Release rc2

README.md

@@ -12,11 +12,11 @@ tags:
   - onnx
 ---
 
-# Pipecat Voice Reader (rc1)
+# Pipecat Voice Reader (rc2)
 
-> **⚠️ Experimental.** This is a first release candidate, published for trying out and feedback. Its outputs,
-> names and calibration may change between releases, and its readings are cues for an LLM to weigh, not verdicts
-> to act on alone. Pin a revision, and read the limitations below before relying on it.
+> **⚠️ Experimental.** This is a release candidate, published for trying out and feedback. Its outputs, names and
+> calibration may change between releases, and its readings are cues for an LLM to weigh, not verdicts to act on
+> alone. Pin a revision, and read the limitations below before relying on it.
 
 A small CPU model that hears how a user *sounds*: tone of voice, sarcasm in the delivery, energy, laughter and
 speech rate. It is the voice half of a pair: it is made to run beside Pipecat's
@@ -28,6 +28,24 @@ writes the reading as RCS perception at the head of the user's message, after th
 ([emotion happy=0.62, neutral=0.30],[sarcasm value=0.08],[tone frustrated=0.71, neutral=0.22],[arousal value=0.34],[laughter value=0.02],[speech_rate value=3.1],[sarcasm_voice value=0.63]) Oh, great.
 ```
 
+## What changed since rc1
+
+- **Licence-clean targets.** Every training target in rc2 is a judge's verdict or vote share, a cue from LAION's
+  CC BY 4.0 model, or a measurement of the audio. rc1 trained its tone head, where the two judges disagreed, on the
+  mean of two third-party audio emotion models; rc2 uses their vote share instead and has no such model in its
+  targets, so its lineage is Whisper (MIT), LAION's affect model (CC BY 4.0), the corpora under Training and the
+  judges. One thread remains: the pseudo-labelled windows below carry the rc1 base model's readings, so rc1's
+  lineage reaches rc2 second-hand through those labels.
+- **More judged conversation.** A second round of about 24,000 conversational turns rated by both judges went into
+  training, and its held-out turns are a new test set (below).
+- **Pseudo-labelled speech.** 159,000 windows read by the rc1 base model (Emolia clips no judge has rated, and
+  conversation turns) joined the training set. The sarcasm and conversation-tone gains come from them.
+- **`tiny` now has the `sad` output** (int8 14.1 MB, from 13.3).
+- Sarcasm in conversation AUROC 0.917 → 0.921 (base) and 0.904 → 0.908 (tiny); conversation tone macro-F1
+  0.706 → 0.754 and 0.712 → 0.739; acted sadness read 23 → 30 of 34 CREMA-D clips (base). Tone on EmoNet-Voice's
+  synthetic clips slips for base (0.563 → 0.510).
+- The tone head leans neutral a little more on borderline turns than rc1 did: see `tone_bias` under Outputs.
+
 ## Made to pair with the Tone Reader
 
 The two readers split the work, and each declares its own signals so the LLM sees both side by side:
@@ -41,7 +59,7 @@ The two readers split the work, and each declares its own signals so the LLM see
 The voice reader never writes `emotion` or `sarcasm`. Its `sarcasm_voice` is a weak cue on its own, and its
 declared description tells the LLM to weigh it with the Tone Reader's `sarcasm`: the case it earns its place in is
 words that read sincere, said in a voice that isn't. For example, from our test clips, "He's the greatest, I'd
-say." said sarcastically reads as happy and 2 % sarcastic from the words, and 15 % (above the 6.8 % threshold)
+say." said sarcastically reads as happy and 2 % sarcastic from the words, and 10 % (above the 6.1 % threshold)
 from the voice.
