Skip to content

[WIP] Update pretrained embs example #1089

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
wants to merge 16 commits into from
Closed

[WIP] Update pretrained embs example #1089

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
16 commits
Select commit Hold shift + click to select a range
97404a0
Created PretrainedEmbeddings and changed other blocks to support pre-…
gabrielspmoreira May 4, 2023
37a06fd
Using SequenceAggregator with pre-trained embeddings
gabrielspmoreira May 9, 2023
6cadf44
Added test for aggregating sequences of pre-trained embeddings
gabrielspmoreira May 9, 2023
260ab05
Fixed bugs in graph mode when using EmbeddingOperator, as last dim wa…
gabrielspmoreira May 10, 2023
eb5d648
Reducing cardinality of testing and sequence_testing fixture from 519…
gabrielspmoreira May 10, 2023
295d88d
Changing test of Transformers with pre-trained embeddings to use sequ…
gabrielspmoreira May 10, 2023
0dc187d
Fixed tests
gabrielspmoreira May 10, 2023
13c7b12
Fixed tests and added test of pre-trained embeddings with masked lang…
gabrielspmoreira May 10, 2023
1e7e283
Added tests with pretrained embeddings for DLRM and DCN
gabrielspmoreira May 10, 2023
5e09725
Fixed PopularityBasedSamplerV2 that was raising error when the sample…
gabrielspmoreira May 10, 2023
c535e83
Linting fix
gabrielspmoreira May 15, 2023
993e4c8
Fixed failing test
gabrielspmoreira May 15, 2023
e8e6bf5
Implementing Sara's suggestions on pretrained embeddings
gabrielspmoreira May 15, 2023
e433341
update pretrained embs example
radekosmulski May 15, 2023
4628b00
update
radekosmulski May 16, 2023
12085f0
update
radekosmulski May 16, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1,245 changes: 1,153 additions & 92 deletions examples/usecases/entertainment-with-pretrained-embeddings.ipynb
View file
Open in desktop

Large diffs are not rendered by default.

278 changes: 278 additions & 0 deletions merlin/datasets/entertainment/movielens/100k/schema.json
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,278 @@
{
"feature": [
{
"name": "movieId",
"type": "INT",
"intDomain": {
"name": "movieId",
"min": "1",
"max": "1680",
"isCategorical": true
},
"annotation": {
"tag": [
"item",
"categorical",
"item_id"
],
"extraMetadata": [
{
"_dims": [
[
0.0,
null
]
],
"dtype_item_size": 32.0,
"freq_threshold": 0.0,
"is_ragged": false,
"max_size": 0.0,
"is_list": false,
"embedding_sizes": {
"dimension": 102.0,
"cardinality": 1680.0
},
"cat_path": ".//categories/unique.movieId.parquet",
"num_buckets": null,
"start_index": 0.0
}
]
}
},
{
"name": "userId",
"type": "INT",
"intDomain": {
"name": "userId",
"max": "943",
"isCategorical": true
},
"annotation": {
"tag": [
"categorical",
"user",
"user_id"
],
"extraMetadata": [
{
"embedding_sizes": {
"cardinality": 943.0,
"dimension": 74.0
},
"num_buckets": null,
"is_ragged": false,
"max_size": 0.0,
"dtype_item_size": 32.0,
"is_list": false,
"start_index": 0.0,
"freq_threshold": 0.0,
"cat_path": ".//categories/unique.userId.parquet"
}
]
}
},
{
"name": "genres",
"type": "INT",
"intDomain": {
"name": "genres",
"max": "216",
"isCategorical": true
},
"annotation": {
"tag": [
"item",
"categorical"
],
"extraMetadata": [
{
"max_size": 0.0,
"start_index": 0.0,
"dtype_item_size": 32.0,
"num_buckets": null,
"freq_threshold": 0.0,
"embedding_sizes": {
"cardinality": 216.0,
"dimension": 32.0
},
"cat_path": ".//categories/unique.genres.parquet",
"is_list": false,
"is_ragged": false
}
]
}
},
{
"name": "TE_movieId_rating",
"type": "FLOAT",
"annotation": {
"tag": [
"continuous"
],
"extraMetadata": [
{
"is_list": false,
"dtype_item_size": 64.0,
"is_ragged": false
}
]
}
},
{
"name": "userId_count",
"type": "FLOAT",
"annotation": {
"tag": [
"continuous"
],
"extraMetadata": [
{
"is_list": false,
"dtype_item_size": 32.0,
"is_ragged": false
}
]
}
},
{
"name": "gender",
"type": "INT",
"intDomain": {
"name": "gender",
"max": "2",
"isCategorical": true
},
"annotation": {
"tag": [
"categorical",
"user"
],
"extraMetadata": [
{
"max_size": 0.0,
"dtype_item_size": 32.0,
"is_list": false,
"is_ragged": false,
"freq_threshold": 0.0,
"embedding_sizes": {
"dimension": 16.0,
"cardinality": 2.0
},
"num_buckets": null,
"start_index": 0.0,
"cat_path": ".//categories/unique.gender.parquet"
}
]
}
},
{
"name": "zip_code",
"type": "INT",
"intDomain": {
"name": "zip_code",
"max": "795",
"isCategorical": true
},
"annotation": {
"tag": [
"categorical",
"user"
],
"extraMetadata": [
{
"embedding_sizes": {
"cardinality": 795.0,
"dimension": 67.0
},
"freq_threshold": 0.0,
"max_size": 0.0,
"cat_path": ".//categories/unique.zip_code.parquet",
"is_ragged": false,
"is_list": false,
"start_index": 0.0,
"dtype_item_size": 32.0,
"num_buckets": null
}
]
}
},
{
"name": "rating",
"type": "INT",
"annotation": {
"tag": [
"regression",
"target"
],
"extraMetadata": [
{
"is_list": false,
"dtype_item_size": 64.0,
"is_ragged": false
}
]
}
},
{
"name": "rating_binary",
"type": "INT",
"annotation": {
"tag": [
"target",
"binary_classification"
],
"extraMetadata": [
{
"dtype_item_size": 32.0,
"is_list": false,
"is_ragged": false
}
]
}
},
{
"name": "age",
"type": "INT",
"intDomain": {
"name": "age",
"max": "8",
"isCategorical": true
},
"annotation": {
"tag": [
"categorical",
"user"
],
"extraMetadata": [
{
"dtype_item_size": 32.0,
"embedding_sizes": {
"dimension": 16.0,
"cardinality": 8.0
},
"max_size": 0.0,
"start_index": 0.0,
"num_buckets": null,
"cat_path": ".//categories/unique.age.parquet",
"is_list": false,
"freq_threshold": 0.0,
"is_ragged": false
}
]
}
},
{
"name": "title",
"type": "BYTES",
"annotation": {
"extraMetadata": [
{
"is_list": false,
"is_ragged": false,
"dtype_item_size": 64.0
}
]
}
}
]
}
Loading