Upload 3 files

config.json

@@ -1,9 +1,20 @@
 {
-  "model_type": "PreTrainedTokenizerFast",
-  "model_type": "gpt2",
+  "architectures": ["BVVBestMoeForCausalLM"],
+  "auto_map": {
+    "AutoConfig": "modeling_bvv_best_moe.BVVBestConfig",
+    "AutoModel": "modeling_bvv_best_moe.BVVBestMoeForCausalLM",
+    "AutoModelForCausalLM": "modeling_bvv_best_moe.BVVBestMoeForCausalLM"
+  },
+  "model_type": "bvv_best",
+  "vocab_size": 131072,
+  "block_size ": 1024,
+  "n_embd": 1024,
+  "n_layer": 16,
+  "n_head": 32,
+  "pad_id": 57344,
   "bos_token": "<s>",
   "eos_token": "</s>",
   "unk_token": "<unk>",
   "pad_token": "<pad>",
-  "vocab_size": 131072
+  "torch_dtype": "float32"
 }

modeling_bvv_best_moe.py

@@ -0,0 +1,332 @@
+from transformers import PreTrainedModel, PretrainedConfig
+from transformers.modeling_outputs import CausalLMOutputWithCrossAttentions
+import torch
+import torch.nn as nn
+from torch.nn import functional as F
+from transformers.modeling_outputs import CausalLMOutput
+
+class BVVBestConfig(PretrainedConfig):
+    model_type = "bvv_best"
+    
+    def __init__(
+        self,
+        vocab_size = 131072,
+        n_embd = 1024,
+        n_head = 32,  
+        n_layer = 16, 
+        block_size = 1024,
+        pad_id = 57344,
+        **kwargs
+    ):
+        super().__init__(**kwargs)
+        self.vocab_size = vocab_size
+        self.block_size  = block_size 
+        self.n_embd = n_embd
+        self.n_layer = n_layer
+        self.n_head = n_head
+        self.pad_id = pad_id 
+
+class RotaryEmbedding(nn.Module):
+    def __init__(self, dim):  # dim = head_dim (?? n_embd!)
+        super().__init__()
+        inv_freq = 1.0 / (10000 ** (torch.arange(0, dim, 2).float() / dim))
+        self.register_buffer("inv_freq", inv_freq, persistent=False)
+
+    def forward(self, seq_len, device):
+        t = torch.arange(seq_len, device=device, dtype=self.inv_freq.dtype)
+        freqs = torch.einsum('i,j->ij', t, self.inv_freq)
+        emb = torch.cat([freqs, freqs], dim=-1)    # (seq_len, dim)
+        return emb
+
+def apply_rotary_emb(x, rot_emb):
+    # x: (B, n_head, seq_len, head_dim)
+    # rot_emb: (seq_len, head_dim)
+    seq_len = x.shape[-2]
+    rot_emb = rot_emb[:seq_len]
+    
+    cos = torch.cos(rot_emb).unsqueeze(0).unsqueeze(0)  # (1, 1, seq_len, head_dim)
+    sin = torch.sin(rot_emb).unsqueeze(0).unsqueeze(0)
+    
+    x_shape = x.shape
+    x = x.reshape(*x_shape[:-1], -1, 2)  # (..., head_dim/2, 2)
+    x1 = x[..., 0]  
+    x2 = x[..., 1] 
+    
+    cos = cos.reshape(*cos.shape[:-1], -1, 2)[..., 0]
+    sin = sin.reshape(*sin.shape[:-1], -1, 2)[..., 0]
+    
+    x1_rot = x1 * cos - x2 * sin
+    x2_rot = x1 * sin + x2 * cos
+    
+    x_rot = torch.stack([x1_rot, x2_rot], dim=-1)
+    return x_rot.reshape(x_shape)
+
+class MultiHeadSelfAttention(nn.Module):
+    def __init__(self, n_embd, n_head, block_size):
+        super().__init__()
+        assert n_embd % n_head == 0
+        self.n_embd = n_embd
+        self.n_head = n_head
+        self.head_dim = n_embd // n_head
+
+        self.q_proj = nn.Linear(n_embd, n_embd, bias=False)
+        self.k_proj = nn.Linear(n_embd, n_embd, bias=False)
+        self.v_proj = nn.Linear(n_embd, n_embd, bias=False)
+        self.o_proj = nn.Linear(n_embd, n_embd, bias=False)
+
+        self.rotary_emb = RotaryEmbedding(self.head_dim)
+        self.dropout = nn.Dropout(0.0)
+
+        self.register_buffer(
+            "tril", torch.tril(torch.ones(block_size, block_size)), persistent=False
+        )
+
+    def forward(self, x):
+        # x: (B, T, n_embd)
+        B, T, C = x.shape
+
+        q = self.q_proj(x)    # (B, T, n_embd)
+        k = self.k_proj(x)
+        v = self.v_proj(x)
+
+        q = q.view(B, T, self.n_head, self.head_dim).transpose(1, 2)  # (B, n_head, T, head_dim)
+        k = k.view(B, T, self.n_head, self.head_dim).transpose(1, 2)
+        v = v.view(B, T, self.n_head, self.head_dim).transpose(1, 2)
+
+        # Rotary embeddings
+        rot_emb = self.rotary_emb(seq_len=T, device=x.device)   # (T, head_dim)
+        q = apply_rotary_emb(q, rot_emb)
