Rasmus

Example of controlled dynamical system?

2

I'm trying to fit a Neural ODE to some observations of a dynamical system, which has some control inputs. That is, the control inputs are not part of the state...

Enable HF SpeedMonitor

4

Enable SpeedMonitor on HF models by using PyTorch FlopCounterMode to calculate model FLOPs.

RoPE implementation differences

5

I've been working with the pretrained Llama 3 weights, and found out that the RoPE implementation here does not match the one found in other places. The difference is whether...

RuntimeError on ROCm

7

Example of command: ```python benchmark_throughput.py --model gpt2 --input-len 256 --output-len 256``` Output: ```Namespace(backend='vllm', dataset=None, input_len=256, output_len=256, model='gpt2', tokenizer='gpt2', quantization=None, tensor_parallel_size=1, n=1, use_beam_search=False, num_prompts=1000, seed=0, hf_max_batch_size=None, trust_remote_code=False, max_model_len=None, dtype='auto', enforce_eager=False) INFO...

[Doc]: ROCm installation instructions do not work

4

### 📚 The doc issue Following the instructions at https://docs.vllm.ai/en/latest/getting_started/amd-installation.html#build-from-source-rocm, using the exact Docker image mentioned (pytorch_rocm6.1.2_ubuntu20.04_py3.9_pytorch_staging.sif, although with a custom Python venv and Pytorch install), and run into the...