Hugging Face's logo

Datasets:
pumaproject
/
puma-community-submissions

Languages:
English
Size:
n<1K
Tags:
benchmark
llm-evaluation
project-management
local-llm
sustainability
issue-triage
License:
Dataset card Files
xet
 Community
1
Dataset Viewer
Duplicate
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
hardware_profile: struct<cpu_cores: int64, cpu_model: string, gpu_model: string, gpu_vram_gb: int64, os: string, profi (... 29 chars omitted)
  child 0, cpu_cores: int64
  child 1, cpu_model: string
  child 2, gpu_model: string
  child 3, gpu_vram_gb: int64
  child 4, os: string
  child 5, profile_id: string
  child 6, ram_gb: int64
integrity: struct<payload_signature: null, predictions_summary_hash: string, verification_status: string>
  child 0, payload_signature: null
  child 1, predictions_summary_hash: string
  child 2, verification_status: string
metrics: struct<accuracy: double, confusion_matrix: null, ece: null, f1_macro: double, f1_per_class: null, ma (... 20 chars omitted)
  child 0, accuracy: double
  child 1, confusion_matrix: null
  child 2, ece: null
  child 3, f1_macro: double
  child 4, f1_per_class: null
  child 5, mae: null
  child 6, mdae: null
notes: null
puma_version: string
raw_predictions_url: null
run_metadata: struct<completed_at: string, latency_ms_p50: int64, latency_ms_p95: int64, latency_ms_total: int64,  (... 148 chars omitted)
  child 0, completed_at: string
  child 1, latency_ms_p50: int64
  child 2, latency_ms_p95: int64
  child 3, latency_ms_total: int64
  child 4, model: string
  child 5, n_instances: int64
  child 6, ollama_version: string
  child 7, scenario: string
  child 8, seed: int64
  child 9, started_at: string
  child 10, strategy: string
  child 11, temperature: double
schema_version: string
submission_id: string
submitted_at: str
...
l
      child 1, description: string
      child 2, properties: struct<codecarbon_version: struct<maxLength: int64, title: string, type: string>, co2_grams_total: s (... 305 chars omitted)
          child 0, codecarbon_version: struct<maxLength: int64, title: string, type: string>
              child 0, maxLength: int64
              child 1, title: string
              child 2, type: string
          child 1, co2_grams_total: struct<minimum: double, title: string, type: string>
              child 0, minimum: double
              child 1, title: string
              child 2, type: string
          child 2, energy_kwh_total: struct<minimum: double, title: string, type: string>
              child 0, minimum: double
              child 1, title: string
              child 2, type: string
          child 3, tracking_mode: struct<enum: list<item: string>, title: string, type: string>
              child 0, enum: list<item: string>
                  child 0, item: string
              child 1, title: string
              child 2, type: string
          child 4, country_iso: struct<maxLength: int64, minLength: int64, pattern: string, title: string, type: string>
              child 0, maxLength: int64
              child 1, minLength: int64
              child 2, pattern: string
              child 3, title: string
              child 4, type: string
      child 3, required: list<item: string>
          child 0, item: string
      child 4, title: string
      child 5, type: string
to
{'$schema': Value('string'), '$id': Value('string'), '$defs': {'HardwareProfile': {'additionalProperties': Value('bool'), 'description': Value('string'), 'properties': {'profile_id': {'maxLength': Value('int64'), 'title': Value('string'), 'type': Value('string')}, 'cpu_model': {'maxLength': Value('int64'), 'title': Value('string'), 'type': Value('string')}, 'cpu_cores': {'maximum': Value('int64'), 'minimum': Value('int64'), 'title': Value('string'), 'type': Value('string')}, 'ram_gb': {'maximum': Value('int64'), 'minimum': Value('int64'), 'title': Value('string'), 'type': Value('string')}, 'gpu_model': {'anyOf': List({'maxLength': Value('int64'), 'type': Value('string')}), 'default': Value('null'), 'title': Value('string')}, 'gpu_vram_gb': {'anyOf': List({'maximum': Value('int64'), 'minimum': Value('int64'), 'type': Value('string')}), 'default': Value('null'), 'title': Value('string')}, 'os': {'maxLength': Value('int64'), 'title': Value('string'), 'type': Value('string')}}, 'required': List(Value('string')), 'title': Value('string'), 'type': Value('string')}, 'Integrity': {'additionalProperties': Value('bool'), 'description': Value('string'), 'properties': {'predictions_summary_hash': {'pattern': Value('string'), 'title': Value('string'), 'type': Value('string')}, 'payload_signature': {'anyOf': List({'maxLength': Value('int64'), 'type': Value('string')}), 'default': Value('null'), 'title': Value('string')}, 'verification_status': {'default': Value('string'), 'enum': List(Valu
...
