Update S3 links

Peter Izsak · Peter Izsak · commit 66863f850822 · 2020-09-16T08:44:55.000+03:00
diff --git a/docs-source/source/model_zoo.rst b/docs-source/source/model_zoo.rst
@@ -27,27 +27,27 @@ NLP Architect Model Zoo
      - Links
    * - :doc:`Sparse GNMT <sparse_gnmt>`
      - 90% sparse GNMT model and a 2x2 block sparse translating German to English trained on Europarl-v7 [#]_ , Common Crawl and News Commentary 11 datasets
-     -  | `model <https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/sparse_gnmt/gnmt_sparse.zip>`_
-        | `2x2 block sparse model <https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/sparse_gnmt/gnmt_blocksparse2x2.zip>`_
+     -  | `model <https://d2zs9tzlek599f.cloudfront.net/models/sparse_gnmt/gnmt_sparse.zip>`_
+        | `2x2 block sparse model <https://d2zs9tzlek599f.cloudfront.net/models/sparse_gnmt/gnmt_blocksparse2x2.zip>`_
    * - :doc:`Intent Extraction <intent>`
      - A :py:class:`MultiTaskIntentModel <nlp_architect.models.intent_extraction.MultiTaskIntentModel>` intent extraction and slot tagging model, trained on SNIPS NLU dataset
-     - | `model <https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/intent/model.h5>`_
-       | `params <https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/intent/model_info.dat>`_
+     - | `model <https://d2zs9tzlek599f.cloudfront.net/models/intent/model.h5>`_
+       | `params <https://d2zs9tzlek599f.cloudfront.net/models/intent/model_info.dat>`_
    * - :doc:`Named Entity Recognition <ner_crf>`
      - A :py:class:`NERCRF <nlp_architect.models.ner_crf.NERCRF>` model trained on CoNLL 2003 dataset
-     - | `model <https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/ner/model.h5>`_
-       | `params <https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/ner/model_info.dat>`_
+     - | `model <https://d2zs9tzlek599f.cloudfront.net/models/ner/model.h5>`_
+       | `params <https://d2zs9tzlek599f.cloudfront.net/models/ner/model_info.dat>`_
    * - :doc:`Dependency parser <bist_parser>`
      - Graph-based dependency parser using BiLSTM feature extractors
-     - `model <https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/dep_parse/bist-pretrained.zip>`_
+     - `model <https://d2zs9tzlek599f.cloudfront.net/models/dep_parse/bist-pretrained.zip>`_
    * - :doc:`Machine comprehension <reading_comprehension>`
      - Match LSTM model trained on SQuAD dataset
-     - | `model <https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/mrc/mrc_model.zip>`_
-       | `data <https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/mrc/mrc_data.zip>`_
+     - | `model <https://d2zs9tzlek599f.cloudfront.net/models/mrc/mrc_model.zip>`_
+       | `data <https://d2zs9tzlek599f.cloudfront.net/models/mrc/mrc_data.zip>`_
    * - :doc:`Word chunker <chunker>`
      - A word chunker model trained on CoNLL 2000 dataset
-     - | `model <https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/chunker/model.h5>`_
-       | `params <https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/chunker/model_info.dat.params>`_
+     - | `model <https://d2zs9tzlek599f.cloudfront.net/models/chunker/model.h5>`_
+       | `params <https://d2zs9tzlek599f.cloudfront.net/models/chunker/model_info.dat.params>`_
 
 References
 ----------
diff --git a/docs-source/source/publications.rst b/docs-source/source/publications.rst
@@ -53,4 +53,4 @@ Demos
 -----
 - NeurIPS 2018:
 
