chore: lint with ruff

.github/workflows/ci.yml

@@ -32,3 +32,7 @@ jobs:
           run: |
             source env/bin/activate
             ruff format --check
+        - name: Run Ruff linting
+          run: |
+            source env/bin/activate
+            ruff check

TEST_get_emission.py

@@ -1,6 +1,4 @@
 import bittensor as bt
-from bittensor_wallet import Wallet
-from bittensor import Balance, tao
 import time
 import random
 

docs/stream_tutorial/client.py

@@ -48,9 +48,6 @@ async def main():
                     print()
                 if isinstance(chunk, list):
                     print(chunk[0], end="", flush=True)
-                else:
-                    # last object yielded is the synapse itself with completion filled
-                    synapse = chunk
             break
 
     # Run the main function with asyncio

docs/stream_tutorial/miner.py

@@ -10,7 +10,7 @@
 
 import bittensor as bt
 from transformers import GPT2Tokenizer
-from typing import List, Dict, Tuple, Union, Callable, Awaitable
+from typing import Dict, Tuple
 
 from protocol import StreamPrompting
 from config import get_config, check_config
@@ -62,7 +62,7 @@ def __init__(self, config=None, axon=None, wallet=None, subtensor=None):
         # The axon handles request processing, allowing validators to send this process requests.
         self.axon = axon or bt.axon(wallet=self.wallet, port=self.config.axon.port)
         # Attach determiners which functions are called when servicing a request.
-        bt.logging.info(f"Attaching forward function to axon.")
+        bt.logging.info("Attaching forward function to axon.")
         print(f"Attaching forward function to axon. {self._prompt}")
         self.axon.attach(
             forward_fn=self._prompt,
@@ -166,12 +166,10 @@ def run(self):
         bt.logging.info(f"Miner starting at block: {self.last_epoch_block}")
 
         # This loop maintains the miner's operations until intentionally stopped.
-        bt.logging.info(f"Starting main loop")
+        bt.logging.info("Starting main loop")
         step = 0
         try:
             while not self.should_exit:
-                start_epoch = time.time()
-
                 # --- Wait until next epoch.
                 current_block = self.subtensor.get_current_block()
                 while (
@@ -215,7 +213,7 @@ def run(self):
             exit()
 
         # In case of unforeseen errors, the miner will log the error and continue operations.
-        except Exception as e:
+        except Exception:
             bt.logging.error(traceback.format_exc())
 
     def run_in_background_thread(self):

docs/stream_tutorial/protocol.py

@@ -1,8 +1,7 @@
 import pydantic
 import bittensor as bt
 
-from abc import ABC, abstractmethod
-from typing import List, Union, Callable, Awaitable
+from typing import List
 from starlette.responses import StreamingResponse
 
 

neurons/miner.py

@@ -22,10 +22,7 @@
 os.environ["USE_TORCH"] = "1"
 
 import time
-import json
 import typing
-import requests
-import asyncio
 import bittensor as bt
 
 # Bittensor Miner Template:

neurons/validator.py

@@ -48,10 +48,8 @@
     MIN_LENGTH_BOOST_TOKEN_COUNT,
     MAX_LENGTH_BOOST_TOKEN_COUNT,
     STUFFED_DESCRIPTION_PUNISHMENT,
-    FOCUS_MIN_SCORE,
     MIN_AUDIO_LENGTH_SECONDS,
     MAX_AUDIO_LENGTH_SECONDS,
-    MIN_AUDIO_LENGTH_SCORE,
     SPEECH_CONTENT_SCALING_FACTOR,
     SPEAKER_DOMINANCE_SCALING_FACTOR,
     BACKGROUND_NOISE_SCALING_FACTOR,
@@ -78,7 +76,7 @@
 import traceback
 import random
 import datetime as dt
-from typing import List, Tuple, Optional, BinaryIO, Dict
+from typing import List, Tuple, Optional, Dict
 import asyncio
 from aiohttp import ClientSession, BasicAuth
 import os
@@ -918,15 +916,17 @@ async def check_videos_and_calculate_rewards_youtube(
             score = max(score, MIN_SCORE)
 
             # Log all our scores
-            bt.logging.info(f"""
+            bt.logging.info(
+                f"""
                 is_unique: {[not is_sim for is_sim in is_too_similar]},
                 video cosine sim: {video_description_relevance_scores},
                 audio cosine sim: {audio_description_relevance_scores},
                 description relevance scores: {description_relevance_scores},
                 query relevance scores: {query_relevance_scores},
                 length scalers: {length_scalers},
                 total score: {score}
-            """)
+            """
+            )
 
