xinference 1.0.1__py3-none-any.whl → 1.1.1__py3-none-any.whl
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Potentially problematic release.
This version of xinference might be problematic. Click here for more details.
- xinference/_compat.py +2 -0
- xinference/_version.py +3 -3
- xinference/api/restful_api.py +28 -6
- xinference/core/utils.py +10 -6
- xinference/deploy/cmdline.py +3 -1
- xinference/deploy/test/test_cmdline.py +56 -0
- xinference/isolation.py +24 -0
- xinference/model/audio/core.py +10 -0
- xinference/model/audio/cosyvoice.py +25 -3
- xinference/model/audio/f5tts.py +200 -0
- xinference/model/audio/f5tts_mlx.py +260 -0
- xinference/model/audio/fish_speech.py +36 -111
- xinference/model/audio/model_spec.json +27 -3
- xinference/model/audio/model_spec_modelscope.json +18 -0
- xinference/model/audio/utils.py +32 -0
- xinference/model/embedding/core.py +203 -142
- xinference/model/embedding/model_spec.json +7 -0
- xinference/model/embedding/model_spec_modelscope.json +8 -0
- xinference/model/image/core.py +69 -1
- xinference/model/image/model_spec.json +127 -4
- xinference/model/image/model_spec_modelscope.json +130 -4
- xinference/model/image/stable_diffusion/core.py +45 -13
- xinference/model/llm/__init__.py +2 -2
- xinference/model/llm/llm_family.json +219 -53
- xinference/model/llm/llm_family.py +15 -36
- xinference/model/llm/llm_family_modelscope.json +167 -20
- xinference/model/llm/mlx/core.py +287 -51
- xinference/model/llm/sglang/core.py +1 -0
- xinference/model/llm/transformers/chatglm.py +9 -5
- xinference/model/llm/transformers/core.py +1 -0
- xinference/model/llm/transformers/qwen2_vl.py +2 -0
- xinference/model/llm/transformers/utils.py +16 -8
- xinference/model/llm/utils.py +5 -1
- xinference/model/llm/vllm/core.py +16 -2
- xinference/thirdparty/cosyvoice/bin/average_model.py +92 -0
- xinference/thirdparty/cosyvoice/bin/export_jit.py +12 -2
- xinference/thirdparty/cosyvoice/bin/export_onnx.py +112 -0
- xinference/thirdparty/cosyvoice/bin/export_trt.sh +9 -0
- xinference/thirdparty/cosyvoice/bin/inference.py +5 -7
- xinference/thirdparty/cosyvoice/bin/train.py +42 -8
- xinference/thirdparty/cosyvoice/cli/cosyvoice.py +96 -25
- xinference/thirdparty/cosyvoice/cli/frontend.py +77 -30
- xinference/thirdparty/cosyvoice/cli/model.py +330 -80
- xinference/thirdparty/cosyvoice/dataset/dataset.py +6 -2
- xinference/thirdparty/cosyvoice/dataset/processor.py +76 -14
- xinference/thirdparty/cosyvoice/flow/decoder.py +92 -13
- xinference/thirdparty/cosyvoice/flow/flow.py +99 -9
- xinference/thirdparty/cosyvoice/flow/flow_matching.py +110 -13
- xinference/thirdparty/cosyvoice/flow/length_regulator.py +5 -4
- xinference/thirdparty/cosyvoice/hifigan/discriminator.py +140 -0
- xinference/thirdparty/cosyvoice/hifigan/generator.py +58 -42
- xinference/thirdparty/cosyvoice/hifigan/hifigan.py +67 -0
- xinference/thirdparty/cosyvoice/llm/llm.py +139 -6
- xinference/thirdparty/cosyvoice/tokenizer/assets/multilingual_zh_ja_yue_char_del.tiktoken +58836 -0
- xinference/thirdparty/cosyvoice/tokenizer/tokenizer.py +279 -0
- xinference/thirdparty/cosyvoice/transformer/embedding.py +2 -2
- xinference/thirdparty/cosyvoice/transformer/encoder_layer.py +7 -7
- xinference/thirdparty/cosyvoice/transformer/upsample_encoder.py +318 -0
- xinference/thirdparty/cosyvoice/utils/common.py +28 -1
