Upload 19 files

README.md

@@ -1,13 +1,2 @@
----
-title: MediNER
-emoji: 🌍
-colorFrom: green
-colorTo: gray
-sdk: streamlit
-sdk_version: 1.43.1
-app_file: app.py
-pinned: false
-short_description: Performs named entity recognition for medical entities.
----
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# medNER_V2
+ This app performs Named Entity REcognition of medical entties.

checkpoint-92/config.json

@@ -0,0 +1,195 @@
+{
+  "_name_or_path": "dslim/distilbert-NER",
+  "activation": "gelu",
+  "architectures": [
+    "DistilBertForTokenClassification"
+  ],
+  "attention_dropout": 0.1,
+  "dim": 768,
+  "dropout": 0.1,
+  "hidden_dim": 3072,
+  "id2label": {
+    "0": "O",
+    "1": "B-ACTIVITY",
+    "2": "I-ACTIVITY",
+    "3": "I-ADMINISTRATION",
+    "4": "B-ADMINISTRATION",
+    "5": "B-AGE",
+    "6": "I-AGE",
+    "7": "I-AREA",
+    "8": "B-AREA",
+    "9": "B-BIOLOGICAL_ATTRIBUTE",
+    "10": "I-BIOLOGICAL_ATTRIBUTE",
+    "11": "I-BIOLOGICAL_STRUCTURE",
+    "12": "B-BIOLOGICAL_STRUCTURE",
+    "13": "B-CLINICAL_EVENT",
+    "14": "I-CLINICAL_EVENT",
+    "15": "B-COLOR",
+    "16": "I-COLOR",
+    "17": "I-COREFERENCE",
+    "18": "B-COREFERENCE",
+    "19": "B-DATE",
+    "20": "I-DATE",
+    "21": "I-DETAILED_DESCRIPTION",
+    "22": "B-DETAILED_DESCRIPTION",
+    "23": "I-DIAGNOSTIC_PROCEDURE",
+    "24": "B-DIAGNOSTIC_PROCEDURE",
+    "25": "I-DISEASE_DISORDER",
+    "26": "B-DISEASE_DISORDER",
+    "27": "B-DISTANCE",
+    "28": "I-DISTANCE",
+    "29": "B-DOSAGE",
+    "30": "I-DOSAGE",
+    "31": "I-DURATION",
+    "32": "B-DURATION",
+    "33": "I-FAMILY_HISTORY",
+    "34": "B-FAMILY_HISTORY",
+    "35": "B-FREQUENCY",
+    "36": "I-FREQUENCY",
+    "37": "I-HEIGHT",
+    "38": "B-HEIGHT",
+    "39": "B-HISTORY",
+    "40": "I-HISTORY",
+    "41": "I-LAB_VALUE",
+    "42": "B-LAB_VALUE",
+    "43": "I-MASS",
+    "44": "B-MASS",
+    "45": "I-MEDICATION",
+    "46": "B-MEDICATION",
+    "47": "I-NONBIOLOGICAL_LOCATION",
+    "48": "B-NONBIOLOGICAL_LOCATION",
+    "49": "I-OCCUPATION",
+    "50": "B-OCCUPATION",
+    "51": "B-OTHER_ENTITY",
+    "52": "I-OTHER_ENTITY",
+    "53": "B-OTHER_EVENT",
+    "54": "I-OTHER_EVENT",
+    "55": "I-OUTCOME",
+    "56": "B-OUTCOME",
+    "57": "I-PERSONAL_BACKGROUND",
+    "58": "B-PERSONAL_BACKGROUND",
+    "59": "B-QUALITATIVE_CONCEPT",
+    "60": "I-QUALITATIVE_CONCEPT",
+    "61": "I-QUANTITATIVE_CONCEPT",
+    "62": "B-QUANTITATIVE_CONCEPT",
+    "63": "B-SEVERITY",
+    "64": "I-SEVERITY",
+    "65": "B-SEX",
+    "66": "I-SEX",
+    "67": "B-SHAPE",
+    "68": "I-SHAPE",
+    "69": "B-SIGN_SYMPTOM",
+    "70": "I-SIGN_SYMPTOM",
+    "71": "B-SUBJECT",
+    "72": "I-SUBJECT",
+    "73": "B-TEXTURE",
