arcadex-llm-g1-0.5b / model.yaml
ogmatrixllm's picture
Upload model.yaml
d937cc9 verified
raw
history blame contribute delete
314 Bytes
name: arcadex-llm-g1-0.5b
description: "Fine-tuned Pokémon LLM using Qwen-2.5 and PEFT LoRA."
framework: transformers
model:
path: ./ # Path to the local directory containing model files
tokenizer: ./tokenizer_config.json # Path to tokenizer
dtype: fp16 # Specify `fp16` for GPUs or `fp32` for CPUs