{"schema_version":"onlylabs.public_analysis_evidence.v1","title":"Nous Research analysis evidence pack","description":"Public onlylabs evidence pack for cited agent analysis: captured pages, ranked public signals, and stored web-search provenance used by the background analysis workflow.","url":"https://onlylabs.fyi/labs/nous","json_url":"https://onlylabs.fyi/analysis/nous/evidence.json","generated_at":"2026-06-11T18:06:29.215Z","org":{"slug":"nous","name":"Nous Research","category":"neolab","category_label":"Neolab","dossier_url":"https://onlylabs.fyi/labs/nous"},"analysis":null,"workflow":{"version":"onlylabs-deepagents-analysis-v3","provider":null,"model":null,"agent":null,"public_pack_mode":"local-pages-and-events","live_web_fetches":false,"note":"Public evidence exports do not trigger live Exa calls; stored Exa provenance is included when analysis metadata contains it."},"stats":{"pages":28,"events":137,"web":0,"evidence":88,"signal_desks":{"hiring":12,"forks":12,"releases":27,"talking":0,"repos":9},"data_radar_lanes":null,"data_radar_matches":null,"stored_analysis_evidence":null,"stored_analysis_web":null,"stored_analysis_signal_desks":null,"stored_analysis_data_radar_lanes":null,"stored_analysis_data_radar_matches":null},"stored_web_provenance":null,"evidence":[{"ref":"P1","kind":"page","title":"NousResearch/llama.cpp repository metadata","date":"2026-06-11T03:17:53.849432+00:00","date_source":null,"source_url":"https://github.com/NousResearch/llama.cpp","signal_url":null,"signal_json_url":null,"text":"# NousResearch/llama.cpp\n\nLanguage: C++\n\nLicense: MIT\n\nStars: 4\n\nForks: 1\n\nOpen issues: 1\n\nCreated: 2023-07-31T15:39:28Z\n\nPushed: 2024-03-10T05:26:15Z\n\nDefault branch: master\n\nFork: no\n\nArchived: no\n\nREADME:\n# llama.cpp\n\n![llama](https://user-images.githubusercontent.com/1991296/230134379-7181e485-c521-4d23-a0d6-f7b3b61ba524.png)\n\n[![License: MIT](https://img.shields.io/badge/license-MIT-blue.svg)](https://opensource.org/licenses/MIT)\n\n[Roadmap](https://github.com/users/ggerganov/projects/7) / [Project status](https://github.com/ggerganov/llama.cpp/discussions/3471) / [Manifesto](https://github.com/ggerganov/llama.cpp/discussions/205) / [ggml](https://github.com/ggerganov/ggml)\n\nInference of Meta's [LLaMA](https://arxiv.org/abs/2302.13971) model (and others) in pure C/C++\n\n> [!IMPORTANT]\n> **Quantization blind testing: https://github.com/ggerganov/llama.cpp/discussions/5962**\n>\n> Vote for which quantization type provides better responses, all other parameters being the same.\n\n### Recent API changes\n\n- [2024 Mar 8] `llama_kv_cache_seq_rm()` returns a `bool` instead of `void`, and new `llama_n_max_seq()` returns the upper limit of acceptable `seq_id` in batches (relevant when dealing with multiple sequences) https://github.com/ggerganov/llama.cpp/pull/5328\n- [2024 Mar 4] Embeddings API updated https://github.com/ggerganov/llama.cpp/pull/5796\n- [2024 Mar 3] `struct llama_context_params` https://github.com/ggerganov/llama.cpp/pull/5849\n\n### Hot topics\n\n- Initial Mamba support has been added: https://github.com/ggerganov/llama.cpp/pull/5328\n\n----\n\n
\nTable of Contents\n
    \n
  1. \nDescription\n
    2. \n
  2. \nUsage\n
      \n
    • Get the Code
      • \n
    • Build
      • \n
    • BLAS Build
      • \n
    • Prepare and Quantize
      • \n
    • Run the quantized model
      • \n
    • Memory/Disk Requirements
      • \n
    • Quantization
      • \n
    • Interactive mode
      • \n
    • \n
      \n

      \n\n

      Obsidian is a joint work between Nous Research and Virtual Interactive. Special thanks to LDJ and qnguyen3 for making this work possible.

      \n\nEasiest way to try out: [Colab](https://colab.research.google.com/drive/1C1FkoeZYBv3dZELaKgxahoZzWPfz0En8?usp=sharing) - After open the Gradio, give the model about 2 minutes to load then refresh the Gradio.\n## Usage\n1. Install Obsidian\n\n- Clone this project and navigate to the Obsidian folder\n\n```bash\ngit clone https://github.com/NousResearch/Obsidian.git\ncd Obsidian\n```\n\n- Download the multimodal projector from Huggingface\n\n```bash\nsh script/download_mm_projector.sh\n```\n\n- Install packages\n\n```bash\nconda create -n obsidian python=3.10 -y\nconda activate obsidian\npip install --upgrade pip # enable PEP 660 support\npip install -e .\n```\n\n- Install additional packages for training cases (required)\n\n```bash\npip install ninja\npip install flash-attn --no-build-isolation\n```\n\n- Install the latest version of `transformers`\n\n```bash\npip install --upgrade transformers==4.34.0\n```\n\n2. Run the Demo UI\n\n#### Launch a controller\n```Shell\npython -m llava.serve.controller --host 0.0.0.0 --port 10000\n```\n\n#### Launch a gradio web server.\n```Shell\npython -m llava.serve.gradio_web_server --controller http://localhost:10000 --model-list-mode reload\n```\nYou just launched the Gradio web interface. Now, you can open the web interface with the URL printed on the screen. You may notice that there is no model in the model list. Do not worry, as we have not launched any model worker yet. It will be automatically updated when you launch a model worker.\n\n#### Launch a model worker\n\nThis is the actual *worker* that performs the i"},{"ref":"P4","kind":"page","title":"NousResearch/StripedHyenaTrainer repository metadata","date":"2026-06-11T03:17:53.095245+00:00","date_source":null,"source_url":"https://github.com/NousResearch/StripedHyenaTrainer","signal_url":null,"signal_json_url":null,"text":"# NousResearch/StripedHyenaTrainer\n\nLanguage: Python\n\nLicense: Apache-2.0\n\nStars: 68\n\nForks: 11\n\nOpen issues: 0\n\nCreated: 2023-11-22T19:44:23Z\n\nPushed: 2023-12-08T20:26:46Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\nThis is the training code used to train [StripedHyena-Nous-7B](https://huggingface.co/togethercomputer/StripedHyena-Nous-7B).\n\nFirst, tokenize your data\n\n```sh\npython tokenization.py \\\n--dataset your-super-cool-sharegpt-format-dataset \\\n--tokenizer togethercomputer/StripedHyena-Hessian-7B \\\n--output tokenized \\\n--num-proc 32 \\\n--pad-to-length 4096 \\\n--truncate\n```\n\nMake sure you have done `accelerate config` -- we used the provided DeepSpeed configuration.\nThen, train!\n\n```sh\naccelerate launch finetune.py \\\n--model togethercomputer/StripedHyena-Hessian-7B \\\n--dataset tokenized \\\n--output-dir output \\\n--epochs 4 \\\n--batch-size 12 \\\n--gradient-accumulate-every 12 \\\n--warmup-steps 350 \\\n--learning-rate 0.000004 \\\n--lr-schedule linear \\\n--weight-decay 0.1 \\\n--checkpointing-steps 1000 \\\n--no-decay poles residues\n```\n\nThe `--no-decay` option disables weight decay on *only* the specified parameters.\nFor StripedHyena, we've found that disabling weight decay on the Hyena operator's `poles` and `residues` parameters improves performance.\nThere is also an option `--frozen` that can completely freeze select parameter groups."},{"ref":"P5","kind":"page","title":"NousResearch/Open-Reasoning-Tasks repository metadata","date":"2026-06-11T03:17:52.758909+00:00","date_source":null,"source_url":"https://github.com/NousResearch/Open-Reasoning-Tasks","signal_url":null,"signal_json_url":null,"text":"# NousResearch/Open-Reasoning-Tasks\n\nDescription: A comprehensive repository of reasoning tasks for LLMs (and beyond)\n\nLanguage: JavaScript\n\nLicense: Apache-2.0\n\nStars: 493\n\nForks: 72\n\nOpen issues: 11\n\nCreated: 2024-07-22T03:40:31Z\n\nPushed: 2024-09-27T17:09:42Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n![image](https://github.com/user-attachments/assets/2527a05e-afbc-4145-9daa-96f0229600f6)\n\n[![Netlify Status](https://api.netlify.com/api/v1/badges/03dad91c-2330-4213-8cfc-db14c113da16/deploy-status)](https://app.netlify.com/sites/openreasoningtasks/deploys)\n\nWelcome to the **LLM Reasoning Task Collection** repository! This project is an open collaboration to create a comprehensive master list of reasoning tasks that can teach, elicit, or show reasoning samples to large language models (LLMs) for training purposes.\n\n## Contents\n\n- [Open Reasoning Tasks: LLM Reasoning Tasks Collection](#open-reasoning-tasks-llm-reasoning-tasks-collection)\n- [Contents](#contents)\n- [Introduction](#introduction)\n- [Contributing](#contributing)\n- [License](#license)\n- [Citation](#citation)\n\n## Introduction\n\nThe goal of this repository is to gather a diverse set of reasoning tasks designed to improve the reasoning capabilities of LLMs. Contributors are encouraged to submit tasks, provide examples, and optionally include diagrams or workflows to illustrate how the tasks function.\n\n## Resources\n\n### [Master Reasoning Tasks List](https://github.com/NousResearch/Open-Reasoning-Tasks/blob/main/tasks.md)\nYou can access the [main tasks list table by clicking here (or open tasks.md file in the top level directory)](https://github.com/NousResearch/Open-Reasoning-Tasks/blob/main/tasks.md)\n\"image\"\n\n### [Web Based Directory](https://reasoning.nousresearch.com)\nYou can access the full [table of reasoning tasks from our quarto based website by clicking here.](https://reasoning.nousresearch.com)\n\"image\"\n\n### AI Reasoning Papers Master List\nComing Soon\n\n### AI Reasoning Format"},{"ref":"P6","kind":"page","title":"NousResearch/finetuning-subnet repository metadata","date":"2026-06-11T03:17:52.738958+00:00","date_source":null,"source_url":"https://github.com/NousResearch/finetuning-subnet","signal_url":null,"signal_json_url":null,"text":"# NousResearch/finetuning-subnet\n\nLanguage: Python\n\nLicense: MIT\n\nStars: 128\n\nForks: 42\n\nOpen issues: 13\n\nCreated: 2024-01-10T15:17:39Z\n\nPushed: 2024-05-19T19:27:56Z\n\nDefault branch: master\n\nFork: no\n\nArchived: no\n\nREADME:\n
      \n\n# **Nous Finetuning Subnet** \n[![Nous](/docs/nous.png)](https://nousresearch.com/)\n[![Bittensor](/docs/taologo.png)](https://bittensor.com/)\n\n---\n\n[Nous Discord](https://discord.gg/rJajyT2fYs) โ€ข [Bittensor Discord](https://discord.gg/bittensor) โ€ข [Network](https://taostats.io/)\n\n---\n\n[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT) \n\n
