Merge branch 'main' into HideLord-main

.github/ISSUE_TEMPLATE/bug_report_template.yml

@@ -0,0 +1,53 @@
+name: "Bug report"
+description: Report a bug
+labels: [ "bug" ]
+body:
+  - type: markdown
+    attributes:
+      value: |
+        Thanks for taking the time to fill out this bug report!
+  - type: textarea
+    id: bug-description
+    attributes:
+      label: Describe the bug
+      description: A clear and concise description of what the bug is.
+      placeholder: Bug description
+    validations:
+      required: true
+  - type: checkboxes
+    attributes:
+      label: Is there an existing issue for this?
+      description: Please search to see if an issue already exists for the issue you encountered.
+      options:
+        - label: I have searched the existing issues
+          required: true
+  - type: textarea
+    id: reproduction
+    attributes:
+      label: Reproduction
+      description: Please provide the steps necessary to reproduce your issue.
+      placeholder: Reproduction
+    validations:
+      required: true
+  - type: textarea
+    id: screenshot
+    attributes:
+      label: Screenshot
+      description: "If possible, please include screenshot(s) so that we can understand what the issue is."
+  - type: textarea
+    id: logs
+    attributes:
+      label: Logs
+      description: "Please include the full stacktrace of the errors you get in the command-line (if any)."
+      render: shell
+    validations:
+      required: true
+  - type: textarea
+    id: system-info
+    attributes:
+      label: System Info
+      description: "Please share your system info with us: operating system, GPU brand, and GPU model. If you are using a Google Colab notebook, mention that instead."
+      render: shell
+      placeholder: 
+    validations:
+      required: true

.github/ISSUE_TEMPLATE/feature_request.md

@@ -0,0 +1,16 @@
+---
+name: Feature request
+about: Suggest an improvement or new feature for the web UI
+title: ''
+labels: 'enhancement'
+assignees: ''
+
+---
+
+**Description**
+
+A clear and concise description of what you want to be implemented.
+
+**Additional Context**
+
+If applicable, please provide any extra information, external links, or screenshots that could be useful.

.github/dependabot.yml

@@ -0,0 +1,11 @@
+# To get started with Dependabot version updates, you'll need to specify which
+# package ecosystems to update and where the package manifests are located.
+# Please see the documentation for all configuration options:
+# https://docs.github.com/github/administering-a-repository/configuration-options-for-dependency-updates
+
+version: 2
+updates:
+  - package-ecosystem: "pip" # See documentation for possible values
+    directory: "/" # Location of package manifests
+    schedule:
+      interval: "weekly"

.github/workflows/stale.yml

@@ -0,0 +1,22 @@
+name: Close inactive issues
+on:
+  schedule:
+    - cron: "10 23 * * *"
+
+jobs:
+  close-issues:
+    runs-on: ubuntu-latest
+    permissions:
+      issues: write
+      pull-requests: write
+    steps:
+      - uses: actions/stale@v5
+        with:
+          stale-issue-message: ""
+          close-issue-message: "This issue has been closed due to inactivity for 30 days. If you believe it is still relevant, you can reopen it (if you are the author) or leave a comment below."
+          days-before-issue-stale: 30
+          days-before-issue-close: 0
+          stale-issue-label: "stale"
+          days-before-pr-stale: -1
+          days-before-pr-close: -1
+          repo-token: ${{ secrets.GITHUB_TOKEN }}

README.md

@@ -60,7 +60,9 @@ 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).
+> **Note**
+> 1. If you are on Windows, it may be easier to run the commands above in a WSL environment. The performance may also be better.
+> 2. For a more detailed, user-contributed guide, see: [Installation instructions for human beings](https://github.com/oobabooga/text-generation-webui/wiki/Installation-instructions-for-human-beings).
 
