XPK Start: Fri Apr 24 20:23:08 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.
2026-04-24 20:23:32.853320: 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)
I0424 20:23:33.064299 133601579358016 max_utils.py:273] Attempting to initialize the jax distributed system...
I0424 20:23:42.105354 133601579358016 distributed.py:149] Starting JAX distributed service on [::]:8482
I0424 20:23:42.107809 133601579358016 distributed.py:172] Connecting to JAX distributed service on mt-06-grad-accum-c9ljf-slice-job-0-0.mt-06-grad-accum-c9ljf:8482
I0424 20:23:43.491119 133601579358016 max_utils.py:284] Jax distributed system initialized!
I0424 20:23:48.693057 133601579358016 max_utils.py:800] System Information: Jax Version: 0.9.2
I0424 20:23:48.693166 133601579358016 max_utils.py:801] System Information: Jaxlib Version: 0.9.2
I0424 20:23:48.693219 133601579358016 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
I0424 20:23:48.693265 133601579358016 train_utils.py:361] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0424 20:23:49.388982 133601579358016 maxtext_utils.py:1565] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0424 20:23:49.389274 133601579358016 checkpointing.py:677] Setting up checkpoint logger...
I0424 20:23:49.389331 133601579358016 checkpointing.py:233] Creating checkpoint manager with ocdbt=True and zarr3=True
I0424 20:23:49.389377 133601579358016 pytree_checkpoint_handler.py:592] save_device_host_concurrent_bytes=None
I0424 20:23:49.389745 133601579358016 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 0x7981f1b1ff50>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0424 20:23:52.196730 133601579358016 checkpointing.py:265] Enabling policy for fixed interval checkpointing.
I0424 20:23:52.196971 133601579358016 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 0x796e68300260>}, handler_registry=None
I0424 20:23:52.197208 133601579358016 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x796e68300260>` 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`.
I0424 20:23:52.197255 133601579358016 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x796e68302180>` 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`.
I0424 20:23:52.197291 133601579358016 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 0x796e68300260>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x796e68300260>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x796e68302180>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x796e68302180>}).
I0424 20:23:52.197612 133601579358016 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.34
I0424 20:23:52.197696 133601579358016 async_checkpointer.py:192] [process=5][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x796e487ef560> timeout: 1200 secs and primary_host=0 for async checkpoint writes
I0424 20:23:53.396942 133601579358016 checkpoint_manager.py:1812] Found 0 checkpoint steps in gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_06_grad_accum/checkpoints
I0424 20:23:53.797089 133601579358016 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_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_06_grad_accum/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x7980c6c8c4a0>
I0424 20:23:53.797262 133601579358016 checkpointing.py:301] Checkpoint manager created!
I0424 20:23:55.496061 133601579358016 nnx_wrappers.py:453] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0424 20:23:55.496184 133601579358016 nnx_wrappers.py:453] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0424 20:23:55.879639 133601579358016 attentions.py:1088] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0424 20:23:55.879745 133601579358016 attentions.py:1088] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0424 20:23:55.896213 133601579358016 attentions.py:1089] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0424 20:23:55.896283 133601579358016 attentions.py:1089] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0424 20:23:55.920816 133601579358016 attentions.py:1154] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0424 20:23:55.920924 133601579358016 attentions.py:1154] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0424 20:23:55.941836 133601579358016 attentions.py:1155] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0424 20:23:55.941904 133601579358016 attentions.py:1155] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0424 20:23:55.958324 133601579358016 attentions.py:1156] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0424 20:23:55.958382 133601579358016 attentions.py:1156] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0424 20:23:55.983432 133601579358016 attentions.py:1198] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0424 20:23:55.983502 133601579358016 attentions.py:1198] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0424 20:23:56.004632 133601579358016 linears.py:525] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0424 20:23:56.004715 133601579358016 linears.py:525] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0424 20:23:56.224375 133601579358016 checkpointing.py:577] checkpoint manager exists so trying to load this run's existing checkpoint
