## Gemma Model Card

### Model Information

#### Description

Gemma is a family of lightweight, state-of-the art open models from Google,
built from the same research and technology used to create the Gemini models.
Gemma models are text-to-text, decoder-only large language models available in English,
with open weights, pre-trained variants, and instruction-tuned variants.

Gemma models are well-suited for a variety of text generation tasks, including
question answering, summarization, and reasoning. Their relatively small size
makes it possible to deploy them in environments with limited resources such as
a laptop, desktop or your own cloud infrastructure, democratizing access to
state of the art AI models and helping foster innovation for everyone.

### References:

**Author**: Google

**Model Page**: [Gemma](https://ai.google.dev/gemma/docs)

**Model Card**: https://ai.google.dev/gemma/docs/model_card

**Resources and Technical Documentation**:

* [Responsible Generative AI Toolkit][rai-toolkit]
* [Gemma on Kaggle][kaggle-gemma]
* [Gemma on Vertex Model Garden][vertex-mg-gemma]

**Terms of Use**: [Terms][terms]

### Third-Party Community Consideration

This model is not owned or developed by NVIDIA. This model has been developed and built to a third-party’s requirements for this application and use case.

#### Inputs and outputs

* **Input Type:** Text string (such as a question, a prompt, or a document to be
  summarized)
* **Output Type:** Text string (Generated text in English language in response to the input, such
  as an answer to a question, or a summary of a document)

### Model Data

#### Training Dataset

These models were trained on a text dataset that includes a wide variety
of sources, totaling 6 trillion tokens. Here are the primary training data sources:

* Web Documents: A diverse collection of web text ensures the model is exposed
  to a broad range of linguistic styles, topics, and vocabulary. Primarily
  English-language content.
* Code: Exposing the model to code helps it to learn the syntax and patterns of
  programming languages, which improves its ability to generate code or
  understand code-related questions.
* Mathematics: Training on mathematical text helps the model learn logical
  reasoning, symbolic representation, and to address mathematical queries.

The combination of these diverse data sources is crucial for training a powerful
language model that can handle a wide variety of different tasks and text
formats.

#### Data Preprocessing

Here are the key data cleaning and filtering methods applied to the training
data:

* CSAM Filtering: Rigorous CSAM (Child Sexual Abuse Material) filtering was
  applied at multiple stages in the data preparation process to ensure the
  exclusion of harmful and illegal content.
* Sensitive Data Filtering: As part of making Gemma pre-trained models safe and
  reliable, automated techniques were used to filter out certain personal
  information and other sensitive data from training sets.
* Additional methods: Filtering based on content quality and safely in line with
  [our policies][safety-policies].

### Implementation Information

#### TensorRT-LLM

The endpoint available on NGC catalog is accelerated by TensorRT-LLM, an open-source library for optimizing inference performance. Gemma is compatible across NVIDIA AI platforms—from the datacenter, cloud, to the local PC with RTX GPU systems.  

Gemma models use a vocabulary size of 256K and support a context length of up to 8K while using rotary positional embedding (RoPE). With support for Position Interpolation (PI) available in TensorRT-LLM, Gemma models using RoPE can support longer output sequence lengths at inference time while retaining original model architecture. 

#### Model Customization

The model is converted to .nemo for easy customization with NVIDIA NeMo framework – an end-to-end framework to curate data, tune models, and deploy anywhere. It supports various customization techniques including RLHF, SFT, LoRA, and Steer-LM.

### Evaluation

Model evaluation metrics and results.

#### Benchmark Results

These models were evaluated against a large collection of different datasets and
metrics to cover different aspects of text generation:

| Benchmark                      | Metric        | 2B Params   | 7B Params |
| ------------------------------ | ------------- | ----------- | --------- |
| [MMLU][mmlu]                   | 5-shot, top-1 | 42.3        | 64.3      |
| [HellaSwag][hellaswag]         | 0-shot        | 71.4        | 81.2      |
| [PIQA][piqa]                   | 0-shot        | 77.3        | 81.2      |
| [SocialIQA][socialiqa]         | 0-shot        | 59.7        | 51.8      |
| [BooIQ][booiq]                 | 0-shot        | 69.4        | 83.2      |
| [WinoGrande][winogrande]       | partial score | 65.4        | 72.3      |
| [CommonsenseQA][commonsenseqa] | 7-shot        | 65.3        | 71.3      |
| [OpenBookQA][openbookqa]       |               | 47.8        | 52.8      |
| [ARC-e][arc]                   |               | 73.2        | 81.5      |
| [ARC-c][arc]                   |               | 42.1        | 53.2      |
| [TriviaQA][triviaqa]           | 5-shot        | 53.2        | 63.4      |
| [Natural Questions][naturalq]  | 5-shot        |             | 23        |
| [HumanEval][humaneval]         | pass@1        | 22.0        | 32.3      |
| [MBPP][mbpp]                   | 3-shot        | 29.2        | 44.4      |
| [GSM8K][gsm8k]                 | maj@1         | 17.7        | 46.4      |
| [MATH][math]                   | 4-shot        | 11.8        | 24.3      |
| [AGIEval][agieval]             |               | 24.2        | 41.7      |
| [BIG-Bench][big-bench]         |               | 35.2        | 55.1      |
| ------------------------------ | ------------- | ----------- | --------- |
| **Average**                    |               | **54.0**    | **56.4**  |

### Ethics and Safety

#### Evaluation Approach

Our evaluation methods include structured evaluations and internal red-teaming
testing of relevant content policies. Red-teaming was conducted by a number of
different teams, each with different goals and human evaluation metrics. These
models were evaluated against a number of different categories relevant to
ethics and safety, including:

* Text-to-Text Content Safety: Human evaluation on prompts covering safety
  policies including child sexual abuse and exploitation, harassment, violence
  and gore, and hate speech.
* Text-to-Text Representational Harms: Benchmark against relevant academic
  datasets such as [WinoBias][winobias] and [BBQ Dataset][bbq].
* Memorization: Automated evaluation of memorization of training data, including
  the risk of personally identifiable information exposure.
* Large-scale harm: Tests for "dangerous capabilities," such as chemical,
  biological, radiological, and nuclear (CBRN) risks.

#### Evaluation Results

The results of ethics and safety evaluations are within acceptable thresholds
for meeting [internal policies][safety-policies] for categories such as child
safety, content safety, representational harms, memorization, large-scale harms.
On top of robust internal evaluations, the results of well known safety
benchmarks like BBQ, BOLD, Winogender, Winobias, RealToxicity, and TruthfulQA
are shown here.

| Benchmark                      | Metric        | 2B Params   | 7B Params |
| ------------------------------ | ------------- | ----------- | --------- |
| [RealToxicity][realtox]        | average       | 6.86        | 7.90      |
| [BOLD][bold]                   |               | 45.57       | 49.08     |
| [CrowS-Pairs][crows]           | top-1         | 45.82       | 51.33     |
| [BBQ Ambig][bbq]               | 1-shot, top-1 | 62.58       | 92.54     |
| [BBQ Disambig][bbq]            | top-1         | 54.62       | 71.99     |
| [Winogender][winogender]       | top-1         | 51.25       | 54.17     |
| [TruthfulQA][truthfulqa]       |               | 44.84       | 31.81     |
| [Winobias 1_2][winobias]       |               | 56.12       | 59.09     |
| [Winobias 2_2][winobias]       |               | 91.10       | 92.23     |
| [Toxigen][toxigen]             |               | 29.77       | 39.59     |
| ------------------------------ | ------------- | ----------- | --------- |

### Usage and Limitations

These models have certain limitations that users should be aware of.

#### Intended Usage

Open Large Language Models (LLMs) have a wide range of applications across
various industries and domains. The following list of potential uses is not
comprehensive. The purpose of this list is to provide contextual information
about the possible use-cases that the model creators considered as part of model
training and development.

* Content Creation and Communication
  * Text Generation: These models can be used to generate creative text formats
    such as poems, scripts, code, marketing copy, and email drafts.
  * Chatbots and Conversational AI: Power conversational interfaces for customer
    service, virtual assistants, or interactive applications.
  * Text Summarization: Generate concise summaries of a text corpus, research
    papers, or reports.
* Research and Education
  * Natural Language Processing (NLP) Research: These models can serve as a
    foundation for researchers to experiment with NLP techniques, develop
    algorithms, and contribute to the advancement of the field.
  * Language Learning Tools: Support interactive language learning experiences,
    aiding in grammar correction or providing writing practice.
  * Knowledge Exploration: Assist researchers in exploring large bodies of text
    by generating summaries or answering questions about specific topics.

#### Limitations

* Training Data
  * The quality and diversity of the training data significantly influence the
    model's capabilities. Biases or gaps in the training data can lead to
    limitations in the model's responses.
  * The scope of the training dataset determines the subject areas the model can
    handle effectively.
* Context and Task Complexity
  * LLMs are better at tasks that can be framed with clear prompts and
    instructions. Open-ended or highly complex tasks might be challenging.
  * A model's performance can be influenced by the amount of context provided
    (longer context generally leads to better outputs, up to a certain point).
* Language Ambiguity and Nuance
  * Natural language is inherently complex. LLMs might struggle to grasp subtle
    nuances, sarcasm, or figurative language.
* Factual Accuracy
  * LLMs generate responses based on information they learned from their
    training datasets, but they are not knowledge bases. They may generate
    incorrect or outdated factual statements.
* Common Sense
  * LLMs rely on statistical patterns in language. They might lack the ability
    to apply common sense reasoning in certain situations.

#### Ethical Considerations and Risks

The development of large language models (LLMs) raises several ethical concerns.
In creating an open model, we have carefully considered the following:

* Bias and Fairness
  * LLMs trained on large-scale, real-world text data can reflect socio-cultural
    biases embedded in the training material. These models underwent careful
    scrutiny, input data pre-processing described and posterior evaluations
    reported in this card.
* Misinformation and Misuse
  * LLMs can be misused to generate text that is false, misleading, or harmful.
  * Guidelines are provided for responsible use with the model, see the
    [Responsible Generative AI Toolkit][rai-toolkit].
* Transparency and Accountability:
  * This model card summarizes details on the models' architecture,
    capabilities, limitations, and evaluation processes.
  * A responsibly developed open model offers the opportunity to share
    innovation by making LLM technology accessible to developers and researchers
    across the AI ecosystem.

Risks identified and mitigations:

* Perpetuation of biases: It's encouraged to perform continuous monitoring
  (using evaluation metrics, human review) and the exploration of de-biasing
  techniques during model training, fine-tuning, and other use cases.
* Generation of harmful content: Mechanisms and guidelines for content safety
  are essential. Developers are encouraged to exercise caution and implement
  appropriate content safety safeguards based on their specific product policies
  and application use cases.
* Misuse for malicious purposes: Technical limitations and developer and
  end-user education can help mitigate against malicious applications of LLMs.
  Educational resources and reporting mechanisms for users to flag misuse are
  provided. Prohibited uses of Gemma models are outlined in the
  [Gemma Prohibited Use Policy][prohibited-use].
* Privacy violations: Models were trained on data filtered for removal of PII
  (Personally Identifiable Information). Developers are encouraged to adhere to
  privacy regulations with privacy-preserving techniques.

[Google's commitments to operate sustainably][sustainability].

#### Benefits

At the time of release, this family of models provides high-performance open
large language model implementations designed from the ground up for Responsible
AI development compared to similarly sized models.

Using the benchmark evaluation metrics described in this document, these models
have shown to provide superior performance to other, comparably-sized open model
alternatives.