      \n\n# Introduction\n\n> **Note:** The following documentation assumes you are familiar with basic Bittensor concepts: Miners, Validators, and incentives. If you need a primer, please check out https://docs.bittensor.com/learn/bittensor-building-blocks.\n\nThe Nous-Bittensor subnet rewards miners for fine-tuning Large Language Models (LLMs) with data generated from a continuous stream of synthetic data provided by subnet 18 (also on Bittensor). It is the first-ever continuous fine-tuning benchmark, with new data generated daily, and the first incentivized fine-tuning benchmark. Additionally, it is the first Bittensor subnet to perform true cross-boundary communication, where data from one subnet is utilized in a secondary subnet.\n\nThe mechanism works like this:\n\n1. Miners train and periodically publish models to ๐Ÿค— Hugging Face and commit the metadata for that model to the Bittensor chain to prove the time of training.\n2. Validators download the models from ๐Ÿค— Hugging Face for each miner based on the Bittensor chain metadata and continuously evaluate them, setting weights based on the performance of each model against the synthetic data. \n3. The Bittensor chain aggregates weights from all active validators using Yuma Consensus to determine the proportion of TAO emission rewarded to miners and validators.\n\nSee the [Miner](docs/miner.md) and [Validator](docs/validator.md) docs for more information about how they work, as well as setup instructions.\n\n---\n\n## Incentive Mechanism\n\nBittensor hosts multiple incentive mecha"},{"ref":"P7","kind":"page","title":"NousResearch/infini-attention repository metadata","date":"2026-06-11T03:17:52.344193+00:00","date_source":null,"source_url":"https://github.com/NousResearch/infini-attention","signal_url":null,"signal_json_url":null,"text":"# NousResearch/infini-attention\n\nLanguage: Python\n\nStars: 5\n\nForks: 4\n\nOpen issues: 0\n\nCreated: 2024-07-08T18:51:03Z\n\nPushed: 2024-07-09T12:54:54Z\n\nDefault branch: master\n\nFork: no\n\nArchived: no\n\nREADME: none published or not readable through the GitHub API."},{"ref":"P8","kind":"page","title":"NousResearch/Hermes-Function-Calling repository metadata","date":"2026-06-11T03:17:52.342186+00:00","date_source":null,"source_url":"https://github.com/NousResearch/Hermes-Function-Calling","signal_url":null,"signal_json_url":null,"text":"# NousResearch/Hermes-Function-Calling\n\nLanguage: Jupyter Notebook\n\nLicense: MIT\n\nStars: 1386\n\nForks: 198\n\nOpen issues: 29\n\nCreated: 2024-02-24T09:00:28Z\n\nPushed: 2025-12-22T14:09:27Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# Hermes-Function-Calling\n\nThis repository contains code for the Hermes Pro Large Language Model to perform function calling based on the provided schema. It allows users to query the model and retrieve information related to stock prices, company fundamentals, financial statements, and more.\n\n## Installation\n\nTo install the required packages, run the following command:\n\n```bash\npip install -r requirements.txt\n```\n\n## Usage\n### Function calling\n\nTo run the function call inference with a query, use the following command:\n\n```bash\npython functioncall.py --query \"I need the current stock price of Tesla (TSLA)\"\n```\n\n### Json mode\n\nTo run the json mode inference with a query, use the following command:\n\n```bash\npython jsonmode.py --query \"Please return a json object to represent Goku from the anime Dragon Ball Z?\"\n\n```\n\n#### Command Line Arguments\n\n- `--model_path`: Path to the model folder (default: \"NousResearch/Hermes-2-Pro-Llama-3-8B\").\n- `--chat_template`: Chat template for prompt formatting (default: \"chatml\").\n- `--num_fewshot`: Option to include few-shot examples (default: None).\n- `--load_in_4bit`: Option to load in 4bit with bitsandbytes (default: \"False\").\n- `--query`: Query to be used for function call inference (default: \"I need the current stock price of Tesla (TSLA)\").\n- `--max_depth`: Maximum number of recursive iterations (default: 5).\n\n## Adding Custom Functions\n\nTo add your own functions for the model to use, you can modify the `functions.py` script. This script contains various functions that retrieve stock-related information using the `yfinance` library.\n\nHere's an example of how to add a new function:\n\n```python\n@tool\ndef get_new_function(symbol: str) -> dict:\n\"\"\"\nDescription of the new function.\nArgs:\nsymbol (str): The stock symbol.\nReturns:\ndict: Dictionary containing the desired information.\n\"\"\"\ntry:\n# Implement the logic to retrieve the desired information\n# using the yfinance library or any other relevant"},{"ref":"P9","kind":"page","title":"NousResearch/DisTrO repository metadata","date":"2026-06-11T03:17:51.742528+00:00","date_source":null,"source_url":"https://github.com/NousResearch/DisTrO","signal_url":null,"signal_json_url":null,"text":"# NousResearch/DisTrO\n\nDescription: Distributed Training Over-The-Internet\n\nStars: 1034\n\nForks: 56\n\nOpen issues: 1\n\nCreated: 2024-08-26T15:30:54Z\n\nPushed: 2025-10-14T15:37:16Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# DisTrO\n\nThis is the repository for DisTrO (Distributed Training Over-The-Internet), a family of low latency distributed optimizers that reduce inter-GPU communication requirements by three to four orders of magnitude.\n\n- [x] Aug. 26th, 2024: DisTrO [(Preliminary Report)](https://github.com/NousResearch/DisTrO/raw/main/A_Preliminary_Report_on_DisTrO.pdf)\n- [x] Dec. 2nd, 2024: DeMo Optimization [(Paper)](https://arxiv.org/abs/2411.19870) [(Code)](https://github.com/bloc97/DeMo), original seed research/idea for DisTrO\n- [x] Dec. 2nd, 2024: [Nous trains a 15b model using DisTrO](https://distro.nousresearch.com/)\n- [x] May 14th, 2025: [Psyche Network](https://nousresearch.com/nous-psyche/)\n- [x] May 14th, 2025: [Nous Consilience 40b LLM](https://psyche.network/runs/consilience-40b-1/0), [Huggingface](https://huggingface.co/PsycheFoundation/consilience-40b-7Y9v38s5)\n- [x] Oct. 14th, 2025: DeMo Optimization [(Paper v2)](https://openreview.net/pdf?id=U9oewpa7cn), [(Production Code)](https://github.com/PsycheFoundation/psyche/blob/b13ff76f879796a071850bae2d82084f360d608d/shared/modeling/src/distro.rs)\n\n[Join us on Discord](https://discord.com/invite/jqVphNsB4H) if you're interested in helping research and build the future of distributed training."},{"ref":"P10","kind":"page","title":"NousResearch/solana-flake repository metadata","date":"2026-06-11T03:17:51.560593+00:00","date_source":null,"source_url":"https://github.com/NousResearch/solana-flake","signal_url":null,"signal_json_url":null,"text":"# NousResearch/solana-flake\n\nLanguage: Nix\n\nStars: 6\n\nForks: 4\n\nOpen issues: 1\n\nCreated: 2024-10-21T15:28:55Z\n\nPushed: 2026-03-16T19:40:47Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# Solana Flake\n\nBuilding on the work of\n\n- https://github.com/nasadorian/solflake\n- https://github.com/itsfarseen/solana-flake\n\nThis flake provides:\n\n## `solana-cli`\n\nThe full Solana CLI, including a functional build toolchain for the `cargo sbf` command.\n\n## `solana-rust`\n\nA `rustc` and `cargo` that can build for the Rust target `sbf-solana-solana`\n\n## `anchor`\n\nThe full Anchor CLI, including a functional build toolchain for the `anchor build` command."},{"ref":"P11","kind":"page","title":"NousResearch/wandb-rs repository metadata","date":"2026-06-11T03:17:51.374651+00:00","date_source":null,"source_url":"https://github.com/NousResearch/wandb-rs","signal_url":null,"signal_json_url":null,"text":"# NousResearch/wandb-rs\n\nLanguage: Rust\n\nStars: 16\n\nForks: 2\n\nOpen issues: 0\n\nCreated: 2024-10-26T03:12:43Z\n\nPushed: 2026-02-24T15:22:31Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# Weights & Biases for Rust\n\nSimple run creation & logging implemented!\n\n```rs\nlet wandb = WandB::new(BackendOptions::new(api_key));\n\nlet run = wandb\n.new_run(\nRunInfo::new(\"wandb-rs\")\n.entity(\"nous_research\")\n.name(\"node-25\")\n.build()?,\n)\n.await?;\n\nfor i in 0..100 {\nrun.log(((\"_step\", i), (\"loss\", 1.0 / (i as f64).sqrt())))\n.await;\n}\n```\n\nsee `examples/test.rs` for an example :)"},{"ref":"P12","kind":"page","title":"NousResearch/forge-feedback repository metadata","date":"2026-06-11T03:17:51.35661+00:00","date_source":null,"source_url":"https://github.com/NousResearch/forge-feedback","signal_url":null,"signal_json_url":null,"text":"# NousResearch/forge-feedback\n\nDescription: Feedback and issues for the Forge Beta Reasoning API. Submit issues here!\n\nStars: 3\n\nForks: 3\n\nOpen issues: 1\n\nCreated: 2024-10-28T20:08:52Z\n\nPushed: 2024-10-28T20:08:53Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# forge-feedback\nFeedback and issues for the Forge Beta Reasoning API. Submit issues here!"},{"ref":"P13","kind":"page","title":"NousResearch/forge-api-demo repository metadata","date":"2026-06-11T03:17:51.165408+00:00","date_source":null,"source_url":"https://github.com/NousResearch/forge-api-demo","signal_url":null,"signal_json_url":null,"text":"# NousResearch/forge-api-demo\n\nDescription: Simple demo showing how to use the Forge API by Nous Research\n\nLanguage: Python\n\nLicense: MIT\n\nStars: 18\n\nForks: 9\n\nOpen issues: 4\n\nCreated: 2024-11-12T05:05:05Z\n\nPushed: 2024-11-12T05:40:07Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# forge-api-demo\n\nThis is a simple demo of the Forge API by Nous Research. It demonstrates how to use the asynchronous API to trigger a reasoning completion and wait for the result.\n\nIf you want to jump straigt to the code, see: [src/forge-api-demo/cli.py](https://github.com/NousResearch/forge-api-demo/blob/main/src/forge-api-demo/cli.py).\n\nSee also:\n* [About Forge API](https://gist.github.com/DamascusGit/c1523fb166109aa6af81af986f856f2d)\n* [API reference](https://forge-api.nousresearch.com/docs)\n\n## Setup\n\nInstall Poetry if you donโ€™t have it:\n\n```bash\ncurl -sSL https://install.python-poetry.org | python3 -\n```\n\nInstall dependencies:\n\n```bash\npoetry install\n```\n\n## Run\n\nHere's how to run the command and see its help output:\n```bash\n> poetry run python src/forge-api-demo/cli.py -h\n```\n\nWhich shows:\n```\nusage: forge-api-demo [-h] -p PROMPT [-r {fast,medium,slow}] [-t]\n\nSimple demo of the Forge API by Nous Research. Run with env var FORGE_API_KEY set.\n\noptions:\n-h, --help show this help message and exit\n-p PROMPT, --prompt PROMPT\nYour prompt.\n-r {fast,medium,slow}, --reasoning-speed {fast,medium,slow}\nThe reasoning preset to use: fast/medium/slow. Defaults to medium.\n-t, --track Whether to return detailed information about the\n\nFor more information, see: https://forge-api.nousresearch.com/docs\n```\n\nExample completion run:\n\n```bash\nFORGE_API_KEY= poetry run python src/forge-api-demo/cli.py -p \"How much wood would a theoretical 80kg woodchuck chuck assuming lunar gravity and a competitive woodchucking environment?