File size: 18,447 Bytes
07423df |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 |
import gc
import logging
from typing import List
import torch
from h2o_wave import Q
from llm_studio.app_utils.sections.chat import chat_tab
from llm_studio.app_utils.sections.chat_update import chat_update
from llm_studio.app_utils.sections.common import delete_dialog
from llm_studio.app_utils.sections.dataset import (
dataset_delete_current_datasets,
dataset_delete_single,
dataset_display,
dataset_edit,
dataset_import,
dataset_import_uploaded_file,
dataset_list,
dataset_list_delete,
dataset_merge,
dataset_newexperiment,
)
from llm_studio.app_utils.sections.experiment import (
experiment_delete,
experiment_display,
experiment_download_logs,
experiment_download_model,
experiment_download_predictions,
experiment_list,
experiment_push_to_huggingface_dialog,
experiment_rename_ui_workflow,
experiment_run,
experiment_start,
experiment_stop,
)
from llm_studio.app_utils.sections.home import home
from llm_studio.app_utils.sections.project import (
current_experiment_compare,
current_experiment_list_compare,
current_experiment_list_delete,
current_experiment_list_stop,
experiment_rename_action_workflow,
list_current_experiments,
)
from llm_studio.app_utils.sections.settings import settings
from llm_studio.app_utils.setting_utils import (
load_default_user_settings,
load_user_settings_and_secrets,
save_user_settings_and_secrets,
)
from llm_studio.app_utils.utils import add_model_type
from llm_studio.app_utils.wave_utils import report_error, wave_utils_handle_error
logger = logging.getLogger(__name__)
async def handle(q: Q) -> None:
"""Handles all requests in application and calls according functions."""
# logger.info(f"args: {q.args}")
# logger.info(f"events: {q.events}")
if not (
q.args.__wave_submission_name__ == "experiment/display/chat/chatbot"
or q.args.__wave_submission_name__ == "experiment/display/chat/clear_history"
):
if "experiment/display/chat/cfg" in q.client:
del q.client["experiment/display/chat/cfg"]
if "experiment/display/chat/model" in q.client:
del q.client["experiment/display/chat/model"]
if "experiment/display/chat/tokenizer" in q.client:
del q.client["experiment/display/chat/tokenizer"]
torch.cuda.empty_cache()
gc.collect()
try:
if q.args.__wave_submission_name__ == "home":
await home(q)
elif q.args.__wave_submission_name__ == "settings":
await settings(q)
elif q.args.__wave_submission_name__ == "save_settings":
logger.info("Saving user settings")
await save_user_settings_and_secrets(q)
await settings(q)
elif q.args.__wave_submission_name__ == "load_settings":
load_user_settings_and_secrets(q)
await settings(q)
elif q.args.__wave_submission_name__ == "restore_default_settings":
load_default_user_settings(q)
await settings(q)
elif q.args.__wave_submission_name__ == "report_error":
await report_error(q)
elif q.args.__wave_submission_name__ == "dataset/import":
await dataset_import(q, step=1)
elif q.args.__wave_submission_name__ == "dataset/list":
await dataset_list(q)
elif q.args.__wave_submission_name__ == "dataset/list/delete/abort":
q.page["dataset/list"].items[0].table.multiple = False
await dataset_list(q, reset=True)
elif q.args.__wave_submission_name__ == "dataset/list/abort":
q.page["dataset/list"].items[0].table.multiple = False
await dataset_list(q, reset=True)
elif q.args.__wave_submission_name__ == "dataset/list/delete":
await dataset_list_delete(q)
elif q.args.__wave_submission_name__ == "dataset/delete/single":
dataset_id = q.client["dataset/delete/single/id"]
dataset_id = q.client["dataset/list/df_datasets"]["id"].iloc[dataset_id]
await dataset_delete_single(q, int(dataset_id))
elif q.args.__wave_submission_name__ == "dataset/delete/dialog/single":