 
 It does run without the Tone Reader: the LLM then gets `tone` instead of `emotion`, and no `sarcasm` from the
@@ -53,15 +71,16 @@ This release carries two sizes of the same model, one folder each, read by the s
 
 | Folder | `variant` | Params | int8 ONNX | Read time, p50 (M1 Max, 4 / 2 / 1 threads) | 1 thread, x86 server | Choose it for |
 |---|---|---|---|---|---|---|
-| `voice-reader-base/` | `"base"` (default) | 26M | 28.6 MB | 33 / 56 / 100 ms | ~100–150 ms | the best tone and sadness reading; adds the `sad` output |
-| `voice-reader-tiny/` | `"tiny"` | 12.5M | 13.3 MB | 16 / 26 / 47 ms | ~45 ms | one-vCPU deployments: sarcasm about as well, half the CPU |
+| `voice-reader-base/` | `"base"` (default) | 26M | 28.6 MB | 28 / 49 / 92 ms | ~90–160 ms | the best tone and sadness reading |
+| `voice-reader-tiny/` | `"tiny"` | 12.5M | 14.1 MB | 13 / 23 / 42 ms | ~45–70 ms | one-vCPU deployments: sarcasm about as well, half the CPU |
 
 The read starts when the VAD says the user stopped and runs while the speech-to-text finalises, so it rarely
 delays the transcription; the cost to watch is CPU at the end of each turn, beside the VAD, the turn detector and
 the reply. Give `base` two threads where the machine allows.
 
 Each folder has `model.int8.onnx` (the default), `model.onnx` (fp32), `config.json` (calibration, threshold,
-output names) and `MANIFEST.json` (SHA-256 of every file, checked by the runtime on download).
+output names) and `MANIFEST.json` (SHA-256 of every file, checked by the runtime on download). The `config.json` at
+the root is the release index: the variants, their folders and files. The runtime reads it first on every load.
 
 ## Use
 
@@ -75,7 +94,7 @@ pipeline = Pipeline([
     transport.input(),
     stt,
     PipecatToneReader(),  # the words: emotion, sarcasm
-    VoiceReaderProcessor(params=VoiceReaderParams(revision="rc1", variant="base", threads=2)),  # the voice
+    VoiceReaderProcessor(params=VoiceReaderParams(revision="rc2", variant="base", threads=2)),  # the voice
     user_aggregator,
     llm,
     tts,
@@ -93,16 +112,23 @@ this repository at the pinned revision the first time.
 | Signal | Published to the LLM | Meaning |
 |---|---|---|
 | `tone` | yes, top two | how the voice sounds: positive, neutral, sad, frustrated (temperature-calibrated) |
-| `sarcasm_voice` | yes | the chance the delivery alone sounds sarcastic, Platt-calibrated on real conversation; below the release threshold (0.068) it is published under 0.5, as a guess; dropped under laughter |
+| `sarcasm_voice` | yes | the chance the delivery alone sounds sarcastic, Platt-calibrated on real conversation; below the release threshold (0.061 for `base`, 0.063 for `tiny`) it is published under 0.5, as a guess; dropped under laughter |
 | `arousal` | yes | vocal energy, 0–1 |
 | `laughter` | yes | 0–1 |
 | `speech_rate` | yes | syllables per second |
 | `upset` | no, raw reading only | how clearly the speaker is upset, 0–1 |
-| `sad` (`base` only) | no, raw reading only | how likely listeners would hear sadness, 0–1 |
+| `sad` | no, raw reading only | how likely listeners would hear sadness, 0–1 |
 
 Nothing is written for a turn with under 0.8 s of speech. The raw reading, including `upset` and `sad`, reaches
 the application through the processor's `on_reading` event.
 
+**`tone_bias`.** `tone` is calibrated on real conversation, where most turns are neutral, so on a borderline turn
+it leans neutral: a sad or frustrated delivery a listener would notice can come out as neutral 0.45, sad 0.40.
+`VoiceReaderParams(tone_bias={"sad": 0.5, "frustrated": 0.5})` adds to those tones' logits before the softmax
+(+0.5 multiplies a tone's odds by about 1.65), so they are published more readily; our test harness runs with
+exactly that. The graded `sad` and `upset` readings are not affected by it. Leave it unset to read the calibrated
+probabilities.