+        k = apply_rotary_emb(k, rot_emb)
+
+        # Attention
+        attn_scores = torch.matmul(q, k.transpose(-2, -1)) * (self.head_dim ** -0.5)  # (B, n_head, T, T)
+        attn_scores = attn_scores.masked_fill(self.tril[:T, :T] == 0, float('-inf'))
+        attn_probs = F.softmax(attn_scores, dim=-1)
+        attn_probs = self.dropout(attn_probs)
+
+        out = torch.matmul(attn_probs, v)  # (B, n_head, T, head_dim)
+        out = out.transpose(1, 2).contiguous().view(B, T, C)  # (B, T, n_embd)
+
+        return self.o_proj(out)
+
+
+class TransformerMLP(nn.Module):
+    def __init__(self, n_embd):
+        super().__init__()
+        self.net = nn.Sequential(
+            nn.Linear(n_embd, 4 * n_embd),
+            nn.GELU(),
+            nn.Linear(4 * n_embd, n_embd),
+            nn.Dropout(0.0),
+        )
+
+    def forward(self, x):
+        return self.net(x)
+
+class TransformerBlock(nn.Module):
+    def __init__(self, n_embd, n_head, block_size):
+        super().__init__()
+        self.self_attn = MultiHeadSelfAttention(n_embd, n_head, block_size)
+        self.mlp = TransformerMLP(n_embd)
+        self.input_layernorm = nn.LayerNorm(n_embd)
+        self.post_attention_layernorm = nn.LayerNorm(n_embd)
+
+    def forward(self, x):
+        x = x + self.self_attn(self.input_layernorm(x))
+        x = x + self.mlp(self.post_attention_layernorm(x))
+        return x
+
+class BVVBestForCausalLM(PreTrainedModel):
+    config_class = BVVBestConfig
+    
+    def __init__(self, config):
+        super().__init__(config)
+        self.token_embeddings = nn.Embedding(config.vocab_size, config.n_embd)
+        
+        self.transformer_layers = nn.Sequential(*[
+            TransformerBlock(config.n_embd, n_head=config.n_head, block_size=config.block_size) for _ in range(config.n_layer)
+        ])
+        self.final_layernorm = nn.LayerNorm(config.n_embd)
+        self.lm_head = nn.Linear(config.n_embd, config.vocab_size)
+
+        self.apply(self._init_weights)
+
+    def _init_weights(self, module):
+        if isinstance(module, nn.Linear):
+            torch.nn.init.normal_(module.weight, mean=0.0, std=0.02)
+            if module.bias is not None:
+                torch.nn.init.zeros_(module.bias)
+        elif isinstance(module, nn.Embedding):
+            torch.nn.init.normal_(module.weight, mean=0.0, std=0.02)
+
+
+    def forward(self, idx, targets=None):
+        B, T = idx.shape
+
+        x = self.token_embeddings(idx)
+        
+        x = self.transformer_layers(x)
+        x = self.final_layernorm(x)
+        logits = self.lm_head(x)
+
+        loss = None
+        if targets is not None:
+            #logits_flat = logits.view(-1, logits.size(-1))
+            #targets_flat = targets.view(-1)
+            logits_flat = logits.reshape(-1, logits.size(-1))
+            targets_flat = targets.reshape(-1)
+            loss = F.cross_entropy(logits_flat, targets_flat, ignore_index = 57344)
+
+        return CausalLMOutput(
+            logits=logits,
+            loss=loss,
+        )
+
+    def generate(self, 
+                input_ids=None,
+                max_new_tokens=None,
+                max_length=None,
+                temperature=1.0,
+                top_k=None,
+                top_p=None,
+                do_sample=True,
+                pad_token_id=None,
+                eos_token_id=None,
+                **kwargs):
+        
+        if input_ids is None:
+            raise ValueError("Input_ids must be provided")
+        
+        idx = input_ids
+        
+        if max_new_tokens is None:
+            if max_length is not None:
+                max_new_tokens = max_length - idx.shape[1]
+            else:
+                max_new_tokens = 50  
+        
+        with torch.no_grad():
+            for _ in range(max_new_tokens):
+                idx_cond = idx[:, -self.config.block_size:]
+                
+                outputs = self(idx_cond)
+                logits = outputs.logits[:, -1, :] / temperature
+                
+                if top_k is not None:
+                    v, _ = torch.topk(logits, min(top_k, logits.size(-1)))