dataset_id = int(q.args["dataset/delete/dialog/single"])
q.client["dataset/delete/single/id"] = dataset_id
name = q.client["dataset/list/df_datasets"]["name"].iloc[dataset_id]
if q.client["delete_dialogs"]:
await delete_dialog(q, [name], "dataset/delete/single", "dataset")
else:
dataset_id = q.client["dataset/list/df_datasets"]["id"].iloc[dataset_id]
await dataset_delete_single(q, int(dataset_id))
elif q.args["dataset/delete/dialog"]:
names = list(
q.client["dataset/list/df_datasets"]["name"].iloc[
list(map(int, q.client["dataset/list/table"]))
]
)
if not names:
return
if q.client["delete_dialogs"]:
await delete_dialog(q, names, "dataset/delete", "dataset")
else:
await dataset_delete_current_datasets(q)
elif q.args.__wave_submission_name__ == "dataset/delete":
await dataset_delete_current_datasets(q)
elif q.args.__wave_submission_name__ == "dataset/edit":
if q.client["dataset/list/df_datasets"] is not None:
dataset_id = int(q.args["dataset/edit"])
dataset_id = q.client["dataset/list/df_datasets"]["id"].iloc[dataset_id]
await dataset_edit(q, int(dataset_id))
elif q.args.__wave_submission_name__ == "dataset/newexperiment":
if q.client["dataset/list/df_datasets"] is not None:
dataset_id = int(q.args["dataset/newexperiment"])
dataset_id = q.client["dataset/list/df_datasets"]["id"].iloc[dataset_id]
await dataset_newexperiment(q, int(dataset_id))
elif q.args.__wave_submission_name__ == "dataset/newexperiment/from_current":
idx = q.client["dataset/display/id"]
dataset_id = q.client["dataset/list/df_datasets"]["id"].iloc[idx]
await dataset_newexperiment(q, dataset_id)
elif q.args.__wave_submission_name__ == "dataset/list/table":
q.client["dataset/display/id"] = int(q.args["dataset/list/table"][0])
await dataset_display(q)
elif q.args.__wave_submission_name__ == "dataset/display/visualization":
await dataset_display(q)
elif q.args.__wave_submission_name__ == "dataset/display/data":
await dataset_display(q)
elif q.args.__wave_submission_name__ == "dataset/display/statistics":
await dataset_display(q)
elif q.args["dataset/display/summary"]:
await dataset_display(q)
elif (
q.args.__wave_submission_name__ == "experiment/start/run"
or q.args.__wave_submission_name__ == "experiment/start/error/proceed"
):
# add model type to cfg file name here
q.client["experiment/start/cfg_file"] = add_model_type(
q.client["experiment/start/cfg_file"],
q.client["experiment/start/cfg_sub"],
)
q.client.delete_cards.add("experiment/start")
await experiment_run(q, pre="experiment/start")
q.client["experiment/list/mode"] = "train"
elif (
q.args.__wave_submission_name__ == "experiment/start_experiment"
or q.args.__wave_submission_name__ == "experiment/list/new"
):
if q.client["experiment/list/df_experiments"] is not None:
selected_idx = int(q.args["experiment/list/new"])
experiment_id = q.client["experiment/list/df_experiments"]["id"].iloc[
selected_idx
]
q.client["experiment/start/cfg_category"] = "experiment"
q.client["experiment/start/cfg_file"] = "experiment"
q.client["experiment/start/cfg_experiment"] = str(experiment_id)
await experiment_start(q)
elif q.args.__wave_submission_name__ == "experiment/start":
q.client["experiment/start/cfg_category"] = None
q.client["experiment/start/cfg_file"] = None
datasets_df = q.client.app_db.get_datasets_df()
if datasets_df.shape[0] == 0:
info = "Import dataset before you create an experiment. "
await dataset_import(q, step=1, info=info)
else:
await experiment_start(q)
elif q.args.__wave_submission_name__ == "experiment/display/download_logs":
await experiment_download_logs(q)
elif (
q.args.__wave_submission_name__ == "experiment/display/download_predictions"
):
await experiment_download_predictions(q)
elif q.args.__wave_submission_name__ == "experiment/list":
