From bdf2ba61577a0aed3d5b259c249dccc2c81b6761 Mon Sep 17 00:00:00 2001
From: Daniel King <daking@hey.com>
Date: Wed, 17 Aug 2022 14:03:12 -0700
Subject: [PATCH 01/17] update spacy and scispacy versions

---
 requirements.in     | 2 +-
 scispacy/version.py | 2 +-
 setup.py            | 2 +-
 3 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/requirements.in b/requirements.in
index 8351a3a..3cc4a36 100644
--- a/requirements.in
+++ b/requirements.in
@@ -1,5 +1,5 @@
 numpy
-spacy>=3.2.0,<3.3.0
+spacy>=3.4.0,<3.5.0
 spacy-lookups-data
 pandas
 requests>=2.0.0,<3.0.0
diff --git a/scispacy/version.py b/scispacy/version.py
index c8b11da..b7139a8 100644
--- a/scispacy/version.py
+++ b/scispacy/version.py
@@ -1,6 +1,6 @@
 _MAJOR = "0"
 _MINOR = "5"
-_REVISION = "0"
+_REVISION = "1"
 
 VERSION_SHORT = "{0}.{1}".format(_MAJOR, _MINOR)
 VERSION = "{0}.{1}.{2}".format(_MAJOR, _MINOR, _REVISION)
diff --git a/setup.py b/setup.py
index c972d5c..91373dc 100644
--- a/setup.py
+++ b/setup.py
@@ -41,7 +41,7 @@
     packages=find_packages(exclude=["*.tests", "*.tests.*", "tests.*", "tests"]),
     license="Apache",
     install_requires=[
-        "spacy>=3.2.0,<3.3.0",
+        "spacy>=3.4.0,<3.5.0",
         "requests>=2.0.0,<3.0.0",
         "conllu",
         "numpy",

From bed5d79325fa7f0fa520c741a1e6fefacaf619e2 Mon Sep 17 00:00:00 2001
From: Daniel King <daking@hey.com>
Date: Wed, 17 Aug 2022 14:04:55 -0700
Subject: [PATCH 02/17] Update configs with new tagger version

---
 configs/base_parser_tagger.cfg         | 3 ++-
 configs/base_parser_tagger_scibert.cfg | 3 ++-
 2 files changed, 4 insertions(+), 2 deletions(-)

diff --git a/configs/base_parser_tagger.cfg b/configs/base_parser_tagger.cfg
index cec9e47..738c85c 100644
--- a/configs/base_parser_tagger.cfg
+++ b/configs/base_parser_tagger.cfg
@@ -55,8 +55,9 @@ upstream = "*"
 factory = "tagger"
 
 [components.tagger.model]
-@architectures = "spacy.Tagger.v1"
+@architectures = "spacy.Tagger.v2"
 nO = null
+normalize = False
 
 [components.tagger.model.tok2vec]
 @architectures = "spacy.Tok2VecListener.v1"
diff --git a/configs/base_parser_tagger_scibert.cfg b/configs/base_parser_tagger_scibert.cfg
index 016a395..65b8f2f 100644
--- a/configs/base_parser_tagger_scibert.cfg
+++ b/configs/base_parser_tagger_scibert.cfg
@@ -54,8 +54,9 @@ upstream = "*"
 factory = "tagger"
 
 [components.tagger.model]
-@architectures = "spacy.Tagger.v1"
+@architectures = "spacy.Tagger.v2"
 nO = null
+normalize = False
 
 [components.tagger.model.tok2vec]
 @architectures = "spacy-transformers.TransformerListener.v1"

From a0d54b149cc66cdfbc8ba6e00c9cb9d9decda4eb Mon Sep 17 00:00:00 2001
From: Daniel King <daking@hey.com>
Date: Wed, 17 Aug 2022 14:13:45 -0700
Subject: [PATCH 03/17] Add gpu args everywhere

---
 project.yml | 84 ++++++++++++++++++++++++++---------------------------
 1 file changed, 41 insertions(+), 43 deletions(-)

diff --git a/project.yml b/project.yml
index 2f11d1a..42eae35 100644
--- a/project.yml
+++ b/project.yml
@@ -2,8 +2,8 @@ title: "scispaCy pipeline"
 description: "All the steps needed in the scispaCy pipeline"
 
