Download performance/optimizations.json from laion/humaneness-voice-small-studies: direct link, hf CLI and curl.
- Browser
- Download file 12.6 kB
-
https://huggingface.co/spaces/laion/humaneness-voice-small-studies/resolve/main/performance/optimizations.json
- Command line
-
hf download hf://spaces/laion/humaneness-voice-small-studies/performance/optimizations.json
-
curl -L -o optimizations.json https://huggingface.co/spaces/laion/humaneness-voice-small-studies/resolve/main/performance/optimizations.json
12.6 kB
| { | |
| "environment": { | |
| "date_utc": "2026-09-30T09:52:06Z", | |
| "gpu": "NVIDIA GeForce RTX 3090", | |
| "vram_gib": 23.68426513671875, | |
| "torch": "2.6.0+cu124", | |
| "cuda": "12.4", | |
| "transformers": "5.15.0", | |
| "python": "3.12.13", | |
| "model": "laion/Humaneness-Voice-Small", | |
| "revision": "5de86032771c37a3008d14e78ee7cdb39368e2a8", | |
| "stage": "S3", | |
| "codec_revision": "f6e20e543b33d2c252a7ef71bdf8aa71e5ff9169", | |
| "tts_dtype": "bfloat16", | |
| "codec_baseline_weights": "float32", | |
| "codec_compute": "bfloat16", | |
| "frames": 100, | |
| "fixed_audio_seconds_per_stream": 8.0, | |
| "cfg": 1.5, | |
| "temperature": 1.0, | |
| "reference_frames": 36, | |
| "caption": { | |
| "text": "The train was late, but we made it home before the rain started. I never thought we would be this lucky.", | |
| "caption": "Warm, natural, softly relieved conversation.", | |
| "general": "Warm natural conversation." | |
| }, | |
| "gpu_isolated": true, | |
| "server_stopped": true, | |
| "gpu_processes": "4181376, 12542 MiB" | |
| }, | |
| "notes": [ | |
| "Codec encoder/decoder weights BF16, quantizer unchanged FP32, using codec's supported setter" | |
| ], | |
| "trials": [ | |
| { | |
| "variant": "legacy", | |
| "n": 1, | |
| "streaming": false, | |
| "chunk_frames": null, | |
| "guidance": 1.5, | |
| "seed": 777, | |
| "fixed_length": true, | |
| "requested_frames": 100, | |
| "audio_seconds": [ | |
| 8.0 | |
| ], | |
| "aggregate_audio_seconds": 8.0, | |
| "device_seconds": 7.8813369140625, | |
| "model_wall_seconds": 7.881344620138407, | |
| "device_rtf": 0.9851671142578124, | |
| "per_stream_rtf": 0.9851671142578124, | |
| "audio_seconds_per_second": 1.0150562128267568, | |
| "token_loop_seconds": 7.7421865234375, | |
| "codec_seconds": 0.1391503906249998, | |
| "first_audio_seconds": 7.8813369140625, | |
| "chunks": [], | |
| "max_inter_chunk_seconds": null, | |
| "peak_allocated_gib": 11.633395195007324, | |
| "peak_reserved_gib": 12.279296875, | |
| "graph_capture_seconds": 0.0, | |
| "codes_sha256": "c3c092d0564717204edf46d466720e781850c7c60cbbfd6bf62436341ec14c3e", | |
| "status": "ok" | |
| }, | |
| { | |
| "variant": "legacy", | |
| "n": 1, | |
| "streaming": false, | |
| "chunk_frames": null, | |
| "guidance": 1.5, | |
| "seed": 778, | |
| "fixed_length": true, | |
| "requested_frames": 100, | |
| "audio_seconds": [ | |
| 8.0 | |
| ], | |
| "aggregate_audio_seconds": 8.0, | |
| "device_seconds": 7.809603515625, | |
| "model_wall_seconds": 7.8096122071146965, | |
| "device_rtf": 0.976200439453125, | |
| "per_stream_rtf": 0.976200439453125, | |
| "audio_seconds_per_second": 1.0243797888067, | |
| "token_loop_seconds": 7.67300927734375, | |
| "codec_seconds": 0.1365942382812495, | |