-  - Unsupervised Aspect-based Sentiment Analysis (`video <https://s3-us-west-2.amazonaws.com/nlp-architect-data/content/absa_kingsman_demo.mp4>`_)
+  - Unsupervised Aspect-based Sentiment Analysis (`video <https://d2zs9tzlek599f.cloudfront.net/content/absa_kingsman_demo.mp4>`_)
diff --git a/docs-source/source/sparse_gnmt.rst b/docs-source/source/sparse_gnmt.rst
@@ -151,7 +151,7 @@ Run inference using our pre-trained models:
 .. code-block:: bash
 
     # Download pre-trained model zip file, e.g. gnmt_sparse.zip
-    wget https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/sparse_gnmt/gnmt_sparse.zip
+    wget https://d2zs9tzlek599f.cloudfront.net/models/sparse_gnmt/gnmt_sparse.zip
 
     # Unzip checkpoint + vocabulary files
     unzip gnmt_sparse.zip -d /tmp/gnmt_sparse_checkpoint
@@ -386,5 +386,5 @@ References
 .. _`Shared Task`: http://www.statmt.org/wmt16/translation-task.html
 .. _`Model Zoo`: http://nlp_architect.nervanasys.com/model_zoo.html
 .. _`TensorFlow API`: https://www.tensorflow.org/api_docs/python/tf/quantize
-.. _`Sparse`: https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/sparse_gnmt/gnmt_sparse.zip
-.. _`2x2 Block Sparse`: https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/sparse_gnmt/gnmt_blocksparse2x2.zip
+.. _`Sparse`: https://d2zs9tzlek599f.cloudfront.net/models/sparse_gnmt/gnmt_sparse.zip
+.. _`2x2 Block Sparse`: https://d2zs9tzlek599f.cloudfront.net/models/sparse_gnmt/gnmt_blocksparse2x2.zip
diff --git a/docs-source/source/term_set_expansion.rst b/docs-source/source/term_set_expansion.rst
@@ -84,19 +84,19 @@ size, min_count, window and hs hyperparameters. Please refer to the np2vec modul
     --corpus_format txt
 
 
-A `pretrained model <https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/term_set/enwiki-20171201_pretrained_set_expansion.txt.tar.gz>`__
+A `pretrained model <https://d2zs9tzlek599f.cloudfront.net/models/term_set/enwiki-20171201_pretrained_set_expansion.txt.tar.gz>`__
 on English Wikipedia dump (``enwiki-20171201-pages-articles-multistream.xml.bz2``) is available under
 Apache 2.0 license. It has been trained with hyperparameters values
-recommended above. Full English Wikipedia `raw corpus <https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/term_set/enwiki-20171201.txt.gz>`_ and
-`marked corpus <https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/term_set/enwiki-20171201_spacy_marked.txt.tar.gz>`_
+recommended above. Full English Wikipedia `raw corpus <https://d2zs9tzlek599f.cloudfront.net/models/term_set/enwiki-20171201.txt.gz>`_ and
+`marked corpus <https://d2zs9tzlek599f.cloudfront.net/models/term_set/enwiki-20171201_spacy_marked.txt.tar.gz>`_
 are also available under the
 `Creative Commons Attribution-Share-Alike 3.0 License <https://creativecommons.org/licenses/by-sa/3.0/>`__.
 
-A `pretrained model with grouping <https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/term_set/enwiki-20171201_grouping_pretrained_set_expansion.tar.gz>`__
+A `pretrained model with grouping <https://d2zs9tzlek599f.cloudfront.net/models/term_set/enwiki-20171201_grouping_pretrained_set_expansion.tar.gz>`__
 on the same English Wikipedia dump is also
 available under
 Apache 2.0 license. It has been trained with hyperparameters values
-recommended above. `Marked corpus <https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/term_set/enwiki-20171201_grouping_marked.txt.tar.gz>`_
+recommended above. `Marked corpus <https://d2zs9tzlek599f.cloudfront.net/models/term_set/enwiki-20171201_grouping_marked.txt.tar.gz>`_
 is also available under the
 `Creative Commons Attribution-Share-Alike 3.0 License <https://creativecommons.org/licenses/by-sa/3.0/>`__.
 
diff --git a/docs-source/source/trend_analysis.rst b/docs-source/source/trend_analysis.rst
@@ -37,7 +37,7 @@ In this stage, the algorithm will also train a W2V model on the joint corpora to
 In the second stage the topic lists are being compared and analyzed.
 Finally the UI reads the analysis data and generates automatic reports for extracted topics, “Hot” and “Cold” trends, and topic clustering in 2D space.
 
