XPK Start: Fri Apr 17 02:50:06 UTC 2026
PyTorch was not found. Models won't be available and only tokenizers, configuration and file/data utilities can be used.
2026-04-17 02:50:30.424239: 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)
I0417 02:50:30.957382 135942929073984 max_utils.py:273] Attempting to initialize the jax distributed system...
I0417 02:50:39.996928 135942929073984 distributed.py:149] Starting JAX distributed service on [::]:8482
I0417 02:50:39.999186 135942929073984 distributed.py:172] Connecting to JAX distributed service on mt-04-int8-r84b3-slice-job-0-0.mt-04-int8-r84b3:8482
I0417 02:50:41.239719 135942929073984 max_utils.py:284] Jax distributed system initialized!
I0417 02:50:47.272218 135942929073984 max_utils.py:800] System Information: Jax Version: 0.9.2
I0417 02:50:47.272317 135942929073984 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0417 02:50:47.272357 135942929073984 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
I0417 02:50:47.272392 135942929073984 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0417 02:50:47.950896 135942929073984 maxtext_utils.py:1687] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0417 02:50:47.951478 135942929073984 maxtext_utils.py:1687] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0417 02:50:47.951787 135942929073984 checkpointing.py:688] Setting up checkpoint logger...
I0417 02:50:47.951855 135942929073984 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0417 02:50:47.951900 135942929073984 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0417 02:50:47.952230 135942929073984 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 0x7ba3101ef6b0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0417 02:50:50.929097 135942929073984 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0417 02:50:50.929434 135942929073984 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 0x7b8eac544d70>}, handler_registry=None
I0417 02:50:50.929718 135942929073984 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b8eac544d70>` 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`.
I0417 02:50:50.929772 135942929073984 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b8eac548e30>` 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`.
I0417 02:50:50.929809 135942929073984 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 0x7b8eac544d70>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b8eac544d70>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b8eac548e30>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b8eac548e30>}).
I0417 02:50:50.930147 135942929073984 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0417 02:50:50.930218 135942929073984 async_checkpointer.py:192] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7b8eac2e8680> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0417 02:50:52.530929 135942929073984 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_04_int8/checkpoints
I0417 02:50:52.995921 135942929073984 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_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_04_int8/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7b8eac5459a0>
I0417 02:50:52.996090 135942929073984 checkpointing.py:302] Checkpoint manager created!
I0417 02:50:54.137891 135942929073984 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0417 02:50:54.137994 135942929073984 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0417 02:50:54.520264 135942929073984 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0417 02:50:54.520354 135942929073984 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0417 02:50:54.536641 135942929073984 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0417 02:50:54.536698 135942929073984 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0417 02:50:54.608378 135942929073984 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0417 02:50:54.608484 135942929073984 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0417 02:50:54.625034 135942929073984 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0417 02:50:54.625094 135942929073984 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0417 02:50:54.641339 135942929073984 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0417 02:50:54.641397 135942929073984 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0417 02:50:54.665962 135942929073984 attentions.py:1197] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0417 02:50:54.666029 135942929073984 attentions.py:1197] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0417 02:50:54.731933 135942929073984 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0417 02:50:54.732016 135942929073984 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0417 02:50:55.243602 135942929073984 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0417 02:50:55.243715 135942929073984 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0417 02:50:57.272744 135942929073984 maxtext_utils.py:1796] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: P('norm',)
Physical: (None,)
I0417 02:50:57.272888 135942929073984 maxtext_utils.py:1796] params/params/decoder/layers/mlp/wi_0/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0417 02:50:57.272944 135942929073984 maxtext_utils.py:1796] params/params/decoder/layers/mlp/wi_1/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0417 02:50:57.273004 135942929073984 maxtext_utils.py:1796] params/params/decoder/layers/mlp/wo/kernel
Shape: float32[7168,16,2048]
Logical: P('mlp', 'layers', 'embed')
Physical: (None, None, 'fsdp')
I0417 02:50:57.273056 135942929073984 maxtext_utils.py:1796] params/params/decoder/layers/post_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0417 02:50:57.273094 135942929073984 maxtext_utils.py:1796] params/params/decoder/layers/pre_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0417 02:50:57.273144 135942929073984 maxtext_utils.py:1796] 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)
I0417 02:50:57.273198 135942929073984 maxtext_utils.py:1796] 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')
I0417 02:50:57.273238 135942929073984 maxtext_utils.py:1796] 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)
I0417 02:50:57.273273 135942929073984 maxtext_utils.py:1796] 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)
I0417 02:50:57.273319 135942929073984 maxtext_utils.py:1796] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: P('embed_vocab', 'vocab')
Physical: ('fsdp', None)
I0417 02:50:57.273363 135942929073984 maxtext_utils.py:1796] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: P('vocab', 'embed_vocab')
Physical: (None, 'fsdp')
I0417 02:50:59.105373 135942929073984 train.py:157] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0417 02:50:59.105470 135942929073984 train.py:157] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0417 02:50:59.121172 135942929073984 train.py:164] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0417 02:50:59.121232 135942929073984 train.py:164] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0417 02:51:12.677149 135942929073984 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.
