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

openai/whisper-large-v3-turbo

📊 Model Parameters

Total Parameters 238,297,600
Context Length 2,048
Hidden Size 1280
Layers 32
Attention Heads 20
KV Heads 20

💾 Memory Requirements

FP32 (Full) 909.0 MB
FP16 (Half) 454.5 MB
INT8 (Quantized) 227.3 MB
INT4 (Quantized) 113.6 MB

🔑 KV Cache (Inference)

Per Token (FP16) 163.84 KB
Max Context FP32 640.0 MB
Max Context FP16 320.0 MB
Max Context INT8 160.0 MB

⚙️ Model Configuration

Core Architecture

Vocabulary Size51,866
Number of Layers32

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 Dtypefloat16
Layer Types:
Attention
MLP/FFN
Normalization
Embedding