- xinference/thirdparty/cosyvoice/utils/executor.py +69 -7
- xinference/thirdparty/cosyvoice/utils/file_utils.py +2 -12
- xinference/thirdparty/cosyvoice/utils/frontend_utils.py +9 -5
- xinference/thirdparty/cosyvoice/utils/losses.py +20 -0
- xinference/thirdparty/cosyvoice/utils/scheduler.py +1 -2
- xinference/thirdparty/cosyvoice/utils/train_utils.py +101 -45
- xinference/thirdparty/f5_tts/api.py +166 -0
- xinference/thirdparty/f5_tts/configs/E2TTS_Base_train.yaml +44 -0
- xinference/thirdparty/f5_tts/configs/E2TTS_Small_train.yaml +44 -0
- xinference/thirdparty/f5_tts/configs/F5TTS_Base_train.yaml +46 -0
- xinference/thirdparty/f5_tts/configs/F5TTS_Small_train.yaml +46 -0
- xinference/thirdparty/f5_tts/eval/README.md +49 -0
- xinference/thirdparty/f5_tts/eval/ecapa_tdnn.py +330 -0
- xinference/thirdparty/f5_tts/eval/eval_infer_batch.py +207 -0
- xinference/thirdparty/f5_tts/eval/eval_infer_batch.sh +13 -0
- xinference/thirdparty/f5_tts/eval/eval_librispeech_test_clean.py +84 -0
- xinference/thirdparty/f5_tts/eval/eval_seedtts_testset.py +84 -0
- xinference/thirdparty/f5_tts/eval/utils_eval.py +405 -0
- xinference/thirdparty/f5_tts/infer/README.md +191 -0
- xinference/thirdparty/f5_tts/infer/SHARED.md +74 -0
- xinference/thirdparty/f5_tts/infer/examples/basic/basic.toml +11 -0
- xinference/thirdparty/f5_tts/infer/examples/basic/basic_ref_en.wav +0 -0
- xinference/thirdparty/f5_tts/infer/examples/basic/basic_ref_zh.wav +0 -0
- xinference/thirdparty/f5_tts/infer/examples/multi/country.flac +0 -0
- xinference/thirdparty/f5_tts/infer/examples/multi/main.flac +0 -0
- xinference/thirdparty/f5_tts/infer/examples/multi/story.toml +19 -0
- xinference/thirdparty/f5_tts/infer/examples/multi/story.txt +1 -0
- xinference/thirdparty/f5_tts/infer/examples/multi/town.flac +0 -0
- xinference/thirdparty/f5_tts/infer/examples/vocab.txt +2545 -0
- xinference/thirdparty/f5_tts/infer/infer_cli.py +226 -0
- xinference/thirdparty/f5_tts/infer/infer_gradio.py +851 -0
- xinference/thirdparty/f5_tts/infer/speech_edit.py +193 -0
- xinference/thirdparty/f5_tts/infer/utils_infer.py +538 -0
- xinference/thirdparty/f5_tts/model/__init__.py +10 -0
- xinference/thirdparty/f5_tts/model/backbones/README.md +20 -0
- xinference/thirdparty/f5_tts/model/backbones/dit.py +163 -0
- xinference/thirdparty/f5_tts/model/backbones/mmdit.py +146 -0
- xinference/thirdparty/f5_tts/model/backbones/unett.py +219 -0
- xinference/thirdparty/f5_tts/model/cfm.py +285 -0
- xinference/thirdparty/f5_tts/model/dataset.py +319 -0
- xinference/thirdparty/f5_tts/model/modules.py +658 -0
- xinference/thirdparty/f5_tts/model/trainer.py +366 -0
- xinference/thirdparty/f5_tts/model/utils.py +185 -0
- xinference/thirdparty/f5_tts/scripts/count_max_epoch.py +33 -0
- xinference/thirdparty/f5_tts/scripts/count_params_gflops.py +39 -0
- xinference/thirdparty/f5_tts/socket_server.py +159 -0
- xinference/thirdparty/f5_tts/train/README.md +77 -0
- xinference/thirdparty/f5_tts/train/datasets/prepare_csv_wavs.py +139 -0
- xinference/thirdparty/f5_tts/train/datasets/prepare_emilia.py +230 -0
- xinference/thirdparty/f5_tts/train/datasets/prepare_libritts.py +92 -0
- xinference/thirdparty/f5_tts/train/datasets/prepare_ljspeech.py +65 -0
- xinference/thirdparty/f5_tts/train/datasets/prepare_wenetspeech4tts.py +125 -0
- xinference/thirdparty/f5_tts/train/finetune_cli.py +174 -0
- xinference/thirdparty/f5_tts/train/finetune_gradio.py +1846 -0
