File size: 12,839 Bytes
0549051
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
"""CLI commands for Hugging Face Inference Endpoints."""

import json
from typing import Annotated, Optional

import typer

from huggingface_hub._inference_endpoints import InferenceEndpoint, InferenceEndpointScalingMetric
from huggingface_hub.errors import HfHubHTTPError

from ._cli_utils import TokenOpt, get_hf_api, typer_factory


ie_cli = typer_factory(help="Manage Hugging Face Inference Endpoints.")

catalog_app = typer_factory(help="Interact with the Inference Endpoints catalog.")

NameArg = Annotated[
    str,
    typer.Argument(help="Endpoint name."),
]
NameOpt = Annotated[
    Optional[str],
    typer.Option(help="Endpoint name."),
]

NamespaceOpt = Annotated[
    Optional[str],
    typer.Option(
        help="The namespace associated with the Inference Endpoint. Defaults to the current user's namespace.",
    ),
]


def _print_endpoint(endpoint: InferenceEndpoint) -> None:
    typer.echo(json.dumps(endpoint.raw, indent=2, sort_keys=True))


@ie_cli.command()
def ls(
    namespace: NamespaceOpt = None,
    token: TokenOpt = None,
) -> None:
    """Lists all Inference Endpoints for the given namespace."""
    api = get_hf_api(token=token)
    try:
        endpoints = api.list_inference_endpoints(namespace=namespace, token=token)
    except HfHubHTTPError as error:
        typer.echo(f"Listing failed: {error}")
        raise typer.Exit(code=error.response.status_code) from error

    typer.echo(
        json.dumps(
            {"items": [endpoint.raw for endpoint in endpoints]},
            indent=2,
            sort_keys=True,
        )
    )


@ie_cli.command(name="deploy")
def deploy(
    name: NameArg,
    repo: Annotated[
        str,
        typer.Option(
            help="The name of the model repository associated with the Inference Endpoint (e.g. 'openai/gpt-oss-120b').",
        ),
    ],
    framework: Annotated[
        str,
        typer.Option(
            help="The machine learning framework used for the model (e.g. 'vllm').",
        ),
    ],
    accelerator: Annotated[
        str,
        typer.Option(
            help="The hardware accelerator to be used for inference (e.g. 'cpu').",
        ),
    ],
    instance_size: Annotated[
        str,
        typer.Option(
            help="The size or type of the instance to be used for hosting the model (e.g. 'x4').",
        ),
    ],
    instance_type: Annotated[
        str,
        typer.Option(
            help="The cloud instance type where the Inference Endpoint will be deployed (e.g. 'intel-icl').",
        ),
    ],
    region: Annotated[
        str,
        typer.Option(
            help="The cloud region in which the Inference Endpoint will be created (e.g. 'us-east-1').",
        ),
    ],
    vendor: Annotated[
        str,
        typer.Option(
            help="The cloud provider or vendor where the Inference Endpoint will be hosted (e.g. 'aws').",
        ),
    ],
    *,
    namespace: NamespaceOpt = None,
    task: Annotated[
        Optional[str],
        typer.Option(
            help="The task on which to deploy the model (e.g. 'text-classification').",
        ),
    ] = None,
    token: TokenOpt = None,
    min_replica: Annotated[
        int,
        typer.Option(
            help="The minimum number of replicas (instances) to keep running for the Inference Endpoint.",
        ),
    ] = 1,
    max_replica: Annotated[
        int,
        typer.Option(
            help="The maximum number of replicas (instances) to scale to for the Inference Endpoint.",
        ),
    ] = 1,
    scale_to_zero_timeout: Annotated[
        Optional[int],
        typer.Option(
            help="The duration in minutes before an inactive endpoint is scaled to zero.",
        ),
    ] = None,
    scaling_metric: Annotated[
        Optional[InferenceEndpointScalingMetric],
        typer.Option(
            help="The metric reference for scaling.",
        ),
    ] = None,
    scaling_threshold: Annotated[
        Optional[float],
        typer.Option(
            help="The scaling metric threshold used to trigger a scale up. Ignored when scaling metric is not provided.",
        ),
    ] = None,
) -> None:
    """Deploy an Inference Endpoint from a Hub repository."""
    api = get_hf_api(token=token)
    endpoint = api.create_inference_endpoint(
        name=name,
        repository=repo,
        framework=framework,
        accelerator=accelerator,
        instance_size=instance_size,
        instance_type=instance_type,
        region=region,
        vendor=vendor,
        namespace=namespace,
        task=task,
        token=token,
        min_replica=min_replica,
        max_replica=max_replica,
        scaling_metric=scaling_metric,
        scaling_threshold=scaling_threshold,
        scale_to_zero_timeout=scale_to_zero_timeout,
    )