\"\n```\n\n## Example outputs\n\n### Without tracking\n\nIf I want to ask:\n\n> How much wood would a theoretical 80kg woodchuck chuck assuming lunar gravity and a competitive woodchucking environment?\n\nI can run the command:\n\n```bash\n> FORGE_API_KEY= poetry run python src/forge-api-demo/cli.py -p \"How much wood would a theoretical 80kg woodchuck chuck as"},{"ref":"P14","kind":"page","title":"NousResearch/lm-eval-harness repository metadata","date":"2026-06-11T03:17:50.783568+00:00","date_source":null,"source_url":"https://github.com/NousResearch/lm-eval-harness","signal_url":null,"signal_json_url":null,"text":"# NousResearch/lm-eval-harness\n\nLanguage: Python\n\nLicense: MIT\n\nStars: 11\n\nForks: 3\n\nOpen issues: 2\n\nCreated: 2024-12-20T01:13:30Z\n\nPushed: 2025-06-29T00:23:31Z\n\nDefault branch: nous\n\nFork: no\n\nArchived: no\n\nREADME:\n# Language Model Evaluation Harness\n\n[![DOI](https://zenodo.org/badge/DOI/10.5281/zenodo.10256836.svg)](https://doi.org/10.5281/zenodo.10256836)\n\n---\n\n*Latest News ๐Ÿ“ฃ*\n- [2025/02] Added [SGLang](https://docs.sglang.ai/) support!\n- [2024/09] We are prototyping allowing users of LM Evaluation Harness to create and evaluate on text+image multimodal input, text output tasks, and have just added the `hf-multimodal` and `vllm-vlm` model types and `mmmu` task as a prototype feature. We welcome users to try out this in-progress feature and stress-test it for themselves, and suggest they check out [`lmms-eval`](https://github.com/EvolvingLMMs-Lab/lmms-eval), a wonderful project originally forking off of the lm-evaluation-harness, for a broader range of multimodal tasks, models, and features.\n- [2024/07] [API model](docs/API_guide.md) support has been updated and refactored, introducing support for batched and async requests, and making it significantly easier to customize and use for your own purposes. **To run Llama 405B, we recommend using VLLM's OpenAI-compliant API to host the model, and use the `local-completions` model type to evaluate the model.**\n- [2024/07] New Open LLM Leaderboard tasks have been added ! You can find them under the [leaderboard](lm_eval/tasks/leaderboard/README.md) task group.\n\n---\n\n## Announcement\n**A new v0.4.0 release of lm-evaluation-harness is available** !\n\nNew updates and features include:\n\n- **New Open LLM Leaderboard tasks have been added ! You can find them under the [leaderboard](lm_eval/tasks/leaderboard/README.md) task group.**\n- Internal refactoring\n- Config-based task creation and configuration\n- Easier import and sharing of externally-defined task config YAMLs\n- Support for Jinja2 prompt design, easy modification of prompts + prompt imports from Promptsource\n- More advanced configuration options, including output post-processing, answer extraction, and multiple LM generations per document, configurable fewshot settin"},{"ref":"P15","kind":"page","title":"NousResearch/nousflash-agents repository metadata","date":"2026-06-11T03:17:50.510089+00:00","date_source":null,"source_url":"https://github.com/NousResearch/nousflash-agents","signal_url":null,"signal_json_url":null,"text":"# NousResearch/nousflash-agents\n\nDescription: Modular Agentic AI Architecture - NousResearch x Teleport (Flashbots)\n\nLanguage: Python\n\nStars: 99\n\nForks: 19\n\nOpen issues: 10\n\nCreated: 2024-11-05T04:47:28Z\n\nPushed: 2025-01-08T09:14:30Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# nousflash\nhehe\nhehe 2 just cuz\n\n(small guest appearance by somewhere systems)\n\n### TODO:\n\n- wallet actions: the agent need to be able to decide when to use the wallet for transfer of assets (how much and to which address), Iโ€™ve added functions in wallet_send.py for this, TODO: make the agent decide when/if send transactions from the address\n\n- thinking about replies or subtweets to previous tweets: instead of news, I made the external data to be replies to the agentโ€™s previous tweets and recent mentions of the agent on twitter, this may need yall to change the agent prompt to make it aware how to respond\n\n### basics:\n\nDB folder has scripts to create and seed the database with some fake data. dokcer should automatically run all of this for you.\n\nengines contains all the functions that generate the content for the agent pipeline.\n\nThe pipeline.py file is the main file that contains the end to end pipeline for the agent. You can see the flow here.\n\n**run_pipeline.py** is the main file that runs the pipeline. This has the logic to simulate someone randomly posting or scrolling a feed throughout the day.\nThis is also the file that runs continuously in the background in the container.\n\n### Running the agent:\n\ndocker-compose up -d\n\nThis will start the agent in the background and run continuously.\n\nYou can also run the agent manually by running:\n\npython run_pipeline.py\n\nThis will run the pipeline LOCALLY and not in the container.\n\nenjoy"},{"ref":"P16","kind":"page","title":"NousResearch/kaida repository metadata","date":"2026-06-11T03:17:50.304843+00:00","date_source":null,"source_url":"https://github.com/NousResearch/kaida","signal_url":null,"signal_json_url":null,"text":"# NousResearch/kaida\n\nDescription: A prototype Kotlin LLM library.\n\nLanguage: Kotlin\n\nLicense: MIT\n\nStars: 9\n\nForks: 3\n\nOpen issues: 0\n\nCreated: 2025-03-18T21:19:02Z\n\nPushed: 2025-03-28T23:40:18Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# Kaida\n\nKaida is a prototype Kotlin library designed to simplify and standardize interactions with multiple large language model (LLM) APIs including OpenAI-compatible completion and chat endpoints as well as Anthropic. It also features a strongly typed Directed Acyclic Graph (DAG) pipeline for structuring complex multi-step workflows.\n\nDesigned for extremely rapid iteration, Kaida prioritizes:\n\n- Unified LLM abstractions that fully model the union set of all supported features using sum types.\n- Swap models (e.g. chat to completion, different APIs, sampler settings) without changing your application code at all.\n- Get compile time errors or fail fast at runtime - never silently do the wrong thing.\n- Write flexible code that can account for a variety of different features.\n- Fully asynchronous streaming API built on Kotlin's [asynchronous flows](https://kotlinlang.org/docs/flow.html) that supports canceling requests and parallelism\n- Automatic serialization of all intermediate states as well as outputs using kotlinx.serialization\n- Observable and debuggable: a pipeline's internal state is easily inspected at any point during execution\n- Decoupled design that lends itself well to dynamically driven UI\n- Every level of tool: low level LLM completion API, high level template and pipeline API built on top\n\nThis document introduces Kaida incrementally through practical examples.\n\n# Table of Contents\n\n- [Installation](#installation)\n- [Configuration](#configuration)\n- [Example Project](#example-project)\n- [Motivation: Why Use Kaida?](#motivation-why-use-kaida)\n- [Typed Feature Flags and Fail-Fast](#typed-feature-flags-and-fail-fast)\n- [Templates for Structured Prompts](#templates-for-structured-prompts)\n- [Building Complex Workflows: The Pipeline DSL](#building-complex-workflows-the-pipeline-dsl)\n- [Retry Policies](#retry-policies)\n- [Serialization and Reloading Pipelines](#serialization-and-reloading-pipelines)\n- [Simpl"},{"ref":"P17","kind":"page","title":"NousResearch/kaida-gencritique repository metadata","date":"2026-06-11T03:17:50.16759+00:00","date_source":null,"source_url":"https://github.com/NousResearch/kaida-gencritique","signal_url":null,"signal_json_url":null,"text":"# NousResearch/kaida-gencritique\n\nDescription: An example project for the Nous Research Kaida library.\n\nLanguage: Kotlin\n\nLicense: MIT\n\nStars: 4\n\nForks: 1\n\nOpen issues: 0\n\nCreated: 2025-03-19T18:40:18Z\n\nPushed: 2025-03-19T18:40:32Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# kaida-gencritique\n\nThis is an example project demonstrating Nous Research's [Kaida library](https://github.com/NousResearch/kaida). It will generate spells for a roleplaying game using a variety of LLM models, then critique them using a different set of LLMs. It will split the spells up so that each critique has one spell from each generative model, and then record which generative model each critic model preferred.\n\nFor more accurate statistics, of course a larger run would need to be performed. This is left as an exercise to interested readers.\n\nTo run this example as written you will need to provide valid API keys in `config\\auth.yaml` for OpenAI, Anthropic, and Fireworks. See `config\\auth.example.yaml` for more details."},{"ref":"P18","kind":"page","title":"NousResearch/storywriter repository metadata","date":"2026-06-11T03:17:49.514774+00:00","date_source":null,"source_url":"https://github.com/NousResearch/storywriter","signal_url":null,"signal_json_url":null,"text":"# NousResearch/storywriter\n\nDescription: Nous Research's experimental LLM story generator\n\nLanguage: Kotlin\n\nLicense: MIT\n\nStars: 8\n\nForks: 4\n\nOpen issues: 0\n\nCreated: 2025-03-19T21:09:53Z\n\nPushed: 2025-03-19T21:15:31Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\nThis is the [Kaida](https://github.com/NousResearch/kaida)-based backend for Nous Research's experimental LLM story generator. \n\nPlease see [here](https://github.com/NousResearch/storywriter-frontend) for the frontend!\n\nTo run:\n\n`./gradlew run`\n\nFor production:\n\n`./gradlew shadowJar`\n\nYou must configure authentication keys for Anthropic and Fireworks to use the repository as presented. Please review `config/auth.example.yaml`."},{"ref":"P19","kind":"page","title":"NousResearch/scaling-transformer repository metadata","date":"2026-06-11T03:17:49.362008+00:00","date_source":null,"source_url":"https://github.com/NousResearch/scaling-transformer","signal_url":null,"signal_json_url":null,"text":"# NousResearch/scaling-transformer\n\nLanguage: Python\n\nLicense: MIT\n\nStars: 4\n\nForks: 2\n\nOpen issues: 0\n\nCreated: 2025-07-14T12:15:17Z\n\nPushed: 2025-10-22T13:43:28Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# scaling-transformer"},{"ref":"P20","kind":"page","title":"NousResearch/atropos repository metadata","date":"2026-06-11T03:17:49.309585+00:00","date_source":null,"source_url":"https://github.com/NousResearch/atropos","signal_url":null,"signal_json_url":null,"text":"# NousResearch/atropos\n\nDescription: Atropos is a Language Model Reinforcement Learning Environments framework for collecting and evaluating LLM trajectories through diverse environments\n\nLanguage: Python\n\nLicense: MIT\n\nStars: 1273\n\nForks: 367\n\nOpen issues: 82\n\nCreated: 2025-04-29T19:02:06Z\n\nPushed: 2026-06-08T16:41:48Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# Atropos - Nous Research's LLM RL Gym\n\n![newatr-02](banner-image.jpg)\n\n
