Spaces:
Sleeping
Sleeping
import pytorch_lightning as pl | |
import torch | |
import torch.nn.functional as F | |
from torch.optim import AdamW | |
from torch.optim.lr_scheduler import OneCycleLR | |
from transformers import AutoTokenizer | |
import torch.nn as nn | |
import math | |
from torch.utils.data import DataLoader, Dataset | |
from datasets import load_dataset | |
import os | |
def _init_weights(module, std=0.02): | |
if isinstance(module, nn.Linear): | |
module.weight.data.normal_(mean=0.0, std=std) | |
elif isinstance(module, nn.Embedding): | |
module.weight.data.normal_(mean=0.0, std=std) | |
class RMSNorm(nn.Module): | |
def __init__(self, dim, eps=1e-5): | |
super().__init__() | |
self.eps = float(eps) # Ensure eps is a float | |
self.weight = nn.Parameter(torch.ones(dim)) | |
def forward(self, x): | |
norm = torch.rsqrt(x.pow(2).mean(-1, keepdim=True) + self.eps) | |
return x * norm * self.weight | |
class RotaryEmbedding(nn.Module): | |
def __init__(self, dim, max_position_embeddings=2048, base=10000): | |
super().__init__() | |
self.dim = dim | |
self.max_position_embeddings = int(max_position_embeddings) # Convert to int | |
self.base = base | |
inv_freq = 1.0 / (base ** (torch.arange(0, dim, 2).float() / dim)) | |
self.register_buffer("inv_freq", inv_freq) | |
t = torch.arange(self.max_position_embeddings).type_as(self.inv_freq) | |
freqs = torch.einsum("i,j->ij", t, self.inv_freq) | |
emb = torch.cat((freqs, freqs), dim=-1) | |
self.register_buffer("cos_cached", emb.cos()[None, None, :, :]) | |
self.register_buffer("sin_cached", emb.sin()[None, None, :, :]) | |
def forward(self, x, seq_len=None): | |
# Convert seq_len to int and ensure it's a valid value | |
seq_len = int(seq_len) if seq_len is not None else x.size(1) | |
if seq_len > self.max_position_embeddings: | |
seq_len = self.max_position_embeddings | |
return ( | |
self.cos_cached[:,:,:seq_len,:], | |
self.sin_cached[:,:,:seq_len,:] | |
) | |
def rotate_half(x): | |
"""Rotates half the hidden dims of the input.""" | |
x1, x2 = x.chunk(2, dim=-1) | |
return torch.cat((-x2, x1), dim=-1) | |
def apply_rotary_pos_emb(q, k, cos, sin): | |
# Ensure proper broadcasting | |
cos = cos[:, :, :q.size(2), :] # [batch, 1, seq_len, dim] | |
sin = sin[:, :, :q.size(2), :] # [batch, 1, seq_len, dim] | |
q_embed = (q * cos) + (rotate_half(q) * sin) | |
k_embed = (k * cos) + (rotate_half(k) * sin) | |
return q_embed, k_embed | |
class Attention(nn.Module): | |
def __init__(self, config): | |
super().__init__() | |
self.hidden_size = config.hidden_size | |
self.num_attention_heads = config.num_attention_heads | |
self.num_key_value_heads = config.num_key_value_heads | |
self.head_dim = self.hidden_size // self.num_attention_heads | |
self.q_proj = nn.Linear(self.hidden_size, self.hidden_size, bias=False) | |
self.k_proj = nn.Linear(self.hidden_size, self.num_key_value_heads * self.head_dim, bias=False) | |
self.v_proj = nn.Linear(self.hidden_size, self.num_key_value_heads * self.head_dim, bias=False) | |
self.o_proj = nn.Linear(self.hidden_size, self.hidden_size, bias=False) | |
def forward(self, hidden_states, cos, sin, attention_mask=None): | |
batch_size, seq_length, _ = hidden_states.shape | |
q = self.q_proj(hidden_states) | |
k = self.k_proj(hidden_states) | |
v = self.v_proj(hidden_states) | |
# Reshape for attention computation | |
q = q.view(batch_size, seq_length, self.num_attention_heads, self.head_dim) | |
k = k.view(batch_size, seq_length, self.num_key_value_heads, self.head_dim) | |
