Open xddun opened 1 day ago
I follow this page, it works:
https://www.dong-blog.fun/post/1863
By the way, I have an additional question. I noticed that the interface for accessing this Triton deployment is quite stiff. Is there a question-and-answer interface similar to OpenAI's available?
When I access it this way, the model's responses seem to be completing my sentences rather than the usual question-and-answer format.
# curl -X POST http://101.136.8.66:8000/v2/models/ensemble/generate -d '{"text_input": "Who are you?", "max_tokens": 200, "bad_words": "", "stop_words": ""}'
{"model_name":"ensemble","model_version":"1","sequence_end":false,"sequence_id":0,"sequence_index":0,"sequence_start":false,"text_output":"Who are you? Where do you come from? Where are you going? These are the questions that philosophers ponder. For businesses, these three questions are equally important. Where a business comes from determines its genes; where a business is going determines its strategy; and who a business is determines its culture. Corporate culture is the soul of a business and its intrinsic driving force for development. Corporate culture is the sum of the business's values, spirit, system, and code of conduct, and it forms a unique, stable, and distinctive corporate culture system over the course of long-term development.\nCorporate culture is the soul of a business and its intrinsic driving force for development. Corporate culture is the sum of the business's values, spirit, system, and code of conduct, and it forms a unique, stable, and distinctive corporate culture system over the course of long-term development.\nCorporate culture is the intrinsic driving force for a business's development. Corporate culture is the sum of the business's values, spirit, system, and code of conduct, and it forms a unique, stable, and distinctive corporate culture system over the course of long-term development. Corporate culture"}
System Info
env:
ubuntu22 RTX3090 Linux euler-MS-7D30 6.8.0-45-generic #45~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Wed Sep 11 15:25:05 UTC 2 x86_64 x86_64 x86_64 GNU/Linux
I wanted to build an image, but unexpectedly encountered an error. My process was as follows in 4steps:
git clone https://github.com/NVIDIA/TensorRT-LLM.git
cd TensorRT-LLM
git lfs pull
make -C docker release_build
error log:
Who can help?
Is it possible to provide a pre-configured image with the environment already set up? Compiling the image is really challenging!
Information
Tasks
examples
folder (such as GLUE/SQuAD, ...)Reproduction
env:
ubuntu22 RTX3090 Linux euler-MS-7D30 6.8.0-45-generic #45~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Wed Sep 11 15:25:05 UTC 2 x86_64 x86_64 x86_64 GNU/Linux
I wanted to build an image, but unexpectedly encountered an error. My process was as follows in 4steps:
git clone https://github.com/NVIDIA/TensorRT-LLM.git
cd TensorRT-LLM
git lfs pull
make -C docker release_build
error log:
Expected behavior
#
actual behavior
#
additional notes
#