I0417 02:51:12.678050 135942929073984 metric_logger.py:289] number parameters: 1.104 billion
I0417 02:51:27.908560 135942929073984 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0417 02:51:28.151398 135942929073984 checkpointing.py:798] Waited 0.24281001091003418 seconds for step 0 to finish before starting checkpointing.
I0417 02:51:28.153887 135942929073984 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0417 02:51:28.155972 135942929073984 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 0
I0417 02:51:28.157533 135942929073984 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_04_int8/checkpoints/0.
I0417 02:51:28.910336 135942929073984 signaling_client.py:364] Using JaxDistributedSignalingClient
I0417 02:51:28.911273 135942929073984 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0417 02:51:28.911326 135942929073984 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
I0417 02:51:29.197266 135942929073984 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.287006s
I0417 02:51:29.197435 135942929073984 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 5.270 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.292691707611084 s) (per-host)
I0417 02:51:29.197487 135942929073984 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.292753s (batch_requests_ready=0.002381s, total_serialization_initiated=0.290303s, others=0.000069s)
I0417 02:51:29.197578 135942929073984 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.296939s (all_items=0.000019s, per_item={'items': '0.00001884'}, temp_paths=0.296921)
I0417 02:51:29.198350 135942929073984 event_tracking.py:125] [process=6] [async] Finished blocking save in 1.04 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_04_int8/checkpoints/0.
I0417 02:51:29.198656 135811266074368 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-17 03:11:29.198618
I0417 02:51:29.200686 135942929073984 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0417 02:51:29.200929 135811236407040 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0417 02:51:29.201050 135942929073984 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_04_int8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776394288.15387, 'wait_for_prev_duration_secs': 5.817413330078125e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776394288.156009, 'checkpointer_blocking_duration_secs': 1.0428013801574707, 'get_old_steps_start_time': 1776394289.1988435, 'get_old_steps_duration_secs': 3.0517578125e-05, 'checkpoint_manager_blocking_start_time': 1776394288.1519516, 'checkpoint_manager_blocking_duration_secs': 1.049062728881836}
I0417 02:51:29.201151 135942929073984 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0417 02:51:29.201199 135942929073984 max_utils.py:750]
Memstats: After params initialized:
I0417 02:51:29.201244 135942929073984 max_utils.py:756] Using (GB) 0.44 / 31.25 (1.408000%) on TPU_24(process=6,(0,6,0,0))
I0417 02:51:29.201274 135942929073984 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_25(process=6,(1,6,0,0))
I0417 02:51:29.201299 135942929073984 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_28(process=6,(0,7,0,0))
I0417 02:51:29.201323 135942929073984 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_29(process=6,(1,7,0,0))
I0417 02:51:29.533047 135942929073984 metric_logger.py:185] completed step: 0, seconds: 15.230, TFLOP/s/device: 0.892, Tokens/s/device: 134.468, total_weights: 65536, loss: 10.874
I0417 02:51:29.772216 135942929073984 metric_logger.py:185] completed step: 1, seconds: 1.623, TFLOP/s/device: 8.371, Tokens/s/device: 1261.751, total_weights: 65536, loss: 10.874
I0417 02:51:30.181439 135942929073984 metric_logger.py:185] completed step: 2, seconds: 0.010, TFLOP/s/device: 1315.690, Tokens/s/device: 198315.096, total_weights: 65536, loss: 10.262
I0417 02:51:30.411034 135942929073984 metric_logger.py:185] completed step: 3, seconds: 0.410, TFLOP/s/device: 33.141, Tokens/s/device: 4995.305, total_weights: 65536, loss: 9.731
I0417 02:51:30.871564 135942929073984 metric_logger.py:185] completed step: 4, seconds: 0.234, TFLOP/s/device: 58.050, Tokens/s/device: 8749.931, total_weights: 65536, loss: 9.272
I0417 02:51:30.877177 135942929073984 metric_logger.py:185] completed step: 5, seconds: 0.229, TFLOP/s/device: 59.212, Tokens/s/device: 8925.147, total_weights: 65536, loss: 8.887
I0417 02:51:33.781581 2925 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0417 02:51:37.170301 135811244799744 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_04_int8/checkpoints/0/items/array_metadatas/process_6
I0417 02:52:01.048890 135942929073984 metric_logger.py:185] completed step: 6, seconds: 0.461, TFLOP/s/device: 29.484, Tokens/s/device: 4444.203, total_weights: 65536, loss: 8.588
I0417 02:52:01.278325 135942929073984 metric_logger.py:185] completed step: 7, seconds: 29.944, TFLOP/s/device: 0.454, Tokens/s/device: 68.394, total_weights: 65536, loss: 8.380
I0417 02:52:01.508954 135942929073984 metric_logger.py:185] completed step: 8, seconds: 0.233, TFLOP/s/device: 58.277, Tokens/s/device: 8784.195, total_weights: 65536, loss: 8.251
I0417 02:52:01.738216 135942929073984 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0417 02:52:01.738815 135942929073984 checkpointing.py:798] Waited 0.0006198883056640625 seconds for step 9 to finish before starting checkpointing.