I0424 20:23:56.224489 133601579358016 checkpointing.py:665] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0424 20:23:57.654486 133601579358016 maxtext_utils.py:1668] params/params/decoder/decoder_norm/scale
Shape: float32[2048]
Logical: P('norm',)
Physical: (None,)
I0424 20:23:57.654615 133601579358016 maxtext_utils.py:1668] params/params/decoder/layers/mlp/wi_0/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0424 20:23:57.654683 133601579358016 maxtext_utils.py:1668] params/params/decoder/layers/mlp/wi_1/kernel
Shape: float32[2048,16,7168]
Logical: P('embed', 'layers', 'mlp')
Physical: ('fsdp', None, None)
I0424 20:23:57.654746 133601579358016 maxtext_utils.py:1668] params/params/decoder/layers/mlp/wo/kernel
Shape: float32[7168,16,2048]
Logical: P('mlp', 'layers', 'embed')
Physical: (None, None, 'fsdp')
I0424 20:23:57.654805 133601579358016 maxtext_utils.py:1668] params/params/decoder/layers/post_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0424 20:23:57.654847 133601579358016 maxtext_utils.py:1668] params/params/decoder/layers/pre_self_attention_layer_norm/scale
Shape: float32[2048,16]
Logical: P('norm', 'layers')
Physical: (None, None)
I0424 20:23:57.654901 133601579358016 maxtext_utils.py:1668] params/params/decoder/layers/self_attention/key/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'kv_heads', 'kv_head_dim')
Physical: ('fsdp', None, None, None)
I0424 20:23:57.654968 133601579358016 maxtext_utils.py:1668] params/params/decoder/layers/self_attention/out/kernel
Shape: float32[16,16,128,2048]
Logical: P('heads', 'layers', 'kv', 'embed')
Physical: (None, None, None, 'fsdp')
I0424 20:23:57.655011 133601579358016 maxtext_utils.py:1668] params/params/decoder/layers/self_attention/query/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'q_heads', 'kv')
Physical: ('fsdp', None, None, None)
I0424 20:23:57.655049 133601579358016 maxtext_utils.py:1668] params/params/decoder/layers/self_attention/value/kernel
Shape: float32[2048,16,16,128]
Logical: P('embed', 'layers', 'kv_heads', 'kv_head_dim')
Physical: ('fsdp', None, None, None)
I0424 20:23:57.655098 133601579358016 maxtext_utils.py:1668] params/params/decoder/logits_dense/kernel
Shape: float32[2048,32000]
Logical: P('embed_vocab', 'vocab')
Physical: ('fsdp', None)
I0424 20:23:57.655147 133601579358016 maxtext_utils.py:1668] params/params/token_embedder/embedding
Shape: float32[32000,2048]
Logical: P('vocab', 'embed_vocab')
Physical: (None, 'fsdp')
I0424 20:23:57.678830 133601579358016 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[2048]............................................... Unknown.
I0424 20:23:57.678898 133601579358016 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[2048]............................................... (None,).
I0424 20:23:57.693778 133601579358016 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[2048,16,7168]....................................... Unknown.
I0424 20:23:57.693837 133601579358016 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[2048,16,7168]....................................... ('fsdp', None, None).
I0424 20:23:57.723725 133601579358016 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[7168,16,2048]....................................... Unknown.
I0424 20:23:57.723789 133601579358016 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[7168,16,2048]....................................... (None, None, 'fsdp').
I0424 20:23:57.738632 133601579358016 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[2048,16]............................................ Unknown.
I0424 20:23:57.738699 133601579358016 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[2048,16]............................................ (None, None).
I0424 20:23:57.768301 133601579358016 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[2048,16,16,128]..................................... Unknown.
I0424 20:23:57.768372 133601579358016 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[2048,16,16,128]..................................... ('fsdp', None, None, None).
I0424 20:23:57.783203 133601579358016 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[16,16,128,2048]..................................... Unknown.
I0424 20:23:57.783260 133601579358016 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[16,16,128,2048]..................................... (None, None, None, 'fsdp').
I0424 20:23:57.827843 133601579358016 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[2048,32000]......................................... Unknown.
I0424 20:23:57.827913 133601579358016 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[2048,32000]......................................... ('fsdp', None).
I0424 20:23:57.842839 133601579358016 gradient_accumulation.py:68] gradient_accumulation/inputs Logical: float32[32000,2048]......................................... Unknown.
I0424 20:23:57.842897 133601579358016 gradient_accumulation.py:68] gradient_accumulation/inputs Physical: float32[32000,2048]......................................... (None, 'fsdp').
