MaxView

XPK Start: Tue Apr 21 06:33:25 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 06:33:50.535394: 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 06:33:50.744940 139326110558016 max_utils.py:273] Attempting to initialize the jax distributed system...
I0421 06:33:59.784782 139326110558016 distributed.py:149] Starting JAX distributed service on [::]:8482
I0421 06:33:59.787045 139326110558016 distributed.py:172] Connecting to JAX distributed service on mt-07-eval-20zob-slice-job-0-0.mt-07-eval-20zob:8482
I0421 06:34:01.331996 139326110558016 max_utils.py:284] Jax distributed system initialized!
I0421 06:34:07.546679 139326110558016 max_utils.py:800] System Information: Jax Version: 0.9.2
I0421 06:34:07.546800 139326110558016 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0421 06:34:07.546841 139326110558016 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 06:34:07.546877 139326110558016 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0421 06:34:08.320462 139326110558016 maxtext_utils.py:1565] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0421 06:34:08.320755 139326110558016 checkpointing.py:677] Setting up checkpoint logger...
I0421 06:34:08.320810 139326110558016 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0421 06:34:08.320855 139326110558016 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0421 06:34:08.321183 139326110558016 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 0x7eb6ee9ae6c0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0421 06:34:11.235323 139326110558016 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0421 06:34:11.235561 139326110558016 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 0x7ea244083cb0>}, handler_registry=None
I0421 06:34:11.235809 139326110558016 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ea244083cb0>` 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 06:34:11.235860 139326110558016 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ea244089220>` 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 06:34:11.235896 139326110558016 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 0x7ea244083cb0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7ea244083cb0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ea244089220>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7ea244089220>}).
I0421 06:34:11.236268 139326110558016 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0421 06:34:11.236344 139326110558016 async_checkpointer.py:192] [process=4][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7ea1504714e0> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0421 06:34:12.288941 139326110558016 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_07_eval/checkpoints
I0421 06:34:12.736588 139326110558016 checkpoint_manager.py:929] [process=4][thread=MainThread] CheckpointManager created,  primary_host=0, CheckpointManagerOptions=CheckpointManagerOptions(save_interval_steps=1, max_to_keep=None, keep_time_interval=None, keep_period=None, should_keep_fn=None, best_fn=None, best_mode='max', keep_checkpoints_without_metrics=True, step_prefix=None, step_format_fixed_length=None, step_name_format=None, create=True, cleanup_tmp_directories=False, save_on_steps=frozenset(), single_host_load_and_broadcast=False, todelete_subdir=None, todelete_full_path=None, enable_background_delete=False, read_only=False, enable_async_checkpointing=True, async_options=None, multiprocessing_options=MultiprocessingOptions(primary_host=0, active_processes=None, barrier_sync_key_prefix=None), should_save_fn=None, file_options=FileOptions(path_permission_mode=None), save_root_metadata=True, temporary_path_class=None, save_decision_policy=FixedIntervalPolicy(interval=10), preservation_policy=LatestN(n=None), prevent_write_metrics=False, enable_should_save_is_saving_in_progress_check=True, enable_per_process_directory_creation=False, lightweight_initialize=False), root_directory=gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_07_eval/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7ea244086690>
I0421 06:34:12.736780 139326110558016 checkpointing.py:301] Checkpoint manager created!
