MaxView

XPK Start: Wed Apr 22 07:52:27 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 07:52:52.680124: 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 07:52:52.891762 138272577578816 max_utils.py:273] Attempting to initialize the jax distributed system...
I0422 07:53:01.935559 138272577578816 distributed.py:149] Starting JAX distributed service on [::]:8482
I0422 07:53:01.938032 138272577578816 distributed.py:172] Connecting to JAX distributed service on mt-11-optimizer-offload-6hsa4-slice-job-0-0.mt-11-optimizer-offload-6hsa4:8482
I0422 07:53:03.732848 138272577578816 max_utils.py:284] Jax distributed system initialized!
I0422 07:53:08.889832 138272577578816 max_utils.py:800] System Information: Jax Version: 0.9.2
I0422 07:53:08.889937 138272577578816 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0422 07:53:08.889977 138272577578816 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 07:53:08.890011 138272577578816 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0422 07:53:09.585953 138272577578816 maxtext_utils.py:1565] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0422 07:53:09.586241 138272577578816 checkpointing.py:677] Setting up checkpoint logger...
I0422 07:53:09.586299 138272577578816 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0422 07:53:09.586344 138272577578816 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0422 07:53:09.586700 138272577578816 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 0x7dc179fe7680>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0422 07:53:12.416175 138272577578816 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0422 07:53:12.416415 138272577578816 checkpoint_manager.py:708] [process=2][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7dade41d9d30>}, handler_registry=None
I0422 07:53:12.416665 138272577578816 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7dade41d9d30>` 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 07:53:12.416727 138272577578816 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7dade41dd880>` 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 07:53:12.416766 138272577578816 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 0x7dade41d9d30>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7dade41d9d30>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7dade41dd880>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7dade41dd880>}).
I0422 07:53:12.417097 138272577578816 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0422 07:53:12.417165 138272577578816 async_checkpointer.py:192] [process=2][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7dadb46e9440> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0422 07:53:13.201467 138272577578816 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_false/checkpoints
I0422 07:53:13.227210 138272577578816 checkpoint_manager.py:929] [process=2][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_false/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7dade41da4b0>
I0422 07:53:13.227347 138272577578816 checkpointing.py:301] Checkpoint manager created!
I0422 07:53:14.147528 138272577578816 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0422 07:53:14.147648 138272577578816 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0422 07:53:14.535388 138272577578816 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0422 07:53:14.535484 138272577578816 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0422 07:53:14.551950 138272577578816 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0422 07:53:14.552015 138272577578816 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0422 07:53:14.575846 138272577578816 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 07:53:14.575913 138272577578816 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 07:53:14.592558 138272577578816 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 07:53:14.592618 138272577578816 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 07:53:14.609143 138272577578816 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 07:53:14.609205 138272577578816 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 07:53:14.634343 138272577578816 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0422 07:53:14.634418 138272577578816 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0422 07:53:14.655630 138272577578816 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0422 07:53:14.655697 138272577578816 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0422 07:53:14.869659 138272577578816 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0422 07:53:14.869790 138272577578816 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0422 07:53:16.316678 138272577578816 maxtext_utils.py:1668]  params/params/decoder/decoder_norm/scale
    Shape:     float32[2048]
    Logical:   P('norm',)
    Physical:  (None,)
I0422 07:53:16.316821 138272577578816 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)
I0422 07:53:16.316875 138272577578816 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)
I0422 07:53:16.316939 138272577578816 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')
I0422 07:53:16.316995 138272577578816 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)
I0422 07:53:16.317034 138272577578816 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)
I0422 07:53:16.317087 138272577578816 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)
I0422 07:53:16.317139 138272577578816 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')
I0422 07:53:16.317180 138272577578816 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)
I0422 07:53:16.317216 138272577578816 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)
I0422 07:53:16.317263 138272577578816 maxtext_utils.py:1668]  params/params/decoder/logits_dense/kernel
    Shape:     float32[2048,32000]
    Logical:   P('embed_vocab', 'vocab')
    Physical:  ('fsdp', None)
I0422 07:53:16.317311 138272577578816 maxtext_utils.py:1668]  params/params/token_embedder/embedding
    Shape:     float32[32000,2048]
    Logical:   P('vocab', 'embed_vocab')
    Physical:  (None, 'fsdp')

I0422 07:53:16.819324 138272577578816 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0422 07:53:16.819423 138272577578816 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0422 07:53:16.835054 138272577578816 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0422 07:53:16.835117 138272577578816 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0422 07:53:27.906997 138272577578816 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.
