alihan/knowledge-harvest-from-lms
1
0
Fork 0
mirror of https://github.com/tanyuqian/knowledge-harvest-from-lms.git synced 2023-06-02 01:35:42 +03:00
Code Issues Packages Projects Releases Wiki Activity

updated.

Browse Source
This commit is contained in:
bwt09
2022-06-27 22:31:08 -07:00
parent 828d36ba4b
commit 9596cf1646
1 changed files with 3 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
README.md
View File

@@ -37,6 +37,7 @@ The prompt searching reads the relation definitions from `"init_prompts"` and `"
(Files in the [relation_info/](relation_info/) folder have contained the results of this step.)
#### Efficient Search for Knowledge Tuples
This step corresponds to [paper]() Section 3.2.
```
python main.py --rel_set conceptnet --model_name roberta-large --n_ent_tuples 1000 --n_prompts 20
```
@@ -57,7 +58,7 @@ The results will be saved into `summary.txt` in `{result_dir}`, e.g., [results/c
#### Example -- "A can B but not good at" (RoBERTaNet-large)
#### Example -- "A can B but not good at" (RoBERTaNet)
```
+------------------------------------+--------------------------------+-------------------------------------------+
| Seed samples | Ours (Top 20) | Ours (Random samples over top 200 tuples) |
@@ -85,7 +86,7 @@ The results will be saved into `summary.txt` in `{result_dir}`, e.g., [results/c
+------------------------------------+--------------------------------+-------------------------------------------+
```
#### Example -- "A needs B to C"
#### Example -- "A needs B to C" (BertNet)
```
+------------------------------------------------+--------------------------------------------------------+--------------------------------------------------+
| Seed samples | Ours (Top 20) | Ours (Random samples over top 200 tuples) |