 vars:
-  version_string: "0.5.0"
-  gpu_id: "0"
+  version_string: "0.5.1"
+  gpu_id: 0
   freqs_loc_s3: "s3://ai2-s2-scispacy/data/gorc_subset.freqs"
   freqs_loc_local: "assets/gorc_subset.freqs"
   vectors_loc_s3: "s3://ai2-s2-scispacy/data/pubmed_with_header.txt.gz"
@@ -260,7 +260,7 @@ commands:
   - name: parser-tagger-train-sm
     help: "Train the base models"
     script:
-      - "spacy train ${vars.parser_tagger_config_loc} --output ${vars.parser_tagger_sm_loc} --code ${vars.code_loc} --paths.vocab_path ${vars.vocab_sm_loc} --vars.include_static_vectors False"
+      - "spacy train ${vars.parser_tagger_config_loc} --output ${vars.parser_tagger_sm_loc} --code ${vars.code_loc} --paths.vocab_path ${vars.vocab_sm_loc} --vars.include_static_vectors False --gpu-id ${vars.gpu_id}"
     deps:
       - "${vars.parser_tagger_config_loc}"
       - "${vars.genia_train_spacy_loc}"
@@ -273,7 +273,7 @@ commands:
   - name: parser-tagger-train-md
     help: "Train the base models"
     script:
-      - "spacy train ${vars.parser_tagger_config_loc} --output ${vars.parser_tagger_md_loc} --code ${vars.code_loc} --paths.vectors ${vars.vectors_md_loc} --paths.vocab_path ${vars.vocab_md_loc} --vars.include_static_vectors True"
+      - "spacy train ${vars.parser_tagger_config_loc} --output ${vars.parser_tagger_md_loc} --code ${vars.code_loc} --paths.vectors ${vars.vectors_md_loc} --paths.vocab_path ${vars.vocab_md_loc} --vars.include_static_vectors True --gpu-id ${vars.gpu_id}"
     deps:
       - "${vars.parser_tagger_config_loc}"
       - "${vars.genia_train_spacy_loc}"
@@ -287,7 +287,7 @@ commands:
   - name: parser-tagger-train-lg
     help: "Train the base models"
     script:
-      - "spacy train ${vars.parser_tagger_config_loc} --output ${vars.parser_tagger_lg_loc} --code ${vars.code_loc} --paths.vectors ${vars.vectors_lg_loc} --paths.vocab_path ${vars.vocab_lg_loc} --vars.include_static_vectors True"
+      - "spacy train ${vars.parser_tagger_config_loc} --output ${vars.parser_tagger_lg_loc} --code ${vars.code_loc} --paths.vectors ${vars.vectors_lg_loc} --paths.vocab_path ${vars.vocab_lg_loc} --vars.include_static_vectors True --gpu-id ${vars.gpu_id}"
     deps:
       - "${vars.parser_tagger_config_loc}"
       - "${vars.genia_train_spacy_loc}"
@@ -314,7 +314,7 @@ commands:
   - name: ner-train-sm
     help: "Train the main ner"
     script:
-      - "spacy train ${vars.ner_config_loc} --output ${vars.ner_sm_loc} --code ${vars.code_loc} --paths.parser_tagger_path ${vars.parser_tagger_sm_loc}/model-best --paths.vocab_path ${vars.vocab_sm_loc} --vars.include_static_vectors False"
+      - "spacy train ${vars.ner_config_loc} --output ${vars.ner_sm_loc} --code ${vars.code_loc} --paths.parser_tagger_path ${vars.parser_tagger_sm_loc}/model-best --paths.vocab_path ${vars.vocab_sm_loc} --vars.include_static_vectors False --gpu-id ${vars.gpu_id}"
     deps:
       - "${vars.ner_config_loc}"
       - "${vars.parser_tagger_sm_loc}/model-best"
@@ -325,7 +325,7 @@ commands:
   - name: ner-train-md
     help: "Train the main ner"
     script:
-      - "spacy train ${vars.ner_config_loc} --output ${vars.ner_md_loc} --code ${vars.code_loc} --paths.vectors ${vars.vectors_md_loc} --paths.parser_tagger_path ${vars.parser_tagger_md_loc}/model-best --paths.vocab_path ${vars.vocab_md_loc} --vars.include_static_vectors True"
+      - "spacy train ${vars.ner_config_loc} --output ${vars.ner_md_loc} --code ${vars.code_loc} --paths.vectors ${vars.vectors_md_loc} --paths.parser_tagger_path ${vars.parser_tagger_md_loc}/model-best --paths.vocab_path ${vars.vocab_md_loc} --vars.include_static_vectors True --gpu-id ${vars.gpu_id}"
     deps:
       - "${vars.ner_config_loc}"
       - "${vars.parser_tagger_md_loc}/model-best"
@@ -337,7 +337,7 @@ commands:
   - name: ner-train-lg
     help: "Train the main ner"
     script:
-      - "spacy train ${vars.ner_config_loc} --output ${vars.ner_lg_loc} --code ${vars.code_loc} --paths.vectors ${vars.vectors_lg_loc} --paths.parser_tagger_path ${vars.parser_tagger_lg_loc}/model-best --paths.vocab_path ${vars.vocab_lg_loc} --vars.include_static_vectors True"
+      - "spacy train ${vars.ner_config_loc} --output ${vars.ner_lg_loc} --code ${vars.code_loc} --paths.vectors ${vars.vectors_lg_loc} --paths.parser_tagger_path ${vars.parser_tagger_lg_loc}/model-best --paths.vocab_path ${vars.vocab_lg_loc} --vars.include_static_vectors True --gpu-id ${vars.gpu_id}"
     deps:
       - "${vars.ner_config_loc}"
       - "${vars.parser_tagger_lg_loc}/model-best"
@@ -360,10 +360,10 @@ commands:
   - name: ner-train-specialized
     help: "Train the specialized NER models"
     script:
-      - "spacy train ${vars.specialized_ner_config_loc} --output ${vars.bc5cdr_md_loc} --code ${vars.code_loc} --paths.vectors ${vars.vectors_md_loc} --paths.parser_tagger_path ${vars.parser_tagger_md_loc}/model-best --paths.train_path ${vars.bc5cdr_loc_local}/train.tsv --paths.dev_path ${vars.bc5cdr_loc_local}/devel.tsv --paths.vocab_path ${vars.vocab_md_loc} --vars.include_static_vectors True"
-      - "spacy train ${vars.specialized_ner_config_loc} --output ${vars.bionlp13cg_md_loc} --code ${vars.code_loc} --paths.vectors ${vars.vectors_md_loc} --paths.parser_tagger_path ${vars.parser_tagger_md_loc}/model-best --paths.train_path ${vars.bionlp13cg_loc_local}/train.tsv --paths.dev_path ${vars.bionlp13cg_loc_local}/devel.tsv --paths.vocab_path ${vars.vocab_md_loc} --vars.include_static_vectors True"
-      - "spacy train ${vars.specialized_ner_config_loc} --output ${vars.craft_md_loc} --code ${vars.code_loc} --paths.vectors ${vars.vectors_md_loc} --paths.parser_tagger_path ${vars.parser_tagger_md_loc}/model-best --paths.train_path ${vars.craft_loc_local}/train.tsv --paths.dev_path ${vars.craft_loc_local}/devel.tsv --paths.vocab_path ${vars.vocab_md_loc} --vars.include_static_vectors True"
-      - "spacy train ${vars.specialized_ner_config_loc} --output ${vars.jnlpba_md_loc} --code ${vars.code_loc} --paths.vectors ${vars.vectors_md_loc} --paths.parser_tagger_path ${vars.parser_tagger_md_loc}/model-best --paths.train_path ${vars.jnlpba_loc_local}/train.tsv --paths.dev_path ${vars.jnlpba_loc_local}/devel.tsv --paths.vocab_path ${vars.vocab_md_loc} --vars.include_static_vectors True"
+      - "spacy train ${vars.specialized_ner_config_loc} --output ${vars.bc5cdr_md_loc} --code ${vars.code_loc} --paths.vectors ${vars.vectors_md_loc} --paths.parser_tagger_path ${vars.parser_tagger_md_loc}/model-best --paths.train_path ${vars.bc5cdr_loc_local}/train.tsv --paths.dev_path ${vars.bc5cdr_loc_local}/devel.tsv --paths.vocab_path ${vars.vocab_md_loc} --vars.include_static_vectors True --gpu-id ${vars.gpu_id}"
+      - "spacy train ${vars.specialized_ner_config_loc} --output ${vars.bionlp13cg_md_loc} --code ${vars.code_loc} --paths.vectors ${vars.vectors_md_loc} --paths.parser_tagger_path ${vars.parser_tagger_md_loc}/model-best --paths.train_path ${vars.bionlp13cg_loc_local}/train.tsv --paths.dev_path ${vars.bionlp13cg_loc_local}/devel.tsv --paths.vocab_path ${vars.vocab_md_loc} --vars.include_static_vectors True --gpu-id ${vars.gpu_id}"
+      - "spacy train ${vars.specialized_ner_config_loc} --output ${vars.craft_md_loc} --code ${vars.code_loc} --paths.vectors ${vars.vectors_md_loc} --paths.parser_tagger_path ${vars.parser_tagger_md_loc}/model-best --paths.train_path ${vars.craft_loc_local}/train.tsv --paths.dev_path ${vars.craft_loc_local}/devel.tsv --paths.vocab_path ${vars.vocab_md_loc} --vars.include_static_vectors True --gpu-id ${vars.gpu_id}"
+      - "spacy train ${vars.specialized_ner_config_loc} --output ${vars.jnlpba_md_loc} --code ${vars.code_loc} --paths.vectors ${vars.vectors_md_loc} --paths.parser_tagger_path ${vars.parser_tagger_md_loc}/model-best --paths.train_path ${vars.jnlpba_loc_local}/train.tsv --paths.dev_path ${vars.jnlpba_loc_local}/devel.tsv --paths.vocab_path ${vars.vocab_md_loc} --vars.include_static_vectors True --gpu-id ${vars.gpu_id}"
     deps:
       - "${vars.corpus_pubtator_loc_local}"
       - "${vars.bc5cdr_loc_local}/train.tsv"
@@ -389,8 +389,8 @@ commands:
   - name: evaluate-parser-tagger-sm
     help: "Evaluate the parser and tagger"
     script:
-      - "spacy evaluate ${vars.parser_tagger_sm_loc}/model-best ${vars.genia_test_spacy_loc} --output ${vars.parser_tagger_sm_loc}/model_best_results.json"
-      - "spacy evaluate ${vars.parser_tagger_sm_loc}/model-best ${vars.ontonotes_test_spacy_loc} --output ${vars.parser_tagger_sm_loc}/model_best_results_onto.json"
+      - "spacy evaluate ${vars.parser_tagger_sm_loc}/model-best ${vars.genia_test_spacy_loc} --output ${vars.parser_tagger_sm_loc}/model_best_results.json --gpu-id ${vars.gpu_id}"
+      - "spacy evaluate ${vars.parser_tagger_sm_loc}/model-best ${vars.ontonotes_test_spacy_loc} --output ${vars.parser_tagger_sm_loc}/model_best_results_onto.json --gpu-id ${vars.gpu_id}"
     deps:
       - "${vars.parser_tagger_sm_loc}/model-best"
       - "${vars.genia_test_spacy_loc}"
@@ -402,8 +402,8 @@ commands:
   - name: evaluate-parser-tagger-md
     help: "Evaluate the parser and tagger"
     script:
-      - "spacy evaluate ${vars.parser_tagger_md_loc}/model-best ${vars.genia_test_spacy_loc} --output ${vars.parser_tagger_md_loc}/model_best_results.json"
-      - "spacy evaluate ${vars.parser_tagger_md_loc}/model-best ${vars.ontonotes_test_spacy_loc} --output ${vars.parser_tagger_md_loc}/model_best_results_onto.json"
+      - "spacy evaluate ${vars.parser_tagger_md_loc}/model-best ${vars.genia_test_spacy_loc} --output ${vars.parser_tagger_md_loc}/model_best_results.json --gpu-id ${vars.gpu_id}"
+      - "spacy evaluate ${vars.parser_tagger_md_loc}/model-best ${vars.ontonotes_test_spacy_loc} --output ${vars.parser_tagger_md_loc}/model_best_results_onto.json --gpu-id ${vars.gpu_id}"
     deps:
       - "${vars.parser_tagger_md_loc}/model-best"
       - "${vars.genia_test_spacy_loc}"
@@ -415,8 +415,8 @@ commands:
   - name: evaluate-parser-tagger-lg
     help: "Evaluate the parser and tagger"
     script:
-      - "spacy evaluate ${vars.parser_tagger_lg_loc}/model-best ${vars.genia_test_spacy_loc} --output ${vars.parser_tagger_lg_loc}/model_best_results.json"
-      - "spacy evaluate ${vars.parser_tagger_lg_loc}/model-best ${vars.ontonotes_test_spacy_loc} --output ${vars.parser_tagger_lg_loc}/model_best_results_onto.json"
+      - "spacy evaluate ${vars.parser_tagger_lg_loc}/model-best ${vars.genia_test_spacy_loc} --output ${vars.parser_tagger_lg_loc}/model_best_results.json --gpu-id ${vars.gpu_id}"
+      - "spacy evaluate ${vars.parser_tagger_lg_loc}/model-best ${vars.ontonotes_test_spacy_loc} --output ${vars.parser_tagger_lg_loc}/model_best_results_onto.json --gpu-id ${vars.gpu_id}"
     deps:
       - "${vars.parser_tagger_lg_loc}/model-best"
       - "${vars.genia_test_spacy_loc}"
@@ -428,8 +428,8 @@ commands:
   - name: evaluate-parser-tagger-scibert
     help: "Evaluate the parser and tagger scibert model"
     script:
-      - "spacy evaluate ${vars.parser_tagger_scibert_loc}/model-best ${vars.genia_test_spacy_loc} --output ${vars.parser_tagger_scibert_loc}/model_best_results.json --gpu-id ${vars.gpu_id}"
-      - "spacy evaluate ${vars.parser_tagger_scibert_loc}/model-best ${vars.ontonotes_test_spacy_loc} --output ${vars.parser_tagger_scibert_loc}/model_best_results_onto.json --gpu-id ${vars.gpu_id}"
+      - "spacy evaluate ${vars.parser_tagger_scibert_loc}/model-best ${vars.genia_test_spacy_loc} --output ${vars.parser_tagger_scibert_loc}/model_best_results.json --gpu-id ${vars.gpu_id} --gpu-id ${vars.gpu_id}"
+      - "spacy evaluate ${vars.parser_tagger_scibert_loc}/model-best ${vars.ontonotes_test_spacy_loc} --output ${vars.parser_tagger_scibert_loc}/model_best_results_onto.json --gpu-id ${vars.gpu_id} --gpu-id ${vars.gpu_id}"
     deps:
       - "${vars.parser_tagger_scibert_loc}/model-best"
       - "${vars.genia_test_spacy_loc}"
@@ -441,7 +441,7 @@ commands:
   - name: evaluate-ner-sm
     help: "Evaluate NER"
     script:
-      - "python scripts/evaluate_ner.py --model_path ${vars.ner_sm_loc}/model-best --dataset medmentions-test --output ${vars.ner_sm_loc}/model_best_results.json --med_mentions_folder_path assets/"
+      - "python scripts/evaluate_ner.py --model_path ${vars.ner_sm_loc}/model-best --dataset medmentions-test --output ${vars.ner_sm_loc}/model_best_results.json --med_mentions_folder_path assets/ --gpu_id ${vars.gpu_id}"
     deps:
       - "${vars.ner_sm_loc}"
       - "${vars.corpus_pubtator_loc_local}"
@@ -451,7 +451,7 @@ commands:
   - name: evaluate-ner-md
     help: "Evaluate NER"
     script:
-      - "python scripts/evaluate_ner.py --model_path ${vars.ner_md_loc}/model-best --dataset medmentions-test --output ${vars.ner_md_loc}/model_best_results.json --med_mentions_folder_path assets/"
+      - "python scripts/evaluate_ner.py --model_path ${vars.ner_md_loc}/model-best --dataset medmentions-test --output ${vars.ner_md_loc}/model_best_results.json --med_mentions_folder_path assets/ --gpu_id ${vars.gpu_id}"
     deps:
       - "${vars.ner_md_loc}"
       - "${vars.corpus_pubtator_loc_local}"
@@ -461,7 +461,7 @@ commands:
   - name: evaluate-ner-lg
     help: "Evaluate NER"
     script:
-      - "python scripts/evaluate_ner.py --model_path ${vars.ner_lg_loc}/model-best --dataset medmentions-test --output ${vars.ner_lg_loc}/model_best_results.json --med_mentions_folder_path assets/"
+      - "python scripts/evaluate_ner.py --model_path ${vars.ner_lg_loc}/model-best --dataset medmentions-test --output ${vars.ner_lg_loc}/model_best_results.json --med_mentions_folder_path assets/ --gpu_id ${vars.gpu_id}"
     deps:
       - "${vars.ner_lg_loc}"
       - "${vars.corpus_pubtator_loc_local}"
@@ -481,10 +481,10 @@ commands:
   - name: evaluate-specialized-ner
     help: "Evaluate specialize NER"
     script:
-      - "python scripts/evaluate_ner.py --model_path ${vars.bc5cdr_md_loc}/model-best --dataset ${vars.bc5cdr_loc_local}/test.tsv --output ${vars.bc5cdr_md_loc}/model_best_results.json"
-      - "python scripts/evaluate_ner.py --model_path ${vars.bionlp13cg_md_loc}/model-best --dataset ${vars.bionlp13cg_loc_local}/test.tsv --output ${vars.bionlp13cg_md_loc}/model_best_results.json"
-      - "python scripts/evaluate_ner.py --model_path ${vars.craft_md_loc}/model-best --dataset ${vars.craft_loc_local}/test.tsv --output ${vars.craft_md_loc}/model_best_results.json"
-      - "python scripts/evaluate_ner.py --model_path ${vars.jnlpba_md_loc}/model-best --dataset ${vars.jnlpba_loc_local}/test.tsv --output ${vars.jnlpba_md_loc}/model_best_results.json"
+      - "python scripts/evaluate_ner.py --model_path ${vars.bc5cdr_md_loc}/model-best --dataset ${vars.bc5cdr_loc_local}/test.tsv --output ${vars.bc5cdr_md_loc}/model_best_results.json --gpu_id ${vars.gpu_id}"
+      - "python scripts/evaluate_ner.py --model_path ${vars.bionlp13cg_md_loc}/model-best --dataset ${vars.bionlp13cg_loc_local}/test.tsv --output ${vars.bionlp13cg_md_loc}/model_best_results.json --gpu_id ${vars.gpu_id}"
+      - "python scripts/evaluate_ner.py --model_path ${vars.craft_md_loc}/model-best --dataset ${vars.craft_loc_local}/test.tsv --output ${vars.craft_md_loc}/model_best_results.json --gpu_id ${vars.gpu_id}"
+      - "python scripts/evaluate_ner.py --model_path ${vars.jnlpba_md_loc}/model-best --dataset ${vars.jnlpba_loc_local}/test.tsv --output ${vars.jnlpba_md_loc}/model_best_results.json --gpu_id ${vars.gpu_id}"
     deps:
       - "${vars.bc5cdr_md_loc}/model-best"
       - "${vars.bionlp13cg_md_loc}/model-best"
@@ -512,9 +512,9 @@ commands:
   - name: evaluate-package-sm
     help: "Evaluate the packaged models"
     script:
-      - "spacy evaluate ${vars.package_sm_loc} ${vars.genia_test_spacy_loc} --output packages/sm_genia_results.json"
-      - "spacy evaluate ${vars.package_sm_loc} ${vars.ontonotes_test_spacy_loc} --output packages/sm_onto_results.json"
-      - "python scripts/evaluate_ner.py --model_path ${vars.package_sm_loc} --dataset medmentions-test --output packages/sm_mm_results.json --med_mentions_folder_path assets/"
+      - "spacy evaluate ${vars.package_sm_loc} ${vars.genia_test_spacy_loc} --output packages/sm_genia_results.json --gpu-id ${vars.gpu_id}"
+      - "spacy evaluate ${vars.package_sm_loc} ${vars.ontonotes_test_spacy_loc} --output packages/sm_onto_results.json --gpu-id ${vars.gpu_id}"
+      - "python scripts/evaluate_ner.py --model_path ${vars.package_sm_loc} --dataset medmentions-test --output packages/sm_mm_results.json --med_mentions_folder_path assets/ --gpu_id ${vars.gpu_id}"
     deps:
       - "${vars.package_sm_loc}"
     outputs:
@@ -553,9 +553,9 @@ commands:
   - name: evaluate-package-md
     help: "Evaluate the packaged models"
     script:
-      - "spacy evaluate ${vars.package_md_loc} ${vars.genia_test_spacy_loc} --output packages/md_genia_results.json"
-      - "spacy evaluate ${vars.package_md_loc} ${vars.ontonotes_test_spacy_loc} --output packages/md_onto_results.json"
-      - "python scripts/evaluate_ner.py --model_path ${vars.package_md_loc} --dataset medmentions-test --output packages/md_mm_results.json --med_mentions_folder_path assets/"
+      - "spacy evaluate ${vars.package_md_loc} ${vars.genia_test_spacy_loc} --output packages/md_genia_results.json --gpu-id ${vars.gpu_id}"
+      - "spacy evaluate ${vars.package_md_loc} ${vars.ontonotes_test_spacy_loc} --output packages/md_onto_results.json --gpu-id ${vars.gpu_id}"
+      - "python scripts/evaluate_ner.py --model_path ${vars.package_md_loc} --dataset medmentions-test --output packages/md_mm_results.json --med_mentions_folder_path assets/ --gpu_id ${vars.gpu_id}"
     deps:
       - "${vars.package_md_loc}"
     outputs:
@@ -566,9 +566,9 @@ commands:
   - name: evaluate-package-lg
     help: "Evaluate the packaged models"
     script:
-      - "spacy evaluate ${vars.package_lg_loc} ${vars.genia_test_spacy_loc} --output packages/lg_genia_results.json"
-      - "spacy evaluate ${vars.package_lg_loc} ${vars.ontonotes_test_spacy_loc} --output packages/lg_onto_results.json"
-      - "python scripts/evaluate_ner.py --model_path ${vars.package_lg_loc} --dataset medmentions-test --output packages/lg_mm_results.json --med_mentions_folder_path assets/"
+      - "spacy evaluate ${vars.package_lg_loc} ${vars.genia_test_spacy_loc} --output packages/lg_genia_results.json --gpu-id ${vars.gpu_id}"
+      - "spacy evaluate ${vars.package_lg_loc} ${vars.ontonotes_test_spacy_loc} --output packages/lg_onto_results.json --gpu-id ${vars.gpu_id}"
+      - "python scripts/evaluate_ner.py --model_path ${vars.package_lg_loc} --dataset medmentions-test --output packages/lg_mm_results.json --med_mentions_folder_path assets/ --gpu_id ${vars.gpu_id}"
     deps:
       - "${vars.package_lg_loc}"
     outputs:
@@ -580,7 +580,7 @@ commands:
     help: "Evaluate the packaged scibert model"
     script:
       - "spacy evaluate ${vars.package_scibert_loc} ${vars.genia_test_spacy_loc} --output packages/scibert_genia_results.json --gpu-id ${vars.gpu_id}"
-      - "spacy evaluate ${vars.package_scibert_loc} ${vars.ontonotes_test_spacy_loc} --output packages/scibert_onto_results.json --gpu-id ${var.gpu_id}"
+      - "spacy evaluate ${vars.package_scibert_loc} ${vars.ontonotes_test_spacy_loc} --output packages/scibert_onto_results.json --gpu-id ${vars.gpu_id}"
       - "python scripts/evaluate_ner.py --model_path ${vars.package_scibert_loc} --dataset medmentions-test --output packages/scibert_mm_results.json --med_mentions_folder_path assets/ --gpu_id ${vars.gpu_id}"
     deps:
       - "${vars.package_scibert_loc}"
@@ -588,8 +588,6 @@ commands:
       - "packages/scibert_genia_results.json"
       - "packages/scibert_onto_results.json"
       - "packages/scibert_mm_results.json"
- 
-      
 