+
 ## Evaluation
 
 Labels on real speech come from two audio LLM judges (Gemini 3.1 Pro preview and OpenAI gpt-audio-1.5) hearing
@@ -111,28 +137,33 @@ speakers never appear in training.
 
 | | `base` | `tiny` |
 |---|---|---|
-| **Sarcasm**, everyday conversation (4,366 turns, 203 sarcastic; Open Yap, otoSpeech, HarperValley), AUROC | 0.917 | 0.904 |
-| — at the release threshold: F1 / recall / false alarms on sincere turns | 0.575 / 0.50 / 1.3 % | 0.560 / 0.48 / 1.3 % |
-| — mean AUROC over four test sets (two conversational, Emolia, MUStARD++) | 0.791 | 0.788 |
-| — MUStARD++ (TV, audio only, evaluation only), AUROC | 0.656 | 0.654 |
-| **Tone** macro-F1: conversation / Emolia / CREMA-D (acted) / EmoNet-Voice (synthetic) | 0.706 / 0.661 / 0.887 / 0.563 | 0.712 / 0.627 / 0.872 / 0.457 |
-| **Sadness**: tone recall for sad, mean of four sets | 0.53 | 0.38 |
-| — `sad` output, AUROC: conversation (any judge heard sadness) / Emolia | 0.762 / 0.929 | – |
-| **Upset**, mean AUROC (three sets) | 0.867 | 0.856 |
+| **Sarcasm**, everyday conversation (3,945 turns both judges rated, 203 sarcastic; Open Yap, otoSpeech, HarperValley), AUROC | 0.921 | 0.908 |
+| — at the release threshold: F1 / recall / false alarms on sincere turns | 0.571 / 0.47 / 1.0 % | 0.548 / 0.45 / 1.1 % |
+| — second conversational test, held out from the rc2 judging round (5,037 turns, 272 sarcastic), AUROC | 0.862 | 0.853 |
+| — mean AUROC over four test sets (two conversational, Emolia, MUStARD++) | 0.798 | 0.797 |
+| — MUStARD++ (TV, audio only, evaluation only), AUROC | 0.669 | 0.662 |
+| **Tone** macro-F1: conversation / Emolia / CREMA-D (acted) / EmoNet-Voice (synthetic) | 0.754 / 0.667 / 0.902 / 0.510 | 0.739 / 0.647 / 0.873 / 0.477 |
+| **Sadness**: tone recall for sad, mean of four sets | 0.54 | 0.45 |
+| — `sad` output, AUROC: conversation (any judge heard sadness) / Emolia | 0.755 / 0.928 | 0.740 / 0.923 |
+| **Upset**, mean AUROC (three sets) | 0.869 | 0.857 |
 | **Arousal**, CCC against the training target | 0.89 | 0.88 |
-| **Laughter**, AUROC / F1 | 0.994 / 0.72 | 0.991 / 0.70 |
-| int8 vs fp32: tone agreement / sarcasm p95 \|Δp\| | 0.983 / 0.011 | 0.977 / 0.010 |
+| **Laughter**, AUROC / F1 | 0.993 / 0.64 | 0.991 / 0.64 |
+| int8 vs fp32: tone agreement / sarcasm p95 \|Δp\| | 0.983 / 0.008 | 0.983 / 0.009 |
 
 ## Limitations
 
 - **Sarcasm from the voice alone is a weak cue.** At the threshold it catches about half of the sarcasm both
-  judges heard, at about 1 false alarm in 75 sincere turns. Weigh it with the words, never alone. On synthetic
-  (TTS) sarcasm it is near chance (AUROC 0.60 on our 20 demo pairs).
+  judges heard, at about 1 false alarm in 100 sincere turns. Weigh it with the words, never alone. On synthetic
+  (TTS) sarcasm it is near chance (AUROC 0.57 on our 20 demo pairs).