I0421 06:34:13.661294 139326110558016 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0421 06:34:13.661409 139326110558016 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 06:34:14.044863 139326110558016 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0421 06:34:14.044956 139326110558016 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 06:34:14.062156 139326110558016 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0421 06:34:14.062224 139326110558016 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 06:34:14.086185 139326110558016 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 06:34:14.086257 139326110558016 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 06:34:14.102818 139326110558016 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 06:34:14.102880 139326110558016 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 06:34:14.119268 139326110558016 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 06:34:14.119332 139326110558016 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 06:34:14.144003 139326110558016 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0421 06:34:14.144074 139326110558016 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 06:34:14.165123 139326110558016 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0421 06:34:14.165186 139326110558016 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0421 06:34:14.374298 139326110558016 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0421 06:34:14.374404 139326110558016 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0421 06:34:15.807862 139326110558016 maxtext_utils.py:1668]  params/params/decoder/decoder_norm/scale
    Shape:     float32[2048]
    Logical:   P('norm',)
    Physical:  (None,)
I0421 06:34:15.807992 139326110558016 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 06:34:15.808046 139326110558016 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 06:34:15.808105 139326110558016 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 06:34:15.808158 139326110558016 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 06:34:15.808202 139326110558016 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 06:34:15.808254 139326110558016 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 06:34:15.808304 139326110558016 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 06:34:15.808342 139326110558016 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 06:34:15.808379 139326110558016 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 06:34:15.808427 139326110558016 maxtext_utils.py:1668]  params/params/decoder/logits_dense/kernel
    Shape:     float32[2048,32000]
    Logical:   P('embed_vocab', 'vocab')
    Physical:  ('fsdp', None)
I0421 06:34:15.808476 139326110558016 maxtext_utils.py:1668]  params/params/token_embedder/embedding
    Shape:     float32[32000,2048]
    Logical:   P('vocab', 'embed_vocab')
    Physical:  (None, 'fsdp')
I0421 06:34:16.301322 139326110558016 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0421 06:34:16.301421 139326110558016 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0421 06:34:16.316770 139326110558016 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0421 06:34:16.316828 139326110558016 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0421 06:34:27.419350 139326110558016 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 06:34:27.420174 139326110558016 metric_logger.py:301] number parameters: 1.104 billion
I0421 06:34:38.666330 139326110558016 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0421 06:34:38.835984 139326110558016 checkpointing.py:776] Waited 0.16963768005371094 seconds for step 0 to finish before starting checkpointing.
I0421 06:34:38.838309 139326110558016 checkpoint_manager.py:2009] [process=4][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0421 06:34:38.840063 139326110558016 checkpoint_manager.py:1512] [process=4] Saving checkpoint at step 0
I0421 06:34:38.841532 139326110558016 event_tracking.py:70] [process=4] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_07_eval/checkpoints/0.
I0421 06:34:39.190538 139326110558016 signaling_client.py:364] Using JaxDistributedSignalingClient
I0421 06:34:39.191474 139326110558016 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 06:34:39.191533 139326110558016 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 06:34:39.469198 139326110558016 base_pytree_checkpoint_handler.py:154] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.278736s
I0421 06:34:39.469368 139326110558016 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/blocking_gbytes_per_sec: 5.427 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.28424072265625 s) (per-host)
I0421 06:34:39.469419 139326110558016 base_pytree_checkpoint_handler.py:768] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.284302s (batch_requests_ready=0.002303s, total_serialization_initiated=0.281930s, others=0.000069s)
I0421 06:34:39.469512 139326110558016 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.288336s (all_items=0.000019s, per_item={'items': '0.00001860'}, temp_paths=0.288317)
I0421 06:34:39.470309 139326110558016 event_tracking.py:125] [process=4] [async] Finished blocking save in 0.63 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_07_eval/checkpoints/0.
I0421 06:34:39.470638 139197620508416 async_checkpointer.py:76] [process=4][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 06:54:39.470600
I0421 06:34:39.479470 139326110558016 checkpoint_manager.py:1560] [process=4][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 06:34:39.479743 139196552066816 async_checkpointer.py:280] [process=4][thread=save_finalize] Waiting for background save thread=async_save.