 ## Installation option 2: one-click installers
 
@@ -140,8 +142,9 @@ Optionally, you can use the following command-line flags:
 | `--cai-chat`  | Launch the web UI in chat mode with a style similar to Character.AI's. If the file `img_bot.png` or `img_bot.jpg` exists in the same folder as server.py, this image will be used as the bot's profile picture. Similarly, `img_me.png` or `img_me.jpg` will be used as your profile picture. |
 | `--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 GPTQ_BITS`  |  Load a pre-quantized model with specified precision. 2, 3, 4 and 8 (bit) are supported. Currently only works with LLaMA. |
+| `--load-in-4bit`  | DEPRECATED: use `--gptq-bits 4` instead. |
+| `--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 and OPT. |
+| `--gptq-model-type MODEL_TYPE`  |  Model type of pre-quantized model. Currently only LLaMa and OPT are supported. |
 | `--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. |

api-example-stream.py

@@ -26,6 +26,7 @@ async def run(context):
         'top_p': 0.9,
         'typical_p': 1,
         'repetition_penalty': 1.05,
+        'encoder_repetition_penalty': 1.0,
         'top_k': 0,
         'min_length': 0,
         'no_repeat_ngram_size': 0,
@@ -59,6 +60,7 @@ async def run(context):
                             params['top_p'],
                             params['typical_p'],
                             params['repetition_penalty'],
+                            params['encoder_repetition_penalty'],
                             params['top_k'],
                             params['min_length'],
                             params['no_repeat_ngram_size'],

api-example.py

@@ -24,6 +24,7 @@
     'top_p': 0.9,
     'typical_p': 1,
     'repetition_penalty': 1.05,
+    'encoder_repetition_penalty': 1.0,
     'top_k': 0,
     'min_length': 0,
     'no_repeat_ngram_size': 0,
@@ -45,6 +46,7 @@
         params['top_p'],
         params['typical_p'],
         params['repetition_penalty'],
+        params['encoder_repetition_penalty'],
         params['top_k'],
         params['min_length'],
         params['no_repeat_ngram_size'],

extensions/gallery/script.py

@@ -76,7 +76,7 @@ def generate_html():
     return container_html
 
 def ui():
-    with gr.Accordion("Character gallery"):
+    with gr.Accordion("Character gallery", open=False):
         update = gr.Button("Refresh")
         gallery = gr.HTML(value=generate_html())
     update.click(generate_html, [], gallery)

extensions/silero_tts/script.py

@@ -81,6 +81,7 @@ def input_modifier(string):
     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>')]
 
+    shared.processing_message = "*Is recording a voice message...*"
     return string
 
 def output_modifier(string):
@@ -119,6 +120,7 @@ def output_modifier(string):
         if params['show_text']:
             string += f'\n\n{original_string}'
 
+    shared.processing_message = "*Is typing...*"
     return string
 
 def bot_prefix_modifier(string):

modules/quantized_LLaMA.py → modules/GPTQ_loader.py

@@ -7,28 +7,40 @@
 import modules.shared as shared
 
 sys.path.insert(0, str(Path("repositories/GPTQ-for-LLaMa")))
-from llama import load_quant
+import llama
+import opt
 
 
-# 4-bit LLaMA
-def load_quantized_LLaMA(model_name):
-    if shared.args.load_in_4bit:
-        bits = 4
+def load_quantized(model_name):
+    if not shared.args.gptq_model_type:
+        # Try to determine model type from model name
+        model_type = model_name.split('-')[0].lower()
+        if model_type not in ('llama', 'opt'):
+            print("Can't determine model type from model name. Please specify it manually using --gptq-model-type "
+                  "argument")
+            exit()
     else:
-        bits = shared.args.gptq_bits
+        model_type = shared.args.gptq_model_type.lower()
+
+    if model_type == 'llama':
+        load_quant = llama.load_quant
+    elif model_type == 'opt':
+        load_quant = opt.load_quant
+    else:
+        print("Unknown pre-quantized model type specified. Only 'llama' and 'opt' are supported")
+        exit()
 