+    "74": "I-TEXTURE",
+    "75": "B-THERAPEUTIC_PROCEDURE",
+    "76": "I-THERAPEUTIC_PROCEDURE",
+    "77": "I-TIME",
+    "78": "B-TIME",
+    "79": "B-VOLUME",
+    "80": "I-VOLUME",
+    "81": "I-WEIGHT",
+    "82": "B-WEIGHT"
+  },
+  "initializer_range": 0.02,
+  "label2id": {
+    "B-ACTIVITY": 1,
+    "B-ADMINISTRATION": 4,
+    "B-AGE": 5,
+    "B-AREA": 8,
+    "B-BIOLOGICAL_ATTRIBUTE": 9,
+    "B-BIOLOGICAL_STRUCTURE": 12,
+    "B-CLINICAL_EVENT": 13,
+    "B-COLOR": 15,
+    "B-COREFERENCE": 18,
+    "B-DATE": 19,
+    "B-DETAILED_DESCRIPTION": 22,
+    "B-DIAGNOSTIC_PROCEDURE": 24,
+    "B-DISEASE_DISORDER": 26,
+    "B-DISTANCE": 27,
+    "B-DOSAGE": 29,
+    "B-DURATION": 32,
+    "B-FAMILY_HISTORY": 34,
+    "B-FREQUENCY": 35,
+    "B-HEIGHT": 38,
+    "B-HISTORY": 39,
+    "B-LAB_VALUE": 42,
+    "B-MASS": 44,
+    "B-MEDICATION": 46,
+    "B-NONBIOLOGICAL_LOCATION": 48,
+    "B-OCCUPATION": 50,
+    "B-OTHER_ENTITY": 51,
+    "B-OTHER_EVENT": 53,
+    "B-OUTCOME": 56,
+    "B-PERSONAL_BACKGROUND": 58,
+    "B-QUALITATIVE_CONCEPT": 59,
+    "B-QUANTITATIVE_CONCEPT": 62,
+    "B-SEVERITY": 63,
+    "B-SEX": 65,
+    "B-SHAPE": 67,
+    "B-SIGN_SYMPTOM": 69,
+    "B-SUBJECT": 71,
+    "B-TEXTURE": 73,
+    "B-THERAPEUTIC_PROCEDURE": 75,
+    "B-TIME": 78,
+    "B-VOLUME": 79,
+    "B-WEIGHT": 82,
+    "I-ACTIVITY": 2,
+    "I-ADMINISTRATION": 3,
+    "I-AGE": 6,
+    "I-AREA": 7,
+    "I-BIOLOGICAL_ATTRIBUTE": 10,
+    "I-BIOLOGICAL_STRUCTURE": 11,
+    "I-CLINICAL_EVENT": 14,
+    "I-COLOR": 16,
+    "I-COREFERENCE": 17,
+    "I-DATE": 20,
+    "I-DETAILED_DESCRIPTION": 21,
+    "I-DIAGNOSTIC_PROCEDURE": 23,
+    "I-DISEASE_DISORDER": 25,
+    "I-DISTANCE": 28,
+    "I-DOSAGE": 30,
+    "I-DURATION": 31,
+    "I-FAMILY_HISTORY": 33,
+    "I-FREQUENCY": 36,
+    "I-HEIGHT": 37,
+    "I-HISTORY": 40,
+    "I-LAB_VALUE": 41,
+    "I-MASS": 43,
+    "I-MEDICATION": 45,
+    "I-NONBIOLOGICAL_LOCATION": 47,
+    "I-OCCUPATION": 49,
+    "I-OTHER_ENTITY": 52,
+    "I-OTHER_EVENT": 54,
+    "I-OUTCOME": 55,
+    "I-PERSONAL_BACKGROUND": 57,
+    "I-QUALITATIVE_CONCEPT": 60,
+    "I-QUANTITATIVE_CONCEPT": 61,
+    "I-SEVERITY": 64,
+    "I-SEX": 66,
+    "I-SHAPE": 68,
+    "I-SIGN_SYMPTOM": 70,
+    "I-SUBJECT": 72,
+    "I-TEXTURE": 74,
+    "I-THERAPEUTIC_PROCEDURE": 76,
+    "I-TIME": 77,
+    "I-VOLUME": 80,
+    "I-WEIGHT": 81,
+    "O": 0
+  },
+  "max_position_embeddings": 512,
+  "model_type": "distilbert",
+  "n_heads": 12,
+  "n_layers": 6,
+  "output_past": true,
+  "pad_token_id": 0,
+  "qa_dropout": 0.1,
+  "seq_classif_dropout": 0.2,
+  "sinusoidal_pos_embds": false,
+  "tie_weights_": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.48.3",
+  "vocab_size": 28996
+}