      \n\n*In Greek mythology, Atropos was the eldest of the three Fates. While her sisters spun and measured the threads of mortal lives, Atropos alone held the shears that would cut these threads, determining the final destiny of each soul. Just as Atropos guided souls to their ultimate fate, this system guides language models toward their optimal potential through reinforcement learning.*\n\n
      \n\n
      \n
      \n
      \n\n\"HuggingFace\"/\n\n\n\n\"Hermes\n

      \n\n# Hermes Agent โ˜ค\n

      \nHermes Agent | Hermes Desktop\n

      \n

      \n\"Documentation\"\n\"Discord\"\n\"License:\n\"Built\n\"ไธญๆ–‡\"\n\"ุงุฑุฏูˆ\"\n

      \n\n**The self-improving AI agent built by [Nous Research](https://nousresearch.com).** It's the only agent with a built-in learning loop โ€” it creates skills from experience, improves them during use, nudges itself to persist knowledge, searches its own past conversations, and builds a deepening model of who you are across sessions. Run it on a $5 VPS, a GPU cluster, or serverless infrastructure that costs nearly nothing when idle. It's not tied to your laptop โ€” talk to it from Telegram while it works on a cloud VM.\n\nUse any model you want โ€” [Nous Portal](https://portal.nousresearch.com), [OpenRouter](https://openrouter.ai) (200+ models), [NovitaAI](https://novita.ai) (AI-native cloud for Model API, Agent Sandbox, and"},{"ref":"P24","kind":"page","title":"NousResearch/nomos repository metadata","date":"2026-06-11T03:17:48.270498+00:00","date_source":null,"source_url":"https://github.com/NousResearch/nomos","signal_url":null,"signal_json_url":null,"text":"# NousResearch/nomos\n\nLanguage: Python\n\nLicense: MIT\n\nStars: 194\n\nForks: 21\n\nOpen issues: 8\n\nCreated: 2025-12-09T20:14:55Z\n\nPushed: 2025-12-18T06:05:44Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# Nomos\n

      \n\"X \"MIT \"Hugging\n

      \n

      \n\"image\"\n

      \n\nA reasoning harness for mathematical problem-solving and proof-writing in natural language.\n\n## Installation\n\n```bash\npip install -r requirements.txt\n```\n\n## Usage\n\n```bash\npython solve_agent.py [options]\n```\n\n### Required Argumentsmassiveaxe\n\n- `problems_dir`: Directory containing `.md` problem files\n\n### Options\n\n| Flag | Default | Description |\n|------|---------|-------------|\n| `--submissions_dir` | `submissions/{problems_dir}-{timestamp}` | Output directory for final submissions |\n| `--judge_prompt` | `prompts/score.md` | Judge prompt file |\n| `--solve_prompt` | `None` | Solver system prompt |\n| `--consolidation_prompt` | `prompts/consolidation.md` | Consolidation prompt |\n| `--pairwise_prompt` | `prompts/pairwise.md` | Pairwise comparison prompt |\n| `--time_limit_hours` | `3.0` | Total time limit |\n| `--max_concurrent` | `32` | Max parallel API requests |\n| `--target_perfect_scores` | `4` | Number of 7/7 scores needed per problem |\n| `--model` | `nomos-1` | Model for solving |\n| `--judge_model` | `nomos-1` | Model for judging |\n| `--base_url` | `http://localhost:30000/v1` | OpenAI-compatible API endpoint |\n\n## Workflow\n\nNomos keeps working on the problems you give it until its time limit runs out or it reaches a target number of self-critiqued perfect scores on every problem. Once either termination condi"},{"ref":"P25","kind":"page","title":"NousResearch/tinker-atropos repository metadata","date":"2026-06-11T03:17:48.26349+00:00","date_source":null,"source_url":"https://github.com/NousResearch/tinker-atropos","signal_url":null,"signal_json_url":null,"text":"# NousResearch/tinker-atropos\n\nDescription: Standalone repo for our Atropos integration with Thinking Machines Tinker API (https://thinkingmachines.ai/tinker/)\n\nLanguage: Python\n\nStars: 88\n\nForks: 25\n\nOpen issues: 5\n\nCreated: 2025-10-07T17:30:26Z\n\nPushed: 2026-03-22T21:35:09Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# tinker-atropos\n\nAn integration layer connecting Atropos (https://github.com/NousResearch/atropos) with the Thinking Machines Tinker API (https://thinkingmachines.ai/tinker/). This package enables seamless model training with Atropos environments from your local machine, abstracting away compute management and infrastructure concerns.\n\n## Installation\n\n```bash\npip install -e .\n```\n\n## Quickstart\n\nFirst, obtain a Tinker API key from https://tinker-console.thinkingmachines.ai/keys.\n\nRun the following commands in separate terminal windows to start a training run:\n\n```bash\n# Terminal 1: Start Atropos API\nrun-api\n\n# Terminal 2: Start training\nexport TINKER_API_KEY=\"\"\npython launch_training.py --config configs/default.yaml\n\n# Terminal 3: Start environment (use built-in or any Atropos environment)\npython tinker_atropos/environments/gsm8k_tinker.py serve --config configs/default.yaml\n```\n\nThis runs a 50-step training example with Llama-3.1-8B-Instruct on the GSM8k environment.\n\n## Using Any Atropos Environment\n\n**You can use any existing Atropos environment directly with Tinker!** Just point to it and pass your Tinker config:\n\n```bash\n# Use any Atropos environment with Tinker training\npython /path/to/atropos/environment.py serve --config /path/to/your_tinker_config.yaml\n```\n\n### Example with Atropos Math Environment\n\n```bash\n# Terminal 1: Start Atropos API\nrun-api\n\n# Terminal 2: Start Tinker training with math config\nexport TINKER_API_KEY=\"\"\npython launch_training.py --config configs/math_config.yaml\n\n# Terminal 3: Use Atropos math environment with Tinker config\npython ~/atropos/environments/math_server.py serve --config configs/math_config.yaml\n```\n\n### How It Works\n\nAtropos environments support a `--config` flag that loads your Tinker config (which follows the standard Atropos format with a `tinker` section). The environm"},{"ref":"P26","kind":"page","title":"NousResearch/iroh-fake-store repository metadata","date":"2026-06-11T03:17:48.251424+00:00","date_source":null,"source_url":"https://github.com/NousResearch/iroh-fake-store","signal_url":null,"signal_json_url":null,"text":"# NousResearch/iroh-fake-store\n\nLanguage: Rust\n\nStars: 2\n\nForks: 3\n\nOpen issues: 0\n\nCreated: 2025-10-02T16:03:23Z\n\nPushed: 2026-03-12T16:15:30Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# iroh-fake-store\n\n[![Crates.io](https://img.shields.io/crates/v/iroh-fake-store.svg)](https://crates.io/crates/iroh-fake-store)\n[![Documentation](https://docs.rs/iroh-fake-store/badge.svg)](https://docs.rs/iroh-fake-store)\n[![License](https://img.shields.io/crates/l/iroh-fake-store.svg)](LICENSE)\n\nfake `iroh-blobs` store for testing. generates data on-the-fly without storing anything in RAM or disk.\n\ncompatible with `iroh-blobs` 0.98 and `iroh` 0.96.\n\n## what it does\n\nfor testing with large blobs (like 2TB) when you don't care about actual content. features:\n\n- zero allocation: generates data on-the-fly\n- deterministic: same config = same hashes\n- configurable: zeros, ones, or pseudo-random data (per-store or per-blob)\n- safe limits: 10GB default max to prevent accidents\n- full protocol support: implements complete `iroh-blobs` store protocol\n- bandwidth throttling: simulate slow peers\n- real data round-trip: blobs added via `add_bytes`/`add_byte_stream` serve back their original data\n\n## installation\n\n```toml\n[dev-dependencies]\niroh-fake-store = \"0.1\"\n```\n\n## usage\n\n### basic\n\n```rust\nuse iroh_fake_store::FakeStore;\n\n#[tokio::main]\nasync fn main() {\nlet store = FakeStore::new([\n1024, // 1KB\n1024 * 1024, // 1MB\n1024 * 1024 * 1024, // 1GB\n]);\n\nlet hashes = store.blobs().list().hashes().await.unwrap();\nfor hash in hashes {\nlet status = store.blobs().status(hash).await.unwrap();\nprintln!(\"blob {} status: {:?}\", hash, status);\n}\n}\n```\n\n### with builder\n\n```rust\nuse iroh_fake_store::{FakeStore, DataStrategy};\n\nlet store = FakeStore::builder()\n.strategy(DataStrategy::PseudoRandom { seed: 42 })\n.max_blob_size(Some(100 * 1024 * 1024)) // 100MB max\n.with_blob(1024)\n.with_blob(2048)\n.build();\n```\n\n### unique blobs with distinct hashes\n\nwhen you need many same-sized blobs that are distinguishable by hash:\n\n```rust\nlet store = FakeStore::builder()\n.with_unique_blobs(100, 1024 * 1024) // 100 x 1MB blobs, each with a unique seed\n.build();\n```\n\n### bandwidth throttling\n\nsimulate s"},{"ref":"P27","kind":"page","title":"NousResearch/huskyholdem-bench repository metadata","date":"2026-06-11T03:17:48.13514+00:00","date_source":null,"source_url":"https://github.com/NousResearch/huskyholdem-bench","signal_url":null,"signal_json_url":null,"text":"# NousResearch/huskyholdem-bench\n\nLanguage: Python\n\nStars: 21\n\nForks: 2\n\nOpen issues: 1\n\nCreated: 2025-07-27T22:59:00Z\n\nPushed: 2025-09-05T05:48:56Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# ๐Ÿƒ Husky Hold'em Benchmark\n\nA comprehensive benchmark for evaluating Large Language Models' ability to generate strategic poker bots through iterative code generation, optimization, and competitive multi-agent gameplay.\n\n## Overview\n\nThe benchmark challenges LLMs to generate functional Python poker bots, refine them through five iterative improvement rounds based on performance feedback, and compete in comprehensive tournaments where success is measured by net monetary gains across a series of six-handed games.\n\n๐ŸŒ **Bench Website**: [huskybench.com](https://huskybench.com/) \n๐Ÿ“Š **API Documentation**: [api.huskybench.com/docs](https://api.huskybench.com/docs)\n\n## Architecture\n\n![System architecture diagram showing the interconnected components of the Husky Hold'em Benchmark](Pokerbot-arch.png)\n\n### Directory Overview\n\n- Frontend: Our web client that showing leaderboard and visualization of all games ran.\n- Server (the-betting-edge): Server that handle queueing and orchestrating muliple round of game on a large scale for our benchmark.