[devops] remove post commit ci (#5566)

* [devops] remove post commit ci

* [misc] run pre-commit on all files

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci



---------
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

.github/pull_request_template.md

@@ -3,6 +3,7 @@
 - [ ] I have created an issue for this PR for traceability
 - [ ] The title follows the standard format: `[doc/gemini/tensor/...]: A concise description`
 - [ ] I have added relevant tags if possible for us to better distinguish different PRs
+- [ ] I have installed pre-commit: `pip install pre-commit && pre-commit install`
 
 
 ## 🚨 Issue number

.github/workflows/post_commit.yml

-name: post-commit
-
-on:
-  pull_request:
-    types:
-        - closed
-
-jobs:
-  # this job will run after a PR is merged to run pre-commit on any changed file
-  # so that the user does not need to learn pre-commit and pre-commit can still
-  # be auto-executed by the workflow
-  pre-commit:
-    runs-on: ubuntu-latest
-    if: github.event.pull_request.merged == true && github.repository == 'hpcaitech/ColossalAI'
-    steps:
-    - uses: actions/checkout@v2
-      with:
-          fetch-depth: 0
-          ref: ${{ github.event.pull_request.head.sha }}
-
-    # the PR branch and the hpcaitech/colossal-ai main branch
-    # must share a common commit, we need to locate that commit,
-    # which is the commit checked-out or forked when the PR branch is created
-    # such that we can look for files changed since that commit
-    - name: Locate base commit
-      id: locate-base-sha
-      run: |
-          curBranch=$(git rev-parse --abbrev-ref HEAD)
-          commonCommit=$(git merge-base origin/main $curBranch)
-          echo $commonCommit
-          echo "baseSHA=$commonCommit" >> $GITHUB_OUTPUT
-
-    - name: Find the changed files
-      id: find-changed-files
-      uses: tj-actions/changed-files@v35
-      with:
-        base_sha: ${{ steps.locate-base-sha.outputs.baseSHA }}
-
-    - name: List all changed files
-      run: |
-        for file in ${{ steps.find-changed-files.outputs.all_changed_files }}; do
-          echo "$file was changed"
-        done
-
-    # check out the main branch
-    - uses: actions/checkout@v2
-      with:
-        ref: 'main'
-
-    - uses: actions/setup-python@v3
-
-    - name: Cache pre-commit hooks
-      uses: actions/cache@v3
-      with:
-        path: ~/.cache/pre-commit
-        key: ${{ runner.os }}-pre-commit-hooks
-
-    - name: Set up pre-commit
-      run: |
-        pip install pre-commit
-        pre-commit install
-
-    # run pre-commit on changed files
-    - name: Run Pre-commit
-      run: |
-        for file in ${{ steps.find-changed-files.outputs.all_changed_files }}; do
-          pre-commit run --files $file || true
-        done
-
-    # create commit for pre-commit
-    # when all files are well formatted, there is no need to create a commit
-    # therefore, this step will produce an error, which should be allowed
-    - name: Create commits
-      id: commit
-      continue-on-error: true
-      run: |
-        git config --global user.name 'github-actions'
-        git config --global user.email 'github-actions@github.com'
-        git remote set-url origin https://x-access-token:${{ secrets.GITHUB_TOKEN }}@github.com/${{ github.repository }}
-        git add -A
-        git commit -am "[format] applied code formatting on changed files in pull request ${{ github.event.pull_request.number }}"
-
-    # create pull request
-    - name: Create Pull Request
-      if: steps.commit.outcome == 'success'
-      id: cpr
-      uses: peter-evans/create-pull-request@v4
-      with:
-        branch: pre-commit-${{ github.event.pull_request.number }}
-        title: "[format] applied code formatting on changed files in PR ${{ github.event.pull_request.number }}"
-
-    - name: Enable Auto-merge for the New PR
-      if: steps.commit.outcome == 'success'
-      uses: peter-evans/enable-pull-request-automerge@v2
-      with:
-        pull-request-number: ${{ steps.cpr.outputs.pull-request-number }}
-        merge-method: squash

applications/Colossal-LLaMA-2/colossal_llama2/model/init_model.py

@@ -8,11 +8,10 @@ import argparse
 
 import numpy as np
 import torch
-from transformers import LlamaTokenizer, LlamaForCausalLM
+from transformers import LlamaForCausalLM, LlamaTokenizer
 
 from colossalai.logging import get_dist_logger
 
-
 logger = get_dist_logger()
 
