app.py

# app.py (second app by claude)
import gradio as gr
import torch
from transformers import PreTrainedTokenizerFast
from pathlib import Path
import sys

# sys.path.append(str(Path('gpt_model_code.py').resolve()))
# from gpt_model_code import load_model_n_tokenizer, generate

# ==================================================================================-
# ==================================================================================-
# ==================================================================================-
# ==================================================================================-
# ==================================================================================-
# could not make importing from gpt_model_code.py work, so i copied the code here

# Copyright (c) Sebastian Raschka under Apache License 2.0 (see LICENSE.txt).
# Source for "Build a Large Language Model From Scratch"
#   - https://www.manning.com/books/build-a-large-language-model-from-scratch
# Code: https://github.com/rasbt/LLMs-from-scratch
#
# This file collects all the relevant code that we covered thus far
# throughout Chapters 2-5.


import torch
import torch.nn as nn

from huggingface_hub import PyTorchModelHubMixin
from transformers import PreTrainedTokenizerFast

GPT_CONFIG_124M = {
    "vocab_size": 50000,     # Vocabulary size
    "context_length": 1024,  # Context length
    "emb_dim": 768,          # Embedding dimension
    "n_heads": 12,           # Number of attention heads
    "n_layers": 12,          # Number of layers
    "drop_rate": 0.1,        # Dropout rate
    "qkv_bias": False        # Query-key-value bias
}

#####################################
# Chapter 3
#####################################
class MultiHeadAttention(nn.Module):
    def __init__(self, d_in, d_out, context_length, dropout, num_heads, qkv_bias=False):
        super().__init__()
        assert d_out % num_heads == 0, "d_out must be divisible by n_heads"

        self.d_out = d_out
        self.num_heads = num_heads
        self.head_dim = d_out // num_heads  # Reduce the projection dim to match desired output dim

        self.W_query = nn.Linear(d_in, d_out, bias=qkv_bias)
        self.W_key = nn.Linear(d_in, d_out, bias=qkv_bias)
        self.W_value = nn.Linear(d_in, d_out, bias=qkv_bias)
        self.out_proj = nn.Linear(d_out, d_out)  # Linear layer to combine head outputs
        self.dropout = nn.Dropout(dropout)
        self.register_buffer('mask', torch.triu(torch.ones(context_length, context_length), diagonal=1))

    def forward(self, x):
        b, num_tokens, d_in = x.shape

        keys = self.W_key(x)  # Shape: (b, num_tokens, d_out)
        queries = self.W_query(x)
        values = self.W_value(x)

        # We implicitly split the matrix by adding a `num_heads` dimension
        # Unroll last dim: (b, num_tokens, d_out) -> (b, num_tokens, num_heads, head_dim)
        keys = keys.view(b, num_tokens, self.num_heads, self.head_dim)
        values = values.view(b, num_tokens, self.num_heads, self.head_dim)
        queries = queries.view(b, num_tokens, self.num_heads, self.head_dim)

        # Transpose: (b, num_tokens, num_heads, head_dim) -> (b, num_heads, num_tokens, head_dim)
        keys = keys.transpose(1, 2)
        queries = queries.transpose(1, 2)
        values = values.transpose(1, 2)

        # Compute scaled dot-product attention (aka self-attention) with a causal mask
        attn_scores = queries @ keys.transpose(2, 3)  # Dot product for each head

        # Original mask truncated to the number of tokens and converted to boolean
        mask_bool = self.mask.bool()[:num_tokens, :num_tokens]

        # Use the mask to fill attention scores
        attn_scores.masked_fill_(mask_bool, -torch.inf)

        attn_weights = torch.softmax(attn_scores / keys.shape[-1]**0.5, dim=-1)
        attn_weights = self.dropout(attn_weights)

