Rachit-Tw commited on
Commit
0dab1eb
·
1 Parent(s): 1b17e7e

fix: implement staggered throttling and sequential file processing for free tier

Browse files
Files changed (3) hide show
  1. index.ts +2 -2
  2. logs.txt +57 -29
  3. src/bot.ts +1 -1
index.ts CHANGED
@@ -63,8 +63,8 @@ export default (app: Probot, {getRouter}: any) => {
63
  process.env.MODEL_TEMPERATURE || '0.0',
64
  process.env.RETRIES || '3',
65
  process.env.TIMEOUT_MS || '120000',
66
- process.env.CONCURRENCY_LIMIT || '6',
67
- process.env.GITHUB_CONCURRENCY_LIMIT || '6',
68
  process.env.API_BASE_URL || 'https://api.groq.com/openai/v1',
69
  process.env.LANGUAGE || 'en-US',
70
  true,
 
63
  process.env.MODEL_TEMPERATURE || '0.0',
64
  process.env.RETRIES || '3',
65
  process.env.TIMEOUT_MS || '120000',
66
+ process.env.CONCURRENCY_LIMIT || '1',
67
+ process.env.GITHUB_CONCURRENCY_LIMIT || '1',
68
  process.env.API_BASE_URL || 'https://api.groq.com/openai/v1',
69
  process.env.LANGUAGE || 'en-US',
70
  true,
logs.txt CHANGED
@@ -1,9 +1,29 @@
1
- INFO (probot): Received pull_request.opened for PR #36
2
- Cloning Rachit-Tiwari-7/TEST-PRIX into /tmp/prix_tmp_3488822808_1775301716913...
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3
  Cloning Rachit-Tiwari-7/TEST-PRIX with depth 1...
4
  From https://github.com/Rachit-Tiwari-7/TEST-PRIX
5
- * branch 63ea3e628a06c997cc96d6d50a73586547336668 -> FETCH_HEAD
6
- Note: switching to '63ea3e628a06c997cc96d6d50a73586547336668'.
7
 
8
  You are in 'detached HEAD' state. You can look around, make experimental
9
  changes and commit them, and you can discard any commits you make in this
@@ -20,41 +40,49 @@ Or undo this operation with:
20
 
21
  Turn off this advice by setting config variable advice.detachedHead to false
22
 
23
- HEAD is now at 63ea3e6 Chceking PR creation
24
  Initialized AI client with model: llama-3.1-8b-instant
25
  Initialized AI client with model: llama-3.3-70b-versatile
26
- INFO (event): PRIX Audit starting for PR #36
27
- id: "7cace510-3018-11f1-95b6-e238a4e12a40"
28
  [Graph] Discovered: 1 TS/JS, 0 Go, 2 Python, 0 Rust, 0 C++/Java files
29
  [GraphCache] TS: 1/1 files need parsing
30
  [GraphCache] Go: All 0 files served from cache
31
  [GraphCache] Python: 2/2 files need parsing
32
  INFO (event): Will review from the base commit: 136795cc20e35ffa9a345771e8eeaaf4bce69ae2
33
- id: "7cace510-3018-11f1-95b6-e238a4e12a40"
34
  checking path: E.js => true
35
  checking path: e.py => true
36
  INFO (event): summarize: E.js
37
- id: "7cace510-3018-11f1-95b6-e238a4e12a40"
38
  INFO (event): summarize: e.py
39
- id: "7cace510-3018-11f1-95b6-e238a4e12a40"
40
- AI response time: 271 ms
41
  INFO (event): filename: E.js, triage: NEEDS_REVIEW
42
- id: "7cace510-3018-11f1-95b6-e238a4e12a40"
43
- AI response time: 299 ms
44
  INFO (event): filename: e.py, triage: NEEDS_REVIEW
45
- id: "7cace510-3018-11f1-95b6-e238a4e12a40"
46
- AI response time: 338 ms
47
- prixExec failed: "git grep -lE "import.*add_to_inventory|require.*add_to_inventory|add_to_inventory\." -- . ":(exclude)e.py"" in /tmp/prix_tmp_3488822808_1775301716913. Error: Command failed: git grep -lE "import.*add_to_inventory|require.*add_to_inventory|add_to_inventory\." -- . ":(exclude)e.py"
48
- AI response time: 3291 ms
49
- INFO (http): POST / 202 - 9001ms
50
- AI response time: 3211 ms
51
- AI response time: 3336 ms
52
- AI response time: 3449 ms
53
- AI response time: 6264 ms
54
- Attempt 1 failed. 3 retries left.
55
- Attempt 2 failed. 2 retries left.
56
- Attempt 3 failed. 1 retries left.
57
- AI response time: 17104 ms
58
- Attempt 1 failed. 3 retries left.
59
- Attempt 2 failed. 2 retries left.
60
- Attempt 3 failed. 1 retries left.
 
