File size: 4,257 Bytes
f732d7c 135325d f732d7c b555022 f732d7c b555022 f732d7c b555022 f732d7c d6c0cbe f732d7c b555022 f732d7c b555022 f732d7c 135325d f732d7c |
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 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 |
from typing import Optional
import gradio as gr
import torch
from peft import PeftModel
from transformers import GenerationConfig
from transformers import LlamaForCausalLM
from transformers import LlamaTokenizer
print("starting server ...")
BASE_MODEL = "decapoda-research/llama-13b-hf"
LORA_WEIGHTS = "izumi-lab/llama-13b-japanese-lora-v0-1ep"
tokenizer = LlamaTokenizer.from_pretrained(BASE_MODEL)
if torch.cuda.is_available():
device = "cuda"
else:
device = "cpu"
try:
if torch.backends.mps.is_available():
device = "mps"
except Exception:
pass
if device == "cuda":
model = LlamaForCausalLM.from_pretrained(
BASE_MODEL,
load_in_8bit=False,
torch_dtype=torch.float16,
device_map="auto",
)
model = PeftModel.from_pretrained(model, LORA_WEIGHTS, torch_dtype=torch.float16)
elif device == "mps":
model = LlamaForCausalLM.from_pretrained(
BASE_MODEL,
device_map={"": device},
torch_dtype=torch.float16,
)
model = PeftModel.from_pretrained(
model,
LORA_WEIGHTS,
device_map={"": device},
torch_dtype=torch.float16,
)
else:
model = LlamaForCausalLM.from_pretrained(
BASE_MODEL, device_map={"": device}, low_cpu_mem_usage=True
)
model = PeftModel.from_pretrained(
model,
LORA_WEIGHTS,
device_map={"": device},
)
def generate_prompt(instruction: str, input: Optional[str] = None):
if input:
return f"""Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.
### Instruction:
{instruction}
### Input:
{input}
### Response:"""
else:
return f"""Below is an instruction that describes a task. Write a response that appropriately completes the request.
### Instruction:
{instruction}
### Response:"""
if device != "cpu":
model.half()
model.eval()
if torch.__version__ >= "2":
model = torch.compile(model)
def evaluate(
instruction: str,
input: Optional[str] = None,
temperature: float = 0.7,
top_p: float = 1.0,
top_k: int = 40,
num_beams: int = 4,
max_new_tokens: int = 256,
**kwargs,
):
prompt = generate_prompt(instruction, input)
inputs = tokenizer(prompt, return_tensors="pt")
input_ids = inputs["input_ids"].to(device)
generation_config = GenerationConfig(
temperature=temperature,
top_p=top_p,
top_k=top_k,
num_beams=num_beams,
**kwargs,
)
with torch.no_grad():
generation_output = model.generate(
input_ids=input_ids,
generation_config=generation_config,
return_dict_in_generate=True,
output_scores=True,
max_new_tokens=max_new_tokens,
)
s = generation_output.sequences[0]
output = tokenizer.decode(s)
return output.split("### Response:")[1].strip()
g = gr.Interface(
fn=evaluate,
inputs=[
gr.components.Textbox(lines=2, label="Instruction", placeholder="東京から大阪に行くには？"),
gr.components.Textbox(lines=2, label="Input", placeholder="none"),
gr.components.Slider(minimum=0, maximum=1, value=0.7, label="Temperature"),
gr.components.Slider(minimum=0, maximum=1, value=1.0, label="Top p"),
gr.components.Slider(minimum=0, maximum=100, step=1, value=40, label="Top k"),
gr.components.Slider(minimum=1, maximum=4, step=1, value=4, label="Beams"),
gr.components.Slider(
minimum=1, maximum=512, step=1, value=128, label="Max tokens"
),
],
outputs=[
gr.inputs.Textbox(
lines=5,
label="Output",
)
],
title="izumi-lab/calm-7b-lora-v0-1ep",
description="izumi-lab/calm-7b-lora-v0-1ep is a 7B-parameter Calm model finetuned to follow instructions. It is trained on the [izumi-lab/llm-japanese-dataset](https://huggingface.co/datasets/izumi-lab/llm-japanese-dataset) dataset and makes use of the Huggingface Calm-7b implementation. For more information, please visit [the project's website](https://llm.msuzuki.me).",
)
g.queue(concurrency_count=1)
print("loading completed")
g.launch()
|