I0422 07:53:27.907801 138272577578816 metric_logger.py:301] number parameters: 1.104 billion
I0422 07:53:39.679344 138272577578816 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0422 07:53:39.832394 138272577578816 checkpointing.py:776] Waited 0.15303492546081543 seconds for step 0 to finish before starting checkpointing.
I0422 07:53:39.834942 138272577578816 checkpoint_manager.py:2009] [process=2][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0422 07:53:39.836760 138272577578816 checkpoint_manager.py:1512] [process=2] Saving checkpoint at step 0
I0422 07:53:39.838347 138272577578816 event_tracking.py:70] [process=2] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260422_071422/linen_xpk_main_20260422_071422_11_optimizer_offload_false/checkpoints/0.
I0422 07:53:40.568942 138272577578816 signaling_client.py:364] Using JaxDistributedSignalingClient
I0422 07:53:40.569882 138272577578816 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0422 07:53:40.569941 138272577578816 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 07:53:40.852207 138272577578816 base_pytree_checkpoint_handler.py:154] [process=2][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.283340s
I0422 07:53:40.852382 138272577578816 base_pytree_checkpoint_handler.py:130] [process=2] /jax/orbax/write/blocking_gbytes_per_sec: 5.336 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2890758514404297 s) (per-host)
I0422 07:53:40.852436 138272577578816 base_pytree_checkpoint_handler.py:768] [process=2][thread=MainThread] Initiated Pytree async_save. Time taken: 0.289140s (batch_requests_ready=0.002398s, total_serialization_initiated=0.286669s, others=0.000073s)
I0422 07:53:40.852538 138272577578816 composite_checkpoint_handler.py:715] [process=2][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.293185s (all_items=0.000019s, per_item={'items': '0.00001884'}, temp_paths=0.293166)
I0422 07:53:40.853371 138272577578816 event_tracking.py:125] [process=2] [async] Finished blocking save in 1.02 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260422_071422/linen_xpk_main_20260422_071422_11_optimizer_offload_false/checkpoints/0.
I0422 07:53:40.853699 138146324334336 async_checkpointer.py:76] [process=2][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-22 08:13:40.853661
I0422 07:53:40.855660 138272577578816 checkpoint_manager.py:1560] [process=2][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0422 07:53:40.855961 138145244899072 async_checkpointer.py:280] [process=2][thread=save_finalize] Waiting for background save thread=async_save.
I0422 07:53:40.856118 138272577578816 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_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776844419.8349245, 'wait_for_prev_duration_secs': 6.318092346191406e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776844419.836798, 'checkpointer_blocking_duration_secs': 1.0170648097991943, 'get_old_steps_start_time': 1776844420.853886, 'get_old_steps_duration_secs': 3.337860107421875e-05, 'checkpoint_manager_blocking_start_time': 1776844419.8328857, 'checkpoint_manager_blocking_duration_secs': 1.023195505142212}
I0422 07:53:40.856222 138272577578816 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0422 07:53:40.856277 138272577578816 max_utils.py:750] 
Memstats: After params initialized:
I0422 07:53:40.856324 138272577578816 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_8(process=2,(0,2,0,0))
I0422 07:53:40.856356 138272577578816 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_9(process=2,(1,2,0,0))
I0422 07:53:40.856383 138272577578816 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_12(process=2,(0,3,0,0))
I0422 07:53:40.856409 138272577578816 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_13(process=2,(1,3,0,0))