q.client["experiment/list/mode"] = "train"
await experiment_list(q)
elif q.args.__wave_submission_name__ == "experiment/list/current":
await list_current_experiments(q)
elif q.args.__wave_submission_name__ == "experiment/list/current/noreset":
await list_current_experiments(q, reset=False)
elif q.args.__wave_submission_name__ == "experiment/list/refresh":
await experiment_list(q)
elif q.args.__wave_submission_name__ == "experiment/list/abort":
await list_current_experiments(q)
elif q.args.__wave_submission_name__ == "experiment/list/stop":
await current_experiment_list_stop(q)
elif q.args.__wave_submission_name__ == "experiment/list/delete":
await current_experiment_list_delete(q)
elif q.args.__wave_submission_name__ == "experiment/list/rename":
await experiment_rename_ui_workflow(q)
elif q.args.__wave_submission_name__ == "experiment/list/compare":
await current_experiment_list_compare(q)
elif (
q.args.__wave_submission_name__ == "experiment/stop"
or q.args.__wave_submission_name__ == "experiment/list/stop/table"
):
if q.args["experiment/list/stop/table"]:
idx = int(q.args["experiment/list/stop/table"])
selected_id = q.client["experiment/list/df_experiments"]["id"].iloc[idx]
experiment_ids = [selected_id]
else:
selected_idxs = q.client["experiment/list/table"]
experiment_ids = list(
q.client["experiment/list/df_experiments"]["id"].iloc[
list(map(int, selected_idxs))
]
)
await experiment_stop(q, experiment_ids)
await list_current_experiments(q)
elif q.args.__wave_submission_name__ == "experiment/list/delete/table/dialog":
idx = int(q.args["experiment/list/delete/table/dialog"])
names = [q.client["experiment/list/df_experiments"]["name"].iloc[idx]]
selected_id = q.client["experiment/list/df_experiments"]["id"].iloc[idx]
q.client["experiment/delete/single/id"] = selected_id
if q.client["delete_dialogs"]:
await delete_dialog(
q, names, "experiment/list/delete/table", "experiment"
)
else:
await experiment_delete_all_artifacts(q, [selected_id])
elif q.args.__wave_submission_name__ == "experiment/delete/dialog":
selected_idxs = q.client["experiment/list/table"]
exp_df = q.client["experiment/list/df_experiments"]
names = list(exp_df["name"].iloc[list(map(int, selected_idxs))])
if not names:
return
if q.client["delete_dialogs"]:
await delete_dialog(q, names, "experiment/delete", "experiment")
else:
experiment_ids = list(exp_df["id"].iloc[list(map(int, selected_idxs))])
await experiment_delete_all_artifacts(q, experiment_ids)
elif (
q.args.__wave_submission_name__ == "experiment/delete"
or q.args.__wave_submission_name__ == "experiment/list/delete/table"
):
if q.args["experiment/list/delete/table"]:
selected_id = q.client["experiment/delete/single/id"]
experiment_ids = [selected_id]
else:
selected_idxs = q.client["experiment/list/table"]
exp_df = q.client["experiment/list/df_experiments"]
experiment_ids = list(exp_df["id"].iloc[list(map(int, selected_idxs))])
await experiment_delete_all_artifacts(q, experiment_ids)
elif q.args.__wave_submission_name__ == "experiment/rename/action":
await experiment_rename_action_workflow(q)
elif q.args.__wave_submission_name__ == "experiment/compare":
await current_experiment_compare(q)
elif q.args.__wave_submission_name__ == "experiment/compare/charts":
await current_experiment_compare(q)
elif q.args.__wave_submission_name__ == "experiment/compare/config":
await current_experiment_compare(q)
elif q.args.__wave_submission_name__ == "experiment/compare/diff_toggle":
q.client["experiment/compare/diff_toggle"] = q.args[
"experiment/compare/diff_toggle"
]
await current_experiment_compare(q)
elif q.args.__wave_submission_name__ == "experiment/list/table":
q.client["experiment/display/id"] = int(q.args["experiment/list/table"][0])