| "first_audio_seconds": 7.809603515625, | |
| "chunks": [], | |
| "max_inter_chunk_seconds": null, | |
| "peak_allocated_gib": 11.63448429107666, | |
| "peak_reserved_gib": 12.279296875, | |
| "graph_capture_seconds": 0.0, | |
| "codes_sha256": "364daa7e3388e823da35a8602a83df3a0c53110c5d837e95153811d2c50c298c", | |
| "status": "ok" | |
| }, | |
| { | |
| "variant": "legacy", | |
| "n": 1, | |
| "streaming": false, | |
| "chunk_frames": null, | |
| "guidance": 1.5, | |
| "seed": 779, | |
| "fixed_length": true, | |
| "requested_frames": 100, | |
| "audio_seconds": [ | |
| 8.0 | |
| ], | |
| "aggregate_audio_seconds": 8.0, | |
| "device_seconds": 7.808337890625, | |
| "model_wall_seconds": 7.80834810808301, | |
| "device_rtf": 0.976042236328125, | |
| "per_stream_rtf": 0.976042236328125, | |
| "audio_seconds_per_second": 1.024545826789222, | |
| "token_loop_seconds": 7.67256396484375, | |
| "codec_seconds": 0.13577392578125025, | |
| "first_audio_seconds": 7.808337890625, | |
| "chunks": [], | |
| "max_inter_chunk_seconds": null, | |
| "peak_allocated_gib": 11.63448429107666, | |
| "peak_reserved_gib": 12.279296875, | |
| "graph_capture_seconds": 0.0, | |
| "codes_sha256": "60deacd127640f7b77fc15bacf400838541bfd94cceca012dbe1d7092898cf51", | |
| "status": "ok" | |
| }, | |
| { | |
| "variant": "fused", | |
| "n": 1, | |
| "streaming": false, | |
| "chunk_frames": null, | |
| "guidance": 1.5, | |
| "seed": 777, | |
| "fixed_length": true, | |
| "requested_frames": 100, | |
| "audio_seconds": [ | |
| 8.0 | |
| ], | |
| "aggregate_audio_seconds": 8.0, | |
| "device_seconds": 4.34336962890625, | |
| "model_wall_seconds": 4.343383330851793, | |
| "device_rtf": 0.5429212036132812, | |
| "per_stream_rtf": 0.5429212036132812, | |
| "audio_seconds_per_second": 1.8418879081250483, | |
| "token_loop_seconds": 4.20710498046875, | |
| "codec_seconds": 0.13626464843749986, | |
| "first_audio_seconds": 4.34336962890625, | |
| "chunks": [], | |
| "max_inter_chunk_seconds": null, | |
| "peak_allocated_gib": 11.63448429107666, | |
| "peak_reserved_gib": 12.283203125, | |
| "graph_capture_seconds": 0.0, | |
| "codes_sha256": "be13aa216e31610c6f904374eb171a6b43ac92f260e8c2cca6710ee5092df5ca", | |
| "status": "ok" | |
| }, | |
| { | |
| "variant": "fused", | |
| "n": 1, | |
| "streaming": false, | |
| "chunk_frames": null, | |
| "guidance": 1.5, | |
| "seed": 778, | |
| "fixed_length": true, | |
| "requested_frames": 100, | |
| "audio_seconds": [ | |
| 8.0 | |
| ], | |
| "aggregate_audio_seconds": 8.0, | |
| "device_seconds": 4.34287841796875, | |
| "model_wall_seconds": 4.342893362045288, | |
| "device_rtf": 0.5428598022460938, | |
| "per_stream_rtf": 0.5428598022460938, | |
| "audio_seconds_per_second": 1.8420962389598183, | |
| "token_loop_seconds": 4.207283203125, | |
| "codec_seconds": 0.13559521484375026, | |
| "first_audio_seconds": 4.34287841796875, | |
| "chunks": [], | |
| "max_inter_chunk_seconds": null, | |
| "peak_allocated_gib": 11.63448429107666, | |
| "peak_reserved_gib": 12.283203125, | |
| "graph_capture_seconds": 0.0, | |
| "codes_sha256": "f1e51d34ee91a1add00bed36281b0d049c1a1a6190a38c4a51ae6491b2435488", | |
| "status": "ok" | |
| }, | |
| { | |
| "variant": "fused", | |
| "n": 1, | |
| "streaming": false, | |
| "chunk_frames": null, | |