v = v.view(batch_size, seq_length, self.num_key_value_heads, self.head_dim) | |
# Transpose for attention computation | |
q = q.transpose(1, 2) # [batch, num_heads, seq_len, head_dim] | |
k = k.transpose(1, 2) # [batch, num_kv_heads, seq_len, head_dim] | |
v = v.transpose(1, 2) # [batch, num_kv_heads, seq_len, head_dim] | |
# Apply rotary embeddings | |
q, k = apply_rotary_pos_emb(q, k, cos, sin) | |
# Repeat k/v heads if num_key_value_heads < num_attention_heads | |
if self.num_key_value_heads != self.num_attention_heads: | |
k = k.repeat_interleave(self.num_attention_heads // self.num_key_value_heads, dim=1) | |
v = v.repeat_interleave(self.num_attention_heads // self.num_key_value_heads, dim=1) | |
# Compute attention | |
attn_weights = torch.matmul(q, k.transpose(-2, -1)) / math.sqrt(self.head_dim) | |
if attention_mask is not None: | |
attn_weights = attn_weights + attention_mask | |
attn_weights = F.softmax(attn_weights, dim=-1) | |
# Compute output | |
output = torch.matmul(attn_weights, v) | |
output = output.transpose(1, 2).contiguous() # [batch, seq_len, num_heads, head_dim] | |
output = output.view(batch_size, seq_length, -1) | |
return self.o_proj(output) | |
class MLP(nn.Module): | |
def __init__(self, config): | |
super().__init__() | |
self.gate_proj = nn.Linear(config.hidden_size, config.intermediate_size, bias=False) | |
self.up_proj = nn.Linear(config.hidden_size, config.intermediate_size, bias=False) | |
self.down_proj = nn.Linear(config.intermediate_size, config.hidden_size, bias=False) | |
self.act_fn = nn.SiLU() | |
def forward(self, x): | |
return self.down_proj(self.act_fn(self.gate_proj(x)) * self.up_proj(x)) | |
class DecoderLayer(nn.Module): | |
def __init__(self, config): | |
super().__init__() | |
self.self_attn = Attention(config) | |
self.mlp = MLP(config) | |
self.input_layernorm = RMSNorm(config.hidden_size, eps=config.rms_norm_eps) | |
self.post_attention_layernorm = RMSNorm(config.hidden_size, eps=config.rms_norm_eps) | |
def forward(self, hidden_states, cos, sin, attention_mask=None): | |
# Self attention | |
residual = hidden_states | |
hidden_states = self.input_layernorm(hidden_states) | |
hidden_states = self.self_attn(hidden_states, cos, sin, attention_mask) | |
hidden_states = residual + hidden_states | |
# MLP | |
residual = hidden_states | |
hidden_states = self.post_attention_layernorm(hidden_states) | |
hidden_states = self.mlp(hidden_states) | |
hidden_states = residual + hidden_states | |
return hidden_states | |
class SmolLM2(nn.Module): | |
def __init__(self, config): | |
super().__init__() | |
self.config = config | |
# Token embeddings | |
self.embed_tokens = nn.Embedding(config.vocab_size, config.hidden_size) | |
# Initialize transformer layers | |
self.layers = nn.ModuleList([ | |
DecoderLayer(config) for _ in range(config.num_hidden_layers) | |
]) | |
# Final layer norm | |
self.norm = RMSNorm(config.hidden_size, eps=float(config.rms_norm_eps)) | |
# Initialize rotary embeddings | |
self.rotary_emb = RotaryEmbedding( | |
config.hidden_size // config.num_attention_heads, | |
max_position_embeddings=config.max_position_embeddings | |
) | |
# Initialize weights | |
self.apply(lambda p: _init_weights(p, std=config.initializer_range)) | |
def forward(self, input_ids, attention_mask=None): | |
try: | |
# Ensure inputs are on the correct device | |
device = input_ids.device | |
batch_size, seq_length = input_ids.shape | |
# Input validation | |
if seq_length > self.config.max_position_embeddings: | |