- xinference/thirdparty/f5_tts/train/train.py +75 -0
- xinference/thirdparty/fish_speech/fish_speech/conversation.py +94 -83
- xinference/thirdparty/fish_speech/fish_speech/models/text2semantic/llama.py +63 -20
- xinference/thirdparty/fish_speech/fish_speech/text/clean.py +1 -26
- xinference/thirdparty/fish_speech/fish_speech/text/spliter.py +1 -1
- xinference/thirdparty/fish_speech/fish_speech/tokenizer.py +152 -0
- xinference/thirdparty/fish_speech/fish_speech/train.py +2 -2
- xinference/thirdparty/fish_speech/fish_speech/webui/manage.py +1 -1
- xinference/thirdparty/fish_speech/tools/{post_api.py → api_client.py} +7 -13
- xinference/thirdparty/fish_speech/tools/api_server.py +98 -0
- xinference/thirdparty/fish_speech/tools/download_models.py +5 -5
- xinference/thirdparty/fish_speech/tools/fish_e2e.py +2 -2
- xinference/thirdparty/fish_speech/tools/inference_engine/__init__.py +192 -0
- xinference/thirdparty/fish_speech/tools/inference_engine/reference_loader.py +125 -0
- xinference/thirdparty/fish_speech/tools/inference_engine/utils.py +39 -0
- xinference/thirdparty/fish_speech/tools/inference_engine/vq_manager.py +57 -0
- xinference/thirdparty/fish_speech/tools/llama/eval_in_context.py +2 -2
- xinference/thirdparty/fish_speech/tools/llama/generate.py +117 -89
- xinference/thirdparty/fish_speech/tools/run_webui.py +104 -0
- xinference/thirdparty/fish_speech/tools/schema.py +11 -28
- xinference/thirdparty/fish_speech/tools/server/agent/__init__.py +57 -0
- xinference/thirdparty/fish_speech/tools/server/agent/generate.py +119 -0
- xinference/thirdparty/fish_speech/tools/server/agent/generation_utils.py +122 -0
- xinference/thirdparty/fish_speech/tools/server/agent/pre_generation_utils.py +72 -0
- xinference/thirdparty/fish_speech/tools/server/api_utils.py +75 -0
- xinference/thirdparty/fish_speech/tools/server/exception_handler.py +27 -0
- xinference/thirdparty/fish_speech/tools/server/inference.py +45 -0
- xinference/thirdparty/fish_speech/tools/server/model_manager.py +122 -0
- xinference/thirdparty/fish_speech/tools/server/model_utils.py +129 -0
- xinference/thirdparty/fish_speech/tools/server/views.py +246 -0
- xinference/thirdparty/fish_speech/tools/webui/__init__.py +173 -0
- xinference/thirdparty/fish_speech/tools/webui/inference.py +91 -0
- xinference/thirdparty/fish_speech/tools/webui/variables.py +14 -0
- xinference/thirdparty/matcha/utils/utils.py +2 -2
- xinference/web/ui/build/asset-manifest.json +3 -3
- xinference/web/ui/build/index.html +1 -1
- xinference/web/ui/build/static/js/{main.2f269bb3.js → main.4eb4ee80.js} +3 -3
- xinference/web/ui/build/static/js/main.4eb4ee80.js.map +1 -0
- xinference/web/ui/node_modules/.cache/babel-loader/8c5eeb02f772d02cbe8b89c05428d0dd41a97866f75f7dc1c2164a67f5a1cf98.json +1 -0
- {xinference-1.0.1.dist-info → xinference-1.1.1.dist-info}/METADATA +41 -17
- {xinference-1.0.1.dist-info → xinference-1.1.1.dist-info}/RECORD +160 -88
- xinference/thirdparty/cosyvoice/bin/export_trt.py +0 -8
- xinference/thirdparty/cosyvoice/flow/__init__.py +0 -0
- xinference/thirdparty/cosyvoice/hifigan/__init__.py +0 -0
- xinference/thirdparty/cosyvoice/llm/__init__.py +0 -0
- xinference/thirdparty/fish_speech/tools/__init__.py +0 -0
- xinference/thirdparty/fish_speech/tools/api.py +0 -943
- xinference/thirdparty/fish_speech/tools/msgpack_api.py +0 -95
