XPK Start: Wed Apr 22 12:47:24 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 12:48:04.082899: 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 12:48:04.645382 137009129506624 max_utils.py:273] Attempting to initialize the jax distributed system...
I0422 12:48:13.685555 137009129506624 distributed.py:149] Starting JAX distributed service on [::]:8482
I0422 12:48:13.687934 137009129506624 distributed.py:172] Connecting to JAX distributed service on mt-03-dropout-9ysfu-slice-job-0-0.mt-03-dropout-9ysfu:8482
I0422 12:48:17.351008 137009129506624 max_utils.py:284] Jax distributed system initialized!
I0422 12:48:23.290495 137009129506624 max_utils.py:800] System Information: Jax Version: 0.9.2
I0422 12:48:23.290603 137009129506624 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0422 12:48:23.290645 137009129506624 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 12:48:23.290730 137009129506624 train_utils.py:391] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0422 12:48:23.989313 137009129506624 maxtext_utils.py:1771] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0422 12:48:23.989911 137009129506624 maxtext_utils.py:1771] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0422 12:48:23.990092 137009129506624 checkpointing.py:688] Setting up checkpoint logger...
I0422 12:48:23.990142 137009129506624 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0422 12:48:23.990184 137009129506624 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0422 12:48:23.990528 137009129506624 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 0x7c9b4bb67dd0>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0422 12:48:26.869266 137009129506624 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0422 12:48:26.869504 137009129506624 checkpoint_manager.py:708] [process=5][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c872c10a300>}, handler_registry=None
I0422 12:48:26.869762 137009129506624 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c872c10a300>` 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 12:48:26.869815 137009129506624 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c872c1120f0>` 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 12:48:26.869848 137009129506624 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 0x7c872c10a300>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x7c872c10a300>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c872c1120f0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x7c872c1120f0>}).
I0422 12:48:26.870167 137009129506624 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0422 12:48:26.870240 137009129506624 async_checkpointer.py:192] [process=5][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x7c87046cd940> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0422 12:48:27.987390 137009129506624 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260422_123906/linen_xpk_feat_nnx_post_train_fixes_20260422_123906_03_dropout/checkpoints
I0422 12:48:28.011948 137009129506624 checkpoint_manager.py:929] [process=5][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_20260422_123906/linen_xpk_feat_nnx_post_train_fixes_20260422_123906_03_dropout/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7c872c1092e0>
I0422 12:48:28.012085 137009129506624 checkpointing.py:302] Checkpoint manager created!
I0422 12:48:28.364672 137009129506624 nnx_wrappers.py:437] Unknown Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0422 12:48:28.364781 137009129506624 nnx_wrappers.py:437] Unknown Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0422 12:48:28.741982 137009129506624 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0422 12:48:28.742074 137009129506624 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0422 12:48:28.758624 137009129506624 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,128,2048]....................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0422 12:48:28.758694 137009129506624 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,128,2048]....................................... ('fsdp', None, None).
I0422 12:48:28.782147 137009129506624 attentions.py:1154] attentions/query Logical: bfloat16[32,128,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0422 12:48:28.782213 137009129506624 attentions.py:1154] attentions/query Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0422 12:48:28.798526 137009129506624 attentions.py:1155] attentions/key Logical: bfloat16[32,128,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0422 12:48:28.798583 137009129506624 attentions.py:1155] attentions/key Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0422 12:48:28.814690 137009129506624 attentions.py:1156] attentions/value Logical: bfloat16[32,128,16,128]..................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0422 12:48:28.814753 137009129506624 attentions.py:1156] attentions/value Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0422 12:48:28.839531 137009129506624 attentions.py:1198] attentions/out Logical: bfloat16[32,128,16,128]..................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0422 12:48:28.839602 137009129506624 attentions.py:1198] attentions/out Physical: bfloat16[32,128,16,128]..................................... ('fsdp', None, None, None).
I0422 12:48:28.866078 137009129506624 linears.py:525] linears/x Logical: bfloat16[32,128,7168]....................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0422 12:48:28.866145 137009129506624 linears.py:525] linears/x Physical: bfloat16[32,128,7168]....................................... ('fsdp', None, None).
