Merge branch 'main' into main

TheTerrasque · Mar 13, 2023 · 63c5a13 · 63c5a13
2 parents 683556f + 7ab45fb
commit 63c5a13
Show file tree

Hide file tree

Showing 10 changed files with 159 additions and 82 deletions.
diff --git a/.github/FUNDING.yml b/.github/FUNDING.yml
@@ -0,0 +1 @@
+ko_fi: oobabooga
diff --git a/README.md b/README.md
@@ -27,7 +27,7 @@ Its goal is to become the [AUTOMATIC1111/stable-diffusion-webui](https://github.
 * [FlexGen offload](https://github.com/oobabooga/text-generation-webui/wiki/FlexGen).
 * [DeepSpeed ZeRO-3 offload](https://github.com/oobabooga/text-generation-webui/wiki/DeepSpeed).
 * Get responses via API, [with](https://github.com/oobabooga/text-generation-webui/blob/main/api-example-streaming.py) or [without](https://github.com/oobabooga/text-generation-webui/blob/main/api-example.py) streaming.
-* [Supports the LLaMA model](https://github.com/oobabooga/text-generation-webui/wiki/LLaMA-model).
+* [Supports the LLaMA model, including 4-bit mode](https://github.com/oobabooga/text-generation-webui/wiki/LLaMA-model).
 * [Supports the RWKV model](https://github.com/oobabooga/text-generation-webui/wiki/RWKV-model).
 * Supports softprompts.
 * [Supports extensions](https://github.com/oobabooga/text-generation-webui/wiki/Extensions).
@@ -60,11 +60,13 @@ pip3 install torch torchvision torchaudio --extra-index-url https://download.pyt
 conda install pytorch torchvision torchaudio git -c pytorch
 ```
 
+See also: [Installation instructions for human beings](https://github.com/oobabooga/text-generation-webui/wiki/Installation-instructions-for-human-beings).
+
 ## Installation option 2: one-click installers
 
-[oobabooga-windows.zip](https://github.com/oobabooga/text-generation-webui/releases/download/installers/oobabooga-windows.zip)
+[oobabooga-windows.zip](https://github.com/oobabooga/one-click-installers/archive/refs/heads/oobabooga-windows.zip)
 
-[oobabooga-linux.zip](https://github.com/oobabooga/text-generation-webui/releases/download/installers/oobabooga-linux.zip)
+[oobabooga-linux.zip](https://github.com/oobabooga/one-click-installers/archive/refs/heads/oobabooga-linux.zip)
 
 Just download the zip above, extract it, and double click on "install". The web UI and all its dependencies will be installed in the same folder.
 
@@ -139,7 +141,7 @@ Optionally, you can use the following command-line flags:
 | `--cpu`       | Use the CPU to generate text.|
 | `--load-in-8bit`  | Load the model with 8-bit precision.|
 | `--load-in-4bit`  | Load the model with 4-bit precision. Currently only works with LLaMA.|
-| `--gptq-bits`  |  Load a pre-quantized model with specified precision. 2, 3, 4 and 8bit are supported. Currently only works with LLaMA. |
+| `--gptq-bits GPTQ_BITS`  |  Load a pre-quantized model with specified precision. 2, 3, 4 and 8 (bit) are supported. Currently only works with LLaMA. |
 | `--bf16`  | Load the model with bfloat16 precision. Requires NVIDIA Ampere GPU. |
 | `--auto-devices` | Automatically split the model across the available GPU(s) and CPU.|
 | `--disk` | If the model is too large for your GPU(s) and CPU combined, send the remaining layers to the disk. |
@@ -155,12 +157,13 @@ Optionally, you can use the following command-line flags:
 | `--local_rank LOCAL_RANK`    | DeepSpeed: Optional argument for distributed setups. |
 |  `--rwkv-strategy RWKV_STRATEGY`         |    RWKV: The strategy to use while loading the model. Examples: "cpu fp32", "cuda fp16", "cuda fp16i8". |
 |  `--rwkv-cuda-on`                        |   RWKV: Compile the CUDA kernel for better performance. |
-| `--no-stream`   | Don't stream the text output in real time. This improves the text generation performance.|
+| `--no-stream`   | Don't stream the text output in real time. |
 | `--settings SETTINGS_FILE` | Load the default interface settings from this json file. See `settings-template.json` for an example. If you create a file called `settings.json`, this file will be loaded by default without the need to use the `--settings` flag.|
 |  `--extensions EXTENSIONS [EXTENSIONS ...]` |  The list of extensions to load. If you want to load more than one extension, write the names separated by spaces. |
 | `--listen`   | Make the web UI reachable from your local network.|
 |  `--listen-port LISTEN_PORT` | The listening port that the server will use. |
 | `--share`   | Create a public URL. This is useful for running the web UI on Google Colab or similar. |
+| `--auto-launch` | Open the web UI in the default browser upon launch. |
 | `--verbose`   | Print the prompts to the terminal. |
 
 Out of memory errors? [Check this guide](https://github.com/oobabooga/text-generation-webui/wiki/Low-VRAM-guide).
