MaxView

XPK Start: Tue Apr 21 11:51:43 UTC 2026
PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.
2026-04-21 11:52:07.727415: 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 11:52:07.909141 139049871128384 max_utils.py:273] Attempting to initialize the jax distributed system...
I0421 11:52:16.951106 139049871128384 distributed.py:149] Starting JAX distributed service on [::]:8482
I0421 11:52:16.953478 139049871128384 distributed.py:172] Connecting to JAX distributed service on mt-04-int8-p53vw-slice-job-0-0.mt-04-int8-p53vw:8482
I0421 11:52:18.577575 139049871128384 max_utils.py:284] Jax distributed system initialized!
I0421 11:52:24.609575 139049871128384 max_utils.py:800] System Information: Jax Version: 0.9.2
I0421 11:52:24.609702 139049871128384 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0421 11:52:24.609746 139049871128384 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 11:52:24.609782 139049871128384 train_utils.py:378] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0421 11:52:25.312183 139049871128384 maxtext_utils.py:1718] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0421 11:52:25.312783 139049871128384 maxtext_utils.py:1718] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0421 11:52:25.313077 139049871128384 checkpointing.py:688] Setting up checkpoint logger...
I0421 11:52:25.313127 139049871128384 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0421 11:52:25.313169 139049871128384 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0421 11:52:25.313503 139049871128384 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 0x7e767131c5c0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0421 11:52:28.230890 139049871128384 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0421 11:52:28.231122 139049871128384 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 0x7e62483c19a0>}, handler_registry=None
I0421 11:52:28.231360 139049871128384 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e62483c19a0>` 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 11:52:28.231408 139049871128384 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e6248460b30>` 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 11:52:28.231444 139049871128384 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 0x7e62483c19a0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7e62483c19a0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e6248460b30>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7e6248460b30>}).
I0421 11:52:28.231785 139049871128384 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0421 11:52:28.231856 139049871128384 async_checkpointer.py:192] [process=4][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7e621858c540> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0421 11:52:30.264950 139049871128384 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260421_114057/linen_xpk_feat_nnx_post_train_fixes_20260421_114057_04_int8/checkpoints
I0421 11:52:30.699245 139049871128384 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_feat_nnx_post_train_fixes_20260421_114057/linen_xpk_feat_nnx_post_train_fixes_20260421_114057_04_int8/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7e613070b0e0>
I0421 11:52:30.699414 139049871128384 checkpointing.py:302] Checkpoint manager created!
I0421 11:52:31.846267 139049871128384 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0421 11:52:31.846372 139049871128384 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 11:52:32.233597 139049871128384 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0421 11:52:32.233696 139049871128384 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 11:52:32.250025 139049871128384 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0421 11:52:32.250094 139049871128384 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0421 11:52:32.315829 139049871128384 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 11:52:32.315911 139049871128384 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 11:52:32.332943 139049871128384 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 11:52:32.333003 139049871128384 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 11:52:32.349715 139049871128384 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 11:52:32.349773 139049871128384 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 11:52:32.375121 139049871128384 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0421 11:52:32.375194 139049871128384 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0421 11:52:32.441233 139049871128384 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0421 11:52:32.441318 139049871128384 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0421 11:52:32.959732 139049871128384 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0421 11:52:32.959837 139049871128384 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0421 11:52:35.424394 139049871128384 maxtext_utils.py:1827]  params/params/decoder/decoder_norm/scale
    Shape:     float32[2048]
