Llama v2 finetuned for redaction of PIIs

README.md

@@ -0,0 +1,9 @@
+---
+tags:
+- autotrain
+- text-generation
+widget:
+- text: "I love AutoTrain because "
+---
+
+# Model Trained Using AutoTrain

adapter_config.json

@@ -0,0 +1,21 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "abhishek/llama-2-7b-hf-small-shards",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 32,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 16,
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "v_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07c719f923d3e3878123962fed4f017d052c18a2b2d7203336b974fadfa2b786
+size 33600461

added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "<pad>": 32000
+}

dataset.json

@@ -0,0 +1,112 @@
+[
+    {
+        "instruction": "redact patient_name, age, disease from the text",
+        "input": "John Doe is a patient who is 45 years old and has been diagnosed with COVID-19. He is currently under observation and receiving appropriate medical care. The doctors are closely monitoring his condition and providing necessary treatment.",
+        "output": "<redacted> is a patient who is <redacted> years old and has been diagnosed with <redacted>. He is currently under observation and receiving appropriate medical care. The doctors are closely monitoring his condition and providing necessary treatment."
+    },
+    {
+        "instruction": "redact patient_name, age, disease from the text",
+        "input": "Jane Smith, aged 32, is currently receiving treatment for Diabetes. She has been prescribed medications and advised to follow a specific diet and lifestyle changes to manage her condition effectively. Regular check-ups are scheduled to monitor her progress.",
+        "output": "<redacted>, aged <redacted>, is currently receiving treatment for <redacted>. She has been prescribed medications and advised to follow a specific diet and lifestyle changes to manage her condition effectively. Regular check-ups are scheduled to monitor her progress."
+    },
+    {
+        "instruction": "redact patient_name, age, disease from the text",
+        "input": "Mary Johnson's Patient Information indicates she is 55 years old and suffering from Hypertension. She has been advised to take her prescribed medications regularly and make lifestyle modifications to control her blood pressure levels. Regular follow-ups with her healthcare provider are scheduled.",
+        "output": "<redacted>'s Patient Information indicates she is <redacted> years old and suffering from <redacted>. She has been advised to take her prescribed medications regularly and make lifestyle modifications to control her blood pressure levels. Regular follow-ups with her healthcare provider are scheduled."
+    },
+    {
+        "instruction": "redact patient_name, age, disease from the text",
+        "input": "Mr. Smith is a 60-year-old man with a diagnosis of Heart Disease. He is under the care of a cardiologist and following a treatment plan to manage his condition. Lifestyle changes, including a heart-healthy diet and exercise, are recommended for his overall well-being.",
+        "output": "<redacted> is a <redacted>-year-old man with a diagnosis of <redacted>. He is under the care of a cardiologist and following a treatment plan to manage his condition. Lifestyle changes, including a heart-healthy diet and exercise, are recommended for his overall well-being."
+    },
+    {
+        "instruction": "redact patient_name, age, disease from the text",
+        "input": "Sarah Miller, aged 40, has been diagnosed with Cancer. She is undergoing chemotherapy and radiation treatments as part of her cancer treatment plan. Support from family and friends has been essential during her journey to recovery.",
+        "output": "<redacted>, aged <redacted>, has been diagnosed with <redacted>. She is undergoing chemotherapy and radiation treatments as part of her cancer treatment plan. Support from family and friends has been essential during her journey to recovery."
+    },
+    {