| "guidance": 1.5, | |
| "seed": 779, | |
| "fixed_length": true, | |
| "requested_frames": 100, | |
| "audio_seconds": [ | |
| 8.0 | |
| ], | |
| "aggregate_audio_seconds": 8.0, | |
| "device_seconds": 4.34487890625, | |
| "model_wall_seconds": 4.3448955453932285, | |
| "device_rtf": 0.54310986328125, | |
| "per_stream_rtf": 0.54310986328125, | |
| "audio_seconds_per_second": 1.8412480928967203, | |
| "token_loop_seconds": 4.209115234375, | |
| "codec_seconds": 0.13576367187499994, | |
| "first_audio_seconds": 4.34487890625, | |
| "chunks": [], | |
| "max_inter_chunk_seconds": null, | |
| "peak_allocated_gib": 11.63448429107666, | |
| "peak_reserved_gib": 12.283203125, | |
| "graph_capture_seconds": 0.0, | |
| "codes_sha256": "3bc12c4f23348191fc98801ce233551322739a7deeba75e793637265ccb2e686", | |
| "status": "ok" | |
| }, | |
| { | |
| "variant": "graph", | |
| "n": 1, | |
| "streaming": false, | |
| "chunk_frames": null, | |
| "guidance": 1.5, | |
| "seed": 777, | |
| "fixed_length": true, | |
| "requested_frames": 100, | |
| "audio_seconds": [ | |
| 8.0 | |
| ], | |
| "aggregate_audio_seconds": 8.0, | |
| "device_seconds": 3.307399169921875, | |
| "model_wall_seconds": 3.3074153251945972, | |
| "device_rtf": 0.4134248962402344, | |
| "per_stream_rtf": 0.4134248962402344, | |
| "audio_seconds_per_second": 2.4188190142736747, | |
| "token_loop_seconds": 3.1741552734375, | |
| "codec_seconds": 0.13324389648437496, | |
| "first_audio_seconds": 3.307399169921875, | |
| "chunks": [], | |
| "max_inter_chunk_seconds": null, | |
| "peak_allocated_gib": 11.64991283416748, | |
| "peak_reserved_gib": 12.30859375, | |
| "graph_capture_seconds": 0.0, | |
| "codes_sha256": "be13aa216e31610c6f904374eb171a6b43ac92f260e8c2cca6710ee5092df5ca", | |
| "status": "ok" | |
| }, | |
| { | |
| "variant": "graph", | |
| "n": 1, | |
| "streaming": false, | |
| "chunk_frames": null, | |
| "guidance": 1.5, | |
| "seed": 778, | |
| "fixed_length": true, | |
| "requested_frames": 100, | |
| "audio_seconds": [ | |
| 8.0 | |
| ], | |
| "aggregate_audio_seconds": 8.0, | |
| "device_seconds": 3.30958740234375, | |
| "model_wall_seconds": 3.3096047937870026, | |
| "device_rtf": 0.41369842529296874, | |
| "per_stream_rtf": 0.41369842529296874, | |
| "audio_seconds_per_second": 2.4172197399393776, | |
| "token_loop_seconds": 3.1765185546875, | |
| "codec_seconds": 0.13306884765624982, | |
| "first_audio_seconds": 3.30958740234375, | |
| "chunks": [], | |
| "max_inter_chunk_seconds": null, | |
| "peak_allocated_gib": 11.649073600769043, | |
| "peak_reserved_gib": 12.33203125, | |
| "graph_capture_seconds": 0.0, | |
| "codes_sha256": "f1e51d34ee91a1add00bed36281b0d049c1a1a6190a38c4a51ae6491b2435488", | |
| "status": "ok" | |
| }, | |
| { | |
| "variant": "graph", | |
| "n": 1, | |
| "streaming": false, | |
| "chunk_frames": null, | |
| "guidance": 1.5, | |
| "seed": 779, | |
| "fixed_length": true, | |
| "requested_frames": 100, | |
| "audio_seconds": [ | |
| 8.0 | |
| ], | |
| "aggregate_audio_seconds": 8.0, | |
| "device_seconds": 3.311110107421875, | |
| "model_wall_seconds": 3.311127882450819, | |
| "device_rtf": 0.4138887634277344, | |
| "per_stream_rtf": 0.4138887634277344, | |
| "audio_seconds_per_second": 2.416108114939442, | |
| "token_loop_seconds": 3.178031005859375, | |
