content large_stringlengths 3 20.5k | url large_stringlengths 53 192 β | branch large_stringclasses 4
values | source large_stringclasses 51
values | embeddings listlengths 384 384 | score float64 -0.21 0.65 |
|---|---|---|---|---|---|
"company", "description": "Company name"}]' \ --intent "Find the CEO and annual revenue" ``` ### YAML config run ```bash parallel-cli enrich run config.yaml ``` ### Status / polling ```bash parallel-cli enrich status --json parallel-cli enrich poll --json ``` Use explicit JSON arrays for column definitions when operati... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/research/research-parallel-cli.md | main | hermes-agent | [
-0.039479415863752365,
0.021222664043307304,
-0.09411092102527618,
-0.008257836103439331,
-0.0008235381683334708,
-0.024047479033470154,
-0.009825683198869228,
0.01626959629356861,
-0.001770135248079896,
0.003389173187315464,
0.07733611762523651,
-0.04331713914871216,
0.012422961182892323,
... | 0.120199 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Drug Discovery Pharmaceutical research assistant for drug discovery workflows. Search bioactive compounds on ChEMBL, calculate drug-likeness (Lipinski Ro5, QED, TPSA, syn... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/research/research-drug-discovery.md | main | hermes-agent | [
-0.06401466578245163,
-0.01280280016362667,
-0.11113205552101135,
-0.02371770516037941,
0.018519965931773186,
-0.03233729675412178,
-0.0007796871359460056,
0.11621767282485962,
-0.04599900543689728,
-0.009891810826957226,
-0.03403772786259651,
0.013828755356371403,
0.023622920736670494,
0.... | 0.077045 |
print(f' Both rules met: {\"Yes β good oral absorption predicted\" if tpsa<=140 and rot<=10 else \"No β reduced oral absorption\"}') " ``` ### 3 β Drug Interaction & Safety Lookup (OpenFDA) ```bash DRUG="$1" ENCODED=$(python3 -c "import urllib.parse,sys; print(urllib.parse.quote(sys.argv[1]))" "$DRUG") curl -s "https:/... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/research/research-drug-discovery.md | main | hermes-agent | [
0.004543369635939598,
0.06572078168392181,
-0.09439152479171753,
-0.039456889033317566,
0.004044983070343733,
-0.043779730796813965,
0.02256312593817711,
0.09027687460184097,
0.03170566260814667,
-0.08898385614156723,
0.031247779726982117,
-0.024939926341176033,
0.009574124589562416,
-0.02... | 0.046931 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Qmd Search personal knowledge bases, notes, docs, and meeting transcripts locally using qmd β a hybrid retrieval engine with BM25, vector search, and LLM reranking. Suppo... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/research/research-qmd.md | main | hermes-agent | [
-0.08315979689359665,
0.01857619546353817,
-0.03102419525384903,
-0.012114482000470161,
-0.017944978550076485,
0.02453966811299324,
0.00666763773187995,
-0.02121827006340027,
-0.03846057876944542,
0.01923009753227234,
0.013278125785291195,
0.018009083345532417,
0.09180386364459991,
0.00946... | 0.120358 |
varies | | `qmd status` | Show index health and collection info | instant | | `qmd mcp` | Start MCP server (stdio) | persistent | | `qmd mcp --http --daemon` | Start MCP server (HTTP, warm models) | persistent | ## Setup Workflow ### 1. Add Collections Point qmd at directories containing your documents: ```bash # Add a... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/research/research-qmd.md | main | hermes-agent | [
-0.06178761273622513,
-0.012967298738658428,
-0.05539151653647423,
0.015154421329498291,
-0.06503874063491821,
0.04245610162615776,
-0.01782006397843361,
0.07886851578950882,
0.051028624176979065,
0.01966572366654873,
-0.010214829817414284,
-0.044248562306165695,
0.0632101520895958,
0.0123... | 0.149411 |
load this skill. ### Option A: Stdio Mode (Simple) Add to `~/.hermes/config.yaml`: ```yaml mcp\_servers: qmd: command: "qmd" args: ["mcp"] timeout: 30 connect\_timeout: 45 ``` This registers tools: `mcp\_qmd\_search`, `mcp\_qmd\_vsearch`, `mcp\_qmd\_deep\_search`, `mcp\_qmd\_get`, `mcp\_qmd\_status`. \*\*Tradeoff:\*\* ... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/research/research-qmd.md | main | hermes-agent | [
-0.04764564707875252,
-0.034483879804611206,
-0.036355409771203995,
0.01754569821059704,
-0.035209592431783676,
-0.0025613855104893446,
-0.09623584151268005,
0.03681914508342743,
0.03368767350912094,
0.0694650337100029,
-0.0011470293393358588,
0.0037380799185484648,
0.03288009762763977,
-0... | 0.11191 |
fast keyword lookup (code identifiers, exact names), BM25 is instant and needs no models. 4. \*\*Use `qmd query` for quality\*\* β when the question is conceptual or the user needs the best possible results, use hybrid search. 5. \*\*Prefer MCP integration\*\* β once configured, the agent gets native tools without need... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/research/research-qmd.md | main | hermes-agent | [
-0.06358522176742554,
-0.07041941583156586,
-0.05642480030655861,
0.02974127046763897,
-0.04914479702711105,
-0.01222960650920868,
-0.00449236249551177,
0.03991967439651489,
-0.04174184426665306,
0.02290128730237484,
0.05167291313409805,
-0.047213196754455566,
0.04756186529994011,
-0.03957... | 0.086139 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Searxng Search Free meta-search via SearXNG β aggregates results from 70+ search engines. Self-hosted or use a public instance. No API key needed. Falls back automaticall... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/research/research-searxng-search.md | main | hermes-agent | [
-0.0652017742395401,
0.034635189920663834,
-0.03675235062837601,
0.033771004527807236,
0.006450710818171501,
-0.039385877549648285,
-0.050824861973524094,
-0.03649027645587921,
-0.06612035632133484,
-0.023771746084094048,
-0.0005063564167357981,
-0.02454179897904396,
0.09330196678638458,
-... | 0.018169 |
RuntimeError("SEARXNG\_URL is not set") query = "fastapi deployment guide" params = { "q": query, "format": "json", "limit": 5, "engines": "google,bing", } resp = requests.get(f"{base\_url}/search", params=params, timeout=10) resp.raise\_for\_status() data = resp.json() for r in data.get("results", []): print(r["title"... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/research/research-searxng-search.md | main | hermes-agent | [
-0.05232244357466698,
0.057968392968177795,
0.02236199751496315,
0.009541830979287624,
-0.08261531591415405,
-0.1005486473441124,
-0.032143816351890564,
-0.006766405422240496,
-0.07430033385753632,
-0.08851544559001923,
0.018820123746991158,
-0.07998621463775635,
0.02160048671066761,
-0.06... | -0.039264 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Scrapling Web scraping with Scrapling - HTTP fetching, stealth browser automation, Cloudflare bypass, and spider crawling via CLI and Python. ## Skill metadata | | | |---... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/research/research-scrapling.md | main | hermes-agent | [
-0.12281426042318344,
0.019233647733926773,
-0.04982034116983414,
0.01623869687318802,
0.05807778239250183,
-0.07929659634828568,
-0.027653904631733894,
-0.07142309844493866,
-0.05883437767624855,
-0.007910759188234806,
0.036875832825899124,
0.02173105627298355,
0.032943323254585266,
-0.03... | 0.047809 |
import DynamicFetcher page = DynamicFetcher.fetch('https://example.com', headless=True) data = page.css('.js-loaded-content::text').getall() ``` ### Wait for Specific Element ```python page = DynamicFetcher.fetch( 'https://example.com', wait\_selector=('.results', 'visible'), network\_idle=True, ) ``` ### Disable Resou... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/research/research-scrapling.md | main | hermes-agent | [
-0.09083104878664017,
0.05367370694875717,
0.02793280780315399,
0.07914085686206818,
-0.016980433836579323,
-0.05530533939599991,
0.020965758711099625,
-0.03418687358498573,
0.04096801206469536,
-0.02334449626505375,
-0.006297975312918425,
0.05409001186490059,
-0.005822123493999243,
-0.019... | 0.02938 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Duckduckgo Search Free web search via DuckDuckGo β text, news, images, videos. No API key needed. Prefer the `ddgs` CLI when installed; use the Python DDGS library only a... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/research/research-duckduckgo-search.md | main | hermes-agent | [
-0.08100470155477524,
-0.0038524428382515907,
-0.06553641706705093,
0.002506205113604665,
0.03429878130555153,
-0.03749574348330498,
0.032967887818813324,
-0.03822449967265129,
-0.09021908044815063,
-0.01808452233672142,
-0.0037761833518743515,
0.012988068163394928,
0.05149054154753685,
-0... | 0.032691 |
After Verification) Use the `DDGS` class in `execute\_code` or another Python runtime only after verifying that `ddgs` is installed there. Do not assume `execute\_code` includes third-party packages by default. Safe wording: - "Use `execute\_code` with `ddgs` after installing or verifying the package if needed" Avoid s... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/research/research-duckduckgo-search.md | main | hermes-agent | [