    _print_endpoint(endpoint)


@catalog_app.command(name="deploy")
def deploy_from_catalog(
    repo: Annotated[
        str,
        typer.Option(
            help="The name of the model repository associated with the Inference Endpoint (e.g. 'openai/gpt-oss-120b').",
        ),
    ],
    name: NameOpt = None,
    namespace: NamespaceOpt = None,
    token: TokenOpt = None,
) -> None:
    """Deploy an Inference Endpoint from the Model Catalog."""
    api = get_hf_api(token=token)
    try:
        endpoint = api.create_inference_endpoint_from_catalog(
            repo_id=repo,
            name=name,
            namespace=namespace,
            token=token,
        )
    except HfHubHTTPError as error:
        typer.echo(f"Deployment failed: {error}")
        raise typer.Exit(code=error.response.status_code) from error

    _print_endpoint(endpoint)


def list_catalog(
    token: TokenOpt = None,
) -> None:
    """List available Catalog models."""
    api = get_hf_api(token=token)
    try:
        models = api.list_inference_catalog(token=token)
    except HfHubHTTPError as error:
        typer.echo(f"Catalog fetch failed: {error}")
        raise typer.Exit(code=error.response.status_code) from error

    typer.echo(json.dumps({"models": models}, indent=2, sort_keys=True))


catalog_app.command(name="ls")(list_catalog)
ie_cli.command(name="list-catalog", help="List available Catalog models.", hidden=True)(list_catalog)


ie_cli.add_typer(catalog_app, name="catalog")


@ie_cli.command()
def describe(
    name: NameArg,
    namespace: NamespaceOpt = None,
    token: TokenOpt = None,
) -> None:
    """Get information about an existing endpoint."""
    api = get_hf_api(token=token)
    try:
        endpoint = api.get_inference_endpoint(name=name, namespace=namespace, token=token)
    except HfHubHTTPError as error:
        typer.echo(f"Fetch failed: {error}")
        raise typer.Exit(code=error.response.status_code) from error

    _print_endpoint(endpoint)


@ie_cli.command()
def update(
    name: NameArg,
    namespace: NamespaceOpt = None,
    repo: Annotated[
        Optional[str],
        typer.Option(
            help="The name of the model repository associated with the Inference Endpoint (e.g. 'openai/gpt-oss-120b').",
        ),
    ] = None,
    accelerator: Annotated[
        Optional[str],
        typer.Option(
            help="The hardware accelerator to be used for inference (e.g. 'cpu').",
        ),
    ] = None,
    instance_size: Annotated[
        Optional[str],
        typer.Option(
            help="The size or type of the instance to be used for hosting the model (e.g. 'x4').",
        ),
    ] = None,
    instance_type: Annotated[
        Optional[str],
        typer.Option(
            help="The cloud instance type where the Inference Endpoint will be deployed (e.g. 'intel-icl').",
        ),
    ] = None,
    framework: Annotated[
        Optional[str],
        typer.Option(
            help="The machine learning framework used for the model (e.g. 'custom').",
        ),
    ] = None,
    revision: Annotated[
        Optional[str],
        typer.Option(
            help="The specific model revision to deploy on the Inference Endpoint (e.g. '6c0e6080953db56375760c0471a8c5f2929baf11').",
        ),
    ] = None,
    task: Annotated[
        Optional[str],
        typer.Option(
            help="The task on which to deploy the model (e.g. 'text-classification').",
        ),
    ] = None,
    min_replica: Annotated[
        Optional[int],
        typer.Option(
            help="The minimum number of replicas (instances) to keep running for the Inference Endpoint.",
        ),
    ] = None,
    max_replica: Annotated[
        Optional[int],
        typer.Option(
            help="The maximum number of replicas (instances) to scale to for the Inference Endpoint.",
        ),
    ] = None,
    scale_to_zero_timeout: Annotated[
        Optional[int],
        typer.Option(
            help="The duration in minutes before an inactive endpoint is scaled to zero.",
        ),
    ] = None,
    scaling_metric: Annotated[
        Optional[InferenceEndpointScalingMetric],
        typer.Option(
            help="The metric reference for scaling.",
        ),
    ] = None,
    scaling_threshold: Annotated[
        Optional[float],
        typer.Option(
            help="The scaling metric threshold used to trigger a scale up. Ignored when scaling metric is not provided.",
        ),
    ] = None,
    token: TokenOpt = None,
) -> None:
    """Update an existing endpoint."""
    api = get_hf_api(token=token)
    try:
        endpoint = api.update_inference_endpoint(
            name=name,
            namespace=namespace,
            repository=repo,
            framework=framework,
            revision=revision,
            task=task,
            accelerator=accelerator,
            instance_size=instance_size,
            instance_type=instance_type,
            min_replica=min_replica,
            max_replica=max_replica,
            scale_to_zero_timeout=scale_to_zero_timeout,
            scaling_metric=scaling_metric,
            scaling_threshold=scaling_threshold,
            token=token,
        )
    except HfHubHTTPError as error:
        typer.echo(f"Update failed: {error}")
        raise typer.Exit(code=error.response.status_code) from error
    _print_endpoint(endpoint)