I0422 12:48:29.092581 137009129506624 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0422 12:48:29.092721 137009129506624 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0422 12:48:30.527520 137009129506624 maxtext_utils.py:1880] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: P('norm',)
Physical: (None,)
I0422 12:48:30.527647 137009129506624 maxtext_utils.py:1880] params/params/decoder/layers/mlp/wi_0/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0422 12:48:30.527718 137009129506624 maxtext_utils.py:1880] params/params/decoder/layers/mlp/wi_1/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0422 12:48:30.527777 137009129506624 maxtext_utils.py:1880] params/params/decoder/layers/mlp/wo/kernel
Shape: float32[7168,16,2048]
Logical: P('mlp', 'layers', 'embed')
Physical: (None, None, 'fsdp')
I0422 12:48:30.527828 137009129506624 maxtext_utils.py:1880] params/params/decoder/layers/post_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0422 12:48:30.527866 137009129506624 maxtext_utils.py:1880] params/params/decoder/layers/pre_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0422 12:48:30.527917 137009129506624 maxtext_utils.py:1880] 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 12:48:30.527967 137009129506624 maxtext_utils.py:1880] 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 12:48:30.528004 137009129506624 maxtext_utils.py:1880] 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 12:48:30.528038 137009129506624 maxtext_utils.py:1880] 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 12:48:30.528084 137009129506624 maxtext_utils.py:1880] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: P('embed_vocab', 'vocab')
Physical: ('fsdp', None)
I0422 12:48:30.528133 137009129506624 maxtext_utils.py:1880] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: P('vocab', 'embed_vocab')
Physical: (None, 'fsdp')
I0422 12:48:31.052219 137009129506624 train.py:157] train/xent Logical: float32[32,128]............................................. ('activation_embed_and_logits_batch', 'activation_length').
I0422 12:48:31.052308 137009129506624 train.py:157] train/xent Physical: float32[32,128]............................................. ('fsdp', None).
I0422 12:48:31.067902 137009129506624 train.py:164] train/z_loss Logical: float32[32,128]............................................. ('activation_embed_and_logits_batch', 'activation_length').
I0422 12:48:31.067963 137009129506624 train.py:164] train/z_loss Physical: float32[32,128]............................................. ('fsdp', None).
I0422 12:48:34.634583 137009129506624 max_utils.py:791] Total memory size: 0.8 GB, Output size: 0.4 GB, Temp size: 0.4 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB.
I0422 12:48:34.635322 137009129506624 metric_logger.py:301] number parameters: 1.104 billion
I0422 12:48:38.806698 137009129506624 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0422 12:48:39.260607 137009129506624 checkpointing.py:798] Waited 0.4538850784301758 seconds for step 0 to finish before starting checkpointing.
I0422 12:48:39.263034 137009129506624 checkpoint_manager.py:2009] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0422 12:48:39.265125 137009129506624 checkpoint_manager.py:1512] [process=5] Saving checkpoint at step 0
I0422 12:48:39.266515 137009129506624 event_tracking.py:70] [process=5] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260422_123906/linen_xpk_feat_nnx_post_train_fixes_20260422_123906_03_dropout/checkpoints/0.
I0422 12:48:39.982931 137009129506624 signaling_client.py:364] Using JaxDistributedSignalingClient
I0422 12:48:39.983817 137009129506624 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0422 12:48:39.983874 137009129506624 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 12:48:40.263917 137009129506624 base_pytree_checkpoint_handler.py:154] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.281056s
I0422 12:48:40.264097 137009129506624 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/blocking_gbytes_per_sec: 5.382 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2866194248199463 s) (per-host)
I0422 12:48:40.264154 137009129506624 base_pytree_checkpoint_handler.py:768] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.286686s (batch_requests_ready=0.002435s, total_serialization_initiated=0.284175s, others=0.000075s)
I0422 12:48:40.264246 137009129506624 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.290779s (all_items=0.000016s, per_item={'items': '0.00001645'}, temp_paths=0.290763)
I0422 12:48:40.265068 137009129506624 event_tracking.py:125] [process=5] [async] Finished blocking save in 1.00 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260422_123906/linen_xpk_feat_nnx_post_train_fixes_20260422_123906_03_dropout/checkpoints/0.
I0422 12:48:40.265358 136881258968832 async_checkpointer.py:76] [process=5][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-22 13:08:40.265324
I0422 12:48:40.283301 137009129506624 checkpoint_manager.py:1560] [process=5][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0422 12:48:40.283551 136879657240320 async_checkpointer.py:280] [process=5][thread=save_finalize] Waiting for background save thread=async_save.
