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
|
test_stage:
quant_modifiers:
QuantizationModifier:
ignore:
- model.layers.0.mlp.down_proj
- LlamaRotaryEmbedding
- LlamaRMSNorm
- SiLU
- MatMulLeftInput_QK
- MatMulRightInput_QK
- MatMulOutput_QK
- MatMulLeftInput_PV
- MatMulRightInput_PV
- MatMulOutput_PV
scheme_overrides:
Linear:
weights:
num_bits: 8
symmetric: true
strategy: "tensor"
input_activations:
num_bits: 8
symmetric: false
strategy: "tensor"
output_activations: null
Embedding:
weights:
num_bits: 8
symmetric: true
strategy: "tensor"
input_activations: null
output_activations: null
|
