Skip to content

Commit 5683a09

Browse files
committed
chore: Notion 동기화 2026-08-11 13:01
1 parent 7b97e50 commit 5683a09

13 files changed

Lines changed: 56 additions & 63 deletions

blog/.notion-sync.json

Lines changed: 22 additions & 22 deletions
Original file line numberDiff line numberDiff line change
@@ -29,14 +29,14 @@
2929
},
3030
"39ce15b4-0359-8051-a39e-d78c277e3c7e": {
3131
"file": "blog/2026-07-13-07-langgraph-기반-방송-워크플로우-설계.md",
32-
"last_edited": "2026-07-21T05:39:00.000Z",
33-
"content_hash": "482a6a3c97491c317fa897f1f101b3db8855e1834d4730f0004062363f7ab7b7",
32+
"last_edited": "2026-07-21T08:11:00.000Z",
33+
"content_hash": "28743b88132f9f2a2edb876606894e77a7d81dad5c691cbeedd88b0e78c039c5",
3434
"order": 5,
3535
"parent_id": null
3636
},
3737
"39ce15b4-0359-80cc-afdb-fad276333f56": {
3838
"file": "blog/2026-07-13-07-rag-파이프라인-구축.md",
39-
"last_edited": "2026-07-20T04:22:00.000Z",
39+
"last_edited": "2026-07-23T01:58:00.000Z",
4040
"content_hash": "1eb3f10e3cc805e566149a0ab2278f467be713c18e1ccd3354ff14784c973d66",
4141
"order": 6,
4242
"parent_id": null
@@ -57,36 +57,36 @@
5757
},
5858
"39de15b4-0359-80eb-b617-db767eb5277f": {
5959
"file": "blog/2026-07-14-11-방송-ai-오픈웨이트-모델-활용-사례-1건.md",
60-
"last_edited": "2026-07-21T05:42:00.000Z",
61-
"content_hash": "0115e9be2dcae44cad52368b480b6e699c14aca28b53d74287bbb0a04d820925",
60+
"last_edited": "2026-07-23T01:17:00.000Z",
61+
"content_hash": "552a94476aea86f4b9f9dbf43a26acb6f48009e26563edc99b8b8d5a31ac9a65",
6262
"order": 9,
6363
"parent_id": null
6464
},
6565
"39de15b4-0359-80c8-85de-f9807821f73d": {
6666
"file": "blog/2026-07-14-09-vllm-기본-추론-최적화-노하우-1건.md",
67-
"last_edited": "2026-07-20T07:25:00.000Z",
68-
"content_hash": "852c084ddc9bd569b6a7e02e5379a3d2ab4c14466bcb4dd71571a6dc88b1ca0d",
67+
"last_edited": "2026-07-23T02:15:00.000Z",
68+
"content_hash": "7922d762e1c215af60e5249a6849e81d35d7b12c7975a7d6d50238e743de4e77",
6969
"order": 10,
7070
"parent_id": null
7171
},
7272
"39de15b4-0359-80b7-ae6a-da99070db110": {
7373
"file": "blog/2026-07-14-09-맥락-기반-광고-최적화-기술-해설-1건.md",
74-
"last_edited": "2026-07-20T07:24:00.000Z",
75-
"content_hash": "a0cc47090e003c0194dcf105e60be1cf88a0ebe9bafd55320b393ff78dc4d3d4",
74+
"last_edited": "2026-07-21T08:17:00.000Z",
75+
"content_hash": "a6fd4e6fc8587bfb7378794a9111f03236dce8cb20ceca628c694da335cd6cfe",
7676
"order": 11,
7777
"parent_id": null
7878
},
7979
"39de15b4-0359-8096-8fe0-d8cf81d3e397": {
8080
"file": "blog/2026-07-14-09-db-스키마-공개.md",
81-
"last_edited": "2026-07-21T05:34:00.000Z",