raise ValueError(f"Input sequence length {seq_length} exceeds maximum position embeddings {self.config.max_position_embeddings}") | |
# Get embeddings | |
hidden_states = self.embed_tokens(input_ids) | |
# Get position embeddings | |
cos, sin = self.rotary_emb(hidden_states, seq_length) | |
# Generate attention mask if none provided | |
if attention_mask is None: | |
attention_mask = torch.ones( | |
(batch_size, seq_length), | |
dtype=torch.bool, | |
device=device | |
) | |
else: | |
# Convert to boolean if it's not already and ensure contiguous memory | |
attention_mask = attention_mask.bool().contiguous() | |
# Create causal mask | |
causal_mask = torch.triu( | |
torch.ones((seq_length, seq_length), device=device), | |
diagonal=1 | |
).bool() | |
# Create attention mask [batch_size, 1, seq_length, seq_length] | |
attention_mask = attention_mask.view(batch_size, 1, 1, seq_length) | |
attention_mask = attention_mask.expand(batch_size, 1, seq_length, seq_length) | |
# Prepare causal mask | |
causal_mask = causal_mask.view(1, 1, seq_length, seq_length) | |
# Combine masks | |
mask = attention_mask & ~causal_mask | |
# Convert boolean mask to float mask | |
mask = mask.to(dtype=hidden_states.dtype) | |
mask = (1.0 - mask) * torch.finfo(hidden_states.dtype).min | |
# Apply transformer layers | |
for layer in self.layers: | |
hidden_states = layer(hidden_states, cos, sin, mask) | |
# Apply final normalization | |
hidden_states = self.norm(hidden_states) | |
# Project back to vocabulary | |
logits = F.linear(hidden_states, self.embed_tokens.weight) | |
return logits | |
except Exception as e: | |
print(f"\nForward pass error:") | |
print(f"Input shape: {input_ids.shape}") | |
print(f"Device: {input_ids.device}") | |
print(f"CUDA memory allocated: {torch.cuda.memory_allocated() / 1e9:.2f} GB") | |
print(f"Error: {str(e)}") | |
raise | |
def generate( | |
self, | |
input_ids, | |
attention_mask=None, | |
max_length=100, | |
temperature=0.7, | |
top_p=0.9, | |
top_k=50, | |
num_return_sequences=1, | |
do_sample=True, | |
pad_token_id=None, | |
bos_token_id=None, | |
eos_token_id=None | |
): | |
try: | |
batch_size = input_ids.shape[0] | |
current_length = input_ids.shape[1] | |
device = input_ids.device | |
# Input validation | |
if current_length >= self.config.max_position_embeddings: | |
raise ValueError(f"Input sequence length {current_length} exceeds maximum position embeddings {self.config.max_position_embeddings}") | |
# Ensure we don't exceed maximum position embeddings | |
max_length = min(max_length, self.config.max_position_embeddings) | |
# Initialize attention mask if None | |
if attention_mask is None: | |
attention_mask = torch.ones_like(input_ids, dtype=torch.bool, device=device) | |
for _ in range(max_length - current_length): | |
# Forward pass | |
outputs = self(input_ids, attention_mask) | |
next_token_logits = outputs[:, -1, :] / temperature | |
# Apply top-k filtering | |
if top_k > 0: | |
indices_to_remove = next_token_logits < torch.topk(next_token_logits, top_k)[0][..., -1, None] | |
next_token_logits[indices_to_remove] = float('-inf') | |
# Apply top-p filtering | |
if top_p < 1.0: | |
sorted_logits, sorted_indices = torch.sort(next_token_logits, descending=True) | |
cumulative_probs = torch.cumsum(F.softmax(sorted_logits, dim=-1), dim=-1) | |
sorted_indices_to_remove = cumulative_probs > top_p | |
sorted_indices_to_remove[..., 1:] = sorted_indices_to_remove[..., :-1].clone() | |
sorted_indices_to_remove[..., 0] = 0 | |
indices_to_remove = sorted_indices_to_remove.scatter(1, sorted_indices, sorted_indices_to_remove) | |