I0421 06:34:39.479866 139326110558016 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_07_eval/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776753278.8382912, 'wait_for_prev_duration_secs': 6.175041198730469e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776753278.8401024, 'checkpointer_blocking_duration_secs': 0.6307046413421631, 'get_old_steps_start_time': 1776753279.4708307, 'get_old_steps_duration_secs': 3.170967102050781e-05, 'checkpoint_manager_blocking_start_time': 1776753278.8364296, 'checkpoint_manager_blocking_duration_secs': 0.643397331237793}
I0421 06:34:39.479971 139326110558016 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0421 06:34:39.480018 139326110558016 max_utils.py:750] 
Memstats: After params initialized:
I0421 06:34:39.480068 139326110558016 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_16(process=4,(0,4,0,0))
I0421 06:34:39.480101 139326110558016 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_17(process=4,(1,4,0,0))
I0421 06:34:39.480128 139326110558016 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_20(process=4,(0,5,0,0))
I0421 06:34:39.480161 139326110558016 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_21(process=4,(1,5,0,0))
I0421 06:34:39.790632 139326110558016 metric_logger.py:196] completed step: 0, seconds: 11.246, TFLOP/s/device: 1.208, Tokens/s/device: 182.108, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0421 06:34:39.966933 139326110558016 metric_logger.py:196] completed step: 1, seconds: 1.123, TFLOP/s/device: 12.101, Tokens/s/device: 1823.987, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0421 06:34:40.380750 139326110558016 metric_logger.py:196] completed step: 2, seconds: 0.029, TFLOP/s/device: 466.896, Tokens/s/device: 70375.588, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28637.129
I0421 06:34:40.528169 139326110558016 metric_logger.py:196] completed step: 3, seconds: 0.414, TFLOP/s/device: 32.793, Tokens/s/device: 4942.968, total_weights: 65536, loss: 9.734, lm_loss: 9.734, perplexity: 16889.203
I0421 06:34:40.823216 139326110558016 metric_logger.py:196] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.997, Tokens/s/device: 13414.554, total_weights: 65536, loss: 9.277, lm_loss: 9.277, perplexity: 10694.614
I0421 06:34:40.829180 139326110558016 metric_logger.py:196] completed step: 5, seconds: 0.148, TFLOP/s/device: 92.041, Tokens/s/device: 13873.365, total_weights: 65536, loss: 8.892, lm_loss: 8.892, perplexity: 7270.717
I0421 06:34:42.943864    2822 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0421 06:34:45.382431 139196560459520 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_07_eval/checkpoints/0/items/array_metadatas/process_4
I0421 06:35:02.517500 139326110558016 metric_logger.py:196] completed step: 6, seconds: 0.295, TFLOP/s/device: 46.007, Tokens/s/device: 6934.686, total_weights: 65536, loss: 8.592, lm_loss: 8.592, perplexity: 5390.759
I0421 06:35:02.664850 139326110558016 metric_logger.py:196] completed step: 7, seconds: 21.542, TFLOP/s/device: 0.631, Tokens/s/device: 95.071, total_weights: 65536, loss: 8.384, lm_loss: 8.384, perplexity: 4376.905
I0421 06:35:02.812168 139326110558016 metric_logger.py:196] completed step: 8, seconds: 0.152, TFLOP/s/device: 89.427, Tokens/s/device: 13479.360, total_weights: 65536, loss: 8.255, lm_loss: 8.255, perplexity: 3846.406
I0421 06:35:02.958761 139326110558016 checkpointing.py:772] Waiting for step 9 to finish before checkpoint...
I0421 06:35:02.959372 139326110558016 checkpointing.py:776] Waited 0.0006308555603027344 seconds for step 9 to finish before starting checkpointing.
I0421 06:35:02.961426 139326110558016 checkpoint_manager.py:2020] [process=4][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 06:35:17.450446 139197620508416 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/gbytes_per_sec: 41.280 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 38.26527953147888 s) (per-host)
I0421 06:35:17.450570 139197620508416 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 37.979801s.
I0421 06:35:23.625924 139197620508416 async_checkpointer.py:160] [process=4][thread=async_save] Background save thread done. Time taken: 44.155140s.
I0421 06:35:23.626216 139196552066816 async_checkpointer.py:288] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 06:35:23.626347 139196552066816 async_checkpointer.py:298] [process=4][thread=save_finalize] No errors found in background save thread=async_save.