@ie_cli.command()
def delete(
    name: NameArg,
    namespace: NamespaceOpt = None,
    yes: Annotated[
        bool,
        typer.Option("--yes", help="Skip confirmation prompts."),
    ] = False,
    token: TokenOpt = None,
) -> None:
    """Delete an Inference Endpoint permanently."""
    if not yes:
        confirmation = typer.prompt(f"Delete endpoint '{name}'? Type the name to confirm.")
        if confirmation != name:
            typer.echo("Aborted.")
            raise typer.Exit(code=2)

    api = get_hf_api(token=token)
    try:
        api.delete_inference_endpoint(name=name, namespace=namespace, token=token)
    except HfHubHTTPError as error:
        typer.echo(f"Delete failed: {error}")
        raise typer.Exit(code=error.response.status_code) from error

    typer.echo(f"Deleted '{name}'.")


@ie_cli.command()
def pause(
    name: NameArg,
    namespace: NamespaceOpt = None,
    token: TokenOpt = None,
) -> None:
    """Pause an Inference Endpoint."""
    api = get_hf_api(token=token)
    try:
        endpoint = api.pause_inference_endpoint(name=name, namespace=namespace, token=token)
    except HfHubHTTPError as error:
        typer.echo(f"Pause failed: {error}")
        raise typer.Exit(code=error.response.status_code) from error

    _print_endpoint(endpoint)


@ie_cli.command()
def resume(
    name: NameArg,
    namespace: NamespaceOpt = None,
    fail_if_already_running: Annotated[
        bool,
        typer.Option(
            "--fail-if-already-running",
            help="If `True`, the method will raise an error if the Inference Endpoint is already running.",
        ),
    ] = False,
    token: TokenOpt = None,
) -> None:
    """Resume an Inference Endpoint."""
    api = get_hf_api(token=token)
    try:
        endpoint = api.resume_inference_endpoint(
            name=name,
            namespace=namespace,
            token=token,
            running_ok=not fail_if_already_running,
        )
    except HfHubHTTPError as error:
        typer.echo(f"Resume failed: {error}")
        raise typer.Exit(code=error.response.status_code) from error
    _print_endpoint(endpoint)


@ie_cli.command()
def scale_to_zero(
    name: NameArg,
    namespace: NamespaceOpt = None,
    token: TokenOpt = None,
) -> None:
    """Scale an Inference Endpoint to zero."""
    api = get_hf_api(token=token)
    try:
        endpoint = api.scale_to_zero_inference_endpoint(name=name, namespace=namespace, token=token)
    except HfHubHTTPError as error:
        typer.echo(f"Scale To Zero failed: {error}")
        raise typer.Exit(code=error.response.status_code) from error

    _print_endpoint(endpoint)