{ | |
"RoPE": 1, | |
"act_method": "swiglu", | |
"architectures": [ | |
"GPT" | |
], | |
"block_size": 1024, | |
"group_size": 2, | |
"model_type": "custom-gpt2", | |
"n_embd": 768, | |
"n_head": 12, | |
"n_layer": 12, | |
"norm_method": "rmsnorm", | |
"torch_dtype": "float32", | |
"transformers_version": "4.45.2", | |
"vocab_size": 50257 | |
} | |