-0.06442947685718536,
0.019984031096100807,
-0.014540844596922398,
0.03190848231315613,
-0.018653422594070435,
-0.10934572666883469,
0.03344987332820892,
0.031234411522746086,
-0.1327471137046814,
-0.07259663939476013,
0.04234902188181877,
0.05698679760098457,
0.07529222220182419,
-0.06721... | -0.015085 |
runtime does not have the package installed | Do not use Python DDGS there until that runtime is prepared | | Search returns nothing | Temporary rate limiting or poor query | Wait a few seconds, retry, or adjust the query | | CLI works but `execute\_code` import fails | Terminal and `execute\_code` are different runtim... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/research/research-duckduckgo-search.md | main | hermes-agent | [
-0.04751751571893692,
-0.041846003383398056,
-0.01931839995086193,
0.002501377370208502,
-0.03575436398386955,
-0.12535516917705536,
0.016390589997172356,
0.05605367198586464,
-0.09797515720129013,
-0.06642264127731323,
0.021588359028100967,
-0.016290517523884773,
0.09156777709722519,
-0.1... | -0.018314 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Bioinformatics Gateway to 400+ bioinformatics skills from bioSkills and ClawBio. Covers genomics, transcriptomics, single-cell, variant calling, pharmacogenomics, metagen... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/research/research-bioinformatics.md | main | hermes-agent | [
-0.0725468248128891,
0.005526124499738216,
-0.056620653718709946,
-0.043685395270586014,
-0.025643033906817436,
-0.059485357254743576,
-0.016832103952765465,
0.015033923089504242,
-0.050813861191272736,
0.018931057304143906,
-0.007444565184414387,
-0.013698633760213852,
-0.014863346703350544... | 0.110723 |
spatial-proteomics, image-analysis ### Epigenomics bioSkills: chip-seq/ β peak-calling, differential-binding, motif-analysis, peak-annotation, chipseq-qc, chipseq-visualization, super-enhancers atac-seq/ β atac-peak-calling, atac-qc, differential-accessibility, footprinting, motif-deviation, nucleosome-positioning meth... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/research/research-bioinformatics.md | main | hermes-agent | [
-0.024993078783154488,
-0.03944673016667366,
-0.09874816238880157,
-0.09678235650062561,
-0.010106523521244526,
-0.011826282367110252,
0.01665228046476841,
0.11784624308347702,
-0.07165730744600296,
-0.005205043591558933,
-0.002870072377845645,
0.011103498749434948,
0.0024362285621464252,
... | 0.165838 |
Learning for Omics bioSkills: machine-learning/ β omics-classifiers, biomarker-discovery, survival-analysis, model-validation, prediction-explanation, atlas-mapping ClawBio: claw-semantic-sim β Semantic similarity index for disease literature (PubMedBERT) omics-target-evidence-mapper β Aggregate target-level evidence a... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/research/research-bioinformatics.md | main | hermes-agent | [
-0.0029260674491524696,
0.006924293003976345,
-0.05350692570209503,
-0.09318657964468002,
0.07273358106613159,
-0.06382142007350922,
-0.04002447798848152,
0.07682671397924423,
-0.04414023831486702,
0.010453303344547749,
-0.026559878140687943,
-0.04545654356479645,
-0.027910497039556503,
0.... | 0.072842 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Pinggy Tunnel Zero-install localhost tunnels over SSH via Pinggy. ## Skill metadata | | | |---|---| | Source | Optional β install with `hermes skills install official/dev... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/devops/devops-pinggy-tunnel.md | main | hermes-agent | [
-0.06382346153259277,
0.0696335956454277,
-0.04492846131324768,
0.0035301665775477886,
-0.02902386151254177,
-0.07505947351455688,
0.0250598955899477,
-0.05191987380385399,
-0.03984649479389191,
0.06396432220935822,
0.032552946358919144,
-0.05743797868490219,
0.043807003647089005,
-0.03797... | 0.129787 |
start it first (e.g. `python3 -m http.server 8000 --bind 127.0.0.1`). Pinggy will happily return a URL pointed at nothing β the user will see 502 until the origin comes up. ### 2. Launch the tunnel as a background process Use `terminal(background=True)` and capture output to a logfile (Pinggy prints the URLs on stdout,... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/devops/devops-pinggy-tunnel.md | main | hermes-agent | [
-0.0122810248285532,
0.06894678622484207,
-0.08056434988975525,
-0.027891721576452255,
-0.03821086138486862,
-0.022428054362535477,
-0.04541262239217758,
-0.007047676015645266,
0.010922745801508427,
0.0613287016749382,
-0.087698794901371,
-0.04283538833260536,
0.014436213299632072,
-0.0823... | -0.031762 |
- \*\*`process(action='log')` may miss SSH banner output.\*\* Pinggy prints the URLs and then the SSH session goes interactive. Always redirect to a logfile and `grep` the file directly β same pattern as `cloudflared-quick-tunnel`. - \*\*Host-key prompt on first run.\*\* Default OpenSSH config asks the user to accept P... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/devops/devops-pinggy-tunnel.md | main | hermes-agent | [
0.005560688674449921,
0.08437308669090271,
-0.020470863208174706,
-0.04950026050209999,
0.046522870659828186,
-0.05001933500170708,
0.06238687038421631,
-0.02897186018526554,
0.054783616214990616,
0.10954490303993225,
-0.003982346970587969,
-0.0833539143204689,
-0.003516129218041897,
-0.03... | 0.036519 |
# Pre-req: the model server is already running on 127.0.0.1:11434 (Ollama default) TOKEN=$(openssl rand -hex 16) nohup ssh -p 443 -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null \ -o ServerAliveInterval=30 \ -R0:localhost:11434 "k:$TOKEN+co+free@a.pinggy.io" \ >/tmp/llm-pinggy.log 2>&1 & echo $! >/tmp/llm-p... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/devops/devops-pinggy-tunnel.md | main | hermes-agent | [
0.03264009952545166,
0.0767890065908432,
-0.048856399953365326,
0.01864405907690525,
0.058243706822395325,
-0.07361762225627899,
-0.0510547049343586,
-0.04200490936636925,
0.08543705195188522,
0.045967839658260345,
-0.007191632874310017,
-0.05586063489317894,
0.0010831390973180532,
-0.0778... | 0.028789 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Watchers Poll RSS, JSON APIs, and GitHub with watermark dedup. ## Skill metadata | | | |---|---| | Source | Optional β install with `hermes skills install official/devops... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/devops/devops-watchers.md | main | hermes-agent | [
-0.12231018394231796,
-0.015277941711246967,
-0.05333079397678375,
0.03507024049758911,
0.03563494235277176,
-0.06102238968014717,
0.03366895765066147,
-0.06657328456640244,
0.0014587246114388108,
0.038912978023290634,
0.003536426927894354,
0.02575642056763172,
0.06767866760492325,
-0.0264... | 0.175148 |
job's agent loop; no changes to cron's built-in `--script` flag are needed. ## State files Every watcher writes `$HERMES\_HOME/watcher-state/<name>.json`. Inspect: ```bash cat $HERMES\_HOME/watcher-state/hn.json ``` Force a replay (next run treated as first poll): ```bash rm $HERMES\_HOME/watcher-state/hn.json ``` ## W... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/devops/devops-watchers.md | main | hermes-agent | [
-0.09278794378042221,
0.005567601416260004,
-0.026630103588104248,
0.021383851766586304,
0.06893859803676605,
-0.04685346409678459,
0.008255253545939922,
-0.0063027325086295605,
-0.012048700824379921,
-0.042300887405872345,
-0.010954177007079124,
-0.01633831486105919,
0.01985352672636509,
... | 0.073201 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Inference Sh Cli Run 150+ AI apps via inference.sh CLI (infsh) β image generation, video creation, LLMs, search, 3D, social automation. Uses the terminal tool. Triggers: ... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/devops/devops-cli.md | main | hermes-agent | [
-0.08348996192216873,
-0.027496036142110825,
-0.02785644307732582,
-0.05554475262761116,
0.08625364303588867,
-0.06252258270978928,
-0.0005054016364738345,
0.012682859785854816,
-0.03444811329245567,
0.06232365593314171,
0.02193479798734188,
-0.09171213209629059,
0.07813211530447006,
-0.00... | 0.190331 |
bytedance/omnihuman-1-5 --input '{"audio": "/path/to/audio.mp3", "image": "/path/to/face.jpg"}' --json ``` ### Search & Research ```bash infsh app list --search search infsh app run tavily/tavily-search --input '{"query": "latest AI news"}' --json infsh app run exa/exa-search --input '{"query": "machine learning papers... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/devops/devops-cli.md | main | hermes-agent | [
-0.017794916406273842,
0.016731925308704376,
-0.06940729916095734,
-0.03138163685798645,
0.07173796743154526,
-0.06116457283496857,
-0.015950798988342285,
0.0296099204570055,