 - **Sadness in conversation is rare** (2–6 % of turns), so even a good detector's per-turn flags are mostly
-  wrong (12–25 % precision at a 5 % false-alarm rate). Treat `sad` as a graded cue or average it over a
+  wrong (12–24 % precision at a 5 % false-alarm rate). Treat `sad` as a graded cue or average it over a
   conversation. It is not a measure of depression or of anyone's mental health, and a single turn cannot show
   either.
-- **Acted, theatrical sadness** is under-read: 23 of 34 CREMA-D sad clips.
+- **Borderline tones lean neutral.** Calibrated on conversation, the tone head publishes neutral on turns a
+  listener might call sad or frustrated; `tone_bias` above is the knob. Acted sadness is read well (30 of 34
+  CREMA-D sad clips by `base`, 27 by `tiny`; the misses read as frustrated).
+- **Laughter** ranks well but its published value is conservative: F1 0.64 at the calibrated 0.5 (rc1: 0.72), at
+  the same AUROC.
 - English only; trained on conversational and podcast speech, 8 kHz telephone audio included.
 - Tone labels are LLM judgments of the audio, not human ratings.
 
@@ -142,14 +173,16 @@ speakers never appear in training.
   [`laion/whisper-tiny-base-emotion-voice-burst`](https://huggingface.co/laion/whisper-tiny-base-emotion-voice-burst)
   (CC BY 4.0), base and tiny encoders respectively, cut to an 8 s window, with masked statistics pooling and one
   small head per output.
-- **Data** (about 119k training windows): Open Yap 1K and otoSpeech conversations, HarperValleyBank calls
-  (CC BY 4.0), the English YODAS part of Emilia from `laion/Emolia` (CC BY 4.0), and CREMA-D (ODbL / DbCL).
-  Augmented with gain, noise, telephone band-limiting and codecs; never speed or pitch, since rate and pitch are
-  the signal.
-- **Labels**: the two judges' verdicts on sarcasm, tone and upset; where they disagreed on tone, the mean of two
-  audio emotion models (emotion2vec+ large and the LAION model) at half weight; the `sad` head on the share of
-  judges who heard sadness.
-- **Tiny**: distilled from the base model's predictions at half weight, with base-labelled Emolia clips added.
+- **Data** (about 290k training windows): about 131k judged windows from Open Yap 1K and otoSpeech conversations
+  (two judging rounds), HarperValleyBank calls (CC BY 4.0), the English YODAS part of Emilia from `laion/Emolia`
+  (CC BY 4.0) and CREMA-D (ODbL / DbCL); plus 159k windows pseudo-labelled by the rc1 base model, 112,996 Emolia
+  clips no judge has rated and 46,021 conversation turns. Augmented with gain, noise, telephone band-limiting and
+  codecs; never speed or pitch, since rate and pitch are the signal.
+- **Labels**: the two judges' verdicts on sarcasm, tone and upset; where they disagreed on tone, their vote share
+  over the four tones; the `sad` head on the share of judges who heard sadness. No audio emotion model's output is
+  a direct target in rc2. The pseudo-labelled windows carry rc1's readings, and rc1 did use two such models
+  (emotion2vec+ large and the LAION model) for its disagreed turns.
+- **Tiny**: distilled from the rc2 base model's predictions at half weight, on the same data.
 - **Calibration** on real conversational dev turns: temperature for tone, Platt for sarcasm and laughter; the
   sarcasm threshold keeps false alarms at 2 % of sincere dev turns.
