-
### 🐛 Describe the bug
torchbench_amp_bf16_inference
- [ ] `moco`
Traceback (most recent call last):
File "/home/sdp/actions-runner/_work/torch-xpu-ops/pytorch/benchmarks/dynamo/common.py", …
-
## 🐛 Bug
When benchmarking model: 'Mixtral-8x7B-v0.1' we get OOM errors even with --checkpoint_activations True
The same configurations works for torch.compile.
Might be related to [https://gi…
-
Hi,
I have migrated to `androidx.benchmark:benchmark-macro-junit4:1.2.2` and when running the assembleDebugAndroidTest task for firebase testlab I am getting the following error (Same for 1.2.0-1.…
-
Why this matters:
- regression comparison when new SST versions are released
- evaluation of impact of SST changes on certain aspects of models in use
- identifying scalability bottlenecks (e.g., …
-
I just ran a simple benchmark (`redis-benchmark -q -n 1000000 -s /var/run/redis/redis.sock -t set,get,mset -P 16 -c 100`) with 3 networking configurations
1. redis on host system over tcp
2. redis …
-
Small Models are provided in https://github.com/ServerlessLLM/ServerlessLLM/blob/main/benchmarks/README.md, such as the test data of opt-6.7b. Compared with SafeTensors, sllm has a speed increase of a…
-
I've been running some large-scale benchmarking with minhash deduplication on SLURM clusters, loosely following [this example](https://github.com/huggingface/datatrove/blob/main/examples/minhash_dedup…
-
# Synopsis
OpenSearch Benchmark (OSB) is a performance testing tool for OpenSearch, a community-driven, open source search and analytics suite. It allows users to benchmark various aspects of OpenSea…
-
### What needs to happen?
Benchmark different machine configurations for RunInference benchmark test on DataflowRunner.
### Issue Priority
Priority: 2
### Issue Component
Component: sdk-py-core
-
Thanos with Memcached enabled plus MiniO as Long-term
**Thanos, Prometheus and Golang version used**:
**Object Storage Provider**: S3 MiniO
**What happened**:
I have configured my Th…