+        "instruction": "redact patient_name, age, disease from the text",
+        "input": "James Johnson is 28 years old and currently undergoing treatment for COVID-19. He is in isolation to prevent further spread of the virus. The healthcare team is closely monitoring his symptoms and providing supportive care.",
+        "output": "<redacted> is <redacted> years old and currently undergoing treatment for <redacted>. He is in isolation to prevent further spread of the virus. The healthcare team is closely monitoring his symptoms and providing supportive care."
+    },
+    {
+        "instruction": "redact patient_name, age, disease from the text",
+        "input": "Emma Smith, aged 38, has been diagnosed with Asthma. She carries an inhaler with her at all times for quick relief during asthma attacks. Emma is advised to avoid triggers that worsen her symptoms.",
+        "output": "<redacted>, aged <redacted>, has been diagnosed with <redacted>. She carries an inhaler with her at all times for quick relief during asthma attacks. <redacted> is advised to avoid triggers that worsen her symptoms."
+    },
+    {
+        "instruction": "redact patient_name, age, disease from the text",
+        "input": "Laura Johnson's Diagnosis indicates she is 50 years old and suffering from Diabetes. She is managing her blood sugar levels through medication and lifestyle changes, including a balanced diet and regular exercise.",
+        "output": "<redacted>'s Diagnosis indicates she is <redacted> years old and suffering from <redacted>. She is managing her blood sugar levels through medication and lifestyle changes, including a balanced diet and regular exercise."
+    },
+    {
+        "instruction": "redact patient_name, age, disease from the text",
+        "input": "Lorem ipsum dolor sit amet, consectetur adipiscing elit. Patient Name: John Doe, Age: 45, Disease: COVID-19. Nulla venenatis accumsan id a enim. Integer quis dui sit amet est pellentesque consectetur. Etiam lobortis, quam et lacinia condimentum, urna quam varius velit, ut ultricies neque quam vitae odio.",
+        "output": "Lorem ipsum dolor sit amet, consectetur adipiscing elit. Patient Name: <redacted>, Age: <redacted>, Disease: <redacted>. Nulla venenatis accumsan id a enim. Integer quis dui sit amet est pellentesque consectetur. Etiam lobortis, quam et lacinia condimentum, urna quam varius velit, ut ultricies neque quam vitae odio."
+    },
+    {
+        "instruction": "redact patient_name, age, disease from the text",
+        "input": "Nullam tristique purus ac sapien egestas. Patient: Jane Smith, Age: 32, Diagnosis: Diabetes. Phasellus venenatis nisi eget ligula varius luctus. Aenean ac metus dui. Sed sit amet volutpat elit. Mauris eleifend, nulla vel rhoncus tempus, odio arcu tempus libero, ut scelerisque arcu lectus at mi.",
+        "output": "Nullam tristique purus ac sapien egestas. Patient: <redacted>, Age: <redacted>, Diagnosis: <redacted>. Phasellus venenatis nisi eget ligula varius luctus. Aenean ac metus dui. Sed sit amet volutpat elit. Mauris eleifend, nulla vel rhoncus tempus, odio arcu tempus libero, ut scelerisque arcu lectus at mi."
+    },
+    {
+        "instruction": "redact patient_name, age, disease from the text",
+        "input": "Proin sagittis sapien vitae lorem dictum. Patient Information: Mary Johnson, Age: 55, Disease: Hypertension. Fusce sit amet quam nec odio posuere vestibulum vel in erat. Vivamus eleifend nunc purus, a fringilla ligula venenatis ac. In semper quam quis quam varius, quis rutrum lacus interdum.",
+        "output": "Proin sagittis sapien vitae lorem dictum. Patient Information: <redacted>, Age: <redacted>, Disease: <redacted>. Fusce sit amet quam nec odio posuere vestibulum vel in erat. Vivamus eleifend nunc purus, a fringilla ligula venenatis ac. In semper quam quis quam varius, quis rutrum lacus interdum."
+    },
+    {
+        "instruction": "redact patient_name, age, disease from the text",
