Spaces:
Runtime error
Runtime error
Update app.py
Browse files
app.py
CHANGED
|
@@ -21,13 +21,14 @@ class SmolLM2Config(PretrainedConfig):
|
|
| 21 |
num_key_value_heads=3,
|
| 22 |
hidden_act="silu",
|
| 23 |
max_position_embeddings=2048,
|
| 24 |
-
initializer_range=0.
|
| 25 |
rms_norm_eps=1e-5,
|
| 26 |
use_cache=True,
|
| 27 |
pad_token_id=None,
|
| 28 |
bos_token_id=0,
|
| 29 |
eos_token_id=0,
|
| 30 |
tie_word_embeddings=True,
|
|
|
|
| 31 |
**kwargs
|
| 32 |
):
|
| 33 |
self.vocab_size = vocab_size
|
|
@@ -41,6 +42,7 @@ class SmolLM2Config(PretrainedConfig):
|
|
| 41 |
self.initializer_range = initializer_range
|
| 42 |
self.rms_norm_eps = rms_norm_eps
|
| 43 |
self.use_cache = use_cache
|
|
|
|
| 44 |
super().__init__(
|
| 45 |
pad_token_id=pad_token_id,
|
| 46 |
bos_token_id=bos_token_id,
|
|
@@ -64,54 +66,125 @@ class RMSNorm(nn.Module):
|
|
| 64 |
x = x * torch.rsqrt(variance + self.eps)
|
| 65 |
return self.weight * x
|
| 66 |
|
| 67 |
-
|
| 68 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 69 |
super().__init__()
|
| 70 |
self.hidden_size = config.hidden_size
|
| 71 |
self.num_heads = config.num_attention_heads
|
|
|
|
| 72 |
self.head_dim = config.hidden_size // config.num_attention_heads
|
| 73 |
|
| 74 |
-
|
| 75 |
-
self.
|
| 76 |
-
self.
|
| 77 |
-
self.
|
|
|
|
| 78 |
|
| 79 |
-
|
| 80 |
-
|
| 81 |
-
|
| 82 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 83 |
)
|
|
|
|
|
|
|
|
|
|
| 84 |
|
| 85 |
-
|
| 86 |
-
|
|
|
|
|
|
|
| 87 |
|
| 88 |
-
|
| 89 |
-
|
| 90 |
-
|
| 91 |
-
hidden_states = self.input_layernorm(hidden_states)
|
| 92 |
|
| 93 |
-
#
|
| 94 |
-
|
| 95 |
-
|
| 96 |
-
|
| 97 |
-
v = self.v_proj(hidden_states).view(batch_size, seq_length, self.num_heads, self.head_dim).transpose(1, 2)
|
| 98 |
|
| 99 |
-
#
|
| 100 |
-
|
| 101 |
-
|
|
|
|
|
|
|
|
|
|
| 102 |
|
| 103 |
if attention_mask is not None:
|
| 104 |
-
|
| 105 |
|
| 106 |
-
|
| 107 |
-
|
|
|
|
|
|
|
|
|
|
| 108 |
|
| 109 |
-
|
| 110 |
-
|
| 111 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 112 |
hidden_states = residual + hidden_states
|
| 113 |
|
| 114 |
-
# MLP
|
| 115 |
residual = hidden_states
|
| 116 |
hidden_states = self.post_attention_layernorm(hidden_states)
|
| 117 |
hidden_states = self.mlp(hidden_states)
|
|
@@ -125,18 +198,48 @@ class SmolLM2ForCausalLM(PreTrainedModel):
|
|
| 125 |
def __init__(self, config):
|
| 126 |
super().__init__(config)
|
| 127 |
self.config = config
|
|
|
|
| 128 |
self.embed_tokens = nn.Embedding(config.vocab_size, config.hidden_size)
|
| 129 |
self.layers = nn.ModuleList([LlamaDecoderLayer(config) for _ in range(config.num_hidden_layers)])
|
| 130 |
self.norm = RMSNorm(config.hidden_size, config.rms_norm_eps)
|
|
|
|
|
|
|
| 131 |
self.lm_head = nn.Linear(config.hidden_size, config.vocab_size, bias=False)
|
