Add validated speech alignment and natural interaction data v1

LICENSE

@@ -0,0 +1,7 @@
+CC BY 4.0
+https://creativecommons.org/licenses/by/4.0/legalcode
+
+LibriTTS-R: Yuma Koizumi and colleagues, 2023. https://www.openslr.org/141/
+otoSpeech-full-duplex-processed-141h: otoearth, 2026.
+https://huggingface.co/datasets/otoearth/otoSpeech-full-duplex-processed-141h
+Changes and processing are described in README.md and reports/.

MANIFEST.json

@@ -0,0 +1,328 @@
+{
+  "files": [
+    {
+      "bytes": 134766737,
+      "path": "alignment/test-00000.parquet",
+      "sha256": "0eb86a4dd7d26d96b44a2605da943415691c480a88835e85d6f3696db76e2268"
+    },
+    {
+      "bytes": 134764261,
+      "path": "alignment/test-00001.parquet",
+      "sha256": "eef79de35474becc82a98017659b59ad5f83bf0a5d0636f33cee93b727ae5b47"
+    },
+    {
+      "bytes": 134890750,
+      "path": "alignment/test-00002.parquet",
+      "sha256": "d668294828c9a7d0a8d1e25691bde52e81dfdff1924224c74cade396ee533099"
+    },
+    {
+      "bytes": 31548974,
+      "path": "alignment/test-00003.parquet",
+      "sha256": "0b039503b892f2856299967a271b11bfa37fab4ba6c4c76f338b8f917b61cd12"
+    },
+    {
+      "bytes": 134779541,
+      "path": "alignment/train-00000.parquet",
+      "sha256": "b28d86ed081689ed5a6fddb75329e2fc99ec33b8075ce80d4a24057dd728805f"
+    },
+    {
+      "bytes": 134697630,
+      "path": "alignment/train-00001.parquet",
+      "sha256": "e35c8fe191f2905800229eaf469827249bf0070cd3a12fd18e5fa32bfd42e0af"
+    },
+    {
+      "bytes": 134680170,
+      "path": "alignment/train-00002.parquet",
+      "sha256": "5913fd8cac9d6df865ed18def388541eb15d23b24cc7b955c7a29d1e7a358b79"
+    },
+    {
+      "bytes": 134662633,
+      "path": "alignment/train-00003.parquet",
+      "sha256": "4428f97042ee6eafd725fce1a9fdd9f2f71ba096a55d134de5c2cfc3598d73bb"
+    },
+    {
+      "bytes": 134653215,
+      "path": "alignment/train-00004.parquet",
+      "sha256": "6057816f048780e69ab0a3aedb70245a88d4c6fcb790e3ba57351952728118f5"
+    },
+    {
+      "bytes": 134746994,
+      "path": "alignment/train-00005.parquet",
+      "sha256": "5458fe1b579466ba46ed5528177733c1cd88367f50309acb1fa18258a9daa1be"
+    },
+    {
+      "bytes": 134865022,
+      "path": "alignment/train-00006.parquet",
+      "sha256": "8489c1ec2aa837a5d82b5669c03d45442a496f7f19c3ec1d44bdf86031fd9d7a"
+    },
+    {
+      "bytes": 134784352,
+      "path": "alignment/train-00007.parquet",
+      "sha256": "71d7120cdf6532903ac16392132ecd12cd81d6b2b6d53588b0253915ea9f5e64"
+    },
+    {
+      "bytes": 134760239,
+      "path": "alignment/train-00008.parquet",
+      "sha256": "d12e47bea296d54012b729be9279ce1546f3760a006257e47c22cd840a0edeab"
+    },
+    {
+      "bytes": 134786585,
+      "path": "alignment/train-00009.parquet",
+      "sha256": "bf68eb8d956eac20cbeb664f16d7bac9d97e6cd350e62ce8dd0bc345b1cc13fc"
