vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

llama : add n_expert and n_expert_used to hparams + change quants

Browse source
This commit is contained in:
Georgi Gerganov 2023-12-10 13:57:54 +02:00
parent d1259b7b35
commit e640cbe055
No known key found for this signature in database
GPG key ID: 449E073F9DC10735
6 changed files with 111 additions and 54 deletions
Download patch file Download diff file Expand all files Collapse all files

2
gguf-py/gguf/constants.py
View file

@ -38,6 +38,8 @@ class Keys:
FEED_FORWARD_LENGTH = "{arch}.feed_forward_length"
USE_PARALLEL_RESIDUAL = "{arch}.use_parallel_residual"
TENSOR_DATA_LAYOUT = "{arch}.tensor_data_layout"
EXPERT_COUNT = "{arch}.expert_count"
EXPERT_USED_COUNT = "{arch}.expert_used_count"
class Attention:
HEAD_COUNT = "{arch}.attention.head_count"

6
gguf-py/gguf/gguf_writer.py
View file

@ -339,6 +339,12 @@ class GGUFWriter:
def add_clamp_kqv(self, value: float) -> None:
self.add_float32(Keys.Attention.CLAMP_KQV.format(arch=self.arch), value)
def add_expert_count(self, count: int) -> None:
self.add_uint32(Keys.LLM.EXPERT_COUNT.format(arch=self.arch), count)
def add_expert_used_count(self, count: int) -> None:
self.add_uint32(Keys.LLM.EXPERT_USED_COUNT.format(arch=self.arch), count)
def add_layer_norm_eps(self, value: float) -> None:
self.add_float32(Keys.Attention.LAYERNORM_EPS.format(arch=self.arch), value)