m': List(Value('string')), 'title': Value('string'), 'type': Value('string')}, 'country_iso': {'maxLength': Value('int64'), 'minLength': Value('int64'), 'pattern': Value('string'), 'title': Value('string'), 'type': Value('string')}}, 'required': List(Value('string')), 'title': Value('string'), 'type': Value('string')}}, 'additionalProperties': Value('bool'), 'description': Value('string'), 'properties': {'schema_version': {'const': Value('string'), 'default': Value('string'), 'title': Value('string'), 'type': Value('string')}, 'submission_id': {'format': Value('string'), 'title': Value('string'), 'type': Value('string')}, 'submitted_at': {'format': Value('string'), 'title': Value('string'), 'type': Value('string')}, 'submitter': {'$ref': Value('string')}, 'puma_version': {'pattern': Value('string'), 'title': Value('string'), 'type': Value('string')}, 'run_metadata': {'$ref': Value('string')}, 'hardware_profile': {'$ref': Value('string')}, 'metrics': {'$ref': Value('string')}, 'sustainability': {'$ref': Value('string')}, 'integrity': {'$ref': Value('string')}, 'raw_predictions_url': {'anyOf': List({'format': Value('string'), 'maxLength': Value('int64'), 'minLength': Value('int64'), 'type': Value('string')}), 'default': Value('null'), 'title': Value('string')}, 'notes': {'anyOf': List({'maxLength': Value('int64'), 'type': Value('string')}), 'default': Value('null'), 'title': Value('string')}}, 'required': List(Value('string')), 'title': Value('string'), 'type': Value('string')}
because column names don't match
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/src/worker/utils.py", line 99, in get_rows_or_raise
                  return get_rows(
                         ^^^^^^^^^
                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 77, in get_rows
                  rows_plus_one = list(itertools.islice(ds, rows_max_number + 1))
                                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 2690, in __iter__
                  for key, example in ex_iterable:
                                      ^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 2227, in __iter__
                  for key, pa_table in self._iter_arrow():
                                       ^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 2251, in _iter_arrow
                  for key, pa_table in self.ex_iterable._iter_arrow():
                                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 494, in _iter_arrow
                  for key, pa_table in iterator:
                                       ^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 384, in _iter_arrow
                  for key, pa_table in self.generate_tables_fn(**gen_kwags):
                                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/json/json.py", line 299, in _generate_tables
                  self._cast_table(pa_table, json_field_paths=json_field_paths),
                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/json/json.py", line 128, in _cast_table
                  pa_table = table_cast(pa_table, self.info.features.arrow_schema)
                             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2321, in table_cast
                  return cast_table_to_schema(table, schema)
                         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2249, in cast_table_to_schema
                  raise CastError(
              datasets.table.CastError: Couldn't cast
              hardware_profile: struct<cpu_cores: int64, cpu_model: string, gpu_model: string, gpu_vram_gb: int64, os: string, profi (... 29 chars omitted)
                child 0, cpu_cores: int64
                child 1, cpu_model: string
                child 2, gpu_model: string
                child 3, gpu_vram_gb: int64
                child 4, os: string
                child 5, profile_id: string
                child 6, ram_gb: int64
              integrity: struct<payload_signature: null, predictions_summary_hash: string, verification_status: string>
                child 0, payload_signature: null
                child 1, predictions_summary_hash: string
                child 2, verification_status: string
              metrics: struct<accuracy: double, confusion_matrix: null, ece: null, f1_macro: double, f1_per_class: null, ma (... 20 chars omitted)
                child 0, accuracy: double
                child 1, confusion_matrix: null
                child 2, ece: null
                child 3, f1_macro: double
                child 4, f1_per_class: null
                child 5, mae: null
                child 6, mdae: null
              notes: null
              puma_version: string
              raw_predictions_url: null
              run_metadata: struct<completed_at: string, latency_ms_p50: int64, latency_ms_p95: int64, latency_ms_total: int64,  (... 148 chars omitted)
                child 0, completed_at: string
                child 1, latency_ms_p50: int64
                child 2, latency_ms_p95: int64
                child 3, latency_ms_total: int64
                child 4, model: string
                child 5, n_instances: int64
                child 6, ollama_version: string
                child 7, scenario: string
                child 8, seed: int64
                child 9, started_at: string
                child 10, strategy: string
                child 11, temperature: double
              schema_version: string