 

config.json

@@ -0,0 +1,64 @@
+{
+ "model": "voice-reader",
+ "release": "rc2",
+ "staged": "2026-11-04T17:14:31+00:00",
+ "window_secs": 8.0,
+ "sample_rate": 16000,
+ "tones": [
+  "positive",
+  "neutral",
+  "sad",
+  "frustrated"
+ ],
+ "outputs": [
+  "tone",
+  "sarcasm",
+  "laughter",
+  "arousal",
+  "speech_rate",
+  "upset",
+  "sad"
+ ],
+ "variants": {
+  "base": {
+   "folder": "voice-reader-base",
+   "run": "v4-base-all-sad",
+   "init": "laion-base",
+   "sarcasm_tau": 0.060534259700932624,
+   "files": {
+    "model.onnx": {
+     "sha256": "8a6aa25f4a0ce43d7e1289a15c0dfa894a8848ee1fbba682c38869782d19ded9",
+     "bytes": 109728612
+    },
+    "model.int8.onnx": {
+     "sha256": "8dfd1f1e271d4d64b0d001220ffb7ebd28935cef630377ac37fb9ce939de1218",
+     "bytes": 28567980
+    },
+    "config.json": {
+     "sha256": "e76664925738674626c87ae92d475a5dcf373d74ce87737ff61b869bef644a84",
+     "bytes": 661
+    }
+   }
+  },
+  "tiny": {
+   "folder": "voice-reader-tiny",
+   "run": "student-tiny-v4-sad",
+   "init": "laion-tiny",
+   "sarcasm_tau": 0.06329724596107583,
+   "files": {
+    "model.onnx": {
+     "sha256": "ad95fc8ae3370f98b91630bc4a12b71aec43d0d652f09b37dfed085e5238a833",
+     "bytes": 53366574
+    },
+    "model.int8.onnx": {
+     "sha256": "31d06f9a42656a3144997efa1f085abe29bbcbf20722e5711ba34d86f430dcbf",
+     "bytes": 14140364
+    },
+    "config.json": {
+     "sha256": "a32094d77c53f0fda50cd5328d4f4fc937535ffe1bfaefba760cb1771b1fb60d",
+     "bytes": 665
+    }
+   }
+  }
+ }
+}

voice-reader-base/MANIFEST.json

@@ -1,46 +1,46 @@
 {
- "release": "rc1",
+ "release": "rc2",
  "folder": "voice-reader-base",
- "run": "v3-base-sad",
- "staged": "2026-11-02T20:40:58+00:00",
+ "run": "v4-base-all-sad",
+ "staged": "2026-11-04T17:14:31+00:00",
  "files": {
   "model.onnx": {
-   "sha256": "99975890aa7f1b0648af82a480233fdf641cee04bfee84370d0bed9c39d63313",
+   "sha256": "8a6aa25f4a0ce43d7e1289a15c0dfa894a8848ee1fbba682c38869782d19ded9",
    "bytes": 109728612
   },
   "model.int8.onnx": {
-   "sha256": "5c25605ea019521853c92347b48ced6f7706c14f21183f5de9da2a607c9fae95",
+   "sha256": "8dfd1f1e271d4d64b0d001220ffb7ebd28935cef630377ac37fb9ce939de1218",
    "bytes": 28567980
   },
   "config.json": {
-   "sha256": "72ed694b3ca24b368baa0d68e47fa90ccf271bf0d74895bff2a20aefb53cd165",
-   "bytes": 658
+   "sha256": "e76664925738674626c87ae92d475a5dcf373d74ce87737ff61b869bef644a84",
+   "bytes": 661
   }
  },
  "parity_int8": {
   "windows": 2000,
-  "tone_argmax_agreement": 0.983,
-  "tone_max_abs_dp": 0.14630481600761414,
-  "sarcasm_p95_abs_dp": 0.011146726086735725,
-  "sarcasm_max_abs_dp": 0.053686708211898804,
-  "laughter_p95_abs_dp": 0.0013878615573048592,
-  "arousal_p95_abs_d": 0.026246830821037292,
-  "speech_rate_p95_rel_d": 0.04056451842188835,
+  "tone_argmax_agreement": 0.9825,
+  "tone_max_abs_dp": 0.11225080490112305,
+  "sarcasm_p95_abs_dp": 0.007848226465284824,