+    },
+    {
+      "bytes": 134791969,
+      "path": "alignment/train-00010.parquet",
+      "sha256": "a113a18ddaa61efd1bdbe557eef1e2817d819716c8f519eda018ac4e01718b9b"
+    },
+    {
+      "bytes": 134750291,
+      "path": "alignment/train-00011.parquet",
+      "sha256": "cdd271a25639976385a3f6a886a04731c530709e4be9fdd411783e6ac1da3a8c"
+    },
+    {
+      "bytes": 134779402,
+      "path": "alignment/train-00012.parquet",
+      "sha256": "e8d656596dc5bdc36b93d96510899f2264905ddec1422b3ae61faa73bdafc62d"
+    },
+    {
+      "bytes": 134686119,
+      "path": "alignment/train-00013.parquet",
+      "sha256": "63e20e0eb2cc0f4949a7eed3d4fb991ae77d76971b212515e42953a359a7267c"
+    },
+    {
+      "bytes": 134785850,
+      "path": "alignment/train-00014.parquet",
+      "sha256": "37270a7e01d2e3f302d64bda242c7e8b5281b368cc36a11d7fddbeac63175741"
+    },
+    {
+      "bytes": 134658251,
+      "path": "alignment/train-00015.parquet",
+      "sha256": "33ff03a134d2983799497283c0828229c0f5c22884176b8faaca491b30bafc6a"
+    },
+    {
+      "bytes": 134749072,
+      "path": "alignment/train-00016.parquet",
+      "sha256": "186f9e0e783707f365ec93aa9f80ade0c8a4886b1facf9eb3ca0ac1e26d068be"
+    },
+    {
+      "bytes": 134677156,
+      "path": "alignment/train-00017.parquet",
+      "sha256": "562d437e18b34226355692f4cad3e2503870911e236579c554c26598559f2ab8"
+    },
+    {
+      "bytes": 134650446,
+      "path": "alignment/train-00018.parquet",
+      "sha256": "1004c87d58cda7a8662ebef865b9a3c28a843209aeeb777b5baac4ec75335993"
+    },
+    {
+      "bytes": 134682989,
+      "path": "alignment/train-00019.parquet",
+      "sha256": "897556ba4d60fc9789adcaafe88c4e921f060443f1398d1c1daabcc90ea170fe"
+    },
+    {
+      "bytes": 134628479,
+      "path": "alignment/train-00020.parquet",
+      "sha256": "fd37fc4efe28217e7d6c98caa700fcef4e0a6f045817113a82fe020e84f944ce"
+    },
+    {
+      "bytes": 134743483,
+      "path": "alignment/train-00021.parquet",
+      "sha256": "d8326a953abd693c8cbe3211d1d2ec7b3487e9ac10dd52943d8358fbb6fd0d85"
+    },
+    {
+      "bytes": 134662708,
+      "path": "alignment/train-00022.parquet",
+      "sha256": "2ed31b0e85914e065f18b987946f7606cebab2b7b52a6b70510882b28c2a6c99"
+    },
+    {
+      "bytes": 48335509,
+      "path": "alignment/train-00023.parquet",
+      "sha256": "eb27d1be30c928bb6f25956b53b50a24bc35e21f6d235aee9ba964f172e36d3d"
+    },
+    {
+      "bytes": 134664319,
+      "path": "alignment/validation-00000.parquet",
+      "sha256": "de0b6ca6e0af3f8370f471c8917b279d2cda0cbae21e28a89ca99a4ebee114d1"
+    },
+    {
+      "bytes": 134830455,
+      "path": "alignment/validation-00001.parquet",
+      "sha256": "0875be2405e1bf44b73ab45ec1a5fdd27c75a25dc5c8f9e277a46d9049046333"
+    },
+    {
+      "bytes": 134800212,
+      "path": "alignment/validation-00002.parquet",