| 132 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 133 |
if config.tie_word_embeddings:
|
| 134 |
self.lm_head.weight = self.embed_tokens.weight
|
| 135 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 136 |
def forward(self, input_ids=None, attention_mask=None, labels=None, **kwargs):
|
| 137 |
hidden_states = self.embed_tokens(input_ids)
|
| 138 |
|
| 139 |
-
#
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 140 |
for layer in self.layers:
|
| 141 |
hidden_states = layer(hidden_states, attention_mask)
|
| 142 |
|
|
@@ -155,15 +258,76 @@ class SmolLM2ForCausalLM(PreTrainedModel):
|
|
| 155 |
"attention_mask": kwargs.get("attention_mask", None)
|
| 156 |
}
|
| 157 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 158 |
# Register the model
|
| 159 |
AutoModelForCausalLM.register(SmolLM2Config, SmolLM2ForCausalLM)
|
| 160 |
|
| 161 |
# Cache for model and tokenizer
|
| 162 |
MODEL = None
|
| 163 |
TOKENIZER = None
|
|
|
|
| 164 |
|
| 165 |
def initialize():
|
| 166 |
-
global MODEL, TOKENIZER
|
| 167 |
|
| 168 |
if MODEL is None:
|
| 169 |
print("Loading model and tokenizer...")
|
|
@@ -175,17 +339,24 @@ def initialize():
|
|
| 175 |
config_path = hf_hub_download(repo_id=model_id, filename="config.json")
|
| 176 |
with open(config_path, 'r') as f:
|
| 177 |
config_dict = json.load(f)
|
| 178 |
-
|
| 179 |
|
| 180 |
# Load tokenizer
|
| 181 |
print("Loading tokenizer...")
|
| 182 |
-
TOKENIZER = AutoTokenizer.from_pretrained(
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 183 |
|
| 184 |
-
#
|
| 185 |
special_tokens = {
|
| 186 |
-
'
|
| 187 |
-
'eos_token': '
|
| 188 |
-
'
|
|
|
|
| 189 |
}
|
| 190 |
TOKENIZER.add_special_tokens(special_tokens)
|
| 191 |
|
|
@@ -194,7 +365,10 @@ def initialize():
|
|
| 194 |
weights_path = hf_hub_download(repo_id=model_id, filename="pytorch_model.bin")
|
| 195 |
|
| 196 |
# Initialize model
|
| 197 |
-
MODEL = SmolLM2ForCausalLM(
|
|
|
|
|
|
|
|
|
|
| 198 |
|
| 199 |
# Load state dict
|
| 200 |
state_dict = torch.load(weights_path, map_location="cpu")
|
|
@@ -228,14 +402,23 @@ def generate_text(prompt, max_length=100, temperature=0.7, top_k=50):
|
|
| 228 |
prompt = TOKENIZER.bos_token + prompt
|
| 229 |
|
| 230 |
# Encode prompt
|
| 231 |
-
|
| 232 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 233 |
|
| 234 |
# Generate
|
| 235 |
with torch.no_grad():
|
| 236 |
outputs = MODEL.generate(
|
| 237 |
input_ids,
|
| 238 |
-
|
|
|
|
| 239 |
temperature=max(0.1, min(temperature, 1.0)), # Clamp temperature
|
| 240 |
top_k=max(1, min(top_k, 100)), # Clamp top_k
|
| 241 |
do_sample=True if temperature > 0 else False,
|
|
|
|
| 21 |
num_key_value_heads=3,
|
| 22 |
hidden_act="silu",
|
| 23 |
max_position_embeddings=2048,
|
| 24 |
+
initializer_range=0.041666666666666664,
|
| 25 |
rms_norm_eps=1e-5,
|
