Update spaCy pipeline

README.md

@@ -25,11 +25,11 @@ model-index:
 | Feature | Description |
 | --- | --- |
 | **Name** | `en_tako_query_analyzer` |
-| **Version** | `0.0.4` |
-| **spaCy** | `>=3.7.5,<3.8.0` |
+| **Version** | `0.0.5` |
+| **spaCy** | `>=3.8.11,<3.9.0` |
 | **Default Pipeline** | `tok2vec_small`, `tagger`, `parser`, `attribute_ruler`, `senter`, `tok2vec`, `ner`, `textcat_multilabel`, `textcat_classify` |
 | **Components** | `tok2vec_small`, `tagger`, `parser`, `attribute_ruler`, `senter`, `tok2vec`, `ner`, `textcat_multilabel`, `textcat_classify` |
-| **Vectors** | 514157 keys, 514157 unique vectors (300 dimensions) |
+| **Vectors** | 684830 keys, 342918 unique vectors (300 dimensions) |
 | **Sources** | n/a |
 | **License** | n/a |
 | **Author** | [n/a]() |
@@ -58,12 +58,12 @@ model-index:
 | `ENTS_P` | 0.00 |
 | `ENTS_R` | 0.00 |
 | `ENTS_PER_TYPE` | 0.00 |
-| `CATS_SCORE` | 82.56 |
-| `CATS_MICRO_P` | 82.30 |
-| `CATS_MICRO_R` | 82.30 |
-| `CATS_MICRO_F` | 82.30 |
-| `CATS_MACRO_P` | 82.33 |
-| `CATS_MACRO_R` | 82.30 |
-| `CATS_MACRO_F` | 82.30 |
-| `CATS_MACRO_AUC` | 89.93 |
-| `TEXTCAT_CLASSIFY_LOSS` | 218.20 |
+| `CATS_SCORE` | 84.62 |
+| `CATS_MICRO_P` | 83.50 |
+| `CATS_MICRO_R` | 83.50 |
+| `CATS_MICRO_F` | 83.50 |
+| `CATS_MACRO_P` | 84.23 |
+| `CATS_MACRO_R` | 83.50 |
+| `CATS_MACRO_F` | 83.41 |
+| `CATS_MACRO_AUC` | 83.96 |
+| `TEXTCAT_CLASSIFY_LOSS` | 1108.70 |

config.cfg

@@ -1,13 +1,13 @@
 [paths]
-train = "corpus/filter-train.spacy"
-dev = "corpus/filter-test.spacy"
+train = "/Users/noahjax/Work/tako-entity-extractor/corpus/filter-train.spacy"
+dev = "/Users/noahjax/Work/tako-entity-extractor/corpus/filter-test.spacy"
 vectors = "en_core_web_lg"
 init_tok2vec = null
 
 [variables]
 wandb_project_name = "tako-query-filter"
 wandb_team_name = "tako-team"
-base_model = "topic/upbeat-bush"
+base_model = "/Users/noahjax/Work/tako-entity-extractor/training/topic/model-best"
 
 [system]
 gpu_allocator = "pytorch"
@@ -296,7 +296,7 @@ project_name = ${variables.wandb_project_name}
 remove_config_values = []
 model_log_interval = null
 log_dataset_dir = null
-entity = null
+entity = ${variables.wandb_team_name}
 run_name = null
 log_best_dir = null
 log_latest_dir = null
@@ -315,6 +315,10 @@ learn_rate = 0.001
 
 [training.score_weights]
 tag_acc = 0.25
+pos_acc = 0.0
+tag_micro_p = null
+tag_micro_r = null
+tag_micro_f = null
 dep_uas = 0.12
 dep_las = 0.12
 dep_las_per_type = null
@@ -353,7 +357,7 @@ positive_label = "ACCEPT"
 
 [initialize.components.textcat_classify.labels]
 @readers = "spacy.read_labels.v1"
-path = "corpus/labels/filter-labels/textcat_classify.json"
+path = "/Users/noahjax/Work/tako-entity-extractor/corpus/labels/filter-labels/textcat_classify.json"
 require = false
 
