Ecosyste.ms: Issues
An open API service for providing issue and pull request metadata for open source projects.
GitHub / hao-ai-lab/LookaheadDecoding issues and pull requests
#64 - Yi Models support
Issue -
State: open - Opened by fayuge 3 months ago
- 1 comment
#63 - How to optimize the scene with relatively short output?
Issue -
State: open - Opened by yangbohust 4 months ago
#62 - Bug in Greedy Search
Issue -
State: open - Opened by david-wei-01001 4 months ago
#61 - Cannot use save_log and log_history
Issue -
State: open - Opened by david-wei-01001 4 months ago
- 1 comment
#60 - about verification branch
Issue -
State: open - Opened by dhcode-cpp 4 months ago
- 1 comment
#59 - TypeError: LlamaSdpaAttention.forward() got an unexpected keyword argument 'lookahead'
Issue -
State: open - Opened by zev123456 5 months ago
- 5 comments
#58 - Tensor parallel
Issue -
State: closed - Opened by wangyuwen1999 5 months ago
- 1 comment
#57 - How to use LADE in single-node multi-process way?
Issue -
State: open - Opened by sjrrr13 5 months ago
- 2 comments
#56 - [BUG Report] jacobi_greedy_search_multilevel function bug
Issue -
State: open - Opened by yangbohust 6 months ago
#55 - Do lookahead and repetition_penalty conflict?
Issue -
State: open - Opened by zhanweiw 6 months ago
#54 - BUG Report
Issue -
State: open - Opened by yangbohust 6 months ago
#53 - Why jacobi_sample_multilevel() fill window with argmax instead of also using sampling?
Issue -
State: open - Opened by yangbohust 6 months ago
#52 - question on verification beginning pos
Issue -
State: open - Opened by felixdae 6 months ago
- 1 comment
#51 - Adaptive LEVEL/GUESS size?
Issue -
State: open - Opened by sahel-sh 7 months ago
#50 - flash attention and sampling
Pull Request -
State: closed - Opened by Viol2000 7 months ago
#49 - add paper
Pull Request -
State: closed - Opened by Viol2000 7 months ago
#48 - specific model
Issue -
State: open - Opened by qspang 8 months ago
- 13 comments
#47 - Multiple GPUS
Issue -
State: open - Opened by qspang 8 months ago
- 9 comments
#46 - Compatibility with Flash Attention 2
Issue -
State: closed - Opened by jasonli0707 8 months ago
- 2 comments
#45 - Related work: Prompt lookup decoding
Issue -
State: open - Opened by shermansiu 8 months ago
- 7 comments
#44 - Questions on combined attention mask structure for Jacobi iteration
Issue -
State: open - Opened by learning-chip 8 months ago
- 23 comments
#43 - support hf v4.36
Pull Request -
State: closed - Opened by Viol2000 9 months ago
#42 - Undefined symbols in lade/decoding.py
Issue -
State: closed - Opened by hjmus 9 months ago
- 1 comment
#41 - Run time error with `lade.config_lade(LEVEL=2)`
Issue -
State: closed - Opened by learning-chip 9 months ago
- 3 comments
#40 - TypeError: 'NoneType' object is not subscriptable at line 427 of decoding.py
Issue -
State: closed - Opened by henryxiao1997 9 months ago
- 5 comments
#39 - Any plan to support other LLMs, and integrated into huggingface?
Issue -
State: closed - Opened by henryxiao1997 9 months ago
- 2 comments
#38 - fix llama kv cache
Pull Request -
State: closed - Opened by jiqing-feng 9 months ago
- 2 comments
#37 - Qs on Understanding Lookahead and Jacobi
Issue -
State: closed - Opened by RonanKMcGovern 9 months ago
- 7 comments
#36 - Can I get a MT-bench evaluation code for reproduction of acceleration?
Issue -
State: open - Opened by je1lee 9 months ago
- 11 comments
#35 - Incompatible with LlamaSdpaAttention in transformers v4.36
Issue -
State: closed - Opened by learning-chip 9 months ago
- 1 comment
#34 - Why the tup needs to be added to the tail of token_map when it found in the token_map?
Issue -
State: closed - Opened by kevinoldching 9 months ago
- 2 comments
#33 - Why using Jacobi decoding? What are the advantages besides the fact that Jacobi decoding can reduce some of the decode steps?