checkpoint-92/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:691568d1730026b95fe5a7d1005a616ca010681b8ec88899b99314c510979a77
+size 261031300

checkpoint-92/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bde04177a1dff0cc8be3fc543d4f5a3b313a0501cff485928d58f0838b00f3da
+size 522123450

checkpoint-92/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4c96f5a47b163e72e28fbde7afe1320d6bcc042926cd5cb52bdc6f70d90c6d4d
+size 14244

checkpoint-92/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b2c3a566d1aa24c99821b905999ebcfce96b1faae6ee23e3f85708be8e34a3d
+size 1064

checkpoint-92/special_tokens_map.json

@@ -0,0 +1,37 @@
+{
+  "cls_token": {
+    "content": "[CLS]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "mask_token": {
+    "content": "[MASK]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "[PAD]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "sep_token": {
+    "content": "[SEP]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "[UNK]",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

checkpoint-92/tokenizer_config.json

@@ -0,0 +1,60 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "100": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "101": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "102": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "103": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "do_lower_case": false,
+  "extra_special_tokens": {},
+  "mask_token": "[MASK]",
+  "max_length": 512,
+  "model_max_length": 512,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "stride": 0,
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "DistilBertTokenizer",
+  "truncation_side": "right",
+  "truncation_strategy": "longest_first",
+  "unk_token": "[UNK]"
+}

checkpoint-92/trainer_state.json

@@ -0,0 +1,81 @@
+{
+  "best_metric": 0.76803058385849,
+  "best_model_checkpoint": "/content/drive/MyDrive/Files/checkpoint-92",
+  "epoch": 4.0,
+  "eval_steps": 500,
+  "global_step": 92,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.7553191489361702,
+      "eval_f1": 0.5810435319543069,
+      "eval_loss": 0.9222098588943481,
+      "eval_precision": 0.5859277708592777,
+      "eval_recall": 0.5762400489895897,
+      "eval_runtime": 0.369,
+      "eval_samples_per_second": 54.199,
+      "eval_steps_per_second": 8.13,
+      "step": 23
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.7700754975978037,
+      "eval_f1": 0.6098380690498014,
+      "eval_loss": 0.8292961120605469,
+      "eval_precision": 0.6085365853658536,
+      "eval_recall": 0.6111451316595223,
+      "eval_runtime": 0.3909,
+      "eval_samples_per_second": 51.158,
+      "eval_steps_per_second": 7.674,
+      "step": 46
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.7865477007549759,
+      "eval_f1": 0.6364756623536661,
+      "eval_loss": 0.782767653465271,
+      "eval_precision": 0.6404215747055176,
+      "eval_recall": 0.6325780771586038,
+      "eval_runtime": 0.3839,
+      "eval_samples_per_second": 52.099,
+      "eval_steps_per_second": 7.815,
+      "step": 69
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.789293067947838,
+      "eval_f1": 0.641846153846154,
+      "eval_loss": 0.76803058385849,
+      "eval_precision": 0.645021645021645,
+      "eval_recall": 0.6387017758726271,
+      "eval_runtime": 0.4404,
+      "eval_samples_per_second": 45.415,
+      "eval_steps_per_second": 6.812,
+      "step": 92
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 92,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 94208005324800.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-92/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce032d36c503c8ec4cc7b6976897685f8c25accd569b89148aacf93a4d442372
+size 5368

