hfppl
hfppl copied to clipboard
probcomp
Probabilistic programming with HuggingFace language models
Simply adding the following lines to the top of the method should do the trick: ```python for particle in particles: particle.start() ```
 I modify .lm.s ---> lm it can run.But have a new issue.  [/usr/local/lib/python3.10/dist-packages/hfppl/distributions/lmcontext.py](https://localhost:8080/#) in sample(self) 24 async def sample(self): 25 probs = np.exp(self.ctx.next_token_logprobs) ---> 26 token_id = np.random.choice(len(probs),...
https://colab.research.google.com/drive/1uJEC-U8dcwsTWccCDGVexpgXexzZ642n?usp=sharing#scrollTo=gsi7R6D0NcGn&uniqifier=1  10 self.lm = LMContext(LLM, prompt) 11 self.q = LMContext(LLM, prompt) ---> 12 self.prompt_len = len(str(self.lm.s)) 13 self.max_tokens = max_tokens 14 AttributeError: 'LMContext' object has no attribute 's'
Hi, Great work. Would it be possible to get more examples? I'd like to see how to use Infilling and Prompt Intersection are used. I think it would really help...
thank you for releasing this amazing repo - SMC steering looks very promising I think this will probably be the first time a lot of DL Engineers come across the...
I thought it'd be fun to add functionality for automatically generating a token graph from the `TokenTrie` class, sort of like the one in the original LLaMPPL paper! A screenshot...
Could you provide an example of how to use `Model.twist()`? Currently, I'm calling it in `step()` with values between 0 and 1. I'd imagine that the effect size should be...
I have a question about the role of N (the number of particles), and K (the factor) in the task of prompt intersection. I am trying to replicate Fig. 4...
This PR adds stratified resampling to `smc_standard`. The resampling method can be specified by a new `resampling_method` argument to `smc_standard` which defaults to `multinomial` (and thus preserves the past behaviour)....
updates: - [github.com/astral-sh/ruff-pre-commit: v0.9.9 → v0.11.4](https://github.com/astral-sh/ruff-pre-commit/compare/v0.9.9...v0.11.4)
![pre-commit-ci[bot] avatar](https://avatars.githubusercontent.com/in/68672?v=4 "pre-commit-ci[bot] avatar"){kind=avatar}