Datasets:

---

arcadia-impact

/

scimt-dispatch-graft-dose-v1

like 0

Follow

Arcadia Impact 8

Error code:   StreamingRowsError
Exception:    CastError
Message:      Couldn't cast
best_global_step: null
best_metric: null
best_model_checkpoint: null
epoch: double
eval_steps: int64
global_step: int64
is_hyper_param_search: bool
is_local_process_zero: bool
is_world_process_zero: bool
log_history: list<item: struct<epoch: double, grad_norm: double, learning_rate: double, loss: double, memory/devi (... 206 chars omitted)
  child 0, item: struct<epoch: double, grad_norm: double, learning_rate: double, loss: double, memory/device_reserved (... 194 chars omitted)
      child 0, epoch: double
      child 1, grad_norm: double
      child 2, learning_rate: double
      child 3, loss: double
      child 4, memory/device_reserved (GiB): double
      child 5, memory/max_active (GiB): double
      child 6, memory/max_allocated (GiB): double
      child 7, ppl: double
      child 8, step: int64
      child 9, tokens/total: int64
      child 10, tokens/train_per_sec_per_gpu: double
      child 11, tokens/trainable: int64
logging_steps: int64
max_steps: int64
num_input_tokens_seen: int64
num_train_epochs: int64
save_steps: int64
stateful_callbacks: struct<TrainerControl: struct<args: struct<should_epoch_stop: bool, should_evaluate: bool, should_lo (... 79 chars omitted)
  child 0, TrainerControl: struct<args: struct<should_epoch_stop: bool, should_evaluate: bool, should_log: bool, should_save: b (... 55 chars omitted)
      child 0, args: struct<should_epoch_stop: bool, should_evaluate: bool, should_log: bool, should_save: bool, should_t (... 19 chars omitted)
          child 0, should_epoch_stop: bool
          child 1, should_evaluate: bool
          child 2, should_log: bool
          child 3, should_save: bool
          child 4, should_training_stop: bool
      child 1, attributes: struct<>
total_flos: double
train_batch_size: int64
trial_name: null
trial_params: null
retained_checkpoints: list<item: int64>
  child 0, item: int64
stage: string
dataset_sha256: string
seconds_per_step: double
version: string
minutes: double
label: string
adapter_payload_audit: struct<tensor_count: int64, lora_b_tensor_count: int64, nonzero_lora_b: bool, vision_tensor_count: i (... 34 chars omitted)
  child 0, tensor_count: int64
  child 1, lora_b_tensor_count: int64
  child 2, nonzero_lora_b: bool
  child 3, vision_tensor_count: int64
  child 4, audited_exact_targets: null
gpus: int64
phase: string
seed: int64
lora: struct<r: int64, alpha: int64, dropout: double, target_linear: bool, target_modules: list<item: stri (... 79 chars omitted)
  child 0, r: int64
  child 1, alpha: int64
  child 2, dropout: double
  child 3, target_linear: bool
  child 4, target_modules: list<item: string>
      child 0, item: string
  child 5, target_parameters: null
  child 6, triton_kernels: bool
  child 7, initial_adapter_path: null
schema_version: string
optimizer_steps: null
completed_at: timestamp[s]
required_checkpoints: list<item: int64>
  child 0, item: int64
to
{'schema_version': Value('string'), 'version': Value('string'), 'label': Value('string'), 'phase': Value('string'), 'stage': Value('string'), 'gpus': Value('int64'), 'seed': Value('int64'), 'dataset_sha256': Value('string'), 'lora': {'r': Value('int64'), 'alpha': Value('int64'), 'dropout': Value('float64'), 'target_linear': Value('bool'), 'target_modules': List(Value('string')), 'target_parameters': Value('null'), 'triton_kernels': Value('bool'), 'initial_adapter_path': Value('null')}, 'global_step': Value('int64'), 'optimizer_steps': Value('null'), 'retained_checkpoints': List(Value('int64')), 'required_checkpoints': List(Value('int64')), 'adapter_payload_audit': {'tensor_count': Value('int64'), 'lora_b_tensor_count': Value('int64'), 'nonzero_lora_b': Value('bool'), 'vision_tensor_count': Value('int64'), 'audited_exact_targets': Value('null')}, 'minutes': Value('float64'), 'seconds_per_step': Value('float64'), 'completed_at': Value('timestamp[s]')}
because column names don't match
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/src/worker/utils.py", line 147, in get_rows_or_raise
                  return get_rows(
                      dataset=dataset,
                  ...<4 lines>...
                      column_names=column_names,
                  )
                File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
                  return func(*args, **kwargs)
                File "/src/services/worker/src/worker/utils.py", line 127, in get_rows
                  rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
                File "/src/services/worker/src/worker/utils.py", line 483, in safe_iter
                  yield from ds.decode(False) if ds.features else ds
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2840, in __iter__
                  for key, example in ex_iterable:
                                      ^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2373, in __iter__
                  for key, pa_table in self._iter_arrow():
                                       ~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2398, in _iter_arrow
                  for key, pa_table in self.ex_iterable._iter_arrow():
                                       ~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 536, in _iter_arrow
                  for key, pa_table in iterator:
                                       ^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 419, in _iter_arrow
                  for key, pa_table in self.generate_tables_fn(**gen_kwags):