I0422 12:48:40.283702 137009129506624 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260422_123906/linen_xpk_feat_nnx_post_train_fixes_20260422_123906_03_dropout/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776862119.2630172, 'wait_for_prev_duration_secs': 5.8650970458984375e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776862119.2651646, 'checkpointer_blocking_duration_secs': 1.0003373622894287, 'get_old_steps_start_time': 1776862120.2655265, 'get_old_steps_duration_secs': 3.0040740966796875e-05, 'checkpoint_manager_blocking_start_time': 1776862119.2612147, 'checkpoint_manager_blocking_duration_secs': 1.0224263668060303}
I0422 12:48:40.283804 137009129506624 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0422 12:48:40.283854 137009129506624 max_utils.py:750]
Memstats: After params initialized:
I0422 12:48:40.283906 137009129506624 max_utils.py:756] Using (GB) 0.41 / 31.25 (1.312000%) on TPU_18(process=5,(2,4,0,0))
I0422 12:48:40.283937 137009129506624 max_utils.py:756] Using (GB) 0.41 / 31.25 (1.312000%) on TPU_19(process=5,(3,4,0,0))
I0422 12:48:40.283965 137009129506624 max_utils.py:756] Using (GB) 0.41 / 31.25 (1.312000%) on TPU_22(process=5,(2,5,0,0))
I0422 12:48:40.283989 137009129506624 max_utils.py:756] Using (GB) 0.41 / 31.25 (1.312000%) on TPU_23(process=5,(3,5,0,0))
I0422 12:48:40.594197 137009129506624 metric_logger.py:196] completed step: 0, seconds: 4.171, TFLOP/s/device: 0.192, Tokens/s/device: 30.686, total_weights: 4096, loss: 10.889, lm_loss: 10.889, perplexity: 53560.699
I0422 12:48:40.699947 137009129506624 metric_logger.py:196] completed step: 1, seconds: 1.786, TFLOP/s/device: 0.448, Tokens/s/device: 71.670, total_weights: 4096, loss: 10.902, lm_loss: 10.902, perplexity: 54298.523
I0422 12:48:41.099136 137009129506624 metric_logger.py:196] completed step: 2, seconds: 0.036, TFLOP/s/device: 22.451, Tokens/s/device: 3588.248, total_weights: 4096, loss: 9.956, lm_loss: 9.956, perplexity: 21082.051
I0422 12:48:41.169824 137009129506624 metric_logger.py:196] completed step: 3, seconds: 0.399, TFLOP/s/device: 2.006, Tokens/s/device: 320.603, total_weights: 4096, loss: 9.139, lm_loss: 9.139, perplexity: 9314.691
I0422 12:48:41.311857 137009129506624 metric_logger.py:196] completed step: 4, seconds: 0.076, TFLOP/s/device: 10.544, Tokens/s/device: 1685.231, total_weights: 4096, loss: 8.459, lm_loss: 8.459, perplexity: 4715.493
I0422 12:48:41.317899 137009129506624 metric_logger.py:196] completed step: 5, seconds: 0.071, TFLOP/s/device: 11.314, Tokens/s/device: 1808.267, total_weights: 4096, loss: 7.924, lm_loss: 7.924, perplexity: 2762.373
I0422 12:48:43.182598 2828 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0422 12:48:45.657625 136879665633024 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260422_123906/linen_xpk_feat_nnx_post_train_fixes_20260422_123906_03_dropout/checkpoints/0/items/array_metadatas/process_5
I0422 12:48:51.550645 137009129506624 metric_logger.py:196] completed step: 6, seconds: 0.143, TFLOP/s/device: 5.616, Tokens/s/device: 897.502, total_weights: 4096, loss: 7.553, lm_loss: 7.553, perplexity: 1906.617
I0422 12:48:51.621322 137009129506624 metric_logger.py:196] completed step: 7, seconds: 10.163, TFLOP/s/device: 0.079, Tokens/s/device: 12.594, total_weights: 4096, loss: 7.267, lm_loss: 7.267, perplexity: 1431.653
I0422 12:48:51.692066 137009129506624 metric_logger.py:196] completed step: 8, seconds: 0.075, TFLOP/s/device: 10.684, Tokens/s/device: 1707.600, total_weights: 4096, loss: 7.111, lm_loss: 7.111, perplexity: 1225.983
I0422 12:48:51.762151 137009129506624 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0422 12:48:51.762800 137009129506624 checkpointing.py:798] Waited 0.0006687641143798828 seconds for step 9 to finish before starting checkpointing.
I0422 12:48:51.764831 137009129506624 checkpoint_manager.py:2020] [process=5][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0422 12:49:12.276750 136881258968832 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/gbytes_per_sec: 48.905 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 32.2992217540741 s) (per-host)
I0422 12:49:12.276862 136881258968832 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 32.011398s.
I0422 12:49:21.241104 136881258968832 async_checkpointer.py:160] [process=5][thread=async_save] Background save thread done. Time taken: 40.975621s.
I0422 12:49:21.241392 136879657240320 async_checkpointer.py:288] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save.
I0422 12:49:21.241510 136879657240320 async_checkpointer.py:298] [process=5][thread=save_finalize] No errors found in background save thread=async_save.
