mirror of
https://github.com/ggml-org/llama.cpp.git
synced 2026-08-27 07:31:24 +02:00
Xuan-Son Nguyen
fa88ae9368
* convert: add @ModelBase.example * add docs * add more variants * BailingMoeV3ForCausalLM * rm pocket-tts
36 lines
1.1 KiB
Python
36 lines
1.1 KiB
Python
from __future__ import annotations
|
|
|
|
from typing import Callable, TYPE_CHECKING
|
|
|
|
if TYPE_CHECKING:
|
|
from torch import Tensor
|
|
|
|
from .base import MmprojModel, ModelBase, gguf
|
|
|
|
from .llama import LlamaModel
|
|
|
|
|
|
@ModelBase.register("CogVLMForCausalLM")
|
|
@ModelBase.example("THUDM/cogvlm2-llama3-chat-19B", "THUDM/cogvlm-chat-hf")
|
|
class CogVLMVisionModel(MmprojModel):
|
|
|
|
def set_gguf_parameters(self):
|
|
super().set_gguf_parameters()
|
|
self.gguf_writer.add_vision_attention_layernorm_eps(self.hparams.get("layer_norm_eps", 1e-6))
|
|
self.gguf_writer.add_clip_projector_type(gguf.VisionProjectorType.COGVLM)
|
|
|
|
@classmethod
|
|
def filter_tensors(cls, item: tuple[str, Callable[[], Tensor]]) -> tuple[str, Callable[[], Tensor]] | None:
|
|
name, gen = item
|
|
|
|
if not name.startswith("model.vision."):
|
|
return None
|
|
|
|
return super().filter_tensors(item)
|
|
|
|
|
|
@ModelBase.register("CogVLMForCausalLM")
|
|
@ModelBase.example("THUDM/cogvlm2-llama3-chat-19B", "THUDM/cogvlm-chat-hf")
|
|
class CogVLMModel(LlamaModel):
|
|
model_arch = gguf.MODEL_ARCH.COGVLM
|