MaxView

XPK Start: Fri Apr 17 03:18:08 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 03:18:32.098774: 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 03:18:32.279453 135493274621760 max_utils.py:273] Attempting to initialize the jax distributed system...
I0417 03:18:41.321873 135493274621760 distributed.py:149] Starting JAX distributed service on [::]:8482
I0417 03:18:41.324425 135493274621760 distributed.py:172] Connecting to JAX distributed service on mt-09-pdb-lt-1-tdsdi-slice-job-0-0.mt-09-pdb-lt-1-tdsdi:8482
I0417 03:18:42.431430 135493274621760 max_utils.py:284] Jax distributed system initialized!
I0417 03:18:48.602266 135493274621760 max_utils.py:800] System Information: Jax Version: 0.9.2
I0417 03:18:48.602367 135493274621760 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0417 03:18:48.602409 135493274621760 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 03:18:48.602443 135493274621760 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0417 03:18:49.285691 135493274621760 maxtext_utils.py:1687] Num_devices: 32, shape (1, 1, 1, 8, 1, 1, 1, 1, 4, 1, 1, 1, 1)
I0417 03:18:49.286305 135493274621760 maxtext_utils.py:1687] Num_devices: 32, shape (1, 1, 1, 8, 1, 1, 1, 1, 4, 1, 1, 1, 1)
I0417 03:18:49.286482 135493274621760 checkpointing.py:688] Setting up checkpoint logger...
I0417 03:18:49.286532 135493274621760 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0417 03:18:49.286575 135493274621760 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0417 03:18:49.286931 135493274621760 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 0x7b3a5b91b650>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0417 03:18:52.217269 135493274621760 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0417 03:18:52.217503 135493274621760 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 0x7b264011c1a0>}, handler_registry=None
I0417 03:18:52.217742 135493274621760 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b264011c1a0>` 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 03:18:52.217803 135493274621760 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b264011f4a0>` 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 03:18:52.217851 135493274621760 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 0x7b264011c1a0>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7b264011c1a0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b264011f4a0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7b264011f4a0>}).
I0417 03:18:52.218167 135493274621760 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0417 03:18:52.218237 135493274621760 async_checkpointer.py:192] [process=6][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7b26102c07c0> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0417 03:18:53.765765 135493274621760 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_09_pdb_lt_1/checkpoints
I0417 03:18:54.196897 135493274621760 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_09_pdb_lt_1/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7b264011e3f0>
I0417 03:18:54.197066 135493274621760 checkpointing.py:302] Checkpoint manager created!
I0417 03:18:55.148059 135493274621760 nnx_wrappers.py:437] Unknown Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0417 03:18:55.148167 135493274621760 nnx_wrappers.py:437] Unknown Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor').
I0417 03:18:55.525008 135493274621760 attentions.py:1088] attentions/inputs_q Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0417 03:18:55.525099 135493274621760 attentions.py:1088] attentions/inputs_q Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor').
I0417 03:18:55.541372 135493274621760 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[8,2048,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0417 03:18:55.541430 135493274621760 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[8,2048,2048]....................................... ('fsdp', None, 'tensor').
I0417 03:18:55.565026 135493274621760 attentions.py:1154] attentions/query Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0417 03:18:55.565092 135493274621760 attentions.py:1154] attentions/query Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None).
I0417 03:18:55.581407 135493274621760 attentions.py:1155] attentions/key Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0417 03:18:55.581467 135493274621760 attentions.py:1155] attentions/key Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None).
I0417 03:18:55.597705 135493274621760 attentions.py:1156] attentions/value Logical: bfloat16[8,2048,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0417 03:18:55.597766 135493274621760 attentions.py:1156] attentions/value Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None).
I0417 03:18:55.622306 135493274621760 attentions.py:1197] attentions/out Logical: bfloat16[8,2048,16,128]..................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0417 03:18:55.622375 135493274621760 attentions.py:1197] attentions/out Physical: bfloat16[8,2048,16,128]..................................... ('fsdp', None, 'tensor', None).
I0417 03:18:55.643938 135493274621760 linears.py:525] linears/x Logical: bfloat16[8,2048,7168]....................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0417 03:18:55.644006 135493274621760 linears.py:525] linears/x Physical: bfloat16[8,2048,7168]....................................... ('fsdp', None, 'tensor').