\n- Poker-engine: Dockerize game server that ensures all games follow standard No-Limit Texas Hold'em rules, managing betting rounds, hand evaluation, and pot distribution.\n- Poker-client: Dockerize game client that execute bot's code file following common API in a sandbox environments.\n- LLM-engine: Iteration engine that generate bots using a variety of model and play test them to make sure they are functionaly good.\n\n## Workflow\n\n1. **Bot Generation**: The LLM-engine component creates `player.py` and `requirements.txt` files for each target model, implementing the poker bot logic and dependencies\n\n2. **Submission Process**: Generated code is automatically uploaded to model-specific accounts via the-betting-edge API, marking them as tournament-ready submissions\n\n3. **Tournament Execution**: The server orchestrates games by pairing bots from different accounts, with each bot running in its dedicated poker-client container. Each gam"},{"ref":"P28","kind":"page","title":"NousResearch/atropos v0.2.1","date":"2026-06-11T03:17:03.742392+00:00","date_source":null,"source_url":"https://github.com/NousResearch/atropos/releases/tag/v0.2.1","signal_url":null,"signal_json_url":null,"text":"# v0.2.1\n\nRepository: NousResearch/atropos\n\nTag: v0.2.1\n\nPublished: 2025-05-18T14:58:32Z\n\nPrerelease: no\n\nRelease notes:\n## What's Changed\n* Make run api not reload by @dmahan93 in https://github.com/NousResearch/atropos/pull/43\n* add code execution environment by @JoeLi12345 in https://github.com/NousResearch/atropos/pull/26\n* Blackjack2 env by @shannonsands in https://github.com/NousResearch/atropos/pull/38\n* fix validation errors by @hjc-puro in https://github.com/NousResearch/atropos/pull/45\n* Llms txt update by @shannonsands in https://github.com/NousResearch/atropos/pull/47\n* updated APIServerConfig and added requirements.txt and install instruโ€ฆ by @shannonsands in https://github.com/NousResearch/atropos/pull/46\n* Instruction following algo environment by @teknium1 in https://github.com/NousResearch/atropos/pull/44\n* Kernelbench env with parallel compilation by @sumo43 in https://github.com/NousResearch/atropos/pull/51\n* Added new env info by @shannonsands in https://github.com/NousResearch/atropos/pull/50\n* add an SFT data loading env by @dmahan93 in https://github.com/NousResearch/atropos/pull/21\n* changed health check to chat completions since all oai models are comโ€ฆ by @shannonsands in https://github.com/NousResearch/atropos/pull/56\n* version bump to 0.2.1 by @hjc-puro in https://github.com/NousResearch/atropos/pull/57\n\n## New Contributors\n* @JoeLi12345 made their first contribution in https://github.com/NousResearch/atropos/pull/26\n* @shannonsands made their first contribution in https://github.com/NousResearch/atropos/pull/38\n\n**Full Changelog**: https://github.com/NousResearch/atropos/compare/v0.2.0...v0.2.1"},{"ref":"E1","kind":"event","title":"NousResearch/hermes-agent","date":"2025-07-22T22:22:28+00:00","date_source":"source","source_url":"https://github.com/NousResearch/hermes-agent","signal_url":"https://onlylabs.fyi/signals/21de0dcc-1d3a-43d4-98d0-499f5b97220c","signal_json_url":"https://onlylabs.fyi/signals/21de0dcc-1d3a-43d4-98d0-499f5b97220c/signal.json","text":"repo_new ยท NousResearch/hermes-agent ยท signal_desk=repos ยท occurred_at=2025-07-22T22:22:28+00:00 ยท url=https://github.com/NousResearch/hermes-agent ยท stars=190295 ยท raw={\"repo\":\"NousResearch/hermes-agent\",\"description\":\"The agent that grows with you\",\"language\":\"Python\"}"},{"ref":"E2","kind":"event","title":"NousResearch/Hermes-3-Llama-3.1-8B","date":"2024-07-28T06:00:57+00:00","date_source":"source","source_url":"https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-8B","signal_url":"https://onlylabs.fyi/signals/6b977483-7095-470b-9c7f-e081f375ef79","signal_json_url":"https://onlylabs.fyi/signals/6b977483-7095-470b-9c7f-e081f375ef79/signal.json","text":"model_released ยท NousResearch/Hermes-3-Llama-3.1-8B ยท signal_desk=releases ยท occurred_at=2024-07-28T06:00:57+00:00 ยท url=https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-8B ยท hf_downloads=227506 ยท hf_likes=451 ยท hf_params=8030261248 ยท pipeline=text-generation ยท license=llama3 ยท raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E3","kind":"event","title":"NousResearch/DisTrO","date":"2024-08-26T15:30:54+00:00","date_source":"source","source_url":"https://github.com/NousResearch/DisTrO","signal_url":"https://onlylabs.fyi/signals/92ee7c18-3727-47be-b1e5-33bc6ee72008","signal_json_url":"https://onlylabs.fyi/signals/92ee7c18-3727-47be-b1e5-33bc6ee72008/signal.json","text":"repo_new ยท NousResearch/DisTrO ยท signal_desk=repos ยท occurred_at=2024-08-26T15:30:54+00:00 ยท url=https://github.com/NousResearch/DisTrO ยท stars=1034 ยท hn=88 points/26 comments ยท raw={\"repo\":\"NousResearch/DisTrO\",\"description\":\"Distributed Training Over-The-Internet\"}"},{"ref":"E4","kind":"event","title":"NousResearch/DeepHermes-3-Llama-3-8B-Preview","date":"2025-02-12T04:09:45+00:00","date_source":"source","source_url":"https://huggingface.co/NousResearch/DeepHermes-3-Llama-3-8B-Preview","signal_url":"https://onlylabs.fyi/signals/a0cfe58b-ada1-41f9-87d4-23c4fb8216b1","signal_json_url":"https://onlylabs.fyi/signals/a0cfe58b-ada1-41f9-87d4-23c4fb8216b1/signal.json","text":"model_released ยท NousResearch/DeepHermes-3-Llama-3-8B-Preview ยท signal_desk=releases ยท occurred_at=2025-02-12T04:09:45+00:00 ยท url=https://huggingface.co/NousResearch/DeepHermes-3-Llama-3-8B-Preview ยท hf_downloads=1135 ยท hf_likes=362 ยท hf_params=8030261248 ยท pipeline=text-generation ยท license=llama3 ยท raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E5","kind":"event","title":"NousResearch/Hermes-3-Llama-3.1-405B","date":"2024-08-13T04:57:53+00:00","date_source":"source","source_url":"https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-405B","signal_url":"https://onlylabs.fyi/signals/6eb86113-07ff-4c45-bd6e-1d27d70e915d","signal_json_url":"https://onlylabs.fyi/signals/6eb86113-07ff-4c45-bd6e-1d27d70e915d/signal.json","text":"model_released ยท NousResearch/Hermes-3-Llama-3.1-405B ยท signal_desk=releases ยท occurred_at=2024-08-13T04:57:53+00:00 ยท url=https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-405B ยท hf_downloads=275 ยท hf_likes=269 ยท hf_params=405853388800 ยท pipeline=text-generation ยท license=llama3 ยท raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E6","kind":"event","title":"NousResearch/Hermes-4.3-36B","date":"2025-11-17T19:48:42+00:00","date_source":"source","source_url":"https://huggingface.co/NousResearch/Hermes-4.3-36B","signal_url":"https://onlylabs.fyi/signals/fb53e2a4-9276-4485-bac6-11faf29ed343","signal_json_url":"https://onlylabs.fyi/signals/fb53e2a4-9276-4485-bac6-11faf29ed343/signal.json","text":"model_released ยท NousResearch/Hermes-4.3-36B ยท signal_desk=releases ยท occurred_at=2025-11-17T19:48:42+00:00 ยท url=https://huggingface.co/NousResearch/Hermes-4.3-36B ยท hf_downloads=27351 ยท hf_likes=235 ยท hf_params=36151104512 ยท pipeline=text-generation ยท license=apache-2.0 ยท raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E7","kind":"event","title":"NousResearch/NousCoder-14B","date":"2025-12-23T07:44:59+00:00","date_source":"source","source_url":"https://huggingface.co/NousResearch/NousCoder-14B","signal_url":"https://onlylabs.fyi/signals/b211e465-3c5d-4816-96bf-7047473270bf","signal_json_url":"https://onlylabs.fyi/signals/b211e465-3c5d-4816-96bf-7047473270bf/signal.json","text":"model_released ยท NousResearch/NousCoder-14B ยท signal_desk=releases ยท occurred_at=2025-12-23T07:44:59+00:00 ยท url=https://huggingface.co/NousResearch/NousCoder-14B ยท hf_downloads=158 ยท hf_likes=209 ยท hf_params=14768307200 ยท pipeline=text-generation ยท license=apache-2.0 ยท raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E8","kind":"event","title":"NousResearch/Hermes-4-70B","date":"2025-08-18T15:39:17+00:00","date_source":"source","source_url":"https://huggingface.co/NousResearch/Hermes-4-70B","signal_url":"https://onlylabs.fyi/signals/98d1e964-f99b-40c6-91d7-377919ea7a88","signal_json_url":"https://onlylabs.fyi/signals/98d1e964-f99b-40c6-91d7-377919ea7a88/signal.json","text":"model_released ยท NousResearch/Hermes-4-70B ยท signal_desk=releases ยท occurred_at=2025-08-18T15:39:17+00:00 ยท url=https://huggingface.co/NousResearch/Hermes-4-70B ยท hf_downloads=2098 ยท hf_likes=194 ยท hf_params=70553706496 ยท pipeline=text-generation ยท license=llama3 ยท raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E9","kind":"event","title":"NousResearch/Hermes-3-Llama-3.2-3B","date":"2024-12-03T23:47:18+00:00","date_source":"source","source_url":"https://huggingface.co/NousResearch/Hermes-3-Llama-3.2-3B","signal_url":"https://onlylabs.fyi/signals/9143ab26-3a27-4da4-a667-f4acf8975ddf","signal_json_url":"https://onlylabs.fyi/signals/9143ab26-3a27-4da4-a667-f4acf8975ddf/signal.json","text":"model_released ยท NousResearch/Hermes-3-Llama-3.2-3B ยท signal_desk=releases ยท occurred_at=2024-12-03T23:47:18+00:00 ยท url=https://huggingface.co/NousResearch/Hermes-3-Llama-3.2-3B ยท hf_downloads=6150 ยท hf_likes=179 ยท hf_params=3212749824 ยท pipeline=text-generation ยท license=llama3"},{"ref":"E10","kind":"event","title":"NousResearch/Minos-v1","date":"2025-04-24T04:38:23+00:00","date_source":"source","source_url":"https://huggingface.co/NousResearch/Minos-v1","signal_url":"https://onlylabs.fyi/signals/8541abd0-b00e-42a5-919a-15cdab8731a2","signal_json_url":"https://onlylabs.fyi/signals/8541abd0-b00e-42a5-919a-15cdab8731a2/signal.json","text":"model_released ยท NousResearch/Minos-v1 ยท signal_desk=releases ยท occurred_at=2025-04-24T04:38:23+00:00 ยท url=https://huggingface.co/NousResearch/Minos-v1 ยท hf_downloads=1176 ยท hf_likes=174 ยท hf_params=395833346 ยท pipeline=text-classification ยท license=apache-2.0 ยท raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E11","kind":"event","title":"NousResearch/Hermes-4-14B","date":"2025-08-30T20:12:05+00:00","date_source":"source","source_url":"https://huggingface.co/NousResearch/Hermes-4-14B","signal_url":"https://onlylabs.fyi/signals/bfc5d6f1-65fb-47e0-82cc-52f862267946","signal_json_url":"https://onlylabs.fyi/signals/bfc5d6f1-65fb-47e0-82cc-52f862267946/signal.json","text":"model_released ยท NousResearch/Hermes-4-14B ยท signal_desk=releases ยท occurred_at=2025-08-30T20:12:05+00:00 ยท url=https://huggingface.co/NousResearch/Hermes-4-14B ยท hf_downloads=37541 ยท hf_likes=154 ยท hf_params=424960 