| "codec_seconds": 0.13307910156250014, | |
| "first_audio_seconds": 3.311110107421875, | |
| "chunks": [], | |
| "max_inter_chunk_seconds": null, | |
| "peak_allocated_gib": 11.648791313171387, | |
| "peak_reserved_gib": 12.33203125, | |
| "graph_capture_seconds": 0.0, | |
| "codes_sha256": "3bc12c4f23348191fc98801ce233551322739a7deeba75e793637265ccb2e686", | |
| "status": "ok" | |
| }, | |
| { | |
| "variant": "graph_codec_bf16", | |
| "n": 1, | |
| "streaming": false, | |
| "chunk_frames": null, | |
| "guidance": 1.5, | |
| "seed": 777, | |
| "fixed_length": true, | |
| "requested_frames": 100, | |
| "audio_seconds": [ | |
| 8.0 | |
| ], | |
| "aggregate_audio_seconds": 8.0, | |
| "device_seconds": 3.319486572265625, | |
| "model_wall_seconds": 3.3195039369165897, | |
| "device_rtf": 0.41493582153320313, | |
| "per_stream_rtf": 0.41493582153320313, | |
| "audio_seconds_per_second": 2.410011255005565, | |
| "token_loop_seconds": 3.188209716796875, | |
| "codec_seconds": 0.13127685546875023, | |
| "first_audio_seconds": 3.319486572265625, | |
| "chunks": [], | |
| "max_inter_chunk_seconds": null, | |
| "peak_allocated_gib": 5.636284828186035, | |
| "peak_reserved_gib": 12.33203125, | |
| "graph_capture_seconds": 0.0, | |
| "codes_sha256": "be13aa216e31610c6f904374eb171a6b43ac92f260e8c2cca6710ee5092df5ca", | |
| "status": "ok" | |
| }, | |
| { | |
| "variant": "graph_codec_bf16", | |
| "n": 1, | |
| "streaming": false, | |
| "chunk_frames": null, | |
| "guidance": 1.5, | |
| "seed": 778, | |
| "fixed_length": true, | |
| "requested_frames": 100, | |
| "audio_seconds": [ | |
| 8.0 | |
| ], | |
| "aggregate_audio_seconds": 8.0, | |
| "device_seconds": 3.311013916015625, | |
| "model_wall_seconds": 3.3110322915017605, | |
| "device_rtf": 0.41387673950195314, | |
| "per_stream_rtf": 0.41387673950195314, | |
| "audio_seconds_per_second": 2.4161783075883174, | |
| "token_loop_seconds": 3.17962353515625, | |
| "codec_seconds": 0.13139038085937527, | |
| "first_audio_seconds": 3.311013916015625, | |
| "chunks": [], | |
| "max_inter_chunk_seconds": null, | |
| "peak_allocated_gib": 5.637994289398193, | |
| "peak_reserved_gib": 12.33203125, | |
| "graph_capture_seconds": 0.0, | |
| "codes_sha256": "f1e51d34ee91a1add00bed36281b0d049c1a1a6190a38c4a51ae6491b2435488", | |
| "status": "ok" | |
| }, | |
| { | |
| "variant": "graph_codec_bf16", | |
| "n": 1, | |
| "streaming": false, | |
| "chunk_frames": null, | |
| "guidance": 1.5, | |
| "seed": 779, | |
| "fixed_length": true, | |
| "requested_frames": 100, | |
| "audio_seconds": [ | |
| 8.0 | |
| ], | |
| "aggregate_audio_seconds": 8.0, | |
| "device_seconds": 3.315505126953125, | |
| "model_wall_seconds": 3.315530363470316, | |
| "device_rtf": 0.4144381408691406, | |
| "per_stream_rtf": 0.4144381408691406, | |
| "audio_seconds_per_second": 2.4129053322718947, | |
| "token_loop_seconds": 3.1838095703125, | |
| "codec_seconds": 0.13169555664062482, | |
| "first_audio_seconds": 3.315505126953125, | |
| "chunks": [], | |
| "max_inter_chunk_seconds": null, | |
| "peak_allocated_gib": 5.637994289398193, | |
| "peak_reserved_gib": 12.33203125, | |
| "graph_capture_seconds": 0.0, | |
| "codes_sha256": "3bc12c4f23348191fc98801ce233551322739a7deeba75e793637265ccb2e686", | |
| "status": "ok" | |
| } | |
| ] | |
| } |