← All Models | GPT - OpenAI's Generative Pre-trained Transformers:

openai/whisper-small

📊 Model Parameters

Total Parameters 193,413,120
Context Length 2,048
Hidden Size 768
Layers 12
Attention Heads 12
KV Heads 12

💾 Memory Requirements

FP32 (Full) 737.8 MB
FP16 (Half) 368.9 MB
INT8 (Quantized) 184.5 MB
INT4 (Quantized) 92.2 MB

🔑 KV Cache (Inference)

Per Token (FP16) 36.86 KB
Max Context FP32 144.0 MB
Max Context FP16 72.0 MB
Max Context INT8 36.0 MB

⚙️ Model Configuration

Core Architecture

Vocabulary Size51,865
Number of Layers12

Attention Configuration

Attention Dropout0%
Tied EmbeddingsYes

Activation & Normalization

Activation Functiongelu

Special Tokens

BOS Token ID50,257
Pad Token ID50,257
EOS Token ID50257

Data Type

Model Dtypefloat32
Layer Types:
Attention
MLP/FFN
Normalization
Embedding