I0424 20:23:58.517005 133601579358016 train.py:155] train/xent Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0424 20:23:58.517096 133601579358016 train.py:155] train/xent Physical: float32[32,2048]............................................ ('fsdp', None).
I0424 20:23:58.532726 133601579358016 train.py:162] train/z_loss Logical: float32[32,2048]............................................ ('activation_embed_and_logits_batch', 'activation_length').
I0424 20:23:58.532791 133601579358016 train.py:162] train/z_loss Physical: float32[32,2048]............................................ ('fsdp', None).
I0424 20:24:10.016345 133601579358016 max_utils.py:791] Total memory size: 1.7 GB, Output size: 0.4 GB, Temp size: 1.3 GB, Argument size: 0.4 GB, Host temp size: 0.0 GB.
I0424 20:24:10.017140 133601579358016 metric_logger.py:301] number parameters: 1.104 billion
I0424 20:24:12.358215 133601579358016 checkpointing.py:772] Waiting for step 0 to finish before checkpoint...
I0424 20:24:22.903255 133601579358016 checkpointing.py:776] Waited 10.545024871826172 seconds for step 0 to finish before starting checkpointing.
I0424 20:24:22.905629 133601579358016 checkpoint_manager.py:2009] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0424 20:24:22.907415 133601579358016 checkpoint_manager.py:1512] [process=5] Saving checkpoint at step 0
I0424 20:24:22.908919 133601579358016 event_tracking.py:70] [process=5] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_06_grad_accum/checkpoints/0.
I0424 20:24:23.642823 133601579358016 signaling_client.py:364] Using JaxDistributedSignalingClient
I0424 20:24:23.643758 133601579358016 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0424 20:24:23.643817 133601579358016 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
I0424 20:24:23.921152 133601579358016 base_pytree_checkpoint_handler.py:154] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.278403s
I0424 20:24:23.921335 133601579358016 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/blocking_gbytes_per_sec: 5.434 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.28389549255371094 s) (per-host)
I0424 20:24:23.921388 133601579358016 base_pytree_checkpoint_handler.py:768] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.283959s (batch_requests_ready=0.002209s, total_serialization_initiated=0.281678s, others=0.000071s)
I0424 20:24:23.921478 133601579358016 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.288241s (all_items=0.000018s, per_item={'items': '0.00001812'}, temp_paths=0.288223)
I0424 20:24:23.922264 133601579358016 event_tracking.py:125] [process=5] [async] Finished blocking save in 1.01 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_06_grad_accum/checkpoints/0.
I0424 20:24:23.922605 133476485416704 async_checkpointer.py:76] [process=5][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-24 20:44:23.922566
I0424 20:24:23.929034 133601579358016 checkpoint_manager.py:1560] [process=5][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0424 20:24:23.929287 133475919181568 async_checkpointer.py:280] [process=5][thread=save_finalize] Waiting for background save thread=async_save.
I0424 20:24:23.929410 133601579358016 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_06_grad_accum/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777062262.9056108, 'wait_for_prev_duration_secs': 7.748603820800781e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777062262.907452, 'checkpointer_blocking_duration_secs': 1.0153191089630127, 'get_old_steps_start_time': 1777062263.922797, 'get_old_steps_duration_secs': 3.0040740966796875e-05, 'checkpoint_manager_blocking_start_time': 1777062262.9038353, 'checkpoint_manager_blocking_duration_secs': 1.0255348682403564}
I0424 20:24:23.929517 133601579358016 checkpointing.py:408] Started an asynchronous checkpoint save for step 0
I0424 20:24:23.929565 133601579358016 max_utils.py:750]
Memstats: After params initialized:
I0424 20:24:23.929614 133601579358016 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_18(process=5,(2,4,0,0))
I0424 20:24:23.929647 133601579358016 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_19(process=5,(3,4,0,0))