     path_to_model = Path(f'models/{model_name}')
-    pt_model = ''
     if path_to_model.name.lower().startswith('llama-7b'):
-        pt_model = f'llama-7b-{bits}bit.pt'
+        pt_model = f'llama-7b-{shared.args.gptq_bits}bit.pt'
     elif path_to_model.name.lower().startswith('llama-13b'):
-        pt_model = f'llama-13b-{bits}bit.pt'
+        pt_model = f'llama-13b-{shared.args.gptq_bits}bit.pt'
     elif path_to_model.name.lower().startswith('llama-30b'):
-        pt_model = f'llama-30b-{bits}bit.pt'
+        pt_model = f'llama-30b-{shared.args.gptq_bits}bit.pt'
     elif path_to_model.name.lower().startswith('llama-65b'):
-        pt_model = f'llama-65b-{bits}bit.pt'
+        pt_model = f'llama-65b-{shared.args.gptq_bits}bit.pt'
     else:
-        pt_model = f'{model_name}-{bits}bit.pt'
+        pt_model = f'{model_name}-{shared.args.gptq_bits}bit.pt'
 
     # Try to find the .pt both in models/ and in the subfolder
     pt_path = None
@@ -40,7 +52,7 @@ def load_quantized_LLaMA(model_name):
         print(f"Could not find {pt_model}, exiting...")
         exit()
 
-    model = load_quant(str(path_to_model), str(pt_path), bits)
+    model = load_quant(str(path_to_model), str(pt_path), shared.args.gptq_bits)
 
     # Multiple GPUs or GPU+CPU
     if shared.args.gpu_memory:

modules/chat.py

@@ -97,7 +97,7 @@ def extract_message_from_reply(question, reply, name1, name2, check, impersonate
 def stop_everything_event():
     shared.stop_everything = True
 
-def chatbot_wrapper(text, max_new_tokens, do_sample, temperature, top_p, typical_p, repetition_penalty, top_k, min_length, no_repeat_ngram_size, num_beams, penalty_alpha, length_penalty, early_stopping, name1, name2, context, check, chat_prompt_size, chat_generation_attempts=1, regenerate=False):
+def chatbot_wrapper(text, max_new_tokens, do_sample, temperature, top_p, typical_p, repetition_penalty, encoder_repetition_penalty, top_k, min_length, no_repeat_ngram_size, num_beams, penalty_alpha, length_penalty, early_stopping, name1, name2, context, check, chat_prompt_size, chat_generation_attempts=1, regenerate=False):
     shared.stop_everything = False
     just_started = True
     eos_token = '\n' if check else None
@@ -126,13 +126,14 @@ def chatbot_wrapper(text, max_new_tokens, do_sample, temperature, top_p, typical
     else:
         prompt = custom_generate_chat_prompt(text, max_new_tokens, name1, name2, context, chat_prompt_size)
 
+    # Yield *Is typing...*
     if not regenerate:
-        yield shared.history['visible']+[[visible_text, '*Is typing...*']]
+        yield shared.history['visible']+[[visible_text, shared.processing_message]]
 
     # Generate
     reply = ''
     for i in range(chat_generation_attempts):
-        for reply in generate_reply(f"{prompt}{' ' if len(reply) > 0 else ''}{reply}", max_new_tokens, do_sample, temperature, top_p, typical_p, repetition_penalty, top_k, min_length, no_repeat_ngram_size, num_beams, penalty_alpha, length_penalty, early_stopping, eos_token=eos_token, stopping_string=f"\n{name1}:"):
+        for reply in generate_reply(f"{prompt}{' ' if len(reply) > 0 else ''}{reply}", max_new_tokens, do_sample, temperature, top_p, typical_p, repetition_penalty, encoder_repetition_penalty, top_k, min_length, no_repeat_ngram_size, num_beams, penalty_alpha, length_penalty, early_stopping, eos_token=eos_token, stopping_string=f"\n{name1}:"):
 
