← All Models
|
architecture:
Trinity-Nano-Base
Trinity-Mini
Trinity-Large-TrueBase
arcee-ai/Trinity-Mini
📊 Model Parameters
Total Parameters
26,123,974,400
Context Length
131,072
Hidden Size
2048
Layers
32
Attention Heads
32
KV Heads
4
💾 Memory Requirements
FP32 (Full)
97.32 GB
FP16 (Half)
48.66 GB
INT8 (Quantized)
24.33 GB
INT4 (Quantized)
12.16 GB
🔑 KV Cache (Inference)
Per Token (FP16)
65.54 KB
Max Context FP32
16.00 GB
Max Context FP16
8.00 GB
Max Context INT8
4.00 GB
⚙️ Model Configuration
Core Architecture
Vocabulary Size
200,192
Hidden Size
2,048
FFN Intermediate Size
6,144
Number of Layers
32
Attention Heads
32
Head Dimension
128
KV Heads
4
Context & Position
Max Context Length
131,072
RoPE Base Frequency
10,000
Sliding Window Size
2,048
Layer Attention Types
[32 items]
Attention Configuration
Attention Bias
No
Attention Dropout
0%
Tied Embeddings
No
Mixture of Experts
Expert FFN Size
1,024
Experts per Token
8
Number of Experts
128
Expert Groups
1
Groups per Token
1
Activation & Normalization
Activation Function
silu
RMSNorm Epsilon
1e-05
Special Tokens
EOS Token ID
Not set
Pad Token ID
Not set
BOS Token ID
Not set
Data Type
Model Dtype
bfloat16
Layer Types:
Attention
MLP/FFN
Normalization
Embedding
Attention
MLP
Norm
Embedding
Clear
Expand All
Collapse All