I0424 20:24:23.929699 133601579358016 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_22(process=5,(2,5,0,0))
I0424 20:24:23.929724 133601579358016 max_utils.py:756] Using (GB) 0.43 / 31.25 (1.376000%) on TPU_23(process=5,(3,5,0,0))
I0424 20:24:24.246626 133601579358016 metric_logger.py:196] completed step: 0, seconds: 2.341, TFLOP/s/device: 23.216, Tokens/s/device: 3499.413, total_weights: 262144, loss: 10.877, lm_loss: 10.877, perplexity: 52959.059
I0424 20:24:24.851211 133601579358016 metric_logger.py:196] completed step: 1, seconds: 11.887, TFLOP/s/device: 4.572, Tokens/s/device: 689.160, total_weights: 262144, loss: 10.877, lm_loss: 10.877, perplexity: 52959.059
I0424 20:24:25.429447 133601579358016 metric_logger.py:196] completed step: 2, seconds: 0.027, TFLOP/s/device: 2045.484, Tokens/s/device: 308317.651, total_weights: 262144, loss: 10.563, lm_loss: 10.563, perplexity: 38662.707
I0424 20:24:26.007688 133601579358016 metric_logger.py:196] completed step: 3, seconds: 0.584, TFLOP/s/device: 93.101, Tokens/s/device: 14033.188, total_weights: 262144, loss: 10.272, lm_loss: 10.272, perplexity: 28909.668
I0424 20:24:26.826536 2550 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0424 20:24:27.164228 133601579358016 metric_logger.py:196] completed step: 4, seconds: 0.578, TFLOP/s/device: 93.996, Tokens/s/device: 14168.034, total_weights: 262144, loss: 10.022, lm_loss: 10.022, perplexity: 22524.992
I0424 20:24:27.170389 133601579358016 metric_logger.py:196] completed step: 5, seconds: 0.578, TFLOP/s/device: 93.981, Tokens/s/device: 14165.805, total_weights: 262144, loss: 9.820, lm_loss: 9.820, perplexity: 18401.865
I0424 20:24:29.300336 133475967850240 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_06_grad_accum/checkpoints/0/items/array_metadatas/process_5
I0424 20:25:01.082912 133476485416704 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/gbytes_per_sec: 42.184 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 37.44543933868408 s) (per-host)
I0424 20:25:01.083021 133476485416704 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 37.160289s.
I0424 20:25:08.752895 133476485416704 async_checkpointer.py:160] [process=5][thread=async_save] Background save thread done. Time taken: 44.830147s.
I0424 20:25:08.753112 133475919181568 async_checkpointer.py:288] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save.
I0424 20:25:08.753170 133475919181568 async_checkpointer.py:298] [process=5][thread=save_finalize] No errors found in background save thread=async_save.
I0424 20:25:08.753216 133475919181568 checkpoint_manager.py:2137] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0424 20:25:08.755157 133475919181568 checkpoint_manager.py:2146] [process=5][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0424 20:25:13.715607 133601579358016 metric_logger.py:196] completed step: 6, seconds: 1.157, TFLOP/s/device: 46.963, Tokens/s/device: 7078.814, total_weights: 262144, loss: 9.667, lm_loss: 9.667, perplexity: 15787.604
I0424 20:25:14.293955 133601579358016 metric_logger.py:196] completed step: 7, seconds: 45.967, TFLOP/s/device: 1.182, Tokens/s/device: 178.213, total_weights: 262144, loss: 9.561, lm_loss: 9.561, perplexity: 14203.827
I0424 20:25:14.871963 133601579358016 metric_logger.py:196] completed step: 8, seconds: 0.583, TFLOP/s/device: 93.142, Tokens/s/device: 14039.441, total_weights: 262144, loss: 9.496, lm_loss: 9.496, perplexity: 13302.920
I0424 20:25:15.449464 133601579358016 checkpointing.py:772] Waiting for step 9 to finish before checkpoint...
I0424 20:25:15.450047 133601579358016 checkpointing.py:776] Waited 0.0006046295166015625 seconds for step 9 to finish before starting checkpointing.
I0424 20:25:15.452154 133601579358016 checkpoint_manager.py:2009] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0424 20:25:15.454068 133601579358016 checkpoint_manager.py:1512] [process=5] Saving checkpoint at step 9
I0424 20:25:15.455457 133601579358016 event_tracking.py:70] [process=5] [async] Started save checkpoint @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_06_grad_accum/checkpoints/9.
I0424 20:25:16.166639 133601579358016 jax_array_handlers.py:360] Scheduling D2H of 39 prioritized jax.Array.
