MaxView

XPK Start: Fri Apr 24 17:03:36 UTC 2026
2026-04-24 17:03:40.434169: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:467] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
WARNING: All log messages before absl::InitializeLog() is called are written to STDERR
E0000 00:00:1777050220.447314      10 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
E0000 00:00:1777050220.451028      10 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
W0000 00:00:1777050220.462579      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1777050220.462600      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1777050220.462602      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
W0000 00:00:1777050220.462604      10 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once.
2026-04-24 17:03:59.727227: 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 17:04:00.251428 132939609249600 max_utils.py:273] Attempting to initialize the jax distributed system...
INFO:2026-04-24 17:04:09,292:jax._src.distributed:140: Starting JAX distributed service on [::]:8482
I0424 17:04:09.292767 132939609249600 distributed.py:140] Starting JAX distributed service on [::]:8482
INFO:2026-04-24 17:04:09,295:jax._src.distributed:157: Connecting to JAX distributed service on mt-11-optimizer-offload-zjm3w-slice-job-0-0.mt-11-optimizer-offload-zjm3w:8482
I0424 17:04:09.295211 132939609249600 distributed.py:157] Connecting to JAX distributed service on mt-11-optimizer-offload-zjm3w-slice-job-0-0.mt-11-optimizer-offload-zjm3w:8482
I0424 17:04:10.230368 132939609249600 max_utils.py:284] Jax distributed system initialized!
I0424 17:04:16.992288 132939609249600 max_utils.py:800] System Information: Jax Version: 0.8.1
I0424 17:04:16.992394 132939609249600 max_utils.py:801] System Information: Jaxlib Version: 0.8.1
I0424 17:04:16.992440 132939609249600 max_utils.py:802] System Information: Jax Backend: PJRT C API
TFRT TPU v6 lite
Built on Nov 12 2025 14:16:36 (1762985796) cl/831091709
I0424 17:04:16.992475 132939609249600 train_utils.py:377] WARNING: Sequence packing is essentially ignored for synthetic data. Please use a real dataset to use sequence packing.
I0424 17:04:17.612093 132939609249600 maxtext_utils.py:1631] Num_devices: 32, shape (1, 1, 1, 32, 1, 1, 1, 1, 1, 1, 1, 1, 1)
I0424 17:04:17.726521 132939609249600 checkpointing.py:688] Setting up checkpoint logger...
I0424 17:04:17.726640 132939609249600 checkpointing.py:234] Creating checkpoint manager with ocdbt=True and zarr3=True
I0424 17:04:17.726686 132939609249600 pytree_checkpoint_handler.py:589] save_device_host_concurrent_bytes=None
I0424 17:04:17.726890 132939609249600 base_pytree_checkpoint_handler.py:415] 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 0x78e7edd78f20>, enable_pinned_host_transfer=False, save_concurrent_bytes: 96000000000 (89.4 GiB), restore_concurrent_bytes: 96000000000 (89.4 GiB)
I0424 17:04:20.638476 132939609249600 checkpointing.py:266] Enabling policy for fixed interval checkpointing.
I0424 17:04:20.638718 132939609249600 checkpoint_manager.py:709] [process=3][thread=MainThread] CheckpointManager init: checkpointers=None, item_names=('items',), item_handlers={'items': <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x78d46f9eb200>}, handler_registry=None
I0424 17:04:20.638970 132939609249600 composite_checkpoint_handler.py:237] Deferred registration for item: "items". Adding handler `<orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x78d46f9eb200>` 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 17:04:20.639021 132939609249600 composite_checkpoint_handler.py:237] Deferred registration for item: "metrics". Adding handler `<orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x78d3784768d0>` 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 17:04:20.639058 132939609249600 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 0x78d46f9eb200>, ('items', <class 'orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeRestoreArgs'>): <orbax.checkpoint._src.handlers.pytree_checkpoint_handler.PyTreeCheckpointHandler object at 0x78d46f9eb200>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonSaveArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x78d3784768d0>, ('metrics', <class 'orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonRestoreArgs'>): <orbax.checkpoint._src.handlers.json_checkpoint_handler.JsonCheckpointHandler object at 0x78d3784768d0>}).