             # Upload our final results to API endpoint for index and dataset insertion. Include leaderboard statistics
             miner_hotkey = videos.axon.hotkey
@@ -1031,7 +1031,7 @@ async def upload_video_metadata(
                     json=payload,
                 ) as response:
                     response.raise_for_status()
-                    result = await response.json()
+                    await response.json()
             return True
         except Exception as e:
             bt.logging.debug(f"Error trying upload_video_metadata_endpoint: {e}")
@@ -1084,7 +1084,7 @@ async def upload_audio_metadata(
                     json=payload,
                 ) as response:
                     response.raise_for_status()
-                    result = await response.json()
+                    await response.json()
             return True
         except Exception as e:
             bt.logging.debug(f"Error trying upload_audio_metadata_endpoint: {e}")

omega/__init__.py

@@ -27,7 +27,7 @@
 )
 
 # Import all submodules.
-from . import protocol
-from . import base
-from . import validator
-from .subnet_links import SUBNET_LINKS
+from . import protocol  # noqa: E402, F401
+from . import base  # noqa: E402, F401
+from . import validator  # noqa: E402, F401
+from .subnet_links import SUBNET_LINKS  # noqa: E402, F401

omega/audio_scoring.py

@@ -5,7 +5,6 @@
     np.NAN = np.nan
 from pyannote.audio import Pipeline
 import librosa
-import os
 import dotenv
 import pandas as pd
 import torch
@@ -127,7 +126,6 @@ def total_score(self, audio_arr, sr, timestamps_start, timestamps_end, speakers)
 
 if __name__ == "__main__":
     from datasets import load_dataset
-    import huggingface_hub
 
     repo_id = "diarizers-community/voxconverse"
 

omega/base/miner.py

@@ -57,7 +57,7 @@ def __init__(self, config=None):
         self.axon = bt.axon(wallet=self.wallet, config=self.config)
 
         # Attach determiners which functions are called when servicing a request.
-        bt.logging.info(f"Attaching forward function to miner axon.")
+        bt.logging.info("Attaching forward function to miner axon.")
         self.axon.attach(
             forward_fn=self.forward_videos,
             blacklist_fn=self.blacklist_videos,
@@ -137,7 +137,7 @@ def run(self):
             exit()
 
         # In case of unforeseen errors, the miner will log the error and continue operations.
-        except Exception as e:
+        except Exception:
             bt.logging.error(traceback.format_exc())
 
     def run_in_background_thread(self):

omega/base/neuron.py

@@ -16,7 +16,6 @@
 # DEALINGS IN THE SOFTWARE.
 
 import copy
-import typing
 import datetime as dt
 
 import bittensor as bt

omega/base/validator.py

@@ -34,7 +34,7 @@
 from omega.base.neuron import BaseNeuron
 from omega.mock import MockDendrite
 from omega.utils.config import add_validator_args
-from omega.constants import FOCUS_REWARDS_PERCENT, AUDIO_REWARDS_PERCENT
+from omega.constants import AUDIO_REWARDS_PERCENT
 
 
 class BaseValidatorNeuron(BaseNeuron):
@@ -191,7 +191,7 @@ def run(self):
                     break
 
                 if self.config.neuron.auto_update and self.should_restart():
-                    bt.logging.info(f"Validator is out of date, quitting to restart.")
+                    bt.logging.info("Validator is out of date, quitting to restart.")
                     raise KeyboardInterrupt
 
                 # Sync metagraph and potentially set weights.
@@ -320,7 +320,7 @@ def set_weights(self):
         # Check if self.scores contains any NaN values and log a warning if it does.
         if torch.isnan(self.scores).any():
             bt.logging.warning(
-                f"Scores contain NaN values. This may be due to a lack of responses from miners, or a bug in your reward functions."
+                "Scores contain NaN values. This may be due to a lack of responses from miners, or a bug in your reward functions."
             )
 
         self.scores, self.focus_scores, self.audio_score_arr = self.pad_tensors(

omega/diarization_metric.py

@@ -1,4 +1,4 @@
-from pyannote.core import Segment, Timeline, Annotation
+from pyannote.core import Segment, Annotation
 from pyannote.metrics.diarization import DiarizationErrorRate
 from omega.diarization_pipeline import CustomDiarizationPipeline
 import numpy as np

omega/imagebind_wrapper.py

@@ -7,7 +7,7 @@
 from imagebind import data
 from imagebind.models import imagebind_model
 from imagebind.models.imagebind_model import ModalityType
-from imagebind.models.multimodal_preprocessors import SimpleTokenizer, TextPreprocessor
+from imagebind.models.multimodal_preprocessors import SimpleTokenizer
 from pydantic import BaseModel
 import torch
 
@@ -135,7 +135,7 @@ def generate_text_embeddings(self, text: str):
     def get_inputs(self, video_file: BinaryIO) -> dict:
         audio_file = video_utils.copy_audio(video_file.name)
         try:
-            duration = video_utils.get_video_duration(video_file.name)
+            video_utils.get_video_duration(video_file.name)
             video_data = data.load_and_transform_video_data(
                 [video_file.name],
                 self.device,
@@ -180,7 +180,6 @@ def embed(self, descriptions: List[str], video_files: List[BinaryIO]) -> Embeddi
     @torch.no_grad()
     def embed_only_video(self, video_files: List[BinaryIO]) -> Embeddings:
         video_filepaths = [video_file.name for video_file in video_files]
-        durations = [video_utils.get_video_duration(f.name) for f in video_files]
         embeddings = self.imagebind(
             {
                 ModalityType.VISION: [
@@ -201,7 +200,6 @@ def embed_video_and_text(
         self, video_files: List[BinaryIO], descriptions: List[str]
     ) -> Embeddings:
         video_filepaths = [video_file.name for video_file in video_files]
-        durations = [video_utils.get_video_duration(f.name) for f in video_files]
         embeddings = self.imagebind(
             {
                 ModalityType.VISION: [

omega/miner_utils.py

@@ -12,7 +12,6 @@
     MAX_VIDEO_LENGTH,
     FIVE_MINUTES,
     MAX_AUDIO_LENGTH_SECONDS,
-    MIN_AUDIO_LENGTH_SECONDS,
 )
 from omega import video_utils
 from omega.diarization_pipeline import CustomDiarizationPipeline

omega/unstuff.py

@@ -5,7 +5,6 @@
 import random
 import torch.nn.functional as F
 from omega.imagebind_wrapper import (
-    split_text_by_token_limit,
     SimpleTokenizer,
     BPE_PATH,
     split_text_by_token_limit,

omega/utils/__init__.py

@@ -1,3 +1,3 @@
-from . import config
-from . import misc
-from . import uids
+from . import config  # noqa: F401
+from . import misc  # noqa: F401
+from . import uids  # noqa: F401

omega/utils/misc.py

@@ -17,8 +17,6 @@
 # DEALINGS IN THE SOFTWARE.
 
 import time
-import math
-import hashlib as rpccheckhealth
 from math import floor
 from typing import Callable, Any
 from functools import lru_cache, update_wrapper

omega/video_utils.py

@@ -8,8 +8,6 @@
 import ffmpeg
 from pydantic import BaseModel
 from yt_dlp import YoutubeDL
-import librosa
-import numpy as np
 
 from omega.constants import FIVE_MINUTES
 

purchase_focus_video.py

@@ -128,7 +128,7 @@ def display_videos(videos_data):
     for idx, video in enumerate(videos_data, 1):
         # Convert created_at to a more readable format
         created_at = datetime.fromisoformat(video["created_at"].replace("Z", "+00:00"))
-        formatted_date = created_at.strftime("%Y-%m-%d %H:%M:%S")
+        created_at.strftime("%Y-%m-%d %H:%M:%S")
 
         table_data.append(
             [
@@ -222,7 +222,7 @@ def get_wallet(wallet_name=None, wallet_hotkey=None, wallet_path=None):
 
     wallet = btcli_wallet(name=name, hotkey=hotkey_name, path=path)
     try:
-        hotkey = wallet.get_hotkey()
+        wallet.get_hotkey()
     except Exception as e:
         print(f"{RED}Error loading hotkey: {e} {RESET}")
         return
@@ -427,7 +427,7 @@ def select_order_for_verification():
     while True:
         if purchases:
             print(
-                f"*** NOTE: A purchase is finalized when the purchase state is 'verified'. ***"
+                "*** NOTE: A purchase is finalized when the purchase state is 'verified'. ***"
             )
             choice = input(
                 f"{CYAN}Enter the number of the order to verify, 'm' for manual input, or 'n' to cancel: {RESET}"

ruff.toml

@@ -0,0 +1,2 @@
+[lint]
+ignore=["E722"]

test_audio_dataset.py

@@ -1,8 +1,9 @@
 import os
 from datasets import load_dataset
-from huggingface_hub import login
 from io import BytesIO
 import soundfile as sf
+import numpy as np
+import pandas as pd
 
 # Set HF_TOKEN environment variable or pass directly
 HF_TOKEN = os.getenv("HF_TOKEN")
@@ -26,10 +27,6 @@
     else:
         print(f"{key}: {first_row[key]}")
 
-
-import librosa
-import numpy as np
-
 breakpoint()
 audio_bytes = first_row["audio_bytes"]
 audio_arr, sr = sf.read(BytesIO(audio_bytes))
@@ -81,9 +78,6 @@
         }
     )
 
-# Convert to pandas DataFrame and save as CSV
-import pandas as pd
-
 df = pd.DataFrame(diarization_data)
 os.makedirs("Dataset_audios/Metadata", exist_ok=True)
 df.to_csv(f"Dataset_audios/Metadata/{youtube_id}_diarization.csv", index=False)