I0424 20:25:16.166749 133601579358016 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
I0424 20:25:16.202480 133601579358016 base_pytree_checkpoint_handler.py:154] [process=5][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.036870s
I0424 20:25:16.202661 133601579358016 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/blocking_gbytes_per_sec: 38.207 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.0403742790222168 s) (per-host)
I0424 20:25:16.202715 133601579358016 base_pytree_checkpoint_handler.py:768] [process=5][thread=MainThread] Initiated Pytree async_save. Time taken: 0.040449s (batch_requests_ready=0.001694s, total_serialization_initiated=0.038672s, others=0.000083s)
I0424 20:25:16.202809 133601579358016 composite_checkpoint_handler.py:715] [process=5][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.044658s (all_items=0.000015s, per_item={'items': '0.00001502'}, temp_paths=0.044643)
I0424 20:25:16.203558 133601579358016 event_tracking.py:125] [process=5] [async] Finished blocking save in 0.75 seconds. Continuing save @ gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_06_grad_accum/checkpoints/9.
I0424 20:25:16.203914 133453282535168 async_checkpointer.py:76] [process=5][thread=async_save] Background save thread started. Deadline for this save operation is 2026-04-24 20:45:16.203877
I0424 20:25:16.210950 133601579358016 checkpoint_manager.py:1560] [process=5][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0424 20:25:16.211256 133475967850240 async_checkpointer.py:280] [process=5][thread=save_finalize] Waiting for background save thread=async_save.
I0424 20:25:16.211425 133601579358016 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_06_grad_accum/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777062315.4521205, 'wait_for_prev_duration_secs': 7.319450378417969e-05, 'time_between_consecutive_saves_sec': 6.696922540664673, 'checkpointer_blocking_start_time': 1777062315.4541063, 'checkpointer_blocking_duration_secs': 0.7499542236328125, 'get_old_steps_start_time': 1777062316.2040834, 'get_old_steps_duration_secs': 2.8848648071289062e-05, 'checkpoint_manager_blocking_start_time': 1777062315.4502797, 'checkpoint_manager_blocking_duration_secs': 0.7611103057861328}
I0424 20:25:16.211529 133601579358016 checkpointing.py:408] Started an asynchronous checkpoint save for step 9
I0424 20:25:16.211573 133601579358016 checkpoint_manager.py:2020] [process=5][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0424 20:25:21.314315 133480260298496 array_metadata_store.py:203] [process=5][thread=array_type_handler] Wrote 39 array_metadata.ArrayMetadata to gs://lance-maxtext/linen_ckpt_xpk_test_pipeline_scan_nnx_20260424_200834/linen_xpk_test_pipeline_scan_nnx_20260424_200834_06_grad_accum/checkpoints/9/items/array_metadatas/process_5
I0424 20:25:58.990206 133453282535168 base_pytree_checkpoint_handler.py:130] [process=5] /jax/orbax/write/gbytes_per_sec: 36.882 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.82789325714111 s) (per-host)
I0424 20:25:58.990343 133453282535168 async_checkpointer.py:90] [process=5][thread=async_save] 3 Handler Commit operations completed. Time taken: 42.786318s.
I0424 20:26:05.658918 133453282535168 async_checkpointer.py:160] [process=5][thread=async_save] Background save thread done. Time taken: 49.454877s.
I0424 20:26:05.659214 133475967850240 async_checkpointer.py:288] [process=5][thread=save_finalize] Done with waiting for background save thread=async_save.
I0424 20:26:05.659346 133475967850240 async_checkpointer.py:298] [process=5][thread=save_finalize] No errors found in background save thread=async_save.
I0424 20:26:05.659397 133475967850240 checkpoint_manager.py:2137] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0424 20:26:05.661595 133475967850240 checkpoint_manager.py:2146] [process=5][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0424 20:26:05.661783 133601579358016 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.
I0424 20:26:05.661936 133601579358016 checkpoint_manager.py:2009] [process=5][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0424 20:26:05.662889 133601579358016 metric_logger.py:196] completed step: 9, seconds: 0.578, TFLOP/s/device: 93.986, Tokens/s/device: 14166.613, total_weights: 262144, loss: 9.457, lm_loss: 9.457, perplexity: 12802.546
Per train step:
Total TFLOPs: 54.35
split as 93.93% learnable weight flops and 6.07% attention flops
XPK End: Fri Apr 24 20:26:18 UTC 2026
EXIT_CODE=0