| 26 |
use_cache=True,
|
| 27 |
pad_token_id=None,
|
| 28 |
bos_token_id=0,
|
| 29 |
eos_token_id=0,
|
| 30 |
tie_word_embeddings=True,
|
| 31 |
+
rope_theta=10000.0,
|
| 32 |
**kwargs
|
| 33 |
):
|
| 34 |
self.vocab_size = vocab_size
|
|
|
|
| 42 |
self.initializer_range = initializer_range
|
| 43 |
self.rms_norm_eps = rms_norm_eps
|
| 44 |
self.use_cache = use_cache
|
| 45 |
+
self.rope_theta = rope_theta
|
| 46 |
super().__init__(
|
| 47 |
pad_token_id=pad_token_id,
|
| 48 |
bos_token_id=bos_token_id,
|
|
|
|
| 66 |
x = x * torch.rsqrt(variance + self.eps)
|
| 67 |
return self.weight * x
|
| 68 |
|
| 69 |
+
def precompute_rope_frequencies(dim: int, max_position_embeddings: int, theta: float = 10000.0):
|
| 70 |
+
position = torch.arange(max_position_embeddings).unsqueeze(1) # [seq_len, 1]
|
| 71 |
+
div_term = theta ** (torch.arange(0, dim, 2).float() / dim) # [dim/2]
|
| 72 |
+
freqs = position / div_term # [seq_len, dim/2]
|
| 73 |
+
return freqs
|
| 74 |
+
|
| 75 |
+
def apply_rotary_embeddings(x: torch.Tensor, freqs: torch.Tensor):
|
| 76 |
+
# x shape: [batch, seq_len, heads, head_dim]
|
| 77 |
+
# freqs shape: [seq_len, head_dim/2]
|
| 78 |
+
x_rot = x.float()
|
| 79 |
+
|
| 80 |
+
# Reshape freqs to match x's dimensions
|
| 81 |
+
freqs = freqs.unsqueeze(0).unsqueeze(2) # [1, seq_len, 1, dim/2]
|
| 82 |
+
|
| 83 |
+
# Split channels for rotation
|
| 84 |
+
x1, x2 = x_rot[..., :x_rot.shape[-1]//2], x_rot[..., x_rot.shape[-1]//2:]
|
| 85 |
+
|
| 86 |
+
# Apply rotary embeddings
|
| 87 |
+
cos = torch.cos(freqs).to(x.device)
|
| 88 |
+
sin = torch.sin(freqs).to(x.device)
|
| 89 |
+
|
| 90 |
+
# Ensure broadcasting dimensions match
|
| 91 |
+
cos = cos.expand_as(x1)
|
| 92 |
+
sin = sin.expand_as(x1)
|
| 93 |
+
|
| 94 |
+
# Rotate x1 and x2
|
| 95 |
+
x1_rot = x1 * cos - x2 * sin
|
| 96 |
+
x2_rot = x2 * cos + x1 * sin
|
| 97 |
+
|
| 98 |
+
# Concatenate back
|
| 99 |
+
return torch.cat([x1_rot, x2_rot], dim=-1).to(x.dtype)
|
| 100 |
+
|
| 101 |
+
class LlamaAttention(nn.Module):
|
| 102 |
+
def __init__(self, config: SmolLM2Config):
|
| 103 |
super().__init__()
|
| 104 |
self.hidden_size = config.hidden_size
|
| 105 |
self.num_heads = config.num_attention_heads
|
| 106 |
+
self.num_kv_heads = config.num_key_value_heads
|
| 107 |
self.head_dim = config.hidden_size // config.num_attention_heads
|
| 108 |
|
| 109 |
+
# Adjust projections to match head dimensions
|
| 110 |
+
self.q_proj = nn.Linear(config.hidden_size, self.num_heads * self.head_dim, bias=False)
|
| 111 |
+
self.k_proj = nn.Linear(config.hidden_size, self.num_kv_heads * self.head_dim, bias=False)
|
| 112 |
+
self.v_proj = nn.Linear(config.hidden_size, self.num_kv_heads * self.head_dim, bias=False)
|
| 113 |
+
self.o_proj = nn.Linear(self.num_heads * self.head_dim, config.hidden_size, bias=False)
