MaxView

XPK Start: Wed Apr 22 07:54:04 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. 
2026-04-22 07:54:29.306583: 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)
I0422 07:54:29.520171 137948336392000 max_utils.py:273] Attempting to initialize the jax distributed system...
I0422 07:54:38.559193 137948336392000 distributed.py:149] Starting JAX distributed service on [::]:8482
I0422 07:54:38.561521 137948336392000 distributed.py:172] Connecting to JAX distributed service on mt-11-optimizer-offload-va803-slice-job-0-0.mt-11-optimizer-offload-va803:8482
I0422 07:54:39.560363 137948336392000 max_utils.py:284] Jax distributed system initialized!
I0422 07:54:44.871335 137948336392000 max_utils.py:800] System Information: Jax Version: 0.9.2
I0422 07:54:44.871441 137948336392000 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0422 07:54:44.871480 137948336392000 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
I0422 07:54:44.871516 137948336392000 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0422 07:54:45.566043 137948336392000 maxtext_utils.py:1604] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0422 07:54:45.566344 137948336392000 checkpointing.py:677] Setting up checkpoint logger...
I0422 07:54:45.566402 137948336392000 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0422 07:54:45.566447 137948336392000 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0422 07:54:45.566787 137948336392000 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 0x7d75e1f1fe90>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0422 07:54:48.504633 137948336392000 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0422 07:54:48.504870 137948336392000 checkpoint_manager.py:708] [process=6][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d622c589a60>}, handler_registry=None
I0422 07:54:48.505103 137948336392000 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d622c589a60>` 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`.
I0422 07:54:48.505151 137948336392000 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d622c591760>` 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`.
I0422 07:54:48.505187 137948336392000 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 0x7d622c589a60>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7d622c589a60>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d622c591760>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7d622c591760>}).
I0422 07:54:48.505545 137948336392000 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0422 07:54:48.505615 137948336392000 async_checkpointer.py:192] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7d622c1b9580> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0422 07:54:50.074041 137948336392000 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_main_20260422_071422/linen_xpk_main_20260422_071422_11_optimizer_offload_true/checkpoints
I0422 07:54:51.365875 137948336392000 checkpoint_manager.py:929] [process=6][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_20260422_071422/linen_xpk_main_20260422_071422_11_optimizer_offload_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7d622c58b080>
I0422 07:54:51.366053 137948336392000 checkpointing.py:301] Checkpoint manager created!
I0422 07:54:52.289310 137948336392000 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0422 07:54:52.289431 137948336392000 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0422 07:54:52.671364 137948336392000 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0422 07:54:52.671460 137948336392000 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0422 07:54:52.687880 137948336392000 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0422 07:54:52.687941 137948336392000 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0422 07:54:52.711562 137948336392000 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0422 07:54:52.711637 137948336392000 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 07:54:52.727935 137948336392000 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0422 07:54:52.727996 137948336392000 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 07:54:52.744332 137948336392000 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0422 07:54:52.744397 137948336392000 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 07:54:52.774186 137948336392000 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0422 07:54:52.774303 137948336392000 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 07:54:52.795945 137948336392000 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0422 07:54:52.796011 137948336392000 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0422 07:54:53.006311 137948336392000 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0422 07:54:53.006421 137948336392000 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0422 07:54:54.597636 137948336392000 maxtext_utils.py:1707]  params/params/decoder/decoder_norm/scale
    Shape:     float32[2048]
    Logical:   P('norm',)
    Physical:  (None,)
I0422 07:54:54.597763 137948336392000 maxtext_utils.py:1707]  params/params/decoder/layers/mlp/wi_0/kernel
    Shape:     float32[2048,16,7168]
    Logical:   P('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0422 07:54:54.597817 137948336392000 maxtext_utils.py:1707]  params/params/decoder/layers/mlp/wi_1/kernel
    Shape:     float32[2048,16,7168]
    Logical:   P('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0422 07:54:54.597875 137948336392000 maxtext_utils.py:1707]  params/params/decoder/layers/mlp/wo/kernel
    Shape:     float32[7168,16,2048]
    Logical:   P('mlp', 'layers', 'embed')
    Physical:  (None, None, 'fsdp')
I0422 07:54:54.597930 137948336392000 maxtext_utils.py:1707]  params/params/decoder/layers/post_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   P('norm', 'layers')
    Physical:  (None, None)
I0422 07:54:54.597969 137948336392000 maxtext_utils.py:1707]  params/params/decoder/layers/pre_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   P('norm', 'layers')
    Physical:  (None, None)
I0422 07:54:54.598020 137948336392000 maxtext_utils.py:1707]  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)
I0422 07:54:54.598072 137948336392000 maxtext_utils.py:1707]  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')
I0422 07:54:54.598111 137948336392000 maxtext_utils.py:1707]  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)
I0422 07:54:54.598148 137948336392000 maxtext_utils.py:1707]  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)
I0422 07:54:54.598195 137948336392000 maxtext_utils.py:1707]  params/params/decoder/logits_dense/kernel
    Shape:     float32[2048,32000]
    Logical:   P('embed_vocab', 'vocab')
    Physical:  ('fsdp', None)
I0422 07:54:54.598241 137948336392000 maxtext_utils.py:1707]  params/params/token_embedder/embedding
    Shape:     float32[32000,2048]
    Logical:   P('vocab', 'embed_vocab')

    Physical:  (None, 'fsdp')
I0422 07:54:55.087557 137948336392000 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0422 07:54:55.087649 137948336392000 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0422 07:54:55.103096 137948336392000 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0422 07:54:55.103157 137948336392000 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0422 07:55:05.989418 137948336392000 max_utils.py:791] Total memory size: 1.3 GB, Output size: 0.1 GB, Temp size: 1.1 GB, Argument size: 0.1 GB, Host temp size: 0.0 GB.