I0417 03:18:55.859155 135493274621760 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0417 03:18:55.859264 135493274621760 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 8
tensor: 4
I0417 03:18:57.282778 135493274621760 maxtext_utils.py:1796]  params/params/decoder/decoder_norm/scale
    Shape:     float32[2048]
    Logical:   P('norm',)
    Physical:  ('tensor',)
I0417 03:18:57.282940 135493274621760 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, 'tensor')
I0417 03:18:57.282998 135493274621760 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, 'tensor')
I0417 03:18:57.283057 135493274621760 maxtext_utils.py:1796]  params/params/decoder/layers/mlp/wo/kernel
    Shape:     float32[7168,16,2048]
    Logical:   P('mlp', 'layers', 'embed')
    Physical:  ('tensor', None, 'fsdp')
I0417 03:18:57.283109 135493274621760 maxtext_utils.py:1796]  params/params/decoder/layers/post_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   P('norm', 'layers')
    Physical:  ('tensor', None)
I0417 03:18:57.283148 135493274621760 maxtext_utils.py:1796]  params/params/decoder/layers/pre_self_attention_layer_norm/scale
    Shape:     float32[2048,16]
    Logical:   P('norm', 'layers')
    Physical:  ('tensor', None)
I0417 03:18:57.283198 135493274621760 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, 'tensor', None)
I0417 03:18:57.283249 135493274621760 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:  ('tensor', None, None, 'fsdp')
I0417 03:18:57.283290 135493274621760 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, 'tensor', None)
I0417 03:18:57.283326 135493274621760 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, 'tensor', None)
I0417 03:18:57.283373 135493274621760 maxtext_utils.py:1796]  params/params/decoder/logits_dense/kernel
    Shape:     float32[2048,32000]
    Logical:   P('embed_vocab', 'vocab')
    Physical:  ('fsdp', 'tensor')
I0417 03:18:57.283420 135493274621760 maxtext_utils.py:1796]  params/params/token_embedder/embedding
    Shape:     float32[32000,2048]
    Logical:   P('vocab', 'embed_vocab')

    Physical:  ('tensor', 'fsdp')
I0417 03:18:57.776648 135493274621760 train.py:157] train/xent Logical: float32[8,2048]............................................. ('activation_embed_and_logits_batch', 'activation_length').
I0417 03:18:57.776745 135493274621760 train.py:157] train/xent Physical: float32[8,2048]............................................. ('fsdp', None).
I0417 03:18:57.792433 135493274621760 train.py:164] train/z_loss Logical: float32[8,2048]............................................. ('activation_embed_and_logits_batch', 'activation_length').
I0417 03:18:57.792494 135493274621760 train.py:164] train/z_loss Physical: float32[8,2048]............................................. ('fsdp', None).
I0417 03:19:12.940458 135493274621760 max_utils.py:791] Total memory size: 0.9 GB, Output size: 0.4 GB, Temp size: 0.5 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB.
I0417 03:19:12.941378 135493274621760 metric_logger.py:289] number parameters: 1.104 billion
I0417 03:19:14.121268 135493274621760 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0417 03:19:28.510910 135493274621760 checkpointing.py:798] Waited 14.389620304107666 seconds for step 0 to finish before starting checkpointing.
I0417 03:19:28.513271 135493274621760 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0417 03:19:28.520959 135493274621760 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 0
I0417 03:19:28.522721 135493274621760 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_09_pdb_lt_1/checkpoints/0.
I0417 03:19:29.274852 135493274621760 signaling_client.py:364] Using JaxDistributedSignalingClient
I0417 03:19:29.275815 135493274621760 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0417 03:19:29.275890 135493274621760 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 03:19:29.541250 135493274621760 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.266494s
I0417 03:19:29.541419 135493274621760 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 5.675 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2718195915222168 s) (per-host)
I0417 03:19:29.541472 135493274621760 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.271884s (batch_requests_ready=0.002167s, total_serialization_initiated=0.269644s, others=0.000072s)
I0417 03:19:29.541564 135493274621760 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.276012s (all_items=0.000017s, per_item={'items': '0.00001669'}, temp_paths=0.275995)
I0417 03:19:29.542410 135493274621760 event_tracking.py:125] [process=6] [async] Finished blocking save in 1.02 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_09_pdb_lt_1/checkpoints/0.