-0.03989453241229057,
-0.017066296190023422,
0.04004734009504318,
-0.047855526208877563,
0.03045201115310192,
0.003... | 0.096558 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Docker Management Manage Docker containers, images, volumes, networks, and Compose stacks β lifecycle ops, debugging, cleanup, and Dockerfile optimization. ## Skill metad... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/devops/devops-docker-management.md | main | hermes-agent | [
-0.03389536589384079,
0.06450299173593521,
0.004329690244048834,
0.0005107639590278268,
-0.0005297362222336233,
-0.11997444182634354,
0.02454083040356636,
-0.04727955162525177,
-0.04304194450378418,
0.04453231766819954,
-0.026150228455662727,
-0.005743024870753288,
0.06409381330013275,
-0.... | 0.11151 |
name, `--restart` restart policy. \*\*Manage running containers:\*\* ```bash docker ps # running containers docker ps -a # all (including stopped) docker stop NAME # graceful stop docker start NAME # start stopped container docker restart NAME # stop + start docker rm NAME # remove stopped container docker rm -f NAME #... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/devops/devops-docker-management.md | main | hermes-agent | [
0.023061318323016167,
0.05313738435506821,
-0.005540056619793177,
-0.022683100774884224,
0.05219471827149391,
-0.021066348999738693,
0.03202776238322258,
0.02313072979450226,
-0.00024707327247597277,
0.06899195909500122,
-0.01913359761238098,
-0.030013713985681534,
0.010211016051471233,
0.... | 0.072382 |
a diagnostic before cleaning: ```bash # Check what's using space docker system df # summary docker system df -v # detailed breakdown # Targeted cleanup (safe) docker container prune # stopped containers docker image prune # dangling images docker volume prune # unused volumes docker network prune # unused networks # Ag... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/devops/devops-docker-management.md | main | hermes-agent | [
0.02411523275077343,
0.06037752702832222,
0.012807533144950867,
-0.03481617197394371,
0.08337552845478058,
-0.09636618942022324,
0.010034851729869843,
-0.008128124289214611,
0.040182095021009445,
0.06959563493728638,
0.023516904562711716,
0.0621880479156971,
0.03986193612217903,
0.03789603... | 0.005438 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Web Pentest Authorized web application penetration testing β reconnaissance, vulnerability analysis, proof-based exploitation, and professional reporting. Adapts Shannon'... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/security/security-web-pentest.md | main | hermes-agent | [
-0.06313955783843994,
0.017004862427711487,
-0.06942395865917206,
-0.016970224678516388,
-0.01679379679262638,
-0.12492284178733826,
0.04144114628434181,
-0.025761589407920837,
-0.06561573594808578,
0.020006736740469933,
0.050059154629707336,
0.01101932767778635,
0.06631052494049072,
0.029... | 0.15847 |
host. The recon-scan.sh script enforces this. Don't bypass it without operator approval. 8. \*\*Authority of the report.\*\* This skill produces a security assessment, not a "PASS." Even a clean run is "no exploitable issues FOUND in scope X within time T using methods Y" β not "the application is secure." Mirror that ... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/security/security-web-pentest.md | main | hermes-agent | [
-0.00017195797408930957,
0.08018126338720322,
-0.07502024620771408,
0.03175073117017746,
0.008501492440700531,
-0.048003144562244415,
0.06647016108036041,
0.03917945921421051,
0.0017631370574235916,
0.023196764290332794,
-0.0066053662449121475,
0.00980681087821722,
0.10122226923704147,
0.0... | 0.120216 |
Reference | |-------|------|-----------| | `injection` | SQLi, command, path traversal, SSTI, LFI/RFI, deserialization | `references/vuln-taxonomy.md` (slot types) | | `xss` | Reflected, stored, DOM-based | `references/vuln-taxonomy.md` (render contexts) | | `auth` | Login bypass, JWT confusion, session fixation, OAuth... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/security/security-web-pentest.md | main | hermes-agent | [
-0.06363121420145035,
0.029996497556567192,
-0.14454039931297302,
-0.032255884259939194,
0.013745344243943691,
-0.047274403274059296,
0.03182600066065788,
0.0027620503678917885,
-0.012453804723918438,
0.0026513044722378254,
-0.017599347978830338,
-0.01209869235754013,
0.0653928592801094,
-... | 0.17604 |
operator. Do not pivot without explicit approval β that pivot is what makes pentesting illegal. --- ## What This Skill Does NOT Cover - Network-layer pentesting beyond port scanning (no Metasploit, Cobalt Strike, AD attacks, network protocol fuzzing). - Reverse engineering / binary analysis (see issue #383). - Source-o... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/security/security-web-pentest.md | main | hermes-agent | [
-0.050669293850660324,
0.07543156296014786,
-0.027563638985157013,
-0.010818632319569588,
-0.04502568766474724,
-0.07261878997087479,
0.006660032086074352,
-0.014268502593040466,
-0.06445173919200897,
0.02736915647983551,
-0.006363942753523588,
0.0008620726875960827,
0.013108938001096249,
... | 0.055143 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Sherlock OSINT username search across 400+ social networks. Hunt down social media accounts by username. ## Skill metadata | | | |---|---| | Source | Optional β install w... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/security/security-sherlock.md | main | hermes-agent | [
-0.05836064741015434,
0.009938395582139492,
-0.052678462117910385,
-0.0029989962931722403,
-0.007790112402290106,
-0.06718415021896362,
0.06348375976085663,
-0.03460659459233284,
-0.059510719031095505,
-0.032732684165239334,
0.03483114391565323,
0.03658219426870346,
0.11259551346302032,
-0... | 0.120875 |
https://twitter.com/username [+] GitHub: https://github.com/username ``` Present findings as clickable links when possible. ## Pitfalls ### No Results Found If Sherlock finds no accounts, this is often correct β the username may not be registered on checked platforms. Suggest: - Checking spelling/variation - Trying sim... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/security/security-sherlock.md | main | hermes-agent | [
-0.058340609073638916,
-0.0641074851155281,
-0.029526440426707268,
-0.023754717782139778,
-0.045211825519800186,
-0.039834704250097275,
0.03217986971139908,
-0.006838631816208363,
-0.038763709366321564,
-0.05314401909708977,
0.08529365062713623,
0.022732695564627647,
0.05804600194096565,
-... | 0.044461 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Oss Forensics Supply chain investigation, evidence recovery, and forensic analysis for GitHub repositories. Covers deleted commit recovery, force-push detection, IOC extr... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/security/security-oss-forensics.md | main | hermes-agent | [
-0.08708814531564713,
0.03662809729576111,
-0.06235433369874954,
-0.03236885368824005,
0.08433811366558075,
-0.09279639273881912,
-0.00626327795907855,
-0.043323952704668045,
0.007594725117087364,
0.05942147970199585,
0.0447639562189579,
0.02163022756576538,
0.037734076380729675,
0.0180016... | 0.184556 |
`optional-skills/` equivalent. All script and template references are relative to it. ## Phase 0: Initialization 1. Create investigation working directory: ```bash mkdir investigation\_$(echo "REPO\_NAME" | tr '/' '\_') cd investigation\_$(echo "REPO\_NAME" | tr '/' '\_') ``` 2. Initialize the evidence store: ```bash p... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/security/security-oss-forensics.md | main | hermes-agent | [
-0.024884382262825966,
0.08655931800603867,
-0.1007855013012886,
-0.010771173052489758,
0.083589568734169,
0.010314765386283398,
0.0016475861193612218,
0.06996206939220428,
0.0030424543656408787,
0.07196734100580215,
0.07870297878980637,
-0.041479721665382385,
0.005668003577739,
0.10698848... | 0.075937 |
\*\*Cross-reference targets\*\* (flag discrepancies as evidence): - PR exists in archive but missing from API β evidence of deletion - Contributor in archive events but not in contributors list β evidence of permission revocation - Commit in archive PushEvents but not in API commit list β evidence of force-push/deletio... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/security/security-oss-forensics.md | main | hermes-agent | [
-0.0697924941778183,
0.001347844721749425,
-0.010341270826756954,
-0.029154837131500244,
0.11521636694669724,
-0.004119095392525196,
-0.02038833312690258,
-0.038709793239831924,
0.05349605530500412,
0.03710852935910225,
0.07814443856477737,
-0.02273806557059288,
0.03344302996993065,
-0.015... | 0.094037 |