+      "sha256": "99e36a53aac44f6fd6c184d805b7cb519ccf09b46754c943eef8291961f38384"
+    },
+    {
+      "bytes": 134692460,
+      "path": "alignment/validation-00003.parquet",
+      "sha256": "e6d43796cc5bf30de05f78afa774be380961a0ef9a561f1c8e2b233f91a9c31d"
+    },
+    {
+      "bytes": 23630855,
+      "path": "alignment/validation-00004.parquet",
+      "sha256": "5b105c61af8ddefa4b3c753f28838754253b2d4fa5893dddfaa97158cd1ed4f6"
+    },
+    {
+      "bytes": 1137461,
+      "path": "codec-pilot/asr-train.jsonl",
+      "sha256": "c735912c8371da4adab782525a40f43866b06b4f6cfb5fcd4c2a92dc1022746c"
+    },
+    {
+      "bytes": 861,
+      "path": "codec-pilot/asr.toml",
+      "sha256": "1506bdb9ae07fcc2d90a89c91e3307b2907faaf7fcfd3da4124c71075b46249f"
+    },
+    {
+      "bytes": 330612,
+      "path": "codec-pilot/cache-index.json",
+      "sha256": "9e1ec8079479159df413ad4236584a4f1d40ba5ddb7ef1cbfc8d1a660ea696e8"
+    },
+    {
+      "bytes": 95825920,
+      "path": "codec-pilot/tensors.tar",
+      "sha256": "60cbb9709121f52cba5d0e624d536276240f8a4ec6317a9dba6d6f85706fc4d4"
+    },
+    {
+      "bytes": 1132050,
+      "path": "codec-pilot/tts-train.jsonl",
+      "sha256": "d20c23bb666ddb7e0d9de60b03d90697dc29c7411a6ff3d198663863ee4c38b1"
+    },
+    {
+      "bytes": 795,
+      "path": "codec-pilot/tts.toml",
+      "sha256": "5e62388546ed14ada4fcb727d70032d0e8428b67a480a98a663ab24628f945df"
+    },
+    {
+      "bytes": 133243531,
+      "path": "interaction/test-00000.parquet",
+      "sha256": "bd19377893662d1ebe88a5da75bdda728e4b8051053491affb0739f1dba23904"
+    },
+    {
+      "bytes": 22998066,
+      "path": "interaction/test-00001.parquet",
+      "sha256": "0a80f98147c49131ffba07919f9b1be037d6d69276b9bcbe02a4be4f31cbddf3"
+    },
+    {
+      "bytes": 133702820,
+      "path": "interaction/train-00000.parquet",
+      "sha256": "3211ece2b8d8c6ae1fffdbb2167659906aa37c60b0401cd73cf05c9ab45656e3"
+    },
+    {
+      "bytes": 132979497,
+      "path": "interaction/train-00001.parquet",
+      "sha256": "21b43761ee2e5b677ec4eb449d1898bef17a3c432598c194786c2d46e60a4f7f"
+    },
+    {
+      "bytes": 132973624,
+      "path": "interaction/train-00002.parquet",
+      "sha256": "7db86a25c54a54f9be100f64b10bf74db6b958ff9057e1fa1336d42c7cc8c1a2"
+    },
+    {
+      "bytes": 134072685,
+      "path": "interaction/train-00003.parquet",
+      "sha256": "193f58818df061714d5850586dc810a282a0d698ff2bd1e8ec666e4553a53fad"
+    },
+    {
+      "bytes": 133397209,
+      "path": "interaction/train-00004.parquet",
+      "sha256": "745cf254d78db7165d6aadcc881df10250d525294b42c1534715d5fb910a498e"
+    },
+    {
+      "bytes": 132631632,
+      "path": "interaction/train-00005.parquet",
+      "sha256": "687154713a94a5a070b16f5bcd986b3a048505fce962f49f2327556b9dfbd301"
+    },
+    {
+      "bytes": 133062493,
+      "path": "interaction/train-00006.parquet",