   - name: package-ner
     help: "Package the models"
@@ -612,10 +610,10 @@ commands:
   - name: evaluate-package-ner
     help: "Evaluate the packaged models"
     script:
-      - "python scripts/evaluate_ner.py --model_path ${vars.package_bc5cdr_loc} --dataset ${vars.bc5cdr_loc_local}/test.tsv --output packages/bc5cdr_results.json"
-      - "python scripts/evaluate_ner.py --model_path ${vars.package_bionlp13cg_loc} --dataset ${vars.bionlp13cg_loc_local}/test.tsv --output packages/bionlp13cg_results.json"
-      - "python scripts/evaluate_ner.py --model_path ${vars.package_craft_loc} --dataset ${vars.craft_loc_local}/test.tsv --output packages/craft_results.json"
-      - "python scripts/evaluate_ner.py --model_path ${vars.package_jnlpba_loc} --dataset ${vars.jnlpba_loc_local}/test.tsv --output packages/jnlpba_results.json"
+      - "python scripts/evaluate_ner.py --model_path ${vars.package_bc5cdr_loc} --dataset ${vars.bc5cdr_loc_local}/test.tsv --output packages/bc5cdr_results.json --gpu_id ${vars.gpu_id}"
+      - "python scripts/evaluate_ner.py --model_path ${vars.package_bionlp13cg_loc} --dataset ${vars.bionlp13cg_loc_local}/test.tsv --output packages/bionlp13cg_results.json --gpu_id ${vars.gpu_id}"
+      - "python scripts/evaluate_ner.py --model_path ${vars.package_craft_loc} --dataset ${vars.craft_loc_local}/test.tsv --output packages/craft_results.json --gpu_id ${vars.gpu_id}"
+      - "python scripts/evaluate_ner.py --model_path ${vars.package_jnlpba_loc} --dataset ${vars.jnlpba_loc_local}/test.tsv --output packages/jnlpba_results.json --gpu_id ${vars.gpu_id}"
     deps:
       - "${vars.package_bc5cdr_loc}"
       - "${vars.package_bionlp13cg_loc}"