I0424 17:04:20.639384 132939609249600 abstract_checkpointer.py:35] orbax-checkpoint version: 0.11.33
I0424 17:04:20.639467 132939609249600 async_checkpointer.py:177] [process=3][thread=MainThread] Using barrier_sync_fn: <function get_barrier_sync_fn.<locals>._fn at 0x78d46ce94f40> timeout: 600 secs and primary_host=0 for async checkpoint writes
I0424 17:04:21.953180 132939609249600 checkpoint_manager.py:1818] Found 0 checkpoint steps in gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_11_optimizer_offload_true/checkpoints
I0424 17:04:22.003471 132939609249600 checkpoint_manager.py:929] [process=3][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/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_11_optimizer_offload_true/checkpoints: <orbax.checkpoint.checkpoint_manager.CheckpointManager object at 0x78d46f9eaf00>
I0424 17:04:22.003632 132939609249600 checkpointing.py:302] Checkpoint manager created!
I0424 17:04:22.945620 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='count'), GetAttrKey(name='value')) to host
I0424 17:04:22.945739 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0424 17:04:22.945779 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0424 17:04:22.945814 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0424 17:04:22.945845 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0424 17:04:22.945874 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0424 17:04:22.945901 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0424 17:04:22.945939 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0424 17:04:22.945965 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0424 17:04:22.945991 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0424 17:04:22.946015 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0424 17:04:22.946039 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0424 17:04:22.946062 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to host
I0424 17:04:22.946086 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0424 17:04:22.946110 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0424 17:04:22.946134 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0424 17:04:22.946157 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0424 17:04:22.946180 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0424 17:04:22.946204 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to host
I0424 17:04:22.946228 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0424 17:04:22.946252 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0424 17:04:22.946276 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0424 17:04:22.946299 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0424 17:04:22.946322 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to host
I0424 17:04:22.946346 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to host
I0424 17:04:22.946369 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=2), DictKey(key='count'), GetAttrKey(name='value')) to host
I0424 17:04:22.946391 132939609249600 maxtext_utils_nnx.py:112] max_utils.py: Moving (DictKey(key='step'), GetAttrKey(name='value')) to host
I0424 17:04:22.947467 132939609249600 checkpointing.py:578] checkpoint manager exists so trying to load this run's existing checkpoint
I0424 17:04:22.947529 132939609249600 checkpointing.py:676] No existing checkpoints found, not restoring checkpoint.
fsdp: 32
I0424 17:04:24.846721 132939609249600 maxtext_utils.py:1749]  decoder/decoder_norm/scale/value
    Shape:     float32[2048]
    Physical:  (None,)
I0424 17:04:24.846847 132939609249600 maxtext_utils.py:1749]  decoder/layers/mlp/wi_0/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0424 17:04:24.846904 132939609249600 maxtext_utils.py:1749]  decoder/layers/mlp/wi_1/kernel/value
    Shape:     float32[2048,16,7168]
    Physical:  ('fsdp', None, None)
I0424 17:04:24.846962 132939609249600 maxtext_utils.py:1749]  decoder/layers/mlp/wo/kernel/value
    Shape:     float32[7168,16,2048]
    Physical:  (None, None, 'fsdp')
I0424 17:04:24.847002 132939609249600 maxtext_utils.py:1749]  decoder/layers/post_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0424 17:04:24.847039 132939609249600 maxtext_utils.py:1749]  decoder/layers/pre_self_attention_layer_norm/scale/value
    Shape:     float32[2048,16]
    Physical:  (None, None)
I0424 17:04:24.847079 132939609249600 maxtext_utils.py:1749]  decoder/layers/self_attention/key/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0424 17:04:24.847117 132939609249600 maxtext_utils.py:1749]  decoder/layers/self_attention/out/kernel/value
    Shape:     float32[16,16,128,2048]
    Physical:  (None, None, None, 'fsdp')
I0424 17:04:24.847153 132939609249600 maxtext_utils.py:1749]  decoder/layers/self_attention/query/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0424 17:04:24.847187 132939609249600 maxtext_utils.py:1749]  decoder/layers/self_attention/value/kernel/value
    Shape:     float32[2048,16,16,128]
    Physical:  ('fsdp', None, None, None)
I0424 17:04:24.847220 132939609249600 maxtext_utils.py:1749]  decoder/logits_dense/kernel/value
    Shape:     float32[2048,32000]
    Physical:  ('fsdp', None)
I0424 17:04:24.847252 132939609249600 maxtext_utils.py:1749]  token_embedder/embedding/value
    Shape:     float32[32000,2048]
    Physical:  (None, 'fsdp')

I0424 17:04:25.247516 132939609249600 nnx_decoders.py:616] nnx_decoders/carry Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0424 17:04:25.247662 132939609249600 nnx_decoders.py:616] nnx_decoders/carry Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0424 17:04:25.253172 132939609249600 nnx_decoders.py:616] Unknown Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_norm_length', 'activation_embed').
