Gemma 1 7B
Parameters
7B
Context Length
8K
Modality
Text
Architecture
Dense
KV Cache Rate
336 KB/tok
License
Gemma Terms of Use
Release Date
21 Feb 2024
Knowledge Cutoff
-
API Pricing (per 1M)
Input: $0.05 · Output: $0.08
Hardware Sizing
Precision
Context
Workload Memory: 8.38 GB
INT4 · 8K context
Base Weights: 3.26 GB
KV Cache: 2.63 GB
Runtime & Graphs: 2.50 GB
Minimum Required Hardware
1x NVIDIA L4 24GB
8.4 / 21.6 GB usable
1x RTX 3060 12GB
8.4 / 10.8 GB usable
Apple M4 24GB
8.4 / 18.0 GB usable
Determine requirements for your specific workload:
Open in VRAM Calculator
Architecture Diagram
Evaluation Benchmarks
Rank
#238
| Benchmark | Score | Rank |
|---|---|---|
General Text | 1138 | 192 |
Rankings
Overall Rank
#238
Coding Rank
-
About Gemma 1 7B
Gemma 1 7B is Google's foundational open-weights decoder-only transformer engineered for general text generation and reasoning. Trained on 6T tokens, it offers an efficient balance of capability and accessible deployment on local workstations.
Technical Specifications
Attention
Attention Structure
Multi-Head Attention
Attention Heads
32
Key-Value Heads
32
Position Embedding
ROPE
Normalization
RMS Normalization
Dimensions
Hidden Dimension Size
3,072
Number of Layers
28
About Gemma 1
Gemma 1 is a family of lightweight, decoder-only transformer models from Google, available in 2B and 7B parameter sizes. Designed for various text generation tasks, they incorporate rotary positional embeddings, shared input/output embeddings, GEGLU activation, and RMSNorm. The 2B model uses multi-query attention, while 7B uses multi-head attention.