81+
"last_edited": "2026-07-21T08:05:00.000Z",
8282
"content_hash": "a0e3af442d6a2053402a3e161af1a44724fe96e31b9824152253f89d9ed49de0",
8383
"order": 12,
8484
"parent_id": null
8585
},
8686
"39de15b4-0359-80cf-bde8-f114c7d20501": {
8787
"file": "blog/2026-07-14-10-sbsi-실증-테스트-사례-공유-1건.md",
88-
"last_edited": "2026-07-21T05:42:00.000Z",
89-
"content_hash": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
88+
"last_edited": "2026-07-24T01:10:00.000Z",
89+
"content_hash": "b96656c0195fab5e3376d12c76491ab8cf445b94c3cab92c7c69e6a222355584",
9090
"order": 13,
9191
"parent_id": null
9292
},
@@ -99,8 +99,8 @@
9999
},
100100
"39de15b4-0359-800d-be84-c4e206eda79e": {
101101
"file": "blog/2026-07-14-11-오픈-소스-기여-종합-가이드-1건.md",
102-
"last_edited": "2026-07-21T05:39:00.000Z",
103-
"content_hash": "6d2bcad61622158a844e817760e6ee934002c80a376507b3e832e8b11691d070",
102+
"last_edited": "2026-07-21T08:22:00.000Z",
103+
"content_hash": "f6745dc1d24b7600bada576f543ad452e90c1655ef589112fa344d7071f918f8",
104104
"order": 15,
105105
"parent_id": null
106106
},
@@ -112,23 +112,23 @@
112112
"parent_id": null
113113
},
114114
"39de15b4-0359-8088-be18-d31d76a03c9a": {
115-
"file": "blog/2026-07-14-08-숏폼-자동-생성-파이프라인-해설-1.md",
116-
"last_edited": "2026-07-21T06:56:00.000Z",
117-
"content_hash": "779313c823feac82c950b7310edaf34a37f194de977fad0a87d18010a533978a",
115+
"file": "blog/2026-07-14-08-숏폼-자동-생성-파이프라인-해설-2.md",
116+
"last_edited": "2026-08-11T01:17:00.000Z",
117+
"content_hash": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
118118
"order": 17,
119119
"parent_id": null
120120
},
121121
"39de15b4-0359-801a-95e2-cf6328c4280c": {
122-
"file": "blog/2026-07-14-08-숏폼-자동-생성-파이프라인-해설-2.md",
123-
"last_edited": "2026-07-20T07:25:00.000Z",
124-
"content_hash": "c084eb174c5fab3a4123d990ba97c1dd3bd249ee124f1bf3141b7f064838a823",
122+
"file": "blog/2026-07-14-08-숏폼-자동-생성-파이프라인-해설-1.md",
123+
"last_edited": "2026-08-11T01:20:00.000Z",
124+
"content_hash": "ce8ce8793016791808f0bab2eed394995a0971f818d41b65eb74aade21adea0f",
125125
"order": 18,
126126
"parent_id": null
127127
},
128128
"39de15b4-0359-8002-bed8-f03d0d4763b1": {
129129
"file": "blog/2026-07-14-08-모아보기-서비스-구현기-1건.md",
130-
"last_edited": "2026-07-21T05:42:00.000Z",
131-
"content_hash": "b2df166a77444fe850633aed5cbe1d15cc79e9ef886936297096a2b66c104fcf",
130+
"last_edited": "2026-07-21T08:17:00.000Z",
131+
"content_hash": "7f8c4d1c556860be4c0e7c3813787ffa53ff6f76c904b747565f496120f99b8e",
132132
"order": 19,
133133
"parent_id": null
134134
},