I0422 07:53:41.187255 138272577578816 metric_logger.py:196] completed step: 0, seconds: 11.771, TFLOP/s/device: 1.154, Tokens/s/device: 173.981, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0422 07:53:41.346943 138272577578816 metric_logger.py:196] completed step: 1, seconds: 1.506, TFLOP/s/device: 9.019, Tokens/s/device: 1359.495, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52779.875
I0422 07:53:41.775915 138272577578816 metric_logger.py:196] completed step: 2, seconds: 0.012, TFLOP/s/device: 1094.853, Tokens/s/device: 165028.203, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28637.129
I0422 07:53:41.923183 138272577578816 metric_logger.py:196] completed step: 3, seconds: 0.429, TFLOP/s/device: 31.650, Tokens/s/device: 4770.646, total_weights: 65536, loss: 9.734, lm_loss: 9.734, perplexity: 16889.203
I0422 07:53:42.217951 138272577578816 metric_logger.py:196] completed step: 4, seconds: 0.153, TFLOP/s/device: 88.866, Tokens/s/device: 13394.813, total_weights: 65536, loss: 9.277, lm_loss: 9.277, perplexity: 10694.614
I0422 07:53:42.223763 138272577578816 metric_logger.py:196] completed step: 5, seconds: 0.147, TFLOP/s/device: 92.304, Tokens/s/device: 13913.043, total_weights: 65536, loss: 8.892, lm_loss: 8.892, perplexity: 7270.717
I0422 07:53:44.108445    2730 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0422 07:53:46.169079 138146296325888 array_metadata_store.py:203] [process=2][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_false/checkpoints/0/items/array_metadatas/process_2
I0422 07:54:04.238450 138272577578816 metric_logger.py:196] completed step: 6, seconds: 0.295, TFLOP/s/device: 46.027, Tokens/s/device: 6937.740, total_weights: 65536, loss: 8.592, lm_loss: 8.592, perplexity: 5390.759
I0422 07:54:04.385786 138272577578816 metric_logger.py:196] completed step: 7, seconds: 21.868, TFLOP/s/device: 0.621, Tokens/s/device: 93.653, total_weights: 65536, loss: 8.384, lm_loss: 8.384, perplexity: 4376.905
I0422 07:54:04.533074 138272577578816 metric_logger.py:196] completed step: 8, seconds: 0.152, TFLOP/s/device: 89.167, Tokens/s/device: 13440.172, total_weights: 65536, loss: 8.255, lm_loss: 8.255, perplexity: 3846.406
I0422 07:54:04.679812 138272577578816 checkpointing.py:772] Waiting for step 9 to finish before checkpoint...
I0422 07:54:04.680451 138272577578816 checkpointing.py:776] Waited 0.0006597042083740234 seconds for step 9 to finish before starting checkpointing.
I0422 07:54:04.682404 138272577578816 checkpoint_manager.py:2020] [process=2][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0422 07:54:18.410327 138146324334336 base_pytree_checkpoint_handler.py:130] [process=2] /jax/orbax/write/gbytes_per_sec: 41.736 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 37.84698247909546 s) (per-host)
I0422 07:54:18.410449 138146324334336 async_checkpointer.py:90] [process=2][thread=async_save] 3 Handler Commit operations completed. Time taken: 37.556624s.
I0422 07:54:27.603103 138146324334336 async_checkpointer.py:160] [process=2][thread=async_save] Background save thread done. Time taken: 46.749262s.
I0422 07:54:27.603390 138145244899072 async_checkpointer.py:288] [process=2][thread=save_finalize] Done with waiting for background save thread=async_save.
I0422 07:54:27.603514 138145244899072 async_checkpointer.py:298] [process=2][thread=save_finalize] No errors found in background save thread=async_save.
I0422 07:54:27.603579 138145244899072 checkpoint_manager.py:2137] [process=2][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0422 07:54:27.605643 138145244899072 checkpoint_manager.py:2146] [process=2][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0422 07:54:27.605858 138272577578816 checkpoint_manager.py:2032] [process=2][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0422 07:54:27.606017 138272577578816 checkpoint_manager.py:1452] Waiting for previous save to complete took 22.923604 seconds. If this number is high, consider checkpointing less frequently.
