Datasets:

Heisenger
/

sft-benchmark-trajectories

# Published v0.3.0 handover

The source package and federated canonical release are public:

- code: <https://github.com/heisenger/nvidia_benchmark>, tag `v0.3.0`;
- dataset: <https://huggingface.co/datasets/Heisenger/sft-benchmark-trajectories>;
- immutable dataset revision:
  `36c9f25415617c20d7367e58c8ed059bb06dfcbe`;
- canonical manifest SHA-256:
  `11358b302b25481d6315cfa08eac1bd95c663946d294def383153b2b15da3a73`.

The Hugging Face repository contains 76 files occupying 9,536,452,678 bytes.
Its public metadata inventory covers all 155,278 canonical trajectories and
67,610 task keys. Canonical event payloads are hosted for 136,373 trajectories.
The other 18,905 records retain scores, ranks, lineage, hashes and pinned
upstream pointers but are metadata-only:

- `agentsuite_tau2`: 8,340;
- `browsecomp_rollouts`: 5,063;
- `swegym_sampled`: 5,502.

These records were not discarded. Their transformed-redistribution permission
remains uncertain, so their raw events are not in the public release.
`registry/permissions.json` separately records public access, local research
use, model training, commercial training, original redistribution and
transformed redistribution, including verification state, conditions and
evidence URLs for every source. This is an engineering audit rather than legal
advice.

## Install and download

Python 3.11 or newer is required.

~~~bash
python3 -m venv .venv
.venv/bin/pip install \
  'git+https://github.com/heisenger/nvidia_benchmark.git@v0.3.0'
.venv/bin/pip install huggingface_hub
.venv/bin/hf download \
  Heisenger/sft-benchmark-trajectories \
  --repo-type dataset \
  --revision 36c9f25415617c20d7367e58c8ed059bb06dfcbe \
  --local-dir ./sft-benchmark-trajectories
~~~

Normal training reads this persisted snapshot and does not redownload upstream
sources. Initial ingestion and explicit refresh remain separate source-adapter
commands documented in [trajectory_sources.md](trajectory_sources.md).

## Consume a policy-selected subset

~~~python
from agent_trajectories import (
    CanonicalDatasetAdapter,
    DatasetQuery,
    TrainingView,
)
dataset = CanonicalDatasetAdapter("./sft-benchmark-trajectories")
query = DatasetQuery.for_training(
    splits=("train",),
    content_availability=("hosted",),
    training_permissions=("permitted", "permitted_with_conditions"),
    training_verifications=("verified",),
)
for example in dataset.iter_examples(
    query, view=TrainingView.FULL_TRAJECTORY
):
    row = example.to_dict()
    print(row["sample_id"], row["metadata"]["lineage"])
    break
~~~

That train-split query returns 77,168 trajectories in this release. It combines the
persisted structural/training-quality gates with verified training permission;
projects should additionally filter `commercial_training_permissions` when
applicable.

Use `TrainingView.ASSISTANT_TURN` for causal turn samples. Selection,
deterministic sampling, alternatives, batching, evaluation bundles and optional
PyTorch integration are documented in
[dataset_adapter.md](dataset_adapter.md).

If a consumer is authorized to retrieve a metadata-only source, first run its
explicit pinned source adapter and then federate that local content:

~~~python
dataset = CanonicalDatasetAdapter(
    "./sft-benchmark-trajectories",
    content_overlays={
        "agentsuite_tau2": "./authorized/agentsuite_tau2-canonical"
    },
)
~~~

The overlay is joined by canonical IDs and is never uploaded or copied by the
consumer. Without one, a payload request for a metadata-only record raises
`MissingTrajectoryContentError` while metadata and permission queries still
work.