From f3e25283ccfead9784a204d0c1aee6c600f7e71c Mon Sep 17 00:00:00 2001
From: Daniel King <daking@hey.com>
Date: Wed, 17 Aug 2022 14:15:19 -0700
Subject: [PATCH 04/17] Fix numpy warning in candidate generation

---
 scispacy/candidate_generation.py | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/scispacy/candidate_generation.py b/scispacy/candidate_generation.py
index 7bcf934..e7f3981 100644
--- a/scispacy/candidate_generation.py
+++ b/scispacy/candidate_generation.py
@@ -281,8 +281,12 @@ def nmslib_knn_with_zero_vectors(
         neighbors.append([])
         distances.append([])
         # interleave `neighbors` and Nones in `extended_neighbors`
-        extended_neighbors[empty_vectors_boolean_flags] = numpy.array(neighbors)[:-1]
-        extended_distances[empty_vectors_boolean_flags] = numpy.array(distances)[:-1]
+        extended_neighbors[empty_vectors_boolean_flags] = numpy.array(
+            neighbors, dtype=object
+        )[:-1]
+        extended_distances[empty_vectors_boolean_flags] = numpy.array(
+            distances, dtype=object
+        )[:-1]
 
         return extended_neighbors, extended_distances
 

From a5a826cb89cff12ca756309af7159c9830763e1f Mon Sep 17 00:00:00 2001
From: Daniel King <daking@hey.com>
Date: Wed, 17 Aug 2022 14:20:46 -0700
Subject: [PATCH 05/17] Add scibert to all workflow

---
 project.yml | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/project.yml b/project.yml
index 42eae35..75cf02f 100644
--- a/project.yml
+++ b/project.yml
@@ -131,25 +131,31 @@ workflows:
     - parser-tagger-train-sm
     - parser-tagger-train-md
     - parser-tagger-train-lg
+    - parser-tagger-train-scibert
     - ner-train-sm
     - ner-train-md
     - ner-train-lg
     - ner-train-specialized
+    - ner-train-scibert
     - evaluate-parser-tagger-sm
     - evaluate-parser-tagger-md
     - evaluate-parser-tagger-lg
+    - evaluate-parser-tagger-scibert
     - evaluate-ner-sm
     - evaluate-ner-md
     - evaluate-ner-lg
     - evaluate-specialized-ner
+    - evaluate-ner-scibert
     - package-sm
     - package-md
     - package-lg
     - package-ner
+    - package-scibert
     - evaluate-package-sm
     - evaluate-package-md
     - evaluate-package-lg
     - evaluate-package-ner
+    - evaluate-package-scibert
 
 commands:
   - name: download

From 483a6a954b32ee4b2020de723ae34833d4b1809d Mon Sep 17 00:00:00 2001
From: Daniel King <daking@hey.com>
Date: Sat, 27 Aug 2022 21:35:25 -0700
Subject: [PATCH 06/17] Add convenience scripts for testing models before
 release

