Hugging Face's logo

codefuse-ai
/
CodeFuse-DeepSeek-33B-4bits

Text Generation
Transformers
PyTorch
llama
text-generation-inference
Inference Endpoints
Model card Files Community
2
codefuse-admin commited on
Commit
8ab8766
·
verified ·
1 Parent(s): a3879ec

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -2
README.md CHANGED
@@ -175,7 +175,7 @@ def load_model_tokenizer(model_path):
175
  model = AutoGPTQForCausalLM.from_quantized("codefuse-ai/CodeFuse-DeepSeek-33B-4bits",
176
  inject_fused_attention=False,
177
  inject_fused_mlp=False,
178
- use_safetensors=False,
179
  use_cuda_fp16=True,
180
  disable_exllama=False,
181
  device_map='auto' # Support multi-gpus
@@ -386,7 +386,7 @@ def load_model_tokenizer(model_path):
386
  model = AutoGPTQForCausalLM.from_quantized("codefuse-ai/CodeFuse-DeepSeek-33B-4bits",
387
  inject_fused_attention=False,
388
  inject_fused_mlp=False,
389
- use_safetensors=False,
390
  use_cuda_fp16=True,
391
  disable_exllama=False,
392
  device_map='auto' # Support multi-gpus
 
175
  model = AutoGPTQForCausalLM.from_quantized("codefuse-ai/CodeFuse-DeepSeek-33B-4bits",
176
  inject_fused_attention=False,
177
  inject_fused_mlp=False,
178
+ use_safetensors=True,
179
  use_cuda_fp16=True,
180
  disable_exllama=False,
181
  device_map='auto' # Support multi-gpus
 
386
  model = AutoGPTQForCausalLM.from_quantized("codefuse-ai/CodeFuse-DeepSeek-33B-4bits",
387
  inject_fused_attention=False,
388
  inject_fused_mlp=False,
389
+ use_safetensors=True,
390
  use_cuda_fp16=True,
391
  disable_exllama=False,
392
  device_map='auto' # Support multi-gpus