I0422 07:54:27.607906 138272577578816 checkpoint_manager.py:1512] [process=2] Saving checkpoint at step 9
I0422 07:54:27.609938 138272577578816 event_tracking.py:70] [process=2] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_main_20260422_071422/linen_xpk_main_20260422_071422_11_optimizer_offload_false/checkpoints/9.
I0422 07:54:27.903514 138272577578816 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0422 07:54:27.903605 138272577578816 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 07:54:27.944075 138272577578816 base_pytree_checkpoint_handler.py:154] [process=2][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.041621s
I0422 07:54:27.944244 138272577578816 base_pytree_checkpoint_handler.py:130] [process=2] /jax/orbax/write/blocking_gbytes_per_sec: 34.085 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.045256614685058594 s) (per-host)
I0422 07:54:27.944301 138272577578816 base_pytree_checkpoint_handler.py:768] [process=2][thread=MainThread] Initiated Pytree async_save. Time taken: 0.045324s (batch_requests_ready=0.001954s, total_serialization_initiated=0.043296s, others=0.000073s)
I0422 07:54:27.944398 138272577578816 composite_checkpoint_handler.py:715] [process=2][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.049564s (all_items=0.000016s, per_item={'items': '0.00001574'}, temp_paths=0.049548)
I0422 07:54:27.945116 138272577578816 event_tracking.py:125] [process=2] [async] Finished blocking save in 0.34 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_main_20260422_071422/linen_xpk_main_20260422_071422_11_optimizer_offload_false/checkpoints/9.
I0422 07:54:27.945426 138145244899072 async_checkpointer.py:76] [process=2][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-22 08:14:27.945387
I0422 07:54:27.947298 138272577578816 checkpoint_manager.py:1560] [process=2][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0422 07:54:27.947521 138119303132928 async_checkpointer.py:280] [process=2][thread=save_finalize] Waiting for background save thread=async_save.
I0422 07:54:27.947626 138272577578816 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_false/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776844444.6823728, 'wait_for_prev_duration_secs': 22.923604011535645, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776844467.6079452, 'checkpointer_blocking_duration_secs': 0.33762431144714355, 'get_old_steps_start_time': 1776844467.9455914, 'get_old_steps_duration_secs': 3.337860107421875e-05, 'checkpoint_manager_blocking_start_time': 1776844444.6806843, 'checkpoint_manager_blocking_duration_secs': 23.266910076141357}
I0422 07:54:27.947803 138272577578816 checkpointing.py:408] Started an asynchronous checkpoint save for step 9
I0422 07:54:27.947848 138272577578816 checkpoint_manager.py:2020] [process=2][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0422 07:54:33.467936 138146296325888 array_metadata_store.py:203] [process=2][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_false/checkpoints/9/items/array_metadatas/process_2
I0422 07:55:10.931235 138145244899072 base_pytree_checkpoint_handler.py:130] [process=2] /jax/orbax/write/gbytes_per_sec: 36.707 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 43.03221559524536 s) (per-host)
I0422 07:55:10.931346 138145244899072 async_checkpointer.py:90] [process=2][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.985807s.
I0422 07:55:19.227466 138145244899072 async_checkpointer.py:160] [process=2][thread=async_save] Background save thread done. Time taken: 51.281910s.
I0422 07:55:19.227699 138119303132928 async_checkpointer.py:288] [process=2][thread=save_finalize] Done with waiting for background save thread=async_save.
I0422 07:55:19.227805 138119303132928 async_checkpointer.py:298] [process=2][thread=save_finalize] No errors found in background save thread=async_save.
I0422 07:55:19.227856 138119303132928 checkpoint_manager.py:2137] [process=2][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0422 07:55:19.229488 138119303132928 checkpoint_manager.py:2146] [process=2][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0422 07:55:19.229651 138272577578816 checkpoint_manager.py:2032] [process=2][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0422 07:55:19.229817 138272577578816 checkpoint_manager.py:2009] [process=2][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0422 07:55:19.230808 138272577578816 metric_logger.py:196] completed step: 9, seconds: 0.147, TFLOP/s/device: 92.282, Tokens/s/device: 13909.736, 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: Wed Apr 22 07:55:27 UTC 2026
EXIT_CODE=0