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

Remove manual importing ipex in all-in-one benchmark (#11272)

Browse source
This commit is contained in:
Yuwen Hu 2024-06-11 09:32:13 +08:00 committed by GitHub
parent 70b17c87be
commit fac49f15e3
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 0 additions and 9 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

9
python/llm/dev/benchmark/all-in-one/run.py
View file

@ -446,7 +446,6 @@ def run_transformer_int4_gpu(repo_id,
lookahead=False):
from ipex_llm.transformers import AutoModel, AutoModelForCausalLM
from transformers import AutoTokenizer, GPTJForCausalLM, LlamaTokenizer
import intel_extension_for_pytorch as ipex
model_path = get_model_path(repo_id, local_model_hub)
# Load model in 4 bit,
# which convert the relevant layers in the model into INT4 format
@ -556,7 +555,6 @@ def run_optimize_model_gpu(repo_id,
batch_size):
from transformers import AutoModel, AutoModelForCausalLM, AutoTokenizer, GPTJForCausalLM, LlamaTokenizer
from ipex_llm import optimize_model
import intel_extension_for_pytorch as ipex
model_path = get_model_path(repo_id, local_model_hub)
# Load model in 4 bit,
# which convert the relevant layers in the model into INT4 format
@ -635,7 +633,6 @@ def run_ipex_fp16_gpu(repo_id,
batch_size):
from transformers import AutoModel, AutoModelForCausalLM
from transformers import AutoTokenizer, GPTJForCausalLM, LlamaTokenizer
import intel_extension_for_pytorch as ipex
model_path = get_model_path(repo_id, local_model_hub)
st = time.perf_counter()
if repo_id in CHATGLM_IDS:
@ -707,7 +704,6 @@ def run_bigdl_fp16_gpu(repo_id,
batch_size):
from ipex_llm.transformers import AutoModel, AutoModelForCausalLM
from transformers import AutoTokenizer, GPTJForCausalLM, LlamaTokenizer
import intel_extension_for_pytorch as ipex
model_path = get_model_path(repo_id, local_model_hub)
st = time.perf_counter()
if repo_id in CHATGLM_IDS:
@ -878,7 +874,6 @@ def run_transformer_int4_gpu_win(repo_id,
streaming):
from ipex_llm.transformers import AutoModel, AutoModelForCausalLM
from transformers import AutoTokenizer, GPTJForCausalLM, LlamaTokenizer, TextStreamer
import intel_extension_for_pytorch as ipex
model_path = get_model_path(repo_id, local_model_hub)
# Load model in 4 bit,
# which convert the relevant layers in the model into INT4 format
@ -983,7 +978,6 @@ def run_transformer_int4_fp16_gpu_win(repo_id,
streaming):
from ipex_llm.transformers import AutoModel, AutoModelForCausalLM
from transformers import AutoTokenizer, GPTJForCausalLM, LlamaTokenizer, TextStreamer
import intel_extension_for_pytorch as ipex
model_path = get_model_path(repo_id, local_model_hub)
# Load model in 4 bit,
# which convert the relevant layers in the model into INT4 format
@ -1092,7 +1086,6 @@ def run_transformer_int4_loadlowbit_gpu_win(repo_id,
streaming):
from ipex_llm.transformers import AutoModel, AutoModelForCausalLM
from transformers import AutoTokenizer, GPTJForCausalLM, LlamaTokenizer, TextStreamer
import intel_extension_for_pytorch as ipex
model_path = get_model_path(repo_id, local_model_hub)
# Load BigDL-LLM optimized low bit model
st = time.perf_counter()
@ -1487,7 +1480,6 @@ def run_deepspeed_optimize_model_gpu(repo_id,
from transformers import AutoModel, AutoModelForCausalLM, AutoTokenizer, GPTJForCausalLM, LlamaTokenizer
from ipex_llm import optimize_model
import intel_extension_for_pytorch as ipex
import deepspeed
from deepspeed.accelerator.cpu_accelerator import CPU_Accelerator
from deepspeed.accelerator import set_accelerator, get_accelerator
@ -1749,7 +1741,6 @@ def run_pipeline_parallel_gpu(repo_id,
n_gpu=2):
from ipex_llm.transformers import AutoModel, AutoModelForCausalLM
from transformers import AutoTokenizer, GPTJForCausalLM, LlamaTokenizer
import intel_extension_for_pytorch as ipex
model_path = get_model_path(repo_id, local_model_hub)
# Load model in 4 bit,
# which convert the relevant layers in the model into INT4 format