Evidence Consolidation After all investigators complete: 1. Run `python3 SKILL\_DIR/scripts/evidence-store.py --store evidence.json list` to see all collected evidence. 2. For each piece of evidence, verify the `content\_sha256` hash matches the original source. 3. Group evidence by: - \*\*Timeline\*\*: Sort all timest... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/security/security-oss-forensics.md | main | hermes-agent | [
-0.07209808379411697,
-0.007325086742639542,
0.011027653701603413,
-0.006237206514924765,
0.10474327951669693,
-0.004552262835204601,
-0.0294251199811697,
-0.022500136867165565,
0.044568948447704315,
0.0454694926738739,
0.07571089267730713,
0.005529006943106651,
0.040008995682001114,
0.016... | 0.02946 |
protecting open-source software from supply chain attacks. It must not be used for: - \*\*Harassment or stalking\*\* of contributors or maintainers - \*\*Doxing\*\* β correlating GitHub activity to real identities for malicious purposes - \*\*Competitive intelligence\*\* β investigating proprietary or internal reposito... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/security/security-oss-forensics.md | main | hermes-agent | [
-0.06037215515971184,
0.026271913200616837,
-0.020969701930880547,
0.005483643617480993,
0.0710601955652237,
0.000883040891494602,
-0.03073830157518387,
0.026089537888765335,
0.060074809938669205,
0.04098972678184509,
0.020444022491574287,
0.03140805661678314,
0.09355486929416656,
-0.03143... | 0.140663 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # 1Password Set up and use 1Password CLI (op). Use when installing the CLI, enabling desktop app integration, signing in, and reading/injecting secrets for commands. ## Ski... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/security/security-1password.md | main | hermes-agent | [
-0.09593664109706879,
0.024221599102020264,
-0.05741482600569725,
-0.023229680955410004,
-0.036995504051446915,
-0.06077083200216293,
0.021757420152425766,
-0.04426605999469757,
-0.061773501336574554,
0.011344632133841515,
0.0008817921043373644,
-0.018741710111498833,
0.1105763390660286,
-... | 0.107891 |
### Run a command with secret env var ```bash export DB\_PASSWORD="op://app-prod/db/password" op run -- sh -c '[ -n "$DB\_PASSWORD" ] && echo "DB\_PASSWORD is set" || echo "DB\_PASSWORD missing"' ``` ## Guardrails - Never print raw secrets back to user unless they explicitly request the value. - Prefer `op run` / `op i... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/security/security-1password.md | main | hermes-agent | [
-0.02556511014699936,
0.0488898903131485,
-0.08133610337972641,
-0.04214368015527725,
-0.05014580115675926,
-0.003951491322368383,
0.045046333223581314,
0.08036446571350098,
0.014763622544705868,
0.03584374487400055,
0.009853390045464039,
-0.08253713697195053,
0.13846264779567719,
-0.01143... | -0.031512 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Code Wiki Generate wiki docs + Mermaid diagrams for any codebase. ## Skill metadata | | | |---|---| | Source | Optional β install with `hermes skills install official/sof... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/software-development/software-development-code-wiki.md | main | hermes-agent | [
-0.050828270614147186,
0.025761518627405167,
-0.022501099854707718,
-0.02735910750925541,
0.0324600450694561,
-0.0795285627245903,
-0.013487640768289566,
-0.011634107679128647,
-0.04853013530373573,
0.028639769181609154,
0.0685950368642807,
-0.019101582467556,
0.07037519663572311,
-0.00132... | 0.102415 |
"uncommitted") REPO\_NAME=$(basename "$PWD") ``` Then set the output dir: ```bash OUTPUT\_DIR="$HOME/.hermes/wikis/$REPO\_NAME" mkdir -p "$OUTPUT\_DIR/modules" "$OUTPUT\_DIR/diagrams" ``` ### 2. Scan repo structure Use the `terminal` tool for the shell work, `read\_file` for manifests: ```bash # Shallow tree first ls -... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/software-development/software-development-code-wiki.md | main | hermes-agent | [
0.01912885159254074,
-0.015007007867097855,
-0.007553422823548317,
-0.02906874194741249,
0.06317964941263199,
-0.06751373410224915,
0.05532613769173622,
-0.009322384372353554,
-0.01814013347029686,
0.006793721113353968,
0.07850605994462967,
-0.015136174857616425,
0.05950329825282097,
0.032... | 0.006722 |
Diagram ## Core Types ```mermaid classDiagram class Agent { +string name +list~Tool~ tools +chat(message) string } class Tool { <> +name string +execute(args) any } Agent --> Tool : uses Tool <|-- TerminalTool Tool <|-- WebTool ``` ## Notes ```` For languages without classes (Go, C, Rust): use the diagram for struct re... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/software-development/software-development-code-wiki.md | main | hermes-agent | [
0.00613182270899415,
-0.007571856956928968,
0.017354531213641167,
-0.06780660897493362,
-0.04417935758829117,
-0.030601270496845245,
-0.00539181986823678,
0.05407332628965378,
0.0031504712533205748,
-0.07816049456596375,
-0.02290194481611252,
-0.039372704923152924,
0.008143304847180843,
0.... | 0.180364 |
by calling `process\_item` on each item" is worse than just linking to the function. - \*\*Mermaid > 50 nodes.\*\* They don't render legibly. Split them. - \*\*Documenting tests, generated code, or vendored deps as if they were product code.\*\* Skip them. - \*\*In-repo output without asking.\*\* Default is `~/.hermes/... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/software-development/software-development-code-wiki.md | main | hermes-agent | [
-0.08732641488313675,
0.06197801232337952,
0.033953361213207245,
0.009718752466142178,
0.05436329171061516,
-0.0790850892663002,
0.008146870881319046,
0.004221049137413502,
-0.05137913301587105,
0.022699913010001183,
0.013296548277139664,
-0.06499288231134415,
0.053772781044244766,
-0.0531... | 0.05588 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Rest Graphql Debug Debug REST/GraphQL APIs: status codes, auth, schemas, repro. ## Skill metadata | | | |---|---| | Source | Optional β install with `hermes skills instal... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/software-development/software-development-rest-graphql-debug.md | main | hermes-agent | [
-0.05178367719054222,
0.05739562213420868,
-0.04677102342247963,
0.02653350681066513,
-0.06337301433086395,
-0.06124582886695862,
-0.07765733450651169,
-0.010474245063960552,
-0.10317537933588028,
0.0910370796918869,
0.016895398497581482,
-0.024541795253753662,
0.08184559643268585,
-0.0307... | 0.043198 |
import requests from requests.exceptions import ConnectTimeout, ReadTimeout try: requests.get(url, timeout=(3.05, 30)) except ConnectTimeout: print("Cannot reach host β DNS, firewall, VPN") except ReadTimeout: print("Connected but server is slow") ''') ``` Diagnosis: high `time\_connect` is network/firewall; high `time... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/software-development/software-development-rest-graphql-debug.md | main | hermes-agent | [
-0.05587293207645416,
0.06413016468286514,
-0.06166435778141022,
0.0452038012444973,
-0.053992997854948044,
-0.14009283483028412,
-0.06110648065805435,
-0.03665197640657425,
0.049827758222818375,
-0.022436583414673805,
0.016182055696845055,
0.012088784947991371,
0.02010347507894039,
0.0067... | -0.018394 |
url, \*\*kwargs): for attempt in range(5): resp = requests.request(method, url, \*\*kwargs) if resp.status\_code != 429: return resp wait = int(resp.headers.get("Retry-After", 2 \*\* attempt)) time.sleep(wait) return resp ''') ``` ### 5xx β server-side, usually not your fault - \*\*500\*\* β server bug. Capture correla... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/software-development/software-development-rest-graphql-debug.md | main | hermes-agent | [
-0.09699966013431549,
0.04023890942335129,
-0.006269673351198435,
0.035066623240709305,
-0.02344290167093277,
-0.09041999280452728,
-0.019983546808362007,
-0.014020312577486038,
-0.03385329991579056,
0.018198823556303978,
-0.013297862373292446,
-0.005833703093230724,
0.03545733913779259,
-... | 0.050891 |
include the auth token in error details. Verify they don't. - [ ] \*\*Verbose `Server` / `X-Powered-By`.\*\* Stack-info leaks. Note for security review. ## Hermes Tool Patterns ### terminal β for curl, dig, openssl ```python terminal('curl -sI https://api.example.com') terminal('openssl s\_client -connect api.example.c... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/software-development/software-development-rest-graphql-debug.md | main | hermes-agent | [
-0.07268314063549042,
0.07602068781852722,
-0.04642949998378754,
0.021858200430870056,
-0.022772597149014473,
-0.1442096084356308,
-0.06233762204647064,
0.010644922964274883,
0.01869678869843483,
-0.036689650267362595,
-0.003825566964223981,
-0.022797664627432823,
0.10915353149175644,
-0.0... | -0.004331 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Fitness Nutrition Gym workout planner and nutrition tracker. Search 690+ exercises by muscle, equipment, or category via wger. Look up macros and calories for 380,000+ fo... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/health/health-fitness-nutrition.md | main | hermes-agent | [