next_token_logits[indices_to_remove] = float('-inf') | |
# Sample from the filtered distribution | |
if do_sample: | |
probs = F.softmax(next_token_logits, dim=-1) | |
next_tokens = torch.multinomial(probs, num_samples=1).squeeze(1) | |
else: | |
next_tokens = torch.argmax(next_token_logits, dim=-1) | |
# Append new tokens | |
input_ids = torch.cat([input_ids, next_tokens.unsqueeze(-1)], dim=-1) | |
attention_mask = torch.cat([attention_mask, torch.ones_like(next_tokens.unsqueeze(-1))], dim=-1) | |
# Stop if we've hit special tokens | |
if (pad_token_id is not None and (next_tokens == pad_token_id).all()) or \ | |
(eos_token_id is not None and (next_tokens == eos_token_id).all()): | |
break | |
return input_ids | |
except Exception as e: | |
print(f"\nGeneration error:") | |
print(f"Input shape: {input_ids.shape}") | |
print(f"Device: {input_ids.device}") | |
print(f"Error: {str(e)}") | |
raise | |
class TextDataset(Dataset): | |
def __init__(self, config, split="train"): | |
self.config = config | |
# Load dataset from HuggingFace | |
full_dataset = load_dataset( | |
config.data.datasets[0].path, | |
config.data.datasets[0].subset, | |
split=split | |
) | |
# Apply split ratio if less than 1 | |
if config.data.datasets[0].split_ratio < 1.0: | |
num_samples = int(len(full_dataset) * config.data.datasets[0].split_ratio) | |
self.dataset = full_dataset.select(range(num_samples)) | |
else: | |
self.dataset = full_dataset | |
# Initialize tokenizer | |
self.tokenizer = AutoTokenizer.from_pretrained(config.model.tokenizer_name) | |
if self.tokenizer.pad_token is None: | |
self.tokenizer.pad_token = self.tokenizer.eos_token | |
def __len__(self): | |
return len(self.dataset) | |
def __getitem__(self, idx): | |
# Get text from dataset | |
text = self.dataset[idx]["text"] | |
# Tokenize | |
encodings = self.tokenizer( | |
text, | |
truncation=True, | |
max_length=self.config.model.max_position_embeddings, | |
padding="max_length", | |
return_tensors="pt" | |
) | |
return { | |
"input_ids": encodings.input_ids.squeeze(), | |
"attention_mask": encodings.attention_mask.squeeze(), | |
"labels": encodings.input_ids.squeeze() | |
} | |
class SmolLM2Lightning(pl.LightningModule): | |
def __init__(self, config): | |
super().__init__() | |
self.save_hyperparameters() | |
self.config = config | |
# Initialize tokenizer | |
self.tokenizer = AutoTokenizer.from_pretrained(config.model.tokenizer_name) | |
if self.tokenizer.pad_token is None: | |
self.tokenizer.pad_token = self.tokenizer.eos_token | |
# Initialize the base model | |
self.model = SmolLM2(config.model) | |
def forward(self, input_ids, attention_mask=None): | |
return self.model(input_ids, attention_mask) | |
def training_step(self, batch, batch_idx): | |
try: | |
input_ids = batch["input_ids"] | |
labels = batch["labels"] | |
attention_mask = batch.get("attention_mask", None) | |
# Ensure tensors are contiguous and on the correct device | |
inputs = input_ids[..., :-1].contiguous() | |
labels = input_ids[..., 1:].contiguous() | |
if attention_mask is not None: | |
attention_mask = attention_mask[..., :-1].contiguous() | |
# Forward pass | |
logits = self(inputs, attention_mask) | |
# Calculate loss | |
loss = F.cross_entropy( | |
logits.view(-1, self.config.model.vocab_size), | |
labels.view(-1), | |
ignore_index=self.config.model.pad_token_id if self.config.model.pad_token_id is not None else -100, | |
reduction='mean' | |
) | |
# Detach loss for logging | |
loss_value = loss.detach().float() | |
