granite-3.0-1b-a400m-instruct/README.md

pipeline_tag, inference, license, library_name, tags, model-index, base_model, new_version

pipeline_tag	inference	license	library_name	tags	model-index	base_model	new_version
text-generation	false	apache-2.0	transformers	language granite-3.0	name results granite-3.0-2b-instruct task dataset metrics type text-generation type name instruction-following IFEval name type value veriefied pass@1 pass@1 32.39 false task dataset metrics type text-generation type name instruction-following MT-Bench name type value veriefied pass@1 pass@1 6.17 false task dataset metrics type text-generation type name human-exams AGI-Eval name type value veriefied pass@1 pass@1 20.35 false task dataset metrics type text-generation type name human-exams MMLU name type value veriefied pass@1 pass@1 32 false task dataset metrics type text-generation type name human-exams MMLU-Pro name type value veriefied pass@1 pass@1 12.21 false task dataset metrics type text-generation type name commonsense OBQA name type value veriefied pass@1 pass@1 38.4 false task dataset metrics type text-generation type name commonsense SIQA name type value veriefied pass@1 pass@1 47.55 false task dataset metrics type text-generation type name commonsense Hellaswag name type value veriefied pass@1 pass@1 65.59 false task dataset metrics type text-generation type name commonsense WinoGrande name type value veriefied pass@1 pass@1 61.17 false task dataset metrics type text-generation type name commonsense TruthfulQA name type value veriefied pass@1 pass@1 49.11 false task dataset metrics type text-generation type name reading-comprehension BoolQ name type value veriefied pass@1 pass@1 70.12 false task dataset metrics type text-generation type name reading-comprehension SQuAD 2.0 name type value veriefied pass@1 pass@1 1.27 false task dataset metrics type text-generation type name reasoning ARC-C name type value veriefied pass@1 pass@1 41.21 false task dataset metrics type text-generation type name reasoning GPQA name type value veriefied pass@1 pass@1 23.07 false task dataset metrics type text-generation type name reasoning BBH name type value veriefied pass@1 pass@1 31.77 false task dataset metrics type text-generation type name code HumanEvalSynthesis name type value veriefied pass@1 pass@1 30.18 false task dataset metrics type text-generation type name code HumanEvalExplain name type value veriefied pass@1 pass@1 26.22 false task dataset metrics type text-generation type name code HumanEvalFix name type value veriefied pass@1 pass@1 21.95 false task dataset metrics type text-generation type name code MBPP name type value veriefied pass@1 pass@1 15.4 false task dataset metrics type text-generation type name math GSM8K name type value veriefied pass@1 pass@1 26.31 false task dataset metrics type text-generation type name math MATH name type value veriefied pass@1 pass@1 10.88 false task dataset metrics type text-generation type name multilingual PAWS-X (7 langs) name type value veriefied pass@1 pass@1 45.84 false task dataset metrics type text-generation type name multilingual MGSM (6 langs) name type value veriefied pass@1 pass@1 11.8 false	ibm-granite/granite-3.0-1b-a400m-base	ibm-granite/granite-3.1-1b-a400m-instruct

Granite-3.0-1B-A400M-Instruct

Model Summary: Granite-3.0-1B-A400M-Instruct is an 1B parameter model finetuned from Granite-3.0-1B-A400M-Base using a combination of open source instruction datasets with permissive license and internally collected synthetic datasets. This model is developed using a diverse set of techniques with a structured chat format, including supervised finetuning, model alignment using reinforcement learning, and model merging.

• Developers: Granite Team, IBM
• GitHub Repository: ibm-granite/granite-3.0-language-models
• Website: Granite Docs
• Paper: Granite 3.0 Language Models
• Release Date: October 21st, 2024
• License: Apache 2.0

Supported Languages: English, German, Spanish, French, Japanese, Portuguese, Arabic, Czech, Italian, Korean, Dutch, and Chinese. Users may finetune Granite 3.0 models for languages beyond these 12 languages.

