{"id":30701043,"url":"https://github.com/solaceharmony/solacelive","last_synced_at":"2026-05-08T07:31:36.566Z","repository":{"id":310945448,"uuid":"1041853126","full_name":"SolaceHarmony/SolaceLive","owner":"SolaceHarmony","description":"Solace Live: real-time packet streaming voice interface and services","archived":false,"fork":false,"pushed_at":"2025-09-01T10:20:50.000Z","size":23361,"stargazers_count":0,"open_issues_count":2,"forks_count":0,"subscribers_count":0,"default_branch":"master","last_synced_at":"2025-09-01T12:38:03.320Z","etag":null,"topics":["audio","react","realtime","solace","typescript","vite","websocket","whisper"],"latest_commit_sha":null,"homepage":null,"language":"TypeScript","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/SolaceHarmony.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2025-08-21T05:35:42.000Z","updated_at":"2025-09-01T10:20:54.000Z","dependencies_parsed_at":null,"dependency_job_id":"d2b71806-c568-4997-b66e-f00b3107d732","html_url":"https://github.com/SolaceHarmony/SolaceLive","commit_stats":null,"previous_names":["solaceharmony/solacelive"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/SolaceHarmony/SolaceLive","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/SolaceHarmony%2FSolaceLive","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/SolaceHarmony%2FSolaceLive/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/SolaceHarmony%2FSolaceLive/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/SolaceHarmony%2FSolaceLive/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/SolaceHarmony","download_url":"https://codeload.github.com/SolaceHarmony/SolaceLive/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/SolaceHarmony%2FSolaceLive/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":273287682,"owners_count":25078575,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","status":"online","status_checked_at":"2025-09-02T02:00:09.530Z","response_time":77,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["audio","react","realtime","solace","typescript","vite","websocket","whisper"],"created_at":"2025-09-02T13:01:46.989Z","updated_at":"2026-05-08T07:31:36.311Z","avatar_url":"https://github.com/SolaceHarmony.png","language":"TypeScript","funding_links":[],"categories":[],"sub_categories":[],"readme":"# SolaceLive - Packetized Speech-to-Speech (Next.js + MLX)\n\nThis project is migrating to a Next.js app with a packetized 80 ms speech-to-speech loop using an MLX-backed realtime server. For the current architecture and contracts, see ARCHITECTURE.md and PLAN.md.\n\nQuickstart (current stack)\n- Install deps: `npm -C next-server install`\n- Start packet WS server: `npm -C next-server run packet:server` (health: http://localhost:8788/health)\n- Start Next.js UI: `npm -C next-server run dev` (open: http://localhost:3000/packet-voice and /whisperx)\n- Smoke test: `npm -C next-server run smoke`\n\nKey environment variables\n- NEXT_PUBLIC_PACKET_WS (default ws://localhost:8788)\n- PACKET_SERVER_URL (override base URL for Next API tests if packet server runs on a non-default host/port)\n- HF_TOKEN (server-side for /api/hf proxy); NEXT_PUBLIC_HF_TOKEN (optional client-side, for public models only)\n- NEXT_PUBLIC_HF_MIRROR=/api/hf (recommended to mirror model assets via local proxy)\n- LM_REPO (HF repo id for LM safetensors) or LM_GGUF (absolute GGUF path for text-only LM mode)\n- MIMI_REPO (HF repo id for Mimi/tokenizer) or MIMI_LOCAL (absolute local safetensors path)\n- Notes: Model readiness visible at http://localhost:8788/weights; encode/decode require real Mimi; LM step requires loaded weights\n\nDocumentation\n- Architecture and runtime contracts: ARCHITECTURE.md\n- Execution plan and milestones: PLAN.md\n\n---\n\n## Additional docs\n- Architecture and runtime contracts: ARCHITECTURE.md\n- Execution plan and milestones: PLAN.md\n- CSM integration details: docs/CSM.md\n\n## Repository layout highlights\n- original_csm/: Reference copy of the original CSM code for study and integration planning.\n- models/: Contains local model artifacts. The .gguf file here corresponds to the included CSM model (used for LM text-only experiments in this repo).\n\n## Acknowledgments\n- Sesame AI Labs for CSM (Conversational Speech Model).\n- Kyutai for Moshi (speech–text) and Mimi (streaming codec).\n- Hugging Face for model hosting and Transformers ecosystem.\n\nNote: Any legacy content referencing LM Studio has been archived and is not part of this Next.js packetized implementation.\n\n## Profiling\nA lightweight LM profiling endpoint is available on the packet server to help measure per‑step generation latency.\n\n- Endpoint: GET http://localhost:8788/profile/lm\n- Query params:\n  - steps: number of timed steps (default 16; min 1; max 1000)\n  - warmup: number of warmup steps before timing (default 2; min 0; max 100)\n  - timer or debug: set to 1 (default) to include perStepMs array; 0 to omit\n\nExample:\n```\ncurl \"http://localhost:8788/profile/lm?steps=32\u0026warmup=4\u0026timer=1\"\n```\nResponse:\n```\n{\n  \"ok\": true,\n  \"steps\": 32,\n  \"warmup\": 4,\n  \"totalMs\": 420,\n  \"avgMs\": 13.1,\n  \"perStepMs\": [12, 13, ...],\n  \"tokens\": [123, 456, ...]\n}\n```\nNotes:\n- Returns 503 if the LM is not ready (weights not loaded).\n- In text‑only GGUF mode (audio_codebooks=0) the endpoint still works, feeding only the text channel.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsolaceharmony%2Fsolacelive","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fsolaceharmony%2Fsolacelive","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsolaceharmony%2Fsolacelive/lists"}