-The noun phrase extraction module is using a pre-trained `model <https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/chunker/model.h5>`__ which is available under the Apache 2.0 license.
+The noun phrase extraction module is using a pre-trained `model <https://d2zs9tzlek599f.cloudfront.net/models/chunker/model.h5>`__ which is available under the Apache 2.0 license.
 
 Flow diagram
 ============
diff --git a/examples/reading_comprehension/match_lstm_mrc/machine_comprehension_api.py b/examples/reading_comprehension/match_lstm_mrc/machine_comprehension_api.py
@@ -85,13 +85,13 @@ def download_model(self):
             if self.prompt is True:
                 license_prompt(
                     "mrc_data",
-                    "https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/mrc"
+                    "https://d2zs9tzlek599f.cloudfront.net/models/mrc"
                     "/mrc_data.zip",
                     self.data_dir,
                 )
                 license_prompt(
                     "mrc_model",
-                    "https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/mrc"
+                    "https://d2zs9tzlek599f.cloudfront.net/models/mrc"
                     "/mrc_model.zip",
                     self.model_dir,
                 )
@@ -100,12 +100,12 @@ def download_model(self):
             makedirs(self.data_dir, exist_ok=True)
             makedirs(self.model_dir, exist_ok=True)
             download_unlicensed_file(
-                "https://s3-us-west-2.amazonaws.com/nlp-architect-data" "/models/mrc/",
+                "https://d2zs9tzlek599f.cloudfront.net" "/models/mrc/",
                 "mrc_data.zip",
                 data_zipfile,
             )
             download_unlicensed_file(
-                "https://s3-us-west-2.amazonaws.com/nlp-architect-data" "/models/mrc/",
+                "https://d2zs9tzlek599f.cloudfront.net" "/models/mrc/",
                 "mrc_model.zip",
                 model_zipfile,
             )
diff --git a/examples/sparse_gnmt/README.md b/examples/sparse_gnmt/README.md
@@ -32,8 +32,8 @@ You can use these models to [Run Inference with Pre-Trained Model](#run-inferenc
 | Model                      | Sparsity | BLEU| Non-Zero Parameters | Data Type |
 |----------------------------|:--------:|:----:|:-------------------:|:---------:|
 | Baseline                   |    0%    | 29.9 |        ~210M        |  Float32  |
-| [Sparse](https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/sparse_gnmt/gnmt_sparse.zip)                     |    90%   | 28.4 |         ~22M        |  Float32  |
-| [2x2 Block Sparse](https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/sparse_gnmt/gnmt_blocksparse2x2.zip)           |    90%   | 27.8 |         ~22M        |  Float32  |
+| [Sparse](https://d2zs9tzlek599f.cloudfront.net/models/sparse_gnmt/gnmt_sparse.zip)                     |    90%   | 28.4 |         ~22M        |  Float32  |
+| [2x2 Block Sparse](https://d2zs9tzlek599f.cloudfront.net/models/sparse_gnmt/gnmt_blocksparse2x2.zip)           |    90%   | 27.8 |         ~22M        |  Float32  |
 | Quantized Sparse           |    90%   | 28.4 |         ~22M        |  Integer8 |
 | Quantized 2x2 Block Sparse |    90%   | 27.6 |         ~22M        |  Integer8 |
 