I0421 06:35:23.626397 139196552066816 checkpoint_manager.py:2137] [process=4][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0421 06:35:23.628110 139196552066816 checkpoint_manager.py:2146] [process=4][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0421 06:35:23.628322 139326110558016 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 06:35:23.628485 139326110558016 checkpoint_manager.py:1452] Waiting for previous save to complete took 20.667051 seconds. If this number is high, consider checkpointing less frequently.
I0421 06:35:23.630450 139326110558016 checkpoint_manager.py:1512] [process=4] Saving checkpoint at step 9
I0421 06:35:23.632408 139326110558016 event_tracking.py:70] [process=4] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_07_eval/checkpoints/9.
I0421 06:35:23.912885 139326110558016 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 06:35:23.912995 139326110558016 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 06:35:23.953052 139326110558016 base_pytree_checkpoint_handler.py:154] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.041307s
I0421 06:35:23.953219 139326110558016 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/blocking_gbytes_per_sec: 33.852 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04556870460510254 s) (per-host)
I0421 06:35:23.953270 139326110558016 base_pytree_checkpoint_handler.py:768] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.045629s (batch_requests_ready=0.002613s, total_serialization_initiated=0.042948s, others=0.000068s)
I0421 06:35:23.953362 139326110558016 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.049850s (all_items=0.000015s, per_item={'items': '0.00001550'}, temp_paths=0.049835)
I0421 06:35:23.954077 139326110558016 event_tracking.py:125] [process=4] [async] Finished blocking save in 0.32 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_07_eval/checkpoints/9.
I0421 06:35:23.954369 139196552066816 async_checkpointer.py:76] [process=4][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 06:55:23.954337
I0421 06:35:23.956332 139326110558016 checkpoint_manager.py:1560] [process=4][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 06:35:23.956615 139172601898752 async_checkpointer.py:280] [process=4][thread=save_finalize] Waiting for background save thread=async_save.
I0421 06:35:23.956769 139326110558016 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260421_061359/linen_xpk_main_20260421_061359_07_eval/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776753302.9613967, 'wait_for_prev_duration_secs': 20.667051076889038, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776753323.6304967, 'checkpointer_blocking_duration_secs': 0.32401132583618164, 'get_old_steps_start_time': 1776753323.9545307, 'get_old_steps_duration_secs': 3.337860107421875e-05, 'checkpoint_manager_blocking_start_time': 1776753302.959602, 'checkpoint_manager_blocking_duration_secs': 20.9971342086792}
I0421 06:35:23.956929 139326110558016 checkpointing.py:408] Started an asynchronous checkpoint save for step 9
I0421 06:35:23.956977 139326110558016 checkpoint_manager.py:2020] [process=4][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 06:35:29.995953 139196560459520 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_07_eval/checkpoints/9/items/array_metadatas/process_4
I0421 06:36:07.000305 139196552066816 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/gbytes_per_sec: 36.656 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 43.09262275695801 s) (per-host)
I0421 06:36:07.000414 139196552066816 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 43.045936s.
I0421 06:36:14.599174 139196552066816 async_checkpointer.py:160] [process=4][thread=async_save] Background save thread done. Time taken: 50.644682s.
I0421 06:36:14.599477 139172601898752 async_checkpointer.py:288] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 06:36:14.599600 139172601898752 async_checkpointer.py:298] [process=4][thread=save_finalize] No errors found in background save thread=async_save.
I0421 06:36:14.599648 139172601898752 checkpoint_manager.py:2137] [process=4][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0421 06:36:14.603534 139172601898752 checkpoint_manager.py:2146] [process=4][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0421 06:36:14.603757 139326110558016 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 06:36:14.603899 139326110558016 checkpoint_manager.py:2009] [process=4][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0421 06:36:14.604885 139326110558016 metric_logger.py:196] completed step: 9, seconds: 0.148, TFLOP/s/device: 92.089, Tokens/s/device: 13880.699, total_weights: 65536, loss: 8.179, lm_loss: 8.179, perplexity: 3564.635
Per train step:
 Total TFLOPs: 13.59 
 split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Tue Apr 21 06:36:25 UTC 2026
EXIT_CODE=0