---
 scripts/install_local_packages.py   | 33 +++++++++++
 scripts/smoke_test.py               | 86 +++++++++++++++++++++++++++++
 scripts/uninstall_local_packages.py | 23 ++++++++
 3 files changed, 142 insertions(+)
 create mode 100644 scripts/install_local_packages.py
 create mode 100644 scripts/smoke_test.py
 create mode 100644 scripts/uninstall_local_packages.py

diff --git a/scripts/install_local_packages.py b/scripts/install_local_packages.py
new file mode 100644
index 0000000..6f1e293
--- /dev/null
+++ b/scripts/install_local_packages.py
@@ -0,0 +1,33 @@
+import os
+
+from scispacy.version import VERSION
+
+
+def main():
+    model_names = [
+        "en_core_sci_sm",
+        "en_core_sci_md",
+        "en_core_sci_lg",
+        "en_core_sci_scibert",
+        "en_ner_bc5cdr_md",
+        "en_ner_craft_md",
+        "en_ner_bionlp13cg_md",
+        "en_ner_jnlpba_md",
+    ]
+
+    full_package_paths = [
+        os.path.join(
+            "packages",
+            f"{model_name}-{VERSION}",
+            "dist",
+            f"{model_name}-{VERSION}.tar.gz",
+        )
+        for model_name in model_names
+    ]
+
+    for package_path in full_package_paths:
+        os.system(f"pip install {package_path}")
+
+
+if __name__ == "__main__":
+    main()
diff --git a/scripts/smoke_test.py b/scripts/smoke_test.py
new file mode 100644
index 0000000..cd488d6
--- /dev/null
+++ b/scripts/smoke_test.py
@@ -0,0 +1,86 @@
+import spacy
+from tqdm import tqdm
+
+from scispacy.abbreviation import AbbreviationDetector
+from scispacy.linking import EntityLinker
+
+
+def main():
+    print("Testing core models...")
+    print()
+    model_names = [
+        "en_core_sci_sm",
+        "en_core_sci_md",
+        "en_core_sci_lg",
+        "en_core_sci_scibert",
+        "en_ner_bc5cdr_md",
+        "en_ner_craft_md",
+        "en_ner_bionlp13cg_md",
+        "en_ner_jnlpba_md",
+    ]
+
+    models = [
+        spacy.load(model_name)
+        for model_name in tqdm(model_names, desc="Loading core models")
+    ]
+
+    text = (
+        "This first sentence mentions John. "
+        "John uses IL gene and interleukin-2 to treat diabetes and "
+        "aspirin as proteins for arms and legs on lemurs and humans."
+    )
+
+    for model_name, model in zip(model_names, models):
+        print(f"Testing {model_name}")
+        doc = model(text)
+        for sentence in doc.sents:
+            print([t.text for t in sentence])
+            print([t.lemma_ for t in sentence])
+            print([t.pos_ for t in sentence])
+            print([t.tag_ for t in sentence])
+            print([t.dep_ for t in sentence])
+            print([t.ent_type_ for t in sentence])
+            print()
+        print()
+
+    print("Testing abbreivation detector...")
+    abbreviation_nlp = spacy.load("en_core_sci_sm")
+    abbreviation_nlp.add_pipe("abbreviation_detector")
+    abbreviation_text = (
+        "Spinal and bulbar muscular atrophy (SBMA) is an inherited "
+        "motor neuron disease caused by the expansion of a polyglutamine "
+        "tract within the androgen receptor (AR). SBMA can be caused by this easily."
+    )
+    abbreviation_doc = abbreviation_nlp(abbreviation_text)
+    for abbrevation in abbreviation_doc._.abbreviations:
+        print(
+            f"{abbrevation} \t ({abbrevation.start}, {abbrevation.end}) {abbrevation._.long_form}"
+        )
+    print()
+
+    print("Testing entity linkers...")
+    print()
+    ontology_names = ["umls", "mesh", "rxnorm", "go", "hpo"]
+    ontology_models = [spacy.load("en_core_sci_sm") for _ in ontology_names]
+    for ontology_name, ontology_model in tqdm(
+        zip(ontology_names, ontology_models), desc="Adding entity linker pipes"
+    ):
+        ontology_model.add_pipe(
+            "scispacy_linker",
+            config={"resolve_abbreviations": False, "linker_name": ontology_name},
+        )
+
+    linking_text = "Diabetes is a disease that affects humans."
+    for ontology_name, ontology_model in zip(ontology_names, ontology_models):
+        print(f"Testing {ontology_name} linker...")
+        linker_pipe = ontology_model.get_pipe("scispacy_linker")
+        doc = ontology_model(linking_text)
+        for entity in doc.ents:
+            print("Entity name: ", entity)
+            for ontology_entity in entity._.kb_ents[:1]:
+                print(linker_pipe.kb.cui_to_entity[ontology_entity[0]])
+        print()
+
+
+if __name__ == "__main__":
+    main()
diff --git a/scripts/uninstall_local_packages.py b/scripts/uninstall_local_packages.py
new file mode 100644
index 0000000..d14d69f
--- /dev/null
+++ b/scripts/uninstall_local_packages.py
@@ -0,0 +1,23 @@
+import os
+
+from scispacy.version import VERSION
+
+
+def main():
+    model_names = [
+        "en_core_sci_sm",
+        "en_core_sci_md",
+        "en_core_sci_lg",
+        "en_core_sci_scibert",
+        "en_ner_bc5cdr_md",
+        "en_ner_craft_md",
+        "en_ner_bionlp13cg_md",
+        "en_ner_jnlpba_md",
+    ]
+
+    for package_name in model_names:
+        os.system(f"pip uninstall {package_name}")
+
+
+if __name__ == "__main__":
+    main()

From bfe27f5cd366104dd5eaba678b579735b08a480f Mon Sep 17 00:00:00 2001
From: Daniel King <daking@hey.com>
Date: Sat, 27 Aug 2022 21:36:02 -0700
Subject: [PATCH 07/17] Fix a couple of file dependencies in project.yml

---
 project.yml | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/project.yml b/project.yml
index 75cf02f..58a8788 100644
--- a/project.yml
+++ b/project.yml
@@ -309,7 +309,7 @@ commands:
     script:
       - "spacy train ${vars.parser_tagger_scibert_config_loc} --output ${vars.parser_tagger_scibert_loc} --code ${vars.code_loc} --paths.vocab_path ${vars.vocab_lg_loc} --gpu-id ${vars.gpu_id}"
     deps:
-      - "${vars.parser_tagger_config_loc}"
+      - "${vars.parser_tagger_scibert_config_loc}"
       - "${vars.genia_train_spacy_loc}"
       - "${vars.genia_dev_spacy_loc}"
       - "${vars.genia_test_spacy_loc}"
@@ -357,7 +357,7 @@ commands:
     script:
       - "spacy train ${vars.ner_scibert_config_loc} --output ${vars.ner_scibert_loc} --code ${vars.code_loc} --paths.parser_tagger_path ${vars.parser_tagger_scibert_loc}/model-best --gpu-id ${vars.gpu_id}"
     deps:
-      - "${vars.ner_config_loc}"
+      - "${vars.ner_scibert_config_loc}"
       - "${vars.parser_tagger_scibert_loc}/model-best"
       - "${vars.corpus_pubtator_loc_local}"
     outputs:

From ffe630de68de9bcd0cc7a5fef15994fe78e38cfe Mon Sep 17 00:00:00 2001
From: Daniel King <daking@hey.com>
Date: Sat, 27 Aug 2022 21:36:42 -0700
Subject: [PATCH 08/17] Update scibert config to better match spacys
 transformer configuration. Attempting to fix scibert sentence splitting

---
 configs/base_parser_tagger_scibert.cfg | 42 ++++++++++++++------------
 1 file changed, 22 insertions(+), 20 deletions(-)

diff --git a/configs/base_parser_tagger_scibert.cfg b/configs/base_parser_tagger_scibert.cfg
index 65b8f2f..fd13a56 100644
--- a/configs/base_parser_tagger_scibert.cfg
+++ b/configs/base_parser_tagger_scibert.cfg
@@ -13,7 +13,6 @@ seed = 0
 [nlp]
 lang = "en"
 pipeline = ["transformer","tagger","attribute_ruler","lemmatizer","parser"]
-batch_size = 256
 tokenizer = {"@tokenizers":"spacy.Tokenizer.v1"}
 disabled = []
 before_creation = null
@@ -39,8 +38,8 @@ update_with_oracle_cut_size = 100
 @architectures = "spacy.TransitionBasedParser.v2"
 state_type = "parser"
 extra_state_tokens = false
-hidden_width = 128
-maxout_pieces = 3
+hidden_width = 64
+maxout_pieces = 2
 use_upper = false
 nO = null
 
@@ -48,10 +47,12 @@ nO = null
 @architectures = "spacy-transformers.TransformerListener.v1"
 grad_factor = 1.0
 pooling = {"@layers":"reduce_mean.v1"}
-upstream = "*"
+upstream = "transformer"
 
 [components.tagger]
 factory = "tagger"
+neg_prefix = "!"
+overwrite = false
 
 [components.tagger.model]
 @architectures = "spacy.Tagger.v2"
@@ -62,7 +63,7 @@ normalize = False
 @architectures = "spacy-transformers.TransformerListener.v1"
 grad_factor = 1.0
 pooling = {"@layers":"reduce_mean.v1"}
-upstream = "*"
+upstream = "transformer"
 