+                    logits[logits < v[:, [-1]]] = float('-inf')
+                
+                if top_p is not None:
+                    sorted_logits, sorted_indices = torch.sort(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)
+                    logits[indices_to_remove] = float('-inf')
+                
+                probs = F.softmax(logits, dim=-1)
+                
+                if do_sample:
+                    idx_next = torch.multinomial(probs, num_samples=1)
+                else:
+                    idx_next = torch.argmax(logits, dim=-1, keepdim=True)
+                
+                idx = torch.cat((idx, idx_next), dim=1)
+                
+    
+                if eos_token_id is not None and (idx_next == eos_token_id).any():
+                    break
+        
+        return idx
+
+
+class BVVBestMoeForCausalLM(PreTrainedModel):
+    config_class = BVVBestConfig
+    
+    def __init__(self, config):
+        super().__init__(config)
+        self.model_a = BVVBestForCausalLM(config)
+        self.model_b = BVVBestForCausalLM(config)
+        self.mode = 'mean'
+        if self.mode == 'stack':
+            self.adapter = nn.Linear(2 * vocab_size, vocab_size)
+
+    def forward(self, idx, targets=None):
+        out_a = self.model_a(idx, targets)
+        out_b = self.model_b(idx, targets)
+        if self.mode == 'mean':
+            logits = (out_a.logits + out_b.logits) / 2
+        elif self.mode == 'stack':
+            logits = torch.cat([out_a.logits, out_b.logits], dim=-1)
+            logits = self.adapter(logits)
+        else:
+            raise NotImplementedError
+        loss = None
+        if targets is not None:
+            logits_flat = logits.reshape(-1, logits.size(-1))
+            targets_flat = targets.reshape(-1)
+            loss = F.cross_entropy(logits_flat, targets_flat, ignore_index = pad_id)
+        return CausalLMOutput(logits=logits, loss=loss)
+
+    def generate(self, 
+                input_ids=None,
+                max_new_tokens=None,
+                max_length=None,
+                temperature=1.0,
+                top_k=None,
+                top_p=None,
+                do_sample=True,
+                pad_token_id=None,
+                eos_token_id=None,
+                **kwargs):
+        
+        if input_ids is None:
+            raise ValueError("Input_ids must be provided")
+        
+        idx = input_ids
+        
+        if max_new_tokens is None:
+            if max_length is not None:
+                max_new_tokens = max_length - idx.shape[1]
+            else:
+                max_new_tokens = 50  
+        
+        with torch.no_grad():
+            for _ in range(max_new_tokens):
+                idx_cond = idx[:, -self.config.block_size:]
+                
+                outputs = self(idx_cond)
+                logits = outputs.logits[:, -1, :] / temperature
+                
+                if top_k is not None:
+                    v, _ = torch.topk(logits, min(top_k, logits.size(-1)))
+                    logits[logits < v[:, [-1]]] = float('-inf')
+                
+                if top_p is not None:
+                    sorted_logits, sorted_indices = torch.sort(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)
+                    logits[indices_to_remove] = float('-inf')
+                
+                probs = F.softmax(logits, dim=-1)
+                
+                if do_sample:
+                    idx_next = torch.multinomial(probs, num_samples=1)
+                else:
+                    idx_next = torch.argmax(logits, dim=-1, keepdim=True)
+                
+                idx = torch.cat((idx, idx_next), dim=1)
+                
+    
+                if eos_token_id is not None and (idx_next == eos_token_id).any():
+                    break
+        
+        return idx

tokenizer_config.json

@@ -1,6 +1,9 @@
 {
+  "tokenizer_class": "PreTrainedTokenizerFast",
+  "model_type": "gpt2",
+  "bos_token": "<s>",
+  "eos_token": "</s>",
   "unk_token": "<unk>",
   "pad_token": "<pad>",
-  "bos_token": "<s>",
-  "eos_token": "</s>"
+  "vocab_size": 131072
 }