              submission_id: string
              submitted_at: str
              ...
              l
                    child 1, description: string
                    child 2, properties: struct<codecarbon_version: struct<maxLength: int64, title: string, type: string>, co2_grams_total: s (... 305 chars omitted)
                        child 0, codecarbon_version: struct<maxLength: int64, title: string, type: string>
                            child 0, maxLength: int64
                            child 1, title: string
                            child 2, type: string
                        child 1, co2_grams_total: struct<minimum: double, title: string, type: string>
                            child 0, minimum: double
                            child 1, title: string
                            child 2, type: string
                        child 2, energy_kwh_total: struct<minimum: double, title: string, type: string>
                            child 0, minimum: double
                            child 1, title: string
                            child 2, type: string
                        child 3, tracking_mode: struct<enum: list<item: string>, title: string, type: string>
                            child 0, enum: list<item: string>
                                child 0, item: string
                            child 1, title: string
                            child 2, type: string
                        child 4, country_iso: struct<maxLength: int64, minLength: int64, pattern: string, title: string, type: string>
                            child 0, maxLength: int64
                            child 1, minLength: int64
                            child 2, pattern: string
                            child 3, title: string
                            child 4, type: string
                    child 3, required: list<item: string>
                        child 0, item: string
                    child 4, title: string
                    child 5, type: string
              to
              {'$schema': Value('string'), '$id': Value('string'), '$defs': {'HardwareProfile': {'additionalProperties': Value('bool'), 'description': Value('string'), 'properties': {'profile_id': {'maxLength': Value('int64'), 'title': Value('string'), 'type': Value('string')}, 'cpu_model': {'maxLength': Value('int64'), 'title': Value('string'), 'type': Value('string')}, 'cpu_cores': {'maximum': Value('int64'), 'minimum': Value('int64'), 'title': Value('string'), 'type': Value('string')}, 'ram_gb': {'maximum': Value('int64'), 'minimum': Value('int64'), 'title': Value('string'), 'type': Value('string')}, 'gpu_model': {'anyOf': List({'maxLength': Value('int64'), 'type': Value('string')}), 'default': Value('null'), 'title': Value('string')}, 'gpu_vram_gb': {'anyOf': List({'maximum': Value('int64'), 'minimum': Value('int64'), 'type': Value('string')}), 'default': Value('null'), 'title': Value('string')}, 'os': {'maxLength': Value('int64'), 'title': Value('string'), 'type': Value('string')}}, 'required': List(Value('string')), 'title': Value('string'), 'type': Value('string')}, 'Integrity': {'additionalProperties': Value('bool'), 'description': Value('string'), 'properties': {'predictions_summary_hash': {'pattern': Value('string'), 'title': Value('string'), 'type': Value('string')}, 'payload_signature': {'anyOf': List({'maxLength': Value('int64'), 'type': Value('string')}), 'default': Value('null'), 'title': Value('string')}, 'verification_status': {'default': Value('string'), 'enum': List(Valu
              ...
              m': List(Value('string')), 'title': Value('string'), 'type': Value('string')}, 'country_iso': {'maxLength': Value('int64'), 'minLength': Value('int64'), 'pattern': Value('string'), 'title': Value('string'), 'type': Value('string')}}, 'required': List(Value('string')), 'title': Value('string'), 'type': Value('string')}}, 'additionalProperties': Value('bool'), 'description': Value('string'), 'properties': {'schema_version': {'const': Value('string'), 'default': Value('string'), 'title': Value('string'), 'type': Value('string')}, 'submission_id': {'format': Value('string'), 'title': Value('string'), 'type': Value('string')}, 'submitted_at': {'format': Value('string'), 'title': Value('string'), 'type': Value('string')}, 'submitter': {'$ref': Value('string')}, 'puma_version': {'pattern': Value('string'), 'title': Value('string'), 'type': Value('string')}, 'run_metadata': {'$ref': Value('string')}, 'hardware_profile': {'$ref': Value('string')}, 'metrics': {'$ref': Value('string')}, 'sustainability': {'$ref': Value('string')}, 'integrity': {'$ref': Value('string')}, 'raw_predictions_url': {'anyOf': List({'format': Value('string'), 'maxLength': Value('int64'), 'minLength': Value('int64'), 'type': Value('string')}), 'default': Value('null'), 'title': Value('string')}, 'notes': {'anyOf': List({'maxLength': Value('int64'), 'type': Value('string')}), 'default': Value('null'), 'title': Value('string')}}, 'required': List(Value('string')), 'title': Value('string'), 'type': Value('string')}
              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.