+      "sha256": "83053bba700e76d3b9dddbeb7b27fe59c70bd5b07c2db1c512c190ca25de655d"
+    },
+    {
+      "bytes": 133550892,
+      "path": "interaction/train-00007.parquet",
+      "sha256": "df0e7d53f3d8d7040e4a5312a137822dcb6c0f085115649c77348e0fc7ca26fa"
+    },
+    {
+      "bytes": 133634730,
+      "path": "interaction/train-00008.parquet",
+      "sha256": "e5902588f2930f701a5b654f6fd1abf009abc8e6e930d7e892d5f21c55e4c3ad"
+    },
+    {
+      "bytes": 123060292,
+      "path": "interaction/train-00009.parquet",
+      "sha256": "a1f0fe5d89397ed51cd0bc87af76c01fb5a4e9bc0d5af296bf19be8a732a41e6"
+    },
+    {
+      "bytes": 133016086,
+      "path": "interaction/validation-00000.parquet",
+      "sha256": "7d6a1dab2ce8ae5ca50db08951c69911d1096bb83639e93ed387ab6184f66f00"
+    },
+    {
+      "bytes": 48162319,
+      "path": "interaction/validation-00001.parquet",
+      "sha256": "89aecbe06e3b6d1cbc9abc18ce8c1d96c3a80182a35fc4a90e64e70b6679acfc"
+    },
+    {
+      "bytes": 344,
+      "path": "LICENSE",
+      "sha256": "ce3102fc1c05e3a5be5c095f2d8d3e6940195e75867e106b3a8bbe55c1524a14"
+    },
+    {
+      "bytes": 7531,
+      "path": "README.md",
+      "sha256": "51eee018612781325e5ef2f00e7920a44d3158c34c81ae75bdb51cd578045a20"
+    },
+    {
+      "bytes": 9381,
+      "path": "reports/alignment.json",
+      "sha256": "516860dc0b5cc0f540722f75779cea62caf95f5c939aea155e1974a488102007"
+    },
+    {
+      "bytes": 634,
+      "path": "reports/codec-pilot.json",
+      "sha256": "210ddb8f452cd8917ac9d08337b1b7a1ea14676411d98a2a6c2268dee270f381"
+    },
+    {
+      "bytes": 34257,
+      "path": "reports/interaction-selection.json",
+      "sha256": "eeda66e6d8332de4ea67a817829b496f7ca144d63da8fe4d7516e9809d9d19c1"
+    },
+    {
+      "bytes": 10166,
+      "path": "reports/interaction.json",
+      "sha256": "8490fa9914206b44f1aba1739c484795924ff0d65d03d6d40f55937e8a6b2b49"
+    },
+    {
+      "bytes": 1212,
+      "path": "reports/processing.json",
+      "sha256": "a203f2f1bbd4e6200977bb42dce2720b8ead37999e5543ac03ad152adb4314ec"
+    },
+    {
+      "bytes": 1278,
+      "path": "reports/validation.json",
+      "sha256": "f499024484f0fd036369a25035126132a09ad1aa99a1ce154a28b17a834e6f5d"
+    },
+    {
+      "bytes": 15302,
+      "path": "tools/hf_training.py",
+      "sha256": "a7c17dfba679cbda48e3d02df6cb28ea97de4c8fd7f8bfc0a1531f7f89c95306"
+    },
+    {
+      "bytes": 4376,
+      "path": "tools/materialize_alignment.py",
+      "sha256": "3b0fc0c711ea96ab78ed6c26c14feee1064f37789247cadf29f5e51e390e0ae9"
+    },
+    {
+      "bytes": 175,
+      "path": "tools/requirements-data.txt",
+      "sha256": "b639dd008ae77af97901c15c34849eb5206f7c3adb8623563b2fb0f4c66bf8d6"
+    }
+  ],
+  "release": "nano-duplex-data-v1",
+  "schema_version": 1,
+  "status": "validated",
+  "total_bytes": 5904585359
+}