I0417 03:19:29.542761 135365423789824 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-17 03:39:29.542721
I0417 03:19:29.553392 135493274621760 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0417 03:19:29.553659 135365398611712 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0417 03:19:29.553786 135493274621760 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_09_pdb_lt_1/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776395968.5132532, 'wait_for_prev_duration_secs': 6.365776062011719e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776395968.5210018, 'checkpointer_blocking_duration_secs': 1.0218825340270996, 'get_old_steps_start_time': 1776395969.542911, 'get_old_steps_duration_secs': 3.0994415283203125e-05, 'checkpoint_manager_blocking_start_time': 1776395968.5114522, 'checkpoint_manager_blocking_duration_secs': 1.0422956943511963}
I0417 03:19:29.553910 135493274621760 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0417 03:19:29.553961 135493274621760 max_utils.py:750] 
Memstats: After params initialized:
I0417 03:19:29.554011 135493274621760 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_24(process=6,(0,6,0,0))
I0417 03:19:29.554045 135493274621760 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_25(process=6,(1,6,0,0))
I0417 03:19:29.554073 135493274621760 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_28(process=6,(0,7,0,0))
I0417 03:19:29.554098 135493274621760 max_utils.py:756] 	Using (GB) 0.43 / 31.25 (1.376000%) on TPU_29(process=6,(1,7,0,0))
I0417 03:19:29.866845 135493274621760 metric_logger.py:185] completed step: 0, seconds: 1.180, TFLOP/s/device: 2.879, Tokens/s/device: 433.984, total_weights: 16384, loss: 10.862
I0417 03:19:29.950943 135493274621760 metric_logger.py:185] completed step: 1, seconds: 15.744, TFLOP/s/device: 0.216, Tokens/s/device: 32.520, total_weights: 16384, loss: 10.862
I0417 03:19:30.379147 135493274621760 metric_logger.py:185] completed step: 2, seconds: 0.026, TFLOP/s/device: 131.348, Tokens/s/device: 19798.152, total_weights: 16384, loss: 9.763
I0417 03:19:30.437647 135493274621760 metric_logger.py:185] completed step: 3, seconds: 0.429, TFLOP/s/device: 7.924, Tokens/s/device: 1194.367, total_weights: 16384, loss: 8.823
I0417 03:19:30.554903 135493274621760 metric_logger.py:185] completed step: 4, seconds: 0.063, TFLOP/s/device: 53.548, Tokens/s/device: 8071.381, total_weights: 16384, loss: 7.947
I0417 03:19:30.560578 135493274621760 metric_logger.py:185] completed step: 5, seconds: 0.058, TFLOP/s/device: 58.105, Tokens/s/device: 8758.275, total_weights: 16384, loss: 7.212
I0417 03:19:33.291452    2518 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0417 03:19:35.799785 135365407004416 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_09_pdb_lt_1/checkpoints/0/items/array_metadatas/process_6
I0417 03:19:45.545507 135493274621760 metric_logger.py:185] completed step: 6, seconds: 0.118, TFLOP/s/device: 28.845, Tokens/s/device: 4347.789, total_weights: 16384, loss: 6.652
I0417 03:19:45.604089 135493274621760 metric_logger.py:185] completed step: 7, seconds: 14.927, TFLOP/s/device: 0.228, Tokens/s/device: 34.299, total_weights: 16384, loss: 6.269
I0417 03:19:45.662440 135493274621760 metric_logger.py:185] completed step: 8, seconds: 0.063, TFLOP/s/device: 53.903, Tokens/s/device: 8124.921, total_weights: 16384, loss: 6.031
I0417 03:19:45.720345 135493274621760 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0417 03:19:45.720973 135493274621760 checkpointing.py:798] Waited 0.0006432533264160156 seconds for step 9 to finish before starting checkpointing.
I0417 03:19:45.723050 135493274621760 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0417 03:20:06.040580 135365423789824 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 42.958 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 36.77094006538391 s) (per-host)
I0417 03:20:06.040714 135365423789824 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 36.497302s.
I0417 03:20:14.267591 135365423789824 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 44.724162s.
