Open-Assistant
Open-Assistant copied to clipboard
LAION-AI
OpenAssistant is a chat-based assistant that understands tasks, can interact with third-party systems, and retrieve information dynamically to do so.
Bumps [typescript](https://github.com/Microsoft/TypeScript) from 4.9.5 to 5.2.2. Release notes Sourced from typescript's releases. TypeScript 5.2 For release notes, check out the release announcement. For the complete list of fixed issues, check...
![dependabot[bot] avatar](https://avatars.githubusercontent.com/in/29110?v=4 "dependabot[bot] avatar"){kind=avatar}
Is there any plan to release the dataset in cycles ? I think in comparison to the V1 dataset it should be grown pretty much
Hello, as someone who has contributed to the data of Open Assistant, I would really appreciate it if I (and others) could make use of it locally on mobile. To...
**Description:** While using Open Assistant for chat-based interactions, I encountered an issue related to changing plugins during the loading of answers. This behavior seems to result in errors and extended...
``` from transformers import AutoTokenizer AutoTokenizer.from_pretrained("OpenAssistant/llama2-13b-orca-8k-3319").padding_side >> 'left' AutoTokenizer.from_pretrained("TheBloke/Llama-2-13B-fp16") >> 'left' AutoTokenizer.from_pretrained("mosaicml/mpt-7b").padding_side >> 'right' AutoTokenizer.from_pretrained("huggyllama/llama-7b").padding_side >> 'left' AutoTokenizer.from_pretrained("OpenAssistant/llama-30b-sft-v8.2-2.4k-steps-system").padding_side >> 'left' ``` Since llama models are using left padding, the supervised...
If the model's num_embeddings is 10000,but we change the tokenizer to 10007. After SFT training the model's num_embeddings will be 10016, that because in model/model_training/utils/utils.py get_model(conf, tokenizer, pad_vocab_size_to_multiple_of=16, check_freeze_layer=True) has...
I am trying to run pretrain of LLaMA 30b. And here is my running cmd: ``` deepspeed trainer_sft.py --configs defaults llama-30b-pretrain pretrain --cache_dir $DATA_PATH --output_dir $MODEL_PATH/llama-30b-pre --deepspeed ``` And after...
Hi, I've seen vicuna model [here](https://lmsys.org/blog/2023-03-30-vicuna/). It seems like a pretty good model and supports lots of other languages (like Persian) other of the box. It speaks farsi but not...
