-
This section: https://qupath.readthedocs.io/en/0.5/docs/deep/gpu.html#creating-a-launcher-for-qupath-pytorch
This line: There are some extra PyTorch options with DJL. There are more details in the …
-
DJL versions >= 0.29.0 doesn't support PyTorch for MacOS x86_64.
Additional context: https://github.com/spring-projects/spring-ai/commit/df17f08f5b7c22b31ac73893b90d621180856774#commitcomment-1468…
-
i use the same code with examples/research_projects/stack_llama_2/scripts/dpo_llama2.py
however, i always got problem when mapping data in DPOTrainer?
```
Map: 17%|██████████████████████████████…
-
问题:使用demo测试发现涉及到视频流检测的,会报错如下:
Exception in thread "main" java.lang.ExceptionInInitializerError
at ai.djl.paddlepaddle.engine.PpEngine.newBaseManager(PpEngine.java:110)
at ai.djl.paddlepaddle.en…
-
### Description
After delivering my session in DevNexus, two devs approached me asking on how to compile a Quarkus + DJL library into native, as DJL library already supports GraalVM compilation https…
-
```
Caused by: java.io.IOException: Resource not found in classpath: native/lib/linux-aarch64/libtokenizers.so
at ai.djl.util.ClassLoaderUtils.getResourceAsStream(ClassLoaderUtils.java:221)
…
-
*Concise Description:*
**[vLLM v0.6.0](https://blog.vllm.ai/2024/09/05/perf-update.html) provides 2.7x throughput improvement and 5x latency reduction over the previous version (v0.5.3)**
*DLC i…
-
### Describe the bug
[2024-08-08T07:17:44,731][WARN ][a.d.h.t.HuggingFaceTokenizer] [opensearch-ml-node] maxLength is not explicitly specified, use modelMaxLength: 512
[2024-08-08T07:17:44,737][ER…
-
when I run this code [https://github.com/deepjavalibrary/djl-demo/blob/master/huggingface/nlp/src/main/java/com/examples/Reranking.java](url), it failed
![WechatIMG18](https://github.com/user-at…
-
hi, We are trying to use the DJL (Deep Java Library) framework and have conducted benchmark tests on different servers, finding that the throughput performance varies on different machines. One can re…