@@ -92,7 +92,7 @@ Follow these instructions in order to use our pre-trained models:
 ```
 
 # Download pre-trained model zip file, e.g. gnmt_sparse.zip
-wget https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/sparse_gnmt/gnmt_sparse.zip
+wget https://d2zs9tzlek599f.cloudfront.net/models/sparse_gnmt/gnmt_sparse.zip
 
 # Unzip checkpoint + vocabulary files
 unzip gnmt_sparse.zip -d /tmp/gnmt_sparse_checkpoint
diff --git a/nlp_architect/api/intent_extraction_api.py b/nlp_architect/api/intent_extraction_api.py
@@ -73,12 +73,12 @@ def _download_pretrained_model(prompt=True):
                 if agreed is False:
                     sys.exit(0)
             download_unlicensed_file(
-                "https://s3-us-west-2.amazonaws.com/nlp-architect-data" "/models/intent/",
+                "https://d2zs9tzlek599f.cloudfront.net" "/models/intent/",
                 "model_info.dat",
                 IntentExtractionApi.pretrained_model_info,
             )
             download_unlicensed_file(
-                "https://s3-us-west-2.amazonaws.com/nlp-architect-data" "/models/intent/",
+                "https://d2zs9tzlek599f.cloudfront.net" "/models/intent/",
                 "model.h5",
                 IntentExtractionApi.pretrained_model,
             )
diff --git a/nlp_architect/api/ner_api.py b/nlp_architect/api/ner_api.py
@@ -72,12 +72,12 @@ def _download_pretrained_model(self, prompt=True):
                 if agreed is False:
                     sys.exit(0)
             download_unlicensed_file(
-                "https://s3-us-west-2.amazonaws.com/nlp-architect-data" "/models/ner/",
+                "https://d2zs9tzlek599f.cloudfront.net" "/models/ner/",
                 "model_v4.h5",
                 self.pretrained_model,
             )
             download_unlicensed_file(
-                "https://s3-us-west-2.amazonaws.com/nlp-architect-data" "/models/ner/",
+                "https://d2zs9tzlek599f.cloudfront.net" "/models/ner/",
                 "model_info_v4.dat",
                 self.pretrained_model_info,
             )
diff --git a/nlp_architect/models/absa/utils.py b/nlp_architect/models/absa/utils.py
@@ -37,7 +37,7 @@ def _download_pretrained_rerank_model(rerank_model_full_path):
         makedirs(rerank_model_dir, exist_ok=True)
         print("dowloading pre-trained reranking model..")
         download_unlicensed_file(
-            "https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/" "absa/",
+            "https://d2zs9tzlek599f.cloudfront.net/models/" "absa/",
             "rerank_model.h5",
             rerank_model_full_path,
         )
diff --git a/nlp_architect/models/pretrained_models.py b/nlp_architect/models/pretrained_models.py
@@ -19,7 +19,7 @@
 
 from nlp_architect import LIBRARY_OUT
 
-S3_PREFIX = "https://s3-us-west-2.amazonaws.com/nlp-architect-data/"
+S3_PREFIX = "https://d2zs9tzlek599f.cloudfront.net/"
 
 
 class PretrainedModel:
diff --git a/nlp_architect/pipelines/spacy_bist.py b/nlp_architect/pipelines/spacy_bist.py
@@ -162,7 +162,7 @@ def _download_pretrained_model():
         zip_path = SpacyBISTParser.dir / "bist-pretrained.zip"
         makedirs(SpacyBISTParser.dir, exist_ok=True)
         download_unlicensed_file(
-            "https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/dep_parse/",
+            "https://d2zs9tzlek599f.cloudfront.net/models/dep_parse/",
             "bist-pretrained.zip",
             zip_path,
         )
diff --git a/solutions/absa_solution/test_absa_solution.py b/solutions/absa_solution/test_absa_solution.py
@@ -27,7 +27,7 @@
