vec2text
vec2text copied to clipboard
utilities for decoding deep representations (like sentence embeddings) back to text
Hi Jack, Thanks for the great work and sharing the code! I am trying to reproduce results from the paper and want to confirm if I am doing it correctly....
Very cool project and I've been having a fun time exploring the repo. I'd like to run some additional examples but I'm having a difficulty reproducing any results using your...
Hello! Thanks for your great paper and sharing the codes. I have a question about the languages it supports. Does it apply to other languages, such as Chinese? Thanks!
You say: "Currently we only support models for inverting OpenAI text-embedding-ada-002 embeddings but are hoping to add more soon. (We can provide the GTR inverters used in the paper upon...
> I ran the following training script python run.py --per_device_train_batch_size 8 --per_device_eval_batch_size 8 --max_seq_length 128 --model_name_or_path t5-small --dataset_name msmarco --embedder_model_name **sentence-transformers/all-MiniLM-L6-v2** --num_repeat_tokens 16 --embedder_no_grad True --num_train_epochs 1 --max_eval_samples 500 --eval_steps...
I have been trying to use your example codes and I can get the `vec2text` to load on the new openai library(`>0.28`) but not the old one. However the `get_embeddings_openai`...
Hello, I would like to look at the hypothesis embeddings, for example, to see how the cosine similarity changes per iteration. It looks to me like `invert_embeddings()` only returns the...
Hi Jack, Thanks for the great work. Do u guys plan to release the inversion models that were trained to invert LLAMA-2 7b embeddings? It would be very helpful for...
Hi Thank you for presenting your research. I have a question regarding the embedding_transform in inversion.py. As per my understanding, this function corresponds to the MLP model described in your...
I didn't find the example that clear but I have a guess at what's happening. Might be worth spelling out something to the effect of trying to map text to...