pages/about.py

@@ -0,0 +1,13 @@
+import streamlit as st
+
+st.title("Info") 
+
+with st.expander("ℹ️ - About this app", expanded=True):
+
+    st.write(
+        """     
+-   This app performs named entity recognition for medical entities. 
+-   myDemo model was developed from dslim/distilbert-NER (a general NER model with 66M parameters) in HuggingFace, and fine-tuned on singh-aditya/MACCROBAT_biomedical_ner (a dataset annotated with medical entity labels in 41 categories). 
+-   The model uses the default pretrained tokenizer in dslim/distilbert-NER.
+       """
+    )

pages/home.py

@@ -0,0 +1,53 @@
+import streamlit as st
+
+st.title("📘Named Entity Recognition")
+
+st.header("Tags the below 41 medical entities")
+
+'ACTIVITY'
+'ADMINISTRATION'
+'AGE'
+'AREA'
+'BIOLOGICAL_ATTRIBUTE'
+'BIOLOGICAL_STRUCTURE'
+'CLINICAL_EVENT'
+'COLOR'
+'COREFERENCE'
+'DATE'
+'DETAILED_DESCRIPTION'
+'DIAGNOSTIC_PROCEDURE'
+'DISEASE_DISORDER'
+'DISTANCE'
+'DOSAGE'
+'DURATION'
+'FAMILY_HISTORY'
+'FREQUENCY'
+'HEIGHT'
+'HISTORY'
+'LAB_VALUE'
+'MASS'
+'MEDICATION'
+'NONBIOLOGICAL_LOCATION'
+'OCCUPATION'
+'OTHER_ENTITY'
+'OUTCOME'
+'PERSONAL_BACKGROUND'
+'QUALITATIVE_CONCEPT'
+'QUANTITATIVE_CONCEPT'
+'SEVERITY'
+'SEX'
+'SHAPE'
+'SIGN_SYMPTOM'
+'SUBJECT'
+'TEXTURE'
+'THERAPEUTIC_PROCEDURE'
+'TIME'
+'VOLUME'
+'WEIGHT'
+
+
+
+
+
+
+

pages/type_text.py

@@ -0,0 +1,71 @@
+import streamlit as st
+import pandas as pd
+from io import StringIO
+import json
+from transformers import pipeline
+#from transformers import AutoTokenizer, AutoModelForTokenClassification
+
+def on_click():
+    st.session_state.user_input = ""
+
+#@st.cache
+def convert_df(df:pd.DataFrame):
+     return df.to_csv(index=False).encode('utf-8')
+
+#@st.cache
+def convert_json(df:pd.DataFrame):
+    result = df.to_json(orient="index")
+    parsed = json.loads(result)
+    json_string = json.dumps(parsed)
+    #st.json(json_string, expanded=True)
+    return json_string
+
+#st.title("📘medical Named Entity Recognition Tagger")
+
+text_input = st.text_input("Type input text and hit Enter", key="user_input")
+st.button("Clear text", on_click=on_click)
+
+my_model_results = pipeline("ner", model= "checkpoint-92")
+HuggingFace_model_results = pipeline("ner", model = "blaze999/Medical-NER")
+
+createNER_button = st.button("Create NER tags")
+
+col1, col2 = st.columns([1,1.5])
+col1.subheader("myDemo Model")
+col2.subheader("blaze999/Medical-NER")
+
+
+dictA = {"word": [], "entity": []}
+dictB = {"word": [], "entity": []}
+
+if text_input is not None and createNER_button == True: 
+
+    with col1:
+        #st.write(my_model_results(text_input))
+        #col1.subheader("myDemo Model")
+        for result in my_model_results(text_input): 
+            st.write(result['word'], result['entity'])
+            dictA["word"].append(result['word']), dictA["entity"].append(result['entity'])
+        dfA = pd.DataFrame.from_dict(dictA)
+        #st.write(dfA)            
+    with col2:
+        #st.write(HuggingFace_model_results(text_input))
+        #col2.subheader("Hugging Face Model")
+        for result in HuggingFace_model_results(text_input):
+            st.write(result['word'], result['entity'])
+            dictB["word"].append(result['word']), dictB["entity"].append(result['entity'])         
+        dfB = pd.DataFrame.from_dict(dictB)
+        #st.write(dfB)
+     
+    bs, b1, b2, b3, bLast = st.columns([0.75, 1.5, 1.5, 1.5, 0.75])
+    with b1:
+        #csvbutton = download_button(results, "results.csv", "📥 Download .csv")
+        csvbutton = st.download_button(label="📥 Download .csv", data=convert_df(dfA), file_name= "results.csv", mime='text/csv', key='csv_b')
+    with b2:
+        #textbutton = download_button(results, "results.txt", "📥 Download .txt")
+        textbutton = st.download_button(label="📥 Download .txt", data=convert_df(dfA), file_name= "results.text", mime='text/plain',  key='text_b')
+    with b3:
+        #jsonbutton = download_button(results, "results.json", "📥 Download .json")
+        jsonbutton = st.download_button(label="📥 Download .json", data=convert_json(dfA), file_name= "results.json", mime='application/json',  key='json_b')
+
+

