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

Merge pull request #471 from nkpz/patch-2

Browse Source 
[gptq_hf_torch] Fix typo in model type check
This commit is contained in:
henk717
2023-09-23 11:26:06 +02:00
committed by GitHub
parent db5ff9e748 d8877b642d
commit 80f34cb2e8
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

@@ -362,7 +362,7 @@ class model_backend(HFTorchInferenceModel):
model = load_quant_offload_device_map(llama_load_quant, location, gptq_file, gptq_bits, gptq_groupsize, device_map, force_bias=v2_bias) model = load_quant_offload_device_map(llama_load_quant, location, gptq_file, gptq_bits, gptq_groupsize, device_map, force_bias=v2_bias)
elif model_type == "opt": elif model_type == "opt":
model = load_quant_offload_device_map(opt_load_quant, location, gptq_file, gptq_bits, gptq_groupsize, device_map, force_bias=v2_bias) model = load_quant_offload_device_map(opt_load_quant, location, gptq_file, gptq_bits, gptq_groupsize, device_map, force_bias=v2_bias)
elif model_tseype == "mpt": elif model_type == "mpt":
model = load_quant_offload_device_map(mpt_load_quant, location, gptq_file, gptq_bits, gptq_groupsize, device_map, force_bias=v2_bias) model = load_quant_offload_device_map(mpt_load_quant, location, gptq_file, gptq_bits, gptq_groupsize, device_map, force_bias=v2_bias)
elif model_type == "gpt_bigcode": elif model_type == "gpt_bigcode":
model = load_quant_offload_device_map(bigcode_load_quant, location, gptq_file, gptq_bits, gptq_groupsize, device_map, force_bias=v2_bias).half() model = load_quant_offload_device_map(bigcode_load_quant, location, gptq_file, gptq_bits, gptq_groupsize, device_map, force_bias=v2_bias).half()