- xinference/thirdparty/fish_speech/tools/webui.py +0 -548
- xinference/web/ui/build/static/js/main.2f269bb3.js.map +0 -1
- xinference/web/ui/node_modules/.cache/babel-loader/bd6ad8159341315a1764c397621a560809f7eb7219ab5174c801fca7e969d943.json +0 -1
- /xinference/thirdparty/{cosyvoice/bin → f5_tts}/__init__.py +0 -0
- /xinference/web/ui/build/static/js/{main.2f269bb3.js.LICENSE.txt → main.4eb4ee80.js.LICENSE.txt} +0 -0
- {xinference-1.0.1.dist-info → xinference-1.1.1.dist-info}/LICENSE +0 -0
- {xinference-1.0.1.dist-info → xinference-1.1.1.dist-info}/WHEEL +0 -0
- {xinference-1.0.1.dist-info → xinference-1.1.1.dist-info}/entry_points.txt +0 -0
- {xinference-1.0.1.dist-info → xinference-1.1.1.dist-info}/top_level.txt +0 -0
|
@@ -0,0 +1,91 @@
|
|
|
1
|
+
import html
|
|
2
|
+
from functools import partial
|
|
3
|
+
from typing import Any, Callable
|
|
4
|
+
|
|
5
|
+
from fish_speech.i18n import i18n
|
|
6
|
+
from tools.schema import ServeReferenceAudio, ServeTTSRequest
|
|
7
|
+
|
|
8
|
+
|
|
9
|
+
def inference_wrapper(
|
|
10
|
+
text,
|
|
11
|
+
normalize,
|
|
12
|
+
reference_id,
|
|
13
|
+
reference_audio,
|
|
14
|
+
reference_text,
|
|
15
|
+
max_new_tokens,
|
|
16
|
+
chunk_length,
|
|
17
|
+
top_p,
|
|
18
|
+
repetition_penalty,
|
|
19
|
+
temperature,
|
|
20
|
+
seed,
|
|
21
|
+
use_memory_cache,
|
|
22
|
+
engine,
|
|
23
|
+
):
|
|
24
|
+
"""
|
|
25
|
+
Wrapper for the inference function.
|
|
26
|
+
Used in the Gradio interface.
|
|
27
|
+
"""
|
|
28
|
+
|
|
29
|
+
if reference_audio:
|
|
30
|
+
references = get_reference_audio(reference_audio, reference_text)
|
|
31
|
+
else:
|
|
32
|
+
references = []
|
|
33
|
+
|
|
34
|
+
req = ServeTTSRequest(
|
|
35
|
+
text=text,
|
|
36
|
+
normalize=normalize,
|
|
37
|
+
reference_id=reference_id if reference_id else None,
|
|
38
|
+
references=references,
|
|
39
|
+
max_new_tokens=max_new_tokens,
|
|
40
|
+
chunk_length=chunk_length,
|
|
41
|
+
top_p=top_p,
|
|
42
|
+
repetition_penalty=repetition_penalty,
|
|
43
|
+
temperature=temperature,
|
|
44
|
+
seed=int(seed) if seed else None,
|
|
45
|
+
use_memory_cache=use_memory_cache,
|
|
46
|
+
)
|
|
47
|
+
|
|
48
|
+
for result in engine.inference(req):
|
|
49
|
+
match result.code:
|
|
50
|
+
case "final":
|
|
51
|
+
return result.audio, None
|
|
52
|
+
case "error":
|
|
53
|
+
return None, build_html_error_message(i18n(result.error))
|
|
54
|
+
case _:
|
|
55
|
+
pass
|
|
56
|
+
|
|
57
|
+
return None, i18n("No audio generated")
|
|
58
|
+
|
|
59
|
+
|
|
60
|
+
def get_reference_audio(reference_audio: str, reference_text: str) -> list:
|
|
61
|
+
"""
|
|
62
|
+
Get the reference audio bytes.
|
|
63
|
+
"""
|
|
64
|
+
|
|
65
|
+
with open(reference_audio, "rb") as audio_file:
|
|
66
|
+
audio_bytes = audio_file.read()
|
|
67
|
+
|
|
68
|
+
return [ServeReferenceAudio(audio=audio_bytes, text=reference_text)]
|
|
69
|
+
|
|
70
|
+
|
|
71
|
+
def build_html_error_message(error: Any) -> str:
|
|
72
|
+
|
|
73
|
+
error = error if isinstance(error, Exception) else Exception("Unknown error")
|
|
74
|
+
|
|
75
|
+
return f"""
|
|
76
|
+
<div style="color: red;
|
|
77
|
+
font-weight: bold;">
|
|
78
|
+
{html.escape(str(error))}
|
|
79
|
+
</div>
|
|
80
|
+
"""
|
|
81
|
+
|
|
82
|
+
|
|
83
|
+
def get_inference_wrapper(engine) -> Callable:
|
|
84
|
+
"""
|
|
85
|
+
Get the inference function with the immutable arguments.