 def test_solution(generate_new=False):
     lexicons_dir = Path(LIBRARY_ROOT) / "examples" / "absa"
     expected_dir = Path(LIBRARY_ROOT) / "tests" / "fixtures" / "data" / "absa_solution"
-    data_url = "https://s3-us-west-2.amazonaws.com/nlp-architect-data/tests/"
+    data_url = "https://d2zs9tzlek599f.cloudfront.net/tests/"
     parsed_data = download_unzip(
         data_url, "tripadvisor_test_parsed.zip", SENTIMENT_OUT / "test" / "tripadvisor_test_parsed"
     )
diff --git a/solutions/set_expansion/README.md b/solutions/set_expansion/README.md
@@ -28,7 +28,7 @@ provide a subset of English Wikipedia at datasets/wikipedia as a sample corpus u
 [Creative Commons Attribution-Share-Alike 3.0 License](https://creativecommons.org/licenses/by-sa/3.0/)- Copyright 2018 Wikimedia Foundation.
 
 The output of this step is the marked corpus where noun phrases are marked with the marking character (default: "\_") as described in the [NLP Architect np2vec module documentation](http://nlp_architect.nervanasys.com/np2vec.html).
-The Noun phrase extraction module is using a pre-trained [model](https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/chunker/model.h5) which is available under the Apache 2.0 license. 
+The Noun phrase extraction module is using a pre-trained [model](https://d2zs9tzlek599f.cloudfront.net/models/chunker/model.h5) which is available under the Apache 2.0 license. 
 
 This is done by running:
 ```
@@ -58,19 +58,19 @@ size, min_count, window and hs hyperparameters. Please refer to the np2vec modul
 python examples/np2vec/train.py --size 100 --min_count 10 --window 10 --hs 0 --corpus MARKED_TRAINING_CORPUS --np2vec_model_file MODEL_PATH --corpus_format txt
 ```
 
-A [pretrained model](https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/term_set/enwiki-20171201_pretrained_set_expansion.txt.tar.gz) 
+A [pretrained model](https://d2zs9tzlek599f.cloudfront.net/models/term_set/enwiki-20171201_pretrained_set_expansion.txt.tar.gz) 
 on English Wikipedia dump (enwiki-20171201-pages-articles-multistream.xml.bz2) is available under the
 Apache 2.0 license. It has been trained with hyperparameters values
-recommended above. Full English Wikipedia [raw corpus](https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/term_set/enwiki-20171201.txt.gz) and 
-[marked corpus](https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/term_set/enwiki-20171201_spacy_marked.txt.tar.gz) 
+recommended above. Full English Wikipedia [raw corpus](https://d2zs9tzlek599f.cloudfront.net/models/term_set/enwiki-20171201.txt.gz) and 
+[marked corpus](https://d2zs9tzlek599f.cloudfront.net/models/term_set/enwiki-20171201_spacy_marked.txt.tar.gz) 
 are also available under the  
 [Creative Commons Attribution-Share-Alike 3.0 License](https://creativecommons.org/licenses/by-sa/3.0/).
 
-A [pretrained model with grouping](https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/term_set/enwiki-20171201_grouping_pretrained_set_expansion.tar.gz)
+A [pretrained model with grouping](https://d2zs9tzlek599f.cloudfront.net/models/term_set/enwiki-20171201_grouping_pretrained_set_expansion.tar.gz)
 on the same English Wikipedia dump is also
 available under
 Apache 2.0 license. It has been trained with hyperparameters values
-recommended above. [Marked corpus](https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/term_set/enwiki-20171201_grouping_marked.txt.tar.gz)
+recommended above. [Marked corpus](https://d2zs9tzlek599f.cloudfront.net/models/term_set/enwiki-20171201_grouping_marked.txt.tar.gz)
 is also available under the
