ayo/ipex-llm
1
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 1

remove load_in_8bit usage as it is not supported a long time ago (#12779)

Browse source
This commit is contained in:
Yishuo Wang 2025-02-07 11:21:29 +08:00 committed by GitHub
parent 9e9b6c9f2b
commit d0d9c9d636
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
7 changed files with 9 additions and 15 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
python/llm/dev/benchmark/harness/README.md
View file

@ -56,7 +56,7 @@ to the following codes to load the low bit models.
class ModifiedAutoModelForCausalLM(AutoModelForCausalLM): class ModifiedAutoModelForCausalLM(AutoModelForCausalLM):
@classmethod @classmethod
def load_low_bit(cls,*args,**kwargs): def load_low_bit(cls,*args,**kwargs):
for k in ['load_in_low_bit', 'device_map', 'max_memory', 'load_in_8bit','load_in_4bit']: for k in ['load_in_low_bit', 'device_map', 'max_memory','load_in_4bit']:
kwargs.pop(k) kwargs.pop(k)
return super().load_low_bit(*args, **kwargs) return super().load_low_bit(*args, **kwargs)

1
python/llm/example/GPU/LLM-Finetuning/axolotl/llama3-qlora.yml
View file

@ -3,7 +3,6 @@ base_model: meta-llama/Meta-Llama-3-8B
model_type: AutoModelForCausalLM model_type: AutoModelForCausalLM
tokenizer_type: AutoTokenizer tokenizer_type: AutoTokenizer
load_in_8bit: false
load_in_4bit: true load_in_4bit: true
strict: false strict: false

1
python/llm/example/GPU/LLM-Finetuning/axolotl/lora.yml
View file

@ -3,7 +3,6 @@ base_model: NousResearch/Llama-2-7b-hf
model_type: LlamaForCausalLM model_type: LlamaForCausalLM
tokenizer_type: LlamaTokenizer tokenizer_type: LlamaTokenizer
load_in_8bit: false
load_in_4bit: true load_in_4bit: true
strict: false strict: false

1
python/llm/example/GPU/LLM-Finetuning/axolotl/qlora.yml
View file

@ -4,7 +4,6 @@ model_type: LlamaForCausalLM
tokenizer_type: LlamaTokenizer tokenizer_type: LlamaTokenizer
is_llama_derived_model: true is_llama_derived_model: true
load_in_8bit: false
load_in_4bit: true load_in_4bit: true
strict: false strict: false

1
python/llm/example/GPU/Speculative-Decoding/EAGLE/evaluation/gen_ea_answer_llama2chat_e2_ipex_optimize.py
View file

@ -312,7 +312,6 @@ def get_model_answers(
torch_dtype=torch.float16, torch_dtype=torch.float16,
# torch_dtype=torch.float32, # torch_dtype=torch.float32,
low_cpu_mem_usage=True, low_cpu_mem_usage=True,
# load_in_8bit=True,
total_token=args.total_token, total_token=args.total_token,
depth=args.depth, depth=args.depth,
top_k=args.top_k, top_k=args.top_k,

1
python/llm/src/ipex_llm/transformers/model.py
View file

@ -233,7 +233,6 @@ class _BaseAutoModelClass:
optimize_model = False optimize_model = False
kwargs["modules_to_not_convert"] = ["lm_head"] kwargs["modules_to_not_convert"] = ["lm_head"]
load_in_8bit = kwargs.pop("load_in_8bit", False)
from ipex_llm.llm_patching import bigdl_patched from ipex_llm.llm_patching import bigdl_patched
if bigdl_patched == 'Train': if bigdl_patched == 'Train':
global patched_training_mode global patched_training_mode

1
python/llm/src/ipex_llm/transformers/npu_model.py
View file

@ -117,7 +117,6 @@ class _BaseAutoModelClass:
# ignore following arguments # ignore following arguments
ignore_argument(kwargs, "model_hub") ignore_argument(kwargs, "model_hub")
ignore_argument(kwargs, "load_in_4bit") ignore_argument(kwargs, "load_in_4bit")
ignore_argument(kwargs, "load_in_8bit")
ignore_argument(kwargs, "imatrix") ignore_argument(kwargs, "imatrix")
ignore_argument(kwargs, "cpu_embedding") ignore_argument(kwargs, "cpu_embedding")
ignore_argument(kwargs, "embedding_qtype") ignore_argument(kwargs, "embedding_qtype")