 [components.transformer]
 factory = "transformer"
@@ -108,24 +109,20 @@ train_corpus = "corpora.train"
 seed = ${system.seed}
 gpu_allocator = ${system.gpu_allocator}
 dropout = 0.1
-accumulate_gradient = 1
-patience = 0
-max_epochs = 8
-max_steps = 0
-eval_frequency = 2300
+accumulate_gradient = 3
+patience = 5000
+max_epochs = 0
+max_steps = 20000
+eval_frequency = 1000
 frozen_components = ["attribute_ruler", "lemmatizer"]
 before_to_disk = null
 
 [training.batcher]
-@batchers = "spacy.batch_by_sequence.v1"
+@batchers = "spacy.batch_by_padded.v1"
 get_length = null
-
-[training.batcher.size]
-@schedules = "compounding.v1"
-start = 4
-stop = 12
-compound = 1.001
-t = 0.0
+discard_oversize = true
+size = 2000
+buffer = 256
 
 [training.logger]
 @loggers = "spacy.ConsoleLogger.v1"
@@ -138,9 +135,14 @@ beta2 = 0.999
 L2_is_weight_decay = true
 L2 = 0.01
 grad_clip = 1.0
-use_averages = false
+use_averages = true
 eps = 0.00000001
-learn_rate = 0.00005
+
+[training.optimizer.learn_rate]
+@schedules = "warmup_linear.v1"
+warmup_steps = 250
+total_steps = 20000
+initial_rate = 0.00005
 
 
 [training.score_weights]

From 00aa1a4f973810a550770e49d7d542da3bfa9850 Mon Sep 17 00:00:00 2001
From: Daniel King <daking@hey.com>
Date: Sun, 28 Aug 2022 01:05:14 -0700
Subject: [PATCH 09/17] Tweak text in smoke test

---
 scripts/smoke_test.py | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/scripts/smoke_test.py b/scripts/smoke_test.py
index cd488d6..0cb1efc 100644
--- a/scripts/smoke_test.py
+++ b/scripts/smoke_test.py
@@ -25,7 +25,7 @@ def main():
     ]
 
     text = (
-        "This first sentence mentions John. "
+        "DNA is a very important part of the cellular structure of the body. "
         "John uses IL gene and interleukin-2 to treat diabetes and "
         "aspirin as proteins for arms and legs on lemurs and humans."
     )
@@ -42,6 +42,7 @@ def main():
             print([t.ent_type_ for t in sentence])
             print()
         print()
+        input("Continue?")
 
     print("Testing abbreivation detector...")
     abbreviation_nlp = spacy.load("en_core_sci_sm")
@@ -57,6 +58,7 @@ def main():
             f"{abbrevation} \t ({abbrevation.start}, {abbrevation.end}) {abbrevation._.long_form}"
         )
     print()
+    input("Continue?")
 
     print("Testing entity linkers...")
     print()
@@ -70,7 +72,7 @@ def main():
             config={"resolve_abbreviations": False, "linker_name": ontology_name},
         )
 
-    linking_text = "Diabetes is a disease that affects humans."
+    linking_text = "Diabetes is a disease that affects humans and is treated with aspirin via a metabolic process."
     for ontology_name, ontology_model in zip(ontology_names, ontology_models):
         print(f"Testing {ontology_name} linker...")
         linker_pipe = ontology_model.get_pipe("scispacy_linker")
@@ -80,6 +82,7 @@ def main():
             for ontology_entity in entity._.kb_ents[:1]:
                 print(linker_pipe.kb.cui_to_entity[ontology_entity[0]])
         print()
+        input("Continue?")
 
 
 if __name__ == "__main__":

From 0fe3d79602f3917aaad4e3eeb45b07c2e278d3c3 Mon Sep 17 00:00:00 2001
From: Daniel King <daking@hey.com>
Date: Sun, 28 Aug 2022 14:14:19 -0700
Subject: [PATCH 10/17] Update version strings in docs

---
 Dockerfile    |  2 +-
 README.md     | 18 +++++++++---------
 docs/index.md | 16 ++++++++--------
 3 files changed, 18 insertions(+), 18 deletions(-)

diff --git a/Dockerfile b/Dockerfile
index 396daba..8417070 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -18,7 +18,7 @@ WORKDIR /work
 COPY requirements.in .
 
 RUN pip install -r requirements.in
-RUN pip install https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.0/en_core_sci_sm-0.5.0.tar.gz
+RUN pip install https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.1/en_core_sci_sm-0.5.1.tar.gz
 RUN python -m spacy download en_core_web_sm
 RUN python -m spacy download en_core_web_md
 
diff --git a/README.md b/README.md
index 210b36c..5bd7394 100644
--- a/README.md
+++ b/README.md
@@ -19,7 +19,7 @@ pip install scispacy
 to install a model (see our full selection of available models below), run a command like the following:
 
 ```bash
-pip install https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.0/en_core_sci_sm-0.5.0.tar.gz
+pip install https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.1/en_core_sci_sm-0.5.1.tar.gz
 ```
 
 Note: We strongly recommend that you use an isolated Python environment (such as virtualenv or conda) to install scispacy.
@@ -76,14 +76,14 @@ pip install CMD-V(to paste the copied URL)
 
 | Model          | Description       | Install URL
 |:---------------|:------------------|:----------|
-| en_core_sci_sm | A full spaCy pipeline for biomedical data with a ~100k vocabulary. |[Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.0/en_core_sci_sm-0.5.0.tar.gz)|
-| en_core_sci_md |  A full spaCy pipeline for biomedical data with a ~360k vocabulary and 50k word vectors. |[Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.0/en_core_sci_md-0.5.0.tar.gz)|
-| en_core_sci_lg |  A full spaCy pipeline for biomedical data with a ~785k vocabulary and 600k word vectors. |[Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.0/en_core_sci_lg-0.5.0.tar.gz)|
-| en_core_sci_scibert |  A full spaCy pipeline for biomedical data with a ~785k vocabulary and `allenai/scibert-base` as the transformer model. You may want to [use a GPU](https://spacy.io/usage#gpu) with this model. |[Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.0/en_core_sci_scibert-0.5.0.tar.gz)|
-| en_ner_craft_md|  A spaCy NER model trained on the CRAFT corpus.|[Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.0/en_ner_craft_md-0.5.0.tar.gz)|
-| en_ner_jnlpba_md | A spaCy NER model trained on the JNLPBA corpus.| [Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.0/en_ner_jnlpba_md-0.5.0.tar.gz)|
-| en_ner_bc5cdr_md |  A spaCy NER model trained on the BC5CDR corpus. | [Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.0/en_ner_bc5cdr_md-0.5.0.tar.gz)|
-| en_ner_bionlp13cg_md |  A spaCy NER model trained on the BIONLP13CG corpus. |[Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.0/en_ner_bionlp13cg_md-0.5.0.tar.gz)|
+| en_core_sci_sm | A full spaCy pipeline for biomedical data with a ~100k vocabulary. |[Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.1/en_core_sci_sm-0.5.1.tar.gz)|
+| en_core_sci_md |  A full spaCy pipeline for biomedical data with a ~360k vocabulary and 50k word vectors. |[Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.1/en_core_sci_md-0.5.1.tar.gz)|
+| en_core_sci_lg |  A full spaCy pipeline for biomedical data with a ~785k vocabulary and 600k word vectors. |[Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.1/en_core_sci_lg-0.5.1.tar.gz)|
+| en_core_sci_scibert |  A full spaCy pipeline for biomedical data with a ~785k vocabulary and `allenai/scibert-base` as the transformer model. You may want to [use a GPU](https://spacy.io/usage#gpu) with this model. |[Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.1/en_core_sci_scibert-0.5.1.tar.gz)|
+| en_ner_craft_md|  A spaCy NER model trained on the CRAFT corpus.|[Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.1/en_ner_craft_md-0.5.1.tar.gz)|
+| en_ner_jnlpba_md | A spaCy NER model trained on the JNLPBA corpus.| [Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.1/en_ner_jnlpba_md-0.5.1.tar.gz)|
+| en_ner_bc5cdr_md |  A spaCy NER model trained on the BC5CDR corpus. | [Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.1/en_ner_bc5cdr_md-0.5.1.tar.gz)|
+| en_ner_bionlp13cg_md |  A spaCy NER model trained on the BIONLP13CG corpus. |[Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.1/en_ner_bionlp13cg_md-0.5.1.tar.gz)|
 
