Update spaCy pipeline

.gitattributes

@@ -30,3 +30,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *strings.json filter=lfs diff=lfs merge=lfs -text
 vectors filter=lfs diff=lfs merge=lfs -text
 model filter=lfs diff=lfs merge=lfs -text
+vocab/key2row filter=lfs diff=lfs merge=lfs -text

README.md

@@ -14,25 +14,25 @@ model-index:
     metrics:
     - name: NER Precision
       type: precision
-      value: 0.8910716705
+      value: 0.8822157434
     - name: NER Recall
       type: recall
-      value: 0.9043035886
+      value: 0.925382263
     - name: NER F Score
       type: f_score
-      value: 0.8976388699
+      value: 0.9032835821
 ---
 | Feature | Description |
 | --- | --- |
 | **Name** | `en_core_med7_trf` |
-| **Version** | `3.1.3.1` |
-| **spaCy** | `>=3.1.4,<3.2.0` |
+| **Version** | `3.4.2.1` |
+| **spaCy** | `>=3.4.2,<3.5.0` |
 | **Default Pipeline** | `transformer`, `ner` |
 | **Components** | `transformer`, `ner` |
-| **Vectors** | 0 keys, 0 unique vectors (0 dimensions) |
+| **Vectors** | 514157 keys, 514157 unique vectors (300 dimensions) |
 | **Sources** | n/a |
 | **License** | `MIT` |
-| **Author** | [Andrey Kormilitzin](kormilitzin.com) |
+| **Author** | [Andrey Kormilitzin](https://www.kormilitzin.com/) |
 
 ### Label Scheme
 
@@ -50,8 +50,8 @@ model-index:
 
 | Type | Score |
 | --- | --- |
-| `ENTS_F` | 89.76 |
-| `ENTS_P` | 89.11 |
-| `ENTS_R` | 90.43 |
-| `TRANSFORMER_LOSS` | 209606.10 |
-| `NER_LOSS` | 874893.84 |
+| `ENTS_F` | 90.33 |
+| `ENTS_P` | 88.22 |
+| `ENTS_R` | 92.54 |
+| `TRANSFORMER_LOSS` | 2502627.06 |
+| `NER_LOSS` | 114576.77 |

config.cfg

@@ -1,8 +1,8 @@
 [paths]
-train = "./data/spacy_format/train_443.spacy"
-dev = "./data/spacy_format/dev_443.spacy"
-vectors = null
-init_tok2vec = null
+train = "./data/spacy_format/train_med7_v34.spacy"
+dev = "./data/spacy_format/dev_med7_v34.spacy"
+vectors = "en_core_web_lg"
+init_tok2vec = "/mnt/sdf/andrey/projects/med7_v3/output_pretrain_lg/model169.bin"
 
 [system]
 gpu_allocator = "pytorch"
@@ -11,7 +11,7 @@ seed = 0
 [nlp]
 lang = "en"
 pipeline = ["transformer","ner"]
-batch_size = 2
+batch_size = 64
 disabled = []
 before_creation = null
 after_creation = null
@@ -24,13 +24,14 @@ tokenizer = {"@tokenizers":"spacy.Tokenizer.v1"}
 factory = "ner"
 incorrect_spans_key = null
 moves = null
+scorer = {"@scorers":"spacy.ner_scorer.v1"}
 update_with_oracle_cut_size = 100
 
 [components.ner.model]
 @architectures = "spacy.TransitionBasedParser.v2"
 state_type = "ner"
 extra_state_tokens = false
-hidden_width = 64
+hidden_width = 128
 maxout_pieces = 2
 use_upper = false
 nO = null
@@ -47,17 +48,22 @@ max_batch_items = 4096
 set_extra_annotations = {"@annotation_setters":"spacy-transformers.null_annotation_setter.v1"}
 
 [components.transformer.model]
-@architectures = "spacy-transformers.TransformerModel.v1"
-name = "roberta-base"
+@architectures = "spacy-transformers.TransformerModel.v3"
+name = "/mnt/sdf/andrey/projects/med7_v3/RoBERTa-base-PM-M3-hf"
+mixed_precision = false
 
