smallthinker right

This commit is contained in:
qiyuxinlin 2025-07-25 12:46:14 +00:00
parent f8719ee7b9
commit 712ad1fa3c
7 changed files with 48 additions and 108 deletions

View file

@ -471,20 +471,17 @@ class KSmallthinkerRotaryEmbedding(BaseInjectedModule, SmallthinkerRotaryEmbeddi
@torch.no_grad()
def forward(self, x, position_ids):
if "dynamic" in self.rope_type:
self._dynamic_frequency_update(position_ids, device=x.device)
# Core RoPE block
inv_freq_expanded = self.inv_freq[None, :, None].float().expand(position_ids.shape[0], -1, 1)
# print(inv_freq_expanded.device)
inv_freq_expanded = self.inv_freq[None, :, None].float().expand(position_ids.shape[0], -1, 1).to(x.device)
position_ids_expanded = position_ids[:, None, :].float()
# Force float32 (see https://github.com/huggingface/transformers/pull/29285)
device_type = x.device.type
device_type = device_type if isinstance(device_type, str) and device_type != "mps" else "cpu"
with torch.autocast(device_type=device_type, enabled=False):
freqs = (inv_freq_expanded.to(x.device) @ position_ids_expanded).transpose(1, 2)
freqs_cis = torch.polar(torch.ones_like(freqs), freqs)
freqs_cis = freqs_cis * self.attention_scaling
return freqs_cis
device_type = x.device.type if isinstance(x.device.type, str) and x.device.type != "mps" else "cpu"
with torch.autocast(device_type=device_type, enabled=False): # Force float32
freqs = (inv_freq_expanded.float() @ position_ids_expanded.float()).transpose(1, 2)
emb = torch.cat((freqs, freqs), dim=-1)
cos = emb.cos() * self.attention_scaling
sin = emb.sin() * self.attention_scaling
return cos.to(dtype=x.dtype), sin.to(dtype=x.dtype)
class KGlm4MoeRotaryEmbedding(BaseInjectedModule, Glm4MoeRotaryEmbedding):
def __init__(