[rai-toolkit]: https://ai.google.dev/responsible
[kaggle-gemma]: https://www.kaggle.com/models/google/gemma
[terms]: https://ai.google.dev/gemma/terms
[vertex-mg-gemma]: https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/335
[sensitive-info]: https://cloud.google.com/dlp/docs/high-sensitivity-infotypes-reference
[safety-policies]: https://storage.googleapis.com/gweb-uniblog-publish-prod/documents/2023_Google_AI_Principles_Progress_Update.pdf#page=11
[prohibited-use]: https://ai.google.dev/gemma/prohibited_use_policy
[tpu]: https://cloud.google.com/tpu/docs/intro-to-tpu
[sustainability]: https://sustainability.google/operating-sustainably/
[jax]: https://github.com/google/jax
[ml-pathways]: https://blog.google/technology/ai/introducing-pathways-next-generation-ai-architecture/
[sustainability]: https://sustainability.google/operating-sustainably/
[foundation-models]: https://ai.google/discover/foundation-models/
[gemini-paper]: https://arxiv.org/abs/2312.11805
[mmlu]: https://arxiv.org/abs/2009.03300
[hellaswag]: https://arxiv.org/abs/1905.07830
[piqa]: https://arxiv.org/abs/1911.11641
[socialiqa]: https://arxiv.org/abs/1904.09728
[booiq]: https://arxiv.org/abs/1905.10044
[winogrande]: https://arxiv.org/abs/1907.10641
[commonsenseqa]: https://arxiv.org/abs/1811.00937
[openbookqa]: https://arxiv.org/abs/1809.02789
[arc]: https://arxiv.org/abs/1911.01547
[triviaqa]: https://arxiv.org/abs/1705.03551
[naturalq]: https://github.com/google-research-datasets/natural-questions
[humaneval]: https://arxiv.org/abs/2107.03374
[mbpp]: https://arxiv.org/abs/2108.07732
[gsm8k]: https://arxiv.org/abs/2110.14168
[realtox]: https://arxiv.org/abs/2009.11462
[bold]: https://arxiv.org/abs/2101.11718
[crows]: https://aclanthology.org/2020.emnlp-main.154/
[bbq]: https://arxiv.org/abs/2110.08193v2
[winogender]: https://arxiv.org/abs/1804.09301
[truthfulqa]: https://arxiv.org/abs/2109.07958
[winobias]: https://arxiv.org/abs/1804.06876
[math]: https://arxiv.org/abs/2103.03874
[agieval]: https://arxiv.org/abs/2304.06364
[big-bench]: https://arxiv.org/abs/2206.04615
[toxigen]: https://arxiv.org/abs/2203.09509

gemma-2b

Gemma is a family of lightweight, state-of-the art LLM open models from Google,

The list of completion choices the model generated for the input prompt.

Choices

Usage statistics for the completion request.

ChatCompletion

ChatCompletionChunk

A word or list of words not to use. The words are case sensitive.

The maximum number of tokens to generate in any given call. Note that the model is not aware of this value, and generation will simply stop at the number of tokens specified.

Max Tokens

A list of messages comprising the conversation so far. The roles of the messages must be alternating between `user` and `assistant`. The last input message should have role `user`. A message with the the `system` role is optional, and must be the very first message if it is present.

Messages

If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

Seed

A string or a list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.

Stop

If set, partial message deltas will be sent. Tokens will be sent as data-only server-sent events (SSE) as they become available (JSON responses are prefixed by `data: `), with the stream terminated by a `data: [DONE]` message.

Stream

The sampling temperature to use for text generation. The higher the temperature value is, the less deterministic the output text will be. It is not recommended to modify both temperature and top_p in the same call.

Temperature

The top-p sampling mass used for text generation. The top-p value determines the probability mass that is sampled at sampling time. For example, if top_p = 0.2, only the most likely tokens (summing to 0.2 cumulative probability) will be sampled. It is not recommended to modify both temperature and top_p in the same call.

Top P

ChatRequest

The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached.

Finish Reason

The index of the choice in the list of choices (always 0).

Index

A chat completion message generated by the model.

Choice

A chat completion delta generated by streamed model responses.

The reason the model stopped generating tokens. This will be `stop` if the model hit a natural stop point or a provided stop sequence, or `length` if the maximum number of tokens specified in the request was reached. Will be `null` if the model has not finished generating.

ChoiceChunk

Function instance used to invoke the request

InvokeError

Content

Role

Message

Number of tokens in the generated completion.

Completion Tokens

Prompt Tokens

Total number of tokens used in the request (prompt + completion).

Total Tokens

Usage

NVCF function

Result is pending. Client should poll using the requestId.


Result is in a different region. Client should use the fully-qualified
endpoint specified in 'Location' response header to fetch the result.
Client can use the same Bearer token in 'Authorization' header when
retrieving the result from the redirected region.


Validation failed, provided entity could not be processed.

requestId

Gets the result of an earlier function invocation request that returned a status of 202.

Gemma 2B