MaxView

XPK Start: Tue Apr 21 15:04:20 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 15:04:45.365256: 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 15:04:45.577939 132465227835200 max_utils.py:273] Attempting to initialize the jax distributed system...
I0421 15:04:54.617842 132465227835200 distributed.py:149] Starting JAX distributed service on [::]:8482
I0421 15:04:54.620313 132465227835200 distributed.py:172] Connecting to JAX distributed service on mt-08-checkpoint-async--cioms-slice-job-0-0.mt-08-checkpoint-async--cioms:8482
I0421 15:04:55.780212 132465227835200 max_utils.py:284] Jax distributed system initialized!
I0421 15:05:01.939207 132465227835200 max_utils.py:800] System Information: Jax Version: 0.9.2
I0421 15:05:01.939317 132465227835200 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0421 15:05:01.939358 132465227835200 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 15:05:01.939393 132465227835200 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0421 15:05:02.632603 132465227835200 maxtext_utils.py:1565] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0421 15:05:02.632906 132465227835200 checkpointing.py:677] Setting up checkpoint logger...
I0421 15:05:02.632963 132465227835200 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0421 15:05:02.633008 132465227835200 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0421 15:05:02.633340 132465227835200 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 0x78796cf889e0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0421 15:05:06.006602 132465227835200 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0421 15:05:06.006876 132465227835200 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 0x7865301b9700>}, handler_registry=None
I0421 15:05:06.007152 132465227835200 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7865301b9700>` 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 15:05:06.007207 132465227835200 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7865301bcb30>` 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 15:05:06.007260 132465227835200 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 0x7865301b9700>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7865301b9700>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7865301bcb30>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7865301bcb30>}).
I0421 15:05:06.007626 132465227835200 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0421 15:05:06.007709 132465227835200 async_checkpointer.py:192] [process=4][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x786408671300> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0421 15:05:07.391685 132465227835200 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_08_checkpoint_async_true/checkpoints
I0421 15:05:07.480459 132465227835200 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_144122/linen_xpk_main_20260421_144122_08_checkpoint_async_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7865301b9970>
I0421 15:05:07.480634 132465227835200 checkpointing.py:301] Checkpoint manager created!
I0421 15:05:08.410668 132465227835200 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0421 15:05:08.410787 132465227835200 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 15:05:08.792992 132465227835200 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0421 15:05:08.793085 132465227835200 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 15:05:08.809713 132465227835200 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0421 15:05:08.809774 132465227835200 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 15:05:08.834557 132465227835200 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 15:05:08.834627 132465227835200 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 15:05:08.851430 132465227835200 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 15:05:08.851494 132465227835200 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 15:05:08.868221 132465227835200 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 15:05:08.868285 132465227835200 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 15:05:08.893679 132465227835200 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0421 15:05:08.893748 132465227835200 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 15:05:08.915337 132465227835200 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0421 15:05:08.915405 132465227835200 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0421 15:05:09.133821 132465227835200 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0421 15:05:09.133933 132465227835200 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0421 15:05:10.579372 132465227835200 maxtext_utils.py:1668]  params/params/decoder/decoder_norm/scale
    Shape:     float32[2048]
    Logical:   P('norm',)
    Physical:  (None,)
I0421 15:05:10.579505 132465227835200 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 15:05:10.579561 132465227835200 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 15:05:10.579623 132465227835200 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 15:05:10.579693 132465227835200 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 15:05:10.579736 132465227835200 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 15:05:10.579792 132465227835200 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 15:05:10.579846 132465227835200 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 15:05:10.579887 132465227835200 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 15:05:10.579924 132465227835200 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 15:05:10.579972 132465227835200 maxtext_utils.py:1668]  params/params/decoder/logits_dense/kernel
    Shape:     float32[2048,32000]
    Logical:   P('embed_vocab', 'vocab')
    Physical:  ('fsdp', None)
I0421 15:05:10.580020 132465227835200 maxtext_utils.py:1668]  params/params/token_embedder/embedding
    Shape:     float32[32000,2048]
    Logical:   P('vocab', 'embed_vocab')
    Physical:  (None, 'fsdp')

I0421 15:05:11.083000 132465227835200 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0421 15:05:11.083095 132465227835200 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0421 15:05:11.098736 132465227835200 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0421 15:05:11.098793 132465227835200 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0421 15:05:22.001513 132465227835200 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 15:05:22.002363 132465227835200 metric_logger.py:301] number parameters: 1.104 billion
I0421 15:05:33.032205 132465227835200 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0421 15:05:33.705476 132465227835200 checkpointing.py:776] Waited 0.6732511520385742 seconds for step 0 to finish before starting checkpointing.
