All bookmarks in the AI Models categoryhttps://michaelbommarito.comen-usCopyright 2026 Michael Bommaritomichael.bommarito@gmail.com (Michael Bommarito)michael.bommarito@gmail.com (Michael Bommarito)1440Astrohttp://www.rssboard.org/rss-specificationhttps://tokenbender.com/post.html?id=avatarlai-models/avatarl-rl-zero-pretrainExperimental project by TokenBender exploring whether reinforcement learning can be successfully implemented from zero pretraining - an interesting research direction in AI/ML | Notes: TokenBender (self-described 'ml hermit') is exploring whether RL from zero pretrain is possible. Also maintains other projects including LLaMA fine-tunes and tools for local LLM hosting. GitHub repo has 49 stars showing community interest. | Tags: reinforcement-learning, rl, experimental, pretrain, ai-research, open-source, github, llm, local-models | Author: TokenBender | Source: tokenbender.comSat, 09 Aug 2025 00:00:00 GMTAI Modelsreinforcement-learningrlexperimentalpretrainai-researchopen-sourcegithubllmlocal-modelsTokenBenderhttps://huggingface.co/Intel/gpt-oss-20b-int4-rtn-AutoRoundgpt-oss/gpt-oss-20b-intel-int4Intel's INT4 quantized version of GPT-OSS 20B using AutoRound technique with symmetric quantization, optimized for efficient inference on CPU/Intel GPU/CUDA hardware. | Notes: 1.8B parameter quantized model with group size 128, using RTN (no algorithm tuning). Mixed int4 model with non-expert layers at 16-bit precision for better performance. | Tags: gpt-oss, llm, quantization, INT4, Intel, AutoRound, model-compression, inference-optimization | Source: HuggingFaceSat, 09 Aug 2025 00:00:00 GMTAI Modelsgpt-ossllmquantizationINT4IntelAutoRoundmodel-compressioninference-optimizationhttps://huggingface.co/openai/gpt-oss-120bgpt-oss/gpt-oss-120b-huggingface117B parameter model with extreme sparsity for production use | Notes: Fits on single H100 GPU with MXFP4 quantization | Tags: gpt-oss, llm, apache-2.0 | Source: HuggingFaceTue, 05 Aug 2025 00:00:00 GMTAI Modelsgpt-ossllmapache-2.0https://huggingface.co/openai/gpt-oss-20bgpt-oss/gpt-oss-20b-huggingface21B parameter model optimized for local deployment | Notes: Runs in ~16GB VRAM, ideal for consumer hardware | Tags: gpt-oss, llm, apache-2.0, local | Source: HuggingFaceTue, 05 Aug 2025 00:00:00 GMTAI Modelsgpt-ossllmapache-2.0localhttps://openai.com/index/introducing-gpt-oss/gpt-oss/gpt-oss-announcementDetailed announcement of GPT-OSS models with benchmarks and examples | Tags: gpt-oss, blog, announcement | Author: OpenAI | Source: OpenAITue, 05 Aug 2025 00:00:00 GMTAI Modelsgpt-ossblogannouncementOpenAIhttps://github.com/openai/gpt-ossgpt-oss/gpt-oss-githubReference PyTorch implementation and tools for GPT-OSS models | Tags: gpt-oss, pytorch, implementation | Author: OpenAI | Source: GitHubTue, 05 Aug 2025 00:00:00 GMTAI Modelsgpt-osspytorchimplementationOpenAIhttps://github.com/openai/harmonygpt-oss/gpt-oss-harmonyStructured response format specification for reasoning and tool use | Tags: harmony, gpt-oss, response-format | Author: OpenAI | Source: GitHubTue, 05 Aug 2025 00:00:00 GMTAI Modelsharmonygpt-ossresponse-formatOpenAIhttps://cdn.openai.com/pdf/419b6906-9da6-406c-a19d-1bb078ac7637/oai_gpt-oss_model_card.pdfgpt-oss/gpt-oss-model-cardTechnical specifications and training details for GPT-OSS models | Tags: gpt-oss, model-card, specifications | Source: OpenAITue, 05 Aug 2025 00:00:00 GMTAI Modelsgpt-ossmodel-cardspecificationshttps://cookbook.openai.com/articles/gpt-oss/run-locally-ollamagpt-oss/gpt-oss-ollama-guideGuide for running GPT-OSS models locally using Ollama | Tags: gpt-oss, ollama, local-deployment | Source: OpenAI CookbookTue, 05 Aug 2025 00:00:00 GMTAI Modelsgpt-ossollamalocal-deploymenthttps://openai.com/open-models/gpt-oss/gpt-oss-open-modelsOfficial announcement page for GPT-OSS open-weight reasoning models | Tags: gpt-oss, openai, announcement | Author: OpenAI | Source: OpenAITue, 05 Aug 2025 00:00:00 GMTAI Modelsgpt-ossopenaiannouncementOpenAIhttps://gpt-oss.com/gpt-oss/gpt-oss-playgroundInteractive demo to try GPT-OSS models directly in browser | Tags: gpt-oss, demo, playground | Source: OpenAITue, 05 Aug 2025 00:00:00 GMTAI Modelsgpt-ossdemoplaygroundhttps://www.kaggle.com/competitions/openai-gpt-oss-20b-red-teaming/gpt-oss/gpt-oss-red-teaming$500k Kaggle competition for community safety testing | Tags: gpt-oss, red-teaming, kaggle, competition | Source: KaggleTue, 05 Aug 2025 00:00:00 GMTAI Modelsgpt-ossred-teamingkagglecompetitionhttps://cdn.openai.com/pdf/231bf018-659a-494d-976c-2efdfc72b652/oai_gpt-oss_Model_Safety.pdfgpt-oss/gpt-oss-safety-paperComprehensive safety analysis including malicious fine-tuning testing | Tags: gpt-oss, safety, research | Source: OpenAITue, 05 Aug 2025 00:00:00 GMTAI Modelsgpt-osssafetyresearchhttps://cookbook.openai.com/articles/gpt-oss/run-transformersgpt-oss/gpt-oss-transformers-guideDevelopment setup guide using HuggingFace Transformers library | Tags: gpt-oss, transformers, huggingface | Source: OpenAI CookbookTue, 05 Aug 2025 00:00:00 GMTAI Modelsgpt-osstransformershuggingfacehttps://cookbook.openai.com/articles/gpt-oss/run-vllmgpt-oss/gpt-oss-vllm-guideProduction deployment guide using vLLM for high-performance inference | Notes: Best for production deployments, supports CUDA 12.1+ | Tags: gpt-oss, vllm, production | Source: OpenAI CookbookTue, 05 Aug 2025 00:00:00 GMTAI Modelsgpt-ossvllmproduction