I0417 02:52:01.740868 135942929073984 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0417 02:52:03.384339 135811266074368 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 45.813 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 34.479557037353516 s) (per-host)
I0417 02:52:03.384464 135811266074368 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 34.185696s.
I0417 02:52:13.572070 135811266074368 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 44.373286s.
I0417 02:52:13.572358 135811236407040 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0417 02:52:13.572476 135811236407040 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0417 02:52:13.572529 135811236407040 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0417 02:52:13.575205 135811236407040 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0417 02:52:13.575381 135942929073984 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.
W0417 02:52:13.575512 135942929073984 checkpoint_manager.py:1452] Waiting for previous save to complete took 11.834644 seconds. If this number is high, consider checkpointing less frequently.
I0417 02:52:13.577309 135942929073984 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 9
I0417 02:52:13.579258 135942929073984 event_tracking.py:70] [process=6] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_04_int8/checkpoints/9.
I0417 02:52:14.275093 135942929073984 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0417 02:52:14.275189 135942929073984 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
I0417 02:52:14.311043 135942929073984 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.037177s
I0417 02:52:14.311201 135942929073984 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 37.832 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.04077410697937012 s) (per-host)
I0417 02:52:14.311251 135942929073984 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.040833s (batch_requests_ready=0.001941s, total_serialization_initiated=0.038825s, others=0.000067s)
I0417 02:52:14.311340 135942929073984 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.044949s (all_items=0.000016s, per_item={'items': '0.00001645'}, temp_paths=0.044933)
I0417 02:52:14.312112 135942929073984 event_tracking.py:125] [process=6] [async] Finished blocking save in 0.73 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_04_int8/checkpoints/9.
I0417 02:52:14.312421 135811236407040 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-17 03:12:14.312384
I0417 02:52:14.318979 135942929073984 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0417 02:52:14.319242 135811244799744 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0417 02:52:14.319383 135942929073984 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_04_int8/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776394321.740839, 'wait_for_prev_duration_secs': 11.834643840789795, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776394333.5773478, 'checkpointer_blocking_duration_secs': 0.7352213859558105, 'get_old_steps_start_time': 1776394334.3125935, 'get_old_steps_duration_secs': 3.266334533691406e-05, 'checkpoint_manager_blocking_start_time': 1776394321.739053, 'checkpoint_manager_blocking_duration_secs': 12.580293416976929}
I0417 02:52:14.319505 135942929073984 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0417 02:52:14.319549 135942929073984 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0417 02:52:19.787811 135787014174464 array_metadata_store.py:203] [process=6][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260417_023710/linen_xpk_feat_nnx_post_train_fixes_20260417_023710_04_int8/checkpoints/9/items/array_metadatas/process_6
I0417 02:52:55.460438 135811236407040 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 38.349 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.18997025489807 s) (per-host)
I0417 02:52:55.460570 135811236407040 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.148031s.
I0417 02:53:06.942631 135811236407040 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 52.630076s.
I0417 02:53:06.942865 135811244799744 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0417 02:53:06.942921 135811244799744 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0417 02:53:06.942973 135811244799744 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0417 02:53:06.944621 135811244799744 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0417 02:53:06.944802 135942929073984 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.
I0417 02:53:06.944973 135942929073984 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0417 02:53:06.945649 135942929073984 metric_logger.py:185] completed step: 9, seconds: 0.229, TFLOP/s/device: 59.213, Tokens/s/device: 8925.264, total_weights: 65536, loss: 8.175
Per train step:
Total TFLOPs: 13.59
split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Fri Apr 17 02:53:17 UTC 2026
EXIT_CODE=0