 [Creative Commons Attribution-Share-Alike 3.0 License](https://creativecommons.org/licenses/by-sa/3.0/).
 
diff --git a/solutions/set_expansion/prepare_data.py b/solutions/set_expansion/prepare_data.py
@@ -37,7 +37,7 @@
 id2group = {}
 id2rep = {}
 np2count = {}
-nlp_chunker_url = "https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/chunker/"
+nlp_chunker_url = "https://d2zs9tzlek599f.cloudfront.net/models/chunker/"
 chunker_path = str(LIBRARY_OUT / "chunker-pretrained")
 chunker_model_dat_file = "model_info.dat.params"
 chunker_model_file = "model.h5"
diff --git a/solutions/trend_analysis/README.md b/solutions/trend_analysis/README.md
@@ -16,7 +16,7 @@ In this stage, the algorithm will also train a W2V model on the joint corpora to
 In the second stage the topic lists are being compared and analysed.
 Finally the UI reads the analysis data and generates automatic reports for extracted topics, “Hot” and “Cold” trends, and topic clustering in 2D space.
 
-The noun phrase extraction module is using a pre-trained [model](https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/chunker/model.h5) which is available under the Apache 2.0 license. 
+The noun phrase extraction module is using a pre-trained [model](https://d2zs9tzlek599f.cloudfront.net/models/chunker/model.h5) which is available under the Apache 2.0 license. 
 
 ### Flow diagram
 
diff --git a/solutions/trend_analysis/np_scorer.py b/solutions/trend_analysis/np_scorer.py
@@ -24,7 +24,7 @@
 from nlp_architect.utils.text import SpacyInstance
 from .scoring_utils import TextSpanScoring
 
-nlp_chunker_url = "https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/chunker/"
+nlp_chunker_url = "https://d2zs9tzlek599f.cloudfront.net/models/chunker/"
 chunker_model_dat_file = "model_info.dat.params"
 chunker_model_file = "model.h5"
 chunker_local_path = str(LIBRARY_OUT / "chunker-pretrained")
diff --git a/tests/test_spacy_np_annotator.py b/tests/test_spacy_np_annotator.py
@@ -27,7 +27,7 @@
 from nlp_architect.utils.io import download_unlicensed_file
 from nlp_architect.utils.text import SpacyInstance, try_to_load_spacy
 
-MODEL_URL = "https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/chunker/"
+MODEL_URL = "https://d2zs9tzlek599f.cloudfront.net/models/chunker/"
 MODEL_FILE = "model.h5"
 MODEL_INFO = "model_info.dat.params"
 local_models_path = path.join(path.dirname(path.realpath(__file__)), "fixtures/data/chunker")
diff --git a/tutorials/Term_Set_Expansion/term_set_expansion.ipynb b/tutorials/Term_Set_Expansion/term_set_expansion.ipynb
@@ -73,7 +73,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "url = 'https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/term_set/enwiki-20171201_pretrained_set_expansion.txt.tar.gz'\n",
+    "url = 'https://d2zs9tzlek599f.cloudfront.net/models/term_set/enwiki-20171201_pretrained_set_expansion.txt.tar.gz'\n",
     "wget.download(url, 'enwiki-20171201_pretrained_set_expansion.txt.tar.gz')\n",
     "import tarfile \n",
     "tarf = tarfile.open('enwiki-20171201_pretrained_set_expansion.txt.tar.gz') \n",

Original file line number	Diff line number	Diff line change
`@@ -37,7 +37,7 @@ def _download_pretrained_rerank_model(rerank_model_full_path):`
`37`	`37`	`makedirs(rerank_model_dir, exist_ok=True)`
`38`	`38`	`print("dowloading pre-trained reranking model..")`
`39`	`39`	`download_unlicensed_file(`
`40`		`- "https://s3-us-west-2.amazonaws.com/nlp-architect-data/models/" "absa/",`
	`40`	`+ "https://d2zs9tzlek599f.cloudfront.net/models/" "absa/",`
`41`	`41`	`"rerank_model.h5",`
`42`	`42`	`rerank_model_full_path,`
`43`	`43`	`)`