|
| 114 |
|
| 115 |
+
# Initialize rotary embeddings
|
| 116 |
+
self.register_buffer(
|
| 117 |
+
"rope_freqs",
|
| 118 |
+
precompute_rope_frequencies(
|
| 119 |
+
self.head_dim, # Use full head_dim for frequencies
|
| 120 |
+
config.max_position_embeddings,
|
| 121 |
+
config.rope_theta
|
| 122 |
+
),
|
| 123 |
+
persistent=False
|
| 124 |
)
|
| 125 |
+
|
| 126 |
+
def forward(self, hidden_states, attention_mask=None):
|
| 127 |
+
batch_size, seq_length, _ = hidden_states.size()
|
| 128 |
|
| 129 |
+
# Project and reshape
|
| 130 |
+
q = self.q_proj(hidden_states).view(batch_size, seq_length, self.num_heads, self.head_dim)
|
| 131 |
+
k = self.k_proj(hidden_states).view(batch_size, seq_length, self.num_kv_heads, self.head_dim)
|
| 132 |
+
v = self.v_proj(hidden_states).view(batch_size, seq_length, self.num_kv_heads, self.head_dim)
|
| 133 |
|
| 134 |
+
# Apply rotary embeddings
|
| 135 |
+
q = apply_rotary_embeddings(q, self.rope_freqs[:seq_length])
|
| 136 |
+
k = apply_rotary_embeddings(k, self.rope_freqs[:seq_length])
|
|
|
|
| 137 |
|
| 138 |
+
# Repeat k/v heads if num_kv_heads < num_heads
|
| 139 |
+
if self.num_kv_heads < self.num_heads:
|
| 140 |
+
k = k.repeat_interleave(self.num_heads // self.num_kv_heads, dim=2)
|
| 141 |
+
v = v.repeat_interleave(self.num_heads // self.num_kv_heads, dim=2)
|
|
|
|
| 142 |
|
| 143 |
+
# Scaled dot-product attention
|
| 144 |
+
q = q.transpose(1, 2) # (batch, num_heads, seq_len, head_dim)
|
| 145 |
+
k = k.transpose(1, 2)
|
| 146 |
+
v = v.transpose(1, 2)
|
| 147 |
+
|
| 148 |
+
attention_scores = torch.matmul(q, k.transpose(-2, -1)) / math.sqrt(self.head_dim)
|
| 149 |
|
| 150 |
if attention_mask is not None:
|
| 151 |
+
attention_scores = attention_scores + attention_mask
|
| 152 |
|
| 153 |
+
attention_probs = F.softmax(attention_scores, dim=-1)
|
| 154 |
+
context = torch.matmul(attention_probs, v)
|
| 155 |
+
|
| 156 |
+
context = context.transpose(1, 2).contiguous()
|
| 157 |
+
context = context.view(batch_size, seq_length, -1)
|
| 158 |
|
| 159 |
+
return self.o_proj(context)
|
| 160 |
+
|
| 161 |
+
class LlamaMLP(nn.Module):
|
| 162 |
+
def __init__(self, config: SmolLM2Config):
|
| 163 |
+
super().__init__()
|
| 164 |
+
self.gate_proj = nn.Linear(config.hidden_size, config.intermediate_size, bias=False)
|
| 165 |
+
self.up_proj = nn.Linear(config.hidden_size, config.intermediate_size, bias=False)
|
| 166 |
+
self.down_proj = nn.Linear(config.intermediate_size, config.hidden_size, bias=False)
|
| 167 |
+
self.act_fn = nn.SiLU()
|
| 168 |
+
|
| 169 |
+
def forward(self, x):
|
| 170 |
+
gate = self.act_fn(self.gate_proj(x))
|
| 171 |
+
up = self.up_proj(x)
|
| 172 |
+
return self.down_proj(gate * up)
|
| 173 |
+
|
| 174 |
+
class LlamaDecoderLayer(nn.Module):
|
| 175 |
+
def __init__(self, config: SmolLM2Config):