pages/upload_file.py

@@ -0,0 +1,76 @@
+import streamlit as st
+import pandas as pd
+from io import StringIO
+import json
+from transformers import pipeline
+#from transformers import AutoTokenizer, AutoModelForTokenClassification
+
+def on_click():
+    st.session_state.user_input = ""
+ 
+#@st.cache
+def convert_df(df:pd.DataFrame):
+     return df.to_csv(index=False).encode('utf-8')
+
+#@st.cache
+def convert_json(df:pd.DataFrame):
+    result = df.to_json(orient="index")
+    parsed = json.loads(result)
+    json_string = json.dumps(parsed)
+    #st.json(json_string, expanded=True)
+    return json_string
+
+#st.title("📘medical Named Entity Recognition Tagger")
+
+uploaded_file = st.file_uploader(label = "Upload single text file")
+if uploaded_file is not None: 
+    stringio = StringIO(uploaded_file.getvalue().decode("utf-8"))
+    string_data = stringio.read()
+    st.success('Your file input is: '+ string_data, icon="✅")
+
+
+my_model_results = pipeline("ner", model= "checkpoint-92")
+HuggingFace_model_results = pipeline("ner", model = "blaze999/Medical-NER")
+
+
+createNER_button = st.button("Create NER tags")
+
+col1, col2 = st.columns([1,1.5])
+col1.subheader("myDemo Model")
+col2.subheader("blaze999/Medical-NER")
+
+if uploaded_file is not None and createNER_button == True: 
+    dict1 = {"word": [], "entity": []}
+    dict2 = {"word": [], "entity": []}
+    #stringio = StringIO(uploaded_file.getvalue().decode("utf-8"))
+    #string_data = stringio.read()
+    #st.write("Your input is: ", string_data)
+    with col1: 
+        #st.write(my_model_results(string_data))  
+        #col1.subheader("myDemo Model")
+        for result in my_model_results(string_data): 
+            st.write(result['word'], result['entity'])
+            dict1["word"].append(result['word']), dict1["entity"].append(result['entity'])         
+        df1 = pd.DataFrame.from_dict(dict1)
+        #st.write(df1)
+    with col2:
+        #st.write(HuggingFace_model_results(string_data))
+        #col2.subheader("Hugging Face Model")
+        for result in HuggingFace_model_results(string_data):
+            st.write(result['word'], result['entity'])     
+            dict2["word"].append(result['word']), dict2["entity"].append(result['entity'])
+        df2 = pd.DataFrame.from_dict(dict2)
+        #st.write(df2)
+
+
+    cs, c1, c2, c3, cLast = st.columns([0.75, 1.5, 1.5, 1.5, 0.75])
+    with c1:
+        #csvbutton = download_button(results, "results.csv", "📥 Download .csv")
+        csvbutton = st.download_button(label="📥 Download .csv", data=convert_df(df1), file_name= "results.csv", mime='text/csv', key='csv')
+    with c2:
+        #textbutton = download_button(results, "results.txt", "📥 Download .txt")
+        textbutton = st.download_button(label="📥 Download .txt", data=convert_df(df1), file_name= "results.text", mime='text/plain',  key='text')
+    with c3:
+        #jsonbutton = download_button(results, "results.json", "📥 Download .json")
+        jsonbutton = st.download_button(label="📥 Download .json", data=convert_json(df1), file_name= "results.json", mime='application/json',  key='json')
+

requirements.txt

@@ -0,0 +1,4 @@
+streamlit==1.41.1
+pandas==2.2.2
+torch==2.4.0
+transformers==4.44.2