        # Shape: (b, num_tokens, num_heads, head_dim)
        context_vec = (attn_weights @ values).transpose(1, 2)

        # Combine heads, where self.d_out = self.num_heads * self.head_dim
        context_vec = context_vec.reshape(b, num_tokens, self.d_out)
        context_vec = self.out_proj(context_vec)  # optional projection

        return context_vec


#####################################
# Chapter 4
#####################################
class LayerNorm(nn.Module):
    def __init__(self, emb_dim):
        super().__init__()
        self.eps = 1e-5
        self.scale = nn.Parameter(torch.ones(emb_dim))
        self.shift = nn.Parameter(torch.zeros(emb_dim))

    def forward(self, x):
        mean = x.mean(dim=-1, keepdim=True)
        var = x.var(dim=-1, keepdim=True, unbiased=False)
        norm_x = (x - mean) / torch.sqrt(var + self.eps)
        return self.scale * norm_x + self.shift


class GELU(nn.Module):
    def __init__(self):
        super().__init__()

    def forward(self, x):
        return 0.5 * x * (1 + torch.tanh(
            torch.sqrt(torch.tensor(2.0 / torch.pi)) *
            (x + 0.044715 * torch.pow(x, 3))
        ))


class FeedForward(nn.Module):
    def __init__(self, cfg):
        super().__init__()
        self.layers = nn.Sequential(
            nn.Linear(cfg["emb_dim"], 4 * cfg["emb_dim"]),
            GELU(),
            nn.Linear(4 * cfg["emb_dim"], cfg["emb_dim"]),
        )

    def forward(self, x):
        return self.layers(x)


class TransformerBlock(nn.Module):
    def __init__(self, cfg):
        super().__init__()
        self.att = MultiHeadAttention(
            d_in=cfg["emb_dim"],
            d_out=cfg["emb_dim"],
            context_length=cfg["context_length"],
            num_heads=cfg["n_heads"],
            dropout=cfg["drop_rate"],
            qkv_bias=cfg["qkv_bias"])
        self.ff = FeedForward(cfg)
        self.norm1 = LayerNorm(cfg["emb_dim"])
        self.norm2 = LayerNorm(cfg["emb_dim"])
        self.drop_shortcut = nn.Dropout(cfg["drop_rate"])

    def forward(self, x):
        # Shortcut connection for attention block
        shortcut = x
        x = self.norm1(x)
        x = self.att(x)   # Shape [batch_size, num_tokens, emb_size]
        x = self.drop_shortcut(x)
        x = x + shortcut  # Add the original input back

        # Shortcut connection for feed-forward block
        shortcut = x
        x = self.norm2(x)
        x = self.ff(x)
        x = self.drop_shortcut(x)
        x = x + shortcut  # Add the original input back

        return x


class GPTModel(nn.Module,
        PyTorchModelHubMixin, # modified to push the model to the hub (https://huggingface.co/docs/hub/en/models-uploading#upload-a-pytorch-model-using-huggingfacehub)
        repo_url="https://huggingface.co/Aananda-giri/GPT2-Nepali/",
        pipeline_tag="text-generation",
        ):
    def __init__(self, cfg):
        super().__init__()
        self.tok_emb = nn.Embedding(cfg["vocab_size"], cfg["emb_dim"])
        self.pos_emb = nn.Embedding(cfg["context_length"], cfg["emb_dim"])
        self.drop_emb = nn.Dropout(cfg["drop_rate"])

        self.trf_blocks = nn.Sequential(
            *[TransformerBlock(cfg) for _ in range(cfg["n_layers"])])

        self.final_norm = LayerNorm(cfg["emb_dim"])
        self.out_head = nn.Linear(cfg["emb_dim"], cfg["vocab_size"], bias=False)

    def forward(self, in_idx):
        batch_size, seq_len = in_idx.shape
        tok_embeds = self.tok_emb(in_idx)
        pos_embeds = self.pos_emb(torch.arange(seq_len, device=in_idx.device))
        x = tok_embeds + pos_embeds  # Shape [batch_size, num_tokens, emb_size]
        x = self.drop_emb(x)
        x = self.trf_blocks(x)
        x = self.final_norm(x)
        logits = self.out_head(x)
        return logits


