Spaces:
Build error
Build error
File size: 7,330 Bytes
9255bb8 b1c5198 a617135 9255bb8 216d322 9255bb8 08c6275 5fe1bef 310157b 08c6275 ca38a58 9255bb8 6150c8f 9255bb8 a617135 6150c8f 3925884 6150c8f 9255bb8 ca38a58 6150c8f 9255bb8 c8f6eb0 ca38a58 220e795 ca38a58 c8f6eb0 ca38a58 08c6275 2923d3d 08c6275 2923d3d 9255bb8 ad9b4e2 a47d2f3 ad9b4e2 9255bb8 6150c8f 83b80a1 6150c8f 83b80a1 adbf59b 6150c8f 216d322 a47d2f3 216d322 a47d2f3 216d322 a47d2f3 9a1c0e1 6150c8f a47d2f3 216d322 a47d2f3 9255bb8 6150c8f 3e1163e a47d2f3 6150c8f cd7fced a47d2f3 3925884 a47d2f3 9255bb8 6150c8f 3925884 6150c8f 5f1a3d0 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 |
import json
import os
import shutil
import requests
import spaces
import gradio as gr
from huggingface_hub import Repository
from text_generation import Client
from transformers import AutoModelForCausalLM, AutoTokenizer
from share_btn import community_icon_html, loading_icon_html, share_js, share_btn_css
checkpoint = "smallcloudai/Refact-1_6B-fim"
device = "cuda"
#device = "cpu" # for GPU usage or "cpu" for CPU usage
tokenizer = AutoTokenizer.from_pretrained(checkpoint)
model = AutoModelForCausalLM.from_pretrained(checkpoint, trust_remote_code=True).to(device)
FIM_PREFIX = "<fim_prefix>"
FIM_MIDDLE = "<fim_middle>"
FIM_SUFFIX = "<fim_suffix>"
FIM_INDICATOR = "<FILL_HERE>"
theme = gr.themes.Monochrome(
primary_hue="indigo",
secondary_hue="blue",
neutral_hue="slate",
radius_size=gr.themes.sizes.radius_sm,
font=[
gr.themes.GoogleFont("Open Sans"),
"ui-sans-serif",
"system-ui",
"sans-serif",
],
)
@spaces.GPU
def generate(
prompt, temperature=0.9, max_new_tokens=256, top_p=0.95, repetition_penalty=1.0, version="StarCoder",
):
temperature = float(temperature)
if temperature < 1e-2:
temperature = 1e-2
top_p = float(top_p)
fim_mode = False
generate_kwargs = dict(
temperature=temperature,
max_new_tokens=max_new_tokens,
top_p=top_p,
repetition_penalty=repetition_penalty,
do_sample=True,
seed=42,
)
if FIM_INDICATOR in prompt:
fim_mode = True
try:
prefix, suffix = prompt.split(FIM_INDICATOR)
except:
raise ValueError(f"Only one {FIM_INDICATOR} allowed in prompt!")
prompt = f"{FIM_PREFIX}{prefix}{FIM_SUFFIX}{suffix}{FIM_MIDDLE}"
inputs = tokenizer.encode(prompt, return_tensors="pt").to(device)
outputs = model.generate(inputs, max_length=100, temperature=0.2)
final = tokenizer.decode(outputs[0])
return final
examples = [
"X_train, y_train, X_test, y_test = train_test_split(X, y, test_size=0.1)\n\n# Train a logistic regression model, predict the labels on the test set and compute the accuracy score",
"// Returns every other value in the array as a new array.\nfunction everyOther(arr) {",
"Poor English: She no went to the market. Corrected English:",
"def alternating(list1, list2):\n results = []\n for i in range(min(len(list1), len(list2))):\n results.append(list1[i])\n results.append(list2[i])\n if len(list1) > len(list2):\n <FILL_HERE>\n else:\n results.extend(list2[i+1:])\n return results",
]
def process_example(args):
for x in generate(args):
pass
return x
css = ".generating {visibility: hidden}"
monospace_css = """
#q-input textarea {
font-family: monospace, 'Consolas', Courier, monospace;
}
"""
css += share_btn_css + monospace_css + ".gradio-container {color: black}"
description = """
<div style="text-align: center;">
<h1> Refact 1.6B <span style='color: #e6b800;'>Models</span> Playground</h1>
</div>
<div style="text-align: left;">
<p>This is a demo to generate text and code with the following model:</p>
<ul>
<li><a href="https://huggingface.co/smallcloudai/Refact-1_6B-fim" style='color: #e6b800;'>ReFact 1.6B</a>: An Open-Source Coding Assistant with Fine-Tuning on codebase, autocompletion, code refactoring, code analysis, integrated chat and more</li>
</ul>
<p><b>Please note:</b> This space is based on the Big Code Playground, and not all functionality may work. It is running on GPUZero, but can also be run on GPU/CPU.</p>
</div>
"""
with gr.Blocks(theme=theme, analytics_enabled=False, css=css) as demo:
with gr.Column():
gr.Markdown(description)
with gr.Row():
version = gr.Dropdown(
["Refact"],
value="Refact",
label="Model",
info="Choose a model from the list",
)
with gr.Row():
with gr.Column():
instruction = gr.Textbox(
placeholder="Enter your code here",
lines=5,
label="Input",
elem_id="q-input",
)
submit = gr.Button("Generate", variant="primary")
output = gr.Code(elem_id="q-output", lines=30, label="Output")
with gr.Row():
with gr.Column():
with gr.Accordion("Advanced settings", open=False):
with gr.Row():
column_1, column_2 = gr.Column(), gr.Column()
with column_1:
temperature = gr.Slider(
label="Temperature",
value=0.2,
minimum=0.0,
maximum=1.0,
step=0.05,
interactive=True,
info="Higher values produce more diverse outputs",
)
max_new_tokens = gr.Slider(
label="Max new tokens",
value=256,
minimum=0,
maximum=8192,
step=64,
interactive=True,
info="The maximum numbers of new tokens",
)
with column_2:
top_p = gr.Slider(
label="Top-p (nucleus sampling)",
value=0.90,
minimum=0.0,
maximum=1,
step=0.05,
interactive=True,
info="Higher values sample more low-probability tokens",
)
repetition_penalty = gr.Slider(
label="Repetition penalty",
value=1.2,
minimum=1.0,
maximum=2.0,
step=0.05,
interactive=True,
info="Penalize repeated tokens",
)
gr.Examples(
examples=examples,
inputs=[instruction],
cache_examples=False,
fn=process_example,
outputs=[output],
)
submit.click(
generate,
inputs=[instruction, temperature, max_new_tokens, top_p, repetition_penalty, version],
outputs=[output],
)
demo.launch(debug=True) |