|
| 176 |
+
super().__init__()
|
| 177 |
+
self.self_attn = LlamaAttention(config)
|
| 178 |
+
self.mlp = LlamaMLP(config)
|
| 179 |
+
self.input_layernorm = RMSNorm(config.hidden_size, config.rms_norm_eps)
|
| 180 |
+
self.post_attention_layernorm = RMSNorm(config.hidden_size, config.rms_norm_eps)
|
| 181 |
+
|
| 182 |
+
def forward(self, hidden_states, attention_mask=None):
|
| 183 |
+
residual = hidden_states
|
| 184 |
+
hidden_states = self.input_layernorm(hidden_states)
|
| 185 |
+
hidden_states = self.self_attn(hidden_states, attention_mask)
|
| 186 |
hidden_states = residual + hidden_states
|
| 187 |
|
|
|
|
| 188 |
residual = hidden_states
|
| 189 |
hidden_states = self.post_attention_layernorm(hidden_states)
|
| 190 |
hidden_states = self.mlp(hidden_states)
|
|
|
|
| 198 |
def __init__(self, config):
|
| 199 |
super().__init__(config)
|
| 200 |
self.config = config
|
| 201 |
+
|
| 202 |
self.embed_tokens = nn.Embedding(config.vocab_size, config.hidden_size)
|
| 203 |
self.layers = nn.ModuleList([LlamaDecoderLayer(config) for _ in range(config.num_hidden_layers)])
|
| 204 |
self.norm = RMSNorm(config.hidden_size, config.rms_norm_eps)
|
| 205 |
+
|
| 206 |
+
# Add lm_head before weight tying
|
| 207 |
self.lm_head = nn.Linear(config.hidden_size, config.vocab_size, bias=False)
|
| 208 |
|
| 209 |
+
# Initialize weights
|
| 210 |
+
self.apply(self._init_weights)
|
| 211 |
+
|
| 212 |
+
# Tie weights if configured
|
| 213 |
if config.tie_word_embeddings:
|
| 214 |
self.lm_head.weight = self.embed_tokens.weight
|
| 215 |
|
| 216 |
+
def _init_weights(self, module):
|
| 217 |
+
if isinstance(module, nn.Linear):
|
| 218 |
+
torch.nn.init.normal_(module.weight, mean=0.0, std=self.config.initializer_range)
|
| 219 |
+
if module.bias is not None:
|
| 220 |
+
torch.nn.init.zeros_(module.bias)
|
| 221 |
+
elif isinstance(module, nn.Embedding):
|
| 222 |
+
torch.nn.init.normal_(module.weight, mean=0.0, std=self.config.initializer_range)
|
| 223 |
+
|
| 224 |
def forward(self, input_ids=None, attention_mask=None, labels=None, **kwargs):
|
| 225 |
hidden_states = self.embed_tokens(input_ids)
|
| 226 |
|
| 227 |
+
# Create causal attention mask if none provided
|
| 228 |
+
if attention_mask is None:
|
| 229 |
+
# Create causal mask
|
| 230 |
+
seq_length = input_ids.size(1)
|
| 231 |
+
# [batch_size, 1, seq_length, seq_length]
|
| 232 |
+
causal_mask = torch.triu(
|
| 233 |
+
torch.ones((seq_length, seq_length), dtype=torch.bool, device=input_ids.device),
|
| 234 |
+
diagonal=1
|
| 235 |
+
).unsqueeze(0).unsqueeze(0)
|
| 236 |
+
attention_mask = torch.zeros(
|
| 237 |
+
(1, 1, seq_length, seq_length),
|
| 238 |
+
dtype=hidden_states.dtype,
|
| 239 |
+
device=hidden_states.device
|
| 240 |
+
)
|
| 241 |
+
attention_mask.masked_fill_(causal_mask, float("-inf"))