I0424 17:04:25.253268 132939609249600 nnx_decoders.py:616] Unknown Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0424 17:04:25.404994 132939609249600 attentions.py:1084] attentions/inputs_q Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0424 17:04:25.405124 132939609249600 attentions.py:1084] attentions/inputs_q Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0424 17:04:25.555110 132939609249600 attentions.py:1085] attentions/inputs_kv Logical: bfloat16[32,2048,2048]...................................... ('activation_batch', 'activation_attn_length', 'activation_attn_embed').
I0424 17:04:25.555238 132939609249600 attentions.py:1085] attentions/inputs_kv Physical: bfloat16[32,2048,2048]...................................... ('fsdp', None, None).
I0424 17:04:25.711525 132939609249600 attentions.py:1150] attentions/query Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0424 17:04:25.711658 132939609249600 attentions.py:1150] attentions/query Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0424 17:04:25.858254 132939609249600 attentions.py:1151] attentions/key Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0424 17:04:25.858386 132939609249600 attentions.py:1151] attentions/key Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0424 17:04:25.997432 132939609249600 attentions.py:1152] attentions/value Logical: bfloat16[32,2048,16,128].................................... ('activation_kv_batch', 'activation_attn_length', 'activation_kv_heads', 'activation_kv_head_dim').
I0424 17:04:25.997560 132939609249600 attentions.py:1152] attentions/value Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0424 17:04:26.146721 132939609249600 attentions.py:1193] attentions/out Logical: bfloat16[32,2048,16,128].................................... ('activation_batch', 'activation_attn_length', 'activation_heads', 'activation_kv').
I0424 17:04:26.146854 132939609249600 attentions.py:1193] attentions/out Physical: bfloat16[32,2048,16,128].................................... ('fsdp', None, None, None).
I0424 17:04:26.169441 132939609249600 linears.py:541] linears/x Logical: bfloat16[32,2048,7168]...................................... ('activation_batch', 'activation_length', 'activation_mlp').
I0424 17:04:26.169536 132939609249600 linears.py:541] linears/x Physical: bfloat16[32,2048,7168]...................................... ('fsdp', None, None).
I0424 17:04:26.310052 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='count'), GetAttrKey(name='value')) to device
I0424 17:04:26.310161 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0424 17:04:26.310202 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:26.310238 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:26.310268 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:26.310296 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0424 17:04:26.310324 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0424 17:04:26.310350 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:26.310375 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:26.310400 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:26.310432 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:26.310459 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:26.310484 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to device
I0424 17:04:26.310517 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0424 17:04:26.310559 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:26.310598 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:26.310625 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:26.310649 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0424 17:04:26.310678 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0424 17:04:26.310702 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:26.310725 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:26.310748 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:26.310772 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:26.310797 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:26.310824 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to device
I0424 17:04:26.310848 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=2), DictKey(key='count'), GetAttrKey(name='value')) to device
I0424 17:04:26.310870 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='step'), GetAttrKey(name='value')) to device
I0424 17:04:37.160896 132939609249600 max_utils.py:791] Total memory size: 1.3 GB, Output size: 0.1 GB, Temp size: 1.1 GB, Argument size: 0.1 GB, Host temp size: 0.0 GB.
