add Falcon files

tutorials/llm_clinical_trials/configs/ner_dolly.cfg

@@ -0,0 +1,17 @@
+[nlp]
+lang = "en"
+pipeline = ["llm"]
+batch_size = 128
+
+[components]
+
+[components.llm]
+factory = "llm"
+
+[components.llm.model]
+@llm_models = "spacy.Dolly.v1"
+name = "dolly-v2-3b"
+
+[components.llm.task]
+@llm_tasks = "spacy.NER.v2"
+labels = "Drug,Dose"

tutorials/llm_clinical_trials/configs/ner_falcon.cfg

@@ -0,0 +1,17 @@
+[nlp]
+lang = "en"
+pipeline = ["llm"]
+batch_size = 128
+
+[components]
+
+[components.llm]
+factory = "llm"
+
+[components.llm.model]
+@llm_models = "spacy.Falcon.v1"
+name = "falcon-7b-instruct"
+
+[components.llm.task]
+@llm_tasks = "spacy.NER.v2"
+labels = "Drug,Dose"

tutorials/llm_clinical_trials/falcon_requirements.txt

@@ -0,0 +1,4 @@
+cupy-cuda117
+torch==1.13.1+cu117 -f https://download.pytorch.org/whl/torch_stable.html
+transformers
+einops

tutorials/llm_clinical_trials/project.yml

@@ -2,7 +2,7 @@ title: 'Clinical trial results extraction with LLMs'
 description: "Using an LLM in a spaCy pipeline to extract patient groups, treatments and outcomes in clinical trials."
 
 vars:
-  ner_config: "ner_openai.cfg"
+  ner_config: "ner_dolly.cfg" # "ner_falcon.cfg"  # "ner_openai.cfg"
   trial_config: "trial_openai.cfg"
   pmid: 27144689
 

tutorials/llm_clinical_trials/scripts/visualise_entities.py

@@ -13,8 +13,12 @@ def visualise_entities(pmid: int, config_path: Path, verbose: bool = False):
     text = read_trial(pmid, verbose=verbose)
     nlp = assemble(config_path)
     doc = nlp(text)
-    options = {"ents": ["Drug", "Dose"], "colors": {"Drug": "pink", "Dose": "orange"}}
-    displacy.serve(doc, style="ent", options=options)
+    # options = {"ents": ["Drug", "Dose"], "colors": {"Drug": "pink", "Dose": "orange"}}
+    ents = list(doc.ents)
+    print("ents", len(ents))
+    for ent in ents:
+        print(ent.text, ent.label_)
+    #displacy.serve(doc, style="ent", options=options)
 
 
 if __name__ == "__main__":