WesleyGPT-Think / config.json
Wasue's picture
WesleyGPT-Think: SFT step 1038, think-format math
f3e6d61 verified
Raw History Blame Contribute Delete
213 Bytes
{
"format": "nanochat-gpt/1",
"model_config": {
"sequence_len": 2048,
"vocab_size": 32768,
"n_layer": 12,
"n_head": 6,
"n_kv_head": 6,
"n_embd": 768,
"window_pattern": "SSSL"
}
}