I0424 17:04:37.163095 132939609249600 metric_logger.py:289] number parameters: 1.104 billion
I0424 17:04:37.696656 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='count'), GetAttrKey(name='value')) to device
I0424 17:04:37.696778 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0424 17:04:37.696819 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:37.696859 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:37.696889 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:37.696918 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0424 17:04:37.696959 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0424 17:04:37.696985 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:37.697011 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:37.697038 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:37.697066 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:37.697091 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:37.697116 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='mu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to device
I0424 17:04:37.697143 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='decoder_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0424 17:04:37.697167 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_0'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:37.697191 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wi_1'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:37.697215 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='mlp'), DictKey(key='wo'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:37.697239 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='post_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0424 17:04:37.697262 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='pre_self_attention_layer_norm'), DictKey(key='scale'), GetAttrKey(name='value')) to device
I0424 17:04:37.697285 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='key'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:37.697308 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='out'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:37.697332 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='query'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:37.697355 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='layers'), DictKey(key='self_attention'), DictKey(key='value'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:37.697378 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='decoder'), DictKey(key='logits_dense'), DictKey(key='kernel'), GetAttrKey(name='value')) to device
I0424 17:04:37.697401 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=0), DictKey(key='nu'), DictKey(key='token_embedder'), DictKey(key='embedding'), GetAttrKey(name='value')) to device
I0424 17:04:37.697424 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='opt_state'), DictKey(key=2), DictKey(key='count'), GetAttrKey(name='value')) to device
I0424 17:04:37.697446 132939609249600 maxtext_utils_nnx.py:129] max_utils.py: Moving (DictKey(key='step'), GetAttrKey(name='value')) to device
I0424 17:04:48.234549 132939609249600 checkpointing.py:794] Waiting for step 0 to finish before checkpoint...
I0424 17:04:48.401765 132939609249600 checkpointing.py:798] Waited 0.1672039031982422 seconds for step 0 to finish before starting checkpointing.
I0424 17:04:48.404413 132939609249600 checkpoint_manager.py:2013] [process=3][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0424 17:04:48.406310 132939609249600 checkpoint_manager.py:1518] [process=3] Saving checkpoint at step 0
I0424 17:04:48.409576 132939609249600 async_checkpointer.py:452] [process=3] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_11_optimizer_offload_true/checkpoints/0.
I0424 17:04:49.027592 132939609249600 signaling_client.py:364] Using JaxDistributedSignalingClient
I0424 17:04:49.028703 132939609249600 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0424 17:04:49.028771 132939609249600 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 17:04:49.281780 132939609249600 base_pytree_checkpoint_handler.py:153] [process=3][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.255321s
I0424 17:04:49.281981 132939609249600 base_pytree_checkpoint_handler.py:129] [process=3] /jax/checkpoint/write/blocking_gbytes_per_sec: 2.833 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.544553279876709 s) (per-host)
I0424 17:04:49.282038 132939609249600 base_pytree_checkpoint_handler.py:737] [process=3][thread=MainThread] Initiated Pytree async_save. Time taken: 0.544621s (batch_requests_ready=0.281930s, total_serialization_initiated=0.262614s, others=0.000077s)
I0424 17:04:49.282142 132939609249600 composite_checkpoint_handler.py:715] [process=3][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.548768s (all_items=0.000024s, per_item={'items': '0.00002432'}, temp_paths=0.548743)
I0424 17:04:49.283077 132807183537920 async_checkpointer.py:79] [process=3][thread=async_save] Background save thread started.
I0424 17:04:49.283246 132939609249600 async_checkpointer.py:561] Finished blocking save. Time taken: 0.876863s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_11_optimizer_offload_true/checkpoints/0.
