File size: 6,172 Bytes
73e0168
f045267
73e0168
f045267
6017ce1
73e0168
 
f045267
73e0168
f045267
73e0168
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f045267
 
bdb8322
73e0168
 
 
 
 
 
 
 
 
f045267
 
73e0168
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f045267
 
73e0168
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
import re
import subprocess
import yaml

import gradio as gr
import requests
from huggingface_hub import HfApi

CMD = ["python" ,"run.py"]

with open("README.md") as f:
    METADATA = yaml.safe_load(f.read().split("---\n")[1])
TITLE = METADATA["title"]
EMOJI = METADATA["emoji"]

try:
    process = subprocess.run(CMD + ["--help"], stdout=subprocess.PIPE, stderr=subprocess.PIPE)
    HELP = not process.returncode and (process.stdout or process.stderr).decode()
except Exception:
    HELP = False

DRY_RUN = bool(HELP) and bool(m :=re.search("--dry(-|_)run", HELP)) and m.group(0)

def update_pbars(pbars: dict[str, float], line: str):
    if (percent_match := re.search("\\d+(?:\\.\\d+)?%", line)) and any(c in line.split("%")[1][:10] for c in "|█▌"):
        [pbars.pop(desc) for desc, percent in pbars.items() if percent == 1.]
        percent = float(percent_match.group(0)[:-1]) / 100
        desc = line[:percent_match.start()].strip() or "Progress"
        pbars[desc] = percent

def dry_run(src, config, split, dst, query):
    if not all([src, config, split, dst, query]):
        raise gr.Error("Please fill source, destination and query.")
    process = subprocess.Popen(CMD + ["--src", src, "--config", config, "--split", split, "--dst", dst, "--query", query, DRY_RUN], stdout=subprocess.PIPE)
    logs = ""
    for line in iter(process.stdout.readline, b""):
        logs += line.decode()
        yield {output_markdown: logs, progress_labels: gr.Label(visible=False)}

def run(src, config, split, dst, query):
    if not all([src, config, split, dst, query]):
        raise gr.Error("Please fill source, destination and query.")
    raise gr.Error("NotImplemented")

READ_FUNCTIONS = ("pl.read_parquet", "pl.read_csv", "pl.read_json")
NUM_TRENDING_DATASETS = 10

with gr.Blocks() as demo:
    with gr.Row():
        with gr.Column(scale=10):
            gr.Markdown(f"# {TITLE} {EMOJI}")
        with gr.Column():
            gr.LoginButton(scale=0.1)
    with gr.Row():
        with gr.Column():
            with gr.Row():
                loading_codes_json = gr.JSON([], visible=False)
                dataset_dropdown = gr.Dropdown(label="Source Dataset", allow_custom_value=True, scale=10)
                subset_dropdown = gr.Dropdown(info="Subset", allow_custom_value=True, show_label=False, visible=False)
                split_dropdown = gr.Dropdown(info="Split", allow_custom_value=True, show_label=False, visible=False)
        with gr.Column(scale=0.1, min_width=60):
            gr.HTML("<div style='font-size: 4em;'>→</div>")
        with gr.Column():
            dst_dropdown = gr.Dropdown(label="Destination Dataset", allow_custom_value=True)
    query_textarea = gr.TextArea(label="SQL Query", placeholder="SELECT * FROM src;", value="SELECT * FROM src;", container=False, show_label=False)
    with gr.Row():
        run_button = gr.Button("Run", scale=10, variant="primary")
        if DRY_RUN:
            dry_run_button = gr.Button("Dry-Run")
    progress_labels= gr.Label(visible=False, label="Progress")
    output_markdown = gr.Markdown(label="Output logs")
    run_button.click(run, inputs=[dataset_dropdown, subset_dropdown, split_dropdown, dst_dropdown, query_textarea], outputs=[progress_labels, output_markdown])
    if DRY_RUN:
        dry_run_button.click(dry_run, inputs=[dataset_dropdown, subset_dropdown, split_dropdown, dst_dropdown, query_textarea], outputs=[progress_labels, output_markdown])

    def show_subset_dropdown(dataset: str):
        if dataset and "/" not in dataset.strip().strip("/"):
            return []
        resp = requests.get(f"https://datasets-server.huggingface.co/compatible-libraries?dataset={dataset}", timeout=3).json()
        loading_codes = ([lib["loading_codes"] for lib in resp.get("libraries", []) if lib["function"] in READ_FUNCTIONS] or [[]])[0] or []
        subsets = [loading_code["config_name"] for loading_code in loading_codes]
        subset = (subsets or [""])[0]
        return dict(choices=subsets, value=subset, visible=len(subsets) > 1, key=hash(str(loading_codes))), loading_codes

    def show_split_dropdown(subset: str, loading_codes: list[dict]):
        splits = ([list(loading_code["arguments"]["splits"]) for loading_code in loading_codes if loading_code["config_name"] == subset] or [[]])[0]
        split = (splits or [""])[0]
        return dict(choices=splits, value=split, visible=len(splits) > 1, key=hash(str(loading_codes) + subset))

    @demo.load(outputs=[dataset_dropdown, loading_codes_json, subset_dropdown, split_dropdown])
    def _fetch_datasets(request: gr.Request):
        dataset = "CohereForAI/Global-MMLU"
        datasets = [dataset] + [ds.id for ds in HfApi().list_datasets(limit=NUM_TRENDING_DATASETS, sort="trendingScore", direction=-1) if ds.id != dataset]
        subsets, loading_codes = show_subset_dropdown(dataset)
        splits = show_split_dropdown(subsets["value"], loading_codes)
        return {
            dataset_dropdown: gr.Dropdown(choices=datasets, value=dataset),
            loading_codes_json: loading_codes,
            subset_dropdown: gr.Dropdown(**subsets),
            split_dropdown: gr.Dropdown(**splits),
        }
    
    @dataset_dropdown.select(inputs=[dataset_dropdown], outputs=[subset_dropdown, split_dropdown])
    def _show_subset_dropdown(dataset: str):
        subsets, loading_codes = show_subset_dropdown(dataset)
        splits = show_split_dropdown(subsets["value"], loading_codes)
        return {
            subset_dropdown: gr.Dropdown(**subsets),
            split_dropdown: gr.Dropdown(**splits),
        }
    
    @subset_dropdown.select(inputs=[dataset_dropdown, subset_dropdown, loading_codes_json], outputs=[split_dropdown])
    def _show_split_dropdown(dataset: str, subset: str, loading_codes: list[dict]):
        splits = show_split_dropdown(subset, loading_codes)
        return {
            split_dropdown: gr.Dropdown(**splits),
        }

if HELP:
    with demo.route("Help", "/help"):
        gr.Markdown(f"# Help\n\n```\n{HELP}\n```")

with demo.route("Jobs", "/jobs"):
    gr.Markdown("# Jobs")

if __name__ == "__main__":
    demo.launch(server_name="0.0.0.0")