 
 
 
 
 
 
 
 
1
+ ===== Application Startup at 2026-04-04 11:24:31 =====
2
+
3
+ PRIX starting on 0.0.0.0:7860...
4
+
5
+ > prix-ai-auditor@1.1.0 start
6
+ > probot run ./lib/index.js
7
+
8
+ INFO (probot): PRIX module loading...
9
+ INFO (probot): PRIX Service Started
10
+ INFO (server): Running Probot v12.4.0 (Node.js: v20.20.2)
11
+ INFO (server): Listening on http://0.0.0.0:7860
12
+ PRIX server is healthy and ready!
13
+ INFO (http): GET /ping 200 - 3ms
14
+ INFO (http): GET /?logs=container 302 - 2ms
15
+ INFO (http): GET /probot 304 - 16ms
16
+ INFO (http): GET /primer.css 304 - 2ms
17
+ INFO (http): GET /robot.svg 304 - 1ms
18
+ INFO (http): POST / 200 - 3ms
19
+ INFO (http): POST / 200 - 1ms
20
+ INFO (http): POST / 200 - 1ms
21
+ INFO (probot): Received pull_request.opened for PR #37
22
+ Cloning Rachit-Tiwari-7/TEST-PRIX into /tmp/prix_tmp_3488828279_1775302043725...
23
  Cloning Rachit-Tiwari-7/TEST-PRIX with depth 1...
24
  From https://github.com/Rachit-Tiwari-7/TEST-PRIX
25
+ * branch 5ed12124e3697343de80c3f43bd5f9f77f64d260 -> FETCH_HEAD
26
+ Note: switching to '5ed12124e3697343de80c3f43bd5f9f77f64d260'.
27
 
28
  You are in 'detached HEAD' state. You can look around, make experimental
29
  changes and commit them, and you can discard any commits you make in this
 