 [components.transformer.model.get_spans]
 @span_getters = "spacy-transformers.strided_spans.v1"
-window = 128
+window = 256
 stride = 96
 
+[components.transformer.model.grad_scaler_config]
+
 [components.transformer.model.tokenizer_config]
 use_fast = true
 
+[components.transformer.model.transformer_config]
+
 [corpora]
 
 [corpora.dev]
@@ -83,9 +89,9 @@ train_corpus = "corpora.train"
 seed = ${system.seed}
 gpu_allocator = ${system.gpu_allocator}
 dropout = 0.1
-patience = 2000
+patience = 3600
 max_epochs = 0
-max_steps = 60000
+max_steps = 20000
 eval_frequency = 200
 frozen_components = []
 annotating_components = []
@@ -94,13 +100,13 @@ before_to_disk = null
 [training.batcher]
 @batchers = "spacy.batch_by_padded.v1"
 discard_oversize = true
-size = 2000
+size = 1000
 buffer = 256
 get_length = null
 
 [training.logger]
 @loggers = "spacy.ConsoleLogger.v1"
-progress_bar = false
+progress_bar = true
 
 [training.optimizer]
 @optimizers = "Adam.v1"
@@ -130,10 +136,14 @@ ents_per_type = null
 vectors = ${paths.vectors}
 init_tok2vec = ${paths.init_tok2vec}
 vocab_data = null
-lookups = null
 before_init = null
 after_init = null
 
 [initialize.components]
 
+[initialize.lookups]
+@misc = "spacy.LookupsDataLoader.v1"
+lang = ${nlp.lang}
+tables = ["lexeme_norm"]
+
 [initialize.tokenizer]

en_core_med7_trf-any-py3-none-any.whl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:796d6aa625f06a7a532c1716943a7f3881299c87749328c8b407963686da19d3
-size 445043002
+oid sha256:a3e0f27b7379e0fa26d9b23b63f7d24154208ff3fdd04796da4e4bc9c01a6e01
+size 1018868436

meta.json

@@ -1,19 +1,19 @@
 {
   "lang":"en",
   "name":"core_med7_trf",
-  "version":"3.1.3.1",
+  "version":"3.4.2.1",
   "description":"",
   "author":"Andrey Kormilitzin",
   "email":"[email protected]",
-  "url":"kormilitzin.com",
+  "url":"https://www.kormilitzin.com/",
   "license":"MIT",
-  "spacy_version":">=3.1.4,<3.2.0",
-  "spacy_git_version":"8bda39f08",
+  "spacy_version":">=3.4.2,<3.5.0",
+  "spacy_git_version":"Unknown",
   "vectors":{
-    "width":0,
-    "vectors":0,
-    "keys":0,
-    "name":null
+    "width":300,
+    "vectors":514157,
+    "keys":514157,
+    "name":"en_vectors"
   },
   "labels":{
     "transformer":[
@@ -41,50 +41,50 @@
 
   ],
   "performance":{
-    "ents_f":0.8976388699,
-    "ents_p":0.8910716705,
-    "ents_r":0.9043035886,
+    "ents_f":0.9032835821,
+    "ents_p":0.8822157434,
+    "ents_r":0.925382263,
     "ents_per_type":{
       "DRUG":{
-        "p":0.8898463844,
-        "r":0.9310074481,
-        "f":0.9099616858
+        "p":0.8804185351,
+        "r":0.9349206349,
+        "f":0.9068514242
       },
-      "DOSAGE":{
-        "p":0.8890554723,
-        "r":0.8930722892,
-        "f":0.8910593539
+      "ROUTE":{
+        "p":0.9444444444,
+        "r":0.9739583333,
+        "f":0.958974359
       },
       "STRENGTH":{
-        "p":0.9363057325,
-        "r":0.9625818522,
-        "f":0.9492619926
+        "p":0.9190283401,
+        "r":0.9380165289,
+        "f":0.9284253579
       },
       "FREQUENCY":{
-        "p":0.7804597701,
-        "r":0.7786697248,
-        "f":0.7795637199
+        "p":0.7512437811,
+        "r":0.7947368421,
+        "f":0.7723785166
       },
       "FORM":{
-        "p":0.9034676664,
-        "r":0.8950789229,
-        "f":0.8992537313
+        "p":0.9449541284,
+        "r":0.9493087558,
+        "f":0.9471264368
       },
-      "ROUTE":{
-        "p":0.949945593,
-        "r":0.9247881356,
-        "f":0.9371980676
+      "DOSAGE":{
+        "p":0.872611465,
+        "r":0.958041958,
+        "f":0.9133333333
       },
       "DURATION":{
-        "p":0.7111111111,
-        "r":0.6666666667,
-        "f":0.688172043
+        "p":0.64,
+        "r":0.7619047619,
+        "f":0.6956521739
       }
     },
-    "transformer_loss":2096.0609702383,
-    "ner_loss":8748.9384091962
+    "transformer_loss":25026.2706476041,
+    "ner_loss":1145.767683525
   },
   "requirements":[
-    "spacy-transformers>=1.1.2,<1.2.0"
+    "spacy-transformers>=1.1.6,<1.2.0"
   ]
 }

