Datasets:

Atlas3D
/

deepseek-v41-flash-fleet-notes

Tags:

License:

Dataset card Files Files and versions

xet

Community

Dataset Viewer

Duplicate

Split (1)

train

The dataset viewer is not available for this split.

Cannot load the dataset split (in streaming mode) to extract the first rows.

Error code:   StreamingRowsError
Exception:    CastError
Message:      Couldn't cast
id: string
group: string
prompt: string
reasoning_content: string
wall_s: double
usage: struct<completion_tokens: int64, prompt_tokens: int64, total_tokens: int64, prompt_tokens_details: s (... 28 chars omitted)
  child 0, completion_tokens: int64
  child 1, prompt_tokens: int64
  child 2, total_tokens: int64
  child 3, prompt_tokens_details: struct<cached_tokens: int64>
      child 0, cached_tokens: int64
content: string
label: string
finish_reason: string
timings: struct<cache_n: int64, prompt_n: int64, prompt_ms: double, prompt_per_token_ms: double, prompt_per_s (... 118 chars omitted)
  child 0, cache_n: int64
  child 1, prompt_n: int64
  child 2, prompt_ms: double
  child 3, prompt_per_token_ms: double
  child 4, prompt_per_second: double
  child 5, predicted_n: int64
  child 6, predicted_ms: double
  child 7, predicted_per_token_ms: double
  child 8, predicted_per_second: double
to
{'id': Value('string'), 'group': Value('string'), 'label': Value('string'), 'content': Value('string'), 'reasoning_content': Value('string'), 'finish_reason': Value('string'), 'timings': {'cache_n': Value('int64'), 'prompt_n': Value('int64'), 'prompt_ms': Value('float64'), 'prompt_per_token_ms': Value('float64'), 'prompt_per_second': Value('float64'), 'predicted_n': Value('int64'), 'predicted_ms': Value('float64'), 'predicted_per_token_ms': Value('float64'), 'predicted_per_second': Value('float64')}, 'wall_s': Value('float64'), 'usage': {'completion_tokens': Value('int64'), 'prompt_tokens': Value('int64'), 'total_tokens': Value('int64'), 'prompt_tokens_details': {'cached_tokens': Value('int64')}}}
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 2951, in __iter__
                  for key, example in ex_iterable:
                                      ^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2461, in __iter__
                  for key, pa_table in self._iter_arrow():
                                       ~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2486, 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 547, in _iter_arrow
                  for key, pa_table in iterator:
                                       ^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 430, 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
              id: string
              group: string
              prompt: string
              reasoning_content: string
              wall_s: double
              usage: struct<completion_tokens: int64, prompt_tokens: int64, total_tokens: int64, prompt_tokens_details: s (... 28 chars omitted)
                child 0, completion_tokens: int64
                child 1, prompt_tokens: int64
                child 2, total_tokens: int64
                child 3, prompt_tokens_details: struct<cached_tokens: int64>
                    child 0, cached_tokens: int64
              content: string
              label: string
              finish_reason: string
              timings: struct<cache_n: int64, prompt_n: int64, prompt_ms: double, prompt_per_token_ms: double, prompt_per_s (... 118 chars omitted)
                child 0, cache_n: int64
                child 1, prompt_n: int64
                child 2, prompt_ms: double
                child 3, prompt_per_token_ms: double
                child 4, prompt_per_second: double
                child 5, predicted_n: int64
                child 6, predicted_ms: double
                child 7, predicted_per_token_ms: double
                child 8, predicted_per_second: double
              to
              {'id': Value('string'), 'group': Value('string'), 'label': Value('string'), 'content': Value('string'), 'reasoning_content': Value('string'), 'finish_reason': Value('string'), 'timings': {'cache_n': Value('int64'), 'prompt_n': Value('int64'), 'prompt_ms': Value('float64'), 'prompt_per_token_ms': Value('float64'), 'prompt_per_second': Value('float64'), 'predicted_n': Value('int64'), 'predicted_ms': Value('float64'), 'predicted_per_token_ms': Value('float64'), 'predicted_per_second': Value('float64')}, 'wall_s': Value('float64'), 'usage': {'completion_tokens': Value('int64'), 'prompt_tokens': Value('int64'), 'total_tokens': Value('int64'), 'prompt_tokens_details': {'cached_tokens': Value('int64')}}}
              because column names don't match

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

DeepSeek-V4.1-Flash on a heterogeneous llama.cpp fleet — notes and patches

Working notes from serving DeepSeek-V4.1-Flash-1017 at its native Q8_0/MXFP4 precision across four machines (Blackwell + Ampere + three V100s + a 4090, llama.cpp RPC), 2026-10-17/18. Shared for reference, not as polished PRs.

FINDINGS-kv-sharing.md — how V4.1 shares its compressed attention cache across layer groups (owners 2, 8, 14, 20), the index-key and top-k sharing maps, how we found it, and what it means for multi-GPU serving and for interpretability work.
patches/ — against ggml-org/llama.cpp PR #28696 head 2076412: dsv41-kv-replicas.patch (per-device replicas of reused KV layers: 3.9 -> 9.5 tok/s on our fleet), dsv41-rpc-graph-cache.patch (per-uid RPC graph cache), dsv41-rpc-cache-verify.patch (rpc tensor-cache integrity), dsv41-mxfp4-e8m0-vector-scale.patch (in-register E8M0 scale conversion in the x86 MXFP4 kernels, +30-39 % on AVX2). The full branch, including cherry-picks from cliecy/llama.cpp with attribution, is dsv41-fleet on our llama.cpp fork.
quality/ — the 30-prompt set (polymer chemistry, lab technique, coding) and the collection harness we use to compare quants blind.
BRANCH-NOTES.md — what the branch carries and who to credit.

Headline numbers on this fleet, same model file throughout: 1.3 tok/s (naive layer split) -> 3.9 (device boundaries aligned to the KV-sharing groups) -> 9.5 (per-device replicas) -> 9.3-10 (kernel + decode fixes). Everything outside VRAM (15 expert layers in RAM) is what remains; six V100s arriving this week should remove it.

Mirror: https://huggingface.co/datasets/Atlas3D/deepseek-v41-flash-fleet-notes

Credits: Atlas3DSS (owner, direction, hardware); Claude (Anthropic) did the measurement, engine work and writing in this shop; Astra (OpenAI Codex) reviewed every patch; cliecy, nicholasshirley, smalinin, Mermiges and the PR #28696 authors for the pieces we built on. MIT, like llama.cpp.

Added 2026-10-18 (evening)

FINDINGS-depth-loop.md: running layers 30-35 a second time at inference, with controls; free in decode time on our layout, outputs diverge, quality scoring in progress. Patch: patches/dsv41-depth-loop.patch; raw sample answers: quality/depth-loop-sample/.
FINDINGS-serving-economics.md: thinking vs best-of-N wall time at ~9 tok/s, why parallel decode does not scale with CPU-resident experts, per-request reasoning caps, a one-token logprob judge on a side GPU, and gated thinking with a seeded think block (quality/gated_think.py, quality/rejudge.py).

Downloads last month: -

Total file size:

153 kB