mcmoose1900B to LocalLLaMA@poweruser.forumEnglish · 1 year agoNouse-Capybara-34B 200Khuggingface.coexternal-linkmessage-square16fedilinkarrow-up11arrow-down10
arrow-up11arrow-down1external-linkNouse-Capybara-34B 200Khuggingface.comcmoose1900B to LocalLLaMA@poweruser.forumEnglish · 1 year agomessage-square16fedilink
minus-squareWolframRavenwolfBlinkfedilinkEnglisharrow-up1·1 year agoAlso note this important issue that affects this and all other Yi-based models: BOS token as 1 seriously hurts these GGUF Yi models
minus-squarea_beautiful_rhindBlinkfedilinkEnglisharrow-up1·1 year agoSo we can just skip BOS token on all these models?
minus-squareambient_temp_xenoBlinkfedilinkEnglisharrow-up1·1 year agoI did the gguf-py/scripts/gguf-set-metadata.py some-yi-model.gguf tokenizer.ggml.bos_token_id 144 and it’s changed the outputs a lot from yesterday.
Also note this important issue that affects this and all other Yi-based models:
BOS token as 1 seriously hurts these GGUF Yi models
So we can just skip BOS token on all these models?
I did the gguf-py/scripts/gguf-set-metadata.py some-yi-model.gguf tokenizer.ggml.bos_token_id 144
and it’s changed the outputs a lot from yesterday.