Enter a valid HuggingFace model ID (e.g., "mistralai/Mistral-7B-Instruct-v0.3").
For Korean, you might use "beomi/KoAlpaca-Polyglot-12.8B" or "skt/kogpt2-base-v2", etc.
The model must have a tokenizer available and be accessible.
Loaded
Token Visualization
Note: Showing preview of first 8096 characters. Stats are calculated on the full file.
Note: Only showing first 50,000 tokens. Total token count: 0