Merge branch 'langchain' of https://github.com/sensein/senselab into …

…langchain

.github/workflows/tests.yaml

@@ -64,7 +64,7 @@ jobs:
           --verbose
       shell: bash
     - name: Upload coverage to Codecov
-      uses: codecov/codecov-action@v4
+      uses: codecov/codecov-action@v5
       with:
         token: ${{ secrets.CODECOV_TOKEN }}
 

.pre-commit-config.yaml

@@ -37,6 +37,8 @@ repos:
 - repo: https://github.com/pre-commit/pre-commit-hooks
   rev: v4.6.0
   hooks:
+  - id: check-added-large-files
+    args: [--maxkb=15000]
   - id: check-case-conflict
   - id: end-of-file-fixer
   - id: trailing-whitespace

CHANGELOG.md

@@ -1,3 +1,38 @@
+# 0.26.1 (Mon Dec 30 2024)
+
+#### 🐛 Bug Fix
+
+- Update model.py [#228](https://github.com/sensein/senselab/pull/228) ([@fabiocat93](https://github.com/fabiocat93))
+
+#### 📝 Documentation
+
+- Updating tutorial files [#210](https://github.com/sensein/senselab/pull/210) ([@900miles](https://github.com/900miles) [@fabiocat93](https://github.com/fabiocat93))
+
+#### 🔩 Dependency Updates
+
+- Bump codecov/codecov-action from 4 to 5 [#211](https://github.com/sensein/senselab/pull/211) ([@dependabot[bot]](https://github.com/dependabot[bot]) [@fabiocat93](https://github.com/fabiocat93))
+
+#### Authors: 3
+
+- [@900miles](https://github.com/900miles)
+- [@dependabot[bot]](https://github.com/dependabot[bot])
+- Fabio Catania ([@fabiocat93](https://github.com/fabiocat93))
+
+---
+
+# 0.26.0 (Mon Dec 23 2024)
+
+#### 🚀 Enhancement
+
+- Bump the production-dependencies group across 1 directory with 3 updates [#218](https://github.com/sensein/senselab/pull/218) ([@dependabot[bot]](https://github.com/dependabot[bot]) [@fabiocat93](https://github.com/fabiocat93))
+
+#### Authors: 2
+
+- [@dependabot[bot]](https://github.com/dependabot[bot])
+- Fabio Catania ([@fabiocat93](https://github.com/fabiocat93))
+
+---
+
 # 0.25.0 (Tue Dec 10 2024)
 
 #### 🚀 Enhancement

pyproject.toml

@@ -32,7 +32,7 @@ datasets = "~=3"
 torch = "~=2.5"
 torchvision = "~=0.20"
 torchaudio = "~=2.5"
-transformers = "~=4.46.2"
+transformers = "~=4.47"
 pydra = "~=0.25"
 pydantic = "~=2.7"
 accelerate = "*"
@@ -43,7 +43,7 @@ torch-audiomentations = "~=0.11"
 sentence-transformers = "~=3.1"
 jiwer = "~=3.0"
 speechbrain = "~=1"
-pyav = "~=13"
+pyav = "~=14.0"
 pyannote-audio = "~=3.3"
 pycountry = "~=24.6"
 types-requests = "~=2.32"
@@ -130,10 +130,7 @@ target-version = "py310"
 
 [tool.ruff.lint]
 select = ["ANN", "D", "E", "F", "I"]
-ignore = [
-  "ANN101",  # self should not be annotated.
-  "ANN102"  # cls should not be annotated.
-]
+ignore = []
 fixable = ["ALL"]
 unfixable = []
 dummy-variable-rgx = "^(_+|(_+[a-zA-Z0-9_]*[a-zA-Z0-9]+?))$"

src/senselab/text/tasks/evaluate_conversation/metrics.py

@@ -27,9 +27,9 @@ def Rouge(*args: List, **kwargs: Dict) -> rouge_scorer.RougeScorer:
 Rouge.__doc__ = rouge_scorer.RougeScorer.__doc__
 
 
-def sentence_bleu_sacre(*args: List, **kwargs: Dict) -> BLEUScore:
+def sentence_bleu_sacre(*args: List) -> List[BLEUScore]:
     """Wrapper for sacrebleu's sentence_bleu function."""
-    return sb.sentence_bleu(*args, **kwargs)
+    return [sb.sentence_bleu(str(item), str(ref)) for item, ref in args]
 
 
 sentence_bleu_sacre.__doc__ = sb.sentence_bleu.__doc__

src/senselab/text/tasks/llms/transcript_manager.py

@@ -27,7 +27,7 @@ class Transcript:
         convert_json_to_scriptlines(json_obj: Dict) -> List[ScriptLine]: Converts transcript format to LLM format.
     """
 
-    def __init__(self, transcript_path: Path) -> None:
+    def __init__(self: "Transcript", transcript_path: Path) -> None:
         """Initializes the manager with a transcript file path.
 
         Args:
@@ -38,12 +38,12 @@ def __init__(self, transcript_path: Path) -> None:
         json_obj = self._load_transcript(transcript_path)
         self.scriptlines = self.convert_json_to_scriptlines(json_obj)
 
-    def print_human_readable(self) -> None:
+    def print_human_readable(self: "Transcript") -> None:
         """Prints the stored scriptlines in a human-readable format."""
         for message in self.scriptlines:
             print(f"{message.speaker}:\t\t{message.text}\n")
 
-    def get_num_tokens(self) -> int:
+    def get_num_tokens(self: "Transcript") -> int:
         """Returns the total number of OpenAI tokens in the conversation.
 
         Returns:
@@ -56,7 +56,7 @@ def get_num_tokens(self) -> int:
                 c += len(encoding.encode(message.text))
         return c
 
-    def extract_response_opportunities(self) -> List[List[ScriptLine]]:
+    def extract_response_opportunities(self: "Transcript") -> List[List[ScriptLine]]:
         """Extract consecutive sublists from the messages list, ending after every 'user' response.
 
         This is used to compare AI responses to a human's response
@@ -71,7 +71,7 @@ def extract_response_opportunities(self) -> List[List[ScriptLine]]:
         sublists = []
 
         for i, message in enumerate(self.scriptlines):
-            if message.speaker == "user" and i > 0:
+            if message.speaker == "user" and i >= 0:
                 sublist = self.scriptlines[0 : i + 1]
                 sublists.append(sublist)
 

src/senselab/utils/data_structures/device.py

@@ -9,9 +9,9 @@
 class DeviceType(Enum):
     """Device types for PyTorch operations."""
 
-    CPU: str = "cpu"
-    CUDA: str = "cuda"
-    MPS: str = "mps"
+    CPU = "cpu"
+    CUDA = "cuda"
+    MPS = "mps"
 
 
 DTYPE_MAP = {DeviceType.CPU: torch.float32, DeviceType.CUDA: torch.float16, DeviceType.MPS: torch.float32}

src/senselab/utils/data_structures/model.py

@@ -153,7 +153,10 @@ def check_hf_repo_exists(repo_id: str, revision: str = "main", repo_type: str =
     """Private function to check if a Hugging Face repository exists."""
     api = HfApi()
     try:
-        api.list_repo_commits(repo_id=repo_id, revision=revision, repo_type=repo_type)
+        if repo_type == "model":
+            api.model_info(repo_id=repo_id, revision=revision)
+        else:
+            api.list_repo_commits(repo_id=repo_id, revision=revision, repo_type=repo_type)
         return True
     except Exception:
         # raise RuntimeError(f"An error occurred: {e}")

src/tests/text/tasks/transcript_manager_test.py

@@ -94,7 +94,9 @@ def test_get_num_tokens(sample_transcript: Path) -> None:
     def test_response_opportunities_extraction(sample_transcript: Path) -> None:
         """Test the extraction of response opportunities."""
         transcript = Transcript(sample_transcript)
+        print(transcript)
         opportunities = transcript.extract_response_opportunities()
+        print(opportunities)
 
         assert len(opportunities) == 2, "Expected two response opportunities"
         assert opportunities[0][-1].speaker == "user", "Expected last message to be first message from user"

src/tests/utils/data_structures/model_test.py

@@ -9,7 +9,7 @@
 
 def test_check_hf_repo_exists_true() -> None:
     """Test HF repo exists."""
-    with patch("huggingface_hub.HfApi.list_repo_commits") as mock_list_repo_commits:
+    with patch("huggingface_hub.HfApi.model_info") as mock_list_repo_commits:
         mock_list_repo_commits.return_value = True
         assert check_hf_repo_exists("valid_repo") is True
 

tutorials/audio/00_getting_started.ipynb

@@ -6,7 +6,7 @@
             "source": [
                 "# Getting Started with ```senselab```\n",
                 "\n",
-                "[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/sensein/senselab/blob/main/tutorials/audiogetting_started.ipynb)\n",
+                "[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/sensein/senselab/blob/main/tutorials/audio/00_getting_started.ipynb)\n",
                 "\n",
                 "\n",
                 "Welcome to the `senselab` quick start tutorial! \n",
@@ -35,7 +35,7 @@
             },
             "outputs": [],
             "source": [
-                "pip install senselab"
+                "%pip install senselab"
             ]
         },
         {
@@ -48,14 +48,17 @@
         },
         {
             "cell_type": "code",
-            "execution_count": 2,
+            "execution_count": null,
             "metadata": {},
             "outputs": [],
             "source": [
                 "from senselab.audio.data_structures import Audio\n",
+                "!mkdir -p tutorial_audio_files\n",
+                "!wget -O tutorial_audio_files/audio_48khz_mono_16bits.wav https://github.com/sensein/senselab/raw/main/src/tests/data_for_testing/audio_48khz_mono_16bits.wav\n",
+                "!wget -O tutorial_audio_files/audio_48khz_stereo_16bits.wav https://github.com/sensein/senselab/raw/main/src/tests/data_for_testing/audio_48khz_stereo_16bits.wav\n",
                 "\n",
-                "MONO_AUDIO_PATH = \"../../src/tests/data_for_testing/audio_48khz_mono_16bits.wav\"\n",
-                "STEREO_AUDIO_PATH = \"../../src/tests/data_for_testing/audio_48khz_stereo_16bits.wav\"\n",
+                "MONO_AUDIO_PATH = \"tutorial_audio_files/audio_48khz_mono_16bits.wav\"\n",
+                "STEREO_AUDIO_PATH = \"tutorial_audio_files/audio_48khz_stereo_16bits.wav\"\n",
                 "\n",
                 "audio1 = Audio.from_filepath(MONO_AUDIO_PATH)\n",
                 "audio2 = Audio.from_filepath(STEREO_AUDIO_PATH)"
@@ -71,9 +74,18 @@
         },
         {
             "cell_type": "code",
-            "execution_count": null,
+            "execution_count": 3,
             "metadata": {},
-            "outputs": [],
+            "outputs": [
+                {
+                    "name": "stdout",
+                    "output_type": "stream",
+                    "text": [
+                        "The original audio has 2 channels.\n",
+                        "The downmixed audio has 1 channels.\n"
+                    ]
+                }
+            ],
             "source": [
                 "from senselab.audio.tasks.preprocessing import downmix_audios_to_mono\n",
                 "\n",
@@ -331,7 +343,7 @@
             "name": "python",
             "nbconvert_exporter": "python",
             "pygments_lexer": "ipython3",
-            "version": "3.10.10"
+            "version": "3.12.0"
         }
     },
     "nbformat": 4,

tutorials/audio/audio_data_augmentation.ipynb

@@ -16,7 +16,23 @@
             "cell_type": "markdown",
             "metadata": {},
             "source": [
-                "We start by importing the modules required for the augmentation, plotting, and audio processing tasks."
+                "First, we should install senselab if it has not already been installed."
+            ]
+        },
+        {
+            "cell_type": "code",
+            "execution_count": null,
+            "metadata": {},
+            "outputs": [],
+            "source": [
+                "%pip install senselab"
+            ]
+        },
+        {
+            "cell_type": "markdown",
+            "metadata": {},
+            "source": [
+                "Now, we start by importing the modules required for the augmentation, plotting, and audio processing tasks."
             ]
         },
         {
@@ -78,7 +94,10 @@
             "outputs": [],
             "source": [
                 "# Load an audio file\n",
-                "audio = Audio.from_filepath(\"../../src/tests/data_for_testing/audio_48khz_mono_16bits.wav\")\n",
+                "!mkdir -p tutorial_audio_files\n",
+                "!wget -O tutorial_audio_files/audio_48khz_mono_16bits.wav https://github.com/sensein/senselab/raw/main/src/tests/data_for_testing/audio_48khz_mono_16bits.wav\n",
+                "\n",
+                "audio = Audio.from_filepath(\"tutorial_audio_files/audio_48khz_mono_16bits.wav\")\n",
                 "\n",
                 "# Play the audio\n",
                 "play_audio(audio)\n",
@@ -145,7 +164,7 @@
             "name": "python",
             "nbconvert_exporter": "python",
             "pygments_lexer": "ipython3",
-            "version": "3.10.10"
+            "version": "3.12.0"
         }
     },
     "nbformat": 4,

tutorials/audio/extract_speaker_embeddings.ipynb

@@ -17,6 +17,15 @@
                 "First, let's import the necessary libraries and the function we'll be using."
             ]
         },
+        {
+            "cell_type": "code",
+            "execution_count": null,
+            "metadata": {},
+            "outputs": [],
+            "source": [
+                "%pip install senselab"
+            ]
+        },
         {
             "cell_type": "code",
             "execution_count": null,
@@ -45,12 +54,16 @@
         },
         {
             "cell_type": "code",
-            "execution_count": 2,
+            "execution_count": null,
             "metadata": {},
             "outputs": [],
             "source": [
-                "audio1 = Audio.from_filepath(\"../../src/tests/data_for_testing/audio_48khz_mono_16bits.wav\")\n",
-                "audio2 = Audio.from_filepath(\"../../src/tests/data_for_testing/audio_48khz_stereo_16bits.wav\")\n",
+                "!mkdir -p tutorial_audio_files\n",
+                "!wget -O tutorial_audio_files/audio_48khz_mono_16bits.wav https://github.com/sensein/senselab/raw/main/src/tests/data_for_testing/audio_48khz_mono_16bits.wav\n",
+                "!wget -O tutorial_audio_files/audio_48khz_stereo_16bits.wav https://github.com/sensein/senselab/raw/main/src/tests/data_for_testing/audio_48khz_stereo_16bits.wav\n",
+                "\n",
+                "audio1 = Audio.from_filepath(\"tutorial_audio_files/audio_48khz_mono_16bits.wav\")\n",
+                "audio2 = Audio.from_filepath(\"tutorial_audio_files/audio_48khz_stereo_16bits.wav\")\n",
                 "\n",
                 "# Downmix to mono\n",
                 "audio2 = downmix_audios_to_mono([audio2])[0]\n",
@@ -99,7 +112,7 @@
             "metadata": {},
             "outputs": [],
             "source": [
-                "from senselab.utils.tasks.cosine_similarity import cosine_similarity\n",
+                "from senselab.utils.tasks.cosine_similarity import compute_cosine_similarity\n",
                 "\n",
                 "\n",
                 "# DIRECTLY PLOT THE EMBEDDINGS FOR THE TWO FILES\n",
@@ -129,7 +142,7 @@
                 "    \n",
                 "    for i in range(n):\n",
                 "        for j in range(n):\n",
-                "            similarity_matrix[i, j] = cosine_similarity(embeddings[i], embeddings[j])\n",
+                "            similarity_matrix[i, j] = compute_cosine_similarity(embeddings[i], embeddings[j])\n",
                 "    \n",
                 "    fig, ax = plt.subplots(figsize=(8, 6))\n",
                 "    im = ax.imshow(similarity_matrix, cmap='coolwarm', vmin=-1, vmax=1)\n",
@@ -174,7 +187,7 @@
     ],
     "metadata": {
         "kernelspec": {
-            "display_name": "senselab",
+            "display_name": "senselab-KP8v1V64-py3.10",
             "language": "python",
             "name": "python3"
         },
@@ -188,7 +201,7 @@
             "name": "python",
             "nbconvert_exporter": "python",
             "pygments_lexer": "ipython3",
-            "version": "3.12.0"
+            "version": "3.10.10"
         }
     },
     "nbformat": 4,