richarderkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf overview

{
  "metadata": {},
  "card_data": {
    "frontmatter": {},
    "hero_image_url": "",
    "summary": "",
    "quick_links": [],
    "benchmark_table_html": "",
    "readme_markdown": "Quantization made by Richard Erkhov.\n\n[Github](https://github.com/RichardErkhov)\n\n[Discord](https://discord.gg/pvy7H8DZMG)\n\n[Request more models](https://github.com/RichardErkhov/quant_request)\n\n\ngemma-2-2b-it-research-in-a-flash - GGUF\n- Model creator: https://huggingface.co/dwhouse/\n- Original model: https://huggingface.co/dwhouse/gemma-2-2b-it-research-in-a-flash/\n\n\n| Name | Quant method | Size |\n| ---- | ---- | ---- |\n| [gemma-2-2b-it-research-in-a-flash.Q2_K.gguf](https://huggingface.co/RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf/blob/main/gemma-2-2b-it-research-in-a-flash.Q2_K.gguf) | Q2_K | 1.15GB |\n| [gemma-2-2b-it-research-in-a-flash.IQ3_XS.gguf](https://huggingface.co/RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf/blob/main/gemma-2-2b-it-research-in-a-flash.IQ3_XS.gguf) | IQ3_XS | 1.22GB |\n| [gemma-2-2b-it-research-in-a-flash.IQ3_S.gguf](https://huggingface.co/RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf/blob/main/gemma-2-2b-it-research-in-a-flash.IQ3_S.gguf) | IQ3_S | 1.27GB |\n| [gemma-2-2b-it-research-in-a-flash.Q3_K_S.gguf](https://huggingface.co/RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf/blob/main/gemma-2-2b-it-research-in-a-flash.Q3_K_S.gguf) | Q3_K_S | 1.27GB |\n| [gemma-2-2b-it-research-in-a-flash.IQ3_M.gguf](https://huggingface.co/RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf/blob/main/gemma-2-2b-it-research-in-a-flash.IQ3_M.gguf) | IQ3_M | 1.3GB |\n| [gemma-2-2b-it-research-in-a-flash.Q3_K.gguf](https://huggingface.co/RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf/blob/main/gemma-2-2b-it-research-in-a-flash.Q3_K.gguf) | Q3_K | 1.36GB |\n| [gemma-2-2b-it-research-in-a-flash.Q3_K_M.gguf](https://huggingface.co/RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf/blob/main/gemma-2-2b-it-research-in-a-flash.Q3_K_M.gguf) | Q3_K_M | 1.36GB |\n| [gemma-2-2b-it-research-in-a-flash.Q3_K_L.gguf](https://huggingface.co/RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf/blob/main/gemma-2-2b-it-research-in-a-flash.Q3_K_L.gguf) | Q3_K_L | 1.44GB |\n| [gemma-2-2b-it-research-in-a-flash.IQ4_XS.gguf](https://huggingface.co/RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf/blob/main/gemma-2-2b-it-research-in-a-flash.IQ4_XS.gguf) | IQ4_XS | 1.47GB |\n| [gemma-2-2b-it-research-in-a-flash.Q4_0.gguf](https://huggingface.co/RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf/blob/main/gemma-2-2b-it-research-in-a-flash.Q4_0.gguf) | Q4_0 | 1.52GB |\n| [gemma-2-2b-it-research-in-a-flash.IQ4_NL.gguf](https://huggingface.co/RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf/blob/main/gemma-2-2b-it-research-in-a-flash.IQ4_NL.gguf) | IQ4_NL | 1.53GB |\n| [gemma-2-2b-it-research-in-a-flash.Q4_K_S.gguf](https://huggingface.co/RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf/blob/main/gemma-2-2b-it-research-in-a-flash.Q4_K_S.gguf) | Q4_K_S | 1.53GB |\n| [gemma-2-2b-it-research-in-a-flash.Q4_K.gguf](https://huggingface.co/RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf/blob/main/gemma-2-2b-it-research-in-a-flash.Q4_K.gguf) | Q4_K | 1.59GB |\n| [gemma-2-2b-it-research-in-a-flash.Q4_K_M.gguf](https://huggingface.co/RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf/blob/main/gemma-2-2b-it-research-in-a-flash.Q4_K_M.gguf) | Q4_K_M | 1.59GB |\n| [gemma-2-2b-it-research-in-a-flash.Q4_1.gguf](https://huggingface.co/RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf/blob/main/gemma-2-2b-it-research-in-a-flash.Q4_1.gguf) | Q4_1 | 1.64GB |\n| [gemma-2-2b-it-research-in-a-flash.Q5_0.gguf](https://huggingface.co/RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf/blob/main/gemma-2-2b-it-research-in-a-flash.Q5_0.gguf) | Q5_0 | 1.75GB |\n| [gemma-2-2b-it-research-in-a-flash.Q5_K_S.gguf](https://huggingface.co/RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf/blob/main/gemma-2-2b-it-research-in-a-flash.Q5_K_S.gguf) | Q5_K_S | 1.75GB |\n| [gemma-2-2b-it-research-in-a-flash.Q5_K.gguf](https://huggingface.co/RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf/blob/main/gemma-2-2b-it-research-in-a-flash.Q5_K.gguf) | Q5_K | 1.79GB |\n| [gemma-2-2b-it-research-in-a-flash.Q5_K_M.gguf](https://huggingface.co/RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf/blob/main/gemma-2-2b-it-research-in-a-flash.Q5_K_M.gguf) | Q5_K_M | 1.79GB |\n| [gemma-2-2b-it-research-in-a-flash.Q5_1.gguf](https://huggingface.co/RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf/blob/main/gemma-2-2b-it-research-in-a-flash.Q5_1.gguf) | Q5_1 | 1.87GB |\n| [gemma-2-2b-it-research-in-a-flash.Q6_K.gguf](https://huggingface.co/RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf/blob/main/gemma-2-2b-it-research-in-a-flash.Q6_K.gguf) | Q6_K | 2.0GB |\n| [gemma-2-2b-it-research-in-a-flash.Q8_0.gguf](https://huggingface.co/RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf/blob/main/gemma-2-2b-it-research-in-a-flash.Q8_0.gguf) | Q8_0 | 2.59GB |\n\n\n\n\nOriginal model description:\n---\nlibrary_name: transformers\nlicense: apache-2.0\ndatasets:\n- abisee/cnn_dailymail\nlanguage:\n- ko\n- en\nmetrics:\n- rouge\nbase_model:\n- google/gemma-2-2b-it\npipeline_tag: summarization\n---\n\n# Model Card for gemma-2-2b-it-research-in-a-flash\n\n- Fine-tune the Gemma2 2b model for summarizing scientific papers.\n- Filter the dataset for computer science papers to optimize training time.\n- Deploy the model on Hugging Face for easy accessibility.\n\n\n## Model Details\n\n### Model Description\n\nhis model is a fine-tuned version of `google/gemma-2-2b-it` on the `cnn_dailymail` dataset, designed for the task of **summarization**. \nIt can summarize paragraphs of text, especially from research papers or news articles, into concise summaries. \nThe model has been fine-tuned using the LoRA (Low-Rank Adaptation) method for parameter-efficient training.\n\nThis is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.\n\n- **Developed by:** Changjip Moon\n- **Model type:** Summarization\n- **Language(s) (NLP):** Korean, English\n- **License:** Apache 2.0\n- **Finetuned from model [optional]:** google/gemma-2-2b-it\n\n### Model Sources [optional]\n\n- **Demo:** https://colab.research.google.com/drive/1xiyWCnTzXmFFgD7CBL-jq8m2Mv29fg-M?usp=sharing\n\n## Uses\n\n### Direct Use\nThis model can be used to generate concise summaries of long texts. It is designed for summarizing academic papers, research materials, or news articles.\n\n### Downstream Use\nThis model can be fine-tuned further for other languages or summarization-specific tasks like topic-based summarization.\n\n### Out-of-Scope Use\nThis model is not designed for tasks outside of text summarization, such as text classification or question answering. It also may not perform well on non-news or non-research data.\n\n## Bias, Risks, and Limitations\n\nThis model may have biases inherited from the `cnn_dailymail` dataset, which is mainly based on news articles in English. It may not perform well on non-news content or in cases where high precision is required for legal, medical, or sensitive content.\n\n\n## Training Details\n\n### Training Data\nThe model was fine-tuned on the `cnn_dailymail` dataset, which contains articles and summaries from CNN and Daily Mail. The dataset is commonly used for text summarization tasks.\n\n### Training Procedure\nThe model was trained using the following hyperparameters:\n\n- **Learning rate:** 2e-4\n- **Batch size:** 8 (with gradient accumulation steps of 4)\n- **Epochs:** 1\n- **Max sequence length:** 256\n- **Optimization method:** AdamW with 8-bit quantization\n\n#### Preprocessing\nStandard tokenization and truncation were applied. The maximum sequence length was set to 256 to balance memory usage and training speed.\n\n\n\n#### Training Hyperparameters\n\n- **Training regime:** go to google colab pages if you want to know\n\n#### Speeds, Sizes, Times\n\n[2500/2500 22:33, Epoch 1/1] : Cause of timeout issue, I need to make a subset of data.. \n\n",
    "related_quantizations": []
  },
  "tags": [
    "gguf",
    "endpoints_compatible",
    "region:us",
    "conversational"
  ],
  "likes": 0,
  "downloads": 107,
  "gated": false,
  "private": false,
  "last_modified": "2025-02-26T10:40:22.000Z",
  "created_at": "2025-02-26T08:01:34.000Z",
  "pipeline_tag": "",
  "library_name": ""
}

{
  "_id": "67beca5e199448325fa64723",
  "id": "RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf",
  "modelId": "RichardErkhov/dwhouse_-_gemma-2-2b-it-research-in-a-flash-gguf",
  "sha": "209756e0e188acbd3cf16253b2dabcd15d3e85a0",
  "createdAt": "2025-02-26T08:01:34.000Z",
  "lastModified": "2025-02-26T10:40:22.000Z",
  "author": "RichardErkhov",
  "downloads": 107,
  "likes": 0,
  "gated": false,
  "private": false,
  "pipeline_tag": "",
  "library_name": "",
  "siblings_count": 24
}