forked from ed-donner/llm_engineering
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathpricer_service.py
66 lines (51 loc) · 2.17 KB
/
pricer_service.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
import modal
from modal import App, Image
# Setup - define our infrastructure with code!
app = modal.App("pricer-service")
image = Image.debian_slim().pip_install("torch", "transformers", "bitsandbytes", "accelerate", "peft")
secrets = [modal.Secret.from_name("hf-secret")]
# Constants
GPU = "T4"
BASE_MODEL = "meta-llama/Meta-Llama-3.1-8B"
PROJECT_NAME = "pricer"
HF_USER = "ed-donner" # your HF name here! Or use mine if you just want to reproduce my results.
RUN_NAME = "2024-09-13_13.04.39"
PROJECT_RUN_NAME = f"{PROJECT_NAME}-{RUN_NAME}"
REVISION = "e8d637df551603dc86cd7a1598a8f44af4d7ae36"
FINETUNED_MODEL = f"{HF_USER}/{PROJECT_RUN_NAME}"
@app.function(image=image, secrets=secrets, gpu=GPU, timeout=1800)
def price(description: str) -> float:
import os
import re
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig, set_seed
from peft import PeftModel
QUESTION = "How much does this cost to the nearest dollar?"
PREFIX = "Price is $"
prompt = f"{QUESTION}\n{description}\n{PREFIX}"
# Quant Config
quant_config = BitsAndBytesConfig(
load_in_4bit=True,
bnb_4bit_use_double_quant=True,
bnb_4bit_compute_dtype=torch.bfloat16,
bnb_4bit_quant_type="nf4"
)
# Load model and tokenizer
tokenizer = AutoTokenizer.from_pretrained(BASE_MODEL)
tokenizer.pad_token = tokenizer.eos_token
tokenizer.padding_side = "right"
base_model = AutoModelForCausalLM.from_pretrained(
BASE_MODEL,
quantization_config=quant_config,
device_map="auto"
)
fine_tuned_model = PeftModel.from_pretrained(base_model, FINETUNED_MODEL, revision=REVISION)
set_seed(42)
inputs = tokenizer.encode(prompt, return_tensors="pt").to("cuda")
attention_mask = torch.ones(inputs.shape, device="cuda")
outputs = fine_tuned_model.generate(inputs, attention_mask=attention_mask, max_new_tokens=5, num_return_sequences=1)
result = tokenizer.decode(outputs[0])
contents = result.split("Price is $")[1]
contents = contents.replace(',','')
match = re.search(r"[-+]?\d*\.\d+|\d+", contents)
return float(match.group()) if match else 0