README.md

@@ -0,0 +1,143 @@
+---
+language:
+- en
+license: cc-by-4.0
+task_categories:
+- automatic-speech-recognition
+- text-to-speech
+- audio-to-audio
+tags:
+- nano-duplex
+- conversational-speech
+- full-duplex
+- mimi
+configs:
+- config_name: alignment
+  default: true
+  data_files:
+  - split: train
+    path: alignment/train-*.parquet
+  - split: validation
+    path: alignment/validation-*.parquet
+  - split: test
+    path: alignment/test-*.parquet
+- config_name: interaction
+  data_files:
+  - split: train
+    path: interaction/train-*.parquet
+  - split: validation
+    path: interaction/validation-*.parquet
+  - split: test
+    path: interaction/test-*.parquet
+---
+
+# nano-duplex data v1
+
+English speech alignment and a first natural-conversation collection for nano-duplex research.
+All audio is embedded in Parquet as PCM16 FLAC at 24 kHz. No external audio paths are required.
+
+| Configuration | Split | Examples | Audio hours | Speakers |
+| --- | --- | ---: | ---: | ---: |
+| alignment | train | 26,478 | 32.377 | 247 |
+| alignment | validation | 4,708 | 5.817 | 40 |
+| alignment | test | 3,624 | 4.428 | 38 |
+| interaction | train | 946 | 15.240 | 57 |
+| interaction | validation | 122 | 1.978 | 11 |
+| interaction | test | 116 | 1.877 | 11 |
+
+## Sources and processing
+
+**Alignment:** [LibriTTS-R](https://www.openslr.org/141/), by Yuma Koizumi and colleagues.
+The complete official `train_clean_100`, `dev_clean`, and `test_clean` archives passed the
+official MD5 checks before extraction. Keep complete mono 24 kHz PCM16 utterances from
+1 through 10 seconds with normalized transcripts. Preserve official splits and original IDs.
+Reject missing transcripts, out-of-range durations, duplicate audio, and entirely silent clips.
+WAV-to-FLAC conversion is lossless; original WAV and decoded PCM SHA256 are recorded.
+Original and normalized text are included; the source transcript hash refers to the original
+normalized-text file, before trimming its surrounding whitespace. RMS and clipping metrics
+are diagnostics, not human quality judgments. This is a filtered subset, not all hours implied
+by the upstream archive names. Citation: Koizumi et al., "LibriTTS-R: A Restored Multi-Speaker
+Text-to-Speech Corpus," 2023.
+
+**Interaction:** [otoearth/otoSpeech-full-duplex-processed-141h](https://huggingface.co/datasets/otoearth/otoSpeech-full-duplex-processed-141h).
+Use eight complete upstream tar shards, selected by lexicographic filename order, at revision
+`c635c29ad083570fe2cadbafc75770e76eded2da`. Every shard passed its pinned Hub SHA256.
+The selection contains 73 complete sessions. Resample each complete stereo session from
+44.1 kHz to 24 kHz using SciPy polyphase filtering, then split into consecutive 60-second
+windows (including the shorter final window). Both channels retain the shared clock.
+Preserve overlapping speech and silence, and re-mute the original redacted intervals after
+resampling. The two `speaker_ids` are ordered by audio channel. Source-relative timestamps
+allow exact reassembly. Speaker-connected components stay within one split; there is no
+reported speaker-ID or session overlap across train/validation/test. Speaker identities beyond
+the IDs reported by the source have not been independently verified. Omit optional surveys
+and demographic profiles. Citation: otoearth, "otoSpeech-full-duplex-processed-141h," 2026.
+
+## Scope
+
+Natural conversations have no assistant tool traces, cancelled playback plans, certified
+prefix-causal action targets, or transcripts in this release. `causal_verified=false` and
+`has_transcript=false` are explicit. These recordings support interaction preparation and
+analysis; they are not ready-made Stage B-D causal trajectories. No synthetic speech is