-0.04749569296836853,
-0.006154035683721304,
-0.044944293797016144,
0.056599196046590805,
-0.0873771458864212,
-0.04462691769003868,
0.01396999228745699,
0.018106799572706223,
-0.1031363382935524,
-0.04569477215409279,
0.016915541142225266,
-0.06230872496962547,
0.024634098634123802,
-0.04... | 0.103535 |
Gym mat | | 5 | Swiss Ball | | 6 | Pull-up bar | | 7 | none (bodyweight) | | 8 | Bench | | 9 | Incline bench | | 10 | Kettlebell | \*\*Step 3 β Fetch and present results:\*\* ```bash # Search exercises by name QUERY="$1" ENCODED=$(python3 -c "import urllib.parse,sys; print(urllib.parse.quote(sys.argv[1]))" "$QUERY") cu... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/health/health-fitness-nutrition.md | main | hermes-agent | [
-0.0631343424320221,
0.07430335134267807,
-0.022399360314011574,
0.025752240791916847,
-0.07277148962020874,
0.02180028147995472,
0.036206379532814026,
0.02974415011703968,
0.02295343391597271,
-0.09173852205276489,
-0.009515298530459404,
-0.005180141888558865,
0.04942651465535164,
-0.0342... | 0.077117 |
muscle groups, and equipment. After nutrition lookup: confirm per-100g macros are returned with kcal, protein, fat, carbs. After calculators: sanity-check outputs (e.g. TDEE should be 1500-3500 for most adults). --- ## Quick Reference | Task | Source | Endpoint | |------|--------|----------| | Search exercises by name ... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/health/health-fitness-nutrition.md | main | hermes-agent | [
-0.03952554240822792,
0.09266366809606552,
-0.06439362466335297,
0.04033553600311279,
-0.022406097501516342,
-0.013309360481798649,
0.05630099028348923,
0.02936634048819542,
-0.008819298818707466,
-0.05634142458438873,
0.06324706226587296,
-0.10366497933864594,
0.004184794146567583,
-0.045... | 0.069499 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Neuroskill Bci Connect to a running NeuroSkill instance and incorporate the user's real-time cognitive and emotional state (focus, relaxation, mood, cognitive load, drows... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/health/health-neuroskill-bci.md | main | hermes-agent | [
-0.08658226579427719,
0.01633533462882042,
-0.042609184980392456,
-0.00815421063452959,
-0.07378040254116058,
-0.034873899072408676,
0.0052963001653552055,
-0.027657881379127502,
-0.01648421585559845,
-0.03625515475869179,
-0.006939672399312258,
-0.04518625885248184,
0.04224328324198723,
0... | 0.190014 |
profile | | `timer` | Launch focus timer (Pomodoro/Deep Work/Short Focus presets) | | `notify "title" "body"` | Send an OS notification via the NeuroSkill app | | `raw '{json}'` | Raw JSON passthrough to the server | ### Global Flags | Flag | Description | |------|-------------| | `--json` | Raw JSON output (no ANSI, p... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/health/health-neuroskill-bci.md | main | hermes-agent | [
-0.062461577355861664,
-0.013951774686574936,
-0.07466571033000946,
0.013882494531571865,
-0.05703441798686981,
-0.09010296314954758,
0.017655164003372192,
0.055697016417980194,
-0.03066338039934635,
-0.009687734767794609,
0.0076202815398573875,
-0.06862863898277283,
-0.0013616171199828386,
... | 0.250102 |
Use this to describe how a session evolved: > "Your focus started at 0.64 and climbed to 0.76 by the end β a clear upward trend. > Cognitive load dropped from 0.38 to 0.28, suggesting the task became more automatic > as you settled in." ### List All Sessions ```bash npx neuroskill sessions --json npx neuroskill session... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/health/health-neuroskill-bci.md | main | hermes-agent | [
-0.008798898197710514,
0.007251539267599583,
-0.016989266499876976,
0.043329108506441116,
-0.01164738554507494,
-0.004958997014909983,
0.04077477008104324,
0.03663727641105652,
0.015177503228187561,
-0.011495599523186684,
-0.079132080078125,
-0.011089794337749481,
-0.012575113214552402,
0.... | 0.158637 |
Labels are stored in a database and indexed for later retrieval via `search-labels` and `interactive` commands. --- ## 7. Real-Time Streaming ```bash npx neuroskill listen --seconds 30 --json npx neuroskill listen --seconds 5 --json | jq '[.[] | select(.event == "scores")]' ``` Streams live WebSocket events (EXG, PPG, ... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/health/health-neuroskill-bci.md | main | hermes-agent | [
-0.07289910316467285,
0.019282376393675804,
-0.10374363511800766,
0.02850458212196827,
0.01699359528720379,
0.0029836709145456553,
0.05471567437052727,
-0.012201200239360332,
0.09777407348155975,
-0.06590497493743896,
-0.012732845731079578,
-0.029202522709965706,
-0.02429981715977192,
0.02... | 0.247402 |
command not yet mapped to a CLI subcommand. --- ## Error Handling | Error | Likely Cause | Fix | |-------|-------------|-----| | `npx neuroskill status` hangs | NeuroSkill app not running | Open NeuroSkill desktop app | | `device.state: "disconnected"` | BCI device not connected | Check Bluetooth, device battery | | Al... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/health/health-neuroskill-bci.md | main | hermes-agent | [
-0.03355231136083603,
-0.020159587264060974,
-0.004890489857643843,
-0.004386145621538162,
0.02568785659968853,
-0.050395239144563675,
0.0449303537607193,
0.0259288027882576,
-0.052151065319776535,
-0.008066916838288307,
-0.00590658001601696,
-0.11491838842630386,
-0.00332511099986732,
0.0... | 0.247859 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Nemo Curator GPU-accelerated data curation for LLM training. Supports text/image/video/audio. Features fuzzy deduplication (16Γ faster), quality filtering (30+ heuristics... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-nemo-curator.md | main | hermes-agent | [
-0.0461311861872673,
-0.003187979571521282,
-0.029951462522149086,
-0.03563779592514038,
0.004783469717949629,
-0.11118905246257782,
-0.008663911372423172,
-0.05638398230075836,
-0.13384240865707397,
-0.005729972384870052,
0.03386132791638374,
-0.037245895713567734,
0.047112658619880676,
0... | 0.268493 |
nemo\_curator.classifiers import QualityClassifier # Quality classification quality\_clf = QualityClassifier( model\_path="nvidia/quality-classifier-deberta", batch\_size=256, device="cuda" ) # Filter low-quality documents high\_quality = dataset.filter(lambda doc: quality\_clf(doc["text"]) > 0.5) ``` ## GPU accelerati... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-nemo-curator.md | main | hermes-agent | [
-0.05277146399021149,
-0.027862943708896637,
-0.021540135145187378,
-0.000303650158457458,
-0.010260230861604214,
-0.15332767367362976,
-0.002769038313999772,
0.0011520416010171175,
-0.08672919124364853,
-0.04804297536611557,
-0.010029962286353111,
-0.05878804624080658,
0.018345162272453308,... | 0.068395 |
0.4.0+ - \*\*License\*\*: Apache 2.0 | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-nemo-curator.md | main | hermes-agent | [
0.04003936052322388,
0.06523658335208893,
-0.0022670719772577286,
-0.08020991086959839,
-0.005003036465495825,
-0.013782781548798084,
-0.012740250676870346,
-0.05831330269575119,
0.0120242303237319,
0.024719728156924248,
0.056058574467897415,
-0.04048905521631241,
-0.04218988120555878,
0.0... | 0.01187 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Qdrant Vector Search High-performance vector similarity search engine for RAG and semantic search. Use when building production RAG systems requiring fast nearest neighbo... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-qdrant.md | main | hermes-agent | [
-0.11122678965330124,
0.05261026695370674,
-0.029861871153116226,
-0.0028361910954117775,
-0.049835145473480225,
-0.02439328469336033,
-0.020446080714464188,
-0.030905934050679207,
-0.06467201560735703,
-0.01938377320766449,
-0.004578132648020983,
-0.017162542790174484,
0.08164739608764648,
... | 0.131002 |
Connections per node (default 16) ef\_construct=100, # Build-time accuracy (default 100) full\_scan\_threshold=10000 # Switch to brute force below this ), on\_disk\_payload=True # Store payload on disk ) # Collection info info = client.get\_collection("documents") print(f"Points: {info.points\_count}, Vectors: {info.ve... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-qdrant.md | main | hermes-agent | [
-0.011784191243350506,
-0.005138473119586706,
-0.05551174655556679,
0.03470681235194206,
0.012959254905581474,
-0.0429314449429512,
-0.027290241792798042,
0.035042084753513336,
-0.022367829456925392,
0.002026349538937211,
0.046365417540073395,
0.00967768020927906,
0.08613426238298416,
-0.0... | 0.024505 |
RAM ) ) ) # Search with rescoring results = client.search( collection\_name="quantized", query\_vector=query, search\_params={"quantization": {"rescore": True}}, # Rescore top results limit=10 ) ``` ## Payload indexing ```python from qdrant\_client.models import PayloadSchemaType # Create payload index for faster filte... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-qdrant.md | main | hermes-agent | [