#####################################
# Chapter 5
#####################################
def text_to_token_ids(text, tokenizer):
    encoded = tokenizer.encode(text)
    encoded_tensor = torch.tensor(encoded).unsqueeze(0)  # add batch dimension
    return encoded_tensor

def token_ids_to_text(token_ids, tokenizer):
    flat = token_ids.squeeze(0)  # remove batch dimension
    return tokenizer.decode(flat.tolist())

def load_model_n_tokenizer():
    model = GPTModel.from_pretrained("Aananda-giri/GPT2-Nepali")

    device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
    model.to(device)
    
    tokenizer = PreTrainedTokenizerFast.from_pretrained("Aananda-giri/GPT2-Nepali")
    return model, tokenizer

def generate(
    model,
    prompt,
    tokenizer,
    max_new_tokens,
    temperature=0.7,
    top_k=50,
    top_p=None,  # New parameter for nucleus sampling
    eos_id=None,
    repetition_penalty=1.2,
    penalize_len_below=50
):
    context_size = GPT_CONFIG_124M['context_length']
    device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
    
    idx = text_to_token_ids(prompt, tokenizer).to(device)
    
    if not eos_id:
        encoded_endoftext = tokenizer.encode("<|endoftext|>")
        eos_id = encoded_endoftext[0] if encoded_endoftext else None

    token_freq = {}

    for step in range(max_new_tokens):
        idx_cond = idx[:, -context_size:]
        with torch.no_grad():
            logits = model(idx_cond)
        logits = logits[:, -1, :]

        # Apply repetition penalty
        for token_id in idx[0].tolist():
            if token_id in token_freq:
                logits[0, token_id] /= repetition_penalty
            else:
                token_freq[token_id] = 1
        
        # Penalize EOT token for shorter sequences
        if eos_id is not None and step < penalize_len_below:
            logits[0, eos_id] /= (penalize_len_below - step) / penalize_len_below

        # Apply temperature scaling
        if temperature > 0.0:
            logits = logits / temperature

        # Convert logits to probabilities
        probs = torch.softmax(logits, dim=-1)

        # Apply top-p (nucleus) sampling if specified
        if top_p:
            sorted_probs, sorted_indices = torch.sort(probs, descending=True)
            cumulative_probs = torch.cumsum(sorted_probs, dim=-1)
            
            # Remove tokens with cumulative probability above the threshold
            sorted_indices_to_remove = cumulative_probs > top_p
            # Shift the indices to the right to keep also the first token above the threshold
            sorted_indices_to_remove[..., 1:] = sorted_indices_to_remove[..., :-1].clone()
            sorted_indices_to_remove[..., 0] = 0

            # Create a mask for indices to remove
            indices_to_remove = sorted_indices_to_remove.scatter(dim=-1, index=sorted_indices, src=sorted_indices_to_remove)
            probs = probs.masked_fill(indices_to_remove, 0.0)
            
            # Renormalize probabilities
            probs = probs / probs.sum(dim=-1, keepdim=True)

        # If top_p is None, apply top-k sampling
        elif top_k:
            top_probs, top_indices = torch.topk(probs, top_k)
            probs = torch.zeros_like(probs).scatter_(-1, top_indices, top_probs)
            # Renormalize probabilities
            probs = probs / probs.sum(dim=-1, keepdim=True)

        # Sample from the filtered distribution
        if temperature > 0.0:
            idx_next = torch.multinomial(probs, num_samples=1)
        else:
            idx_next = torch.argmax(probs, dim=-1, keepdim=True)

        if idx_next == eos_id:
            break

        idx = torch.cat((idx, idx_next), dim=1)
        text = token_ids_to_text(idx, tokenizer)

    return text

# ==================================================================================-
# ==================================================================================-
# ==================================================================================-
# ==================================================================================-
# ==================================================================================-


