Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Update AIPC_Inference.md to reflect changes in intel-npu-acceleration-library #246

Merged
merged 5 commits into from
Jan 16, 2025
Merged

Update AIPC_Inference.md to reflect changes in intel-npu-acceleration-library #246

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
85fa72e
Update AIPC_Inference.md
Jaylyn-Barbee Jan 13, 2025
b0efcee
Update AIPC_NPU_DEMO.ipynb
Jaylyn-Barbee Jan 13, 2025
09e86c7
Update AIPC_NPU_DEMO.ipynb
Jaylyn-Barbee Jan 13, 2025
f5c8e38
Update AIPC_NPU_DEMO.ipynb
Jaylyn-Barbee Jan 13, 2025
277453d
Update AIPC_NPU_DEMO.ipynb
Jaylyn-Barbee Jan 13, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
13 changes: 6 additions & 7 deletions code/03.Inference/AIPC/AIPC_NPU_DEMO.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,7 +15,8 @@
"outputs": [],
"source": [
"from transformers import AutoTokenizer, pipeline,TextStreamer\n",
"import intel_npu_acceleration_library as npu_lib\n",
"from intel_npu_acceleration_library import NPUModelForCausalLM, int4\n",
"from intel_npu_acceleration_library.compiler import CompilerConfig\n",
"\n",
"\n",
"import warnings"
Expand Down Expand Up @@ -84,12 +85,10 @@
}
],
"source": [
"model = npu_lib.NPUModelForCausalLM.from_pretrained(\n",
" model_id,\n",
" torch_dtype=\"auto\",\n",
" dtype=npu_lib.int4,\n",
" trust_remote_code=True\n",
" )\n",
"compiler_conf = CompilerConfig(dtype=int4)\n",
"model = NPUModelForCausalLM.from_pretrained(\n",
" model_id, use_cache=True, config=compiler_conf, attn_implementation='sdpa'\n",
").eval()\n",
"\n",
"tokenizer = AutoTokenizer.from_pretrained(model_id)\n",
"\n",
Expand Down
38 changes: 16 additions & 22 deletions md/03.Inference/AIPC_Inference.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -46,51 +46,45 @@ Install the Python Library with pip
Using Intel NPU acceleration, this library does not affect the traditional encoding process. You only need to use this library to quantize the original Phi-3 model, such as FP16，INT8，INT4，such as

```python

from transformers import AutoTokenizer, pipeline,TextStreamer
import intel_npu_acceleration_library as npu_lib
from intel_npu_acceleration_library import NPUModelForCausalLM, int4
from intel_npu_acceleration_library.compiler import CompilerConfig
import warnings

model_id = "microsoft/Phi-3-mini-4k-instruct"

model = npu_lib.NPUModelForCausalLM.from_pretrained(
model_id,
torch_dtype="auto",
dtype=npu_lib.int4,
trust_remote_code=True
)
compiler_conf = CompilerConfig(dtype=int4)
model = NPUModelForCausalLM.from_pretrained(
model_id, use_cache=True, config=compiler_conf, attn_implementation="sdpa"
).eval()

tokenizer = AutoTokenizer.from_pretrained(model_id)

text_streamer = TextStreamer(tokenizer, skip_prompt=True)

```
After the quantification is successful, continue execution to call the NPU to run the Phi-3 model.

After the quantification is successful, continue execution to call the NPU to run the Phi-3 model.

```python

generation_args = {
"max_new_tokens": 1024,
"return_full_text": False,
"temperature": 0.3,
"do_sample": False,
"streamer": text_streamer,
}
"max_new_tokens": 1024,
"return_full_text": False,
"temperature": 0.3,
"do_sample": False,
"streamer": text_streamer,
}

pipe = pipeline(
"text-generation",
model=model,
tokenizer=tokenizer,
"text-generation",
model=model,
tokenizer=tokenizer,
)

query = "<|system|>You are a helpful AI assistant.<|end|><|user|>Can you introduce yourself?<|end|><|assistant|>"

with warnings.catch_warnings():
warnings.simplefilter("ignore")
pipe(query, **generation_args)


```

When executing code, we can view the running status of the NPU through Task Manager
Expand Down
Loading