-0.022946316748857498,
0.0685722678899765,
-0.01726837083697319,
0.05155934393405914,
-0.05580167844891548,
-0.02994362823665142,
0.0020228498615324497,
0.013788528740406036,
0.07666035741567612,
-0.0007628027233295143,
-0.0026572379283607006,
-0.06158336251974106,
0.048531271517276764,
-0... | 0.041693 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Sparse Autoencoder Training Provides guidance for training and analyzing Sparse Autoencoders (SAEs) using SAELens to decompose neural network activations into interpretab... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-saelens.md | main | hermes-agent | [
-0.09808814525604248,
-0.04423217475414276,
-0.02755754627287388,
0.036764081567525864,
-0.020539885386824608,
-0.019745755940675735,
-0.020950734615325928,
0.003714041318744421,
-0.12145335972309113,
-0.01685790903866291,
-0.0029399609193205833,
-0.03956562653183937,
0.013206174597144127,
... | 0.123598 |
= model.to\_str\_tokens(tokens[0, pos:pos+1])[0] print(f"Token '{token}': features {top\_features.indices.tolist()}") # 5. Reconstruct activations reconstructed = sae.decode(sae\_features) reconstruction\_error = (activations - reconstructed).norm() ``` ### Available Pre-trained SAEs | Release | Model | Layers | |-----... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-saelens.md | main | hermes-agent | [
-0.10862406343221664,
-0.027220571413636208,
0.00436037965118885,
0.12307868897914886,
0.1020313948392868,
0.010269607417285442,
0.021032582968473434,
0.04775721952319145,
-0.08615057915449142,
-0.041607242077589035,
-0.04629642516374588,
-0.044956788420677185,
-0.002952050417661667,
-0.03... | 0.077799 |
features at final position features = sae.encode(cache["resid\_pre", 8])[0, -1] # [d\_sae] # Get logit attribution per feature # Feature contribution = feature\_activation Γ decoder\_weight Γ unembedding W\_dec = sae.W\_dec # [d\_sae, d\_model] W\_U = model.W\_U # [d\_model, vocab] # Contribution to "Paris" logit paris... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-saelens.md | main | hermes-agent | [
0.014593768864870071,
0.021280992776155472,
0.018711557611823082,
0.03368198871612549,
0.09575232118368149,
-0.011146372184157372,
0.0672461986541748,
0.05528424307703972,
-0.054928503930568695,
-0.0008439303492195904,
0.023433435708284378,
-0.08386792242527008,
0.0245230570435524,
-0.0788... | 0.024272 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Lambda Labs Gpu Cloud Reserved and on-demand GPU cloud instances for ML training and inference. Use when you need dedicated GPU instances with simple SSH access, persiste... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-lambda-labs.md | main | hermes-agent | [
-0.044821932911872864,
-0.04696689918637276,
-0.05581812933087349,
0.017695793882012367,
-0.003744585672393441,
-0.03603007644414902,
-0.027848731726408005,
-0.07174728810787201,
-0.09292744845151901,
0.01892915926873684,
0.008704083040356636,
0.014768382534384727,
0.04772741347551346,
-0.... | 0.030449 |
1x GPU: Fine-tuning, inference, development ``` ### Launch times - Single-GPU: 3-5 minutes - Multi-GPU: 10-15 minutes ## Lambda Stack All instances come with Lambda Stack pre-installed: ```bash # Included software - Ubuntu 22.04 LTS - NVIDIA drivers (latest) - CUDA 12.x - cuDNN 8.x - NCCL (for multi-GPU) - PyTorch (lat... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-lambda-labs.md | main | hermes-agent | [
-0.021441252902150154,
-0.060383863747119904,
-0.05279739573597908,
0.0321355015039444,
-0.00012819432595279068,
0.004317989572882652,
-0.05402495712041855,
-0.011586852371692657,
-0.02664289064705372,
-0.0547148771584034,
-0.0009610448032617569,
-0.07214058935642242,
-0.016178136691451073,
... | -0.024512 |
# Clone repo git clone https://github.com/user/project cd project # Install dependencies pip install -r requirements.txt # Train python train.py --epochs 100 --checkpoint-dir /lambda/nfs/storage/checkpoints ``` ### Multi-GPU training (single node) ```python # train\_ddp.py import torch import torch.distributed as dist ... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-lambda-labs.md | main | hermes-agent | [
-0.014850934967398643,
-0.10385750234127045,
-0.019346794113516808,
0.025431450456380844,
0.018060941249132156,
-0.0692012682557106,
-0.028758225962519646,
0.03059833124279976,
-0.09280738234519958,
-0.06812401860952377,
0.021891266107559204,
-0.021472400054335594,
0.0021697862539440393,
-... | 0.018227 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Whisper OpenAI's general-purpose speech recognition model. Supports 99 languages, transcription, translation to English, and language identification. Six model sizes from... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-whisper.md | main | hermes-agent | [
-0.07122287154197693,
-0.09020194411277771,
-0.025265969336032867,
-0.033521492034196854,
-0.0031077300664037466,
-0.054232943803071976,
-0.0031033519189804792,
-0.05059755966067314,
-0.02971121296286583,
-0.05645069479942322,
0.011829007416963577,
-0.01407388411462307,
0.013238363899290562,... | 0.19952 |
Timestamps ```python # Word-level timestamps result = model.transcribe("audio.mp3", word\_timestamps=True) for segment in result["segments"]: for word in segment["words"]: print(f"{word['word']} ({word['start']:.2f}s - {word['end']:.2f}s)") ``` ### Temperature fallback ```python # Retry with different temperatures if c... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-whisper.md | main | hermes-agent | [
-0.019053321331739426,
-0.03662551939487457,
-0.015389773063361645,
0.06805668026208878,
-0.013328567147254944,
-0.036431778222322464,
0.0020684702321887016,
0.03991566598415375,
-0.0081691425293684,
-0.042391885071992874,
0.016815925016999245,
-0.07553345710039139,
0.028341902419924736,
-... | 0.082221 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Instructor Extract structured data from LLM responses with Pydantic validation, retry failed extractions automatically, parse complex JSON with type safety, and stream pa... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-instructor.md | main | hermes-agent | [
-0.05434662476181984,
0.03262608125805855,
-0.061756681650877,
-0.03268028423190117,
-0.0160259697586298,
-0.0788896381855011,
-0.029932653531432152,
-0.020442306995391846,
-0.08113416284322739,
0.030353570356965065,
0.03511202707886696,
0.0028789634816348553,
0.07476776838302612,
-0.02850... | 0.060609 |
import Enum class Sentiment(str, Enum): POSITIVE = "positive" NEGATIVE = "negative" NEUTRAL = "neutral" class Review(BaseModel): text: str sentiment: Sentiment # Only these 3 values allowed review = client.messages.create( model="claude-sonnet-4-5-20250929", max\_tokens=1024, messages=[{ "role": "user", "content": "Thi... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-instructor.md | main | hermes-agent | [
-0.02573339268565178,
0.030332643538713455,
-0.023687899112701416,
-0.041980866342782974,
0.029661918058991432,
-0.017925594002008438,
-0.018496617674827576,
0.013816311955451965,
0.008554250933229923,
-0.012560490518808365,
0.06564301997423172,
-0.07725097984075546,
0.08552971482276917,
0... | 0.000864 |
founded in 2003. It operates in the automotive and energy industry with approximately 140,000 employees. The company is headquartered in Austin, Texas. """ company = client.messages.create( model="claude-sonnet-4-5-20250929", max\_tokens=1024, messages=[{ "role": "user", "content": f"Extract company information from: {... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-instructor.md | main | hermes-agent | [
-0.01461855135858059,
0.05634499713778496,
0.01729961484670639,
0.062423791736364365,
0.02729957178235054,
-0.03810161352157593,
0.02542661502957344,
-0.02179034985601902,
0.01655399799346924,
-0.08178167790174484,
0.02392517775297165,
-0.035981472581624985,
0.021236810833215714,
-0.040003... | 0.103438 |
Good: Constrain values class Rating(BaseModel): score: int = Field(ge=1, le=5, description="Rating from 1 to 5 stars") review: str = Field(min\_length=10, description="Review text, at least 10 chars") ``` ### 3. Provide Examples in Prompts ```python messages = [{ "role": "user", "content": """Extract person info from: ... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-instructor.md | main | hermes-agent | [
0.021151036024093628,
0.1016281247138977,
-0.0009188556578010321,
0.04248477518558502,
0.00751554686576128,
0.023382073268294334,
0.027769889682531357,
0.03311539068818092,
-0.09108778089284897,
-0.019105149433016777,
0.07566691190004349,
-0.10531270503997803,
0.0479779914021492,
