gpt

GPT API calling assistant. Build chat completion requests, generate embedding payloads, prepare fine-tuning data, estimate costs, batch requests, and convert between formats.

Commands

chat

Build a chat completion request JSON payload from messages.

bash scripts/script.sh chat --system "You are a translator" --user "Translate: hello"
bash scripts/script.sh chat --system "Helper" --user "Hi" --model gpt-4o --temperature 0.7

embed

Generate an embedding API request payload.

bash scripts/script.sh embed --input "Text to embed"
bash scripts/script.sh embed --file texts.txt --model text-embedding-3-small

finetune

Prepare and validate fine-tuning JSONL data from source files.

bash scripts/script.sh finetune --input pairs.csv --output training.jsonl
bash scripts/script.sh finetune --validate training.jsonl

cost

Estimate API costs based on token count and model pricing.

bash scripts/script.sh cost --tokens 5000 --model gpt-4o
bash scripts/script.sh cost --file input.txt --model gpt-4-turbo

batch

Generate multiple API request payloads from a list of inputs.

bash scripts/script.sh batch --file prompts.txt --model gpt-4o --output batch_requests.jsonl

convert

Convert between JSONL and CSV data formats.

bash scripts/script.sh convert --input data.csv --to jsonl --output data.jsonl
bash scripts/script.sh convert --input data.jsonl --to csv --output data.csv

Output

All commands output JSON request bodies or JSONL to stdout by default. Use --output to write to a file. Cost estimates print a breakdown table with per-model pricing.

Requirements

• bash 4+

Feedback

Report issues or suggestions: https://bytesagain.com/feedback/

Powered by BytesAgain | bytesagain.com