@@ -179,14 +182,10 @@ Check the [wiki](https://github.com/oobabooga/text-generation-webui/wiki/System-
 
 Pull requests, suggestions, and issue reports are welcome.
 
-Before reporting a bug, make sure that you have created a conda environment and installed the dependencies exactly as in the *Installation* section above.
-
-These issues are known:
-
-* 8-bit doesn't work properly on Windows or older GPUs.
-* DeepSpeed doesn't work properly on Windows.
+Before reporting a bug, make sure that you have:
 
-For these two, please try commenting on an existing issue instead of creating a new one.
+1. Created a conda environment and installed the dependencies exactly as in the *Installation* section above.
+2. [Searched](https://github.com/oobabooga/text-generation-webui/issues) to see if an issue already exists for the issue you encountered.
 
 ## Credits
 

diff --git a/extensions/silero_tts/script.py b/extensions/silero_tts/script.py
@@ -1,8 +1,12 @@
+import time
 from pathlib import Path
 
 import gradio as gr
 import torch
 
+import modules.chat as chat
+import modules.shared as shared
+
 torch._C._jit_set_profiling_mode(False)
 
 params = {
@@ -12,10 +16,28 @@
     'model_id': 'v3_en',
     'sample_rate': 48000,
     'device': 'cpu',
+    'show_text': False,
+    'autoplay': True,
+    'voice_pitch': 'medium',
+    'voice_speed': 'medium',
 }
+
 current_params = params.copy()
 voices_by_gender = ['en_99', 'en_45', 'en_18', 'en_117', 'en_49', 'en_51', 'en_68', 'en_0', 'en_26', 'en_56', 'en_74', 'en_5', 'en_38', 'en_53', 'en_21', 'en_37', 'en_107', 'en_10', 'en_82', 'en_16', 'en_41', 'en_12', 'en_67', 'en_61', 'en_14', 'en_11', 'en_39', 'en_52', 'en_24', 'en_97', 'en_28', 'en_72', 'en_94', 'en_36', 'en_4', 'en_43', 'en_88', 'en_25', 'en_65', 'en_6', 'en_44', 'en_75', 'en_91', 'en_60', 'en_109', 'en_85', 'en_101', 'en_108', 'en_50', 'en_96', 'en_64', 'en_92', 'en_76', 'en_33', 'en_116', 'en_48', 'en_98', 'en_86', 'en_62', 'en_54', 'en_95', 'en_55', 'en_111', 'en_3', 'en_83', 'en_8', 'en_47', 'en_59', 'en_1', 'en_2', 'en_7', 'en_9', 'en_13', 'en_15', 'en_17', 'en_19', 'en_20', 'en_22', 'en_23', 'en_27', 'en_29', 'en_30', 'en_31', 'en_32', 'en_34', 'en_35', 'en_40', 'en_42', 'en_46', 'en_57', 'en_58', 'en_63', 'en_66', 'en_69', 'en_70', 'en_71', 'en_73', 'en_77', 'en_78', 'en_79', 'en_80', 'en_81', 'en_84', 'en_87', 'en_89', 'en_90', 'en_93', 'en_100', 'en_102', 'en_103', 'en_104', 'en_105', 'en_106', 'en_110', 'en_112', 'en_113', 'en_114', 'en_115']
-wav_idx = 0
+voice_pitches = ['x-low', 'low', 'medium', 'high', 'x-high']
+voice_speeds = ['x-slow', 'slow', 'medium', 'fast', 'x-fast']
+
+# Used for making text xml compatible, needed for voice pitch and speed control
+table = str.maketrans({
+    "<": "&lt;",
+    ">": "&gt;",
+    "&": "&amp;",
+    "'": "&apos;",
+    '"': "&quot;",
+})
+
+def xmlesc(txt):
+    return txt.translate(table)
 
 def load_model():
     model, example_text = torch.hub.load(repo_or_dir='snakers4/silero-models', model='silero_tts', language=params['language'], speaker=params['model_id'])
@@ -33,20 +55,40 @@ def remove_surrounded_chars(string):
             new_string += char
     return new_string
 
+def remove_tts_from_history(name1, name2):
+    for i, entry in enumerate(shared.history['internal']):
+        shared.history['visible'][i] = [shared.history['visible'][i][0], entry[1]]