+substituted for real recordings. Tool, vision, and motion supervision are outside this release.
+
+`codec-pilot/` contains real Mimi ASR features and eight-codebook TTS targets for 773 training
+utterances (57.301 minutes), drawn from `alignment/train`. This pilot has no independent
+evaluation split. Its JSONL manifests reference tensor files packed in `tensors.tar`;
+`cache-index.json` records every member's size and SHA256. The server entry expands and
+verifies them automatically. The full
+alignment and interaction collections are processed audio; only this pilot is codec-encoded.
+The exact Mimi repository, revision, checkpoint digest and cache dimensions are recorded.
+Tensor files must be loaded with `torch.load(..., weights_only=True)` as nano-duplex does.
+The provided `codec-pilot/asr.toml` and `codec-pilot/tts.toml` select matching pinned assets.
+
+## Loading
+
+```python
+from datasets import load_dataset, Audio
+
+alignment = load_dataset("JonesLin/nano-duplex-data-v1", "alignment", token=True)
+interaction = load_dataset("JonesLin/nano-duplex-data-v1", "interaction", token=True)
+# Decode with soundfile without requiring torchcodec:
+audio = alignment["train"].cast_column("audio", Audio(decode=False))[0]["audio"]
+from io import BytesIO
+import soundfile as sf
+pcm, sample_rate = sf.read(BytesIO(audio["bytes"]))
+```
+
+From the nano-duplex project checkout, install `requirements-speech.txt` and
+`requirements-data.txt`, then use the server entry (a copy is included at
+`tools/hf_training.py`; it belongs in the project's `nano_duplex` package):
+
+```bash
+# Download only the small pre-encoded pilot, verify and expand it, then train ASR/TTS.
+python -m nano_duplex.hf_training --profile pilot --steps 400
+
+# Full Stage A: download alignment data, encode each split once, train, evaluate.
+python -m nano_duplex.hf_training --profile full --nproc 4 --steps 20000
+```
+
+The first command resolves a concrete Hub commit and writes `dataset-lock.json`.
+Alternatively pass `--revision` with an exact commit. `--snapshot-dir`, `--work-dir`,
+and `--run-dir` select server storage. `--phase prepare` stops before training;
+`--phase train --resume` reuses data caches and resumes existing task checkpoints.
+An unstarted second task starts fresh. Use `--dry-run` to inspect commands without
+network access or file writes. H200 distributed execution has not been run here;
+20,000 steps is a starting budget, not a measured convergence requirement.
+
+The full profile still performs one Mimi encoding pass over the complete alignment
+collection on first use. Subsequent runs reuse checked prepared manifests. Use
+`tools/materialize_alignment.py` separately if only mono WAV/raw JSONL export is needed.
+The interaction configuration is available through `load_dataset`, but is not fed into
+the causal duplex trainer without the additional labels described above.
+Changing a dataset manifest requires a new training run, not resuming an old optimizer state.
+
+## Provenance and license
+
+Both source corpora are provided under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/).
+This derivative retains that license and the source attributions above. The source-provided
+gated dataset access was obtained through the collecting user's own Hugging Face account.
+Source receipts, split statistics, exclusions, processing parameters, and full validation
+results are in `reports/`. `MANIFEST.json` inventories the exact release files with SHA256.
+The validation reads every packaged audio example, checks its PCM hash, duration, channel
+count, session continuity, redaction masks, and split isolation. No speech-quality or
+conversational-performance claim follows from those structural checks.

