Datasets:

Modalities:
Image
Text
Formats:
parquet
Languages:
English
ArXiv:
Libraries:
Datasets
Dask
License:
elliesleightholm commited on
Commit
0700cd0
β€’
1 Parent(s): d551155

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +5 -5
README.md CHANGED
@@ -127,10 +127,10 @@ marqo-gs-dataset/
127
  β”œβ”€β”€ marqo_gs_wfash_1m/
128
  ```
129
  For each dataset such as marqo_gs_full_10m, there are 4 splits as discussed before.
130
- - query_0_product_id_0 represents in-domain set,
131
- - query_1_product_id_0 represents novel query set,
132
- - query_0_product_id_1 represents novel document set,
133
- - query_1_product_id_1 represents zero shot set,
134
  For each split, there is a ground truth csv containing triplet information,
135
  a set of validation ground truth and a set of test ground truth.
136
 
@@ -250,7 +250,7 @@ with torch.no_grad(), torch.cuda.amp.autocast():
250
  print("Label probs:", text_probs)
251
  ```
252
  ### Quick Demo with Hugging Face for E5 models.
253
- Here is a quick example to load our finetuned e5 text models from hugging face directly.
254
  ```python
255
  import torch.nn.functional as F
256
 
 
127
  β”œβ”€β”€ marqo_gs_wfash_1m/
128
  ```
129
  For each dataset such as marqo_gs_full_10m, there are 4 splits as discussed before.
130
+ - `query_0_product_id_0` represents in-domain set,
131
+ - `query_1_product_id_0` represents novel query set,
132
+ - `query_0_product_id_1` represents novel document set,
133
+ - `query_1_product_id_1` represents zero shot set,
134
  For each split, there is a ground truth csv containing triplet information,
135
  a set of validation ground truth and a set of test ground truth.
136
 
 
250
  print("Label probs:", text_probs)
251
  ```
252
  ### Quick Demo with Hugging Face for E5 models.
253
+ Here is a quick example to load our finetuned e5 text models from Hugging Face directly.
254
  ```python
255
  import torch.nn.functional as F
256