+    return chat.generate_chat_output(shared.history['visible'], name1, name2, shared.character)
+
+def toggle_text_in_history(name1, name2):
+    for i, entry in enumerate(shared.history['visible']):
+        visible_reply = entry[1]
+        if visible_reply.startswith('<audio'):
+            if params['show_text']:
+                reply = shared.history['internal'][i][1]
+                shared.history['visible'][i] = [shared.history['visible'][i][0], f"{visible_reply.split('</audio>')[0]}</audio>\n\n{reply}"]
+            else:
+                shared.history['visible'][i] = [shared.history['visible'][i][0], f"{visible_reply.split('</audio>')[0]}</audio>"]
+    return chat.generate_chat_output(shared.history['visible'], name1, name2, shared.character)
+
 def input_modifier(string):
     """
     This function is applied to your text inputs before
     they are fed into the model.
     """
 
+    # Remove autoplay from the last reply
+    if (shared.args.chat or shared.args.cai_chat) and len(shared.history['internal']) > 0:
+        shared.history['visible'][-1] = [shared.history['visible'][-1][0], shared.history['visible'][-1][1].replace('controls autoplay>','controls>')]
+
     return string
 
 def output_modifier(string):
     """
     This function is applied to the model outputs.
     """
 
-    global wav_idx, model, current_params
+    global model, current_params
 
     for i in params:
         if params[i] != current_params[i]:
@@ -57,20 +99,25 @@ def output_modifier(string):
     if params['activate'] == False:
         return string
 
+    original_string = string
     string = remove_surrounded_chars(string)
     string = string.replace('"', '')
     string = string.replace('“', '')
     string = string.replace('\n', ' ')
     string = string.strip()
 
     if string == '':
-        string = 'empty reply, try regenerating'
-
-    output_file = Path(f'extensions/silero_tts/outputs/{wav_idx:06d}.wav')
-    model.save_wav(text=string, speaker=params['speaker'], sample_rate=int(params['sample_rate']), audio_path=str(output_file))
-
-    string = f'<audio src="file/{output_file.as_posix()}" controls></audio>'
-    wav_idx += 1
+        string = '*Empty reply, try regenerating*'
+    else:
+        output_file = Path(f'extensions/silero_tts/outputs/{shared.character}_{int(time.time())}.wav')
+        prosody = '<prosody rate="{}" pitch="{}">'.format(params['voice_speed'], params['voice_pitch'])
+        silero_input = f'<speak>{prosody}{xmlesc(string)}</prosody></speak>'
+        model.save_wav(ssml_text=silero_input, speaker=params['speaker'], sample_rate=int(params['sample_rate']), audio_path=str(output_file))
+
+        autoplay = 'autoplay' if params['autoplay'] else ''
+        string = f'<audio src="file/{output_file.as_posix()}" controls {autoplay}></audio>'
+        if params['show_text']:
+            string += f'\n\n{original_string}'
 
     return string
 
@@ -85,9 +132,36 @@ def bot_prefix_modifier(string):
 
 def ui():
     # Gradio elements
-    activate = gr.Checkbox(value=params['activate'], label='Activate TTS')
-    voice = gr.Dropdown(value=params['speaker'], choices=voices_by_gender, label='TTS voice')
+    with gr.Accordion("Silero TTS"):
+        with gr.Row():
+            activate = gr.Checkbox(value=params['activate'], label='Activate TTS')
+            autoplay = gr.Checkbox(value=params['autoplay'], label='Play TTS automatically')
