Llama 3.1 70B
Model Overview¶
The Meta Llama 3.1 collection of multilingual large language models (LLMs) is a collection of pretrained and instruction tuned generative models in 8B, 70B and 405B sizes (text in/text out). The Llama 3.1 instruction tuned text only models (8B, 70B, 405B) are optimized for multilingual dialogue use cases and outperform many of the available open source and closed chat models on common industry benchmarks.
- Model Architecture: Llama 3.1 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align with human preferences for helpfulness and safety.
- Model Release Date: July 23, 2024.
- Repository: llama-models/models/llama3_1
- Model Source: meta-llama/Meta-Llama-3.1-70B-Instruct
- License: A custom commercial license
QPC Configurations¶
| Precision | SoCs / Tensor slicing | NSP-Cores (per SoC) | Full Batch Size | Chunking Prompt Length | Context Length (CL) | Generated URL | Download |
|---|---|---|---|---|---|---|---|
| MXFP6 | 8 | 16 | 8 | 128 | 8192 | https://qualcom-qpc-models.s3-accelerate.amazonaws.com/SDK1.20.4/meta-llama/Llama-3.1-70B-Instruct/qpc_16cores_128pl_8192cl_8fbs_8devices_mxfp6_mxint8.tar.gz | Download |
| MXFP6 | 4 | 16 | 1 | 128 | 8192 | https://qualcom-qpc-models.s3-accelerate.amazonaws.com/SDK1.20.4/meta-llama/Llama-3.1-70B-Instruct/qpc_16cores_128pl_8192cl_1fbs_4devices_mxfp6_mxint8.tar.gz | Download |
Run This Model¶
# Download QPC
mkdir -p meta-llama/Llama-3.1-70B-Instruct
cd meta-llama/Llama-3.1-70B-Instruct
wget <Download URL>
tar xzvf <downloaded filename.tar.gz>
# Run QPC
python3 -m QEfficient.cloud.execute --model_name meta-llama/Llama-3.1-70B-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 meta-llama/Llama-3.1-70B-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