Predict Output

How many tokens does "hello world" use with a GPT tokenizer?

PremiumCalculate whether…

Calculate how many tokens the rainbow flag emoji uses in a byte level BPE tokenizer with no emoji in training data.

PremiumCalculate how cl100k_base tokenizes…

The same sentence in English, Spanish, Hindi, and Japanese. Calculate the token ratio on o200k_base for each.

Calculate the approximate token count difference between 'Hello world' and its Arabic equivalent 'مرحبا بالعالم' in a GPT-4 tokenizer.

PremiumCalculate the token count…

Compute the RRF score for a candidate ranked 3 by dense and 5 by BM25

Compute the cost to embed 1 billion tokens with text-embedding-3-small

Compute the cost to embed 1B tokens with text-embedding-3-large via batch API

Compute storage for 1M docs at 768 dim, float32

Compute the raw storage needed for a 10M doc index at dimension 1536 (float32)

Compute the storage savings from truncating a 3072 dim model to 768

What is the typical layer count for a Llama 7B and a Llama 70B

Trace the attention score tensor shape in a 16-head block before softmax

Work out the KV cache reduction when 32 query heads share 8 KV heads

Calculate the per token KV cache footprint of a Llama-3-8B style config in fp16

Derive the SwiGLU hidden width that matches a 4x GELU FFN parameter budget

Compute the head dimension a config file implies for d_model 4096 and 32 heads

Apply the 12-L-d-squared rule to size a 24-layer, d_model 1024 decoder

Compute the trainable parameter count for a Llama style 7B decoder

Compute the embedding table parameter count for a 32k vocab at d_model 512

What value does a 0-indexed causal mask store at row 2, column 3 for a 4-token sequence?

Calculate the softmax output for a causal masked attention row.

Compute the KV cache memory for a single request at 4096 context on Llama-2 70B (MHA) in FP16.