I0422 12:49:21.241560 136879657240320 checkpoint_manager.py:2137] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0422 12:49:21.243662 136879657240320 checkpoint_manager.py:2146] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0422 12:49:21.243836 137009129506624 checkpoint_manager.py:2032] [process=5][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0422 12:49:21.243978 137009129506624 checkpoint_manager.py:1452] Waiting for previous save to complete took 29.479146 seconds. If this number is high, consider checkpointing less frequently.
I0422 12:49:21.246042 137009129506624 checkpoint_manager.py:1512] [process=5] Saving checkpoint at step 9
I0422 12:49:21.248078 137009129506624 event_tracking.py:70] [process=5] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260422_123906/linen_xpk_feat_nnx_post_train_fixes_20260422_123906_03_dropout/checkpoints/9.
I0422 12:49:21.949378 137009129506624 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0422 12:49:21.949470 137009129506624 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 12:49:21.984413 137009129506624 base_pytree_checkpoint_handler.py:154] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.036124s
I0422 12:49:21.984583 137009129506624 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/blocking_gbytes_per_sec: 38.966 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.039588212966918945 s) (per-host)
I0422 12:49:21.984634 137009129506624 base_pytree_checkpoint_handler.py:768] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.039649s (batch_requests_ready=0.001850s, total_serialization_initiated=0.037730s, others=0.000069s)
I0422 12:49:21.984745 137009129506624 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.043838s (all_items=0.000015s, per_item={'items': '0.00001550'}, temp_paths=0.043823)
I0422 12:49:21.985461 137009129506624 event_tracking.py:125] [process=5] [async] Finished blocking save in 0.74 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260422_123906/linen_xpk_feat_nnx_post_train_fixes_20260422_123906_03_dropout/checkpoints/9.
I0422 12:49:21.985793 136879657240320 async_checkpointer.py:76] [process=5][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-22 13:09:21.985760
I0422 12:49:21.987941 137009129506624 checkpoint_manager.py:1560] [process=5][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0422 12:49:21.988276 136879665633024 async_checkpointer.py:280] [process=5][thread=save_finalize] Waiting for background save thread=async_save.
I0422 12:49:21.988419 137009129506624 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260422_123906/linen_xpk_feat_nnx_post_train_fixes_20260422_123906_03_dropout/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1776862131.7647943, 'wait_for_prev_duration_secs': 29.479146242141724, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1776862161.2460816, 'checkpointer_blocking_duration_secs': 0.7398524284362793, 'get_old_steps_start_time': 1776862161.9859676, 'get_old_steps_duration_secs': 3.814697265625e-05, 'checkpoint_manager_blocking_start_time': 1776862131.7630248, 'checkpoint_manager_blocking_duration_secs': 30.22535514831543}
I0422 12:49:21.988597 137009129506624 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0422 12:49:21.988647 137009129506624 checkpoint_manager.py:2020] [process=5][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0422 12:49:28.369987 136856612001536 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_feat_nnx_post_train_fixes_20260422_123906/linen_xpk_feat_nnx_post_train_fixes_20260422_123906_03_dropout/checkpoints/9/items/array_metadatas/process_5
I0422 12:50:02.985193 136879657240320 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/gbytes_per_sec: 38.489 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 41.04015922546387 s) (per-host)
I0422 12:50:02.985320 136879657240320 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 40.999417s.
I0422 12:50:11.418372 136879657240320 async_checkpointer.py:160] [process=5][thread=async_save] Background save thread done. Time taken: 49.432453s.
I0422 12:50:11.418699 136879665633024 async_checkpointer.py:288] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save.
I0422 12:50:11.418832 136879665633024 async_checkpointer.py:298] [process=5][thread=save_finalize] No errors found in background save thread=async_save.
I0422 12:50:11.418883 136879665633024 checkpoint_manager.py:2137] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0422 12:50:11.420145 136879665633024 checkpoint_manager.py:2146] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0422 12:50:11.420329 137009129506624 checkpoint_manager.py:2032] [process=5][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0422 12:50:11.420478 137009129506624 checkpoint_manager.py:2009] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0422 12:50:11.421445 137009129506624 metric_logger.py:196] completed step: 9, seconds: 0.071, TFLOP/s/device: 11.350, Tokens/s/device: 1814.085, total_weights: 4096, loss: 6.981, lm_loss: 6.981, perplexity: 1075.867
Per train step:
Total TFLOPs: 0.80
split as 99.60% learnable weight flops and 0.40% attention flops
XPK End: Wed Apr 22 12:50:19 UTC 2026
EXIT_CODE=0