 
 ## Additional Pipeline Components
diff --git a/docs/index.md b/docs/index.md
index 8579f9e..25b7150 100644
--- a/docs/index.md
+++ b/docs/index.md
@@ -17,14 +17,14 @@ pip install <Model URL>
 
 | Model          | Description       | Install URL
 |:---------------|:------------------|:----------|
-| en_core_sci_sm | A full spaCy pipeline for biomedical data. |[Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.0/en_core_sci_sm-0.5.0.tar.gz)|
-| en_core_sci_md |  A full spaCy pipeline for biomedical data with a larger vocabulary and 50k word vectors. |[Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.0/en_core_sci_md-0.5.0.tar.gz)|
-| en_core_sci_scibert |  A full spaCy pipeline for biomedical data with a ~785k vocabulary and `allenai/scibert-base` as the transformer model. |[Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.0/en_core_sci_scibert-0.5.0.tar.gz)|
-| en_core_sci_lg |  A full spaCy pipeline for biomedical data with a larger vocabulary and 600k word vectors. |[Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.0/en_core_sci_lg-0.5.0.tar.gz)|
-| en_ner_craft_md|  A spaCy NER model trained on the CRAFT corpus.|[Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.0/en_ner_craft_md-0.5.0.tar.gz)|
-| en_ner_jnlpba_md | A spaCy NER model trained on the JNLPBA corpus.| [Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.0/en_ner_jnlpba_md-0.5.0.tar.gz)|
-| en_ner_bc5cdr_md |  A spaCy NER model trained on the BC5CDR corpus. | [Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.0/en_ner_bc5cdr_md-0.5.0.tar.gz)|
-| en_ner_bionlp13cg_md |  A spaCy NER model trained on the BIONLP13CG corpus. | [Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.0/en_ner_bionlp13cg_md-0.5.0.tar.gz)|
+| en_core_sci_sm | A full spaCy pipeline for biomedical data. |[Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.1/en_core_sci_sm-0.5.1.tar.gz)|
+| en_core_sci_md |  A full spaCy pipeline for biomedical data with a larger vocabulary and 50k word vectors. |[Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.1/en_core_sci_md-0.5.1.tar.gz)|
+| en_core_sci_scibert |  A full spaCy pipeline for biomedical data with a ~785k vocabulary and `allenai/scibert-base` as the transformer model. |[Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.1/en_core_sci_scibert-0.5.1.tar.gz)|
+| en_core_sci_lg |  A full spaCy pipeline for biomedical data with a larger vocabulary and 600k word vectors. |[Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.1/en_core_sci_lg-0.5.1.tar.gz)|
+| en_ner_craft_md|  A spaCy NER model trained on the CRAFT corpus.|[Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.1/en_ner_craft_md-0.5.1.tar.gz)|
+| en_ner_jnlpba_md | A spaCy NER model trained on the JNLPBA corpus.| [Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.1/en_ner_jnlpba_md-0.5.1.tar.gz)|
+| en_ner_bc5cdr_md |  A spaCy NER model trained on the BC5CDR corpus. | [Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.1/en_ner_bc5cdr_md-0.5.1.tar.gz)|
+| en_ner_bionlp13cg_md |  A spaCy NER model trained on the BIONLP13CG corpus. | [Download](https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.1/en_ner_bionlp13cg_md-0.5.1.tar.gz)|
 
 
 

From b0e69473b8ad63b5cb2eacd46c55f369f3e4b6b8 Mon Sep 17 00:00:00 2001
From: Daniel King <daking@hey.com>
Date: Sun, 28 Aug 2022 22:52:07 -0700
Subject: [PATCH 11/17] Put transformer config back the way it was

---
 configs/base_parser_tagger_scibert.cfg | 47 ++++++++++++--------------
 1 file changed, 22 insertions(+), 25 deletions(-)

diff --git a/configs/base_parser_tagger_scibert.cfg b/configs/base_parser_tagger_scibert.cfg
index fd13a56..3615299 100644
--- a/configs/base_parser_tagger_scibert.cfg
+++ b/configs/base_parser_tagger_scibert.cfg
@@ -13,6 +13,7 @@ seed = 0
 [nlp]
 lang = "en"
 pipeline = ["transformer","tagger","attribute_ruler","lemmatizer","parser"]
+batch_size = 256
 tokenizer = {"@tokenizers":"spacy.Tokenizer.v1"}
 disabled = []
 before_creation = null
@@ -38,8 +39,8 @@ update_with_oracle_cut_size = 100
 @architectures = "spacy.TransitionBasedParser.v2"
 state_type = "parser"
 extra_state_tokens = false
-hidden_width = 64
-maxout_pieces = 2
+hidden_width = 128
+maxout_pieces = 3
 use_upper = false
 nO = null
 
@@ -47,23 +48,20 @@ nO = null
 @architectures = "spacy-transformers.TransformerListener.v1"
 grad_factor = 1.0
 pooling = {"@layers":"reduce_mean.v1"}
-upstream = "transformer"
+upstream = "*"
 
 [components.tagger]
 factory = "tagger"
-neg_prefix = "!"
-overwrite = false
 
 [components.tagger.model]
-@architectures = "spacy.Tagger.v2"
+@architectures = "spacy.Tagger.v1"
 nO = null
-normalize = False
 
 [components.tagger.model.tok2vec]
 @architectures = "spacy-transformers.TransformerListener.v1"
 grad_factor = 1.0
 pooling = {"@layers":"reduce_mean.v1"}
-upstream = "transformer"
+upstream = "*"
 
 [components.transformer]
 factory = "transformer"
@@ -109,20 +107,24 @@ train_corpus = "corpora.train"
 seed = ${system.seed}
 gpu_allocator = ${system.gpu_allocator}
 dropout = 0.1
-accumulate_gradient = 3
-patience = 5000
-max_epochs = 0
-max_steps = 20000
-eval_frequency = 1000
+accumulate_gradient = 1
+patience = 0
+max_epochs = 8
+max_steps = 0
+eval_frequency = 2300
 frozen_components = ["attribute_ruler", "lemmatizer"]
 before_to_disk = null
 
 [training.batcher]
-@batchers = "spacy.batch_by_padded.v1"
+@batchers = "spacy.batch_by_sequence.v1"
 get_length = null
-discard_oversize = true
-size = 2000
-buffer = 256
+
+[training.batcher.size]
+@schedules = "compounding.v1"
+start = 4
+stop = 12
+compound = 1.001
+t = 0.0
 
 [training.logger]
 @loggers = "spacy.ConsoleLogger.v1"
@@ -135,14 +137,9 @@ beta2 = 0.999
 L2_is_weight_decay = true
 L2 = 0.01
 grad_clip = 1.0
-use_averages = true
+use_averages = false
 eps = 0.00000001
-
-[training.optimizer.learn_rate]
-@schedules = "warmup_linear.v1"
-warmup_steps = 250
-total_steps = 20000
-initial_rate = 0.00005
+learn_rate = 0.00005
 
 
 [training.score_weights]
@@ -163,4 +160,4 @@ ents_r = 0.0
 [initialize]
 
 [initialize.before_init]
-@callbacks = "replace_tokenizer"
+@callbacks = "replace_tokenizer"
\ No newline at end of file

From 3825235e6d689186d3c87cef6c09e1a48e943223 Mon Sep 17 00:00:00 2001
From: Daniel King <daking@hey.com>
Date: Sun, 28 Aug 2022 22:52:16 -0700
Subject: [PATCH 12/17] Update metrics