+  "sarcasm_max_abs_dp": 0.05139094591140747,
+  "laughter_p95_abs_dp": 0.0011351238936185837,
+  "arousal_p95_abs_d": 0.01641533151268959,
+  "speech_rate_p95_rel_d": 0.024630198255181313,
   "auroc_sarcasm": {
-   "reference": 0.8036348918022429,
-   "other": 0.8078700047385879
+   "reference": 0.8032301374190491,
+   "other": 0.8031314168377823
   },
   "auroc_sarcasm_both": {
-   "reference": 0.8481948120698782,
-   "other": 0.8499735309687665
+   "reference": 0.8373954473266279,
+   "other": 0.8395341450502911
   },
   "auroc_sarcasm_voice_only": {
-   "reference": 0.9116690614837782,
-   "other": 0.9187022612115253
+   "reference": 0.8842925206080314,
+   "other": 0.8799818498071542
   },
   "tone_macro_f1": {
-   "reference": 0.5146829335401363,
-   "other": 0.49449348337582855,
+   "reference": 0.5149164346694854,
+   "other": 0.5091877352207718,
    "rows": 1546
   }
  }

voice-reader-base/config.json

@@ -15,20 +15,20 @@
   "upset",
   "sad"
  ],
- "sarcasm_tau": 0.06803736669700389,
+ "sarcasm_tau": 0.060534259700932624,
  "calibration": {
   "sarcasm": {
-   "a": 1.678951463802178,
-   "b": 0.12539543584903914
+   "a": 1.7739810811041519,
+   "b": 0.311429803621114
   },
   "laughter": {
-   "a": 0.8057154083813632,
-   "b": -0.09714794868282932
+   "a": 0.8804135276963087,
+   "b": 0.22502338628323926
   },
-  "tone_temperature": 0.8488899224103036
+  "tone_temperature": 0.9473796186181745
  },
- "run": "v3-base-sad",
+ "run": "v4-base-all-sad",
  "init": "laion-base",
- "best_epoch": 7,
+ "best_epoch": 8,
  "note": "sarcasm_tau is the dev 2 % FPR threshold on the Platt-calibrated probability; tone temperature and laughter Platt curve fitted on dev"
 }

voice-reader-base/model.int8.onnx

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5c25605ea019521853c92347b48ced6f7706c14f21183f5de9da2a607c9fae95
+oid sha256:8dfd1f1e271d4d64b0d001220ffb7ebd28935cef630377ac37fb9ce939de1218
 size 28567980

voice-reader-base/model.onnx

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:99975890aa7f1b0648af82a480233fdf641cee04bfee84370d0bed9c39d63313
+oid sha256:8a6aa25f4a0ce43d7e1289a15c0dfa894a8848ee1fbba682c38869782d19ded9
 size 109728612

voice-reader-tiny/MANIFEST.json

@@ -1,46 +1,46 @@
 {
- "release": "rc1",
+ "release": "rc2",
  "folder": "voice-reader-tiny",
- "run": "student-tiny-v3",
- "staged": "2026-11-02T20:40:59+00:00",
+ "run": "student-tiny-v4-sad",
+ "staged": "2026-11-04T17:14:31+00:00",
  "files": {
   "model.onnx": {
-   "sha256": "873b5e5991f607483331ca7078b175dadaee3a1ec2046507ae65f008db64cad0",
-   "bytes": 50202707
+   "sha256": "ad95fc8ae3370f98b91630bc4a12b71aec43d0d652f09b37dfed085e5238a833",
+   "bytes": 53366574
   },
   "model.int8.onnx": {
-   "sha256": "1823e7d7e861586f3981ffa470150a5c3653f5d017df97a49e606ccc4e6c6c42",
-   "bytes": 13332966
+   "sha256": "31d06f9a42656a3144997efa1f085abe29bbcbf20722e5711ba34d86f430dcbf",
+   "bytes": 14140364
   },
   "config.json": {