I0424 17:04:49.301800 132939609249600 checkpoint_manager.py:1566] [process=3][thread=MainThread][step=0] Starting CheckpointManager Save Finalize thread=save_finalize
I0424 17:04:49.302130 132805489063680 async_checkpointer.py:265] [process=3][thread=save_finalize] Waiting for background save thread=async_save.
I0424 17:04:49.302293 132939609249600 standard_logger.py:34] {'step': 0, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_11_optimizer_offload_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777050288.404395, 'wait_for_prev_duration_secs': 6.67572021484375e-05, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777050288.406351, 'checkpointer_blocking_duration_secs': 0.8770017623901367, 'get_old_steps_start_time': 1777050289.2833736, 'get_old_steps_duration_secs': 2.5033950805664062e-05, 'checkpoint_manager_blocking_start_time': 1777050288.402169, 'checkpoint_manager_blocking_duration_secs': 0.9000844955444336}
I0424 17:04:49.302403 132939609249600 checkpointing.py:409] Started an asynchronous checkpoint save for step 0
I0424 17:04:49.302482 132939609249600 max_utils.py:750] 
Memstats: After params initialized:
I0424 17:04:49.302537 132939609249600 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_10(process=3,(2,2,0,0))
I0424 17:04:49.302570 132939609249600 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_11(process=3,(3,2,0,0))
I0424 17:04:49.302599 132939609249600 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_14(process=3,(2,3,0,0))
I0424 17:04:49.302624 132939609249600 max_utils.py:756] 	Using (GB) 0.17 / 31.25 (0.544000%) on TPU_15(process=3,(3,3,0,0))
I0424 17:04:49.582847 132939609249600 metric_logger.py:185] completed step: 0, seconds: 11.070, TFLOP/s/device: 1.227, Tokens/s/device: 184.998, total_weights: 65536, loss: 10.874
I0424 17:04:49.752717 132939609249600 metric_logger.py:185] completed step: 1, seconds: 1.346, TFLOP/s/device: 10.092, Tokens/s/device: 1521.206, total_weights: 65536, loss: 10.874
I0424 17:04:50.170120 132939609249600 metric_logger.py:185] completed step: 2, seconds: 0.012, TFLOP/s/device: 1116.720, Tokens/s/device: 168324.156, total_weights: 65536, loss: 10.262
I0424 17:04:50.333705 132939609249600 metric_logger.py:185] completed step: 3, seconds: 0.404, TFLOP/s/device: 33.608, Tokens/s/device: 5065.708, total_weights: 65536, loss: 9.732
I0424 17:04:50.661400 132939609249600 metric_logger.py:185] completed step: 4, seconds: 0.179, TFLOP/s/device: 75.875, Tokens/s/device: 11436.741, total_weights: 65536, loss: 9.274
I0424 17:04:50.668615 132939609249600 metric_logger.py:185] completed step: 5, seconds: 0.163, TFLOP/s/device: 83.311, Tokens/s/device: 12557.484, total_weights: 65536, loss: 8.888
I0424 17:05:09.816468 132939609249600 metric_logger.py:185] completed step: 6, seconds: 0.328, TFLOP/s/device: 41.405, Tokens/s/device: 6241.086, total_weights: 65536, loss: 8.589
I0424 17:05:09.980163 132939609249600 metric_logger.py:185] completed step: 7, seconds: 18.985, TFLOP/s/device: 0.716, Tokens/s/device: 107.872, total_weights: 65536, loss: 8.380
I0424 17:05:10.143660 132939609249600 metric_logger.py:185] completed step: 8, seconds: 0.170, TFLOP/s/device: 80.106, Tokens/s/device: 12074.475, total_weights: 65536, loss: 8.251
I0424 17:05:10.308317 132939609249600 checkpointing.py:794] Waiting for step 9 to finish before checkpoint...
I0424 17:05:10.308919 132939609249600 checkpointing.py:798] Waited 0.0006139278411865234 seconds for step 9 to finish before starting checkpointing.