alignment/test-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0eb86a4dd7d26d96b44a2605da943415691c480a88835e85d6f3696db76e2268
+size 134766737

alignment/test-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eef79de35474becc82a98017659b59ad5f83bf0a5d0636f33cee93b727ae5b47
+size 134764261

alignment/test-00002.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d668294828c9a7d0a8d1e25691bde52e81dfdff1924224c74cade396ee533099
+size 134890750

alignment/test-00003.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b039503b892f2856299967a271b11bfa37fab4ba6c4c76f338b8f917b61cd12
+size 31548974

alignment/train-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b28d86ed081689ed5a6fddb75329e2fc99ec33b8075ce80d4a24057dd728805f
+size 134779541

alignment/train-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e35c8fe191f2905800229eaf469827249bf0070cd3a12fd18e5fa32bfd42e0af
+size 134697630

alignment/train-00002.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5913fd8cac9d6df865ed18def388541eb15d23b24cc7b955c7a29d1e7a358b79
+size 134680170

alignment/train-00003.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4428f97042ee6eafd725fce1a9fdd9f2f71ba096a55d134de5c2cfc3598d73bb
+size 134662633

alignment/train-00004.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6057816f048780e69ab0a3aedb70245a88d4c6fcb790e3ba57351952728118f5
+size 134653215

alignment/train-00005.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5458fe1b579466ba46ed5528177733c1cd88367f50309acb1fa18258a9daa1be
+size 134746994

alignment/train-00006.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8489c1ec2aa837a5d82b5669c03d45442a496f7f19c3ec1d44bdf86031fd9d7a
+size 134865022

alignment/train-00007.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:71d7120cdf6532903ac16392132ecd12cd81d6b2b6d53588b0253915ea9f5e64
+size 134784352

alignment/train-00008.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d12e47bea296d54012b729be9279ce1546f3760a006257e47c22cd840a0edeab
+size 134760239

alignment/train-00009.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bf68eb8d956eac20cbeb664f16d7bac9d97e6cd350e62ce8dd0bc345b1cc13fc
+size 134786585

alignment/train-00010.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a113a18ddaa61efd1bdbe557eef1e2817d819716c8f519eda018ac4e01718b9b
+size 134791969

alignment/train-00011.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cdd271a25639976385a3f6a886a04731c530709e4be9fdd411783e6ac1da3a8c
+size 134750291

alignment/train-00012.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e8d656596dc5bdc36b93d96510899f2264905ddec1422b3ae61faa73bdafc62d
+size 134779402

alignment/train-00013.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63e20e0eb2cc0f4949a7eed3d4fb991ae77d76971b212515e42953a359a7267c
+size 134686119

alignment/train-00014.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37270a7e01d2e3f302d64bda242c7e8b5281b368cc36a11d7fddbeac63175741
+size 134785850

alignment/train-00015.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:33ff03a134d2983799497283c0828229c0f5c22884176b8faaca491b30bafc6a
+size 134658251

alignment/train-00016.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:186f9e0e783707f365ec93aa9f80ade0c8a4886b1facf9eb3ca0ac1e26d068be
+size 134749072

alignment/train-00017.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:562d437e18b34226355692f4cad3e2503870911e236579c554c26598559f2ab8
+size 134677156

alignment/train-00018.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1004c87d58cda7a8662ebef865b9a3c28a843209aeeb777b5baac4ec75335993
+size 134650446

alignment/train-00019.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:897556ba4d60fc9789adcaafe88c4e921f060443f1398d1c1daabcc90ea170fe
+size 134682989

alignment/train-00020.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fd37fc4efe28217e7d6c98caa700fcef4e0a6f045817113a82fe020e84f944ce
+size 134628479

alignment/train-00021.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d8326a953abd693c8cbe3211d1d2ec7b3487e9ac10dd52943d8358fbb6fd0d85
+size 134743483

alignment/train-00022.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2ed31b0e85914e065f18b987946f7606cebab2b7b52a6b70510882b28c2a6c99
+size 134662708

alignment/train-00023.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eb27d1be30c928bb6f25956b53b50a24bc35e21f6d235aee9ba964f172e36d3d
+size 48335509

alignment/validation-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de0b6ca6e0af3f8370f471c8917b279d2cda0cbae21e28a89ca99a4ebee114d1
+size 134664319

alignment/validation-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0875be2405e1bf44b73ab45ec1a5fdd27c75a25dc5c8f9e277a46d9049046333
+size 134830455

alignment/validation-00002.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:99e36a53aac44f6fd6c184d805b7cb519ccf09b46754c943eef8291961f38384
+size 134800212

alignment/validation-00003.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e6d43796cc5bf30de05f78afa774be380961a0ef9a561f1c8e2b233f91a9c31d
+size 134692460

