Add atomic composition controls and representative stages

2026-10-05 05:03:40 +09:00
parent 1e7c258631
commit 4c411e446d
56 changed files with 5911 additions and 1601 deletions
@@ -0,0 +1,117 @@
+{
+  "schemaVersion": 1,
+  "runId": "20261004T193323396066Z_c5daf2df",
+  "createdUtc": "2026-10-04T19:33:23.398573+00:00",
+  "modelRevision": "0c0e3051f131929182e2c023b9537f8b1c68adfe",
+  "catalogHash": "1c74dcfc4d67eb2583dac6168e164c242d7af349ef77e74f700426dc43c6a21f",
+  "profileHash": "a5740d9db8a4aaebf08d0ac3b8f3225693f878aa2868143f01191a059e2680e8",
+  "scope": "selected",
+  "regenerateRequested": true,
+  "inputSnapshots": {
+    "catalog": "catalog.json",
+    "profile": "profile.json"
+  },
+  "requestedIds": [
+    "guide_merge",
+    "guide_split",
+    "demo_composition"
+  ],
+  "environment": {
+    "python": "3.12.15 | packaged by Anaconda, Inc. | (main, Oct  1 2026, 20:14:50) [MSC v.1942 64 bit (AMD64)]",
+    "packages": {
+      "qwen-tts": "0.1.1",
+      "torch": "2.11.0+cu128",
+      "torchaudio": "2.11.0",
+      "transformers": "4.57.3",
+      "soundfile": "0.14.0",
+      "huggingface-hub": "0.36.2"
+    },
+    "device": "NVIDIA GeForce RTX 4080 SUPER",
+    "cuda": "12.8",
+    "modelDevice": "cuda:0",
+    "cudaProbe": 140.0,
+    "modelLoadSeconds": 6.364993199938908,
+    "modelPeakVramBytes": 4181138432,
+    "modelDefaults": {
+      "do_sample": true,
+      "repetition_penalty": 1.05,
+      "temperature": 0.9,
+      "top_p": 1.0,
+      "top_k": 50,
+      "subtalker_dosample": true,
+      "subtalker_temperature": 0.9,
+      "subtalker_top_p": 1.0,
+      "subtalker_top_k": 50,
+      "max_new_tokens": 8192
+    },
+    "effectiveGenerationParameters": {
+      "do_sample": true,
+      "top_k": 50,
+      "top_p": 1,
+      "temperature": 0.9,
+      "repetition_penalty": 1.05,
+      "subtalker_dosample": true,
+      "subtalker_top_k": 50,
+      "subtalker_top_p": 1,
+      "subtalker_temperature": 0.9,
+      "max_new_tokens": 8192
+    },
+    "peakVramBytes": 4395508224
+  },
+  "clips": [
+    {
+      "id": "guide_merge",
+      "textHash": "10114a172f0beb4b89e004a2b0d20525eb92e961a3e2f4199d7f1493d52bde07",
+      "settingsHash": "a5740d9db8a4aaebf08d0ac3b8f3225693f878aa2868143f01191a059e2680e8",
+      "originalRunId": "20261004T193323396066Z_c5daf2df",
+      "file": "generated/20261004T193323396066Z_c5daf2df/guide_merge.wav",
+      "review": "not_listened",
+      "sha256": "f521af71107190360187a11d20d0f95f4aac14ee73d79a75bb85018d1bfa164d",
+      "sampleRate": 24000,
+      "channels": 1,
+      "frames": 86400,
+      "duration": 3.6,
+      "peak": 0.38671875,
+      "rms": 0.05468548461794853,
+      "qualityWarnings": [],
+      "result": "generated",
+      "generationSeconds": 8.76143110007979
+    },
+    {
+      "id": "guide_split",
+      "textHash": "9bb7cdc1a8c9cbcb6d2ea132dd92e074289bebdb3835be0dc47cc971807353fa",
+      "settingsHash": "a5740d9db8a4aaebf08d0ac3b8f3225693f878aa2868143f01191a059e2680e8",
+      "originalRunId": "20261004T193323396066Z_c5daf2df",
+      "file": "generated/20261004T193323396066Z_c5daf2df/guide_split.wav",
+      "review": "not_listened",
+      "sha256": "9362a2243fd44b607a50fe4ca7f3a4d64a64e8ceb919b9eb6bb7098adaf67eac",
+      "sampleRate": 24000,
+      "channels": 1,
+      "frames": 128640,
+      "duration": 5.36,
+      "peak": 0.62890625,
+      "rms": 0.06911510229110718,
+      "qualityWarnings": [],
+      "result": "generated",
+      "generationSeconds": 14.070215299958363
+    },
+    {
+      "id": "demo_composition",
+      "textHash": "50b079d47c9bd03bf8b762790ca977271286264eda554c0217b3062fc0f6b560",
+      "settingsHash": "a5740d9db8a4aaebf08d0ac3b8f3225693f878aa2868143f01191a059e2680e8",
+      "originalRunId": "20261004T193323396066Z_c5daf2df",
+      "file": "generated/20261004T193323396066Z_c5daf2df/demo_composition.wav",
+      "review": "not_listened",
+      "sha256": "f1aa3029c780da9005ccf1f173a82025f17225d4ee0952e9a21a75aa5516821e",
+      "sampleRate": 24000,
+      "channels": 1,
+      "frames": 228480,
+      "duration": 9.52,
+      "peak": 0.6328125,
+      "rms": 0.06663522124290466,
+      "qualityWarnings": [],
+      "result": "generated",
+      "generationSeconds": 18.75810609990731
+    }
+  ]
+}