Make dependencies more loose

.github/workflows/build_and_test.yml

@@ -8,7 +8,7 @@ jobs:
  runs-on: ubuntu-20.04
  strategy:
  matrix:
- python-version: [3.6, 3.7, 3.8]
+ python-version: ["3.7", "3.8", "3.9", "3.10", "3.11"]
 
  steps:
  - uses: actions/checkout@v2

.gitignore

@@ -1,3 +1,6 @@
+# ignore files dumped in tests
+tmp/
+
 *.py[cod]
 __pycache__
 

requirements.txt

@@ -1,11 +1,11 @@
-matplotlib~=3.3.1
-numpy~=1.19.1
-ipython~=7.16.1
-scikit-learn~=0.24.2
-seaborn~=0.11.0
-transformers~=4.6.1
-pytest~=6.1.2
-setuptools~=49.6.0
-torch~=1.9.0
-PyYAML==5.4.1
-captum==0.4.1
+matplotlib>=3.3
+numpy>=1.19
+ipython>=7.16
+scikit-learn>=0.24.2,<2
+seaborn>=0.11
+transformers~=4.6
+pytest>=6.1.2
+setuptools>=49.6.0
+torch>=1.9.0,<3
+PyYAML>=6.0
+captum~=0.4.1

setup.py

@@ -52,6 +52,8 @@ def read(*names, **kwargs):
  'Programming Language :: Python :: 3.7',
  'Programming Language :: Python :: 3.8',
  'Programming Language :: Python :: 3.9',
+ 'Programming Language :: Python :: 3.10',
+ 'Programming Language :: Python :: 3.11',
  'Programming Language :: Python :: Implementation :: CPython',
  'Programming Language :: Python :: Implementation :: PyPy',
  'Topic :: Utilities',
@@ -63,12 +65,12 @@ def read(*names, **kwargs):
  keywords=[
  'Natural Language Processing', 'Explainable AI', 'keyword3',
  ],
- python_requires='!=3.0.*, !=3.1.*, !=3.2.*, !=3.3.*, !=3.4.*',
+ python_requires='!=3.0.*, !=3.1.*, !=3.2.*, !=3.3.*, !=3.4.*, !=3.5.*, !=3.6.*',
  install_requires=[
  "transformers ~= 4.2",
- "seaborn ~= 0.11",
- "scikit-learn~=0.23",
- "PyYAML~=5.4",
+ "seaborn >= 0.11",
+ "scikit-learn>=0.23,<2",
+ "PyYAML>=6.0",
  "captum ~= 0.4"
  ],
  extras_require={

src/ecco/lm.py

@@ -199,12 +199,12 @@ def generate(self, input_str: str,
  raise ValueError(
  "max_length set to {} while input token has more tokens ({}). Consider increasing max_length" \
  .format(max_length, cur_len))
- 
+
  # Get decoder input ids
  if self.model_type == 'enc-dec': # FIXME: only done because causal LMs like GPT-2 have the _prepare_decoder_input_ids_for_generation method but do not use it
  assert len(input_ids.size()) == 2 # will break otherwise
  if version.parse(transformers.__version__) >= version.parse('4.13'):
- 
+
  # following the code in https://github.com/huggingface/transformers/blob/d0c1aebea467af499331234e7b285a6bf91ea073/tests/generation/test_utils.py#L2099
  model_kwargs = self.model._prepare_encoder_decoder_kwargs_for_generation(input_ids, {})
  decoder_input_ids, model_kwargs = self.model._prepare_decoder_input_ids_for_generation(
@@ -649,7 +649,7 @@ def display_input_sequence(self, input_ids):
  parentDiv: '{viz_id}',
  data: {json.dumps(data)},
  tokenization_config: {json.dumps(self.model_config['tokenizer_config'])}
- 
+
  }})
  }}, function (err) {{
  console.log(err);
@@ -729,10 +729,11 @@ def sample_output_token(scores, do_sample, temperature, top_k, top_p):
  if temperature != 1.0:
  scores = scores / temperature
  # Top-p/top-k filtering
- next_token_logscores = transformers.generation_utils. \
- top_k_top_p_filtering(scores,
- top_k=top_k,
- top_p=top_p)
+ if version.parse(transformers.__version__) >= version.parse('4.25.1'):
+ top_k_top_p_filtering_fn = transformers.generation.utils.top_k_top_p_filtering
+ else:
+ top_k_top_p_filtering_fn = transformers.generation_utils.top_k_top_p_filtering
+ next_token_logscores = top_k_top_p_filtering_fn(scores, top_k=top_k, top_p=top_p)
  # Sample
  probs = F.softmax(next_token_logscores, dim=-1)