I0421 15:05:33.707861 132465227835200 checkpoint_manager.py:2009] [process=4][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0421 15:05:33.709725 132465227835200 checkpoint_manager.py:1512] [process=4] Saving checkpoint at step 0
I0421 15:05:33.711070 132465227835200 event_tracking.py:70] [process=4] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_08_checkpoint_async_true/checkpoints/0.
I0421 15:05:34.309060 132465227835200 signaling_client.py:364] Using JaxDistributedSignalingClient
I0421 15:05:34.309921 132465227835200 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 15:05:34.309978 132465227835200 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 15:05:34.586334 132465227835200 base_pytree_checkpoint_handler.py:154] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.277350s
I0421 15:05:34.586500 132465227835200 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/blocking_gbytes_per_sec: 5.447 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.28317809104919434 s) (per-host)
I0421 15:05:34.586552 132465227835200 base_pytree_checkpoint_handler.py:768] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.283241s (batch_requests_ready=0.002556s, total_serialization_initiated=0.280614s, others=0.000071s)
I0421 15:05:34.586654 132465227835200 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.287342s (all_items=0.000019s, per_item={'items': '0.00001907'}, temp_paths=0.287323)
I0421 15:05:34.587431 132465227835200 event_tracking.py:125] [process=4] [async] Finished blocking save in 0.88 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_08_checkpoint_async_true/checkpoints/0.
I0421 15:05:34.587784 132336353781504 async_checkpointer.py:76] [process=4][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 15:25:34.587745
I0421 15:05:34.598319 132465227835200 checkpoint_manager.py:1560] [process=4][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 15:05:34.598634 132334749734656 async_checkpointer.py:280] [process=4][thread=save_finalize] Waiting for background save thread=async_save.
I0421 15:05:34.598812 132465227835200 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_08_checkpoint_async_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776783933.707844, 'wait_for_prev_duration_secs': 6.031990051269531e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776783933.7097635, 'checkpointer_blocking_duration_secs': 0.8781702518463135, 'get_old_steps_start_time': 1776783934.5879586, 'get_old_steps_duration_secs': 2.86102294921875e-05, 'checkpoint_manager_blocking_start_time': 1776783933.705941, 'checkpoint_manager_blocking_duration_secs': 0.8928318023681641}
I0421 15:05:34.598922 132465227835200 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0421 15:05:34.598970 132465227835200 max_utils.py:750] 
Memstats: After params initialized:
I0421 15:05:34.599020 132465227835200 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_16(process=4,(0,4,0,0))
I0421 15:05:34.599053 132465227835200 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_17(process=4,(1,4,0,0))
I0421 15:05:34.599081 132465227835200 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_20(process=4,(0,5,0,0))
I0421 15:05:34.599107 132465227835200 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_21(process=4,(1,5,0,0))
I0421 15:05:34.909237 132465227835200 metric_logger.py:196] completed step: 0, seconds: 11.030, TFLOP/s/device: 1.232, Tokens/s/device: 185.680, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0421 15:05:35.174863 132465227835200 metric_logger.py:196] completed step: 1, seconds: 1.876, TFLOP/s/device: 7.244, Tokens/s/device: 1091.966, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0421 15:05:35.595723 132465227835200 metric_logger.py:196] completed step: 2, seconds: 0.118, TFLOP/s/device: 114.891, Tokens/s/device: 17317.628, total_weights: 65536, loss: 10.560, lm_loss: 10.560, perplexity: 38547.719
I0421 15:05:35.743153 132465227835200 metric_logger.py:196] completed step: 3, seconds: 0.421, TFLOP/s/device: 32.257, Tokens/s/device: 4862.171, total_weights: 65536, loss: 9.980, lm_loss: 9.980, perplexity: 21590.506
I0421 15:05:35.747570 132465227835200 checkpointing.py:772] Waiting for step 5 to finish before checkpoint...
I0421 15:05:36.037224 132465227835200 checkpointing.py:776] Waited 0.28962182998657227 seconds for step 5 to finish before starting checkpointing.
I0421 15:05:36.040132 132465227835200 checkpoint_manager.py:2020] [process=4][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 15:05:37.745204    2843 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0421 15:05:39.765428 132335270553344 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_144122/linen_xpk_main_20260421_144122_08_checkpoint_async_true/checkpoints/0/items/array_metadatas/process_4
I0421 15:06:11.790275 132336353781504 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/gbytes_per_sec: 42.137 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 37.486912965774536 s) (per-host)
I0421 15:06:11.790409 132336353781504 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 37.202515s.
I0421 15:06:19.086285 132336353781504 async_checkpointer.py:160] [process=4][thread=async_save] Background save thread done. Time taken: 44.498375s.
I0421 15:06:19.086600 132334749734656 async_checkpointer.py:288] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 15:06:19.086728 132334749734656 async_checkpointer.py:298] [process=4][thread=save_finalize] No errors found in background save thread=async_save.