# Log metrics | |
self.log('train_loss', loss_value, prog_bar=True, on_step=True, sync_dist=True) | |
return loss | |
except Exception as e: | |
print(f"\nTraining step error:") | |
print(f"Input shape: {input_ids.shape if input_ids is not None else 'None'}") | |
print(f"Device: {input_ids.device if input_ids is not None else 'None'}") | |
print(f"Error: {str(e)}") | |
raise | |
def validation_step(self, batch, batch_idx): | |
try: | |
input_ids = batch["input_ids"] | |
labels = batch["labels"] | |
attention_mask = batch.get("attention_mask", None) | |
# Ensure tensors are contiguous and on the correct device | |
inputs = input_ids[..., :-1].contiguous() | |
labels = input_ids[..., 1:].contiguous() | |
if attention_mask is not None: | |
attention_mask = attention_mask[..., :-1].contiguous() | |
# Forward pass | |
logits = self(inputs, attention_mask) | |
# Calculate loss | |
loss = F.cross_entropy( | |
logits.view(-1, self.config.model.vocab_size), | |
labels.view(-1), | |
ignore_index=self.config.model.pad_token_id if self.config.model.pad_token_id is not None else -100, | |
reduction='mean' | |
) | |
# Detach loss for logging | |
loss_value = loss.detach().float() | |
# Log metrics | |
self.log('val_loss', loss_value, prog_bar=True, on_epoch=True, sync_dist=True) | |
return loss | |
except Exception as e: | |
print(f"\nValidation step error:") | |
print(f"Input shape: {input_ids.shape if input_ids is not None else 'None'}") | |
print(f"Device: {input_ids.device if input_ids is not None else 'None'}") | |
print(f"Error: {str(e)}") | |
raise | |
def configure_optimizers(self): | |
# Create optimizer with explicit type conversion | |
optimizer = AdamW( | |
self.parameters(), | |
lr=float(self.config.scheduler.learning_rate), | |
weight_decay=float(self.config.optimizer.weight_decay), | |
betas=(float(self.config.optimizer.adam_beta1), | |
float(self.config.optimizer.adam_beta2)), | |
eps=float(self.config.optimizer.adam_eps), | |
) | |
# Create scheduler | |
scheduler = OneCycleLR( | |
optimizer, | |
max_lr=float(self.config.scheduler.max_lr), | |
total_steps=int(self.config.training.max_steps), | |
pct_start=float(self.config.scheduler.pct_start), | |
anneal_strategy=self.config.scheduler.anneal_strategy, | |
cycle_momentum=bool(self.config.scheduler.cycle_momentum), | |
div_factor=float(self.config.scheduler.div_factor), | |
final_div_factor=float(self.config.scheduler.final_div_factor), | |
) | |
return { | |
"optimizer": optimizer, | |
"lr_scheduler": { | |
"scheduler": scheduler, | |
"interval": "step", | |
"frequency": 1 | |
} | |
} | |
def generate(self, *args, **kwargs): | |
return self.model.generate(*args, **kwargs) | |
def train_dataloader(self): | |
dataset = TextDataset(self.config, split="train") | |
return DataLoader( | |
dataset, | |
batch_size=self.config.training.batch_size, | |
shuffle=True, | |
num_workers=self.config.data.loading.num_workers, | |
pin_memory=self.config.data.loading.pin_memory, | |
persistent_workers=True, | |
prefetch_factor=self.config.data.loading.prefetch_factor, | |
drop_last=True # Drop incomplete batches | |
) | |
def val_dataloader(self): | |
dataset = TextDataset(self.config, split="validation") | |
return DataLoader( | |
dataset, | |
batch_size=self.config.training.batch_size, | |
shuffle=False, | |
num_workers=self.config.data.loading.num_workers, | |
pin_memory=self.config.data.loading.pin_memory, | |
persistent_workers=True, | |
prefetch_factor=self.config.data.loading.prefetch_factor | |
) |