alignment/validation-00004.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5b105c61af8ddefa4b3c753f28838754253b2d4fa5893dddfaa97158cd1ed4f6
+size 23630855

codec-pilot/asr.toml

@@ -0,0 +1,29 @@
+# Pinned Hub assets. Override --dataset with your snapshot path.
+policy = "Qwen/Qwen3-0.6B-Base"
+policy_revision = "da87bfb608c14b7cf20ba1ce41287e8de496c0cd"
+mimi_repo = "kyutai/moshiko-pytorch-bf16"
+mimi_revision = "2bfc9ae6e89079a5cc7ed2a68436010d91a3d289"
+device = "cuda"
+audio_dim = 512
+vision_dim = 768
+codebooks = 8
+codebook_size = 2048
+audio_head_width = 256
+audio_head_layers = 2
+audio_head_attention_heads = 4
+max_context_tokens = 2048
+precision = "bfloat16"
+freeze_policy = true
+batch_size = 8
+gradient_accumulation = 1
+learning_rate = 0.0003
+weight_decay = 0.01
+steps = 400
+save_every = 400
+dataset = "codec-pilot/asr-train.jsonl"
+output_dir = "runs/hf-pilot-asr"
+stage = "alignment_asr"
+
+# Opt-in: use a new run; effective batch differs from the baseline.
+alignment_batching = "batched"
+frozen_linear_dtype = "bfloat16"

codec-pilot/tensors.tar

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60cbb9709121f52cba5d0e624d536276240f8a4ec6317a9dba6d6f85706fc4d4
+size 95825920

codec-pilot/tts.toml

@@ -0,0 +1,28 @@
+policy = "Qwen/Qwen3-0.6B-Base"
+policy_revision = "da87bfb608c14b7cf20ba1ce41287e8de496c0cd"
+mimi_repo = "kyutai/moshiko-pytorch-bf16"
+mimi_revision = "2bfc9ae6e89079a5cc7ed2a68436010d91a3d289"
+device = "cuda"
+audio_dim = 512
+vision_dim = 768
+codebooks = 8
+codebook_size = 2048
+audio_head_width = 256
+audio_head_layers = 2
+audio_head_attention_heads = 4
+max_context_tokens = 2048
+precision = "bfloat16"
+freeze_policy = true
+batch_size = 8
+gradient_accumulation = 1
+learning_rate = 0.0003
+weight_decay = 0.01
+steps = 400
+save_every = 400
+dataset = "codec-pilot/tts-train.jsonl"
+output_dir = "runs/hf-pilot-tts"
+stage = "alignment_tts"
+
+# Opt-in: use a new run; effective batch differs from the baseline.
+alignment_batching = "batched"
+frozen_linear_dtype = "bfloat16"

interaction/test-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd19377893662d1ebe88a5da75bdda728e4b8051053491affb0739f1dba23904
+size 133243531

interaction/test-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0a80f98147c49131ffba07919f9b1be037d6d69276b9bcbe02a4be4f31cbddf3
+size 22998066

interaction/train-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3211ece2b8d8c6ae1fffdbb2167659906aa37c60b0401cd73cf05c9ab45656e3
+size 133702820

interaction/train-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21b43761ee2e5b677ec4eb449d1898bef17a3c432598c194786c2d46e60a4f7f
+size 132979497

interaction/train-00002.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7db86a25c54a54f9be100f64b10bf74db6b958ff9057e1fa1336d42c7cc8c1a2
+size 132973624

interaction/train-00003.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:193f58818df061714d5850586dc810a282a0d698ff2bd1e8ec666e4553a53fad
+size 134072685

interaction/train-00004.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:745cf254d78db7165d6aadcc881df10250d525294b42c1534715d5fb910a498e
+size 133397209

interaction/train-00005.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:687154713a94a5a070b16f5bcd986b3a048505fce962f49f2327556b9dfbd301
+size 132631632