I0422 07:55:05.990317 137948336392000 metric_logger.py:301] number parameters: 1.104 billion
I0422 07:55:17.669755 137948336392000 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0422 07:55:17.838344 137948336392000 checkpointing.py:776] Waited 0.16853690147399902 seconds for step 0 to finish before starting checkpointing.
I0422 07:55:17.840741 137948336392000 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0422 07:55:17.842523 137948336392000 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 0
I0422 07:55:17.844006 137948336392000 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260422_071422/linen_xpk_main_20260422_071422_11_optimizer_offload_true/checkpoints/0.
I0422 07:55:18.150923 137948336392000 signaling_client.py:364] Using JaxDistributedSignalingClient
I0422 07:55:18.151973 137948336392000 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0422 07:55:18.152029 137948336392000 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
I0422 07:55:18.422970 137948336392000 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.272123s
I0422 07:55:18.423151 137948336392000 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 5.538 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.27854108810424805 s) (per-host)
I0422 07:55:18.423202 137948336392000 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.278609s (batch_requests_ready=0.002281s, total_serialization_initiated=0.276252s, others=0.000076s)
I0422 07:55:18.423312 137948336392000 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.282653s (all_items=0.000018s, per_item={'items': '0.00001836'}, temp_paths=0.282634)
I0422 07:55:18.424104 137948336392000 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.58 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260422_071422/linen_xpk_main_20260422_071422_11_optimizer_offload_true/checkpoints/0.
I0422 07:55:18.424473 137815727777536 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-22 08:15:18.424434
I0422 07:55:18.428472 137948336392000 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0422 07:55:18.428773 137815586113280 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0422 07:55:18.428929 137948336392000 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260422_071422/linen_xpk_main_20260422_071422_11_optimizer_offload_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776844517.840723, 'wait_for_prev_duration_secs': 6.270408630371094e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776844517.842564, 'checkpointer_blocking_duration_secs': 0.5820703506469727, 'get_old_steps_start_time': 1776844518.42466, 'get_old_steps_duration_secs': 2.9325485229492188e-05, 'checkpoint_manager_blocking_start_time': 1776844517.8389087, 'checkpoint_manager_blocking_duration_secs': 0.5899815559387207}
I0422 07:55:18.429037 137948336392000 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0422 07:55:18.429096 137948336392000 max_utils.py:750] 
Memstats: After params initialized:
I0422 07:55:18.429144 137948336392000 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_24(process=6,(0,6,0,0))
I0422 07:55:18.429178 137948336392000 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_25(process=6,(1,6,0,0))
I0422 07:55:18.429206 137948336392000 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_28(process=6,(0,7,0,0))
I0422 07:55:18.429234 137948336392000 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_29(process=6,(1,7,0,0))
I0422 07:55:18.742632 137948336392000 metric_logger.py:196] completed step: 0, seconds: 11.679, TFLOP/s/device: 1.163, Tokens/s/device: 175.352, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0422 07:55:18.929448 137948336392000 metric_logger.py:196] completed step: 1, seconds: 1.071, TFLOP/s/device: 12.682, Tokens/s/device: 1911.637, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0422 07:55:19.339333 137948336392000 metric_logger.py:196] completed step: 2, seconds: 0.024, TFLOP/s/device: 567.620, Tokens/s/device: 85557.923, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28637.129
I0422 07:55:19.502312 137948336392000 metric_logger.py:196] completed step: 3, seconds: 0.410, TFLOP/s/device: 33.139, Tokens/s/device: 4995.049, total_weights: 65536, loss: 9.734, lm_loss: 9.734, perplexity: 16889.047
I0422 07:55:19.828365 137948336392000 metric_logger.py:196] completed step: 4, seconds: 0.169, TFLOP/s/device: 80.539, Tokens/s/device: 12139.749, total_weights: 65536, loss: 9.278, lm_loss: 9.278, perplexity: 10694.968
I0422 07:55:19.834508 137948336392000 metric_logger.py:196] completed step: 5, seconds: 0.163, TFLOP/s/device: 83.374, Tokens/s/device: 12567.039, total_weights: 65536, loss: 8.892, lm_loss: 8.892, perplexity: 7270.520
I0422 07:55:21.499222    2732 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0422 07:55:24.904046 137815594505984 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260422_071422/linen_xpk_main_20260422_071422_11_optimizer_offload_true/checkpoints/0/items/array_metadatas/process_6
I0422 07:55:42.139524 137948336392000 metric_logger.py:196] completed step: 6, seconds: 0.327, TFLOP/s/device: 41.586, Tokens/s/device: 6268.288, total_weights: 65536, loss: 8.592, lm_loss: 8.592, perplexity: 5390.688
I0422 07:55:42.302311 137948336392000 metric_logger.py:196] completed step: 7, seconds: 22.143, TFLOP/s/device: 0.614, Tokens/s/device: 92.490, total_weights: 65536, loss: 8.384, lm_loss: 8.384, perplexity: 4377.010
I0422 07:55:42.465031 137948336392000 metric_logger.py:196] completed step: 8, seconds: 0.168, TFLOP/s/device: 81.039, Tokens/s/device: 12215.125, total_weights: 65536, loss: 8.255, lm_loss: 8.255, perplexity: 3845.943
I0422 07:55:42.627360 137948336392000 checkpointing.py:772] Waiting for step 9 to finish before checkpoint...
