Granite 3.2 8b instruct
Model Overview¶
Granite-3.2-8B-Instruct is an 8-billion-parameter, long-context AI model fine-tuned for thinking capabilities.
This model is designed to handle general instruction-following tasks and can be integrated into AI assistants across various domains, including business applications.
- Model Architecture: Granite-3.2-8B-Instruct is based on a decoder-only dense transformer architecture. Core components of this architecture are: GQA and RoPE, MLP with SwiGLU, RMSNorm, and shared input/output embeddings.
- Website: Granite Docs
- Model Source: ibm-granite/granite-3.2-8b-instruct
- Release Date: February 26th, 2025
- License: Apache 2.0
- Supported Languages: English, German, Spanish, French, Japanese, Portuguese, Arabic, Czech, Italian, Korean, Dutch, and Chinese.
QPC Configurations¶
| Precision | SoCs / Tensor slicing | NSP-Cores (per SoC) | Full Batch Size | Chunking Prompt Length | Context Length (CL) | Generated URL | Download |
|---|---|---|---|---|---|---|---|
| MXFP6 | 1 | 16 | 1 | 128 | 8192 | https://dc00tk1pxen80.cloudfront.net/SDK1.20.2/ibm-granite/granite-3.2-8b-instruct/qpc_16cores_128pl_8192cl_1fbs_1devices_mxfp6_mxint8.tar.gz | Download |
| MXFP6 | 4 | 16 | 1 | 128 | 32768 | https://dc00tk1pxen80.cloudfront.net/SDK1.20.2/ibm-granite/granite-3.2-8b-instruct/qpc_16cores_128pl_32768cl_1fbs_4devices_mxfp6_mxint8.tar.gz | Download |
| MXFP6 | 4 | 16 | 8 | 128 | 32768 | https://dc00tk1pxen80.cloudfront.net/SDK1.20.2/ibm-granite/granite-3.2-8b-instruct/qpc_16cores_128pl_32768cl_8fbs_4devices_mxfp6_mxint8.tar.gz | Download |
Run This Model¶
# Download QPC
mkdir -p ibm-granite/granite-3.2-8b-instruct
cd ibm-granite/granite-3.2-8b-instruct
wget <Download URL>
tar xzvf <downloaded filename.tar.gz>
# Run QPC
python3 -m QEfficient.cloud.execute --model_name ibm-granite/granite-3.2-8b-instruct --qpc_path <path/to/qpc> --prompt "# shortest path algorithm\n" --generation_len 128
API Endpoint¶
# Start REST endpoint with vLLM
VLLM_QAIC_MAX_CPU_THREADS=8 VLLM_QAIC_QPC_PATH=/path/to/qpc python3 -m vllm.entrypoints.openai.api_server \
--host 0.0.0.0 \
--port 8000 \
--model ibm-granite/granite-3.2-8b-instruct \
--max-model-len <Context Length> \
--max-num-seq <Full Batch Size> \
--max-seq_len-to-capture <Chunking Prompt Length> \
--device qaic \
--block-size 32