---
 docs/index.md | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)

diff --git a/docs/index.md b/docs/index.md
index 25b7150..b1aab56 100644
--- a/docs/index.md
+++ b/docs/index.md
@@ -34,18 +34,18 @@ Our models achieve performance within 3% of published state of the art dependenc
 
 | model          | UAS | LAS   | POS   | Mentions (F1) | Web UAS | 
 |:---------------|:----|:------|:------|:---|:---|
-| en_core_sci_sm | 89.27| 87.33  |  98.29  |  68.05  |  87.61  |
-| en_core_sci_md | 89.86| 87.92 |  98.43 |  69.32 |  88.05  |
-| en_core_sci_lg | 89.54| 87.66  |  98.29  |  69.52  |  87.68  |
-| en_core_sci_scibert | 92.28| 90.83  |  98.93  |  67.84  |  92.63  |
+| en_core_sci_sm | 89.03| 87.00  |  98.13  |  67.87  |  87.42  |
+| en_core_sci_md | 89.73| 87.85 |  98.40 |  69.53 |  87.79  |
+| en_core_sci_lg | 89.75| 87.79  |  98.49  |  69.69  |  87.74  |
+| en_core_sci_scibert | 92.21| 90.65  |  98.86  |  68.01  |  92.58  |
 
 
 | model          | F1 |   Entity Types|
 |:---------------|:-----|:--------|
-| en_ner_craft_md | 78.35|GGP, SO, TAXON, CHEBI, GO, CL|
-| en_ner_jnlpba_md | 70.89| DNA, CELL_TYPE, CELL_LINE, RNA, PROTEIN |
-| en_ner_bc5cdr_md | 84.70| DISEASE, CHEMICAL|
-| en_ner_bionlp13cg_md | 76.79| AMINO_ACID, ANATOMICAL_SYSTEM, CANCER, CELL, CELLULAR_COMPONENT, DEVELOPING_ANATOMICAL_STRUCTURE, GENE_OR_GENE_PRODUCT, IMMATERIAL_ANATOMICAL_ENTITY, MULTI-TISSUE_STRUCTURE, ORGAN, ORGANISM, ORGANISM_SUBDIVISION, ORGANISM_SUBSTANCE, PATHOLOGICAL_FORMATION, SIMPLE_CHEMICAL, TISSUE |
+| en_ner_craft_md | 76.75|GGP, SO, TAXON, CHEBI, GO, CL|
+| en_ner_jnlpba_md | 72.28| DNA, CELL_TYPE, CELL_LINE, RNA, PROTEIN |
+| en_ner_bc5cdr_md | 84.53| DISEASE, CHEMICAL|
+| en_ner_bionlp13cg_md | 76.57| AMINO_ACID, ANATOMICAL_SYSTEM, CANCER, CELL, CELLULAR_COMPONENT, DEVELOPING_ANATOMICAL_STRUCTURE, GENE_OR_GENE_PRODUCT, IMMATERIAL_ANATOMICAL_ENTITY, MULTI-TISSUE_STRUCTURE, ORGAN, ORGANISM, ORGANISM_SUBDIVISION, ORGANISM_SUBSTANCE, PATHOLOGICAL_FORMATION, SIMPLE_CHEMICAL, TISSUE |
 
 
 ### Example Usage

From bd8ab1ad6b7fbb3abc804f2a6fc64bb6e22f3d3f Mon Sep 17 00:00:00 2001
From: Daniel King <daking@hey.com>
Date: Sun, 28 Aug 2022 22:53:01 -0700
Subject: [PATCH 13/17] Add new line back

---
 configs/base_parser_tagger_scibert.cfg | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/configs/base_parser_tagger_scibert.cfg b/configs/base_parser_tagger_scibert.cfg
index 3615299..016a395 100644
--- a/configs/base_parser_tagger_scibert.cfg
+++ b/configs/base_parser_tagger_scibert.cfg
@@ -160,4 +160,4 @@ ents_r = 0.0
 [initialize]
 
 [initialize.before_init]
-@callbacks = "replace_tokenizer"
\ No newline at end of file
+@callbacks = "replace_tokenizer"

From 82c51e0ce75cf72ec9ea340a32121681c6090a53 Mon Sep 17 00:00:00 2001
From: Daniel King <daking@hey.com>
Date: Tue, 6 Sep 2022 17:07:27 -0700
Subject: [PATCH 14/17] Convenience script for installing remote packages

---
 scripts/install_remote_packages.py | 28 ++++++++++++++++++++++++++++
 1 file changed, 28 insertions(+)
 create mode 100644 scripts/install_remote_packages.py

diff --git a/scripts/install_remote_packages.py b/scripts/install_remote_packages.py
new file mode 100644
index 0000000..60ff0f5
--- /dev/null
+++ b/scripts/install_remote_packages.py
@@ -0,0 +1,28 @@
+import os
+
+from scispacy.version import VERSION
+
+
+def main():
+    s3_prefix = "https://s3-us-west-2.amazonaws.com/ai2-s2-scispacy/releases/v0.5.1/"
+    model_names = [
+        "en_core_sci_sm",
+        "en_core_sci_md",
+        "en_core_sci_lg",
+        "en_core_sci_scibert",
+        "en_ner_bc5cdr_md",
+        "en_ner_craft_md",
+        "en_ner_bionlp13cg_md",
+        "en_ner_jnlpba_md",
+    ]
+
+    full_package_paths = [
+        f"{s3_prefix}{model_name}-{VERSION}.tar.gz" for model_name in model_names
+    ]
+
+    for package_path in full_package_paths:
+        os.system(f"pip install {package_path}")
+
+
+if __name__ == "__main__":
+    main()

From 2369bdf7941cbcbfdd6ad1c31cbb769049fe2d12 Mon Sep 17 00:00:00 2001
From: Daniel King <daking@hey.com>
Date: Tue, 6 Sep 2022 17:07:45 -0700
Subject: [PATCH 15/17] Convenience script for getting the metrics to report

---
 scripts/print_out_metrics.py | 45 ++++++++++++++++++++++++++++++++++++
 1 file changed, 45 insertions(+)
 create mode 100644 scripts/print_out_metrics.py

diff --git a/scripts/print_out_metrics.py b/scripts/print_out_metrics.py
new file mode 100644
index 0000000..4c634b5
--- /dev/null
+++ b/scripts/print_out_metrics.py
@@ -0,0 +1,45 @@
+import os
+import json
+
+
+def main():
+    core_model_names = ["lg", "md", "sm", "scibert"]
+    ner_model_names = ["bc5cdr", "bionlp13cg", "craft", "jnlpba"]
+
+    base_path = "packages"
+    for core_model_name in core_model_names:
+        print(f"Printing results for {core_model_name}")
+        with open(
+            os.path.join(base_path, f"{core_model_name}_genia_results.json")
+        ) as _genia_results_file:
+            genia_results = json.load(_genia_results_file)
+
+        with open(
+            os.path.join(base_path, f"{core_model_name}_onto_results.json")
+        ) as _onto_results_file:
+            onto_results = json.load(_onto_results_file)
+
+        with open(
+            os.path.join(base_path, f"{core_model_name}_mm_results.json")
+        ) as _mm_results_file:
+            mm_results = json.load(_mm_results_file)
+
+        print(f"Genia tag accuracy: {genia_results['tag_acc']}")
+        print(f"Genia uas: {genia_results['dep_uas']}")
+        print(f"Genia las: {genia_results['dep_las']}")
+        print(f"Ontonotes uas: {onto_results['dep_uas']}")
+        print(f"MedMentions F1: {mm_results['f1-measure-untyped']}")
+        print()
+
+    for ner_model_name in ner_model_names:
+        print(f"Printing results for {ner_model_name}")
+        with open(
+            os.path.join(base_path, f"{ner_model_name}_results.json")
+        ) as _ner_results_file:
+            ner_results = json.load(_ner_results_file)
+
+        print(f"NER F1: {ner_results['f1-measure-overall']}")
+
+
+if __name__ == "__main__":
+    main()

From 05e67037b85b80067c1ed11e0ca18d62a78cdf78 Mon Sep 17 00:00:00 2001
From: Daniel King <daking@hey.com>
Date: Tue, 6 Sep 2022 17:07:54 -0700
Subject: [PATCH 16/17] Fix hyponym test

---
 tests/test_hyponym_detector.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tests/test_hyponym_detector.py b/tests/test_hyponym_detector.py
index e8ab8f6..ff41c54 100644
--- a/tests/test_hyponym_detector.py
+++ b/tests/test_hyponym_detector.py
@@ -20,7 +20,7 @@ def test_sentences(self):
         )
         doc = self.nlp(text)
         fig_trees = doc[21:23]
-        plant_species = doc[17:19]
+        plant_species = doc[18:19]
         assert doc._.hearst_patterns == [("such_as", plant_species, fig_trees)]
 
         doc = self.nlp("SARS, or other coronaviruses, are bad.")

From 40ded231f25b781a9bdf89db932ff675c5156ed2 Mon Sep 17 00:00:00 2001
From: Daniel King <daking@hey.com>
Date: Tue, 6 Sep 2022 17:14:01 -0700
Subject: [PATCH 17/17] Actually fix hyponym test

---
 tests/test_hyponym_detector.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tests/test_hyponym_detector.py b/tests/test_hyponym_detector.py
index ff41c54..f89f8e3 100644
--- a/tests/test_hyponym_detector.py
+++ b/tests/test_hyponym_detector.py
@@ -20,7 +20,7 @@ def test_sentences(self):
         )
         doc = self.nlp(text)
         fig_trees = doc[21:23]
-        plant_species = doc[18:19]
+        plant_species = doc[16:19]
         assert doc._.hearst_patterns == [("such_as", plant_species, fig_trees)]
 
         doc = self.nlp("SARS, or other coronaviruses, are bad.")