PUMA Community Submissions

Community-contributed benchmark results from PUMA — an empirical evaluation platform for local LLM agents on Project Management Office (PMO) tasks.

What you'll find here

Each submission is a JSON file under submissions/ containing the result of a PUMA benchmark run on one of the supported scenarios:

  • triage_jira — issue triage on the Jira Social Repository (Zenodo DOI), reported as F1-macro
  • effort_tawos — story point effort estimation on TAWOS, reported as MAE in story points
  • prioritization_jira — issue prioritization, reported as nDCG@10 (community-eval, optional)

Each submission includes:

  • Run metadata (model, prompting strategy, scenario, seed, temperature)
  • Hardware profile (CPU-only / GPU / Apple Silicon variant)
  • Metrics with bootstrap confidence intervals
  • Sustainability data (kWh consumed, gCO₂eq via CodeCarbon)
  • Optional raw_predictions_url for integrity verification

Live leaderboard

Interactive view with filters, scatter plots, and verified badges:

👉 pumaproject/puma-leaderboard

Schema

The canonical JSON Schema lives in the governance repo:

schema/submission.v1.json

Minimal example:

{
  "schema_version": "1.0.0",
  "submission_id": "sub_2026_001",
  "submitter": {
    "github_handle": "pumacp",
    "affiliation": "your-organization"
  },
  "run_metadata": {
    "scenario": "triage_jira",
    "model": "qwen2.5:3b",
    "prompting": "few_shot_3",
    "seed": 42,
    "temperature": 0.0
  },
  "hardware_profile": {
    "type": "cpu_only",
    "ram_gb": 16
  },
  "metrics": {
    "f1_macro": 0.5867,
    "ci_lower": 0.5612,
    "ci_upper": 0.6122
  },
  "sustainability": {
    "kwh": 0.0074,
    "co2_g": 3.075
  },
  "raw_predictions_url": "https://github.com/pumacp/puma-community/raw/main/raw/sub_2026_001.jsonl",
  "predictions_summary_hash": "sha256:..."
}

How submissions get here

  1. A community member runs puma share-results --run-id <id> locally
  2. PUMA opens a pull request on pumacp/puma-community
  3. Automated validation checks schema, hash integrity, and reproducibility metadata
  4. Once merged, a GitHub Action mirrors the file to this dataset automatically
  5. The leaderboard Space refreshes within ~5 minutes

Trust model

This dataset is built on a transparency, not gatekeeping principle:

  • Every submission carries enough metadata to reproduce the run locally
  • An optional raw_predictions_url allows the verifier Space to recompute the SHA-256 over the predictions and emit a verified: true sidecar in GitHub
  • Verification is integrity-only — it does not re-execute the model
  • Unverified submissions remain visible but are flagged in the leaderboard

Canonical sources

Resource Location
Source code github.com/pumacp/puma
Governance & PR flow github.com/pumacp/puma-community
Live leaderboard pumaproject/puma-leaderboard
Citable snapshots (quarterly) Zenodo DOI (to be published in Q3 2026)

Citation 

@misc{puma2026,
  title  = {PUMA: PUMA Understanding & Management w Agents},
  author = {{PUMA Project Contributors}},
  year   = {2026},
  url    = {https://github.com/pumacp/puma}
}

License

Downloads last month
109
Total file size:
43.7 kB

Space using pumaproject/puma-community-submissions 1