 

applications/Colossal-LLaMA-2/colossal_llama2/utils/ckpt_io.py

@@ -10,8 +10,8 @@ import os
 from typing import Any, Dict, Tuple, Union
 
 import torch
-from torch.optim.optimizer import Optimizer
 from torch.optim.lr_scheduler import _LRScheduler
+from torch.optim.optimizer import Optimizer
 
 from colossalai.booster import Booster
 from colossalai.cluster import DistCoordinator

applications/Colossal-LLaMA-2/colossal_llama2/utils/stream_chat_patch.py

 from copy import deepcopy
-from typing import Optional, List, Dict, Tuple, Callable, Any
+from typing import Any, Callable, Dict, List, Optional, Tuple
 
 import torch
 from torch import nn
-
 from transformers import PreTrainedTokenizer
-from transformers.utils import logging
 from transformers.generation.utils import GenerationConfig, LogitsProcessorList, StoppingCriteriaList
+from transformers.utils import logging
 
 logger = logging.get_logger(__name__)
 
 
 def get_prompt_template(
-    input_query:str, 
-    history:List[Dict]= None, 
-    roles:list = ["", "Human", "Assistant"],
+    input_query: str,
+    history: List[Dict] = None,
+    roles: list = ["", "Human", "Assistant"],
 ) -> str:
     """
     Generates a prompt template for chat models based on input and history.
@@ -48,6 +47,7 @@ def get_prompt_template(
                 prompt += f"{role}: <s>"
     return prompt
 
+
 @torch.inference_mode()
 def streaming_chat(
     model: Any,
@@ -99,14 +99,14 @@ def streaming_chat(
         logits_processor = LogitsProcessorList()
 
     generation_kwargs = {
-        'temperature': temperature,
-        'top_p': top_p,
-        'top_k': top_k,
-        'do_sample': do_sample,
-        'max_new_tokens': max_new_tokens,
-        'length_penalty': length_penalty,
-        'use_cache': True,
-        **kwargs
+        "temperature": temperature,
+        "top_p": top_p,
+        "top_k": top_k,
+        "do_sample": do_sample,
+        "max_new_tokens": max_new_tokens,
+        "length_penalty": length_penalty,
+        "use_cache": True,
+        **kwargs,
     }
 
     prompt_str = get_prompt_template(input_query, history=history, roles=roles)
@@ -116,13 +116,18 @@ def streaming_chat(
     history.append({"role": roles[1], "message": input_query.strip()})
     history.append({"role": roles[2], "message": None})
 
-    for outputs in stream_generate(model, **inputs, past_key_values=past_key_values,
-                            eos_token_id=eos_token_id, return_past_key_values=return_past_key_values,
-                            **generation_kwargs):
+    for outputs in stream_generate(
+        model,
+        **inputs,
+        past_key_values=past_key_values,
+        eos_token_id=eos_token_id,
+        return_past_key_values=return_past_key_values,
+        **generation_kwargs,
+    ):
         if return_past_key_values:
             outputs, past_key_values = outputs
 
