Skip to content

Instantly share code, notes, and snippets.

View nerdalert's full-sized avatar
🐈
🦀 🐿

Brent Salisbury nerdalert

🐈
🦀 🐿
View GitHub Profile
{"date": "20250411-002323", "backend": "vllm", "model_id": "meta-llama/Llama-3.2-1B", "tokenizer_id": "meta-llama/Llama-3.2-1B", "num_prompts": 120, "framework": "vllm", "request_rate": 1.0, "burstiness": 1.0, "max_concurrency": null, "duration": 99.19297069497406, "completed": 120, "total_input_tokens": 120000, "total_output_tokens": 12000, "request_throughput": 1.2097631430861078, "request_goodput:": null, "output_throughput": 120.97631430861077, "total_token_throughput": 1330.7394573947186, "mean_ttft_ms": 56.25359537589247, "median_ttft_ms": 55.28098650393076, "std_ttft_ms": 6.545660891106274, "p99_ttft_ms": 78.70767521642848, "mean_tpot_ms": 7.615017463035274, "median_tpot_ms": 7.524641732229014, "std_tpot_ms": 0.5137661324558762, "p99_tpot_ms": 8.92187871250578, "mean_itl_ms": 7.615019605385871, "median_itl_ms": 7.299988501472399, "std_itl_ms": 3.706068885790247, "p99_itl_ms": 8.394360903184861}
{"date": "20250411-002536", "backend": "vllm", "model_id": "meta-llama/Llama-3.2-1B", "tokenizer_id": "meta-l

| date | backend | model_id | tokenizer_id | num_prompts | framework | request_rate | burstiness | max_concurrency | duration | completed | total_input_tokens | total_output_tokens | request_throughput | request_goodput: | output_throughput | total_token_throughput | mean_ttft_ms | median_ttft_ms | std_ttft_ms | p99_ttft_ms | mean_tpot_ms | median_tpot_ms | std_tpot_ms | p99_tpot_ms | mean_itl_ms | median_itl_ms | std_itl_ms | p99_itl_ms | |:--------------------|:----------|:---------------------------------|:---------------------------------|--------------:|:------------|---------------:|-------------:|------------------:|-----------:|------------:|---------------------:|----------------------:|---------------------:|-------------------:|--------------------:|-------------------------:|---------------:|-----------------:|--------------:|--------------:|---------------:|-----------------:|-----------


Ilab UI API Server


1. GET /models

Purpose:

#!/usr/bin/env python3
"""
Example script that:
1. Converts a document to a DocLing Document.
2. Chunks it using a HybridChunker.
3. Embeds each chunk using a SentenceTransformer.
4. Stores them in a LanceDB index.
5. Searches for a user-provided query and returns the best matching chunk or all matching chunks based on a flag.
"""
curl -X POST "${DS_HOST}/api/orchestrator/api/v1/query/run"      -H "X-Authorization: Bearer ${DS_TOKEN}"      -H "Content-Type: application/json"      -d '{
           "query": {
             "variables": {},
             "template": {
               "version": "1",
               "tasks": [
                 {
                   "id": "QA",
                   "kind": "SemanticRag",
{
"created_by": "amaa1981, fatehks, webdba, youeyoue",
"document": {
"commit": "706f365a256950f408707e2f03fe76e380c734c6",
"patterns": [
"prince_leopold_island.md"
],
"repo": "https://github.com/juliadenham/Summit_knowledge"
},
"domain": "geography",

Here are the steps to open a PR (Pull Request) to a GitHub repository, including how to commit messages with vi and DCO (Developer Certificate of Origin) signoff.

Step-by-Step Guide

  1. Fork the Repository:
    • Go to the GitHub repository you want to contribute to.
    • Click the "Fork" button at the top right of the repository page.
    • This creates a copy of the repository under your GitHub account.

curl -X 'POST' \
--cacert server-ca-crt.pem2 \
--cert client-tls-crt.pem2 \
--key client-tls-key.pem2 \
'https://datagen-service-predictor-fmaas-rhods-testing.apps.fmaas-devstage-backend.fmaas.res.ibm.com/skill/datagen' \
-H 'accept: application/json' \
-H 'Content-Type: application/json' \
-d '{

# Final Working with fwmark set on the interface Configuration

REMOTE_KEY=6/CwH/gzz9jdKnxeVpWFivycFQMBkniLeBNFQq0+f04=
sudo wg set wg0 peer $REMOTE_KEY allowed-ips 0.0.0.0/0 persistent-keepalive 25  endpoint 54.227.102.183:41823

# This command enables the src_valid_mark functionality for all network interfaces. This is required for routing marked packets properly with WireGuard.
sudo sysctl -q net.ipv4.conf.all.src_valid_mark=1
# This command adds a rule to the routing policy database (RPDB) that says, "If a packet does not have the firewall mark 51820, look up the routing table 51820."
command: nexd --stun --username qaa9a7badf-df1b-11ed-9534-0022480a240b --password x6!GHS6g3TUyHH relay https://qa.nexodus.io
2023-04-20T02:12:13.738Z	INFO	nexd/main.go:128	Debug logging enabled
2023-04-20T02:12:13.738Z	INFO	nexd/main.go:68	Starting relay agent with wireguard driver
2023-04-20T02:12:13.742Z	DEBUG	nexodus/stun_linux.go:73	send to 64.233.186.127:19302: (0 bytes)
2023-04-20T02:12:13.911Z	DEBUG	nexodus/stun_linux.go:170	response from 64.233.186.127: (40 bytes)
2023-04-20T02:12:13.911Z	DEBUG	nexodus/stun_linux.go:66	reflexive binding is: 52.90.113.12:51820
2023-04-20T02:12:13.916Z	DEBUG	nexodus/stun_linux.go:73	send to 142.250.15.127:19302: (0 bytes)
2023-04-20T02:12:14.031Z	DEBUG	nexodus/stun_linux.go:170	response from 142.250.15.127: (40 bytes)
2023-04-20T02:12:14.031Z	DEBUG	nexodus/stun_linux.go:66	reflexive binding is: 52.90.113.12:51820