             # Extracting the reply
             reply, next_character_found = extract_message_from_reply(prompt, reply, name1, name2, check)
@@ -159,7 +160,7 @@ def chatbot_wrapper(text, max_new_tokens, do_sample, temperature, top_p, typical
 
     yield shared.history['visible']
 
-def impersonate_wrapper(text, max_new_tokens, do_sample, temperature, top_p, typical_p, repetition_penalty, top_k, min_length, no_repeat_ngram_size, num_beams, penalty_alpha, length_penalty, early_stopping, name1, name2, context, check, chat_prompt_size, chat_generation_attempts=1):
+def impersonate_wrapper(text, max_new_tokens, do_sample, temperature, top_p, typical_p, repetition_penalty, encoder_repetition_penalty, top_k, min_length, no_repeat_ngram_size, num_beams, penalty_alpha, length_penalty, early_stopping, name1, name2, context, check, chat_prompt_size, chat_generation_attempts=1):
     eos_token = '\n' if check else None
 
     if 'pygmalion' in shared.model_name.lower():
@@ -168,28 +169,29 @@ def impersonate_wrapper(text, max_new_tokens, do_sample, temperature, top_p, typ
     prompt = generate_chat_prompt(text, max_new_tokens, name1, name2, context, chat_prompt_size, impersonate=True)
 
     reply = ''
-    yield '*Is typing...*'
+    # Yield *Is typing...*
+    yield shared.processing_message
     for i in range(chat_generation_attempts):
-        for reply in generate_reply(prompt+reply, max_new_tokens, do_sample, temperature, top_p, typical_p, repetition_penalty, top_k, min_length, no_repeat_ngram_size, num_beams, penalty_alpha, length_penalty, early_stopping, eos_token=eos_token, stopping_string=f"\n{name2}:"):
+        for reply in generate_reply(prompt+reply, max_new_tokens, do_sample, temperature, top_p, typical_p, repetition_penalty, encoder_repetition_penalty, top_k, min_length, no_repeat_ngram_size, num_beams, penalty_alpha, length_penalty, early_stopping, eos_token=eos_token, stopping_string=f"\n{name2}:"):
             reply, next_character_found = extract_message_from_reply(prompt, reply, name1, name2, check, impersonate=True)
             yield reply
             if next_character_found:
                 break
         yield reply
 
-def cai_chatbot_wrapper(text, max_new_tokens, do_sample, temperature, top_p, typical_p, repetition_penalty, top_k, min_length, no_repeat_ngram_size, num_beams, penalty_alpha, length_penalty, early_stopping, name1, name2, context, check, chat_prompt_size, chat_generation_attempts=1):
-    for _history in chatbot_wrapper(text, max_new_tokens, do_sample, temperature, top_p, typical_p, repetition_penalty, top_k, min_length, no_repeat_ngram_size, num_beams, penalty_alpha, length_penalty, early_stopping, name1, name2, context, check, chat_prompt_size, chat_generation_attempts):
+def cai_chatbot_wrapper(text, max_new_tokens, do_sample, temperature, top_p, typical_p, repetition_penalty, encoder_repetition_penalty, top_k, min_length, no_repeat_ngram_size, num_beams, penalty_alpha, length_penalty, early_stopping, name1, name2, context, check, chat_prompt_size, chat_generation_attempts=1):
+    for _history in chatbot_wrapper(text, max_new_tokens, do_sample, temperature, top_p, typical_p, repetition_penalty, encoder_repetition_penalty, top_k, min_length, no_repeat_ngram_size, num_beams, penalty_alpha, length_penalty, early_stopping, name1, name2, context, check, chat_prompt_size, chat_generation_attempts):
         yield generate_chat_html(_history, name1, name2, shared.character)
 