## Real Nemotron SFT demonstration

The real successful `official_tau3` trajectory
`trajectory:f9cf8588a314fa9804d57402` has 13 events and six assistant targets.
Using the tokenizer and native chat template from
`nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16` at revision
`a9904d24bcc1d289a1950fa9d2b978c47cf903b9` produced:

| View | Tokens | Supervised assistant tokens | Ignored tokens |
|---|---:|---:|---:|
| Full trajectory | 3,174 | 696 | 2,478 |
| One assistant turn | 1,877 | 60 | 1,817 |

The turn example masks 80 prior assistant tokens, and both examples use `-100`
for every non-target label. The demonstrated right-padded batch has shape
`[2, 3174]`, fields `input_ids`, `attention_mask` and `labels`, and SHA-256
`d680ae7e4ea605b3db96370de6760d86ed227411dac901d74a511f72c3346ae8`.
See [nemotron_sft_demo.md](nemotron_sft_demo.md) and the published
`examples/nemotron_sft_demo.json`.

## Scores and validation

No data was rebuilt or rescored for publication. The release preserves:

- upstream native rewards and source-reported success/failure labels;
- computed structural completeness, action/observation, known-outcome,
  verification-reliability and SFT-eligibility scores;
- deterministic within-task quality ranks.

It contains zero independent environment reproductions and zero new native
benchmark replays. Model NLL has not been computed; the Nemotron result is a
tokenization and loss-mask demonstration, not model evaluation.

A fresh virtual environment installed package 0.3.0 from the public Git commit,
downloaded the complete dataset revision anonymously, instantiated consumer API
1.1, and verified:

- all 155,278 metadata records and all 136,373 hosted payload declarations;
- 96,369 verified-permission, persisted-gate training records;
- portable manifest paths with no original workspace references;
- one full and six causal assistant-turn views for the real trajectory above;
- the metadata-only payload guard and all 13 queryable source permission audits.

The primary limitations are source-reported rather than replayed outcomes,
source-specific use conditions, text-encoded tools in some corpora, and
metadata-only payloads where transformed redistribution remains unresolved.

	# Published v0.3.0 handover

	The source package and federated canonical release are public:

	- code: <https://github.com/heisenger/nvidia_benchmark>, tag `v0.3.0`;
	- dataset: <https://huggingface.co/datasets/Heisenger/sft-benchmark-trajectories>;
	- immutable dataset revision:
	`36c9f25415617c20d7367e58c8ed059bb06dfcbe`;
	- canonical manifest SHA-256:
	`11358b302b25481d6315cfa08eac1bd95c663946d294def383153b2b15da3a73`.

	The Hugging Face repository contains 76 files occupying 9,536,452,678 bytes.
	Its public metadata inventory covers all 155,278 canonical trajectories and
	67,610 task keys. Canonical event payloads are hosted for 136,373 trajectories.
	The other 18,905 records retain scores, ranks, lineage, hashes and pinned
	upstream pointers but are metadata-only:

	- `agentsuite_tau2`: 8,340;
	- `browsecomp_rollouts`: 5,063;
	- `swegym_sampled`: 5,502.

	These records were not discarded. Their transformed-redistribution permission
	remains uncertain, so their raw events are not in the public release.
	`registry/permissions.json` separately records public access, local research
	use, model training, commercial training, original redistribution and
	transformed redistribution, including verification state, conditions and
	evidence URLs for every source. This is an engineering audit rather than legal
	advice.

	## Install and download

	Python 3.11 or newer is required.

	~~~bash
	python3 -m venv .venv
	.venv/bin/pip install \
	'git+https://github.com/heisenger/nvidia_benchmark.git@v0.3.0'
	.venv/bin/pip install huggingface_hub

	.venv/bin/hf download \
	Heisenger/sft-benchmark-trajectories \
	--repo-type dataset \
	--revision 36c9f25415617c20d7367e58c8ed059bb06dfcbe \
	--local-dir ./sft-benchmark-trajectories
	~~~

	Normal training reads this persisted snapshot and does not redownload upstream
	sources. Initial ingestion and explicit refresh remain separate source-adapter
	commands documented in [trajectory_sources.md](trajectory_sources.md).