|
|
86
|
+
"""
|
|
87
|
+
|
|
88
|
+
return partial(
|
|
89
|
+
inference_wrapper,
|
|
90
|
+
engine=engine,
|
|
91
|
+
)
|
|
@@ -0,0 +1,14 @@
|
|
|
1
|
+
from fish_speech.i18n import i18n
|
|
2
|
+
|
|
3
|
+
HEADER_MD = f"""# Fish Speech
|
|
4
|
+
|
|
5
|
+
{i18n("A text-to-speech model based on VQ-GAN and Llama developed by [Fish Audio](https://fish.audio).")}
|
|
6
|
+
|
|
7
|
+
{i18n("You can find the source code [here](https://github.com/fishaudio/fish-speech) and models [here](https://huggingface.co/fishaudio/fish-speech-1.5).")}
|
|
8
|
+
|
|
9
|
+
{i18n("Related code and weights are released under CC BY-NC-SA 4.0 License.")}
|
|
10
|
+
|
|
11
|
+
{i18n("We are not responsible for any misuse of the model, please consider your local laws and regulations before using it.")}
|
|
12
|
+
"""
|
|
13
|
+
|
|
14
|
+
TEXTBOX_PLACEHOLDER = i18n("Put your text here.")
|
|
@@ -7,10 +7,10 @@ from pathlib import Path
|
|
|
7
7
|
from typing import Any, Callable, Dict, Tuple
|
|
8
8
|
|
|
9
9
|
import gdown
|
|
10
|
-
import matplotlib.pyplot as plt
|
|
10
|
+
# import matplotlib.pyplot as plt
|
|
11
11
|
import numpy as np
|
|
12
12
|
import torch
|
|
13
|
-
import wget
|
|
13
|
+
# import wget
|
|
14
14
|
from omegaconf import DictConfig
|
|
15
15
|
|
|
16
16
|
from matcha.utils import pylogger, rich_utils
|
|
@@ -1,14 +1,14 @@
|
|
|
1
1
|
{
|
|
2
2
|
"files": {
|
|
3
3
|
"main.css": "./static/css/main.5061c4c3.css",
|
|
4
|
-
"main.js": "./static/js/main.
|
|
4
|
+
"main.js": "./static/js/main.4eb4ee80.js",
|
|
5
5
|
"static/media/icon.webp": "./static/media/icon.4603d52c63041e5dfbfd.webp",
|
|
6
6
|
"index.html": "./index.html",
|
|
7
7
|
"main.5061c4c3.css.map": "./static/css/main.5061c4c3.css.map",
|
|
8
|
-
"main.
|
|
8
|
+
"main.4eb4ee80.js.map": "./static/js/main.4eb4ee80.js.map"
|
|
9
9
|
},
|
|
10
10
|
"entrypoints": [
|
|
11
11
|
"static/css/main.5061c4c3.css",
|
|
12
|
-
"static/js/main.
|
|
12
|
+
"static/js/main.4eb4ee80.js"
|
|
13
13
|
]
|
|
14
14
|
}
|
|
@@ -1 +1 @@
|
|
|
1
|
-
<!doctype html><html lang="en"><head><meta charset="utf-8"/><link rel="icon" href="./favicon.svg"/><meta name="viewport" content="width=device-width,initial-scale=1"/><meta name="theme-color" content="#000000"/><meta name="description" content="Web site created using create-react-app"/><link rel="apple-touch-icon" href="./logo192.png"/><link rel="manifest" href="./manifest.json"/><title>Xinference</title><script defer="defer" src="./static/js/main.
|
|
1
|
+
<!doctype html><html lang="en"><head><meta charset="utf-8"/><link rel="icon" href="./favicon.svg"/><meta name="viewport" content="width=device-width,initial-scale=1"/><meta name="theme-color" content="#000000"/><meta name="description" content="Web site created using create-react-app"/><link rel="apple-touch-icon" href="./logo192.png"/><link rel="manifest" href="./manifest.json"/><title>Xinference</title><script defer="defer" src="./static/js/main.4eb4ee80.js"></script><link href="./static/css/main.5061c4c3.css" rel="stylesheet"></head><body><noscript>You need to enable JavaScript to run this app.</noscript><div id="root"></div></body></html>
|