+        "input": "Lorem ipsum dolor sit amet, consectetur adipiscing elit. Patient Name: Alice Johnson, Age: 27, Disease: COVID-19. Nulla venenatis accumsan id a enim. Integer quis dui sit amet est pellentesque consectetur. Etiam lobortis, quam et lacinia condimentum, urna quam varius velit, ut ultricies neque quam vitae odio.",
+        "output": "Lorem ipsum dolor sit amet, consectetur adipiscing elit. Patient Name: <redacted>, Age: <redacted>, Disease: <redacted>. Nulla venenatis accumsan id a enim. Integer quis dui sit amet est pellentesque consectetur. Etiam lobortis, quam et lacinia condimentum, urna quam varius velit, ut ultricies neque quam vitae odio."
+    },
+    {
+        "instruction": "redact patient_name, age, disease from the text",
+        "input": "Nullam tristique purus ac sapien egestas. Patient: James Smith, Age: 50, Diagnosis: Asthma. Phasellus venenatis nisi eget ligula varius luctus. Aenean ac metus dui. Sed sit amet volutpat elit. Mauris eleifend, nulla vel rhoncus tempus, odio arcu tempus libero, ut scelerisque arcu lectus at mi.",
+        "output": "Nullam tristique purus ac sapien egestas. Patient: <redacted>, Age: <redacted>, Diagnosis: <redacted>. Phasellus venenatis nisi eget ligula varius luctus. Aenean ac metus dui. Sed sit amet volutpat elit. Mauris eleifend, nulla vel rhoncus tempus, odio arcu tempus libero, ut scelerisque arcu lectus at mi."
+    },
+    {
+        "instruction": "redact patient_name, age, disease from the text",
+        "input": "Proin sagittis sapien vitae lorem dictum. Patient Information: Mary Johnson, Age: 55, Disease: Hypertension. Fusce sit amet quam nec odio posuere vestibulum vel in erat. Vivamus eleifend nunc purus, a fringilla ligula venenatis ac. In semper quam quis quam varius, quis rutrum lacus interdum.",
+        "output": "Proin sagittis sapien vitae lorem dictum. Patient Information: <redacted>, Age: <redacted>, Disease: <redacted>. Fusce sit amet quam nec odio posuere vestibulum vel in erat. Vivamus eleifend nunc purus, a fringilla ligula venenatis ac. In semper quam quis quam varius, quis rutrum lacus interdum."
+    },
+    {
+        "instruction": "redact patient_name, age, disease from the text",
+        "input": "Lorem ipsum dolor sit amet, consectetur adipiscing elit. Patient Name: John Doe, Age: 45, Disease: COVID-19. Nulla venenatis accumsan id a enim. Integer quis dui sit amet est pellentesque consectetur. Etiam lobortis, quam et lacinia condimentum, urna quam varius velit, ut ultricies neque quam vitae odio.",
+        "output": "Lorem ipsum dolor sit amet, consectetur adipiscing elit. Patient Name: <redacted>, Age: <redacted>, Disease: <redacted>. Nulla venenatis accumsan id a enim. Integer quis dui sit amet est pellentesque consectetur. Etiam lobortis, quam et lacinia condimentum, urna quam varius velit, ut ultricies neque quam vitae odio."
+    },
+    {
+        "instruction": "redact patient_name, age, disease from the text",
+        "input": "Nullam tristique purus ac sapien egestas. Patient: Jane Smith, Age: 32, Diagnosis: Diabetes. Phasellus venenatis nisi eget ligula varius luctus. Aenean ac metus dui. Sed sit amet volutpat elit. Mauris eleifend, nulla vel rhoncus tempus, odio arcu tempus libero, ut scelerisque arcu lectus at mi.",
+        "output": "Nullam tristique purus ac sapien egestas. Patient: <redacted>, Age: <redacted>, Diagnosis: <redacted>. Phasellus venenatis nisi eget ligula varius luctus. Aenean ac metus dui. Sed sit amet volutpat elit. Mauris eleifend, nulla vel rhoncus tempus, odio arcu tempus libero, ut scelerisque arcu lectus at mi."
+    },
+    {
+        "instruction": "redact patient_name, age, disease from the text",
+        "input": "Proin sagittis sapien vitae lorem dictum. Patient Information: Mary Johnson, Age: 55, Disease: Hypertension. Fusce sit amet quam nec odio posuere vestibulum vel in erat. Vivamus eleifend nunc purus, a fringilla ligula venenatis ac. In semper quam quis quam varius, quis rutrum lacus interdum.",