-0.041092... | 0.033944 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Fine Tuning With Trl TRL: SFT, DPO, PPO, GRPO, reward modeling for LLM RLHF. ## Skill metadata | | | |---|---| | Source | Optional β install with `hermes skills install o... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-training-trl-fine-tuning.md | main | hermes-agent | [
-0.08720647543668747,
-0.04634105786681175,
-0.052704181522130966,
0.011978315189480782,
-0.05132434517145157,
-0.08436813205480576,
0.03649395331740379,
-0.015970228239893913,
-0.09027411788702011,
-0.0022514292504638433,
-0.018137101083993912,
0.0052964952774345875,
0.043375734239816666,
... | 0.115102 |
Dataset format: ```json { "prompt": "What is the capital of France?", "chosen": "The capital of France is Paris.", "rejected": "I don't know." } ``` Load dataset: ```python from datasets import load\_dataset dataset = load\_dataset("trl-lib/ultrafeedback\_binarized", split="train") # Or load your own # dataset = load\_... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-training-trl-fine-tuning.md | main | hermes-agent | [
-0.018341567367315292,
0.00901122484356165,
0.01580311916768551,
-0.012192605063319206,
-0.04622817412018776,
-0.039021894335746765,
0.015758464112877846,
0.07621883600950241,
-0.04239589348435402,
-0.04067394137382507,
0.03942284733057022,
-0.09599771350622177,
-0.019551852717995644,
-0.0... | 0.02191 |
quality\*\* Tune beta parameter: ```python # Higher beta = more conservative (stays closer to reference) config = DPOConfig(beta=0.5) # Default 0.1 # Lower beta = more aggressive alignment config = DPOConfig(beta=0.01) ``` \*\*Issue: Reward model not learning\*\* Check loss type and learning rate: ```python config = Re... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-training-trl-fine-tuning.md | main | hermes-agent | [
-0.09436452388763428,
-0.03210577741265297,
-0.00013741203292738646,
0.04358934983611107,
-0.06251341849565506,
0.00391944358125329,
-0.00858659390360117,
0.018266066908836365,
-0.035222262144088745,
-0.01965322531759739,
-0.02073482796549797,
-0.046844080090522766,
0.016444962471723557,
-... | -0.021521 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Peft Fine Tuning Parameter-efficient fine-tuning for LLMs using LoRA, QLoRA, and 25+ methods. Use when fine-tuning large models (7B-70B) with limited GPU memory, when you... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-peft.md | main | hermes-agent | [
-0.0752178505063057,
-0.005904872436076403,
-0.01883349008858204,
-0.017923902720212936,
-0.04617569223046303,
-0.07033786922693253,
-0.02774137631058693,
0.019525691866874695,
-0.12620563805103302,
-0.0005018868832848966,
0.01725234091281891,
-0.033932626247406006,
0.06234538182616234,
-0... | 0.087075 |
(enables gradient checkpointing) model = prepare\_model\_for\_kbit\_training(model) # LoRA config for QLoRA lora\_config = LoraConfig( r=64, # Higher rank for 70B lora\_alpha=128, lora\_dropout=0.1, target\_modules=["q\_proj", "v\_proj", "k\_proj", "o\_proj", "gate\_proj", "up\_proj", "down\_proj"], bias="none", task\_... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-peft.md | main | hermes-agent | [
-0.052886079996824265,
-0.05767590180039406,
-0.08326668292284012,
0.011416513472795486,
0.04852830991148949,
-0.017857344821095467,
0.007674288935959339,
0.052661292254924774,
-0.0865413248538971,
-0.08747149258852005,
-0.03820793330669403,
-0.02762500010430813,
0.06763889640569687,
-0.01... | -0.008335 |
SFTTrainer( model=model, args=SFTConfig(output\_dir="./output", max\_seq\_length=512), train\_dataset=dataset, peft\_config=lora\_config, # Pass LoRA config directly ) trainer.train() ``` ### With Axolotl (YAML config) ```yaml # axolotl config.yaml adapter: lora lora\_r: 16 lora\_alpha: 32 lora\_dropout: 0.05 lora\_tar... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-peft.md | main | hermes-agent | [
-0.058432698249816895,
-0.02937174029648304,
-0.022611426189541817,
-0.028467688709497452,
0.040088951587677,
-0.02751901000738144,
0.0014623303432017565,
0.0663711428642273,
-0.0431973896920681,
-0.10172685235738754,
-0.005027168896049261,
-0.06354010850191116,
-0.031000468879938126,
-0.0... | 0.024036 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Faiss Facebook's library for efficient similarity search and clustering of dense vectors. Supports billions of vectors, GPU acceleration, and various index types (Flat, I... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-faiss.md | main | hermes-agent | [
-0.08359623700380325,
-0.028048472478985786,
-0.07973185926675797,
-0.03256703540682793,
0.03245723620057106,
-0.07896330952644348,
-0.016994187608361244,
-0.025585027411580086,
-0.07686765491962433,
0.013877924531698227,
-0.0003984467766713351,
0.028967369347810745,
0.06176459416747093,
-... | 0.145992 |
= faiss.index\_cpu\_to\_gpu(res, 0, index\_cpu) # GPU 0 # Multi-GPU index\_gpu = faiss.index\_cpu\_to\_all\_gpus(index\_cpu) # 10-100Γ faster than CPU ``` ## LangChain integration ```python from langchain\_community.vectorstores import FAISS from langchain\_openai import OpenAIEmbeddings # Create FAISS vector store vec... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-faiss.md | main | hermes-agent | [
-0.015301200561225414,
-0.003985297400504351,
-0.06141797453165054,
0.014248084276914597,
0.0478750504553318,
-0.11231251060962677,
-0.04771742597222328,
0.015833567827939987,
0.04505256563425064,
-0.04120645299553871,
0.008832711726427078,
-0.014377702958881855,
0.027571434155106544,
-0.0... | 0.061837 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Llava Large Language and Vision Assistant. Enables visual instruction tuning and image-based conversations. Combines CLIP vision encoder with Vicuna/LLaMA language models... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-llava.md | main | hermes-agent | [
-0.04391862452030182,
-0.023806311190128326,
-0.013216978870332241,
-0.07819456607103348,
-0.017377551645040512,
-0.07202908396720886,
-0.024138744920492172,
0.004637506790459156,
-0.036023207008838654,
0.0036207367666065693,
0.009884928353130817,
-0.06141766533255577,
0.02075766585767269,
... | 0.165468 |
image?") conv.append\_message(conv.roles[1], None) response1 = generate(conv, model, image) # "A dog playing in a park" # Turn 2 conv.messages[-1][1] = response1 # Add previous response conv.append\_message(conv.roles[0], "What breed is the dog?") conv.append\_message(conv.roles[1], None) response2 = generate(conv, mod... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-llava.md | main | hermes-agent | [
-0.005024392157793045,
0.013338086195290089,
0.02074485458433628,
0.06479236483573914,
0.02519502490758896,
-0.07161717116832733,
0.008136766962707043,
0.03213009238243103,
0.0003378814726602286,
-0.048278018832206726,
0.05241389200091362,
-0.06138569489121437,
0.09507355839014053,
0.05768... | 0.007612 |
{/\* This page is auto-generated from the skill's SKILL.md by website/scripts/generate-skill-docs.py. Edit the source SKILL.md, not this page. \*/} # Pytorch Fsdp Expert guidance for Fully Sharded Data Parallel training with PyTorch FSDP - parameter sharding, mixed precision, CPU offloading, FSDP2 ## Skill metadata | |... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-pytorch-fsdp.md | main | hermes-agent | [
-0.08341444283723831,
-0.015144187957048416,
-0.06427577137947083,
0.011358977295458317,
-0.002491404302418232,
-0.07760430872440338,
-0.019225168973207474,
0.005139274522662163,
-0.10271860659122467,
-0.028600430116057396,
0.014282322488725185,
-0.03421516716480255,
-0.032462771981954575,
... | 0.102539 |
input in inputs: >>> loss = model(input).sum() >>> loss.backward() >>> optim.step() >>> # All ranks reach here without hanging/erroring static notify\_join\_context(joinable)[source]# Notifies the join context manager that the calling process has not yet joined. Then, if throw\_on\_early\_termination=True, checks if un... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-pytorch-fsdp.md | main | hermes-agent | [
-0.04749300330877304,
-0.038591258227825165,
-0.023160360753536224,
0.06568370759487152,
0.012302816845476627,
-0.034935928881168365,
0.014499313198029995,
0.06068548187613487,
-0.03317943215370178,
-0.015954436734318733,
-0.02607111819088459,
-0.02147846296429634,
0.0895790308713913,
-0.0... | 0.087518 |
β β broadcast β β β ? β β β β all\_reduce β β β ? β β β β reduce β β β ? β β β β all\_gather β β β ? β β β β gather β β β ? β β β β scatter β β β ? β β β β reduce\_scatter β β β β β β β β all\_to\_all β β β ? β β β β barrier β β β ? β β β β Backends that come with PyTorch# PyTorch distributed package supports Linux (st... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-pytorch-fsdp.md | main | hermes-agent | [
-0.05554578825831413,
-0.0648590549826622,
-0.09169778972864151,
0.0035296399146318436,
0.03467791527509689,
-0.06596800684928894,
-0.02896738238632679,
-0.023817190900444984,
-0.0331469364464283,
-0.015760593116283417,
-0.001260124146938324,
-0.0490950308740139,