-        outputs = outputs.tolist()[0][len(inputs["input_ids"][0]):-1]
+        outputs = outputs.tolist()[0][len(inputs["input_ids"][0]) : -1]
         response = tokenizer.decode(outputs)
 
         history[-1]["message"] = response.strip()

applications/Colossal-LLaMA-2/inference_example.py

@@ -15,7 +15,7 @@ def load_model(model_path, device="cuda", **kwargs):
     model.to(device)
 
     try:
-        tokenizer = AutoTokenizer.from_pretrained(model_path, padding_side='left')
+        tokenizer = AutoTokenizer.from_pretrained(model_path, padding_side="left")
     except OSError:
         raise ImportError("Tokenizer not found. Please check if the tokenizer exists or the model path is correct.")
 

applications/Colossal-LLaMA-2/requirements.txt

@@ -12,4 +12,3 @@ flash-attn>=2.0.0,<=2.0.5
 tqdm
 sentencepiece==0.1.99
 protobuf<=3.20.0
-

applications/Colossal-LLaMA-2/stream_chat_example.py

-import os
 import argparse
 
-from transformers import AutoTokenizer, AutoModelForCausalLM
 from colossal_llama2.utils.stream_chat_patch import streaming_chat
+from transformers import AutoModelForCausalLM, AutoTokenizer
 
 SYSTEM = "A chat between a curious human and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the human's questions."
 
+
 def main(args):
     model = AutoModelForCausalLM.from_pretrained(args.model_path).cuda().eval()
     tokenizer = AutoTokenizer.from_pretrained(args.tokenizer_path)
@@ -27,29 +27,34 @@ def main(args):
         print(f"\n{roles[2]}: ", end="")
         gen_len = 0
         for response, history, past_key_values in streaming_chat(
-            model, tokenizer, input_query, history=history, roles=roles,
-            temperature = args.temperature,
-            top_p = args.top_p,
-            top_k = args.top_k,
-            do_sample = args.do_sample,
-            length_penalty = args.length_penalty,
-            max_new_tokens = args.max_new_tokens,
+            model,
+            tokenizer,
+            input_query,
+            history=history,
+            roles=roles,
+            temperature=args.temperature,
+            top_p=args.top_p,
+            top_k=args.top_k,
+            do_sample=args.do_sample,
+            length_penalty=args.length_penalty,
+            max_new_tokens=args.max_new_tokens,
             past_key_values=past_key_values,
-            return_past_key_values=True):
-
+            return_past_key_values=True,
+        ):
             output = response[gen_len:]
             print(output, end="", flush=True)
             gen_len = len(response)
 
+
 if __name__ == "__main__":
     parser = argparse.ArgumentParser()
-    parser.add_argument('--model_path', type=str, default=None, help="path to chat version model")
-    parser.add_argument('--tokenizer_path', type=str, default=None, help="path to chat version tokenizer")
-    parser.add_argument('--temperature', type=float, default=0.8, help="set temperature")
-    parser.add_argument('--top_p', type=float, default=0.95, help="set top p value")
-    parser.add_argument('--top_k', type=int, default=50, help="set top k value")
-    parser.add_argument('--do_sample', type=bool, default=True, help="whether turn on do_sample or not")
-    parser.add_argument('--length_penalty', type=float, default=1.2, help="set length penalty")
-    parser.add_argument('--max_new_tokens', type=int, default=512, help="set max new tokens")
+    parser.add_argument("--model_path", type=str, default=None, help="path to chat version model")
+    parser.add_argument("--tokenizer_path", type=str, default=None, help="path to chat version tokenizer")
+    parser.add_argument("--temperature", type=float, default=0.8, help="set temperature")
+    parser.add_argument("--top_p", type=float, default=0.95, help="set top p value")
+    parser.add_argument("--top_k", type=int, default=50, help="set top k value")
+    parser.add_argument("--do_sample", type=bool, default=True, help="whether turn on do_sample or not")
+    parser.add_argument("--length_penalty", type=float, default=1.2, help="set length penalty")
+    parser.add_argument("--max_new_tokens", type=int, default=512, help="set max new tokens")
     args = parser.parse_args()
     main(args)

applications/ColossalChat/examples/training_scripts/train_dpo.py

@@ -20,13 +20,13 @@ import colossalai
 from colossalai.booster import Booster
 from colossalai.booster.plugin import GeminiPlugin, HybridParallelPlugin, LowLevelZeroPlugin
 from colossalai.cluster import DistCoordinator
+from colossalai.logging import get_dist_logger
 from colossalai.nn.lr_scheduler import CosineAnnealingWarmupLR
 from colossalai.nn.optimizer import HybridAdam
-from colossalai.utils import get_current_device
-from colossalai.logging import get_dist_logger
 
 logger = get_dist_logger()
 
+
 def train(args):
     # check lora compatibility
     if "gemini" in args.plugin and args.lora_rank > 0:

applications/ColossalEval/examples/dataset_evaluation/inference.py

@@ -3,7 +3,6 @@ import copy
 import os
 from typing import Dict, List
 
-import torch
 import torch.distributed as dist
 from colossal_eval import dataset, models, utils
 

applications/ColossalMoE/infer.py

@@ -106,6 +106,5 @@ def main():
     print(f"[{coordinator.rank}] {outputs}")
 
 
-
 if __name__ == "__main__":
     main()

applications/ColossalQA/colossalqa/chain/retrieval_qa/base.py

@@ -24,6 +24,7 @@ from langchain.pydantic_v1 import Field
 from langchain.schema import BaseRetriever, Document
 from langchain.schema.language_model import BaseLanguageModel
 
+
 class CustomBaseRetrievalQA(BaseRetrievalQA):
     """Base class for question-answering chains."""
 
@@ -98,7 +99,6 @@ class CustomBaseRetrievalQA(BaseRetrievalQA):
             for k, v in inputs.items()
             if k in ["stop", "temperature", "top_k", "top_p", "max_new_tokens", "doc_prefix"]
         }
-        answers = []
         if self.combine_documents_chain.memory is not None:
             buffered_history_backup, summarized_history_temp_backup = copy.deepcopy(
                 self.combine_documents_chain.memory.buffered_history
@@ -117,10 +117,10 @@ class CustomBaseRetrievalQA(BaseRetrievalQA):
             ) = copy.deepcopy(buffered_history_backup), copy.deepcopy(summarized_history_temp_backup)
 
         # if rejection_trigger_keywords is not given, return the response from LLM directly
-        rejection_trigger_keywords = inputs.get('rejection_trigger_keywords', [])
+        rejection_trigger_keywords = inputs.get("rejection_trigger_keywords", [])
         answer = answer if all([rej not in answer for rej in rejection_trigger_keywords]) else None
         if answer is None:
-            answer = inputs.get('rejection_answer', "抱歉，根据提供的信息无法回答该问题。")
+            answer = inputs.get("rejection_answer", "抱歉，根据提供的信息无法回答该问题。")
         if self.combine_documents_chain.memory is not None:
             self.combine_documents_chain.memory.save_context({"question": question}, {"output": answer})
 
@@ -161,10 +161,14 @@ class CustomBaseRetrievalQA(BaseRetrievalQA):
             input_documents=docs, question=question, callbacks=_run_manager.get_child(), **kwargs
         )
         # if rejection_trigger_keywords is not given, return the response from LLM directly
-        rejection_trigger_keywords = inputs.get('rejection_trigger_keywords', [])
-        answer = answer if all([rej not in answer for rej in rejection_trigger_keywords]) or len(rejection_trigger_keywords)==0 else None
+        rejection_trigger_keywords = inputs.get("rejection_trigger_keywords", [])
+        answer = (
+            answer
+            if all([rej not in answer for rej in rejection_trigger_keywords]) or len(rejection_trigger_keywords) == 0
+            else None
+        )
         if answer is None:
-            answer = inputs.get('rejection_answer', "抱歉，根据提供的信息无法回答该问题。")
+            answer = inputs.get("rejection_answer", "抱歉，根据提供的信息无法回答该问题。")
         self.combine_documents_chain.memory.save_context({"question": question}, {"output": answer})
 
         if self.return_source_documents: