Cleaned up code and refactored to new file

TODO.md

@@ -18,9 +18,8 @@
   - Use (anchor, positive) pairs to train a new model 
   - Use (sentence) + class labels to train a new model
 - Implement BatchAllTripletLoss
-- Implement a two-phase training regime 
+- Train with anchored definitions and/or...
-  1. Train with anchored definitions then...
+- Train with class labels
-  2. Train with class labels
 
 
 ## Evaluate

pyproject.toml

@@ -3,7 +3,7 @@ name = "ucsinfer"
 version = "0.1.0"
 description = "Tools for applying UCS categories to sounds using large-language models"
 authors = [
-    {name = "Jamie Hardt",email = "jamiehardt@me.com"}
+    {name = "Jamie Hardt", email = "jamiehardt@me.com"}
 ]
 readme = "README.md"
 requires-python = ">=3.13"

ucsinfer/__main__.py

@@ -1,13 +1,11 @@
 import os
-# import csv
 import logging
 from itertools import chain
 
-import tqdm
 import click
-# from tabulate import tabulate, SEPARATING_LINE
 
 from .inference import InferenceContext, load_ucs
+from .import_csv import csv_to_data
 from .gather import (build_sentence_class_dataset, print_dataset_stats,
                      ucs_definitions_generator, scan_metadata, walk_path)
 from .recommend import print_recommendation
@@ -22,6 +20,7 @@ formatter = logging.Formatter(
 stream_handler.setFormatter(formatter)
 logger.addHandler(stream_handler)
 
+
 @click.group(epilog="For more information see "
              "<https://git.squad51.us/jamie/ucsinfer>")
 @click.option('--verbose', '-v', flag_value=True, help='Verbose output')
@@ -52,8 +51,8 @@ def ucsinfer(ctx, verbose, no_model_cache, model, complete_ucs):
         stream_handler.setLevel(logging.WARNING)
 
     os.environ['TOKENIZERS_PARALLELISM'] = 'false'
-    logger.info("Setting TOKENIZERS_PARALLELISM environment variable to `false"
+    logger.info("Setting TOKENIZERS_PARALLELISM environment variable to "
-                " explicitly")
+                "`false` explicitly")
 
     ctx.ensure_object(dict)
     ctx.obj['model_cache'] = not no_model_cache
@@ -77,7 +76,7 @@ def ucsinfer(ctx, verbose, no_model_cache, model, complete_ucs):
               help="Recommend a category for given text instead of reading "
               "from a file")
 @click.argument('paths', nargs=-1, metavar='<paths>')
-@click.option('--interactive','-i', flag_value=True, default=False, 
+@click.option('--interactive', '-i', flag_value=True, default=False,
               help="After processing each path in <paths>, prompt for a "
               "recommendation to accept, and then prepend the selection to "
               "the file name.")
@@ -121,7 +120,8 @@ def recommend(ctx, text, paths, interactive, skip_ucs):
             text = os.path.basename(path)
 
         while True:
-            retval = print_recommendation(path, text, inference_ctx, interactive)
+            retval = print_recommendation(
+                path, text, inference_ctx, interactive)
             if not retval:
                 break
             if retval[0] is False:
@@ -137,16 +137,50 @@ def recommend(ctx, text, paths, interactive, skip_ucs):
                 break
 
 
-@ucsinfer.command('gather')
+@ucsinfer.command('csv')
+@click.option('--filename-col', default="FileName",
+              help="Heading or index of the column containing filenames",
+              show_default=True)
+@click.option('--description-col', default="TrackDescription",
+              help="Heading or index of the column containing descriptions",
+              show_default=True)
 @click.option('--out', default='dataset/', show_default=True)
-# @click.option('--ucs-data', flag_value=True, help="Create a dataset based "
-#               "on the UCS category explanations and synonymns (PATHS will "
-#               "be ignored.)")
 @click.argument('paths', nargs=-1)
 @click.pass_context
-def gather(ctx, paths, out, ucs_data):
+def import_csv(ctx, paths: list[str], out, filename_col, description_col):
     """
-    Scan files to build a training dataset
+    Scan training data from CSV files 
+
+    `csv` is used to build a training dataset for finetuning the selected 
+    model, as like the `gather` command, except instead of scanning the 
+    file system it builds a dataset from descriptions and UCS filenames in 
+    columns of a CSV file.
+    """
+    logger.debug("CSV mode")
+
+    logger.debug(f"Loading category list...")
+    ucs = load_ucs(full_ucs=ctx.obj['complete_ucs'])
+
+    catid_list = [cat.catid for cat in ucs]
+
+    logger.info("Building dataset from csv...")
+
+    dataset = build_sentence_class_dataset(
+        chain(csv_to_data(paths, description_col, filename_col, catid_list),
+              ucs_definitions_generator(ucs)), catid_list)
+
+    logger.info(f"Saving dataset to disk at {out}")
+    print_dataset_stats(dataset, catid_list)
+    dataset.save_to_disk(out)
+
+
+@ucsinfer.command('gather')
+@click.option('--out', default='dataset/', show_default=True)
+@click.argument('paths', nargs=-1)
+@click.pass_context
+def gather(ctx, paths, out):
+    """
+    Scan training data from audio files
 
     `gather` is used to build a training dataset for finetuning the selected
     model. Description sentences and UCS categories are collected from '.wav'
@@ -164,19 +198,15 @@ def gather(ctx, paths, out, ucs_data):
     logger.debug(f"Loading category list...")
     ucs = load_ucs(full_ucs=ctx.obj['complete_ucs'])
 
-    scan_list: list[tuple[str,str]] = []
+    scan_list: list[tuple[str, str]] = []
     catid_list = [cat.catid for cat in ucs]
 
-    if ucs_data:
-        logger.info('Creating dataset for UCS categories instead of from PATH')
-        paths = []
-
     for path in paths:
         scan_list += walk_path(path, catid_list)
 
     logger.info(f"Found {len(scan_list)} files to process.")
 
-    logger.info("Building dataset...")
+    logger.info("Building dataset files...")
 
     dataset = build_sentence_class_dataset(
         chain(scan_metadata(scan_list, catid_list),
@@ -187,6 +217,7 @@ def gather(ctx, paths, out, ucs_data):
     print_dataset_stats(dataset, catid_list)
     dataset.save_to_disk(out)
 
+
 @ucsinfer.command('qualify')
 def qualify():
     """
@@ -208,7 +239,6 @@ def finetune(ctx):
     logger.debug("FINETUNE mode")
 
 
-
 @ucsinfer.command('evaluate')
 @click.argument('dataset', default='dataset/')
 @click.pass_context
@@ -220,6 +250,5 @@ def evaluate(ctx, dataset, offset, limit):
     logger.warning("Model evaluation is not currently implemented")
 
 
-
 if __name__ == '__main__':
     ucsinfer(obj={})

ucsinfer/import_csv.py

@@ -0,0 +1,28 @@
+import csv
+
+from typing import Generator
+
+from .util import parse_ucs
+
+
+def csv_to_data(paths, description_key, filename_key, catid_list) -> Generator[tuple[str, str], None, None]:
+    """
+    Accepts a list of paths and returns an iterator of (sentence, class) 
+    tuples.
+    """
+    for path in paths:
+        with open(path, 'r') as f:
+            records = csv.DictReader(f)
+
+            assert filename_key in records.fieldnames, \
+                (f"Filename key `{filename_key}` not present in file "
+                 "{path}")
+
+            assert description_key in records.fieldnames, \
+                (f"Description key `{description_key}` not present in "
+                 "file {path}")
+
+            for record in records:
+                ucs_comps = parse_ucs(record[filename_key], catid_list)
+                if ucs_comps:
+                    yield (record[description_key], ucs_comps.cat_id)