Intended use: The model is designed to respond to general instructions and can be used to build AI assistants for multiple domains, including business applications.

Capabilities

• Summarization
• Text classification
• Text extraction
• Question-answering
• Retrieval Augmented Generation (RAG)
• Code related tasks
• Function-calling tasks
• Multilingual dialog use cases

Generation: This is a simple example of how to use Granite-3.0-1B-A400M-Instruct model.

Install the following libraries:

pip install torch torchvision torchaudio
pip install accelerate
pip install transformers

Then, copy the snippet from the section that is relevant for your use case.

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

device = "auto"
model_path = "ibm-granite/granite-3.0-1b-a400m-instruct"
tokenizer = AutoTokenizer.from_pretrained(model_path)
# drop device_map if running on CPU
model = AutoModelForCausalLM.from_pretrained(model_path, device_map=device)
model.eval()
# change input text as desired
chat = [
    { "role": "user", "content": "Please list one IBM Research laboratory located in the United States. You should only output its name and location." },
]
chat = tokenizer.apply_chat_template(chat, tokenize=False, add_generation_prompt=True)
# tokenize the text
input_tokens = tokenizer(chat, return_tensors="pt").to(device)
# generate output tokens
output = model.generate(**input_tokens, 
                        max_new_tokens=100)
# decode output tokens into text
output = tokenizer.batch_decode(output)
# print output
print(output)

Model Architecture: Granite-3.0-1B-A400M-Instruct is based on a decoder-only sparse Mixture of Experts (MoE) transformer architecture. Core components of this architecture are: Fine-grained Experts, Dropless Token Routing, and Load Balancing Loss.

Model	2B Dense	8B Dense	1B MoE	3B MoE
Embedding size	2048	4096	1024	1536
Number of layers	40	40	24	32
Attention head size	64	128	64	64
Number of attention heads	32	32	16	24
Number of KV heads	8	8	8	8
MLP hidden size	8192	12800	512	512
MLP activation	SwiGLU	SwiGLU	SwiGLU	SwiGLU
Number of Experts	—	—	32	40
MoE TopK	—	—	8	8
Initialization std	0.1	0.1	0.1	0.1
Sequence Length	4096	4096	4096	4096
Position Embedding	RoPE	RoPE	RoPE	RoPE
# Parameters	2.5B	8.1B	1.3B	3.3B
# Active Parameters	2.5B	8.1B	400M	800M
# Training tokens	12T	12T	10T	10T

Training Data: Overall, our SFT data is largely comprised of three key sources: (1) publicly available datasets with permissive license, (2) internal synthetic data targeting specific capabilities, and (3) very small amounts of human-curated data. A detailed attribution of datasets can be found in the Granite Technical Report and Accompanying Author List.

Infrastructure: We train Granite 3.0 Language Models using IBM's super computing cluster, Blue Vela, which is outfitted with NVIDIA H100 GPUs. This cluster provides a scalable and efficient infrastructure for training our models over thousands of GPUs while minimizing environmental impact by utilizing 100% renewable energy sources.

Ethical Considerations and Limitations: Granite 3.0 Instruct Models are primarily finetuned using instruction-response pairs mostly in English, but also multilingual data covering eleven languages. Although this model can handle multilingual dialog use cases, its performance might not be similar to English tasks. In such case, introducing a small number of examples (few-shot) can help the model in generating more accurate outputs. While this model has been aligned by keeping safety in consideration, the model may in some cases produce inaccurate, biased, or unsafe responses to user prompts. So we urge the community to use this model with proper safety testing and tuning tailored for their specific tasks.

Resources

• ⭐️ Learn about the latest updates with Granite: https://www.ibm.com/granite
• 📄 Get started with tutorials, best practices, and prompt engineering advice: https://www.ibm.com/granite/docs/
• 💡 Learn about the latest Granite learning resources: https://ibm.biz/granite-learning-resources