unsloth/r1-1776-distill-llama-70b-gguf overview

{
  "metadata": {},
  "card_data": {
    "base_model": "perplexity-ai/r1-1776-distill-llama-70b",
    "language": [
      "en"
    ],
    "library_name": "transformers",
    "license": "mit",
    "tags": [
      "deepseek",
      "deepseek_v3",
      "unsloth",
      "transformers"
    ],
    "frontmatter": {
      "base_model": "perplexity-ai/r1-1776-distill-llama-70b",
      "language": [
        "en"
      ],
      "library_name": "transformers",
      "license": "mit",
      "tags": [
        "deepseek",
        "deepseek_v3",
        "unsloth",
        "transformers"
      ]
    },
    "hero_image_url": "https://github.com/unslothai/unsloth/raw/main/images/unsloth%20new%20logo.png",
    "summary": "Blog link: https://perplexity.ai/hub/blog/open-sourcing-r1-1776 This is a Llama 70B distilled version of R1 1776. R1 1776 is a DeepSeek-R1 reasoning model that has been post-trained by Perplexity AI to remove Chinese Communist Party censorship. The model provides unbiased, accurate, and factual information while maintaining high reasoning capabilities.",
    "quick_links": [],
    "benchmark_table_html": "",
    "readme_markdown": "---\nbase_model: perplexity-ai/r1-1776-distill-llama-70b\nlanguage:\n- en\nlibrary_name: transformers\nlicense: mit\ntags:\n- deepseek\n- deepseek_v3\n- unsloth\n- transformers\n---\n<div>\n  <p style=\"margin-bottom: 0; margin-top: 0;\">\n    <strong>See <a href=\"https://huggingface.co/collections/unsloth/deepseek-r1-all-versions-678e1c48f5d2fce87892ace5\">our collection</a> for versions of Deepseek-R1 including GGUF & 4-bit formats.</strong>\n  </p>\n  <p style=\"margin-bottom: 0;\">\n    <em>Unsloth's r1-1776 <a href=\"https://unsloth.ai/blog/deepseekr1-dynamic\">2-bit Dynamic Quants</a> is selectively quantized, greatly improving accuracy over standard 1-bit/2-bit.</em>\n  </p>\n  <div style=\"display: flex; gap: 5px; align-items: center; \">\n    <a href=\"https://github.com/unslothai/unsloth/\">\n      <img src=\"https://github.com/unslothai/unsloth/raw/main/images/unsloth%20new%20logo.png\" width=\"133\">\n    </a>\n    <a href=\"https://discord.gg/unsloth\">\n      <img src=\"https://github.com/unslothai/unsloth/raw/main/images/Discord%20button.png\" width=\"173\">\n    </a>\n    <a href=\"https://docs.unsloth.ai/basics/tutorial-how-to-run-deepseek-r1-on-your-own-local-device\">\n      <img src=\"https://raw.githubusercontent.com/unslothai/unsloth/refs/heads/main/images/documentation%20green%20button.png\" width=\"143\">\n    </a>\n  </div>\n<h1 style=\"margin-top: 0rem;\">Finetune your own Reasoning model like R1 with Unsloth!</h2>\n</div>\n\nWe have a free Google Colab notebook for turning Llama 3.1 (8B) into a reasoning model: https://colab.research.google.com/github/unslothai/notebooks/blob/main/nb/Llama3.1_(8B)-GRPO.ipynb\n\n## ✨ Finetune for Free\n\nAll notebooks are **beginner friendly**! Add your dataset, click \"Run All\", and you'll get a 2x faster finetuned model which can be exported to GGUF, vLLM or uploaded to Hugging Face.\n\n| Unsloth supports          |    Free Notebooks                                                                                           | Performance | Memory use |\n|-----------------|--------------------------------------------------------------------------------------------------------------------------|-------------|----------|\n| **GRPO with Phi-4 (14B)**      | [▶️ Start on Colab](https://colab.research.google.com/github/unslothai/notebooks/blob/main/nb/Phi_4_(14B)-GRPO.ipynb)               | 2x faster | 80% less |\n| **Llama-3.2 (3B)**      | [▶️ Start on Colab](https://colab.research.google.com/github/unslothai/notebooks/blob/main/nb/Llama3.2_(1B_and_3B)-Conversational.ipynb)               | 2.4x faster | 58% less |\n| **Llama-3.2 (11B vision)**      | [▶️ Start on Colab](https://colab.research.google.com/github/unslothai/notebooks/blob/main/nb/Llama3.2_(11B)-Vision.ipynb)               | 2x faster | 60% less |\n| **Qwen2 VL (7B)**      | [▶️ Start on Colab](https://colab.research.google.com/github/unslothai/notebooks/blob/main/nb/Qwen2_VL_(7B)-Vision.ipynb)               | 1.8x faster | 60% less |\n| **Qwen2.5 (7B)**      | [▶️ Start on Colab](https://colab.research.google.com/github/unslothai/notebooks/blob/main/nb/Qwen2.5_(7B)-Alpaca.ipynb)               | 2x faster | 60% less |\n| **Llama-3.1 (8B)**      | [▶️ Start on Colab](https://colab.research.google.com/github/unslothai/notebooks/blob/main/nb/Llama3.1_(8B)-Alpaca.ipynb)               | 2.4x faster | 58% less |\n| **Phi-3.5 (mini)** | [▶️ Start on Colab](https://colab.research.google.com/github/unslothai/notebooks/blob/main/nb/Phi_3.5_Mini-Conversational.ipynb)               | 2x faster | 50% less |\n| **Gemma 2 (9B)**      | [▶️ Start on Colab](https://colab.research.google.com/github/unslothai/notebooks/blob/main/nb/Gemma2_(9B)-Alpaca.ipynb)               | 2.4x faster | 58% less |\n| **Mistral (7B)**    | [▶️ Start on Colab](https://colab.research.google.com/github/unslothai/notebooks/blob/main/nb/Mistral_v0.3_(7B)-Conversational.ipynb)               | 2.2x faster | 62% less |\n\n[<img src=\"https://raw.githubusercontent.com/unslothai/unsloth/refs/heads/main/images/documentation%20green%20button.png\" width=\"200\"/>](https://docs.unsloth.ai)\n\n- This [Llama 3.2 conversational notebook](https://colab.research.google.com/github/unslothai/notebooks/blob/main/nb/Llama3.2_(1B_and_3B)-Conversational.ipynb) is useful for ShareGPT ChatML / Vicuna templates.\n- This [text completion notebook](https://colab.research.google.com/github/unslothai/notebooks/blob/main/nb/Mistral_(7B)-Text_Completion.ipynb) is for raw text. This [DPO notebook](https://colab.research.google.com/drive/15vttTpzzVXv_tJwEk-hIcQ0S9FcEWvwP?usp=sharing) replicates Zephyr.\n- \\* Kaggle has 2x T4s, but we use 1. Due to overhead, 1x T4 is 5x faster.\n\n# R1 1776 Distill Llama 70B\n\nBlog link: [https://perplexity.ai/hub/blog/open-sourcing-r1-1776](https://perplexity.ai/hub/blog/open-sourcing-r1-1776 ) \n\nThis is a Llama 70B distilled version of [R1 1776](https://huggingface.co/perplexity-ai/r1-1776).\n\nR1 1776 is a DeepSeek-R1 reasoning model that has been post-trained by Perplexity AI to remove Chinese Communist Party censorship. \nThe model provides unbiased, accurate, and factual information while maintaining high reasoning capabilities.\n\n## Evals\n\nTo ensure our model remains fully “uncensored” and capable of engaging with a broad spectrum of sensitive topics, \nwe curated a diverse, multilingual evaluation set of over a 1000 of examples that comprehensively cover such subjects. \nWe then use human annotators as well as carefully designed LLM judges to measure the likelihood a model will evade or \nprovide overly sanitized responses to the queries.\n\nWe also ensured that the model’s math and reasoning abilities remained intact after the decensoring process. \nEvaluations on multiple benchmarks showed that our post-trained model performed on par with the base R1 model, \nindicating that the decensoring had no impact on its core reasoning capabilities.\n\n| Benchmark | R1-Distill-Llama-70B | R1-1776-Distill-Llama-70B |\n| --- | --- | --- |\n| China Censorship |  80.53 | 0.2 |\n| Internal Benchmarks (avg) | 47.64 |  48.4 |\n| AIME 2024 | 70 | 70 |\n| MATH-500 | 94.5 | 94.8 |\n| MMLU | 88.52 * | 88.40 |\n| DROP | 84.55 * | 84.83 |\n| GPQA | 65.2 | 65.05 |\n\n\\* Evaluated by Perplexity AI since they were not reported in the [paper](https://arxiv.org/abs/2501.12948).",
    "related_quantizations": []
  },
  "tags": [
    "transformers",
    "gguf",
    "llama",
    "text-generation",
    "deepseek",
    "deepseek_v3",
    "unsloth",
    "en",
    "arxiv:2501.12948",
    "base_model:perplexity-ai/r1-1776-distill-llama-70b",
    "base_model:quantized:perplexity-ai/r1-1776-distill-llama-70b",
    "license:mit",
    "endpoints_compatible",
    "region:us",
    "conversational"
  ],
  "likes": 24,
  "downloads": 251,
  "gated": false,
  "private": false,
  "last_modified": "2025-02-27T01:55:59.000Z",
  "created_at": "2025-02-22T09:43:21.000Z",
  "pipeline_tag": "text-generation",
  "library_name": "transformers"
}

{
  "_id": "67b99c39d04d2b1541cc26b6",
  "id": "unsloth/r1-1776-distill-llama-70b-GGUF",
  "modelId": "unsloth/r1-1776-distill-llama-70b-GGUF",
  "sha": "e4485249875bfd948d15b9c2eb5920de54cb2b80",
  "createdAt": "2025-02-22T09:43:21.000Z",
  "lastModified": "2025-02-27T01:55:59.000Z",
  "author": "unsloth",
  "downloads": 251,
  "likes": 24,
  "gated": false,
  "private": false,
  "pipeline_tag": "text-generation",
  "library_name": "transformers",
  "siblings_count": 15
}