I0424 17:05:10.310952 132939609249600 checkpoint_manager.py:2024] [process=3][thread=MainThread][step=0][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0424 17:05:11.253314    2633 google_auth_provider.cc:181] Running on GCE, using service account 562977990677-compute@developer.gserviceaccount.com
I0424 17:05:13.208224 132807191930624 array_metadata_store.py:203] [process=3][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_11_optimizer_offload_true/checkpoints/0/items/array_metadatas/process_3
I0424 17:05:46.156740 132807183537920 base_pytree_checkpoint_handler.py:129] [process=3] /jax/checkpoint/write/gbytes_per_sec: 27.509 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 57.419265031814575 s) (per-host)
I0424 17:05:46.156859 132807183537920 async_checkpointer.py:90] [process=3][thread=async_save] 3 Handler Commit operations completed. Time taken: 56.873671s.
I0424 17:05:54.373109 132807183537920 async_checkpointer.py:144] [process=3][thread=async_save] Background save thread done. Time taken: 65.089905s.
I0424 17:05:54.373408 132805489063680 async_checkpointer.py:273] [process=3][thread=save_finalize] Done with waiting for background save thread=async_save.
I0424 17:05:54.373561 132805489063680 async_checkpointer.py:283] [process=3][thread=save_finalize] No errors found in background save thread=async_save.
I0424 17:05:54.373615 132805489063680 checkpoint_manager.py:2133] [process=3][thread=save_finalize][step=0] CheckpointManager Save Finalize is syncing with other hosts...
I0424 17:05:54.376429 132805489063680 checkpoint_manager.py:2142] [process=3][thread=save_finalize][step=0] CheckpointManager Save Finalize is done on all hosts.
I0424 17:05:54.376674 132939609249600 checkpoint_manager.py:2036] [process=3][thread=MainThread][step=0][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=0.
W0424 17:05:54.376803 132939609249600 checkpoint_manager.py:1458] Waiting for previous save to complete took 44.065874 seconds. If this number is high, consider checkpointing less frequently.
I0424 17:05:54.378801 132939609249600 checkpoint_manager.py:1518] [process=3] Saving checkpoint at step 9
I0424 17:05:54.382664 132939609249600 async_checkpointer.py:452] [process=3] Started async saving checkpoint to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_11_optimizer_offload_true/checkpoints/9.
I0424 17:05:54.908511 132939609249600 jax_array_handlers.py:358] Scheduling D2H of 69 prioritized jax.Array.
I0424 17:05:54.908667 132939609249600 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 17:05:54.954245 132939609249600 base_pytree_checkpoint_handler.py:153] [process=3][thread=MainThread] Initiated "orbax.checkpoint._src.serialization.jax_array_handlers.ArrayHandler".serialize. Time taken: 0.047966s
I0424 17:05:54.954421 132939609249600 base_pytree_checkpoint_handler.py:129] [process=3] /jax/checkpoint/write/blocking_gbytes_per_sec: 5.228 GiB/s (total gbytes: 1.5 GiB) (time elapsed: 0.2950711250305176 s) (per-host)
I0424 17:05:54.954474 132939609249600 base_pytree_checkpoint_handler.py:737] [process=3][thread=MainThread] Initiated Pytree async_save. Time taken: 0.295134s (batch_requests_ready=0.242227s, total_serialization_initiated=0.052835s, others=0.000072s)
I0424 17:05:54.954571 132939609249600 composite_checkpoint_handler.py:715] [process=3][thread=MainThread] Initiated CompositeCheckpointHandler.async_save. Time taken: 0.299238s (all_items=0.000015s, per_item={'items': '0.00001550'}, temp_paths=0.299222)
I0424 17:05:54.956708 132805489063680 async_checkpointer.py:79] [process=3][thread=async_save] Background save thread started.
I0424 17:05:54.956866 132939609249600 async_checkpointer.py:561] Finished blocking save. Time taken: 0.577994s. Continuing background save to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_11_optimizer_offload_true/checkpoints/9.
