Dataset Preview
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed
Error code: DatasetGenerationError
Exception: ArrowNotImplementedError
Message: Cannot write struct type 'A_log_roundtrip_max_abs' with no child field to Parquet. Consider adding a dummy child field.
Traceback: Traceback (most recent call last):
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1868, in _prepare_split_single
num_examples, num_bytes = writer.finalize()
~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 803, in finalize
self.write_rows_on_file()
~~~~~~~~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 663, in write_rows_on_file
self._write_table(table)
~~~~~~~~~~~~~~~~~^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 790, in _write_table
self._build_writer(inferred_schema=pa_table.schema)
~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 834, in _build_writer
self.pa_writer = pq.ParquetWriter(
~~~~~~~~~~~~~~~~^
self.stream,
^^^^^^^^^^^^
...<9 lines>...
},
^^
)
^
File "/usr/local/lib/python3.14/site-packages/pyarrow/parquet/core.py", line 1115, in __init__
self.writer = _parquet.ParquetWriter(
~~~~~~~~~~~~~~~~~~~~~~^
sink, schema,
^^^^^^^^^^^^^
...<20 lines>...
max_rows_per_page=max_rows_per_page,
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
**options)
^^^^^^^^^^
File "pyarrow/_parquet.pyx", line 2430, in pyarrow._parquet.ParquetWriter.__cinit__
File "pyarrow/error.pxi", line 155, in pyarrow.lib.pyarrow_internal_check_status
return check_status(status)
File "pyarrow/error.pxi", line 92, in pyarrow.lib.check_status
raise convert_status(status)
pyarrow.lib.ArrowNotImplementedError: Cannot write struct type 'A_log_roundtrip_max_abs' with no child field to Parquet. Consider adding a dummy child field.
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1369, in compute_config_parquet_and_info_response
parquet_operations, partial, estimated_dataset_info = stream_convert_to_parquet(
~~~~~~~~~~~~~~~~~~~~~~~~~^
builder, max_dataset_size_bytes=max_dataset_size_bytes
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
)
^
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 948, in stream_convert_to_parquet
builder._prepare_split(split_generator=splits_generators[split], file_format="parquet")
~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1694, in _prepare_split
for job_id, done, content in self._prepare_split_single(
~~~~~~~~~~~~~~~~~~~~~~~~~~^
gen_kwargs=gen_kwargs, job_id=job_id, **_prepare_split_args
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
):
^
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1879, in _prepare_split_single
raise DatasetGenerationError("An error occurred while generating the dataset") from e
datasets.exceptions.DatasetGenerationError: An error occurred while generating the datasetNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
source_repo string | source_revision string | source_gguf string | template_repo string | template_revision string | source_format string | source_gguf_sha256 string | template string | template_tensors_checked int64 | all_tensors_bit_exact bool | A_log_roundtrip_max_abs dict | tokenizer_ids_match bool | bpe_merge_ranks_match bool | validation_date timestamp[s] | output_safetensors_sha256 string | reference_safetensors_sha256 string | complete_safetensors_byte_identity bool | output_safetensors_bytes int64 | total_output_tensors int64 | loading_configs_semantically_identical bool | scope string | native_cli_check dict |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
OS-Software/Ternary-Bonsai-2-27B-Uncensored-Heretic-v2-GGUF | 6db27547a8f25667cdb43a28d15cc5d33937b7dc | Ternary-Bonsai-2-27B-Uncensored-Heretic-v2-PTQ1_0.gguf | 3hs4n/Ternary-Bonsai-2-27B-Uncensored-Heretic-v2-mlx-2bit | df26fb79a9ab5bc5af71a1e2a6b81decb33378dd | PTQ1_0 | c30f3862e4809db58366aab0b97a50e0a144dd010061d931a61c000adaf50683 | 3hs4n/Ternary-Bonsai-2-27B-Uncensored-Heretic-v2-mlx-2bit | 2,057 | true | {} | true | true | 2026-10-09T00:00:00 | e56ba66424066818c7c89f0f1ba74111705b9062367905228736013a48aaf08d | e56ba66424066818c7c89f0f1ba74111705b9062367905228736013a48aaf08d | true | 8,595,478,392 | 2,390 | true | All 2057 language tensors checked individually; complete 2390-tensor Safetensors file matches the published PQ2_0-derived file SHA256; vision reused unchanged; no model-quality benchmark. | {
"exit_code": 0,
"stdout": "Paris\n",
"max_tokens": 64,
"temperature": 0,
"thinking": "disabled by native CLI default"
} |
prism-ml/Ternary-Bonsai-2-27B-gguf | b072e1d3b35a0a630cece372c2127528e0994386 | Ternary-Bonsai-2-27B-PQ2_0.gguf | null | fcba37d2117a7077eac6b613b2668d14d9779edd | null | 3907dc1658db1f78a9826bf8d5bcb8dc65db0d466388937af57f2294fae62ec1 | prism-ml/Ternary-Bonsai-2-27B-mlx-2bit | 2,057 | true | {} | true | true | null | null | null | null | null | null | null | Converter validation against the matching official base GGUF/MLX pair; this is not a comparison asserting Heretic weights equal base-model weights. | null |
NInfer Bonsai 2 27B Heretic: artifact and sources
The converted .ninfer artifact for Ternary Bonsai 2 27B Heretic v2, with every source file the
conversion read, so the artifact can be rebuilt or served without the upstream repositories.
What it is for
models/holds the artifact that the V-Sekai-fire/ninfer-all CUDA engine and its Apple silicon MLX loader (v-sekai/mlx/) serve, with the converter'sconversion.jsonreport.sources/holds the converter inputs. Their upstream repositories are fetched without a pinned revision, so this copy is the fixed input set.
Provenance
One folder per upstream repository under sources/, named as v-sekai/fetch-sources.nu names it:
Heretic-v2-GGUF: OS-Software/Ternary-Bonsai-2-27B-Uncensored-Heretic-v2-GGUF (PQ2_0 GGUF).Qwen3.8-27B: Qwen/Qwen3.8-27B configs, tokenizer and shard 1 (the vision tower).Ternary-Bonsai-2-27B-MTP,Ternary-Bonsai-2-27B-DFlash2: ProCreations drafter heads.
The independent MLX conversion that v-sekai/mlx/parity_reference.py compares against
(3hs4n/Ternary-Bonsai-2-27B-Uncensored-Heretic-v2-mlx-2bit) is not archived here; download it
from that public repository when needed.
models/ is the output of v-sekai/convert.nu (recipe bonsai2_27b_ternary) on these sources.
Use
From a ninfer-all checkout:
hf download chibifire/ninfer-bonsai2-heretic --repo-type dataset --local-dir .
Add a folder with hf upload chibifire/ninfer-bonsai2-heretic <dir> <dir> --repo-type dataset from
the checkout root, keeping the same layout.
Licence
Each sources/ folder keeps its upstream licence files and terms. The artifact is derived from
them. The dataset is private because it redistributes third-party weights.
- Downloads last month
- 325