-   "sha256": "79582d3816d963a092dd4d71da61d4444b3127659f846076223d20f240fc821d",
-   "bytes": 655
+   "sha256": "a32094d77c53f0fda50cd5328d4f4fc937535ffe1bfaefba760cb1771b1fb60d",
+   "bytes": 665
   }
  },
  "parity_int8": {
   "windows": 2000,
-  "tone_argmax_agreement": 0.977,
-  "tone_max_abs_dp": 0.2388080656528473,
-  "sarcasm_p95_abs_dp": 0.010020742192864418,
-  "sarcasm_max_abs_dp": 0.04528002440929413,
-  "laughter_p95_abs_dp": 0.0021592299453914165,
-  "arousal_p95_abs_d": 0.019707761704921722,
-  "speech_rate_p95_rel_d": 0.03602607175707817,
+  "tone_argmax_agreement": 0.983,
+  "tone_max_abs_dp": 0.21781721711158752,
+  "sarcasm_p95_abs_dp": 0.009449773468077183,
+  "sarcasm_max_abs_dp": 0.0660611093044281,
+  "laughter_p95_abs_dp": 0.003110053250566125,
+  "arousal_p95_abs_d": 0.019918425008654594,
+  "speech_rate_p95_rel_d": 0.023557225242257118,
   "auroc_sarcasm": {
-   "reference": 0.7930224293160638,
-   "other": 0.7871979150213236
+   "reference": 0.8067347180540199,
+   "other": 0.8055303269625651
   },
   "auroc_sarcasm_both": {
-   "reference": 0.8221916357861302,
-   "other": 0.819777660137639
+   "reference": 0.8388777130757015,
+   "other": 0.8375224986765485
   },
   "auroc_sarcasm_voice_only": {
-   "reference": 0.8642516826741284,
-   "other": 0.8534371927701732
+   "reference": 0.8949557589049384,
+   "other": 0.886863797927853
   },
   "tone_macro_f1": {
-   "reference": 0.45427391583654375,
-   "other": 0.46146068669445456,
+   "reference": 0.47394276820506326,
+   "other": 0.4619192915760122,
    "rows": 1546
   }
  }

voice-reader-tiny/config.json

@@ -12,22 +12,23 @@
   "laughter",
   "arousal",
   "speech_rate",
-  "upset"
+  "upset",
+  "sad"
  ],
- "sarcasm_tau": 0.06590602517797733,
+ "sarcasm_tau": 0.06329724596107583,
  "calibration": {
   "sarcasm": {
-   "a": 1.7731141168712765,
-   "b": 0.06321434642450725
+   "a": 1.778761607691192,
+   "b": 0.2934051131149342
   },
   "laughter": {
-   "a": 0.8303770152621825,
-   "b": -0.17392277961573782
+   "a": 0.9323411247145696,
+   "b": 0.17735405909388247
   },
-  "tone_temperature": 0.7673753909755566
+  "tone_temperature": 0.8733854440751974
  },
- "run": "student-tiny-v3",
+ "run": "student-tiny-v4-sad",
  "init": "laion-tiny",
- "best_epoch": 10,
+ "best_epoch": 15,
  "note": "sarcasm_tau is the dev 2 % FPR threshold on the Platt-calibrated probability; tone temperature and laughter Platt curve fitted on dev"
 }

voice-reader-tiny/model.int8.onnx

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1823e7d7e861586f3981ffa470150a5c3653f5d017df97a49e606ccc4e6c6c42
-size 13332966
+oid sha256:31d06f9a42656a3144997efa1f085abe29bbcbf20722e5711ba34d86f430dcbf
+size 14140364

voice-reader-tiny/model.onnx

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:873b5e5991f607483331ca7078b175dadaee3a1ec2046507ae65f008db64cad0
-size 50202707
+oid sha256:ad95fc8ae3370f98b91630bc4a12b71aec43d0d652f09b37dfed085e5238a833
+size 53366574