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

Initial OpenELM support (270M only so far)

Browse source
This commit is contained in:
Icecream95 2024-05-18 19:41:42 +12:00
parent c1b295eea5
commit 217d8d7b77
4 changed files with 322 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

14
gguf-py/gguf/constants.py
View file

@ -139,6 +139,7 @@ class MODEL_ARCH(IntEnum):
COMMAND_R = auto()
DBRX = auto()
OLMO = auto()
OPENELM = auto()
class MODEL_TENSOR(IntEnum):
@ -217,6 +218,7 @@ MODEL_ARCH_NAMES: dict[MODEL_ARCH, str] = {
MODEL_ARCH.COMMAND_R: "command-r",
MODEL_ARCH.DBRX: "dbrx",
MODEL_ARCH.OLMO: "olmo",
MODEL_ARCH.OPENELM: "openelm",
}
TENSOR_NAMES: dict[MODEL_TENSOR, str] = {
@ -743,6 +745,18 @@ MODEL_TENSORS: dict[MODEL_ARCH, list[MODEL_TENSOR]] = {
MODEL_TENSOR.FFN_DOWN,
MODEL_TENSOR.FFN_UP,
],
MODEL_ARCH.OPENELM: [
MODEL_TENSOR.TOKEN_EMBD,
MODEL_TENSOR.OUTPUT_NORM,
MODEL_TENSOR.ATTN_NORM,
MODEL_TENSOR.ATTN_QKV,
MODEL_TENSOR.ATTN_Q_NORM,
MODEL_TENSOR.ATTN_K_NORM,
MODEL_TENSOR.ATTN_OUT,
MODEL_TENSOR.FFN_NORM,
MODEL_TENSOR.FFN_UP,
MODEL_TENSOR.FFN_DOWN,
],
# TODO
}