carolinaxxxxx
carolinaxxxxx
When the device_index = 1 parameter (GPU 1), the GPU 0 is charged with low value data (in my case about 250 MB) and GPU itself shows signs of work,...
@minhthuc2502 - Does CTranslate2 support openchat models e.g. openchat/openchat-3.5-0106-gemma? I managed to perform the conversion to ct2, but I can't "make" it work properly? THX
@minhthuc2502 I use below code for test: ``` import ctranslate2 import transformers generator = ctranslate2.Generator("/test/openchat35gemma", device="cuda", device_index=1) tokenizer = transformers.AutoTokenizer.from_pretrained("/test/openchat35gemma") prompt = f"GPT4 Correct User: HelloGPT4 Correct Assistant: HiGPT4 Correct...
Hi, thanks for reply. I changed VPS to Aeza. Everything works out of the box on Debian. Good job! środa, 4 września 2024 01:07, Tenebris ***@***.***> napisał(a): > Did you...
@MahmoudAshraf97 good job sir. However I have a problem - it does not respect the --hotwords option at all in standard interference and batching mode. I tested on different materials....
@MahmoudAshraf97 good job sir. However I have a problem - it does not respect the --hotwords option at all in standard interference and batching mode. I tested on different materials....
@MahmoudAshraf97 any chance to make hotwords work as it did before? Thanks. Good job.
Can anyone confirm this issue? Thx.