    Logical:   P('norm',)
    Physical:  (None,)
I0421 11:52:35.424527 139049871128384 maxtext_utils.py:1827]  params/params/decoder/layers/mlp/wi_0/kernel
    Shape:     float32[2048,16,7168]
    Logical:   P('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0421 11:52:35.424582 139049871128384 maxtext_utils.py:1827]  params/params/decoder/layers/mlp/wi_1/kernel
    Shape:     float32[2048,16,7168]
    Logical:   P('embed', 'layers', 'mlp')
    Physical:  ('fsdp', None, None)
I0421 11:52:35.424661 139049871128384 maxtext_utils.py:1827]  params/params/decoder/layers/mlp/wo/kernel
    Shape:     float32[7168,16,2048]
    Logical:   P('mlp', 'layers', 'embed')
    Physical:  (None, None, 'fsdp')
I0421 11:52:35.424722 139049871128384 maxtext_utils.py:1827]  params/params/decoder/layers/post_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   P('norm', 'layers')
    Physical:  (None, None)
I0421 11:52:35.424762 139049871128384 maxtext_utils.py:1827]  params/params/decoder/layers/pre_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   P('norm', 'layers')
    Physical:  (None, None)
I0421 11:52:35.424815 139049871128384 maxtext_utils.py:1827]  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 11:52:35.424869 139049871128384 maxtext_utils.py:1827]  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 11:52:35.424908 139049871128384 maxtext_utils.py:1827]  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 11:52:35.424943 139049871128384 maxtext_utils.py:1827]  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 11:52:35.424990 139049871128384 maxtext_utils.py:1827]  params/params/decoder/logits_dense/kernel
    Shape:     float32[2048,32000]
    Logical:   P('embed_vocab', 'vocab')
    Physical:  ('fsdp', None)
I0421 11:52:35.425036 139049871128384 maxtext_utils.py:1827]  params/params/token_embedder/embedding
    Shape:     float32[32000,2048]
    Logical:   P('vocab', 'embed_vocab')
    Physical:  (None, 'fsdp')

I0421 11:52:36.825304 139049871128384 train.py:157] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0421 11:52:36.825397 139049871128384 train.py:157] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0421 11:52:36.840943 139049871128384 train.py:164] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0421 11:52:36.841006 139049871128384 train.py:164] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0421 11:52:50.249835 139049871128384 max_utils.py:791] Total memory size: 1.8 GB, Output size: 0.4 GB, Temp size: 1.4 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB.
I0421 11:52:50.250685 139049871128384 metric_logger.py:301] number parameters: 1.104 billion
I0421 11:53:05.601267 139049871128384 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0421 11:53:05.843667 139049871128384 checkpointing.py:798] Waited 0.2423548698425293 seconds for step 0 to finish before starting checkpointing.
I0421 11:53:05.847906 139049871128384 checkpoint_manager.py:2009] [process=4][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0421 11:53:05.849133 139049871128384 checkpoint_manager.py:1512] [process=4] Saving checkpoint at step 0
I0421 11:53:05.850747 139049871128384 event_tracking.py:70] [process=4] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260421_114057/linen_xpk_feat_nnx_post_train_fixes_20260421_114057_04_int8/checkpoints/0.
I0421 11:53:06.599261 139049871128384 signaling_client.py:364] Using JaxDistributedSignalingClient
I0421 11:53:06.600234 139049871128384 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 11:53:06.600295 139049871128384 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 11:53:06.888467 139049871128384 base_pytree_checkpoint_handler.py:154] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.289277s
I0421 11:53:06.888667 139049871128384 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/blocking_gbytes_per_sec: 5.230 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2949709892272949 s) (per-host)
I0421 11:53:06.888726 139049871128384 base_pytree_checkpoint_handler.py:768] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.295052s (batch_requests_ready=0.002180s, total_serialization_initiated=0.292780s, others=0.000092s)
I0421 11:53:06.888817 139049871128384 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.299344s (all_items=0.000016s, per_item={'items': '0.00001645'}, temp_paths=0.299328)
I0421 11:53:06.889728 139049871128384 event_tracking.py:125] [process=4] [async] Finished blocking save in 1.04 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260421_114057/linen_xpk_feat_nnx_post_train_fixes_20260421_114057_04_int8/checkpoints/0.
I0421 11:53:06.890066 138917866551040 async_checkpointer.py:76] [process=4][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 12:13:06.890028
I0421 11:53:06.892106 139049871128384 checkpoint_manager.py:1560] [process=4][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 11:53:06.892388 138917841372928 async_checkpointer.py:280] [process=4][thread=save_finalize] Waiting for background save thread=async_save.