I0421 15:06:19.086781 132334749734656 checkpoint_manager.py:2137] [process=4][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0421 15:06:19.088302 132334749734656 checkpoint_manager.py:2146] [process=4][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0421 15:06:19.088480 132465227835200 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 15:06:19.088616 132465227835200 checkpoint_manager.py:1452] Waiting for previous save to complete took 43.048482 seconds. If this number is high, consider checkpointing less frequently.
I0421 15:06:19.090713 132465227835200 checkpoint_manager.py:1512] [process=4] Saving checkpoint at step 5
I0421 15:06:19.093236 132465227835200 event_tracking.py:70] [process=4] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_08_checkpoint_async_true/checkpoints/5.
I0421 15:06:19.801956 132465227835200 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 15:06:19.802055 132465227835200 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 15:06:19.844458 132465227835200 base_pytree_checkpoint_handler.py:154] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.043634s
I0421 15:06:19.844636 132465227835200 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/blocking_gbytes_per_sec: 32.736 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.047121524810791016 s) (per-host)
I0421 15:06:19.844701 132465227835200 base_pytree_checkpoint_handler.py:768] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.047197s (batch_requests_ready=0.001827s, total_serialization_initiated=0.045286s, others=0.000084s)
I0421 15:06:19.844792 132465227835200 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.052058s (all_items=0.000017s, per_item={'items': '0.00001693'}, temp_paths=0.052041)
I0421 15:06:19.845510 132465227835200 event_tracking.py:125] [process=4] [async] Finished blocking save in 0.75 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_08_checkpoint_async_true/checkpoints/5.
I0421 15:06:19.845823 132334749734656 async_checkpointer.py:76] [process=4][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 15:26:19.845790
I0421 15:06:19.847851 132465227835200 checkpoint_manager.py:1560] [process=4][thread=MainThread][step=5] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 15:06:19.848143 132336345388800 async_checkpointer.py:280] [process=4][thread=save_finalize] Waiting for background save thread=async_save.
I0421 15:06:19.848277 132465227835200 standard_logger.py:34] {'step': 5, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_08_checkpoint_async_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776783936.0401032, 'wait_for_prev_duration_secs': 43.048482179641724, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776783979.0907533, 'checkpointer_blocking_duration_secs': 0.7552139759063721, 'get_old_steps_start_time': 1776783979.8459918, 'get_old_steps_duration_secs': 3.0279159545898438e-05, 'checkpoint_manager_blocking_start_time': 1776783936.0378659, 'checkpoint_manager_blocking_duration_secs': 43.81037759780884}
I0421 15:06:19.848438 132465227835200 checkpointing.py:408] Started an asynchronous checkpoint save for step 5
I0421 15:06:19.849287 132465227835200 metric_logger.py:196] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.783, Tokens/s/device: 13382.385, total_weights: 65536, loss: 9.460, lm_loss: 9.460, perplexity: 12840.355
I0421 15:06:19.855455 132465227835200 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.284, Tokens/s/device: 13910.114, total_weights: 65536, loss: 8.959, lm_loss: 8.959, perplexity: 7776.467
I0421 15:06:25.398536 132335270553344 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_144122/linen_xpk_main_20260421_144122_08_checkpoint_async_true/checkpoints/5/items/array_metadatas/process_4
I0421 15:06:42.718101 132465227835200 metric_logger.py:196] completed step: 6, seconds: 44.106, TFLOP/s/device: 0.308, Tokens/s/device: 46.433, total_weights: 65536, loss: 8.469, lm_loss: 8.469, perplexity: 4762.740
I0421 15:06:42.865603 132465227835200 metric_logger.py:196] completed step: 7, seconds: 22.717, TFLOP/s/device: 0.598, Tokens/s/device: 90.154, total_weights: 65536, loss: 8.003, lm_loss: 8.003, perplexity: 2989.151
I0421 15:06:43.012887 132465227835200 metric_logger.py:196] completed step: 8, seconds: 0.152, TFLOP/s/device: 89.364, Tokens/s/device: 13469.874, total_weights: 65536, loss: 7.572, lm_loss: 7.572, perplexity: 1943.098
I0421 15:06:43.159552 132465227835200 checkpointing.py:772] Waiting for step 9 to finish before checkpoint...
I0421 15:06:43.160250 132465227835200 checkpointing.py:776] Waited 0.0007193088531494141 seconds for step 9 to finish before starting checkpointing.
I0421 15:06:43.162408 132465227835200 checkpoint_manager.py:2020] [process=4][thread=MainThread][step=5][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 15:07:01.931316 132334749734656 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/gbytes_per_sec: 37.490 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.13376212120056 s) (per-host)
I0421 15:07:01.931441 132334749734656 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.085506s.