# Load model and tokenizer once at startup
model, tokenizer = load_model_n_tokenizer()
model.eval()
import sys

def generate_text(prompt, max_new_tokens, top_k, top_p, temperature, repetition_penalty, penalize_len_below):
    device = next(model.parameters()).device
    
    # Convert top_k to None if using top_p
    if top_p > 0:
        top_k = None
    else:
        top_p = None
    
    with torch.no_grad():
        if top_k!=None:
            # it expects top_k to be integer not float
            top_k = int(top_k)
        output_text = generate(  # function uses `with torch.no_grad()` internally already
            model=model,
            prompt=prompt,
            tokenizer=tokenizer,
            max_new_tokens=max_new_tokens,
            top_p=top_p,# top p sampling is prefered over top k if top_p != None
            top_k=top_k,
            temperature=0.7,
            repetition_penalty=repetition_penalty,  # New parameter: Repetition penalty factor
            penalize_len_below=penalize_len_below  # New parameter: Minimum content length for penalizing EOT token.
        )
        
        return output_text

css = """
    #bright-textbox {
        background-color: #ffeb3b; /* Bright yellow */
        color: #000000; /* Black text for contrast */
        border: 2px solid #fbc02d; /* Slightly darker yellow for the border */
        font-size: 16px;
        padding: 10px;
        border-radius: 5px;
    }
"""

# Create Gradio interface
with gr.Blocks(title="Nepali GPT-2 Text Generator", css=css) as interface:
    gr.Markdown("# Nepali GPT-2 Text Generator")
    gr.Markdown("Enter Nepali (नेपाली) text to generate content using the custom GPT2-Nepali model.")
    
    with gr.Row():
        with gr.Column():
            prompt = gr.Textbox(
                label="Prompt",
                placeholder="यहाँ नेपाली मा इन्पुट दिनु होस् ... (please Enter Nepali text here...)" #,
                # value="रामले भात"
       )
            max_tokens = gr.Slider(minimum=1, maximum=512, value=50, step=1, label="Max New Tokens")
            
            with gr.Row():
                with gr.Column():
                    temperature = gr.Slider(minimum=0.1, maximum=2.0, value=0.7, step=0.1, label="Temperature")
                    repetition_penalty = gr.Slider(minimum=1.0, maximum=2.0, value=1.2, step=0.1, label="Repetition Penalty")
                with gr.Column():
                    top_k = gr.Slider(minimum=0, maximum=100, value=50, step=1, label="Top K (set to 0 to use Top P)")
                    top_p = gr.Slider(minimum=0, maximum=1.0, value=0.9, step=0.05, label="Top P (set above 0 to use instead of Top K)")
            
            min_length = gr.Slider(minimum=1, maximum=200, value=50, step=1, label="Minimum Length Penalty")
            generate_btn = gr.Button("Generate Text")
        
        with gr.Column():
            output = gr.Textbox(label="Generated Text", lines=10)
    
    # Add examples if you have any
    gr.Examples(
        examples=[
            ["रामले भात", 50, 50, 0, 0.7, 1.2, 50],
            ["नेपाल एउटा", 100, 0, 0.9, 0.8, 1.2, 100],
            ["नेपाल का वर्तमान प्रधानमन्त्री ", 100, 0, 0.9, 0.8, 1.2, 100],
            ["भारतीय  प्रधानमन्त्री  ", 100, 0, 0.9, 0.8, 1.2, 100],
            ["अमिरिकी रास्ट्रपति डोनाल्ड", 100, 0, 0.9, 0.8, 1.2, 100],
        ],
        inputs=[prompt, max_tokens, top_k, top_p, temperature, repetition_penalty, min_length],
        outputs=output,
        fn=generate_text,
        cache_examples=True,
    )
    
    generate_btn.click(
        fn=generate_text,
        inputs=[prompt, max_tokens, top_p, top_k, temperature, repetition_penalty, min_length],
        outputs=output
    )


'''

'''

interface.launch()