AI21 Jamba 1.5 Mini
Version: 1
Jamba 1.5 Mini is a state-of-the-art, hybrid SSM-Transformer instruction following foundation model. It's a Mixture-of-Expert model with 52B total parameters and 12B active parameters. The Jamba family of models are the most powerful & efficient long-context models on the market, offering a 256K context window, the longest available.. For long context input, they deliver up to 2.5X faster inference than leading models of comparable sizes. Jamba supports function calling/tool use, structured output (JSON), and grounded generation with citation mode and documents API. Jamba officially supports English, French, Spanish, Portuguese, German, Arabic and Hebrew, but can also work in many other languages.
Model Developer Name: AI21 Labs
Model Architecture
Jamba 1.5 Mini is a state-of-the-art, hybrid SSM-Transformer instruction following foundation modelModel Variations
52B total parameters and 12B active parametersModel Input
Model inputs text only.Model Output
Model generates text only.Model Dates
Jamba 1.5 Mini was trained in Q3 2024 with data covering through early March 2024.Model Information Table
| Name | Params | Content Length |
|---|---|---|
| Jamba 1.5 Mini | 52B (12B active) | 256K |
| Jamba 1.5 Large | 398B (94B active) | 256K |
License
Please refer to this link for AI21's Terms of Use For more information about how Azure processes data, for privacy and security in relation to Models-as-a-Service (MaaS), please see this Microsoft Docs link .Ethical Considerations and Limitations
Ethical considerations
AI21 Labs is on a mission to supercharge human productivity with machines working alongside humans as thought partners, thereby promoting human welfare and prosperity. To deliver its promise, this technology must be deployed and used in a responsible and sustainable way, taking into consideration potential risks, including malicious use by bad actors, accidental misuse and broader societal harms. We take these risks extremely seriously and put measures in place to mitigate them. AI21 provides open access to Jamba that can be used to power a large variety of useful applications. We believe it is important to ensure that this technology is used in a responsible way, while allowing developers the freedom they need to experiment rapidly and deploy solutions at scale. Overall, we view the safe implementation of this technology as a partnership and collaboration between AI21 and our customers and encourage engagement and dialogue to raise the bar on responsible usage. In order to use Jamba, you are required to comply with our Terms of Use and with the following usage guidelines. Provided you comply with these requirements, you may use Jamba to power applications with live users without any additional approval. We reserve the right to limit or suspend your access to Jamba at any time where we believe these terms or guidelines are violated. Please check these usage guidelines periodically, as they may be updated from time to time. For any questions, clarifications or concerns, please contact safety@ai21.com .Limitations
There are a number of limitations inherent to neural networks technology that apply to Jamba. These limitations require explanation and carry important caveats for the application and usage of Jamba. Accuracy: Jamba, like other large pretrained language models, lacks important context about the world because it is trained on textual data and is not grounded in other modalities of experience such as video, real-world physical interaction, and human feedback. Like all language models, Jamba is far more accurate when responding to inputs similar to its training datasets. Novel inputs have a tendency to generate higher variance in its output. Coherence and consistency: Responses from Jamba are sometimes inconsistent, contradictory, or contain seemingly random sentences and paragraphs. Western/English bias: Jamba is trained primarily on English language text from the internet, and is best suited to classifying, searching, summarizing, and generating English text. Furthermore, Jamba has a tendency to hold and amplify the biases contained in its training dataset. As a result, groups of people who were not involved in the creation of the training data can be underrepresented, and stereotypes and prejudices can be perpetuated. Racial, religious, gender, socioeconomic, and other categorizations of human groups can be considered among these factors. Explainability: It is difficult to explain or predict how Jamba will respond without additional training and fine tuning. This is a common issue with neural networks of this scope and scale. Recency: Jamba was trained on a dataset created in March 2024, and therefore has no knowledge of events that have occurred after that date. We update our models regularly to keep them as current as possible, but there are notable gaps and inaccuracies in responses as a result of this lack of recency.Safety and Responsible Use
AI21 Labs is on a mission to make AI-first experiences, with machines working alongside humans as thought partners, thereby promoting human welfare and prosperity. To deliver its promise, this technology must be deployed and used in a responsible and sustainable way, taking into consideration potential risks, including malicious use by bad actors, accidental misuse and broader societal harms. We take these risks extremely seriously and put measures in place to mitigate them. AI safety is an important challenge with a large surface area, which we believe can be addressed most effectively by working together. We invite anyone interested in conducting research or otherwise promoting AI safety to contact us at safety@ai21.com and explore opportunities for collaboration.Content Filtering
Prompts and completions in GitHub Models are passed through a default configuration of Azure AI Content Safety classification models to detect and prevent the output of harmful content. Learn more about Azure AI Content Safety . Configuration options for content filtering vary when you deploy a model for production in Azure AI; learn more .Training Data
Jamba is trained on an in-house dataset that contains text data from the web, books, and code. The knowledge cutoff date is March 5, 2024.Evaluation Results
| Category | Metric | Score |
|---|---|---|
| General | Arena Hard | 46.1 |
| MMLU | 69.7 | |
| MMLU Pro (CoT) | 42.5 | |
| IFEval | 75.8 | |
| BBH | 53.4 | |
| WildBench | 42.4 | |
| Reasoning | ARC-C | 85.7 |
| GPQA | 32.3 | |
| Math, Code & tool use | GSM8K | 75.8 |
| HumanEval | 62.8 | |
| BFCL | 80.6 |
Evaluation of pretrained LLMs on automatic safety benchmarks
| TruthfulQA | |
|---|---|
| Jamba 1.5 Mini | 54.1 |
| Jamba 1.5 Large | 58.3 |
Evaluation of fine-tuned LLMs on different safety datasets
| RealToxicity* | |
|---|---|
| Jamba 1.5 Mini | 8.1 |
| Jamba 1.5 Large | 6.7 |
Model Specifications
Context Length262144
LicenseCustom
Last UpdatedOctober 2024
Input TypeText
Output TypeText
PublisherAI21 Labs
Languages7 Languages