File size: 941 Bytes
04c8b0f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
{
    "fp8": {
        "model\\.layers\\.[0-2]": [
            "mlp.down_proj",
            "mlp.gate_proj",
            "mlp.up_proj",
            "self_attn.kv_a_proj_with_mqa",
            "self_attn.kv_b_proj",
            "self_attn.o_proj",
            "self_attn.q_a_proj",
            "self_attn.q_b_proj"
        ],
        "model\\.layers\\.(?![0-2](?:[^0-9]|$))\\d+\\.self_attn": [
            "kv_a_proj_with_mqa",
            "kv_b_proj",
            "o_proj",
            "q_a_proj",
            "q_b_proj"
        ],
        "model\\.layers\\.(?![0-2](?:[^0-9]|$))\\d+\\.mlp\\.shared_experts": [
            "down_proj",
            "gate_proj",
            "up_proj"
        ],
        "model\\.layers\\.(?![0-2](?:[^0-9]|$))\\d+\\.mlp\\.experts\\.\\d+": [
            "down_proj",
            "gate_proj",
            "up_proj"
        ]
    },
    "bf16": {
        "model\\.embed_tokens": [],
        "lm_head": []
    }
}