File size: 7,477 Bytes
d89e980
 
9523a2b
 
d93bf74
 
9523a2b
 
 
909aca2
 
 
9523a2b
 
 
42f7557
909aca2
42f7557
5080c22
d89e980
 
 
d93bf74
 
 
 
 
 
 
909aca2
 
c911921
5080c22
0161136
909aca2
d89e980
0161136
d89e980
0161136
 
 
42f7557
 
 
0161136
 
 
 
e433d54
0161136
 
0dc9507
0161136
 
 
 
9523a2b
42f7557
fd0376c
42f7557
 
 
d89e980
42f7557
 
fd0376c
42f7557
 
 
 
 
d89e980
42f7557
 
9523a2b
5080c22
 
909aca2
 
 
 
 
d93bf74
909aca2
 
 
 
5080c22
 
 
 
909aca2
0161136
 
 
 
 
 
 
 
 
 
909aca2
 
0161136
9523a2b
 
 
909aca2
9523a2b
 
 
 
 
 
909aca2
 
 
 
 
 
9523a2b
909aca2
 
9523a2b
0161136
 
 
 
9523a2b
 
 
909aca2
 
9523a2b
 
d93bf74
9523a2b
 
 
 
909aca2
9523a2b
 
909aca2
9523a2b
 
d89e980
 
9523a2b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ec833bc
9523a2b
 
 
 
 
 
 
 
909aca2
 
 
 
 
9523a2b
 
 
 
 
 
 
 
45b40bd
 
9523a2b
 
 
 
 
 
 
c4eee19
 
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
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
"""Run."""
# pylint: disable=invalid-name,line-too-long,broad-except,missing-function-docstring
from __future__ import annotations

import os
import time
from typing import Iterable

import gradio as gr
import pynvml

# import torch
from ctransformers import AutoModelForCausalLM
from gradio.themes.base import Base
from gradio.themes.utils import colors, fonts, sizes
from huggingface_hub import hf_hub_download, hf_hub_url  # snapshot_download,
from loguru import logger
from python_run_cmd import run_cmd

ret = run_cmd("which aria2c", mute_stdout=False)
logger.debug(ret)

os.environ["TZ"] = "Asia/Shanghai"
try:
    time.tzset()  # type: ignore
    logger.debug(f"Timezone set to {os.environ['TZ']=}")
except AttributeError:
    ...  # Windows

repo_id = "TheBloke/openbuddy-mistral-7B-v13-GGUF"
filename = "openbuddy-mistral-7b-v13.Q4_K_S.gguf"  # 4.17G
filename = "openbuddy-mistral-7b-v13.Q4_K_M.gguf"  # 4.39G

model_ready = True
logger.debug("Start dl")

# try to download 5 times:
model_path = f"./{filename}"
for idx in range(5):
    logger.debug(f"attempt {idx + 1}")
    try:
        model_path = hf_hub_download(
            repo_id=repo_id, filename=filename, revision="main"
        )
        break
    except Exception as exc:
        logger.error(f"failed to download {filename}: {exc}")
        # raise SystemExit("hf acting up, can't donwload the model {filename=}, exiting")
        time.sleep(3)
else:
    logger.warning("Tried 5 times to no vain")
    # raise gr.Error(f"hf acting up, can't donwload the model {filename=}, exiting")
    # raise SystemExit("hf acting up, can't donwload the model {filename=}, exiting")
    model_ready = False

logger.debug(f"Done dl, {model_ready=}")

if not model_ready:  # try aria2c
    logger.debug("Try wget...")
    url = hf_hub_url(
        repo_id,
        filename,
        # revision="main",
    )
    logger.debug(f"{url=}")
    ret = run_cmd(f"wget -c {url}", mute_stdout=False)
    logger.debug(ret)

    model_path = f"./{filename}"

    # both successful
    if not ret.returncode:
        model_ready = True

# Set gpu_layers to the number of layers to offload to GPU. Set to 0 if no GPU acceleration is available on your system.
# model = AutoModelForCausalLM.from_pretrained("TheBloke/openbuddy-mistral-7B-v13-GGUF", model_file="openbuddy-mistral-7b-v13.Q4_K_S.gguf", model_type="mistral", gpu_layers=0)

has_cuda = False
try:
    pynvml.nvmlInit()
    has_cuda = True
    logger.debug("has cuda")
except pynvml.nvml.NVMLError_LibraryNotFound:  # type: ignore
    logger.debug("no cuda")

