saymrwulf/transformers
1
0
Fork 0
mirror of https://github.com/saymrwulf/transformers.git synced 2026-05-14 20:58:08 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Fix image preview in multi-GPU inference docs (#35303)

Browse source 
fix: link for img
This commit is contained in:
Jacky Lee 2024-12-17 09:33:50 -08:00 committed by GitHub
parent 4302b27719
commit 927c3e39ec
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
2 changed files with 2 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
docs/source/en/perf_infer_gpu_multi.md
View file

@ -64,5 +64,5 @@ You can benefit from considerable speedups for inference, especially for inputs
For a single forward pass on [Llama](https://huggingface.co/docs/transformers/model_doc/llama#transformers.LlamaModel) with a sequence length of 512 and various batch sizes, the expected speedup is as follows:
<div style="text-align: center">
<img src="huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/Meta-Llama-3-8B-Instruct, seqlen = 512, python, w_ compile.png">
<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/Meta-Llama-3-8B-Instruct%2C%20seqlen%20%3D%20512%2C%20python%2C%20w_%20compile.png">
</div>

2
docs/source/zh/perf_infer_gpu_multi.md
View file

@ -64,5 +64,5 @@ torchrun --nproc-per-node 4 demo.py
以下是 [Llama](https://huggingface.co/docs/transformers/model_doc/llama#transformers.LlamaModel) 模型在序列长度为 512 且不同批量大小情况下的单次前向推理的预期加速效果：
<div style="text-align: center">
<img src="huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/Meta-Llama-3-8B-Instruct, seqlen = 512, python, w_ compile.png">
<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/Meta-Llama-3-8B-Instruct%2C%20seqlen%20%3D%20512%2C%20python%2C%20w_%20compile.png">
</div>