40
 
41
  Turn off this advice by setting config variable advice.detachedHead to false
42
 
43
+ HEAD is now at 5ed1212 ORIGIN
44
  Initialized AI client with model: llama-3.1-8b-instant
45
  Initialized AI client with model: llama-3.3-70b-versatile
46
+ INFO (event): PRIX Audit starting for PR #37
47
+ id: "3f84cee0-3019-11f1-9fbe-e0cb78aae281"
48
  [Graph] Discovered: 1 TS/JS, 0 Go, 2 Python, 0 Rust, 0 C++/Java files
49
  [GraphCache] TS: 1/1 files need parsing
50
  [GraphCache] Go: All 0 files served from cache
51
  [GraphCache] Python: 2/2 files need parsing
52
  INFO (event): Will review from the base commit: 136795cc20e35ffa9a345771e8eeaaf4bce69ae2
53
+ id: "3f84cee0-3019-11f1-9fbe-e0cb78aae281"
54
  checking path: E.js => true
55
  checking path: e.py => true
56
  INFO (event): summarize: E.js
57
+ id: "3f84cee0-3019-11f1-9fbe-e0cb78aae281"
58
  INFO (event): summarize: e.py
59
+ id: "3f84cee0-3019-11f1-9fbe-e0cb78aae281"
60
+ AI response time: 300 ms
61
  INFO (event): filename: E.js, triage: NEEDS_REVIEW
62
+ id: "3f84cee0-3019-11f1-9fbe-e0cb78aae281"
63
+ AI response time: 295 ms
64
  INFO (event): filename: e.py, triage: NEEDS_REVIEW
65
+ id: "3f84cee0-3019-11f1-9fbe-e0cb78aae281"
66
+ AI response time: 722 ms
67
+ prixExec failed: "git grep -lE "import.*add_to_inventory|require.*add_to_inventory|add_to_inventory\." -- . ":(exclude)e.py"" in /tmp/prix_tmp_3488828279_1775302043725. Error: Command failed: git grep -lE "import.*add_to_inventory|require.*add_to_inventory|add_to_inventory\." -- . ":(exclude)e.py"
68
+ AI response time: 3248 ms
69
+ INFO (http): POST / 202 - 9002ms
70
+ AI response time: 3331 ms
71
+ AI response time: 3601 ms
72
+ AI response time: 3607 ms
73
+ AI response time: 3809 ms
74
+ AI response time: 3269 ms
75
+ Attempt 1 failed. 3 retries left. Error: OpenAI error 429: {"error":{"message":"Rate limit reached for model `llama-3.3-70b-versatile` in organization `org_01kgqhwjhvfarvnkqfknpyj35m` service tier `on_demand` on tokens per minute (TPM): Limit 12000, Used 9779, Requested 2326. Please try again in 525ms. Need more tokens? Upgrade to Dev Tier today at https://console.groq.com/settings/billing","type":"tokens","code":"rate_limit_exceeded"}}
76
+
77
+ AI response time: 5118 ms
78
+ Attempt 1 failed. 3 retries left. Error: OpenAI error 429: {"error":{"message":"Rate limit reached for model `llama-3.3-70b-versatile` in organization `org_01kgqhwjhvfarvnkqfknpyj35m` service tier `on_demand` on tokens per minute (TPM): Limit 12000, Used 11081, Requested 2183. Please try again in 6.32s. Need more tokens? Upgrade to Dev Tier today at https://console.groq.com/settings/billing","type":"tokens","code":"rate_limit_exceeded"}}
79
+
80
+ Attempt 2 failed. 2 retries left. Error: OpenAI error 429: {"error":{"message":"Rate limit reached for model `llama-3.3-70b-versatile` in organization `org_01kgqhwjhvfarvnkqfknpyj35m` service tier `on_demand` on tokens per minute (TPM): Limit 12000, Used 10673, Requested 2183. Please try again in 4.28s. Need more tokens? Upgrade to Dev Tier today at https://console.groq.com/settings/billing","type":"tokens","code":"rate_limit_exceeded"}}
81
+
82
+ Attempt 3 failed. 1 retries left. Error: OpenAI error 429: {"error":{"message":"Rate limit reached for model `llama-3.3-70b-versatile` in organization `org_01kgqhwjhvfarvnkqfknpyj35m` service tier `on_demand` on tokens per minute (TPM): Limit 12000, Used 9864, Requested 2183. Please try again in 235ms. Need more tokens? Upgrade to Dev Tier today at https://console.groq.com/settings/billing","type":"tokens","code":"rate_limit_exceeded"}}
83
+
84
+ AI response time: 18646 ms
85
+ AI response time: 5964 ms
86
+ Attempt 1 failed. 3 retries left. Error: OpenAI error 429: {"error":{"message":"Rate limit reached for model `llama-3.3-70b-versatile` in organization `org_01kgqhwjhvfarvnkqfknpyj35m` service tier `on_demand` on tokens per minute (TPM): Limit 12000, Used 11888, Requested 1528. Please try again in 7.08s. Need more tokens? Upgrade to Dev Tier today at https://console.groq.com/settings/billing","type":"tokens","code":"rate_limit_exceeded"}}
87
+
88
+ Attempt 2 failed. 2 retries left. Error: OpenAI error 429: {"error":{"message":"Rate limit reached for model `llama-3.3-70b-versatile` in organization `org_01kgqhwjhvfarvnkqfknpyj35m` service tier `on_demand` on tokens per minute (TPM): Limit 12000, Used 11480, Requested 2181. Please try again in 8.305s. Need more tokens? Upgrade to Dev Tier today at https://console.groq.com/settings/billing","type":"tokens","code":"rate_limit_exceeded"}}
src/bot.ts CHANGED
@@ -136,7 +136,7 @@ IMPORTANT: Entire response must be in ISO code: ${this.options.language}
136
  },
137
  {
138
  retries: this.options.retries,
139
- minTimeout: 2000, // Wait at least 2 seconds between retries to avoid TPM spike
140
  onFailedAttempt: (err: any) => {
141
  logInfo(
142
  `Attempt ${err.attemptNumber} failed. ${err.retriesLeft} retries left. Error: ${err.message}`
 
136
  },
137
  {
138
  retries: this.options.retries,
139
+ minTimeout: 10000, // 10s delay to allow TPM refresh
140
  onFailedAttempt: (err: any) => {
141
  logInfo(
142
  `Attempt ${err.attemptNumber} failed. ${err.retriesLeft} retries left. Error: ${err.message}`