I0417 03:20:14.267881 135365398611712 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0417 03:20:14.268002 135365398611712 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0417 03:20:14.268055 135365398611712 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0417 03:20:14.270468 135365398611712 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0417 03:20:14.270645 135493274621760 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 03:20:14.270775 135493274621760 checkpoint_manager.py:1452] Waiting for previous save to complete took 28.547726 seconds. If this number is high, consider checkpointing less frequently.
I0417 03:20:14.272548 135493274621760 checkpoint_manager.py:1512] [process=6] Saving checkpoint at step 9
I0417 03:20:14.274566 135493274621760 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_09_pdb_lt_1/checkpoints/9.
I0417 03:20:15.403867 135493274621760 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0417 03:20:15.403962 135493274621760 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 03:20:15.430147 135493274621760 base_pytree_checkpoint_handler.py:154] [process=6][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.027506s
I0417 03:20:15.430276 135493274621760 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/blocking_gbytes_per_sec: 50.028 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.030833959579467773 s) (per-host)
I0417 03:20:15.430325 135493274621760 base_pytree_checkpoint_handler.py:768] [process=6][thread=MainThread] Initiated Pytree async_save. Time taken: 0.030892s (batch_requests_ready=0.001737s, total_serialization_initiated=0.029091s, others=0.000063s)
I0417 03:20:15.430403 135493274621760 composite_checkpoint_handler.py:715] [process=6][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.034997s (all_items=0.000015s, per_item={'items': '0.00001478'}, temp_paths=0.034982)
I0417 03:20:15.431066 135493274621760 event_tracking.py:125] [process=6] [async] Finished blocking save in 1.16 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_09_pdb_lt_1/checkpoints/9.
I0417 03:20:15.431335 135364350048000 async_checkpointer.py:76] [process=6][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-17 03:40:15.431309
I0417 03:20:15.433963 135493274621760 checkpoint_manager.py:1560] [process=6][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0417 03:20:15.434192 135365398611712 async_checkpointer.py:280] [process=6][thread=save_finalize] Waiting for background save thread=async_save.
I0417 03:20:15.434318 135493274621760 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_09_pdb_lt_1/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776395985.7230203, 'wait_for_prev_duration_secs': 28.547726154327393, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776396014.2725859, 'checkpointer_blocking_duration_secs': 1.158860445022583, 'get_old_steps_start_time': 1776396015.4314704, 'get_old_steps_duration_secs': 2.9325485229492188e-05, 'checkpoint_manager_blocking_start_time': 1776395985.7211936, 'checkpoint_manager_blocking_duration_secs': 29.71308732032776}
I0417 03:20:15.434436 135493274621760 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0417 03:20:15.434480 135493274621760 checkpoint_manager.py:2020] [process=6][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0417 03:20:21.514464 135364832073472 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_09_pdb_lt_1/checkpoints/9/items/array_metadatas/process_6
I0417 03:20:56.645696 135364350048000 base_pytree_checkpoint_handler.py:130] [process=6] /jax/orbax/write/gbytes_per_sec: 38.297 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.246217250823975 s) (per-host)
I0417 03:20:56.645807 135364350048000 async_checkpointer.py:90] [process=6][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.214383s.
I0417 03:21:04.171983 135364350048000 async_checkpointer.py:160] [process=6][thread=async_save] Background save thread done. Time taken: 48.740544s.
I0417 03:21:04.172211 135365398611712 async_checkpointer.py:288] [process=6][thread=save_finalize] Done with waiting for background save thread=async_save.
I0417 03:21:04.172268 135365398611712 async_checkpointer.py:298] [process=6][thread=save_finalize] No errors found in background save thread=async_save.
I0417 03:21:04.172315 135365398611712 checkpoint_manager.py:2137] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0417 03:21:04.173877 135365398611712 checkpoint_manager.py:2146] [process=6][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0417 03:21:04.174070 135493274621760 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 03:21:04.174230 135493274621760 checkpoint_manager.py:2009] [process=6][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0417 03:21:04.174894 135493274621760 metric_logger.py:185] completed step: 9, seconds: 0.058, TFLOP/s/device: 58.339, Tokens/s/device: 8793.474, total_weights: 16384, loss: 5.892
Per train step:
 Total TFLOPs: 3.40 
 split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Fri Apr 17 03:21:13 UTC 2026
EXIT_CODE=0