dbg4 / src /server /backend.py
justheuristic's picture
dbg
23d0807
"""Code for serving bloom blocks via hivemind-server"""
from typing import Sequence, Tuple
import torch
from hivemind.moe.server.module_backend import ModuleBackend
from hivemind.moe.server.task_pool import TaskPool
from src.bloom.from_pretrained import BloomBlock
from src.server.cache import MemoryCache
MAX_LENGTH = 2048
class TransformerBackend(ModuleBackend):
"""A wrapper for BloomBlock that can process requests for bloom layer forward, forward_incremental, and backward"""
def __init__(self, *args, memory_cache: MemoryCache, **kwargs):
super().__init__(*args, **kwargs)
assert isinstance(self.module, BloomBlock)
self.memory_cache = memory_cache
for name, param in self.module.named_parameters():
assert not param.requires_grad, f"Bloom layer parameters must not accumulate gradients, but {name} does"
for name, buf in self.module.named_buffers():
assert not buf.requires_grad, f"Bloom layer parameters must not accumulate gradients, but {name} does"
self.inference_pool = TaskPool(self.inference_step, max_batch_size=1, name=f"{self.name}_inference")
def inference_step(self, cache_metadata: torch.IntTensor, *inputs: torch.Tensor) -> Tuple[torch.Tensor, ...]:
with torch.inference_mode():
attention_cache_handle = int(cache_metadata[0, 0].item())
prefix_length = int(cache_metadata[0, 1].item())
hidden_states = inputs[0] # todo: in future, it would be best to support attention mask here
assert (
hidden_states.ndim == 3
), "expected hidden states to be 3-dimensional: [batch_size, seq_len, hid_size]"
with self.memory_cache.use_cache(attention_cache_handle) as cache:
assert isinstance(self.module, BloomBlock) and cache.shape[0] == 2 and cache.ndim == 5
layer_past = past_k, past_v = cache[0, :, :prefix_length], cache[1, :, :prefix_length]
print("METADATA:", cache_metadata, past_k.shape, past_v.shape)
hidden_states, (new_k, new_v) = self.module.forward(
hidden_states, layer_past=layer_past, use_cache=True
)
# todo remove these asserts once we pass all tests
new_length = new_v.shape[1]
assert new_length > prefix_length
assert new_k.shape[0] == past_k.shape[0] and new_v.shape[0] == past_v.shape[0]
assert new_k.shape[1] == new_length and new_v.shape[1] == new_length
assert new_k.shape[2:] == past_k.shape[2:] and new_v.shape[2:] == past_v.shape[2:]
assert torch.allclose(new_v[:, : past_v.shape[1]], past_v)
assert torch.allclose(new_k[:, : past_k.shape[1]], past_k)
cache[0, :, prefix_length:new_length, :] = new_k[:, prefix_length:new_length]
cache[1, :, prefix_length:new_length, :] = new_v[:, prefix_length:new_length]
return (hidden_states,)
def get_pools(self) -> Sequence[TaskPool]:
return self.forward_pool, self.backward_pool, self.inference_pool