: 4762.740
I0421 07:18:02.591560 133151012456256 metric_logger.py:196] completed step: 7, seconds: 22.043, TFLOP/s/device: 0.616, Tokens/s/device: 92.908, total_weights: 65536, loss: 8.003, lm_loss: 8.003, perplexity: 2989.151
I0421 07:18:02.738981 133151012456256 metric_logger.py:196] completed step: 8, seconds: 0.153, TFLOP/s/device: 88.910, Tokens/s/device: 13401.475, total_weights: 65536, loss: 7.572, lm_loss: 7.572, perplexity: 1943.098
I0421 07:18:02.885562 133151012456256 checkpointing.py:772] Waiting for step 9 to finish before checkpoint...
I0421 07:18:02.886242 133151012456256 checkpointing.py:776] Waited 0.0006997585296630859 seconds for step 9 to finish before starting checkpointing.
I0421 07:18:02.888272 133151012456256 checkpoint_manager.py:2020] [process=4][thread=MainThread][step=5][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 07:18:21.719510 133021793081088 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/gbytes_per_sec: 38.039 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.52601909637451 s) (per-host)
I0421 07:18:21.719631 133021793081088 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.474259s.
I0421 07:18:30.205688 133021793081088 async_checkpointer.py:160] [process=4][thread=async_save] Background save thread done. Time taken: 49.960300s.
I0421 07:18:30.206021 133025975179008 async_checkpointer.py:288] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 07:18:30.206139 133025975179008 async_checkpointer.py:298] [process=4][thread=save_finalize] No errors found in background save thread=async_save.
I0421 07:18:30.206188 133025975179008 checkpoint_manager.py:2137] [process=4][thread=save_finalize][step=5] CheckpointManager Save Finalize is syncing with other hosts...
I0421 07:18:30.207503 133025975179008 checkpoint_manager.py:2146] [process=4][thread=save_finalize][step=5] CheckpointManager Save Finalize is done on all hosts.
I0421 07:18:30.207685 133151012456256 checkpoint_manager.py:2032] [process=4][thread=MainThread][step=5][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=5.
W0421 07:18:30.207827 133151012456256 checkpoint_manager.py:1452] Waiting for previous save to complete took 27.319560 seconds. If this number is high, consider checkpointing less frequently.
I0421 07:18:30.209271 133151012456256 checkpoint_manager.py:1512] [process=4] Saving checkpoint at step 9
I0421 07:18:30.211270 133151012456256 event_tracking.py:70] [process=4] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/9.
I0421 07:18:30.924988 133151012456256 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 07:18:30.925083 133151012456256 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 07:18:30.950070 133151012456256 base_pytree_checkpoint_handler.py:154] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.026075s
I0421 07:18:30.950209 133151012456256 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/blocking_gbytes_per_sec: 52.448 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.02941131591796875 s) (per-host)
I0421 07:18:30.950256 133151012456256 base_pytree_checkpoint_handler.py:768] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.029468s (batch_requests_ready=0.001617s, total_serialization_initiated=0.027790s, others=0.000061s)
I0421 07:18:30.950328 133151012456256 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.033536s (all_items=0.000028s, per_item={'items': '0.00002813'}, temp_paths=0.033508)
I0421 07:18:30.950986 133151012456256 event_tracking.py:125] [process=4] [async] Finished blocking save in 0.74 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/9.
I0421 07:18:30.951318 133025975179008 async_checkpointer.py:76] [process=4][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 07:38:30.951280
I0421 07:18:30.961097 133151012456256 checkpoint_manager.py:1560] [process=4][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 07:18:30.961403 133015224243968 async_checkpointer.py:280] [process=4][thread=save_finalize] Waiting for background save thread=async_save.
I0421 07:18:30.961565 133151012456256 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776755882.8882408, 'wait_for_prev_duration_secs': 27.319560289382935, 'time_between_consecutive_saves_sec': 23.040175676345825, 'checkpointer_blocking_start_time': 1776755910.20931, 'checkpointer_blocking_duration_secs': 0.7421514987945557, 'get_old_steps_start_time': 1776755910.9514782, 'get_old_steps_duration_secs': 2.5510787963867188e-05, 'checkpoint_manager_blocking_start_time': 1776755882.8864732, 'checkpoint_manager_blocking_duration_secs': 28.075058698654175}
I0421 07:18:30.961679 133151012456256 checkpointing.py:408] Started an asynchronous checkpoint save for step 9
I0421 07:18:30.961738 133151012456256 checkpoint_manager.py:2020] [process=4][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 07:18:36.440145 133019113289472 array_metadata_store.py:203] [process=4][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_15_ocdbt_true/checkpoints/9/items/array_metadatas/process_4
I0421 07:19:13.626617 133025975179008 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/gbytes_per_sec: 36.988 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.705769300460815 s) (per-host)
I0421 07:19:13.626754 133025975179008 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.675327s.
I0421 07:19:21.416924 133025975179008 async_checkpointer.py:160] [process=4][thread=async_save] Background save thread done. Time taken: 50.465481s.
I0421 07:19:21.417205 133015224243968 async_checkpointer.py:288] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 07:19:21.417331 133015224243968 async_checkpointer.py:298] [process=4][thread=save_finalize] No errors found in background save thread=async_save.
I0421 07:19:21.417376 133015224243968 checkpoint_manager.py:2137] [process=4][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0421 07:19:21.418814 133015224243968 checkpoint_manager.py:2146] [process=4][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0421 07:19:21.418983 133151012456256 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 07:19:21.419129 133151012456256 checkpoint_manager.py:2009] [process=4][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0421 07:19:21.420096 133151012456256 metric_logger.py:196] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.483, Tokens/s/device: 13940.033, 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 07:19:30 UTC 2026
EXIT_CODE=0