ner/model

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:efdefbc26ff3a7778078533cf5a62d1b09cf3f8f58830170295ddfad4e3547a1
-size 244778
+oid sha256:3cce46a8bbd4b30427fcc65d07f6d1391d90c46b43ff7d555518c660158e0df9
+size 487730

ner/moves

@@ -1 +1 @@
-��moves�
�{"0":{},"1":{"DRUG":25068,"STRENGTH":19845,"DOSAGE":16674,"FREQUENCY":15332,"FORM":12981,"ROUTE":8181,"DURATION":1940},"2":{"DRUG":25068,"STRENGTH":19845,"DOSAGE":16674,"FREQUENCY":15332,"FORM":12981,"ROUTE":8181,"DURATION":1940},"3":{"DRUG":25068,"STRENGTH":19845,"DOSAGE":16674,"FREQUENCY":15332,"FORM":12981,"ROUTE":8181,"DURATION":1940},"4":{"DRUG":25068,"STRENGTH":19845,"DOSAGE":16674,"FREQUENCY":15332,"FORM":12981,"ROUTE":8181,"DURATION":1940,"":1},"5":{"":1}}�cfg��neg_key�
+��moves�
�{"0":{},"1":{"DRUG":27417,"STRENGTH":21625,"DOSAGE":18350,"FREQUENCY":16642,"FORM":14267,"ROUTE":8996,"DURATION":2140},"2":{"DRUG":27417,"STRENGTH":21625,"DOSAGE":18350,"FREQUENCY":16642,"FORM":14267,"ROUTE":8996,"DURATION":2140},"3":{"DRUG":27417,"STRENGTH":21625,"DOSAGE":18350,"FREQUENCY":16642,"FORM":14267,"ROUTE":8996,"DURATION":2140},"4":{"DRUG":27417,"STRENGTH":21625,"DOSAGE":18350,"FREQUENCY":16642,"FORM":14267,"ROUTE":8996,"DURATION":2140,"":1},"5":{"":1}}�cfg��neg_key�

vocab/key2row

@@ -1 +1,3 @@
-�
+version https://git-lfs.github.com/spec/v1
+oid sha256:31566ae010da3d399eb1d930ae142757afd2601034a4be3bdb00d18881c8c06a
+size 7066303

vocab/lookups.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:76be8b528d0075f7aae98d6fa57a6d3c83ae480a8469e668d7b0af968995ac71
-size 1
+oid sha256:1ddd140ecac6a8c4592e9146d6e30074569ffaed97ee51edc9587dc510f8934c
+size 69982

vocab/strings.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bbbc0cbaea5d8f3124a60adbf573d2826f8294f5d0d10719045d57e792253753
-size 833590
+oid sha256:0b1693baa4ec8d99e20fa93b38ecafa6b4f49f244472d455f35e7605f20345e0
+size 10856674

vocab/vectors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:14772b683e726436d5948ad3fff2b43d036ef2ebbe3458aafed6004e05a40706
-size 128
+oid sha256:234dcf234bfdf01775ae6182715d55eaacfcde8555b189f25440b56d3c39fd5d
+size 616988528

vocab/vectors.cfg

@@ -0,0 +1,3 @@
+{
+  "mode":"default"
+}