Qwen2-72B
Parameters
72B
Context Length
33K
Modality
Text
Architecture
Dense
KV Cache Rate
160 KB/tok
License
Tongyi Qianwen LICENSE AGREEMENT
Release Date
7 Jun 2024
Knowledge Cutoff
-
API Pricing (per 1M)
Input: $0.90 · Output: $0.90
Hardware & System Requirements
Precision
Context
Workload Memory: 37.28 GB
INT4 · 8K context
Base Weights: 33.53 GB
KV Cache: 1.25 GB
Runtime & Graphs: 2.50 GB
Minimum Required Hardware
1x NVIDIA A40 48GB
37.3 / 43.2 GB usable
2x RTX 4090 24GB
37.3 / 43.2 GB usable · TP 2
Apple M4 Max 128GB
37.3 / 96.0 GB usable
Determine requirements for your specific workload:
Open in VRAM Calculator
Runtime requirements
Checkpoint format
FP16
CUDA
>= 11.8
PyTorch
>= 2.1
Transformers
>= 4.40.1
| Architecture | Precision | Attention kernel |
|---|---|---|
Blackwell, Hopper | Native FP16 | FlashAttention-3 |
Ada Lovelace, Ampere | Native FP16 | FlashAttention-2 |
Architecture Diagram
Evaluation Benchmarks
Rank
#188
| Benchmark | Score | Rank |
|---|---|---|
Professional Knowledge | 0.644 | 52 |
Graduate-Level QA | 0.424 | 113 |
General Text | 1262 | 167 |
Intelligence Index | 0.06 | 294 |
General Knowledge Reference | 0.823 | 17 |
Rankings
Overall Rank
#188
Coding Rank
-
About Qwen2-72B
Qwen2-72B is Alibaba Cloud's flagship dense foundation model from the Qwen2 generation, optimized for complex reasoning and enterprise-scale NLP. It delivers state-of-the-art coding, mathematics, and multilingual performance.
Specifications
Attention
Attention Structure
Grouped-Query Attention
Attention Heads
128
Key-Value Heads
8
Position Embedding
ROPE
RoPE Theta
1,000,000
Normalization
RMS Normalization
Activation Function
SwigLU
Dimensions
Base Precision
FP16
Hidden Dimension Size
8,192
Number of Layers
80
FFN Intermediate Size (Dense)
29,568
Tokenizer
Vocabulary Size
152,064
About Qwen2
The Alibaba Qwen2 model family comprises large language models built upon the Transformer architecture. It includes both dense and Mixture-of-Experts (MoE) variants, designed for diverse language tasks. Technical features include Grouped Query Attention and support for extended context lengths up to 131,072 tokens, optimizing memory footprint for inference.