Issue -
State: closed - Opened by ZipECHO 9 months ago
- 19 comments
#32 - Questions on the attention mask, and whether to accept the last element of guess_results when all guess_tokens are accepted
Issue -
State: open - Opened by YingHH1 10 months ago
- 6 comments
#31 - The meaning of the CHAT option in decoding
Issue -
State: closed - Opened by YingHH1 10 months ago
- 1 comment
#30 - The inference results are inconsistent with Huggingface.
Issue -
State: open - Opened by cyfwry 10 months ago
- 6 comments
#29 - Does Lade support topp/topk/temperature sampleing?
Issue -
State: open - Opened by AlvL1225 10 months ago
- 2 comments
#28 - question about attention patterns
Issue -
State: closed - Opened by SUDA-HLT-ywfang 10 months ago
- 6 comments
#27 - Can't run minimal.py on A100
Issue -
State: closed - Opened by jiqing-feng 10 months ago
- 3 comments
#26 - Benchmarks comparing with Medusa
Issue -
State: open - Opened by Rock-Anderson 10 months ago
#25 - Update decoding.py
Pull Request -
State: closed - Opened by eltociear 10 months ago
#24 - Is it similar to ProphetNet, ProphetNet-Ads and BANG?
Issue -
State: closed - Opened by qiweizhen 10 months ago
- 2 comments
#23 - Is it the same as ProphetNet and ProphetNet-Ads?
Issue -
State: closed - Opened by qiweizhen 10 months ago
#22 - amazing work!! any plan support for chatglm or Qwen model?
Issue -
State: open - Opened by white-wolf-tech 10 months ago
- 1 comment
#21 - Can lade accelerate T5?
Issue -
State: closed - Opened by yjdy 10 months ago
- 1 comment
#20 - The Jacobi method and its corresponding code
Issue -
State: closed - Opened by YingHH1 10 months ago
- 2 comments
#19 - No speed up
Issue -
State: open - Opened by Louis-y-nlp 10 months ago
- 5 comments
#18 - Support for bnb nf4 quant?
Issue -
State: open - Opened by col-in-coding 10 months ago
- 1 comment
#17 - NameError: name 'F' is not defined
Issue -
State: closed - Opened by CSWellesSun 10 months ago
- 1 comment
#16 - Any analysis on the impact on accuracy
Issue -
State: closed - Opened by qizzzh 10 months ago
- 2 comments
#15 - Support for GPTQ?
Issue -
State: open - Opened by liHai001 10 months ago
- 4 comments
#14 - How to generate the n-grams - which to keep, which to discard?
Issue -
State: open - Opened by bobqianic 10 months ago
- 27 comments
#13 - Lookahead Decoding Development Roadmap
Issue -
State: open - Opened by Viol2000 10 months ago
- 6 comments
#12 - Does the module support fine-tuned Llama2?
Issue -
State: closed - Opened by spring1915 10 months ago
- 4 comments
#11 - Support Baichuan models
Issue -
State: closed - Opened by ghost 10 months ago
- 1 comment
#10 - How about batching throughput and energy consumption
Issue -
State: open - Opened by Light-of-Hers 10 months ago
- 2 comments
#9 - Is there a plan to rebuild the code in a clear style?
Issue -
State: closed - Opened by hhhh12345678 10 months ago
- 1 comment
#8 - Question on Initial guess tokens
Issue -
State: closed - Opened by DaehanKim 10 months ago
- 1 comment
#7 - AttributeError: module 'lade' has no attribute 'config_pading'
Issue -
State: closed - Opened by liHai001 10 months ago
- 1 comment
#6 - Gumbel noise for sampling in-place of greedy decoding
Pull Request -
State: closed - Opened by knagrecha 10 months ago
- 36 comments
#5 - Fix broken reference
Pull Request -
State: closed - Opened by bilal-aamer 10 months ago
#4 - Broken links in ReadMe
Issue -
State: closed - Opened by bilal-aamer 10 months ago
#3 - import necessary packages
Pull Request -
State: closed - Opened by WrRan 10 months ago
#2 - import necessary packages
Pull Request -
State: closed - Opened by WrRan 10 months ago
#1 - Integration with other open-source libraries
Issue -
State: open - Opened by shermansiu 10 months ago
- 19 comments