-def regenerate_wrapper(text, max_new_tokens, do_sample, temperature, top_p, typical_p, repetition_penalty, top_k, min_length, no_repeat_ngram_size, num_beams, penalty_alpha, length_penalty, early_stopping, name1, name2, context, check, chat_prompt_size, chat_generation_attempts=1):
+def regenerate_wrapper(text, max_new_tokens, do_sample, temperature, top_p, typical_p, repetition_penalty, encoder_repetition_penalty, top_k, min_length, no_repeat_ngram_size, num_beams, penalty_alpha, length_penalty, early_stopping, name1, name2, context, check, chat_prompt_size, chat_generation_attempts=1):
     if (shared.character != 'None' and len(shared.history['visible']) == 1) or len(shared.history['internal']) == 0:
         yield generate_chat_output(shared.history['visible'], name1, name2, shared.character)
     else:
         last_visible = shared.history['visible'].pop()
         last_internal = shared.history['internal'].pop()
-
-        yield generate_chat_output(shared.history['visible']+[[last_visible[0], '*Is typing...*']], name1, name2, shared.character)
-        for _history in chatbot_wrapper(last_internal[0], max_new_tokens, do_sample, temperature, top_p, typical_p, repetition_penalty, top_k, min_length, no_repeat_ngram_size, num_beams, penalty_alpha, length_penalty, early_stopping, name1, name2, context, check, chat_prompt_size, chat_generation_attempts, regenerate=True):
+        # Yield '*Is typing...*'
+        yield generate_chat_output(shared.history['visible']+[[last_visible[0], shared.processing_message]], name1, name2, shared.character)
+        for _history in chatbot_wrapper(last_internal[0], max_new_tokens, do_sample, temperature, top_p, typical_p, repetition_penalty, encoder_repetition_penalty, top_k, min_length, no_repeat_ngram_size, num_beams, penalty_alpha, length_penalty, early_stopping, name1, name2, context, check, chat_prompt_size, chat_generation_attempts, regenerate=True):
             if shared.args.cai_chat:
                 shared.history['visible'][-1] = [last_visible[0], _history[-1][1]]
             else:

modules/models.py

@@ -1,6 +1,5 @@
 import json
 import os
-import sys
 import time
 import zipfile
 from pathlib import Path
@@ -35,14 +34,15 @@
     ds_config = generate_ds_config(shared.args.bf16, 1 * world_size, shared.args.nvme_offload_dir)
     dschf = HfDeepSpeedConfig(ds_config) # Keep this object alive for the Transformers integration
 
+
 def load_model(model_name):
     print(f"Loading {model_name}...")
     t0 = time.time()
 
     shared.is_RWKV = model_name.lower().startswith('rwkv-')
 
     # Default settings
-    if not any([shared.args.cpu, shared.args.load_in_8bit, shared.args.load_in_4bit, shared.args.gptq_bits > 0, shared.args.auto_devices, shared.args.disk, shared.args.gpu_memory is not None, shared.args.cpu_memory is not None, shared.args.deepspeed, shared.args.flexgen, shared.is_RWKV]):
+    if not any([shared.args.cpu, shared.args.load_in_8bit, shared.args.gptq_bits, shared.args.auto_devices, shared.args.disk, shared.args.gpu_memory is not None, shared.args.cpu_memory is not None, shared.args.deepspeed, shared.args.flexgen, shared.is_RWKV]):
         if any(size in shared.model_name.lower() for size in ('13b', '20b', '30b')):
             model = AutoModelForCausalLM.from_pretrained(Path(f"models/{shared.model_name}"), device_map='auto', load_in_8bit=True)
         else:
@@ -87,11 +87,11 @@ def load_model(model_name):
 
         return model, tokenizer
 
-    # 4-bit LLaMA
-    elif shared.args.gptq_bits > 0 or shared.args.load_in_4bit:
-        from modules.quantized_LLaMA import load_quantized_LLaMA
+    # Quantized model
+    elif shared.args.gptq_bits > 0:
+        from modules.GPTQ_loader import load_quantized
 
-        model = load_quantized_LLaMA(model_name)
+        model = load_quantized(model_name)
 
     # Custom
     else: