22 lines
470 B
JSON
22 lines
470 B
JSON
|
|
{
|
||
|
|
"hidden_size": 2560,
|
||
|
|
"intermediate_size": 9728,
|
||
|
|
"num_hidden_layers": 36,
|
||
|
|
"num_attention_heads": 32,
|
||
|
|
"num_key_value_heads": 8,
|
||
|
|
"head_dim": 128,
|
||
|
|
"vocab_size": 151936,
|
||
|
|
"architectures": [
|
||
|
|
"Qwen3ForCausalLM"
|
||
|
|
],
|
||
|
|
"model_type": "qwen3",
|
||
|
|
"max_position_embeddings": 32768,
|
||
|
|
"torch_dtype": "bfloat16",
|
||
|
|
"bos_token_id": 151643,
|
||
|
|
"eos_token_id": 151645,
|
||
|
|
"tie_word_embeddings": false,
|
||
|
|
"rope_scaling": {
|
||
|
|
"type": "dynamic",
|
||
|
|
"factor": 2.0
|
||
|
|
}
|
||
|
|
}
|