|
| 242 |
+
|
| 243 |
for layer in self.layers:
|
| 244 |
hidden_states = layer(hidden_states, attention_mask)
|
| 245 |
|
|
|
|
| 258 |
"attention_mask": kwargs.get("attention_mask", None)
|
| 259 |
}
|
| 260 |
|
| 261 |
+
def generate(
|
| 262 |
+
self,
|
| 263 |
+
input_ids,
|
| 264 |
+
max_length=100,
|
| 265 |
+
temperature=0.7,
|
| 266 |
+
top_k=50,
|
| 267 |
+
do_sample=True,
|
| 268 |
+
num_return_sequences=1,
|
| 269 |
+
pad_token_id=None,
|
| 270 |
+
eos_token_id=None,
|
| 271 |
+
**kwargs
|
| 272 |
+
):
|
| 273 |
+
cur_len = input_ids.shape[1]
|
| 274 |
+
batch_size = input_ids.shape[0]
|
| 275 |
+
|
| 276 |
+
if max_length < cur_len:
|
| 277 |
+
max_length = cur_len
|
| 278 |
+
|
| 279 |
+
unfinished_sequences = torch.ones(batch_size, dtype=torch.long, device=input_ids.device)
|
| 280 |
+
|
| 281 |
+
while cur_len < max_length:
|
| 282 |
+
# Prepare model inputs
|
| 283 |
+
model_inputs = self.prepare_inputs_for_generation(input_ids)
|
| 284 |
+
|
| 285 |
+
# Forward pass
|
| 286 |
+
with torch.no_grad():
|
| 287 |
+
outputs = self(**model_inputs)
|
| 288 |
+
next_token_logits = outputs[:, -1, :]
|
| 289 |
+
|
| 290 |
+
# Temperature scaling
|
| 291 |
+
if temperature != 1.0 and temperature > 0:
|
| 292 |
+
next_token_logits = next_token_logits / temperature
|
| 293 |
+
|
| 294 |
+
# Top-k filtering
|
| 295 |
+
if top_k > 0:
|
| 296 |
+
indices_to_remove = next_token_logits < torch.topk(next_token_logits, top_k)[0][..., -1, None]
|
| 297 |
+
next_token_logits[indices_to_remove] = float('-inf')
|
| 298 |
+
|
| 299 |
+
# Sample or greedy
|
| 300 |
+
if do_sample:
|
| 301 |
+
probs = F.softmax(next_token_logits, dim=-1)
|
| 302 |
+
next_tokens = torch.multinomial(probs, num_samples=1)
|
| 303 |
+
else:
|
| 304 |
+
next_tokens = torch.argmax(next_token_logits, dim=-1)
|
| 305 |
+
next_tokens = next_tokens.unsqueeze(-1)
|
| 306 |
+
|
| 307 |
+
# Append next tokens
|
| 308 |
+
input_ids = torch.cat([input_ids, next_tokens], dim=-1)
|
| 309 |
+
cur_len = input_ids.shape[1]
|
| 310 |
+
|
| 311 |
+
# Early stopping if all sequences have reached the EOS token
|
| 312 |
+
if eos_token_id is not None:
|
| 313 |
+
unfinished_sequences = unfinished_sequences.mul(
|
| 314 |
+
next_tokens.squeeze(-1).ne(eos_token_id).long()
|
| 315 |
+
)
|
| 316 |
+
if unfinished_sequences.max() == 0:
|
| 317 |
+
break
|
| 318 |
+
|
| 319 |
+
return input_ids
|
| 320 |
+
|
| 321 |
# Register the model
|
| 322 |
AutoModelForCausalLM.register(SmolLM2Config, SmolLM2ForCausalLM)
|
| 323 |
|
| 324 |
# Cache for model and tokenizer
|
| 325 |
MODEL = None
|
| 326 |
TOKENIZER = None
|
| 327 |
+
CONFIG = None
|
| 328 |
|
| 329 |
def initialize():
|
| 330 |
+
global MODEL, TOKENIZER, CONFIG
|
| 331 |
|
| 332 |
if MODEL is None:
|
| 333 |
print("Loading model and tokenizer...")