ยท pipeline=text-generation ยท license=apache-2.0 ยท raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E12","kind":"event","title":"NousResearch/nomos-1","date":"2025-12-07T04:00:59+00:00","date_source":"source","source_url":"https://huggingface.co/NousResearch/nomos-1","signal_url":"https://onlylabs.fyi/signals/bd79802c-c731-43a8-980f-870c11e77e99","signal_json_url":"https://onlylabs.fyi/signals/bd79802c-c731-43a8-980f-870c11e77e99/signal.json","text":"model_released ยท NousResearch/nomos-1 ยท signal_desk=releases ยท occurred_at=2025-12-07T04:00:59+00:00 ยท url=https://huggingface.co/NousResearch/nomos-1 ยท hf_downloads=596 ยท hf_likes=152 ยท hf_params=30532122624 ยท pipeline=text-generation ยท license=apache-2.0 ยท raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E13","kind":"event","title":"NousResearch/Hermes-3-Llama-3.1-70B","date":"2024-07-29T14:27:57+00:00","date_source":"source","source_url":"https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-70B","signal_url":"https://onlylabs.fyi/signals/6c2a3dc2-ea82-4685-a759-5f8f4dad062d","signal_json_url":"https://onlylabs.fyi/signals/6c2a3dc2-ea82-4685-a759-5f8f4dad062d/signal.json","text":"model_released ยท NousResearch/Hermes-3-Llama-3.1-70B ยท signal_desk=releases ยท occurred_at=2024-07-29T14:27:57+00:00 ยท url=https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-70B ยท hf_downloads=117191 ยท hf_likes=126 ยท hf_params=70553706496 ยท pipeline=text-generation ยท license=llama3 ยท raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E14","kind":"event","title":"NousResearch/DeepHermes-3-Mistral-24B-Preview","date":"2025-03-02T13:05:55+00:00","date_source":"source","source_url":"https://huggingface.co/NousResearch/DeepHermes-3-Mistral-24B-Preview","signal_url":"https://onlylabs.fyi/signals/71dc8936-bbca-4ded-af23-6814da2d2fba","signal_json_url":"https://onlylabs.fyi/signals/71dc8936-bbca-4ded-af23-6814da2d2fba/signal.json","text":"model_released ยท NousResearch/DeepHermes-3-Mistral-24B-Preview ยท signal_desk=releases ยท occurred_at=2025-03-02T13:05:55+00:00 ยท url=https://huggingface.co/NousResearch/DeepHermes-3-Mistral-24B-Preview ยท hf_downloads=297 ยท hf_likes=122 ยท hf_params=23572464640 ยท pipeline=text-generation ยท license=apache-2.0 ยท raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E15","kind":"event","title":"NousResearch/Hermes-4-405B","date":"2025-08-06T02:51:17+00:00","date_source":"source","source_url":"https://huggingface.co/NousResearch/Hermes-4-405B","signal_url":"https://onlylabs.fyi/signals/f4d1958c-ccfa-40b4-86b9-63dc435c412a","signal_json_url":"https://onlylabs.fyi/signals/f4d1958c-ccfa-40b4-86b9-63dc435c412a/signal.json","text":"model_released ยท NousResearch/Hermes-4-405B ยท signal_desk=releases ยท occurred_at=2025-08-06T02:51:17+00:00 ยท url=https://huggingface.co/NousResearch/Hermes-4-405B ยท hf_downloads=427 ยท hf_likes=91 ยท hf_params=405853388800 ยท pipeline=text-generation ยท license=llama3 ยท raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E16","kind":"event","title":"NousResearch/hermes-agent v2026.6.5","date":"2026-06-06T00:55:58+00:00","date_source":"source","source_url":"https://github.com/NousResearch/hermes-agent/releases/tag/v2026.6.5","signal_url":"https://onlylabs.fyi/signals/8dd91ac5-b880-4dde-80e9-f4c5cc714155","signal_json_url":"https://onlylabs.fyi/signals/8dd91ac5-b880-4dde-80e9-f4c5cc714155/signal.json","text":"release ยท NousResearch/hermes-agent v2026.6.5 ยท signal_desk=releases ยท occurred_at=2026-06-06T00:55:58+00:00 ยท url=https://github.com/NousResearch/hermes-agent/releases/tag/v2026.6.5 ยท raw={\"repo\":\"NousResearch/hermes-agent\"}"},{"ref":"E17","kind":"event","title":"UI/UX Designer","date":"2026-06-05T22:36:23.684+00:00","date_source":"source","source_url":"https://nousresearch.com/careers","signal_url":"https://onlylabs.fyi/signals/f5dab6fa-a64d-46ad-b92d-afbdce05af18","signal_json_url":"https://onlylabs.fyi/signals/f5dab6fa-a64d-46ad-b92d-afbdce05af18/signal.json","text":"job_opened ยท UI/UX Designer ยท signal_desk=hiring ยท occurred_at=2026-06-05T22:36:23.684+00:00 ยท url=https://nousresearch.com/careers ยท raw={\"ats\":\"agent\"}"},{"ref":"E18","kind":"event","title":"General Counsel","date":"2026-06-05T22:36:23.684+00:00","date_source":"source","source_url":"https://nousresearch.com/careers","signal_url":"https://onlylabs.fyi/signals/b72dedc2-a720-483e-a5dc-38abbb5d3167","signal_json_url":"https://onlylabs.fyi/signals/b72dedc2-a720-483e-a5dc-38abbb5d3167/signal.json","text":"job_opened ยท General Counsel ยท signal_desk=hiring ยท occurred_at=2026-06-05T22:36:23.684+00:00 ยท url=https://nousresearch.com/careers ยท raw={\"ats\":\"agent\"}"},{"ref":"E19","kind":"event","title":"Full Stack Engineer","date":"2026-06-05T22:36:23.684+00:00","date_source":"source","source_url":"https://nousresearch.com/careers","signal_url":"https://onlylabs.fyi/signals/766ad46d-9384-4663-907e-6ea254125748","signal_json_url":"https://onlylabs.fyi/signals/766ad46d-9384-4663-907e-6ea254125748/signal.json","text":"job_opened ยท Full Stack Engineer ยท signal_desk=hiring ยท occurred_at=2026-06-05T22:36:23.684+00:00 ยท url=https://nousresearch.com/careers ยท raw={\"ats\":\"agent\"}"},{"ref":"E20","kind":"event","title":"Forward Deployed Engineer","date":"2026-06-05T22:36:23.684+00:00","date_source":"source","source_url":"https://nousresearch.com/careers","signal_url":"https://onlylabs.fyi/signals/1b6685b6-e38c-4dc6-8b5d-96e8edf34bd1","signal_json_url":"https://onlylabs.fyi/signals/1b6685b6-e38c-4dc6-8b5d-96e8edf34bd1/signal.json","text":"job_opened ยท Forward Deployed Engineer ยท signal_desk=hiring ยท occurred_at=2026-06-05T22:36:23.684+00:00 ยท url=https://nousresearch.com/careers ยท raw={\"ats\":\"agent\"}"},{"ref":"E21","kind":"event","title":"Research Scientist","date":"2026-06-05T22:36:23.684+00:00","date_source":"source","source_url":"https://nousresearch.com/careers","signal_url":"https://onlylabs.fyi/signals/6be31e4e-374a-44d5-8668-68a24fbd5b03","signal_json_url":"https://onlylabs.fyi/signals/6be31e4e-374a-44d5-8668-68a24fbd5b03/signal.json","text":"job_opened ยท Research Scientist ยท signal_desk=hiring ยท occurred_at=2026-06-05T22:36:23.684+00:00 ยท url=https://nousresearch.com/careers ยท raw={\"ats\":\"agent\"}"},{"ref":"E22","kind":"event","title":"Machine Learning Engineer","date":"2026-06-05T22:36:23.684+00:00","date_source":"source","source_url":"https://nousresearch.com/careers","signal_url":"https://onlylabs.fyi/signals/043257c1-e295-4561-a2a6-dbb678dc3614","signal_json_url":"https://onlylabs.fyi/signals/043257c1-e295-4561-a2a6-dbb678dc3614/signal.json","text":"job_opened ยท Machine Learning Engineer ยท signal_desk=hiring ยท occurred_at=2026-06-05T22:36:23.684+00:00 ยท url=https://nousresearch.com/careers ยท raw={\"ats\":\"agent\"}"},{"ref":"E23","kind":"event","title":"NousResearch/DeepHermes-3-Llama-3-3B-Preview","date":"2025-02-16T22:03:39+00:00","date_source":"source","source_url":"https://huggingface.co/NousResearch/DeepHermes-3-Llama-3-3B-Preview","signal_url":"https://onlylabs.fyi/signals/cecc0378-5ce5-496b-b67b-d8b11f83c2a4","signal_json_url":"https://onlylabs.fyi/signals/cecc0378-5ce5-496b-b67b-d8b11f83c2a4/signal.json","text":"model_released ยท NousResearch/DeepHermes-3-Llama-3-3B-Preview ยท signal_desk=releases ยท occurred_at=2025-02-16T22:03:39+00:00 ยท url=https://huggingface.co/NousResearch/DeepHermes-3-Llama-3-3B-Preview ยท hf_downloads=593 ยท hf_likes=39 ยท hf_params=3212749824 ยท pipeline=text-generation ยท license=llama3"},{"ref":"E24","kind":"event","title":"NousResearch/hermes-agent v2026.5.29.2","date":"2026-05-29T13:37:26+00:00","date_source":"source","source_url":"https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.29.2","signal_url":"https://onlylabs.fyi/signals/395c00ec-6573-41ef-a267-e4995dd1053a","signal_json_url":"https://onlylabs.fyi/signals/395c00ec-6573-41ef-a267-e4995dd1053a/signal.json","text":"release ยท NousResearch/hermes-agent v2026.5.29.2 ยท signal_desk=releases ยท occurred_at=2026-05-29T13:37:26+00:00 ยท url=https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.29.2 ยท raw={\"repo\":\"NousResearch/hermes-agent\"}"},{"ref":"E25","kind":"event","title":"NousResearch/hermes-agent v2026.5.29","date":"2026-05-29T01:12:15+00:00","date_source":"source","source_url":"https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.29","signal_url":"https://onlylabs.fyi/signals/e5cf6b58-4ac0-435c-bb76-273c57c7ae86","signal_json_url":"https://onlylabs.fyi/signals/e5cf6b58-4ac0-435c-bb76-273c57c7ae86/signal.json","text":"release ยท NousResearch/hermes-agent v2026.5.29 ยท signal_desk=releases ยท occurred_at=2026-05-29T01:12:15+00:00 ยท url=https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.29 ยท raw={\"repo\":\"NousResearch/hermes-agent\"}"},{"ref":"E26","kind":"event","title":"NousResearch/hermes-agent v2026.5.28","date":"2026-05-28T17:46:09+00:00","date_source":"source","source_url":"https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.28","signal_url":"https://onlylabs.fyi/signals/c00a16bd-6bad-4994-bdf8-8beb3a846048","signal_json_url":"https://onlylabs.fyi/signals/c00a16bd-6bad-4994-bdf8-8beb3a846048/signal.json","text":"release ยท NousResearch/hermes-agent v2026.5.28 ยท signal_desk=releases ยท occurred_at=2026-05-28T17:46:09+00:00 ยท url=https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.28 ยท raw={\"repo\":\"NousResearch/hermes-agent\"}"},{"ref":"E27","kind":"event","title":"NousResearch/Automodel","date":"2026-05-27T12:18:10+00:00","date_source":"source","source_url":"https://github.com/NousResearch/Automodel","signal_url":"https://onlylabs.fyi/signals/007a4fd1-4440-4405-853a-0a96c587ccb2","signal_json_url":"https://onlylabs.fyi/signals/007a4fd1-4440-4405-853a-0a96c587ccb2/signal.json","text":"repo_forked ยท NousResearch/Automodel ยท signal_desk=forks ยท occurred_at=2026-05-27T12:18:10+00:00 ยท url=https://github.com/NousResearch/Automodel ยท stars=10 ยท raw={\"repo\":\"NousResearch/Automodel\",\"parent\":\"NVIDIA-NeMo/Automodel\"}"},{"ref":"E28","kind":"event","title":"NousResearch/Megatron-LM","date":"2026-05-27T12:17:29+00:00","date_source":"source","source_url":"https://github.com/NousResearch/Megatron-LM","signal_url":"https://onlylabs.fyi/signals/cded9ac1-e648-4965-8bda-97827b4128d0","signal_json_url":"https://onlylabs.fyi/signals/cded9ac1-e648-4965-8bda-97827b4128d0/signal.json","text":"repo_forked ยท NousResearch/Megatron-LM ยท signal_desk=forks ยท occurred_at=2026-05-27T12:17:29+00:00 ยท url=https://github.com/NousResearch/Megatron-LM ยท stars=5 ยท raw={\"repo\":\"NousResearch/Megatron-LM\",\"parent\":\"NVIDIA/Megatron-LM\"}"},{"ref":"E29","kind":"event","title":"NousResearch/Megatron-Bridge","date":"2026-05-27T12:17:08+00:00","date_source":"source","source_url":"https://github.com/NousResearch/Megatron-Bridge","signal_url":"https://onlylabs.fyi/signals/a66f6267-7e1f-452a-8efd-640d342f9ce6","signal_json_url":"https://onlylabs.fyi/signals/a66f6267-7e1f-452a-8efd-640d342f9ce6/signal.json","text":"repo_forked ยท NousResearch/Megatron-Bridge ยท signal_desk=forks ยท occurred_at=2026-05-27T12:17:08+00:00 ยท url=https://github.com/NousResearch/Megatron-Bridge ยท stars=5 ยท raw={\"repo\":\"NousResearch/Megatron-Bridge\",\"parent\":\"NVIDIA-NeMo/Megatron-Bridge\"}"},{"ref":"E30","kind":"event","title":"NousResearch/hermes-compression-eval","date":"2026-05-16T09:41:37+00:00","date_source":"source","source_url":"https://github.com/NousResearch/hermes-compression-eval","signal_url":"https://onlylabs.fyi/signals/8c537d36-674f-4a6d-8078-e61250d0fead","signal_json_url":"https://onlylabs.fyi/signals/8c537d36-674f-4a6d-8078-e61250d0fead/signal.json","text":"repo_new ยท NousResearch/hermes-compression-eval ยท signal_desk=repos ยท occurred_at=2026-05-16T09:41:37+00:00 ยท url=https://github.com/NousResearch/hermes-compression-eval ยท stars=9 ยท raw={\"repo\":\"NousResearch/hermes-compression-eval\",\"description\":\"Offline probe-based evaluation harness for hermes-agent's ContextCompressor. Methodology adapted from Factory's Dec 2025 'Evaluating Compression'.