-
**Please acknowledge the following before creating a ticket**
- [x] I have read the GitHub issues section of [REPORTING-BUGS](../blob/master/REPORTING-BUGS).
**Description of the bug:**
Passing…
-
SOTA (CUBLAS, CUTLASS) FP8 GEMM kernels are performing poorly for small M (bs*seq_len) < 32 regime.
This work will focus on leveraging the performant pieces of the [Marlin](https://github.com/IST-D…
-
We used FIO to test the performance of the file system, and the summary of the test is as follows:
| | Aster RamFs | Aster Ext2 | Linux Ext2 |
| :-----| :---- | :---- | :---- |
| seq-wri…
-
exec readme bash Pairwise Knowledge Fusion
FuseLLM/FuseChat/train/trainer.py", line 121, in compute_loss
if self.args.distill_loss_type == "ce":
loss_lm = cross_entropy…
-
### 🐛 Describe the bug
We had many dynamic recompilation related to AMP (float vs half}. It happens when we use `TORCHDYNAMO_INLINE_INBUILT_NN_MODULES=1` /cc @ezyang @msaroufim @bdhirsh @anijain230…
bhack updated
3 weeks ago
-
When I try to convert a stream of byte arrays to a seq of byte arrays, it just returns back the original stream without doing any conversion.
```clojure
user=> (require '[manifold.stream :as ms])
u…
-
```scala
import scala.collection.immutable
class Test {
def foo(as: immutable.Seq[Int]) = {
val List(_, bs: _*) = as
val cs: collection.Seq[Int] = bs
}
}
```
```scala
> dotc …
-
**Describe the bug**
It seems that `binary_to_term` got slower.
**To Reproduce**
I used the following simple way to generate a rather large term and then N copies of that term as a binary in a li…
-
作者,你好。
我想使用你的模型在**bert**上,我看到你的代码中model文件中151行这里:
```
def forward(self, x, hidden):
bptt_len, bs = x.shape
vocab_sz = self.embedding.num_embeddings
```
这里的 x 的输入只有 seq_len, bs,…
-
Hi
i tested BitMapperBS recently ,it was really an ultrafast and memory-efficient aligner.
i runned the software successfully.but i have same problems with the output. the output didn't contain Met…