I0421 11:53:06.892544 139049871128384 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260421_114057/linen_xpk_feat_nnx_post_train_fixes_20260421_114057_04_int8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776772385.847886, 'wait_for_prev_duration_secs': 6.389617919921875e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776772385.8491712, 'checkpointer_blocking_duration_secs': 1.04105806350708, 'get_old_steps_start_time': 1776772386.890256, 'get_old_steps_duration_secs': 2.9325485229492188e-05, 'checkpoint_manager_blocking_start_time': 1776772385.8441494, 'checkpoint_manager_blocking_duration_secs': 1.0483567714691162}
I0421 11:53:06.892661 139049871128384 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0421 11:53:06.892711 139049871128384 max_utils.py:750] 
Memstats: After params initialized:
I0421 11:53:06.892763 139049871128384 max_utils.py:756] 	Using (GB) 0.44 / 31.25 (1.408000%) on TPU_16(process=4,(0,4,0,0))
I0421 11:53:06.892796 139049871128384 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_17(process=4,(1,4,0,0))
I0421 11:53:06.892824 139049871128384 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_20(process=4,(0,5,0,0))
I0421 11:53:06.892848 139049871128384 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_21(process=4,(1,5,0,0))
I0421 11:53:07.221758 139049871128384 metric_logger.py:196] completed step: 0, seconds: 15.350, TFLOP/s/device: 0.885, Tokens/s/device: 133.416, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52796.277
I0421 11:53:07.453865 139049871128384 metric_logger.py:196] completed step: 1, seconds: 1.619, TFLOP/s/device: 8.393, Tokens/s/device: 1265.067, total_weights: 65536, loss: 10.874, lm_loss: 10.874, perplexity: 52796.277
I0421 11:53:07.860695 139049871128384 metric_logger.py:196] completed step: 2, seconds: 0.006, TFLOP/s/device: 2353.157, Tokens/s/device: 354693.453, total_weights: 65536, loss: 10.262, lm_loss: 10.262, perplexity: 28634.820
I0421 11:53:08.090284 139049871128384 metric_logger.py:196] completed step: 3, seconds: 0.405, TFLOP/s/device: 33.572, Tokens/s/device: 5060.314, total_weights: 65536, loss: 9.731, lm_loss: 9.731, perplexity: 16827.199
I0421 11:53:08.549528 139049871128384 metric_logger.py:196] completed step: 4, seconds: 0.235, TFLOP/s/device: 57.842, Tokens/s/device: 8718.529, total_weights: 65536, loss: 9.272, lm_loss: 9.272, perplexity: 10638.531
I0421 11:53:08.555716 139049871128384 metric_logger.py:196] completed step: 5, seconds: 0.229, TFLOP/s/device: 59.214, Tokens/s/device: 8925.420, total_weights: 65536, loss: 8.887, lm_loss: 8.887, perplexity: 7233.941
I0421 11:53:10.279115    2909 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0421 11:53:12.235564 138917849765632 array_metadata_store.py:203] [process=4][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260421_114057/linen_xpk_feat_nnx_post_train_fixes_20260421_114057_04_int8/checkpoints/0/items/array_metadatas/process_4
I0421 11:53:38.801610 139049871128384 metric_logger.py:196] completed step: 6, seconds: 0.460, TFLOP/s/device: 29.560, Tokens/s/device: 4455.564, total_weights: 65536, loss: 8.588, lm_loss: 8.588, perplexity: 5365.468
I0421 11:53:39.030885 139049871128384 metric_logger.py:196] completed step: 7, seconds: 30.018, TFLOP/s/device: 0.453, Tokens/s/device: 68.227, total_weights: 65536, loss: 8.380, lm_loss: 8.380, perplexity: 4358.779
I0421 11:53:39.260368 139049871128384 metric_logger.py:196] completed step: 8, seconds: 0.234, TFLOP/s/device: 58.039, Tokens/s/device: 8748.249, total_weights: 65536, loss: 8.251, lm_loss: 8.251, perplexity: 3831.453
I0421 11:53:39.489388 139049871128384 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0421 11:53:39.490110 139049871128384 checkpointing.py:798] Waited 0.0007445812225341797 seconds for step 9 to finish before starting checkpointing.
I0421 11:53:39.492111 139049871128384 checkpoint_manager.py:2020] [process=4][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 11:53:40.608423 138917866551040 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/gbytes_per_sec: 46.439 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 34.01470184326172 s) (per-host)
I0421 11:53:40.608556 138917866551040 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 33.718375s.
I0421 11:53:48.660501 138917866551040 async_checkpointer.py:160] [process=4][thread=async_save] Background save thread done. Time taken: 41.770291s.
I0421 11:53:48.660804 138917841372928 async_checkpointer.py:288] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 11:53:48.660928 138917841372928 async_checkpointer.py:298] [process=4][thread=save_finalize] No errors found in background save thread=async_save.
I0421 11:53:48.660987 138917841372928 checkpoint_manager.py:2137] [process=4][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0421 11:53:48.662390 138917841372928 checkpoint_manager.py:2146] [process=4][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0421 11:53:48.662583 139049871128384 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 11:53:48.662737 139049871128384 checkpoint_manager.py:1452] Waiting for previous save to complete took 9.170623 seconds. If this number is high, consider checkpointing less frequently.