 [initialize.tokenizer]

en_tako_query_analyzer-any-py3-none-any.whl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e8de7f18d2cc25ababfa5d8131bb5a3d3e4ba8415066bf97493f4b388da4c6e9
-size 649860398
+oid sha256:5b34711c5f8f8e7d2de88f759dbaf91c708027038921877018f9a56bd186711d
+size 462333461

meta.json

@@ -1,18 +1,18 @@
 {
   "lang":"en",
   "name":"tako_query_analyzer",
-  "version":"0.0.4",
+  "version":"0.0.5",
   "description":"",
   "author":"",
   "email":"",
   "url":"",
   "license":"",
-  "spacy_version":">=3.7.5,<3.8.0",
-  "spacy_git_version":"a6d0fc360",
+  "spacy_version":">=3.8.11,<3.9.0",
+  "spacy_git_version":"e7a662a",
   "vectors":{
     "width":300,
-    "vectors":514157,
-    "keys":514157,
+    "vectors":342918,
+    "keys":684830,
     "name":"en_vectors"
   },
   "labels":{
@@ -190,30 +190,30 @@
     "ents_p":0.0,
     "ents_r":0.0,
     "ents_per_type":0.0,
-    "cats_score":0.8256157635,
+    "cats_score":0.8462255359,
     "cats_score_desc":"F (ACCEPT)",
-    "cats_micro_p":0.823,
-    "cats_micro_r":0.823,
-    "cats_micro_f":0.823,
-    "cats_macro_p":0.8232909619,
-    "cats_macro_r":0.823,
-    "cats_macro_f":0.822960166,
-    "cats_macro_auc":0.899297,
+    "cats_micro_p":0.835,
+    "cats_micro_r":0.835,
+    "cats_micro_f":0.835,
+    "cats_macro_p":0.8422963898,
+    "cats_macro_r":0.835,
+    "cats_macro_f":0.8341160042,
+    "cats_macro_auc":0.83958575,
     "cats_f_per_type":{
       "ACCEPT":{
-        "p":0.813592233,
-        "r":0.838,
-        "f":0.8256157635
+        "p":0.7923211169,
+        "r":0.908,
+        "f":0.8462255359
       },
       "REJECT":{
-        "p":0.8329896907,
-        "r":0.808,
-        "f":0.8203045685
+        "p":0.8922716628,
+        "r":0.762,
+        "f":0.8220064725
       }
     },
-    "textcat_classify_loss":2.1820269685
+    "textcat_classify_loss":11.0869617185
   },
   "requirements":[
-
+    "spacy>=3.8.11,<3.9.0"
   ]
 }

ner/model

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a18b22cda1f4ebaa4786ba94b7baa9379ecca5475386ddd9c8b258717defc3a2
+oid sha256:b6c629c28cd04f736632f4c75ee41c7b45439a38dc77b72bd20b2c6dd4fcd5f1
 size 2715735

parser/model

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a1836fbc02b3924b2fd5f65325c58ae852ff112db1090ca724e5a801e68b85fd
+oid sha256:e0098910535a863d430082ece57455d5fa071cd4ca3a0054e80582076c752b1e
 size 319909

senter/model

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e62c2504dcdc66144b30c048100af05f12207f933f6d669982b947ae71ffdeef
+oid sha256:b69e7174a0b6307739adec3c8053ac45fdbad7171c4d3d8bfcd9713ecbc01f8c
 size 197089

textcat_classify/model

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:43f147a53892c746a8f24180b5870c414902c640b10f9ca2927ed2b91a8bf5fd
+oid sha256:8fdfd1bf415ae90c99273cb57c6a7ab8e1d430865ff0946c5f3fde3d827c78f9
 size 36071049

textcat_multilabel/model

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c238f141aa54e070622e6506599093673815490619c560feee160d93c37a502a
+oid sha256:d9cf2c46c7490c9aa43b83356e1b2cec3e2db257ad4b36cd22c61ea92d9254fd
 size 15382175

tok2vec/model

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f8f59ffdaeb93abafba597c7089b697ab873950f84b9b9bb5fe16022de2c9702
+oid sha256:ba7649d58640d75ed1b18f347ab2c0dedf316e46840afcb898b75d6db692239b
 size 28290008

tok2vec_small/model

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:42d8414521eaf75f817bd1b351b26039a22a912bb2617f95ead305420f2ebffd
+oid sha256:e84fc06eb319c94d28e460fc334e292120b01f18baa5dc8b50c977459820a090
 size 6269370

vocab/key2row

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:31566ae010da3d399eb1d930ae142757afd2601034a4be3bdb00d18881c8c06a
-size 7066303
+oid sha256:c8163b927a234a675074bb38ce62c17a57182998dc83fb9275d35500559a582a
+size 9311659

vocab/strings.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:33316624b8e8ce9be5d4762929948a8c6bef9a9d6a138aae4110f40cd54858a3
-size 10662076
+oid sha256:ed26b6bae75e5f9901b6d7c87c8293d75a292d81814631f55d0b046d9857888d
+size 10042571

vocab/vectors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:234dcf234bfdf01775ae6182715d55eaacfcde8555b189f25440b56d3c39fd5d
-size 616988528
+oid sha256:dd82f972c4fca3d440c505cdd94c88efdded56457cc86851d584b751f7dea673
+size 411501728