+        "output": "Proin sagittis sapien vitae lorem dictum. Patient Information: <redacted>, Age: <redacted>, Disease: <redacted>. Fusce sit amet quam nec odio posuere vestibulum vel in erat. Vivamus eleifend nunc purus, a fringilla ligula venenatis ac. In semper quam quis quam varius, quis rutrum lacus interdum."
+    },
+    {
+        "instruction": "redact patient_name, age, disease from the text",
+        "input": "Lorem ipsum dolor sit amet, consectetur adipiscing elit. Patient Name: Alice Johnson, Age: 27, Disease: COVID-19. Nulla venenatis accumsan id a enim. Integer quis dui sit amet est pellentesque consectetur. Etiam lobortis, quam et lacinia condimentum, urna quam varius velit, ut ultricies neque quam vitae odio.",
+        "output": "Lorem ipsum dolor sit amet, consectetur adipiscing elit. Patient Name: <redacted>, Age: <redacted>, Disease: <redacted>. Nulla venenatis accumsan id a enim. Integer quis dui sit amet est pellentesque consectetur. Etiam lobortis, quam et lacinia condimentum, urna quam varius velit, ut ultricies neque quam vitae odio."
+    },
+    {
+        "instruction": "redact patient_name, age, disease from the text",
+        "input": "Nullam tristique purus ac sapien egestas. Patient: James Smith, Age: 50, Diagnosis: Asthma. Phasellus venenatis nisi eget ligula varius luctus. Aenean ac metus dui. Sed sit amet volutpat elit. Mauris eleifend, nulla vel rhoncus tempus, odio arcu tempus libero, ut scelerisque arcu lectus at mi.",
+        "output": "Nullam tristique purus ac sapien egestas. Patient: <redacted>, Age: <redacted>, Diagnosis: <redacted>. Phasellus venenatis nisi eget ligula varius luctus. Aenean ac metus dui. Sed sit amet volutpat elit. Mauris eleifend, nulla vel rhoncus tempus, odio arcu tempus libero, ut scelerisque arcu lectus at mi."
+    },
+    {
+        "instruction": "redact patient_name, age, disease from the text",
+        "input": "Proin sagittis sapien vitae lorem dictum. Patient Information: Mary Johnson, Age: 55, Disease: Hypertension. Fusce sit amet quam nec odio posuere vestibulum vel in erat. Vivamus eleifend nunc purus, a fringilla ligula venenatis ac. In semper quam quis quam varius, quis rutrum lacus interdum.",
+        "output": "Proin sagittis sapien vitae lorem dictum. Patient Information: <redacted>, Age: <redacted>, Disease: <redacted>. Fusce sit amet quam nec odio posuere vestibulum vel in erat. Vivamus eleifend nunc purus, a fringilla ligula venenatis ac. In semper quam quis quam varius, quis rutrum lacus interdum."
+    },
+    {
+        "instruction": "redact patient_name, age, disease from the text",
+        "input": "Lorem ipsum dolor sit amet, consectetur adipiscing elit. Patient Name: John Doe, Age: 45, Disease: COVID-19. Nulla venenatis accumsan id a enim. Integer quis dui sit amet est pellentesque consectetur. Etiam lobortis, quam et lacinia condimentum, urna quam varius velit, ut ultricies neque quam vitae odio.",
+        "output": "Lorem ipsum dolor sit amet, consectetur adipiscing elit. Patient Name: <redacted>, Age: <redacted>, Disease: <redacted>. Nulla venenatis accumsan id a enim. Integer quis dui sit amet est pellentesque consectetur. Etiam lobortis, quam et lacinia condimentum, urna quam varius velit, ut ultricies neque quam vitae odio."
+    },
+    {
+        "instruction": "redact patient_name, age, disease from the text",
+        "input": "Nullam tristique purus ac sapien egestas. Patient: Jane Smith, Age: 32, Diagnosis: Diabetes. Phasellus venenatis nisi eget ligula varius luctus. Aenean ac metus dui. Sed sit amet volutpat elit. Mauris eleifend, nulla vel rhoncus tempus, odio arcu tempus libero, ut scelerisque arcu lectus at mi.",
+        "output": "Nullam tristique purus ac sapien egestas. Patient: <redacted>, Age: <redacted>, Diagnosis: <redacted>. Phasellus venenatis nisi eget ligula varius luctus. Aenean ac metus dui. Sed sit amet volutpat elit. Mauris eleifend, nulla vel rhoncus tempus, odio arcu tempus libero, ut scelerisque arcu lectus at mi."
+    }
+]

special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<unk>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}