I0421 11:53:48.664513 139049871128384 checkpoint_manager.py:1512] [process=4] Saving checkpoint at step 9
I0421 11:53:48.666580 139049871128384 event_tracking.py:70] [process=4] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260421_114057/linen_xpk_feat_nnx_post_train_fixes_20260421_114057_04_int8/checkpoints/9.
I0421 11:53:49.368112 139049871128384 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0421 11:53:49.368212 139049871128384 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 11:53:49.402560 139049871128384 base_pytree_checkpoint_handler.py:154] [process=4][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.035580s
I0421 11:53:49.402721 139049871128384 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/blocking_gbytes_per_sec: 39.545 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.03900790214538574 s) (per-host)
I0421 11:53:49.402772 139049871128384 base_pytree_checkpoint_handler.py:768] [process=4][thread=MainThread] Initiated Pytree async_save. Time taken: 0.039069s (batch_requests_ready=0.001763s, total_serialization_initiated=0.037237s, others=0.000068s)
I0421 11:53:49.402855 139049871128384 composite_checkpoint_handler.py:715] [process=4][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.043237s (all_items=0.000015s, per_item={'items': '0.00001478'}, temp_paths=0.043223)
I0421 11:53:49.403563 139049871128384 event_tracking.py:125] [process=4] [async] Finished blocking save in 0.74 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260421_114057/linen_xpk_feat_nnx_post_train_fixes_20260421_114057_04_int8/checkpoints/9.
I0421 11:53:49.403882 138917841372928 async_checkpointer.py:76] [process=4][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-21 12:13:49.403850
I0421 11:53:49.405861 139049871128384 checkpoint_manager.py:1560] [process=4][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0421 11:53:49.406176 138917849765632 async_checkpointer.py:280] [process=4][thread=save_finalize] Waiting for background save thread=async_save.
I0421 11:53:49.406366 139049871128384 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260421_114057/linen_xpk_feat_nnx_post_train_fixes_20260421_114057_04_int8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776772419.4920812, 'wait_for_prev_duration_secs': 9.170623302459717, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776772428.6645548, 'checkpointer_blocking_duration_secs': 0.7394697666168213, 'get_old_steps_start_time': 1776772429.404049, 'get_old_steps_duration_secs': 3.123283386230469e-05, 'checkpoint_manager_blocking_start_time': 1776772419.490357, 'checkpoint_manager_blocking_duration_secs': 9.915974378585815}
I0421 11:53:49.406476 139049871128384 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0421 11:53:49.406518 139049871128384 checkpoint_manager.py:2020] [process=4][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0421 11:53:55.441809 138894024595200 array_metadata_store.py:203] [process=4][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260421_114057/linen_xpk_feat_nnx_post_train_fixes_20260421_114057_04_int8/checkpoints/9/items/array_metadatas/process_4
I0421 11:54:32.041122 138917841372928 base_pytree_checkpoint_handler.py:130] [process=4] /jax/orbax/write/gbytes_per_sec: 37.013 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.67737102508545 s) (per-host)
I0421 11:54:32.041251 138917841372928 async_checkpointer.py:90] [process=4][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.637256s.
I0421 11:54:40.002081 138917841372928 async_checkpointer.py:160] [process=4][thread=async_save] Background save thread done. Time taken: 50.598070s.
I0421 11:54:40.002372 138917849765632 async_checkpointer.py:288] [process=4][thread=save_finalize] Done with waiting for background save thread=async_save.
I0421 11:54:40.002506 138917849765632 async_checkpointer.py:298] [process=4][thread=save_finalize] No errors found in background save thread=async_save.
I0421 11:54:40.002557 138917849765632 checkpoint_manager.py:2137] [process=4][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0421 11:54:40.004396 138917849765632 checkpoint_manager.py:2146] [process=4][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0421 11:54:40.004557 139049871128384 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 11:54:40.004714 139049871128384 checkpoint_manager.py:2009] [process=4][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0421 11:54:40.005854 139049871128384 metric_logger.py:196] completed step: 9, seconds: 0.229, TFLOP/s/device: 59.223, Tokens/s/device: 8926.664, total_weights: 65536, loss: 8.175, lm_loss: 8.175, perplexity: 3552.333
Per train step:
 Total TFLOPs: 13.59 
 split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Tue Apr 21 11:54:53 UTC 2026
EXIT_CODE=0