openvibers/openvibe.ai Commits
.env.example documents AI_MEDIA_STREAM (media.analyze streams through the loopback reader by default; 0 downloads first).
Changes
@@@ -79,6 +79,9 @@ AI_MODEL=# A local model server speaking the OpenAI API (llama.cpp llama-server, Ollama /v1): provider `local`, route# media.local, used by media.analyze's overview. No key; never a paid provider.#AI_LOCAL_LLM_URL=http://127.0.0.1:8090/v1++# media.analyze reads a recording where it lies through a loopback reader (server/media-proxy.js): nothing is++# downloaded, and speech windows seek with range requests. 0 downloads it first (the disk cap then applies).++#AI_MEDIA_STREAM=1#AI_LOCAL_LLM_MODEL=qwen2.5-1.5b-instruct#AI_LOCAL_LLM_TIMEOUT_MS=180000# media.analyze: file cap (also bounded by free temp disk minus the reserve), download timeout, scene threshold
@@ -79,6 +79,9 @@ AI_MODEL=
@@ -79,6 +79,9 @@ AI_MODEL=
# A local model server speaking the OpenAI API (llama.cpp llama-server, Ollama /v1): provider `local`, route
# A local model server speaking the OpenAI API (llama.cpp llama-server, Ollama /v1): provider `local`, route
# media.local, used by media.analyze's overview. No key; never a paid provider.
# media.local, used by media.analyze's overview. No key; never a paid provider.
#AI_LOCAL_LLM_URL=http://127.0.0.1:8090/v1
#AI_LOCAL_LLM_URL=http://127.0.0.1:8090/v1
+# media.analyze reads a recording where it lies through a loopback reader (server/media-proxy.js): nothing is
+# downloaded, and speech windows seek with range requests. 0 downloads it first (the disk cap then applies).
+#AI_MEDIA_STREAM=1
#AI_LOCAL_LLM_MODEL=qwen2.5-1.5b-instruct
#AI_LOCAL_LLM_MODEL=qwen2.5-1.5b-instruct
#AI_LOCAL_LLM_TIMEOUT_MS=180000
#AI_LOCAL_LLM_TIMEOUT_MS=180000
# media.analyze: file cap (also bounded by free temp disk minus the reserve), download timeout, scene threshold
# media.analyze: file cap (also bounded by free temp disk minus the reserve), download timeout, scene threshold