q.client["experiment/display/logs_path"] = None
q.client["experiment/display/preds_path"] = None
q.client["experiment/display/tab"] = None
await experiment_display(q)
elif q.args.__wave_submission_name__ == "experiment/display/refresh":
await experiment_display(q)
elif q.args.__wave_submission_name__ == "experiment/display/charts":
await experiment_display(q)
elif q.args.__wave_submission_name__ == "experiment/display/summary":
await experiment_display(q)
elif (
q.args.__wave_submission_name__ == "experiment/display/train_data_insights"
):
await experiment_display(q)
elif (
q.args.__wave_submission_name__
== "experiment/display/validation_prediction_insights"
):
await experiment_display(q)
elif (
q.args.__wave_submission_name__ == "experiment/display/push_to_huggingface"
):
await experiment_push_to_huggingface_dialog(q)
elif q.args.__wave_submission_name__ == "experiment/display/download_model":
await experiment_download_model(q)
elif (
q.args.__wave_submission_name__
== "experiment/display/push_to_huggingface_submit"
):
await experiment_push_to_huggingface_dialog(q)
elif q.args.__wave_submission_name__ == "experiment/display/config":
await experiment_display(q)
elif q.args.__wave_submission_name__ == "experiment/display/logs":
await experiment_display(q)
elif q.args.__wave_submission_name__ == "experiment/display/chat":
await experiment_display(q)
elif q.args.__wave_submission_name__ == "experiment/display/chat/chatbot":
await chat_update(q)
elif q.args.__wave_submission_name__ == "experiment/display/chat/clear_history":
await chat_tab(q, load_model=False)
elif q.args.__wave_submission_name__ == "dataset/import/local_upload":
await dataset_import_uploaded_file(q)
elif q.args.__wave_submission_name__ == "dataset/import/local_path_list":
await dataset_import(q, step=1)
elif q.args.__wave_submission_name__ == "dataset/import/2":
await dataset_import(q, step=2)
elif q.args.__wave_submission_name__ == "dataset/import/3":
await dataset_import(q, step=3)
elif q.args.__wave_submission_name__ == "dataset/import/3/edit":
await dataset_import(q, step=3, edit=True)
elif q.args.__wave_submission_name__ == "dataset/import/4":
await dataset_import(q, step=4)
elif q.args.__wave_submission_name__ == "dataset/import/4/edit":
await dataset_import(q, step=4, edit=True)
elif q.args.__wave_submission_name__ == "dataset/import/6":
await dataset_import(q, step=6)
elif (
q.args.__wave_submission_name__ == "dataset/import/source"
and not q.args["dataset/list"]
):
await dataset_import(q, step=1)
elif q.args.__wave_submission_name__ == "dataset/merge":
await dataset_merge(q, step=1)
elif q.args.__wave_submission_name__ == "dataset/merge/action":
await dataset_merge(q, step=2)
elif q.args.__wave_submission_name__ == "dataset/import/cfg_file":
await dataset_import(q, step=3)
# leave at the end of dataset import routing,
# would also be triggered if user clicks on
# a continue button in the dataset import wizard
elif q.args.__wave_submission_name__ == "dataset/import/cfg/train_dataframe":
await dataset_import(q, step=3)
elif q.args.__wave_submission_name__ == "experiment/start/cfg_file":
q.client["experiment/start/cfg_file"] = q.args["experiment/start/cfg_file"]
await experiment_start(q)
elif q.args.__wave_submission_name__ == "experiment/start/dataset":
await experiment_start(q)
elif q.client["nav/active"] == "experiment/start":
await experiment_start(q)
except Exception as unknown_exception:
logger.error("Unknown exception", exc_info=True)
await wave_utils_handle_error(
q,
error=unknown_exception,
)
async def experiment_delete_all_artifacts(q: Q, experiment_ids: List[int]):
await experiment_stop(q, experiment_ids)
await experiment_delete(q, experiment_ids)
await list_current_experiments(q)
|