+        show_text = gr.Checkbox(value=params['show_text'], label='Show message text under audio player')
+        voice = gr.Dropdown(value=params['speaker'], choices=voices_by_gender, label='TTS voice')
+        with gr.Row():
+            v_pitch = gr.Dropdown(value=params['voice_pitch'], choices=voice_pitches, label='Voice pitch')
+            v_speed = gr.Dropdown(value=params['voice_speed'], choices=voice_speeds, label='Voice speed')
+        with gr.Row():
+            convert = gr.Button('Permanently replace audios with the message texts')
+            convert_cancel = gr.Button('Cancel', visible=False)
+            convert_confirm = gr.Button('Confirm (cannot be undone)', variant="stop", visible=False)
+
+    # Convert history with confirmation
+    convert_arr = [convert_confirm, convert, convert_cancel]
+    convert.click(lambda :[gr.update(visible=True), gr.update(visible=False), gr.update(visible=True)], None, convert_arr)
+    convert_confirm.click(lambda :[gr.update(visible=False), gr.update(visible=True), gr.update(visible=False)], None, convert_arr)
+    convert_confirm.click(remove_tts_from_history, [shared.gradio['name1'], shared.gradio['name2']], shared.gradio['display'])
+    convert_confirm.click(lambda : chat.save_history(timestamp=False), [], [], show_progress=False)
+    convert_cancel.click(lambda :[gr.update(visible=False), gr.update(visible=True), gr.update(visible=False)], None, convert_arr)
+
+    # Toggle message text in history
+    show_text.change(lambda x: params.update({"show_text": x}), show_text, None)
+    show_text.change(toggle_text_in_history, [shared.gradio['name1'], shared.gradio['name2']], shared.gradio['display'])
+    show_text.change(lambda : chat.save_history(timestamp=False), [], [], show_progress=False)
 
     # Event functions to update the parameters in the backend
     activate.change(lambda x: params.update({"activate": x}), activate, None)
+    autoplay.change(lambda x: params.update({"autoplay": x}), autoplay, None)
     voice.change(lambda x: params.update({"speaker": x}), voice, None)
+    v_pitch.change(lambda x: params.update({"voice_pitch": x}), v_pitch, None)
+    v_speed.change(lambda x: params.update({"voice_speed": x}), v_speed, None)
diff --git a/modules/RWKV.py b/modules/RWKV.py
@@ -25,10 +25,10 @@ def from_pretrained(self, path, dtype="fp16", device="cuda"):
         tokenizer_path = Path(f"{path.parent}/20B_tokenizer.json")
 
         if shared.args.rwkv_strategy is None:
-            model = RWKV(model=os.path.abspath(path), strategy=f'{device} {dtype}')
+            model = RWKV(model=str(path), strategy=f'{device} {dtype}')
         else:
-            model = RWKV(model=os.path.abspath(path), strategy=shared.args.rwkv_strategy)
-        pipeline = PIPELINE(model, os.path.abspath(tokenizer_path))
+            model = RWKV(model=str(path), strategy=shared.args.rwkv_strategy)
+        pipeline = PIPELINE(model, str(tokenizer_path))
 
         result = self()
         result.pipeline = pipeline
@@ -61,7 +61,7 @@ def __init__(self):
     @classmethod
     def from_pretrained(self, path):
         tokenizer_path = path / "20B_tokenizer.json"
-        tokenizer = Tokenizer.from_file(os.path.abspath(tokenizer_path))
+        tokenizer = Tokenizer.from_file(str(tokenizer_path))
 
         result = self()
         result.tokenizer = tokenizer