Initial commit of Healthbot files

cbt-tinyllama/cbt-tinyllama-merged/chat_template.jinja

@@ -0,0 +1,15 @@
+{% for message in messages %}
+{% if message['role'] == 'user' %}
+{{ '<|user|>
+' + message['content'] + eos_token }}
+{% elif message['role'] == 'system' %}
+{{ '<|system|>
+' + message['content'] + eos_token }}
+{% elif message['role'] == 'assistant' %}
+{{ '<|assistant|>
+'  + message['content'] + eos_token }}
+{% endif %}
+{% if loop.last and add_generation_prompt %}
+{{ '<|assistant|>' }}
+{% endif %}
+{% endfor %}

cbt-tinyllama/cbt-tinyllama-merged/config.json

@@ -0,0 +1,29 @@
+{
+  "architectures": [
+    "LlamaForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "head_dim": 64,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "initializer_range": 0.02,
+  "intermediate_size": 5632,
+  "max_position_embeddings": 2048,
+  "mlp_bias": false,
+  "model_type": "llama",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 22,
+  "num_key_value_heads": 4,
+  "pretraining_tp": 1,
+  "rms_norm_eps": 1e-05,
+  "rope_scaling": null,
+  "rope_theta": 10000.0,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.52.2",
+  "use_cache": true,
+  "vocab_size": 32000
+}

cbt-tinyllama/cbt-tinyllama-merged/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b847be0c992dbe7bc7eb870b0d3b67bc1e02cee034771329d6a53da6f580496c
+size 4400276670

cbt-tinyllama/cbt-tinyllama-merged/special_tokens_map.json

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

cbt-tinyllama/cbt-tinyllama-merged/tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
+size 499723

cbt-tinyllama/cbt-tinyllama-merged/tokenizer_config.json

@@ -0,0 +1,43 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "add_prefix_space": null,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "legacy": false,
+  "model_max_length": 2048,
+  "pad_token": "</s>",
+  "padding_side": "right",
+  "sp_model_kwargs": {},
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}

chatbot.py

@@ -0,0 +1,116 @@
+import streamlit as st
+from transformers import AutoTokenizer, AutoModelForCausalLM
+import torch
+
+# --- Streamlit page config (must be first) ---
+st.set_page_config(page_title="TinyLLaMA Chatbot", layout="centered")
+
+# Device: CPU only
+device = torch.device("cpu")
+
+# --- Load the model and tokenizer ---
+@st.cache_resource
+def load_model():
+    model_path = r"C:\Users\HP\OneDrive\Desktop\HealthBot\cbt-tinyllama-merged\cbt-tinyllama-merged"  # adjust if needed
+    tokenizer = AutoTokenizer.from_pretrained(model_path)
+    
+    # Set pad token if missing
+    if tokenizer.pad_token is None:
+        tokenizer.pad_token = tokenizer.eos_token
+
+    model = AutoModelForCausalLM.from_pretrained(model_path)
+    model.to(device)
+    model.eval()
+    return tokenizer, model
+
+tokenizer, model = load_model()
+
+# --- Custom styling for chat bubbles ---
+st.markdown("""
+    <style>
+    .user-bubble {
+        background-color: #DCF8C6;
+        padding: 10px;
+        border-radius: 20px;
+        margin-bottom: 10px;
+        width: fit-content;
+        max-width: 80%;
+        align-self: flex-end;
+    }
+    .bot-bubble {
+        background-color: #F1F0F0;
+        padding: 10px;
+        border-radius: 20px;
+        margin-bottom: 10px;
+        width: fit-content;
+        max-width: 80%;
+        align-self: flex-start;
+    }
+    .chat-container {
+        display: flex;
+        flex-direction: column;
+    }
+    </style>
+""", unsafe_allow_html=True)
+
+# --- Title ---
+st.title("🤖 TinyLLaMA Chatbot")
+st.markdown("A conversational assistant powered by your fine-tuned TinyLLaMA model.")
+
+# --- Initialize chat history ---
+if "messages" not in st.session_state:
+    st.session_state.messages = []
+
+# --- User input ---
+user_input = st.chat_input("Type your message...")
+
+# --- Generate response function ---
+def generate_response(prompt):
+    input_ids = tokenizer.encode(prompt, return_tensors="pt").to(device)
+    attention_mask = (input_ids != tokenizer.pad_token_id).long().to(device)
+
+    # Trim input to max length
+    max_length = model.config.max_position_embeddings
+    if input_ids.size(1) > max_length:
+        input_ids = input_ids[:, -max_length:]
+        attention_mask = attention_mask[:, -max_length:]
+
+    with torch.no_grad():
+        output_ids = model.generate(
+            input_ids,
+            attention_mask=attention_mask,
+            max_new_tokens=100,
+            do_sample=True,
+            top_k=50,
+            top_p=0.95,
+            temperature=0.8,
+            pad_token_id=tokenizer.eos_token_id
+        )
+
+    decoded = tokenizer.decode(output_ids[0], skip_special_tokens=True)
+    # Remove prompt from output and stop at next user prompt if exists
+    response = decoded[len(prompt):].split("User:")[0].strip()
+    return response
+
+# --- Process user input ---
+if user_input:
+    st.session_state.messages.append({"role": "user", "content": user_input})
+
+    # Build full prompt from history
+    prompt = ""
+    for msg in st.session_state.messages:
+        role = "User" if msg["role"] == "user" else "Assistant"
+        prompt += f"{role}: {msg['content']}\n"
+    prompt += "Assistant:"
+
+    bot_reply = generate_response(prompt)
+    st.session_state.messages.append({"role": "assistant", "content": bot_reply})
+
+# --- Display chat ---
+for msg in st.session_state.messages:
+    if msg["role"] == "user":
+        st.markdown(f'<div class="chat-container"><div class="user-bubble"><b>You:</b><br>{msg["content"]}</div></div>', unsafe_allow_html=True)
+    else:
+        st.markdown(f'<div class="chat-container"><div class="bot-bubble"><b>Bot:</b><br>{msg["content"]}</div></div>', unsafe_allow_html=True)
+
+