# if torch.cuda.is_available():
if has_cuda:
    gpu_layers = 50  # set to what you like for GPU
else:
    gpu_layers = 0

logger.debug("Start loading the model")
try:
    model = AutoModelForCausalLM.from_pretrained(
        model_path, model_type="mistral", gpu_layers=gpu_layers
    )
except Exception as exc:
    logger.error(exc)
    model_ready = False
    model = None

logger.debug(f"Done loading the model, {model_ready=}")

ins = """[INST] <<FRIDAY>>
Remember that your English name is "openbuddy" and your name in Chinese is "开友". You are a helpful, respectful and honest assistant. Always answer as helpfully as possible, while being safe.  Your answers should not include any harmful, unethical, racist, sexist, toxic, dangerous, or illegal content. Please ensure that your responses are socially unbiased and positive in nature.
If a question does not make any sense, or is not factually coherent, explain why instead of answering something not correct. If you don't know the answer to a question, please don't share false information.
<</FRIDAY>>
{} [/INST]
"""

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",
    ],
)


def response(question):
    if model is None:
        res = "model not ready (got a problem with downloading the file {filename=} from hf.co)"
    else:
        res = model(ins.format(question))
    yield res


examples = ["Hello!"]


def process_example(args):
    x = None
    for x in response(args):
        pass
    return x


css = ".generating {visibility: hidden}"


# Based on the gradio theming guide and borrowed from https://huggingface.co/spaces/shivi/dolly-v2-demo
class SeafoamCustom(Base):
    """Define."""

    def __init__(
        self,
        *,
        primary_hue: colors.Color | str = colors.emerald,
        secondary_hue: colors.Color | str = colors.blue,
        neutral_hue: colors.Color | str = colors.blue,
        spacing_size: sizes.Size | str = sizes.spacing_md,
        radius_size: sizes.Size | str = sizes.radius_md,
        font: fonts.Font
        | str
        | Iterable[fonts.Font | str] = (
            fonts.GoogleFont("Quicksand"),
            "ui-sans-serif",
            "sans-serif",
        ),
        font_mono: fonts.Font
        | str
        | Iterable[fonts.Font | str] = (
            fonts.GoogleFont("IBM Plex Mono"),
            "ui-monospace",
            "monospace",
        ),
    ):
        """Init."""
        super().__init__(
            primary_hue=primary_hue,
            secondary_hue=secondary_hue,
            neutral_hue=neutral_hue,
            spacing_size=spacing_size,
            radius_size=radius_size,
            font=font,
            font_mono=font_mono,
        )
        super().set(
            button_primary_background_fill="linear-gradient(90deg, *primary_300, *secondary_400)",
            button_primary_background_fill_hover="linear-gradient(90deg, *primary_200, *secondary_300)",
            button_primary_text_color="white",
            button_primary_background_fill_dark="linear-gradient(90deg, *primary_600, *secondary_800)",
            block_shadow="*shadow_drop_lg",
            button_shadow="*shadow_drop_lg",
            input_background_fill="zinc",
            input_border_color="*secondary_300",
            input_shadow="*shadow_drop",
            input_shadow_focus="*shadow_drop_lg",
        )


seafoam = SeafoamCustom()


with gr.Blocks(theme=seafoam, analytics_enabled=False, css=css) as demo:
    with gr.Column():
        gr.Markdown(
            """ ## Testrun

            Type in the box below and click the button to generate answers to your most pressing questions!

      """
        )

        with gr.Row():
            with gr.Column(scale=3):
                instruction = gr.Textbox(
                    placeholder="Enter your question here",
                    label="Question",
                    elem_id="q-input",
                )

                with gr.Box():
                    gr.Markdown("**Answer**")
                    output = gr.Markdown(elem_id="q-output")
                submit = gr.Button("Generate", variant="primary")
                gr.Examples(
                    examples=examples,
                    inputs=[instruction],
                    # cache_examples=True,
                    cache_examples=False,
                    fn=process_example,
                    outputs=[output],
                )

    submit.click(response, inputs=[instruction], outputs=[output])
    instruction.submit(response, inputs=[instruction], outputs=[output])

# demo.queue(concurrency_count=1, max_size=5).launch(debug=False, share=True)
demo.queue(concurrency_count=1, max_size=5).launch(debug=False)