I0422 07:55:42.627936 137948336392000 checkpointing.py:776] Waited 0.0005986690521240234 seconds for step 9 to finish before starting checkpointing.
I0422 07:55:42.630036 137948336392000 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0422 07:55:55.887483 137815727777536 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 41.850 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 37.74284219741821 s) (per-host)
I0422 07:55:55.887603 137815727777536 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 37.463013s.
I0422 07:56:04.708600 137815727777536 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 46.283994s.
I0422 07:56:04.708891 137815586113280 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0422 07:56:04.709008 137815586113280 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0422 07:56:04.709072 137815586113280 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0422 07:56:04.710690 137815586113280 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0422 07:56:04.710876 137948336392000 checkpoint_manager.py:2032] [process=6][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0422 07:56:04.711010 137948336392000 checkpoint_manager.py:1452] Waiting for previous save to complete took 22.080975 seconds. If this number is high, consider checkpointing less frequently.
I0422 07:56:04.712853 137948336392000 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 9
I0422 07:56:04.714958 137948336392000 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260422_071422/linen_xpk_main_20260422_071422_11_optimizer_offload_true/checkpoints/9.
I0422 07:56:05.426498 137948336392000 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0422 07:56:05.426595 137948336392000 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
I0422 07:56:05.460991 137948336392000 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.035435s
I0422 07:56:05.461162 137948336392000 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 39.678 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.038875579833984375 s) (per-host)
I0422 07:56:05.461212 137948336392000 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.038938s (batch_requests_ready=0.001829s, total_serialization_initiated=0.037039s, others=0.000070s)
I0422 07:56:05.461316 137948336392000 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.043830s (all_items=0.000015s, per_item={'items': '0.00001478'}, temp_paths=0.043815)
I0422 07:56:05.462079 137948336392000 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.75 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260422_071422/linen_xpk_main_20260422_071422_11_optimizer_offload_true/checkpoints/9.
I0422 07:56:05.462411 137815586113280 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-22 08:16:05.462377
I0422 07:56:05.464407 137948336392000 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0422 07:56:05.464686 137794013832960 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0422 07:56:05.464846 137948336392000 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_main_20260422_071422/linen_xpk_main_20260422_071422_11_optimizer_offload_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776844542.6300058, 'wait_for_prev_duration_secs': 22.080974817276, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776844564.712894, 'checkpointer_blocking_duration_secs': 0.7496590614318848, 'get_old_steps_start_time': 1776844565.4625773, 'get_old_steps_duration_secs': 3.0279159545898438e-05, 'checkpoint_manager_blocking_start_time': 1776844542.6281776, 'checkpoint_manager_blocking_duration_secs': 22.83663582801819}
I0422 07:56:05.464952 137948336392000 checkpointing.py:408] Started an asynchronous checkpoint save for step 9
I0422 07:56:05.464997 137948336392000 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0422 07:56:10.476979 137815594505984 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_main_20260422_071422/linen_xpk_main_20260422_071422_11_optimizer_offload_true/checkpoints/9/items/array_metadatas/process_6
I0422 07:56:48.469789 137815586113280 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 36.693 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 43.047465085983276 s) (per-host)
I0422 07:56:48.469914 137815586113280 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 43.007388s.
I0422 07:56:58.032939 137815586113280 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 52.570397s.
I0422 07:56:58.033242 137794013832960 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0422 07:56:58.033404 137794013832960 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0422 07:56:58.033454 137794013832960 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0422 07:56:58.035804 137794013832960 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0422 07:56:58.036001 137948336392000 checkpoint_manager.py:2032] [process=6][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0422 07:56:58.036169 137948336392000 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0422 07:56:58.037316 137948336392000 metric_logger.py:196] completed step: 9, seconds: 0.162, TFLOP/s/device: 83.626, Tokens/s/device: 12605.094, total_weights: 65536, loss: 8.179, lm_loss: 8.179, perplexity: 3564.739
Per train step:
 Total TFLOPs: 13.59 
 split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Wed Apr 22 07:57:09 UTC 2026
EXIT_CODE=0