\",\"language\":\"Python\"}"},{"ref":"E31","kind":"event","title":"NousResearch/hermes-agent v2026.5.16","date":"2026-05-16T09:59:15+00:00","date_source":"source","source_url":"https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.16","signal_url":"https://onlylabs.fyi/signals/2dc73fd7-5d9d-4cb8-80f4-bd9307d2e534","signal_json_url":"https://onlylabs.fyi/signals/2dc73fd7-5d9d-4cb8-80f4-bd9307d2e534/signal.json","text":"release ยท NousResearch/hermes-agent v2026.5.16 ยท signal_desk=releases ยท occurred_at=2026-05-16T09:59:15+00:00 ยท url=https://github.com/NousResearch/hermes-agent/releases/tag/v2026.5.16 ยท raw={\"repo\":\"NousResearch/hermes-agent\"}"},{"ref":"E32","kind":"event","title":"NousResearch/agent-governance-toolkit","date":"2026-05-10T22:34:18+00:00","date_source":"source","source_url":"https://github.com/NousResearch/agent-governance-toolkit","signal_url":"https://onlylabs.fyi/signals/8f913bfb-f7f5-4ebc-9d01-416bfaecdb12","signal_json_url":"https://onlylabs.fyi/signals/8f913bfb-f7f5-4ebc-9d01-416bfaecdb12/signal.json","text":"repo_forked ยท NousResearch/agent-governance-toolkit ยท signal_desk=forks ยท occurred_at=2026-05-10T22:34:18+00:00 ยท url=https://github.com/NousResearch/agent-governance-toolkit ยท stars=28 ยท raw={\"repo\":\"NousResearch/agent-governance-toolkit\",\"parent\":\"microsoft/agent-governance-toolkit\"}"},{"ref":"E33","kind":"event","title":"NousResearch/Hermes-3-Llama-3.1-405B-FP8","date":"2024-08-13T01:21:28+00:00","date_source":"source","source_url":"https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-405B-FP8","signal_url":"https://onlylabs.fyi/signals/4439220a-39c4-4108-924f-105c2654a2a7","signal_json_url":"https://onlylabs.fyi/signals/4439220a-39c4-4108-924f-105c2654a2a7/signal.json","text":"model_released ยท NousResearch/Hermes-3-Llama-3.1-405B-FP8 ยท signal_desk=releases ยท occurred_at=2024-08-13T01:21:28+00:00 ยท url=https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-405B-FP8 ยท hf_downloads=42 ยท hf_likes=29 ยท hf_params=405873258496 ยท license=llama3 ยท raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E34","kind":"event","title":"NousResearch/hermes-example-plugins","date":"2026-05-10T13:29:46+00:00","date_source":"source","source_url":"https://github.com/NousResearch/hermes-example-plugins","signal_url":"https://onlylabs.fyi/signals/277c6ce4-3bfd-4913-bea0-5bf8be0cfcc9","signal_json_url":"https://onlylabs.fyi/signals/277c6ce4-3bfd-4913-bea0-5bf8be0cfcc9/signal.json","text":"repo_new ยท NousResearch/hermes-example-plugins ยท signal_desk=repos ยท occurred_at=2026-05-10T13:29:46+00:00 ยท url=https://github.com/NousResearch/hermes-example-plugins ยท stars=19 ยท raw={\"repo\":\"NousResearch/hermes-example-plugins\",\"description\":\"Example plugins for hermes-agent โ€” reference implementations and documentation companions, not bundled with the core repo.\",\"language\":\"Python\"}"},{"ref":"E35","kind":"event","title":"NousResearch/Hermes-3-Llama-3.1-70B-FP8","date":"2024-08-06T15:42:03+00:00","date_source":"source","source_url":"https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-70B-FP8","signal_url":"https://onlylabs.fyi/signals/baf76d07-f271-4037-a182-2dbbad3e9625","signal_json_url":"https://onlylabs.fyi/signals/baf76d07-f271-4037-a182-2dbbad3e9625/signal.json","text":"model_released ยท NousResearch/Hermes-3-Llama-3.1-70B-FP8 ยท signal_desk=releases ยท occurred_at=2024-08-06T15:42:03+00:00 ยท url=https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-70B-FP8 ยท hf_downloads=12892 ยท hf_likes=26 ยท hf_params=70560423936 ยท license=llama3 ยท raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E36","kind":"event","title":"NousResearch/kanban-video-pipeline","date":"2026-05-02T17:01:19+00:00","date_source":"source","source_url":"https://github.com/NousResearch/kanban-video-pipeline","signal_url":"https://onlylabs.fyi/signals/f80092ae-2b05-4d8e-820d-acf5003f94b7","signal_json_url":"https://onlylabs.fyi/signals/f80092ae-2b05-4d8e-820d-acf5003f94b7/signal.json","text":"repo_new ยท NousResearch/kanban-video-pipeline ยท signal_desk=repos ยท occurred_at=2026-05-02T17:01:19+00:00 ยท url=https://github.com/NousResearch/kanban-video-pipeline ยท stars=34 ยท raw={\"repo\":\"NousResearch/kanban-video-pipeline\",\"language\":\"Python\"}"},{"ref":"E37","kind":"event","title":"NousResearch/Nemotron","date":"2026-04-27T14:08:47+00:00","date_source":"source","source_url":"https://github.com/NousResearch/Nemotron","signal_url":"https://onlylabs.fyi/signals/b857381d-51f3-446f-9a61-234abaced3d1","signal_json_url":"https://onlylabs.fyi/signals/b857381d-51f3-446f-9a61-234abaced3d1/signal.json","text":"repo_forked ยท NousResearch/Nemotron ยท signal_desk=forks ยท occurred_at=2026-04-27T14:08:47+00:00 ยท url=https://github.com/NousResearch/Nemotron ยท stars=12 ยท raw={\"repo\":\"NousResearch/Nemotron\",\"parent\":\"NVIDIA-NeMo/Nemotron\"}"},{"ref":"E38","kind":"event","title":"NousResearch/NemoClaw","date":"2026-04-24T18:35:00+00:00","date_source":"source","source_url":"https://github.com/NousResearch/NemoClaw","signal_url":"https://onlylabs.fyi/signals/a0d4e67f-3f35-4ea4-ab24-efba1dfd20b7","signal_json_url":"https://onlylabs.fyi/signals/a0d4e67f-3f35-4ea4-ab24-efba1dfd20b7/signal.json","text":"repo_forked ยท NousResearch/NemoClaw ยท signal_desk=forks ยท occurred_at=2026-04-24T18:35:00+00:00 ยท url=https://github.com/NousResearch/NemoClaw ยท stars=23 ยท raw={\"repo\":\"NousResearch/NemoClaw\",\"parent\":\"NVIDIA/NemoClaw\"}"},{"ref":"E39","kind":"event","title":"NousResearch/hermes-agent-self-evolution","date":"2026-03-09T10:42:48+00:00","date_source":"source","source_url":"https://github.com/NousResearch/hermes-agent-self-evolution","signal_url":"https://onlylabs.fyi/signals/4a222e65-4949-4077-bda4-cb86fd507a66","signal_json_url":"https://onlylabs.fyi/signals/4a222e65-4949-4077-bda4-cb86fd507a66/signal.json","text":"repo_new ยท NousResearch/hermes-agent-self-evolution ยท signal_desk=repos ยท occurred_at=2026-03-09T10:42:48+00:00 ยท url=https://github.com/NousResearch/hermes-agent-self-evolution ยท stars=4023 ยท raw={\"repo\":\"NousResearch/hermes-agent-self-evolution\",\"description\":\"โš’ Evolutionary self-improvement for Hermes Agent โ€” optimize skills, prompts, and code using DSPy + GEPA\",\"language\":\"Python\"}"},{"ref":"E40","kind":"event","title":"NousResearch/Llama-3.2-1B","date":"2024-09-27T04:28:43+00:00","date_source":"source","source_url":"https://huggingface.co/NousResearch/Llama-3.2-1B","signal_url":"https://onlylabs.fyi/signals/f371f565-617c-4b3c-b2f4-2bd593a2aa25","signal_json_url":"https://onlylabs.fyi/signals/f371f565-617c-4b3c-b2f4-2bd593a2aa25/signal.json","text":"model_released ยท NousResearch/Llama-3.2-1B ยท signal_desk=releases ยท occurred_at=2024-09-27T04:28:43+00:00 ยท url=https://huggingface.co/NousResearch/Llama-3.2-1B ยท hf_downloads=21908 ยท hf_likes=19 ยท hf_params=1235814400 ยท pipeline=text-generation ยท license=llama3.2"},{"ref":"E41","kind":"event","title":"NousResearch/tinker-nemogym","date":"2026-04-21T15:20:07+00:00","date_source":"source","source_url":"https://github.com/NousResearch/tinker-nemogym","signal_url":"https://onlylabs.fyi/signals/57c59b7f-0edf-4b01-9a60-1d76004341e9","signal_json_url":"https://onlylabs.fyi/signals/57c59b7f-0edf-4b01-9a60-1d76004341e9/signal.json","text":"repo_new ยท NousResearch/tinker-nemogym ยท signal_desk=repos ยท occurred_at=2026-04-21T15:20:07+00:00 ยท url=https://github.com/NousResearch/tinker-nemogym ยท stars=8 ยท raw={\"repo\":\"NousResearch/tinker-nemogym\",\"language\":\"Python\"}"},{"ref":"E42","kind":"event","title":"NousResearch/wterm","date":"2026-04-20T20:12:38+00:00","date_source":"source","source_url":"https://github.com/NousResearch/wterm","signal_url":"https://onlylabs.fyi/signals/56158b96-7f32-427e-a3d1-d2d2f1de3e78","signal_json_url":"https://onlylabs.fyi/signals/56158b96-7f32-427e-a3d1-d2d2f1de3e78/signal.json","text":"repo_forked ยท NousResearch/wterm ยท signal_desk=forks ยท occurred_at=2026-04-20T20:12:38+00:00 ยท url=https://github.com/NousResearch/wterm ยท stars=4 ยท raw={\"repo\":\"NousResearch/wterm\",\"parent\":\"vercel-labs/wterm\"}"},{"ref":"E43","kind":"event","title":"NousResearch/OpenShell","date":"2026-04-20T15:26:48+00:00","date_source":"source","source_url":"https://github.com/NousResearch/OpenShell","signal_url":"https://onlylabs.fyi/signals/0365d232-2a06-44f7-88a3-8a4553d3eda6","signal_json_url":"https://onlylabs.fyi/signals/0365d232-2a06-44f7-88a3-8a4553d3eda6/signal.json","text":"repo_forked ยท NousResearch/OpenShell ยท signal_desk=forks ยท occurred_at=2026-04-20T15:26:48+00:00 ยท url=https://github.com/NousResearch/OpenShell ยท stars=12 ยท raw={\"repo\":\"NousResearch/OpenShell\",\"parent\":\"NVIDIA/OpenShell\"}"},{"ref":"E44","kind":"event","title":"NousResearch/OpenShell-Community","date":"2026-04-20T15:45:42+00:00","date_source":"source","source_url":"https://github.com/NousResearch/OpenShell-Community","signal_url":"https://onlylabs.fyi/signals/f4453673-903a-4284-ae22-31e9123171a4","signal_json_url":"https://onlylabs.fyi/signals/f4453673-903a-4284-ae22-31e9123171a4/signal.json","text":"repo_forked