	## Consume a policy-selected subset

	~~~python
	from agent_trajectories import (
	CanonicalDatasetAdapter,
	DatasetQuery,
	TrainingView,
	)

	dataset = CanonicalDatasetAdapter("./sft-benchmark-trajectories")
	query = DatasetQuery.for_training(
	splits=("train",),
	content_availability=("hosted",),
	training_permissions=("permitted", "permitted_with_conditions"),
	training_verifications=("verified",),
	)

	for example in dataset.iter_examples(
	query, view=TrainingView.FULL_TRAJECTORY
	):
	row = example.to_dict()
	print(row["sample_id"], row["metadata"]["lineage"])
	break
	~~~

	That train-split query returns 77,168 trajectories in this release. It combines the
	persisted structural/training-quality gates with verified training permission;
	projects should additionally filter `commercial_training_permissions` when
	applicable.

	Use `TrainingView.ASSISTANT_TURN` for causal turn samples. Selection,
	deterministic sampling, alternatives, batching, evaluation bundles and optional
	PyTorch integration are documented in
	[dataset_adapter.md](dataset_adapter.md).

	If a consumer is authorized to retrieve a metadata-only source, first run its
	explicit pinned source adapter and then federate that local content:

	~~~python
	dataset = CanonicalDatasetAdapter(
	"./sft-benchmark-trajectories",
	content_overlays={
	"agentsuite_tau2": "./authorized/agentsuite_tau2-canonical"
	},
	)
	~~~

	The overlay is joined by canonical IDs and is never uploaded or copied by the
	consumer. Without one, a payload request for a metadata-only record raises
	`MissingTrajectoryContentError` while metadata and permission queries still
	work.

	## Real Nemotron SFT demonstration

	The real successful `official_tau3` trajectory
	`trajectory:f9cf8588a314fa9804d57402` has 13 events and six assistant targets.
	Using the tokenizer and native chat template from
	`nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16` at revision
	`a9904d24bcc1d289a1950fa9d2b978c47cf903b9` produced:

	\| View \| Tokens \| Supervised assistant tokens \| Ignored tokens \|
	\|---\|---:\|---:\|---:\|
	\| Full trajectory \| 3,174 \| 696 \| 2,478 \|
	\| One assistant turn \| 1,877 \| 60 \| 1,817 \|

	The turn example masks 80 prior assistant tokens, and both examples use `-100`
	for every non-target label. The demonstrated right-padded batch has shape
	`[2, 3174]`, fields `input_ids`, `attention_mask` and `labels`, and SHA-256
	`d680ae7e4ea605b3db96370de6760d86ed227411dac901d74a511f72c3346ae8`.
	See [nemotron_sft_demo.md](nemotron_sft_demo.md) and the published
	`examples/nemotron_sft_demo.json`.

	## Scores and validation

	No data was rebuilt or rescored for publication. The release preserves:

	- upstream native rewards and source-reported success/failure labels;
	- computed structural completeness, action/observation, known-outcome,
	verification-reliability and SFT-eligibility scores;
	- deterministic within-task quality ranks.

	It contains zero independent environment reproductions and zero new native
	benchmark replays. Model NLL has not been computed; the Nemotron result is a
	tokenization and loss-mask demonstration, not model evaluation.

	A fresh virtual environment installed package 0.3.0 from the public Git commit,
	downloaded the complete dataset revision anonymously, instantiated consumer API
	1.1, and verified:

	- all 155,278 metadata records and all 136,373 hosted payload declarations;
	- 96,369 verified-permission, persisted-gate training records;
	- portable manifest paths with no original workspace references;
	- one full and six causal assistant-turn views for the real trajectory above;
	- the metadata-only payload guard and all 13 queryable source permission audits.

	The primary limitations are source-reported rather than replayed outcomes,
	source-specific use conditions, text-encoded tools in some corpora, and
	metadata-only payloads where transformed redistribution remains unresolved.