Updated model

app.py

@@ -1,44 +1,50 @@
+import os
+import urllib.request
 import gradio as gr
-from transformers import AutoTokenizer, AutoModelForCausalLM
-import torch
-
-model = AutoModelForCausalLM.from_pretrained(
-    "bajrangCoder/BhagavadGita",
-    torch_dtype=torch.bfloat16,
-    trust_remote_code=True,
-    device_map="auto",
-    low_cpu_mem_usage=True,
-)
-tokenizer = AutoTokenizer.from_pretrained("bajrangCoder/BhagavadGita")
+from llama_cpp import Llama
 
 
-def generate_text(input_text):
-    input_ids = tokenizer.encode(input_text, return_tensors="pt")
-    attention_mask = torch.ones(input_ids.shape)
+def download_file(file_link, filename):
+    # Checks if the file already exists before downloading
+    if not os.path.isfile(filename):
+        urllib.request.urlretrieve(file_link, filename)
+        print("File downloaded successfully.")
+    else:
+        print("File already exists.")
+
+
+# Dowloading GGML model from HuggingFace
+ggml_model_path = "https://huggingface.co/bajrangCoder/BhagavadGita/resolve/main/bhagvat_gita-unsloth.Q4_K_M.gguf"
+filename = "bhagvat_gita-unsloth.Q4_K_M.gguf"
+
+download_file(ggml_model_path, filename)
 
-    output = model.generate(
-        input_ids,
-        attention_mask=attention_mask,
-        max_length=200,
-        do_sample=True,
-        top_k=10,
-        num_return_sequences=1,
-        eos_token_id=tokenizer.eos_token_id,
-    )
 
-    output_text = tokenizer.decode(output[0], skip_special_tokens=True)
-    print(output_text)
+llm = Llama(model_path=filename, n_ctx=512, n_batch=126)
+
+
+def generate_text(prompt="how to face failurs in life"):
+    output = llm(
+        prompt,
+        max_tokens=256,
+        temperature=0.1,
+        top_p=0.5,
+        echo=False,
+        stop=["#"],
+    )
+    output_text = output["choices"][0]["text"].strip()
 
     # Remove Prompt Echo from Generated Text
-    cleaned_output_text = output_text.replace(input_text, "")
+    cleaned_output_text = output_text.replace(prompt, "")
     return cleaned_output_text
 
 
-text_generation_interface = gr.Interface(
+description = "BhagavadGita"
+
+gradio_interface = gr.Interface(
     fn=generate_text,
-    inputs=[
-        gr.inputs.Textbox(label="Input Text"),
-    ],
-    outputs=gr.inputs.Textbox(label="Generated Text"),
-    title="BhagavadGita Instruct",
-).launch()
+    inputs="text",
+    outputs="text",
+    title="BhagavadGita",
+)
+gradio_interface.launch()

requirements.txt

@@ -1,6 +1 @@
-huggingface_hub==0.22.2
-datasets
-transformers
-accelerate
-einops
-safetensors
+llama-cpp-python