I0421 15:07:10.105312 132334749734656 async_checkpointer.py:160] [process=4][thread=async_save] Background save thread done. Time taken: 50.259362s.
I0421 15:07:10.105635 132336345388800 async_checkpointer.py:288] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 15:07:10.105820 132336345388800 async_checkpointer.py:298] [process=4][thread=save_finalize] No errors found in background save thread=async_save.
I0421 15:07:10.105870 132336345388800 checkpoint_manager.py:2137] [process=4][thread=save_finalize][step=5] CheckpointManager Save Finalize is syncing with other hosts...
I0421 15:07:10.107278 132336345388800 checkpoint_manager.py:2146] [process=4][thread=save_finalize][step=5] CheckpointManager Save Finalize is done on all hosts.
I0421 15:07:10.107450 132465227835200 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 15:07:10.107608 132465227835200 checkpoint_manager.py:1452] Waiting for previous save to complete took 26.945206 seconds. If this number is high, consider checkpointing less frequently.
I0421 15:07:10.109027 132465227835200 checkpoint_manager.py:1512] [process=4] Saving checkpoint at step 9
I0421 15:07:10.110984 132465227835200 event_tracking.py:70] [process=4] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_08_checkpoint_async_true/checkpoints/9.
I0421 15:07:10.397665 132465227835200 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 15:07:10.397759 132465227835200 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 15:07:10.440495 132465227835200 base_pytree_checkpoint_handler.py:154] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.043616s
I0421 15:07:10.440664 132465227835200 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/blocking_gbytes_per_sec: 32.731 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04712867736816406 s) (per-host)
I0421 15:07:10.440715 132465227835200 base_pytree_checkpoint_handler.py:768] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.047200s (batch_requests_ready=0.001697s, total_serialization_initiated=0.045428s, others=0.000075s)
I0421 15:07:10.440803 132465227835200 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.051131s (all_items=0.000011s, per_item={'items': '0.00001121'}, temp_paths=0.051119)
I0421 15:07:10.441449 132465227835200 event_tracking.py:125] [process=4] [async] Finished blocking save in 0.33 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_08_checkpoint_async_true/checkpoints/9.
I0421 15:07:10.441793 132336345388800 async_checkpointer.py:76] [process=4][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 15:27:10.441757
I0421 15:07:10.443746 132465227835200 checkpoint_manager.py:1560] [process=4][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 15:07:10.444029 132335850465024 async_checkpointer.py:280] [process=4][thread=save_finalize] Waiting for background save thread=async_save.
I0421 15:07:10.444183 132465227835200 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_144122/linen_xpk_main_20260421_144122_08_checkpoint_async_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776784003.16238, 'wait_for_prev_duration_secs': 26.94520616531372, 'time_between_consecutive_saves_sec': 24.074039220809937, 'checkpointer_blocking_start_time': 1776784030.1090658, 'checkpointer_blocking_duration_secs': 0.3328721523284912, 'get_old_steps_start_time': 1776784030.4419544, 'get_old_steps_duration_secs': 2.3126602172851562e-05, 'checkpoint_manager_blocking_start_time': 1776784003.1604922, 'checkpoint_manager_blocking_duration_secs': 27.283660650253296}
I0421 15:07:10.444284 132465227835200 checkpointing.py:408] Started an asynchronous checkpoint save for step 9
I0421 15:07:10.444325 132465227835200 checkpoint_manager.py:2020] [process=4][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 15:07:16.013427 132335861233408 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_144122/linen_xpk_main_20260421_144122_08_checkpoint_async_true/checkpoints/9/items/array_metadatas/process_4
I0421 15:07:53.435534 132336345388800 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/gbytes_per_sec: 36.699 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 43.0419704914093 s) (per-host)
I0421 15:07:53.435668 132336345388800 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.993762s.
I0421 15:08:00.170984 132336345388800 async_checkpointer.py:160] [process=4][thread=async_save] Background save thread done. Time taken: 49.729063s.
I0421 15:08:00.171270 132335850465024 async_checkpointer.py:288] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 15:08:00.171402 132335850465024 async_checkpointer.py:298] [process=4][thread=save_finalize] No errors found in background save thread=async_save.
I0421 15:08:00.171446 132335850465024 checkpoint_manager.py:2137] [process=4][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0421 15:08:00.174329 132335850465024 checkpoint_manager.py:2146] [process=4][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0421 15:08:00.174499 132465227835200 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 15:08:00.174661 132465227835200 checkpoint_manager.py:2009] [process=4][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0421 15:08:00.175807 132465227835200 metric_logger.py:196] completed step: 9, seconds: 0.148, TFLOP/s/device: 91.958, Tokens/s/device: 13860.971, 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 15:08:11 UTC 2026
EXIT_CODE=0