|
|
|
|
| 339 |
config_path = hf_hub_download(repo_id=model_id, filename="config.json")
|
| 340 |
with open(config_path, 'r') as f:
|
| 341 |
config_dict = json.load(f)
|
| 342 |
+
CONFIG = SmolLM2Config(**config_dict)
|
| 343 |
|
| 344 |
# Load tokenizer
|
| 345 |
print("Loading tokenizer...")
|
| 346 |
+
TOKENIZER = AutoTokenizer.from_pretrained(
|
| 347 |
+
model_id,
|
| 348 |
+
model_max_length=CONFIG.max_position_embeddings,
|
| 349 |
+
padding_side="left",
|
| 350 |
+
truncation_side="left",
|
| 351 |
+
trust_remote_code=True
|
| 352 |
+
)
|
| 353 |
|
| 354 |
+
# Make sure we're using the correct special tokens
|
| 355 |
special_tokens = {
|
| 356 |
+
'bos_token': '<|endoftext|>',
|
| 357 |
+
'eos_token': '<|endoftext|>',
|
| 358 |
+
'unk_token': '<|endoftext|>',
|
| 359 |
+
'pad_token': '<|endoftext|>' # Using endoftext as pad token since it's not specified
|
| 360 |
}
|
| 361 |
TOKENIZER.add_special_tokens(special_tokens)
|
| 362 |
|
|
|
|
| 365 |
weights_path = hf_hub_download(repo_id=model_id, filename="pytorch_model.bin")
|
| 366 |
|
| 367 |
# Initialize model
|
| 368 |
+
MODEL = SmolLM2ForCausalLM(CONFIG)
|
| 369 |
+
|
| 370 |
+
# Resize token embeddings to match tokenizer
|
| 371 |
+
MODEL.resize_token_embeddings(len(TOKENIZER))
|
| 372 |
|
| 373 |
# Load state dict
|
| 374 |
state_dict = torch.load(weights_path, map_location="cpu")
|
|
|
|
| 402 |
prompt = TOKENIZER.bos_token + prompt
|
| 403 |
|
| 404 |
# Encode prompt
|
| 405 |
+
encoded = TOKENIZER.encode_plus(
|
| 406 |
+
prompt,
|
| 407 |
+
add_special_tokens=True,
|
| 408 |
+
return_tensors="pt",
|
| 409 |
+
padding=True,
|
| 410 |
+
truncation=True,
|
| 411 |
+
max_length=CONFIG.max_position_embeddings
|
| 412 |
+
)
|
| 413 |
+
input_ids = encoded["input_ids"].to(MODEL.device)
|
| 414 |
+
attention_mask = encoded["attention_mask"].to(MODEL.device)
|
| 415 |
|
| 416 |
# Generate
|
| 417 |
with torch.no_grad():
|
| 418 |
outputs = MODEL.generate(
|
| 419 |
input_ids,
|
| 420 |
+
attention_mask=attention_mask,
|
| 421 |
+
max_length=min(max_length + len(input_ids[0]), CONFIG.max_position_embeddings),
|
| 422 |
temperature=max(0.1, min(temperature, 1.0)), # Clamp temperature
|
| 423 |
top_k=max(1, min(top_k, 100)), # Clamp top_k
|
| 424 |
do_sample=True if temperature > 0 else False,
|