Update app.py
Browse files
app.py
CHANGED
|
@@ -1,5 +1,7 @@
|
|
| 1 |
import gradio as gr
|
| 2 |
from huggingface_hub import InferenceClient
|
|
|
|
|
|
|
| 3 |
|
| 4 |
"""
|
| 5 |
For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
|
|
@@ -7,57 +9,29 @@ For more information on `huggingface_hub` Inference API support, please check th
|
|
| 7 |
client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
|
| 8 |
|
| 9 |
|
| 10 |
-
def respond(
|
| 11 |
-
message,
|
| 12 |
-
history: list[tuple[str, str]],
|
| 13 |
-
system_message,
|
| 14 |
-
max_tokens,
|
| 15 |
-
temperature,
|
| 16 |
-
top_p,
|
| 17 |
-
):
|
| 18 |
-
messages = [{"role": "system", "content": system_message}]
|
| 19 |
|
| 20 |
-
for val in history:
|
| 21 |
-
if val[0]:
|
| 22 |
-
messages.append({"role": "user", "content": val[0]})
|
| 23 |
-
if val[1]:
|
| 24 |
-
messages.append({"role": "assistant", "content": val[1]})
|
| 25 |
|
| 26 |
-
messages.append({"role": "user", "content": message})
|
| 27 |
|
| 28 |
-
|
|
|
|
|
|
|
|
|
|
| 29 |
|
| 30 |
-
|
| 31 |
-
|
| 32 |
-
|
| 33 |
-
|
| 34 |
-
|
| 35 |
-
|
| 36 |
-
):
|
| 37 |
-
token = message.choices[0].delta.content
|
| 38 |
|
| 39 |
-
|
| 40 |
-
|
|
|
|
|
|
|
|
|
|
| 41 |
|
| 42 |
-
|
| 43 |
-
|
| 44 |
-
""
|
| 45 |
-
|
| 46 |
-
|
| 47 |
-
additional_inputs=[
|
| 48 |
-
gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
|
| 49 |
-
gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
|
| 50 |
-
gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
|
| 51 |
-
gr.Slider(
|
| 52 |
-
minimum=0.1,
|
| 53 |
-
maximum=1.0,
|
| 54 |
-
value=0.95,
|
| 55 |
-
step=0.05,
|
| 56 |
-
label="Top-p (nucleus sampling)",
|
| 57 |
-
),
|
| 58 |
-
],
|
| 59 |
-
)
|
| 60 |
-
|
| 61 |
-
|
| 62 |
-
if __name__ == "__main__":
|
| 63 |
-
demo.launch()
|
|
|
|
| 1 |
import gradio as gr
|
| 2 |
from huggingface_hub import InferenceClient
|
| 3 |
+
from transformers import AutoTokenizer, AutoModelForCausalLM
|
| 4 |
+
import torch
|
| 5 |
|
| 6 |
"""
|
| 7 |
For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
|
|
|
|
| 9 |
client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
|
| 10 |
|
| 11 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
|
|
|
|
| 14 |
|
| 15 |
+
model = AutoModelForCausalLM.from_pretrained("Vikhrmodels/Vikhr-7B-instruct_0.4",
|
| 16 |
+
device_map="auto",
|
| 17 |
+
attn_implementation="flash_attention_2",
|
| 18 |
+
torch_dtype=torch.bfloat16)
|
| 19 |
|
| 20 |
+
tokenizer = AutoTokenizer.from_pretrained("Vikhrmodels/Vikhr-7B-instruct_0.4")
|
| 21 |
+
from transformers import AutoTokenizer, pipeline
|
| 22 |
+
pipe = pipeline("text-generation", model=model, tokenizer=tokenizer)
|
| 23 |
+
prompts = [
|
| 24 |
+
"В чем разница между фруктом и овощем?",
|
| 25 |
+
"Годы жизни колмагорова?"]
|
|
|
|
|
|
|
| 26 |
|
| 27 |
+
def test_inference(prompt):
|
| 28 |
+
prompt = pipe.tokenizer.apply_chat_template([{"role": "user", "content": prompt}], tokenize=False, add_generation_prompt=True)
|
| 29 |
+
print(prompt)
|
| 30 |
+
outputs = pipe(prompt, max_new_tokens=512, do_sample=True, num_beams=1, temperature=0.25, top_k=50, top_p=0.98, eos_token_id=79097)
|
| 31 |
+
return outputs[0]['generated_text'][len(prompt):].strip()
|
| 32 |
|
| 33 |
+
|
| 34 |
+
for prompt in prompts:
|
| 35 |
+
print(f" prompt:\n{prompt}")
|
| 36 |
+
print(f" response:\n{test_inference(prompt)}")
|
| 37 |
+
print("-"*50)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|