ยท NousResearch/OpenShell-Community ยท signal_desk=forks ยท occurred_at=2026-04-20T15:45:42+00:00 ยท url=https://github.com/NousResearch/OpenShell-Community ยท stars=7 ยท raw={\"repo\":\"NousResearch/OpenShell-Community\",\"parent\":\"NVIDIA/OpenShell-Community\"}"},{"ref":"E45","kind":"event","title":"NousResearch/DeepHermes-ToolCalling-Specialist-Atropos","date":"2025-04-11T07:16:41+00:00","date_source":"source","source_url":"https://huggingface.co/NousResearch/DeepHermes-ToolCalling-Specialist-Atropos","signal_url":"https://onlylabs.fyi/signals/7983f6b9-fb11-4cb1-831f-688cff1027bd","signal_json_url":"https://onlylabs.fyi/signals/7983f6b9-fb11-4cb1-831f-688cff1027bd/signal.json","text":"model_released ยท NousResearch/DeepHermes-ToolCalling-Specialist-Atropos ยท signal_desk=releases ยท occurred_at=2025-04-11T07:16:41+00:00 ยท url=https://huggingface.co/NousResearch/DeepHermes-ToolCalling-Specialist-Atropos ยท hf_downloads=24 ยท hf_likes=18 ยท hf_params=8030261248 ยท pipeline=reinforcement-learning ยท license=llama3 ยท raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E46","kind":"event","title":"NousResearch/DeepHermes-Financial-Fundamentals-Prediction-Specialist-Atropos","date":"2025-04-28T06:50:28+00:00","date_source":"source","source_url":"https://huggingface.co/NousResearch/DeepHermes-Financial-Fundamentals-Prediction-Specialist-Atropos","signal_url":"https://onlylabs.fyi/signals/8ddd55b5-d655-43f9-b484-5614fa48be15","signal_json_url":"https://onlylabs.fyi/signals/8ddd55b5-d655-43f9-b484-5614fa48be15/signal.json","text":"model_released ยท NousResearch/DeepHermes-Financial-Fundamentals-Prediction-Specialist-Atropos ยท signal_desk=releases ยท occurred_at=2025-04-28T06:50:28+00:00 ยท url=https://huggingface.co/NousResearch/DeepHermes-Financial-Fundamentals-Prediction-Specialist-Atropos ยท hf_downloads=49 ยท hf_likes=16 ยท hf_params=8030261248 ยท pipeline=text-generation ยท license=llama3 ยท raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E47","kind":"event","title":"NousResearch/moe-10b-a1b-8k-wsd-lr3e4-1t","date":"2026-03-06T21:15:44+00:00","date_source":"source","source_url":"https://huggingface.co/NousResearch/moe-10b-a1b-8k-wsd-lr3e4-1t","signal_url":"https://onlylabs.fyi/signals/248fb787-293d-48cd-a62a-cb7087c8c95b","signal_json_url":"https://onlylabs.fyi/signals/248fb787-293d-48cd-a62a-cb7087c8c95b/signal.json","text":"model_released ยท NousResearch/moe-10b-a1b-8k-wsd-lr3e4-1t ยท signal_desk=releases ยท occurred_at=2026-03-06T21:15:44+00:00 ยท url=https://huggingface.co/NousResearch/moe-10b-a1b-8k-wsd-lr3e4-1t ยท hf_downloads=28 ยท hf_likes=14 ยท hf_params=10489144832"},{"ref":"E48","kind":"event","title":"NousResearch/ink","date":"2026-04-09T18:10:36+00:00","date_source":"source","source_url":"https://github.com/NousResearch/ink","signal_url":"https://onlylabs.fyi/signals/7bf09d19-627b-4a50-ba2d-65131ed7a5d0","signal_json_url":"https://onlylabs.fyi/signals/7bf09d19-627b-4a50-ba2d-65131ed7a5d0/signal.json","text":"repo_forked ยท NousResearch/ink ยท signal_desk=forks ยท occurred_at=2026-04-09T18:10:36+00:00 ยท url=https://github.com/NousResearch/ink ยท stars=5 ยท raw={\"repo\":\"NousResearch/ink\",\"parent\":\"vadimdemedes/ink\"}"},{"ref":"E49","kind":"event","title":"NousResearch/RL","date":"2026-04-03T14:35:01+00:00","date_source":"source","source_url":"https://github.com/NousResearch/RL","signal_url":"https://onlylabs.fyi/signals/c908cced-c9d6-467a-9da6-08a2a4218c79","signal_json_url":"https://onlylabs.fyi/signals/c908cced-c9d6-467a-9da6-08a2a4218c79/signal.json","text":"repo_forked ยท NousResearch/RL ยท signal_desk=forks ยท occurred_at=2026-04-03T14:35:01+00:00 ยท url=https://github.com/NousResearch/RL ยท stars=11 ยท raw={\"repo\":\"NousResearch/RL\",\"parent\":\"NVIDIA-NeMo/RL\"}"},{"ref":"E50","kind":"event","title":"NousResearch/autoreason","date":"2026-03-28T19:34:36+00:00","date_source":"source","source_url":"https://github.com/NousResearch/autoreason","signal_url":"https://onlylabs.fyi/signals/2c1f03a4-776d-47a8-baed-3b9b868b7ff1","signal_json_url":"https://onlylabs.fyi/signals/2c1f03a4-776d-47a8-baed-3b9b868b7ff1/signal.json","text":"repo_new ยท NousResearch/autoreason ยท signal_desk=repos ยท occurred_at=2026-03-28T19:34:36+00:00 ยท url=https://github.com/NousResearch/autoreason ยท stars=574 ยท raw={\"repo\":\"NousResearch/autoreason\",\"description\":\"Autoresearch for subjective domains.\",\"language\":\"TeX\"}"},{"ref":"E51","kind":"event","title":"NousResearch/DeepHermes-AscensionMaze-RLAIF-8b-Atropos","date":"2025-04-02T05:04:58+00:00","date_source":"source","source_url":"https://huggingface.co/NousResearch/DeepHermes-AscensionMaze-RLAIF-8b-Atropos","signal_url":"https://onlylabs.fyi/signals/881a162a-ea0c-466f-b876-2114359cdc88","signal_json_url":"https://onlylabs.fyi/signals/881a162a-ea0c-466f-b876-2114359cdc88/signal.json","text":"model_released ยท NousResearch/DeepHermes-AscensionMaze-RLAIF-8b-Atropos ยท signal_desk=releases ยท occurred_at=2025-04-02T05:04:58+00:00 ยท url=https://huggingface.co/NousResearch/DeepHermes-AscensionMaze-RLAIF-8b-Atropos ยท hf_downloads=22 ยท hf_likes=9 ยท hf_params=8030261248 ยท pipeline=reinforcement-learning ยท license=llama3 ยท raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E52","kind":"event","title":"NousResearch/Gym","date":"2026-03-31T01:31:07+00:00","date_source":"source","source_url":"https://github.com/NousResearch/Gym","signal_url":"https://onlylabs.fyi/signals/8e07cdc6-d143-4189-b8b0-a15b26691e8c","signal_json_url":"https://onlylabs.fyi/signals/8e07cdc6-d143-4189-b8b0-a15b26691e8c/signal.json","text":"repo_forked ยท NousResearch/Gym ยท signal_desk=forks ยท occurred_at=2026-03-31T01:31:07+00:00 ยท url=https://github.com/NousResearch/Gym ยท stars=17 ยท raw={\"repo\":\"NousResearch/Gym\",\"parent\":\"NVIDIA-NeMo/Gym\"}"},{"ref":"E53","kind":"event","title":"NousResearch/Kimi-K2-Thinking-Alternate-Tokenizer","date":"2026-01-15T20:00:44+00:00","date_source":"source","source_url":"https://huggingface.co/NousResearch/Kimi-K2-Thinking-Alternate-Tokenizer","signal_url":"https://onlylabs.fyi/signals/ec747bba-2966-4fc4-83f7-f0e193f71d4f","signal_json_url":"https://onlylabs.fyi/signals/ec747bba-2966-4fc4-83f7-f0e193f71d4f/signal.json","text":"model_released ยท NousResearch/Kimi-K2-Thinking-Alternate-Tokenizer ยท signal_desk=releases ยท occurred_at=2026-01-15T20:00:44+00:00 ยท url=https://huggingface.co/NousResearch/Kimi-K2-Thinking-Alternate-Tokenizer ยท hf_likes=8"},{"ref":"E54","kind":"event","title":"NousResearch/hermes-paperclip-adapter","date":"2026-03-11T00:16:55+00:00","date_source":"source","source_url":"https://github.com/NousResearch/hermes-paperclip-adapter","signal_url":"https://onlylabs.fyi/signals/c6f62a5c-1e43-488e-8a85-073711ed0854","signal_json_url":"https://onlylabs.fyi/signals/c6f62a5c-1e43-488e-8a85-073711ed0854/signal.json","text":"repo_new ยท NousResearch/hermes-paperclip-adapter ยท signal_desk=repos ยท occurred_at=2026-03-11T00:16:55+00:00 ยท url=https://github.com/NousResearch/hermes-paperclip-adapter ยท stars=1557 ยท raw={\"repo\":\"NousResearch/hermes-paperclip-adapter\",\"description\":\"Paperclip adapter for Hermes Agent โ€” run Hermes as a managed employee in a Paperclip company\",\"language\":\"TypeScript\"}"},{"ref":"E55","kind":"event","title":"Forward Deployed Engineer","date":"unknown","date_source":"missing","source_url":"https://nousresearch.com/careers","signal_url":"https://onlylabs.fyi/signals/96a480d0-dde7-488f-a2e1-15151824ffb0","signal_json_url":"https://onlylabs.fyi/signals/96a480d0-dde7-488f-a2e1-15151824ffb0/signal.json","text":"job_opened ยท Forward Deployed Engineer ยท signal_desk=hiring ยท occurred_at=unknown ยท url=https://nousresearch.com/careers ยท raw={\"location\":\"Remote\",\"ats\":\"agent\"}"},{"ref":"E56","kind":"event","title":"General Counsel","date":"unknown","date_source":"missing","source_url":"https://nousresearch.com/careers","signal_url":"https://onlylabs.fyi/signals/46a54c2f-e4ae-4f22-9061-d72ff2977147","signal_json_url":"https://onlylabs.fyi/signals/46a54c2f-e4ae-4f22-9061-d72ff2977147/signal.json","text":"job_opened ยท General Counsel ยท signal_desk=hiring ยท occurred_at=unknown ยท url=https://nousresearch.com/careers ยท raw={\"location\":\"Remote\",\"ats\":\"agent\"}"},{"ref":"E57","kind":"event","title":"Machine Learning Engineer","date":"unknown","date_source":"missing","source_url":"https://nousresearch.com/careers","signal_url":"https://onlylabs.fyi/signals/00f64ec7-3cd0-46dc-8032-26d01f4c1df4","signal_json_url":"https://onlylabs.fyi/signals/00f64ec7-3cd0-46dc-8032-26d01f4c1df4/signal.json","text":"job_opened ยท Machine Learning Engineer ยท signal_desk=hiring ยท occurred_at=unknown ยท url=https://nousresearch.com/careers ยท raw={\"location\":\"Remote\",\"ats\":\"agent\"}"},{"ref":"E58","kind":"event","title":"Full Stack Engineer","date":"unknown","date_source":"missing","source_url":"https://nousresearch.com/careers","signal_url":"https://onlylabs.fyi/signals/2d58c534-50c1-482f-b421-ecca547e5714","signal_json_url":"https://onlylabs.fyi/signals/2d58c534-50c1-482f-b421-ecca547e5714/signal.json","text":"job_opened ยท Full Stack Engineer ยท signal_desk=hiring ยท occurred_at=unknown ยท url=https://nousresearch.com/careers ยท raw={\"location\":\"Remote\",\"ats\":\"agent\"}"},{"ref":"E59","kind":"event","title":"UI/UX Designer","date":"unknown","date_source":"missing","source_url":"https://nousresearch.com/careers","signal_url":"https://onlylabs.fyi/signals/4a9f248d-af87-4dac-9a13-833fcbf4550c","signal_json_url":"https://onlylabs.fyi/signals/4a9f248d-af87-4dac-9a13-833fcbf4550c/signal.json","text":"job_opened ยท UI/UX Designer ยท signal_desk=hiring ยท occurred_at=unknown ยท url=https://nousresearch.com/careers ยท raw={\"location\":\"Remote\",\"ats\":\"agent\"}"},{"ref":"E60","kind":"event","title":"Research Scientist","date":"unknown","date_source":"missing","source_url":"https://nousresearch.com/careers","signal_url":"https://onlylabs.fyi/signals/1a2f2e03-3e67-43cf-a20b-d516668debea","signal_json_url":"https://onlylabs.fyi/signals/1a2f2e03-3e67-43cf-a20b-d516668debea/signal.json","text":"job_opened ยท Research Scientist ยท signal_desk=hiring ยท occurred_at=unknown ยท url=https://nousresearch.com/careers ยท raw={\"location\":\"Remote\",\"ats\":\"agent\"}"}]}