0.017942162230610847,
0.00... | 0.147289 |
NCCL\_DEBUG\_SUBSYS=GRAPH to inspect the detailed detection result and save as reference if further help from NCCL team is needed. Performance tuning - NCCL performs automatic tuning based on its topology detection to save usersβ tuning effort. On some socket-based systems, users may still try tuning NCCL\_SOCKET\_NTHR... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-pytorch-fsdp.md | main | hermes-agent | [
-0.04433127120137215,
-0.058184392750263214,
-0.11732789874076843,
0.06488394737243652,
0.007316341623663902,
-0.1086069643497467,
-0.01551769208163023,
0.00027563361800275743,
-0.010736790485680103,
-0.012882115319371223,
-0.1017918512225151,
-0.04091351479291916,
-0.027470139786601067,
-... | 0.138777 |
registered for the device type indicated by the device\_id kwarg (if provided). The known default registrations today are: nccl for cuda, gloo for cpu, xccl for xpu. If neither backend nor device\_id is provided, c10d will detect the accelerator on the run-time machine and use a backend registered for that detected acc... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-pytorch-fsdp.md | main | hermes-agent | [
-0.07826787233352661,
-0.07387277483940125,
-0.14853116869926453,
-0.005278265103697777,
-0.020645463839173317,
-0.05461340397596359,
-0.014590682461857796,
0.032141730189323425,
-0.009233645163476467,
-0.06998451054096222,
-0.0011530079646036029,
-0.04315422847867012,
0.025432400405406952,
... | 0.188505 |
mesh\_shape, and mesh\_dim\_names parameters. This creates a DeviceMesh with an n-dimensional array layout, where n is the length of mesh\_shape. If mesh\_dim\_names is provided, each dimension is labeled as mesh\_dim\_names[i]. Note init\_device\_mesh follows SPMD programming model, meaning the same PyTorch Python pro... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-pytorch-fsdp.md | main | hermes-agent | [
-0.035452645272016525,
-0.035003360360860825,
-0.05017511919140816,
-0.01706085167825222,
0.003060264978557825,
-0.10555735975503922,
-0.02581007219851017,
0.0014290434774011374,
-0.01897062361240387,
-0.018856016919016838,
0.024194929748773575,
-0.015863077715039253,
0.028309842571616173,
... | 0.173552 |
machines in a group, along with a desired world\_size. The URL should start with file:// and contain a path to a non-existent file (in an existing directory) on a shared file system. File-system initialization will automatically create that file if it doesnβt exist, but will not delete the file. Therefore, it is your r... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-pytorch-fsdp.md | main | hermes-agent | [
-0.09074220061302185,
-0.028145793825387955,
-0.034551702439785004,
0.0370345376431942,
0.034672707319259644,
-0.06943269819021225,
-0.030468566343188286,
0.04156208038330078,
0.058120712637901306,
0.016895217821002007,
0.05589916184544563,
0.04242263734340668,
0.018117474392056465,
-0.008... | 0.098872 |
Note The entry Backend.UNDEFINED is present but only used as initial value of some fields. Users should neither use it directly nor assume its existence. classmethod register\_backend(name, func, extended\_api=False, devices=None)[source]# Register a new backend with the given name and instantiating function. This clas... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-pytorch-fsdp.md | main | hermes-agent | [
-0.09087883681058884,
-0.022308681160211563,
-0.1294132024049759,
0.041195064783096313,
-0.01843077316880226,
-0.05300237983465195,
0.0020612638909369707,
0.05478440597653389,
-0.014044170267879963,
-0.03527660295367241,
0.05341263860464096,
-0.02603396587073803,
-0.01494681928306818,
-0.0... | 0.114422 |
to synchronize the trainer processes using some means other than torch.distributed primitives \_after\_ calling destroy and before subsequently initializing. This behavior is currently unsupported/untested, due to the difficulty of achieving this synchronization, and is considered a known issue. Please file a github is... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-pytorch-fsdp.md | main | hermes-agent | [
-0.07936929166316986,
-0.12291811406612396,
-0.02503020316362381,
0.07017700374126434,
0.0023227545898407698,
-0.06710159778594971,
-0.04110735282301903,
-0.07971523702144623,
-0.0036400624085217714,
-0.04181826859712601,
0.02546256221830845,
0.03676074370741844,
-0.0776885524392128,
0.006... | 0.068221 |
or GroupMember.NON\_GROUP\_MEMBER if the rank is not part of ranks. N.B. use\_local\_synchronization doesnβt work with MPI. N.B. While use\_local\_synchronization=True can be significantly faster with larger clusters and small process groups, care must be taken since it changes cluster behavior as non-member ranks donβ... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-pytorch-fsdp.md | main | hermes-agent | [
-0.07765862345695496,
-0.11454881727695465,
-0.04746647924184799,
0.05981440842151642,
0.011879958212375641,
-0.02293470688164234,
0.003412393154576421,
-0.044646020978689194,
-0.03547277674078941,
-0.058271657675504684,
0.03682907298207283,
-0.0056971777230501175,
-0.00425606919452548,
-0... | 0.040891 |
2 hosts with 4 GPUs each. A reduction over the first dimension of mesh will reduce across columns (0, 4), .. and (3, 7), a reduction over the second dimension of mesh reduces across rows (0, 1, 2, 3) and (4, 5, 6, 7). Example: >>> from torch.distributed.device\_mesh import DeviceMesh >>> >>> # Initialize device mesh as... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-pytorch-fsdp.md | main | hermes-agent | [
0.00412956066429615,
0.00015217173495329916,
-0.017590150237083435,
0.027427755296230316,
0.04739092290401459,
-0.07895402610301971,
0.013905058614909649,
-0.027250085026025772,
-0.034919966012239456,
0.02203945256769657,
-0.01527191698551178,
-0.05096523091197014,
0.0004000767949037254,
0... | 0.058627 |
Calling mesh\_2d.get\_local\_rank(mesh\_dim=0) on rank 4, 5, 6, 7 would return 1. Calling mesh\_2d.get\_local\_rank(mesh\_dim=1) on rank 0, 4 would return 0. Calling mesh\_2d.get\_local\_rank(mesh\_dim=1) on rank 1, 5 would return 1. Calling mesh\_2d.get\_local\_rank(mesh\_dim=1) on rank 2, 6 would return 2. Calling me... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-pytorch-fsdp.md | main | hermes-agent | [
0.027403827756643295,
-0.03164198249578476,
0.017978888005018234,
0.014102346263825893,
0.028256582096219063,
-0.05002959445118904,
-0.0007573360926471651,
-0.007522708270698786,
-0.024866361171007156,
-0.01885138638317585,
0.02149069495499134,
-0.0808769091963768,
-0.011065163649618626,
0... | 0.023357 |
group (ProcessGroup, optional) β The process group to work on. If None, the default process group will be used. tag (int, optional) β Tag to match recv with remote send group\_src (int, optional) β Destination rank on group. Invalid to specify both src and group\_src. Returns A distributed request object. None, if not ... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-pytorch-fsdp.md | main | hermes-agent | [
-0.07407832145690918,
-0.07990994304418564,
-0.04564127326011658,
0.11557760089635849,
0.009343213401734829,
-0.061399076133966446,
0.06373713165521622,
0.02183184027671814,
-0.007772504352033138,
-0.0299614816904068,
-0.018949098885059357,
-0.0015751669416204095,
-0.03451245278120041,
0.0... | 0.072218 |
device (torch.device, optional) β If not None, receives on this device. Default is None. group\_src (int, optional) β Destination rank on group. Invalid to specify both src and group\_src. use\_batch (bool, optional) β If True, use batch p2p operations instead of regular send operations. This avoids initializing 2-rank... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-pytorch-fsdp.md | main | hermes-agent | [
-0.060358885675668716,
-0.05942109599709511,
-0.049476418644189835,
0.07901197671890259,
0.03636535629630089,
-0.05688779056072235,
0.07008538395166397,
0.03641656041145325,
-0.042489681392908096,
-0.047293223440647125,
0.015185524709522724,
-0.024422558024525642,
0.008243306539952755,
0.0... | 0.083792 |
point-to-point operations for batch\_isend\_irecv. This class builds the type of P2P operation, communication buffer, peer rank, Process Group, and tag. Instances of this class will be passed to batch\_isend\_irecv for point-to-point communications. Parameters op (Callable) β A function to send data to or receive data ... | https://github.com/NousResearch/hermes-agent/blob/main/website/docs/user-guide/skills/optional/mlops/mlops-pytorch-fsdp.md | main | hermes-agent | [
-0.05434297025203705,
-0.07748182117938995,
-0.055084582418203354,
0.06592530757188797,
-0.005563617218285799,
-0.05630355700850487,
0.04997103288769722,
0.01684452034533024,
-0.0162811316549778,
-0.02295936644077301,
0.007541370578110218,
0.07480479776859283,
-0.06889709085226059,
-0.0117... | 0.142874 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.