diff options
Diffstat (limited to 'bitsandbytes/nn/modules.py')
| -rw-r--r-- | bitsandbytes/nn/modules.py | 21 |
1 files changed, 15 insertions, 6 deletions
diff --git a/bitsandbytes/nn/modules.py b/bitsandbytes/nn/modules.py index b222f54..9250fec 100644 --- a/bitsandbytes/nn/modules.py +++ b/bitsandbytes/nn/modules.py @@ -221,6 +221,7 @@ class Linear8bitLt(nn.Linear): output_features, bias=True, has_fp16_weights=True, + memory_efficient_backward=False, threshold=0.0, index=None, ): @@ -232,10 +233,13 @@ class Linear8bitLt(nn.Linear): self.state.threshold = threshold self.state.has_fp16_weights = has_fp16_weights + self.state.memory_efficient_backward = memory_efficient_backward if threshold > 0.0 and not has_fp16_weights: self.state.use_pool = True - self.weight = Int8Params(self.weight.data, has_fp16_weights=has_fp16_weights) + self.weight = Int8Params( + self.weight.data, has_fp16_weights=has_fp16_weights, requires_grad=has_fp16_weights + ) def init_8bit_state(self): self.state.CB = self.weight.CB @@ -255,11 +259,16 @@ class Linear8bitLt(nn.Linear): out = bnb.matmul(x, self.weight, bias=self.bias, state=self.state) - if not self.state.has_fp16_weights and self.state.CB is not None: - # we converted 8-bit row major to turing/ampere format in the first inference pass - # we no longer need the row-major weight - del self.state.CB - self.weight.data = self.state.CxB + if not self.state.has_fp16_weights: + if not self.state.memory_efficient_backward and self.state.CB is not None: + # we converted 8-bit row major to turing/ampere format in the first inference pass + # we no longer need the row-major weight + del self.state.CB + self.weight.data = self.state.CxB + elif self.state.memory_efficient_backward and self.state.CxB is not None: + # For memory efficient backward, we convert 8-bit row major to turing/ampere format at each inference pass. + # Thus, we delete CxB from the state. + del self.state.CxB return out |