I0424 17:05:54.984874 132939609249600 checkpoint_manager.py:1566] [process=3][thread=MainThread][step=9] Starting CheckpointManager Save Finalize thread=save_finalize
I0424 17:05:54.985177 132807191930624 async_checkpointer.py:265] [process=3][thread=save_finalize] Waiting for background save thread=async_save.
I0424 17:05:54.985343 132939609249600 standard_logger.py:34] {'step': 9, 'event_type': 'save', 'directory': 'gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_11_optimizer_offload_true/checkpoints', 'reached_preemption': False, 'preemption_received_at': None, 'synchronous': False, 'wait_for_prev_start_time': 1777050310.3109055, 'wait_for_prev_duration_secs': 44.065874099731445, 'time_between_consecutive_saves_sec': None, 'checkpointer_blocking_start_time': 1777050354.3788414, 'checkpointer_blocking_duration_secs': 0.5781500339508057, 'get_old_steps_start_time': 1777050354.9570093, 'get_old_steps_duration_secs': 3.0517578125e-05, 'checkpoint_manager_blocking_start_time': 1777050310.3092256, 'checkpoint_manager_blocking_duration_secs': 44.676082372665405}
I0424 17:05:54.985451 132939609249600 checkpointing.py:409] Started an asynchronous checkpoint save for step 9
I0424 17:05:54.985509 132939609249600 checkpoint_manager.py:2024] [process=3][thread=MainThread][step=9][wait_until_finished] Waiting for Save Finalize thread (save_finalize) to complete.
I0424 17:05:59.770402 132809890748160 array_metadata_store.py:203] [process=3][thread=array_type_handler] Wrote 69 array_metadata.ArrayMetadata to gs://lance-maxtext/nnx_ckpt_xpk_feat_nnx_set_defaults_true_20260424_145744/nnx_xpk_feat_nnx_set_defaults_true_20260424_145744_11_optimizer_offload_true/checkpoints/9/items/array_metadatas/process_3
I0424 17:06:36.859695 132805489063680 base_pytree_checkpoint_handler.py:129] [process=3] /jax/checkpoint/write/gbytes_per_sec: 37.429 MiB/s (total gbytes: 1.5 GiB) (time elapsed: 42.2003173828125 s) (per-host)
I0424 17:06:36.859799 132805489063680 async_checkpointer.py:90] [process=3][thread=async_save] 3 Handler Commit operations completed. Time taken: 41.902980s.
I0424 17:06:45.819341 132805489063680 async_checkpointer.py:144] [process=3][thread=async_save] Background save thread done. Time taken: 50.862505s.
I0424 17:06:45.819658 132807191930624 async_checkpointer.py:273] [process=3][thread=save_finalize] Done with waiting for background save thread=async_save.
I0424 17:06:45.819782 132807191930624 async_checkpointer.py:283] [process=3][thread=save_finalize] No errors found in background save thread=async_save.
I0424 17:06:45.819831 132807191930624 checkpoint_manager.py:2133] [process=3][thread=save_finalize][step=9] CheckpointManager Save Finalize is syncing with other hosts...
I0424 17:06:45.821149 132807191930624 checkpoint_manager.py:2142] [process=3][thread=save_finalize][step=9] CheckpointManager Save Finalize is done on all hosts.
I0424 17:06:45.821331 132939609249600 checkpoint_manager.py:2036] [process=3][thread=MainThread][step=9][wait_until_finished] Done waiting for Save Finalize thread (save_finalize) running at step=9.
I0424 17:06:45.821511 132939609249600 checkpoint_manager.py:2013] [process=3][thread=MainThread][wait_until_finished] No Save Finalize thread to wait for. Returning.
I0424 17:06:45.822165 132939609249600 metric_logger.py:185] completed step: 9, seconds: 0.163, TFLOP/s/device: 83.296, Tokens/s/device: 12555.251, 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 24 17:06:57 UTC 2026
EXIT_CODE=0