Chen Qian
Chen Qian
@isaacbmiller Mind taking a look?
@adaminsky Thanks for the PR! This paper is very new, and doesn't have reference yet, so it's may not be a good time to ship to DSPy directly. However, if...
Thanks Abi! I am a little unsure if a blocklist of token ids could work as expected, mainly due to subword tokenizer. For common English curse words, they are usually...
@atharvapurdue Thanks for your interest! Feel free to create a colab to get familiar with the sampler APIs and accumulate some stats. We are still making a few changes to...
@mattdangerw Since LLaMa exists in KerasNLP now, I assume the routine input/output checks have been done, so I am more confused about this implementation. here is how another repo do...
@Bhuvanesh09 This is something pretty interesting while complex, if you are interested in this path, here are some guidelines: 1. Pick 1-2 datasets and 3-5 models, and report the benchmark...
@Bhuvanesh09 Thanks for putting up the result and reproducible code! This is a complex one, we will find some time to go through it carefully and loop back to you.
@Cachet23 Thanks for reporting the issue! Deepseek is not yet fully supported in DSPy, we need to change the adapter to parse the Deepseek's response because it has more fields...
@FireMasterK Thanks for reporting the issue! We are working on native function calling for ReAct, please stay tuned.
@drewstone thanks for reporting the issue, could you share the full reproducible code?