rixty/KoboldAI-Client
1
1
Fork 0
mirror of https://github.com/KoboldAI/KoboldAI-Client.git synced 2025-06-05 21:59:24 +02:00
Code Issues Projects Releases Wiki Activity

AutoGPTQ breakmodel

Browse Source
This commit is contained in:
Henk
2023-08-10 17:41:31 +02:00
parent 54addfc234
commit f2d7ef3aca
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
modeling/inference_models/gptq_hf_torch/class.py
View File

@@ -350,7 +350,7 @@ class model_backend(HFTorchInferenceModel):
auto_gptq.modeling._base.AutoConfig = hf_bleeding_edge.AutoConfig auto_gptq.modeling._base.AutoConfig = hf_bleeding_edge.AutoConfig
auto_gptq.modeling._base.AutoModelForCausalLM = hf_bleeding_edge.AutoModelForCausalLM auto_gptq.modeling._base.AutoModelForCausalLM = hf_bleeding_edge.AutoModelForCausalLM
model = AutoGPTQForCausalLM.from_quantized(location, model_basename=Path(gptq_file).stem, use_safetensors=gptq_file.endswith(".safetensors")) model = AutoGPTQForCausalLM.from_quantized(location, model_basename=Path(gptq_file).stem, use_safetensors=gptq_file.endswith(".safetensors"), device_map=device_map)
# Patch in embeddings function # Patch in embeddings function
def get_input_embeddings(self): def get_input_embeddings(self):