                                       ~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 343, in _generate_tables
                  self._cast_table(pa_table, json_field_paths=json_field_paths),
                  ~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
                  pa_table = table_cast(pa_table, self.info.features.arrow_schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2378, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2306, in cast_table_to_schema
                  raise CastError(
                  ...<3 lines>...
                  )
              datasets.table.CastError: Couldn't cast
              best_global_step: null
              best_metric: null
              best_model_checkpoint: null
              epoch: double
              eval_steps: int64
              global_step: int64
              is_hyper_param_search: bool
              is_local_process_zero: bool
              is_world_process_zero: bool
              log_history: list<item: struct<epoch: double, grad_norm: double, learning_rate: double, loss: double, memory/devi (... 206 chars omitted)
                child 0, item: struct<epoch: double, grad_norm: double, learning_rate: double, loss: double, memory/device_reserved (... 194 chars omitted)
                    child 0, epoch: double
                    child 1, grad_norm: double
                    child 2, learning_rate: double
                    child 3, loss: double
                    child 4, memory/device_reserved (GiB): double
                    child 5, memory/max_active (GiB): double
                    child 6, memory/max_allocated (GiB): double
                    child 7, ppl: double
                    child 8, step: int64
                    child 9, tokens/total: int64
                    child 10, tokens/train_per_sec_per_gpu: double
                    child 11, tokens/trainable: int64
              logging_steps: int64
              max_steps: int64
              num_input_tokens_seen: int64
              num_train_epochs: int64
              save_steps: int64
              stateful_callbacks: struct<TrainerControl: struct<args: struct<should_epoch_stop: bool, should_evaluate: bool, should_lo (... 79 chars omitted)
                child 0, TrainerControl: struct<args: struct<should_epoch_stop: bool, should_evaluate: bool, should_log: bool, should_save: b (... 55 chars omitted)
                    child 0, args: struct<should_epoch_stop: bool, should_evaluate: bool, should_log: bool, should_save: bool, should_t (... 19 chars omitted)
                        child 0, should_epoch_stop: bool
                        child 1, should_evaluate: bool
                        child 2, should_log: bool
                        child 3, should_save: bool
                        child 4, should_training_stop: bool
                    child 1, attributes: struct<>
              total_flos: double
              train_batch_size: int64
              trial_name: null
              trial_params: null
              retained_checkpoints: list<item: int64>
                child 0, item: int64
              stage: string
              dataset_sha256: string
              seconds_per_step: double
              version: string
              minutes: double
              label: string
              adapter_payload_audit: struct<tensor_count: int64, lora_b_tensor_count: int64, nonzero_lora_b: bool, vision_tensor_count: i (... 34 chars omitted)
                child 0, tensor_count: int64
                child 1, lora_b_tensor_count: int64
                child 2, nonzero_lora_b: bool
                child 3, vision_tensor_count: int64
                child 4, audited_exact_targets: null
              gpus: int64
              phase: string
              seed: int64
              lora: struct<r: int64, alpha: int64, dropout: double, target_linear: bool, target_modules: list<item: stri (... 79 chars omitted)
                child 0, r: int64
                child 1, alpha: int64
                child 2, dropout: double
                child 3, target_linear: bool
                child 4, target_modules: list<item: string>
                    child 0, item: string
                child 5, target_parameters: null
                child 6, triton_kernels: bool
                child 7, initial_adapter_path: null
              schema_version: string
              optimizer_steps: null
              completed_at: timestamp[s]
              required_checkpoints: list<item: int64>
                child 0, item: int64
              to
              {'schema_version': Value('string'), 'version': Value('string'), 'label': Value('string'), 'phase': Value('string'), 'stage': Value('string'), 'gpus': Value('int64'), 'seed': Value('int64'), 'dataset_sha256': Value('string'), 'lora': {'r': Value('int64'), 'alpha': Value('int64'), 'dropout': Value('float64'), 'target_linear': Value('bool'), 'target_modules': List(Value('string')), 'target_parameters': Value('null'), 'triton_kernels': Value('bool'), 'initial_adapter_path': Value('null')}, 'global_step': Value('int64'), 'optimizer_steps': Value('null'), 'retained_checkpoints': List(Value('int64')), 'required_checkpoints': List(Value('int64')), 'adapter_payload_audit': {'tensor_count': Value('int64'), 'lora_b_tensor_count': Value('int64'), 'nonzero_lora_b': Value('bool'), 'vision_tensor_count': Value('int64'), 'audited_exact_targets': Value('null')}, 'minutes': Value('float64'), 'seconds_per_step': Value('float64'), 'completed_at': Value('timestamp[s]')}
              because column names don't match