blog/2026-07-13-07-langgraph-기반-방송-워크플로우-설계.md

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -312,7 +312,7 @@ BUDGET_TOLERANCE = 1.15 # target_duration_sec 초과 허용 배율
312312
313313
노드별 중간 결과는 `output/<req_id>/<node>.json` 으로 덤프된다.
314314
315-
#### 마무리
315+
### 마무리
316316
317317
---
318318

blog/2026-07-13-07-rag-파이프라인-구축.md

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -6,7 +6,7 @@ authors: [sbin]
66
description: "영상을 벡터로 접어 넣고(색인) 자연어 질의로 클립을 꺼내는(검색) RAG — 타임코드가 대칭축인 파이프라인 구축기"
77
tags: [rag, milvus]
88
last_update:
9-
date: 2026-07-20
9+
date: 2026-07-23
1010
---
1111

1212
### 들어가며

blog/2026-07-14-08-모아보기-서비스-구현기-1건.md

Lines changed: 1 addition & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -12,13 +12,12 @@ last_update:
1212

1313
- 모아보기 서비스에 대한 간단한 소개
1414
- 모아보기 관련 코드 작성 (compilation)
15+
`service=compilation`  + `preset=감성_몽타주`  가 **텍스트로**  LLM에 전달
1516

1617
<!--truncate-->
1718

18-
`service=compilation`  + `preset=감성_몽타주`  가 **텍스트로**  LLM에 전달
1919

2020

21-
2221
### 모아보기
2322

2423
:::note

blog/2026-07-14-08-숏폼-자동-생성-파이프라인-해설-1.md

Lines changed: 3 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -1,18 +1,17 @@
11
---
22
title: "08_숏폼 자동 생성 파이프라인 해설_1"
33
date: 2026-07-14T12:08:00
4-
slug: 17
5-
authors: [sbin]
4+
slug: 18
65
tags: [pipeline, shortform]
76
last_update:
8-
date: 2026-07-21
7+
date: 2026-08-11
98
---
109

1110
### 들어가며
1211

1312
---
1413

15-
- 숏폼 자동 생성 서비스에 대한 간단한 소개이다. (pinpoint)
14+
- 숏폼 자동 생성 서비스에 대한 간단한 소개이다.
1615
- ffmpeg 이용 커팅 되는 코드 부분 구현
1716
- 9/16, 16/9
1817

blog/2026-07-14-08-숏폼-자동-생성-파이프라인-해설-2.md

Lines changed: 0 additions & 24 deletions
This file was deleted.

blog/2026-07-14-09-db-스키마-공개.md

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -3,6 +3,7 @@ title: "09_DB 스키마 공개"
33
date: 2026-07-14T12:09:00
44
slug: 12
55
authors: [sbin]
6+
tags: [업데이트 필요]
67
last_update:
78
date: 2026-07-21
89
---

blog/2026-07-14-09-vllm-기본-추론-최적화-노하우-1건.md

Lines changed: 3 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -4,7 +4,7 @@ date: 2026-07-14T12:09:00
44
slug: 10
55
tags: [vllm, inference]
66
last_update:
7-
date: 2026-07-20
7+
date: 2026-07-23
88
---
99

1010
### 들어가며
@@ -15,6 +15,8 @@ last_update:
1515

1616
<!--truncate-->
1717

18+
/code
19+
1820
### 마무리
1921

2022
---

blog/2026-07-14-09-맥락-기반-광고-최적화-기술-해설-1건.md

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -3,7 +3,7 @@ title: "09_맥락 기반 광고 최적화 기술 해설 1건"
33
date: 2026-07-14T12:09:00
44
slug: 11
55
last_update:
6-
date: 2026-07-20
6+
date: 2026-07-21
77
---
88

99
### 들어가며

blog/2026-07-14-10-sbsi-실증-테스트-사례-공유-1건.md

Lines changed: 12 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -4,6 +4,17 @@ date: 2026-07-14T12:10:00
44
slug: 13
55
tags: [sbs]
66
last_update:
7-
date: 2026-07-21
7+
date: 2026-07-24
88
---
99

10+
11+
**뭐 하려는 거냐:** Real-ESRGAN이라는 AI 초해상도 모델을 쓰려는 겁니다. ffmpeg 스케일링은 픽셀을 늘려서 보간할 뿐이라 흐릿하게 커지기만 하는데, 이 모델은 저화질·압축된 영상에서 잃어버린 윤곽과 질감을 **복원하도록 학습된** 모델이라 결과가 확연히 다릅니다. 파이프라인은 이렇습니다:
12+
13+
<!--truncate-->
14+
15+
```text
16+
디코딩 → GPU 2장에 프레임 분배 → AI 4배 확대(640×360 → 2560×1440)
17+
→ 1920×1080으로 다운샘플(슈퍼샘플링 효과, 더 깔끔) → H.264 고비트레이트 재인코딩
18+
→ 원본 오디오 그대로 합치기
19+
```
20+

0 commit comments

Comments
 (0)